{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 6228,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00016056518946692356,
      "grad_norm": 19.07044219970703,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 2.2765,
      "step": 1
    },
    {
      "epoch": 0.0003211303789338471,
      "grad_norm": 17.860671997070312,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.2325,
      "step": 2
    },
    {
      "epoch": 0.0004816955684007707,
      "grad_norm": 16.178730010986328,
      "learning_rate": 3e-06,
      "loss": 2.1528,
      "step": 3
    },
    {
      "epoch": 0.0006422607578676942,
      "grad_norm": 19.609806060791016,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.2842,
      "step": 4
    },
    {
      "epoch": 0.0008028259473346178,
      "grad_norm": 18.418621063232422,
      "learning_rate": 5e-06,
      "loss": 2.2578,
      "step": 5
    },
    {
      "epoch": 0.0009633911368015414,
      "grad_norm": 17.91242027282715,
      "learning_rate": 6e-06,
      "loss": 2.2732,
      "step": 6
    },
    {
      "epoch": 0.001123956326268465,
      "grad_norm": 19.31393814086914,
      "learning_rate": 7.000000000000001e-06,
      "loss": 2.24,
      "step": 7
    },
    {
      "epoch": 0.0012845215157353885,
      "grad_norm": 20.467914581298828,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.2559,
      "step": 8
    },
    {
      "epoch": 0.001445086705202312,
      "grad_norm": 16.672758102416992,
      "learning_rate": 9e-06,
      "loss": 2.178,
      "step": 9
    },
    {
      "epoch": 0.0016056518946692357,
      "grad_norm": 14.895868301391602,
      "learning_rate": 1e-05,
      "loss": 2.1822,
      "step": 10
    },
    {
      "epoch": 0.0017662170841361592,
      "grad_norm": 14.540231704711914,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 2.1189,
      "step": 11
    },
    {
      "epoch": 0.0019267822736030828,
      "grad_norm": 13.031673431396484,
      "learning_rate": 1.2e-05,
      "loss": 2.0964,
      "step": 12
    },
    {
      "epoch": 0.0020873474630700066,
      "grad_norm": 8.03255844116211,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 1.9923,
      "step": 13
    },
    {
      "epoch": 0.00224791265253693,
      "grad_norm": 11.225895881652832,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 1.9474,
      "step": 14
    },
    {
      "epoch": 0.0024084778420038534,
      "grad_norm": 5.122796058654785,
      "learning_rate": 1.5e-05,
      "loss": 1.8987,
      "step": 15
    },
    {
      "epoch": 0.002569043031470777,
      "grad_norm": 7.303050518035889,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.9861,
      "step": 16
    },
    {
      "epoch": 0.0027296082209377005,
      "grad_norm": 3.677480697631836,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 1.954,
      "step": 17
    },
    {
      "epoch": 0.002890173410404624,
      "grad_norm": 2.2276861667633057,
      "learning_rate": 1.8e-05,
      "loss": 1.9711,
      "step": 18
    },
    {
      "epoch": 0.0030507385998715477,
      "grad_norm": 1.7764992713928223,
      "learning_rate": 1.9e-05,
      "loss": 1.9135,
      "step": 19
    },
    {
      "epoch": 0.0032113037893384713,
      "grad_norm": 1.8619823455810547,
      "learning_rate": 2e-05,
      "loss": 1.8812,
      "step": 20
    },
    {
      "epoch": 0.003371868978805395,
      "grad_norm": 1.7041542530059814,
      "learning_rate": 2.1e-05,
      "loss": 1.8469,
      "step": 21
    },
    {
      "epoch": 0.0035324341682723185,
      "grad_norm": 1.5247529745101929,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 1.9096,
      "step": 22
    },
    {
      "epoch": 0.003692999357739242,
      "grad_norm": 1.4550870656967163,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 1.8792,
      "step": 23
    },
    {
      "epoch": 0.0038535645472061657,
      "grad_norm": 1.0546276569366455,
      "learning_rate": 2.4e-05,
      "loss": 1.7961,
      "step": 24
    },
    {
      "epoch": 0.00401412973667309,
      "grad_norm": 0.8481073379516602,
      "learning_rate": 2.5e-05,
      "loss": 1.8205,
      "step": 25
    },
    {
      "epoch": 0.004174694926140013,
      "grad_norm": 0.8418527245521545,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 1.8691,
      "step": 26
    },
    {
      "epoch": 0.004335260115606936,
      "grad_norm": 0.6527712941169739,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 1.839,
      "step": 27
    },
    {
      "epoch": 0.00449582530507386,
      "grad_norm": 0.5822799205780029,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 1.8679,
      "step": 28
    },
    {
      "epoch": 0.004656390494540783,
      "grad_norm": 0.6079473495483398,
      "learning_rate": 2.9e-05,
      "loss": 1.8828,
      "step": 29
    },
    {
      "epoch": 0.004816955684007707,
      "grad_norm": 0.6239740252494812,
      "learning_rate": 3e-05,
      "loss": 1.7649,
      "step": 30
    },
    {
      "epoch": 0.00497752087347463,
      "grad_norm": 0.5457854866981506,
      "learning_rate": 3.1e-05,
      "loss": 1.7975,
      "step": 31
    },
    {
      "epoch": 0.005138086062941554,
      "grad_norm": 0.7451772093772888,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 1.8618,
      "step": 32
    },
    {
      "epoch": 0.0052986512524084775,
      "grad_norm": 0.665291428565979,
      "learning_rate": 3.3e-05,
      "loss": 1.8,
      "step": 33
    },
    {
      "epoch": 0.005459216441875401,
      "grad_norm": 0.6924399733543396,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 1.8253,
      "step": 34
    },
    {
      "epoch": 0.005619781631342325,
      "grad_norm": 0.7987289428710938,
      "learning_rate": 3.5e-05,
      "loss": 1.9036,
      "step": 35
    },
    {
      "epoch": 0.005780346820809248,
      "grad_norm": 0.5785585641860962,
      "learning_rate": 3.6e-05,
      "loss": 1.8167,
      "step": 36
    },
    {
      "epoch": 0.005940912010276172,
      "grad_norm": 0.576931357383728,
      "learning_rate": 3.7e-05,
      "loss": 1.8636,
      "step": 37
    },
    {
      "epoch": 0.0061014771997430954,
      "grad_norm": 0.4122508466243744,
      "learning_rate": 3.8e-05,
      "loss": 1.7927,
      "step": 38
    },
    {
      "epoch": 0.006262042389210019,
      "grad_norm": 0.36012470722198486,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 1.8464,
      "step": 39
    },
    {
      "epoch": 0.006422607578676943,
      "grad_norm": 0.37618574500083923,
      "learning_rate": 4e-05,
      "loss": 1.7824,
      "step": 40
    },
    {
      "epoch": 0.006583172768143866,
      "grad_norm": 0.32919710874557495,
      "learning_rate": 4.1e-05,
      "loss": 1.6654,
      "step": 41
    },
    {
      "epoch": 0.00674373795761079,
      "grad_norm": 0.3425315320491791,
      "learning_rate": 4.2e-05,
      "loss": 1.7902,
      "step": 42
    },
    {
      "epoch": 0.006904303147077713,
      "grad_norm": 0.4032002091407776,
      "learning_rate": 4.3e-05,
      "loss": 1.7432,
      "step": 43
    },
    {
      "epoch": 0.007064868336544637,
      "grad_norm": 0.37265583872795105,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.8446,
      "step": 44
    },
    {
      "epoch": 0.0072254335260115606,
      "grad_norm": 0.4276624619960785,
      "learning_rate": 4.5e-05,
      "loss": 1.7643,
      "step": 45
    },
    {
      "epoch": 0.007385998715478484,
      "grad_norm": 0.39329883456230164,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.771,
      "step": 46
    },
    {
      "epoch": 0.007546563904945408,
      "grad_norm": 0.3743005394935608,
      "learning_rate": 4.7e-05,
      "loss": 1.7001,
      "step": 47
    },
    {
      "epoch": 0.007707129094412331,
      "grad_norm": 0.3529745638370514,
      "learning_rate": 4.8e-05,
      "loss": 1.8086,
      "step": 48
    },
    {
      "epoch": 0.007867694283879255,
      "grad_norm": 0.33023884892463684,
      "learning_rate": 4.9e-05,
      "loss": 1.7104,
      "step": 49
    },
    {
      "epoch": 0.00802825947334618,
      "grad_norm": 0.3542814552783966,
      "learning_rate": 5e-05,
      "loss": 1.7174,
      "step": 50
    },
    {
      "epoch": 0.008188824662813102,
      "grad_norm": 0.29632100462913513,
      "learning_rate": 5.1000000000000006e-05,
      "loss": 1.6615,
      "step": 51
    },
    {
      "epoch": 0.008349389852280027,
      "grad_norm": 0.3719535171985626,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 1.7675,
      "step": 52
    },
    {
      "epoch": 0.00850995504174695,
      "grad_norm": 0.3312448263168335,
      "learning_rate": 5.300000000000001e-05,
      "loss": 1.7556,
      "step": 53
    },
    {
      "epoch": 0.008670520231213872,
      "grad_norm": 0.31105342507362366,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 1.795,
      "step": 54
    },
    {
      "epoch": 0.008831085420680796,
      "grad_norm": 0.3008953928947449,
      "learning_rate": 5.500000000000001e-05,
      "loss": 1.6527,
      "step": 55
    },
    {
      "epoch": 0.00899165061014772,
      "grad_norm": 0.30014312267303467,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 1.7012,
      "step": 56
    },
    {
      "epoch": 0.009152215799614644,
      "grad_norm": 0.3198379874229431,
      "learning_rate": 5.6999999999999996e-05,
      "loss": 1.7207,
      "step": 57
    },
    {
      "epoch": 0.009312780989081566,
      "grad_norm": 0.3174596428871155,
      "learning_rate": 5.8e-05,
      "loss": 1.7685,
      "step": 58
    },
    {
      "epoch": 0.00947334617854849,
      "grad_norm": 0.3041522800922394,
      "learning_rate": 5.9e-05,
      "loss": 1.6611,
      "step": 59
    },
    {
      "epoch": 0.009633911368015413,
      "grad_norm": 0.30511003732681274,
      "learning_rate": 6e-05,
      "loss": 1.7535,
      "step": 60
    },
    {
      "epoch": 0.009794476557482338,
      "grad_norm": 0.3328545093536377,
      "learning_rate": 6.1e-05,
      "loss": 1.738,
      "step": 61
    },
    {
      "epoch": 0.00995504174694926,
      "grad_norm": 0.29658642411231995,
      "learning_rate": 6.2e-05,
      "loss": 1.6807,
      "step": 62
    },
    {
      "epoch": 0.010115606936416185,
      "grad_norm": 0.29087984561920166,
      "learning_rate": 6.3e-05,
      "loss": 1.6995,
      "step": 63
    },
    {
      "epoch": 0.010276172125883108,
      "grad_norm": 0.3203093111515045,
      "learning_rate": 6.400000000000001e-05,
      "loss": 1.6552,
      "step": 64
    },
    {
      "epoch": 0.010436737315350032,
      "grad_norm": 0.3448745012283325,
      "learning_rate": 6.500000000000001e-05,
      "loss": 1.684,
      "step": 65
    },
    {
      "epoch": 0.010597302504816955,
      "grad_norm": 0.30873870849609375,
      "learning_rate": 6.6e-05,
      "loss": 1.7397,
      "step": 66
    },
    {
      "epoch": 0.01075786769428388,
      "grad_norm": 0.3079901933670044,
      "learning_rate": 6.7e-05,
      "loss": 1.757,
      "step": 67
    },
    {
      "epoch": 0.010918432883750802,
      "grad_norm": 0.30332741141319275,
      "learning_rate": 6.800000000000001e-05,
      "loss": 1.6954,
      "step": 68
    },
    {
      "epoch": 0.011078998073217727,
      "grad_norm": 0.30804821848869324,
      "learning_rate": 6.9e-05,
      "loss": 1.6608,
      "step": 69
    },
    {
      "epoch": 0.01123956326268465,
      "grad_norm": 0.33730414509773254,
      "learning_rate": 7e-05,
      "loss": 1.6891,
      "step": 70
    },
    {
      "epoch": 0.011400128452151574,
      "grad_norm": 0.3238990306854248,
      "learning_rate": 7.1e-05,
      "loss": 1.6052,
      "step": 71
    },
    {
      "epoch": 0.011560693641618497,
      "grad_norm": 0.32166120409965515,
      "learning_rate": 7.2e-05,
      "loss": 1.687,
      "step": 72
    },
    {
      "epoch": 0.011721258831085421,
      "grad_norm": 0.3208368122577667,
      "learning_rate": 7.3e-05,
      "loss": 1.6878,
      "step": 73
    },
    {
      "epoch": 0.011881824020552344,
      "grad_norm": 0.34366941452026367,
      "learning_rate": 7.4e-05,
      "loss": 1.7257,
      "step": 74
    },
    {
      "epoch": 0.012042389210019268,
      "grad_norm": 0.3077256381511688,
      "learning_rate": 7.500000000000001e-05,
      "loss": 1.6705,
      "step": 75
    },
    {
      "epoch": 0.012202954399486191,
      "grad_norm": 0.3369261622428894,
      "learning_rate": 7.6e-05,
      "loss": 1.7493,
      "step": 76
    },
    {
      "epoch": 0.012363519588953115,
      "grad_norm": 0.322565495967865,
      "learning_rate": 7.7e-05,
      "loss": 1.7214,
      "step": 77
    },
    {
      "epoch": 0.012524084778420038,
      "grad_norm": 0.30557385087013245,
      "learning_rate": 7.800000000000001e-05,
      "loss": 1.6356,
      "step": 78
    },
    {
      "epoch": 0.012684649967886963,
      "grad_norm": 0.3298359215259552,
      "learning_rate": 7.900000000000001e-05,
      "loss": 1.8063,
      "step": 79
    },
    {
      "epoch": 0.012845215157353885,
      "grad_norm": 0.3066646158695221,
      "learning_rate": 8e-05,
      "loss": 1.6539,
      "step": 80
    },
    {
      "epoch": 0.01300578034682081,
      "grad_norm": 0.29285869002342224,
      "learning_rate": 8.1e-05,
      "loss": 1.6448,
      "step": 81
    },
    {
      "epoch": 0.013166345536287732,
      "grad_norm": 0.3066501021385193,
      "learning_rate": 8.2e-05,
      "loss": 1.6828,
      "step": 82
    },
    {
      "epoch": 0.013326910725754657,
      "grad_norm": 0.31421324610710144,
      "learning_rate": 8.3e-05,
      "loss": 1.6947,
      "step": 83
    },
    {
      "epoch": 0.01348747591522158,
      "grad_norm": 0.33448484539985657,
      "learning_rate": 8.4e-05,
      "loss": 1.7222,
      "step": 84
    },
    {
      "epoch": 0.013648041104688504,
      "grad_norm": 0.3122430443763733,
      "learning_rate": 8.5e-05,
      "loss": 1.6433,
      "step": 85
    },
    {
      "epoch": 0.013808606294155427,
      "grad_norm": 0.31773683428764343,
      "learning_rate": 8.6e-05,
      "loss": 1.7606,
      "step": 86
    },
    {
      "epoch": 0.013969171483622351,
      "grad_norm": 0.29960861802101135,
      "learning_rate": 8.7e-05,
      "loss": 1.6992,
      "step": 87
    },
    {
      "epoch": 0.014129736673089274,
      "grad_norm": 0.3499071002006531,
      "learning_rate": 8.800000000000001e-05,
      "loss": 1.7152,
      "step": 88
    },
    {
      "epoch": 0.014290301862556198,
      "grad_norm": 0.3297984004020691,
      "learning_rate": 8.900000000000001e-05,
      "loss": 1.731,
      "step": 89
    },
    {
      "epoch": 0.014450867052023121,
      "grad_norm": 0.3231171667575836,
      "learning_rate": 9e-05,
      "loss": 1.6932,
      "step": 90
    },
    {
      "epoch": 0.014611432241490046,
      "grad_norm": 0.31853851675987244,
      "learning_rate": 9.1e-05,
      "loss": 1.7144,
      "step": 91
    },
    {
      "epoch": 0.014771997430956968,
      "grad_norm": 0.31341439485549927,
      "learning_rate": 9.200000000000001e-05,
      "loss": 1.6652,
      "step": 92
    },
    {
      "epoch": 0.014932562620423893,
      "grad_norm": 0.2927674353122711,
      "learning_rate": 9.300000000000001e-05,
      "loss": 1.6714,
      "step": 93
    },
    {
      "epoch": 0.015093127809890815,
      "grad_norm": 0.3104013502597809,
      "learning_rate": 9.4e-05,
      "loss": 1.65,
      "step": 94
    },
    {
      "epoch": 0.01525369299935774,
      "grad_norm": 0.3095431625843048,
      "learning_rate": 9.5e-05,
      "loss": 1.6997,
      "step": 95
    },
    {
      "epoch": 0.015414258188824663,
      "grad_norm": 0.3033742606639862,
      "learning_rate": 9.6e-05,
      "loss": 1.7215,
      "step": 96
    },
    {
      "epoch": 0.015574823378291587,
      "grad_norm": 0.3096693456172943,
      "learning_rate": 9.7e-05,
      "loss": 1.6991,
      "step": 97
    },
    {
      "epoch": 0.01573538856775851,
      "grad_norm": 0.30744585394859314,
      "learning_rate": 9.8e-05,
      "loss": 1.6688,
      "step": 98
    },
    {
      "epoch": 0.015895953757225433,
      "grad_norm": 0.30216723680496216,
      "learning_rate": 9.900000000000001e-05,
      "loss": 1.6727,
      "step": 99
    },
    {
      "epoch": 0.01605651894669236,
      "grad_norm": 0.31961628794670105,
      "learning_rate": 0.0001,
      "loss": 1.6672,
      "step": 100
    },
    {
      "epoch": 0.01621708413615928,
      "grad_norm": 0.29976239800453186,
      "learning_rate": 0.0001,
      "loss": 1.643,
      "step": 101
    },
    {
      "epoch": 0.016377649325626204,
      "grad_norm": 0.335050493478775,
      "learning_rate": 0.0001,
      "loss": 1.6817,
      "step": 102
    },
    {
      "epoch": 0.016538214515093127,
      "grad_norm": 0.31094300746917725,
      "learning_rate": 0.0001,
      "loss": 1.6277,
      "step": 103
    },
    {
      "epoch": 0.016698779704560053,
      "grad_norm": 0.32502052187919617,
      "learning_rate": 0.0001,
      "loss": 1.765,
      "step": 104
    },
    {
      "epoch": 0.016859344894026976,
      "grad_norm": 0.33868634700775146,
      "learning_rate": 0.0001,
      "loss": 1.6906,
      "step": 105
    },
    {
      "epoch": 0.0170199100834939,
      "grad_norm": 0.29988473653793335,
      "learning_rate": 0.0001,
      "loss": 1.6936,
      "step": 106
    },
    {
      "epoch": 0.01718047527296082,
      "grad_norm": 0.3123616874217987,
      "learning_rate": 0.0001,
      "loss": 1.6805,
      "step": 107
    },
    {
      "epoch": 0.017341040462427744,
      "grad_norm": 0.3224555253982544,
      "learning_rate": 0.0001,
      "loss": 1.6686,
      "step": 108
    },
    {
      "epoch": 0.01750160565189467,
      "grad_norm": 0.31106632947921753,
      "learning_rate": 0.0001,
      "loss": 1.6977,
      "step": 109
    },
    {
      "epoch": 0.017662170841361593,
      "grad_norm": 0.30861422419548035,
      "learning_rate": 0.0001,
      "loss": 1.6124,
      "step": 110
    },
    {
      "epoch": 0.017822736030828516,
      "grad_norm": 0.31756776571273804,
      "learning_rate": 0.0001,
      "loss": 1.7024,
      "step": 111
    },
    {
      "epoch": 0.01798330122029544,
      "grad_norm": 0.3358090817928314,
      "learning_rate": 0.0001,
      "loss": 1.7037,
      "step": 112
    },
    {
      "epoch": 0.018143866409762364,
      "grad_norm": 0.3067599833011627,
      "learning_rate": 0.0001,
      "loss": 1.6947,
      "step": 113
    },
    {
      "epoch": 0.018304431599229287,
      "grad_norm": 0.33327704668045044,
      "learning_rate": 0.0001,
      "loss": 1.6335,
      "step": 114
    },
    {
      "epoch": 0.01846499678869621,
      "grad_norm": 0.2796766459941864,
      "learning_rate": 0.0001,
      "loss": 1.5963,
      "step": 115
    },
    {
      "epoch": 0.018625561978163133,
      "grad_norm": 0.3339906632900238,
      "learning_rate": 0.0001,
      "loss": 1.6542,
      "step": 116
    },
    {
      "epoch": 0.01878612716763006,
      "grad_norm": 0.3105831742286682,
      "learning_rate": 0.0001,
      "loss": 1.6735,
      "step": 117
    },
    {
      "epoch": 0.01894669235709698,
      "grad_norm": 0.3182494640350342,
      "learning_rate": 0.0001,
      "loss": 1.6692,
      "step": 118
    },
    {
      "epoch": 0.019107257546563904,
      "grad_norm": 0.3525417745113373,
      "learning_rate": 0.0001,
      "loss": 1.7499,
      "step": 119
    },
    {
      "epoch": 0.019267822736030827,
      "grad_norm": 0.31575801968574524,
      "learning_rate": 0.0001,
      "loss": 1.6368,
      "step": 120
    },
    {
      "epoch": 0.019428387925497753,
      "grad_norm": 0.35428088903427124,
      "learning_rate": 0.0001,
      "loss": 1.8006,
      "step": 121
    },
    {
      "epoch": 0.019588953114964676,
      "grad_norm": 0.3589598536491394,
      "learning_rate": 0.0001,
      "loss": 1.717,
      "step": 122
    },
    {
      "epoch": 0.0197495183044316,
      "grad_norm": 0.32543036341667175,
      "learning_rate": 0.0001,
      "loss": 1.8,
      "step": 123
    },
    {
      "epoch": 0.01991008349389852,
      "grad_norm": 0.298537939786911,
      "learning_rate": 0.0001,
      "loss": 1.6333,
      "step": 124
    },
    {
      "epoch": 0.020070648683365448,
      "grad_norm": 0.3392146825790405,
      "learning_rate": 0.0001,
      "loss": 1.6314,
      "step": 125
    },
    {
      "epoch": 0.02023121387283237,
      "grad_norm": 0.29613935947418213,
      "learning_rate": 0.0001,
      "loss": 1.6209,
      "step": 126
    },
    {
      "epoch": 0.020391779062299293,
      "grad_norm": 0.3168414533138275,
      "learning_rate": 0.0001,
      "loss": 1.5678,
      "step": 127
    },
    {
      "epoch": 0.020552344251766216,
      "grad_norm": 0.2909003794193268,
      "learning_rate": 0.0001,
      "loss": 1.6471,
      "step": 128
    },
    {
      "epoch": 0.020712909441233142,
      "grad_norm": 0.3160809576511383,
      "learning_rate": 0.0001,
      "loss": 1.6526,
      "step": 129
    },
    {
      "epoch": 0.020873474630700065,
      "grad_norm": 0.296406090259552,
      "learning_rate": 0.0001,
      "loss": 1.7082,
      "step": 130
    },
    {
      "epoch": 0.021034039820166987,
      "grad_norm": 0.36247017979621887,
      "learning_rate": 0.0001,
      "loss": 1.7076,
      "step": 131
    },
    {
      "epoch": 0.02119460500963391,
      "grad_norm": 0.294392466545105,
      "learning_rate": 0.0001,
      "loss": 1.6495,
      "step": 132
    },
    {
      "epoch": 0.021355170199100836,
      "grad_norm": 0.33377623558044434,
      "learning_rate": 0.0001,
      "loss": 1.6448,
      "step": 133
    },
    {
      "epoch": 0.02151573538856776,
      "grad_norm": 0.3045656383037567,
      "learning_rate": 0.0001,
      "loss": 1.6668,
      "step": 134
    },
    {
      "epoch": 0.02167630057803468,
      "grad_norm": 0.3585946559906006,
      "learning_rate": 0.0001,
      "loss": 1.6351,
      "step": 135
    },
    {
      "epoch": 0.021836865767501604,
      "grad_norm": 0.29203692078590393,
      "learning_rate": 0.0001,
      "loss": 1.5617,
      "step": 136
    },
    {
      "epoch": 0.02199743095696853,
      "grad_norm": 0.3085068166255951,
      "learning_rate": 0.0001,
      "loss": 1.7582,
      "step": 137
    },
    {
      "epoch": 0.022157996146435453,
      "grad_norm": 0.30218175053596497,
      "learning_rate": 0.0001,
      "loss": 1.6359,
      "step": 138
    },
    {
      "epoch": 0.022318561335902376,
      "grad_norm": 0.32938602566719055,
      "learning_rate": 0.0001,
      "loss": 1.5966,
      "step": 139
    },
    {
      "epoch": 0.0224791265253693,
      "grad_norm": 0.32060807943344116,
      "learning_rate": 0.0001,
      "loss": 1.7,
      "step": 140
    },
    {
      "epoch": 0.022639691714836225,
      "grad_norm": 0.29182660579681396,
      "learning_rate": 0.0001,
      "loss": 1.6386,
      "step": 141
    },
    {
      "epoch": 0.022800256904303148,
      "grad_norm": 0.3093222677707672,
      "learning_rate": 0.0001,
      "loss": 1.6954,
      "step": 142
    },
    {
      "epoch": 0.02296082209377007,
      "grad_norm": 0.30981460213661194,
      "learning_rate": 0.0001,
      "loss": 1.6187,
      "step": 143
    },
    {
      "epoch": 0.023121387283236993,
      "grad_norm": 0.3060247004032135,
      "learning_rate": 0.0001,
      "loss": 1.6366,
      "step": 144
    },
    {
      "epoch": 0.02328195247270392,
      "grad_norm": 0.34533223509788513,
      "learning_rate": 0.0001,
      "loss": 1.6014,
      "step": 145
    },
    {
      "epoch": 0.023442517662170842,
      "grad_norm": 0.32339876890182495,
      "learning_rate": 0.0001,
      "loss": 1.7605,
      "step": 146
    },
    {
      "epoch": 0.023603082851637765,
      "grad_norm": 0.29571667313575745,
      "learning_rate": 0.0001,
      "loss": 1.6328,
      "step": 147
    },
    {
      "epoch": 0.023763648041104687,
      "grad_norm": 0.3205752670764923,
      "learning_rate": 0.0001,
      "loss": 1.6198,
      "step": 148
    },
    {
      "epoch": 0.023924213230571614,
      "grad_norm": 0.29312294721603394,
      "learning_rate": 0.0001,
      "loss": 1.6381,
      "step": 149
    },
    {
      "epoch": 0.024084778420038536,
      "grad_norm": 0.28196433186531067,
      "learning_rate": 0.0001,
      "loss": 1.6155,
      "step": 150
    },
    {
      "epoch": 0.02424534360950546,
      "grad_norm": 0.3127625286579132,
      "learning_rate": 0.0001,
      "loss": 1.7451,
      "step": 151
    },
    {
      "epoch": 0.024405908798972382,
      "grad_norm": 0.3327120244503021,
      "learning_rate": 0.0001,
      "loss": 1.7074,
      "step": 152
    },
    {
      "epoch": 0.024566473988439308,
      "grad_norm": 0.49247097969055176,
      "learning_rate": 0.0001,
      "loss": 1.74,
      "step": 153
    },
    {
      "epoch": 0.02472703917790623,
      "grad_norm": 0.3472457528114319,
      "learning_rate": 0.0001,
      "loss": 1.6715,
      "step": 154
    },
    {
      "epoch": 0.024887604367373153,
      "grad_norm": 0.3496563732624054,
      "learning_rate": 0.0001,
      "loss": 1.6577,
      "step": 155
    },
    {
      "epoch": 0.025048169556840076,
      "grad_norm": 0.335408091545105,
      "learning_rate": 0.0001,
      "loss": 1.6932,
      "step": 156
    },
    {
      "epoch": 0.025208734746307002,
      "grad_norm": 0.31248655915260315,
      "learning_rate": 0.0001,
      "loss": 1.697,
      "step": 157
    },
    {
      "epoch": 0.025369299935773925,
      "grad_norm": 0.2786000967025757,
      "learning_rate": 0.0001,
      "loss": 1.57,
      "step": 158
    },
    {
      "epoch": 0.025529865125240848,
      "grad_norm": 0.2947397530078888,
      "learning_rate": 0.0001,
      "loss": 1.6523,
      "step": 159
    },
    {
      "epoch": 0.02569043031470777,
      "grad_norm": 0.34668445587158203,
      "learning_rate": 0.0001,
      "loss": 1.6965,
      "step": 160
    },
    {
      "epoch": 0.025850995504174697,
      "grad_norm": 0.2984938621520996,
      "learning_rate": 0.0001,
      "loss": 1.6581,
      "step": 161
    },
    {
      "epoch": 0.02601156069364162,
      "grad_norm": 0.3221776783466339,
      "learning_rate": 0.0001,
      "loss": 1.604,
      "step": 162
    },
    {
      "epoch": 0.026172125883108542,
      "grad_norm": 0.28798162937164307,
      "learning_rate": 0.0001,
      "loss": 1.5953,
      "step": 163
    },
    {
      "epoch": 0.026332691072575465,
      "grad_norm": 0.32310038805007935,
      "learning_rate": 0.0001,
      "loss": 1.6954,
      "step": 164
    },
    {
      "epoch": 0.026493256262042388,
      "grad_norm": 0.2939600944519043,
      "learning_rate": 0.0001,
      "loss": 1.6698,
      "step": 165
    },
    {
      "epoch": 0.026653821451509314,
      "grad_norm": 0.2926805019378662,
      "learning_rate": 0.0001,
      "loss": 1.6898,
      "step": 166
    },
    {
      "epoch": 0.026814386640976236,
      "grad_norm": 0.3102514445781708,
      "learning_rate": 0.0001,
      "loss": 1.7258,
      "step": 167
    },
    {
      "epoch": 0.02697495183044316,
      "grad_norm": 0.28639596700668335,
      "learning_rate": 0.0001,
      "loss": 1.7035,
      "step": 168
    },
    {
      "epoch": 0.027135517019910082,
      "grad_norm": 0.34116673469543457,
      "learning_rate": 0.0001,
      "loss": 1.6692,
      "step": 169
    },
    {
      "epoch": 0.027296082209377008,
      "grad_norm": 0.28041908144950867,
      "learning_rate": 0.0001,
      "loss": 1.6578,
      "step": 170
    },
    {
      "epoch": 0.02745664739884393,
      "grad_norm": 0.3725098669528961,
      "learning_rate": 0.0001,
      "loss": 1.6975,
      "step": 171
    },
    {
      "epoch": 0.027617212588310854,
      "grad_norm": 0.31921643018722534,
      "learning_rate": 0.0001,
      "loss": 1.6892,
      "step": 172
    },
    {
      "epoch": 0.027777777777777776,
      "grad_norm": 0.3719729781150818,
      "learning_rate": 0.0001,
      "loss": 1.7543,
      "step": 173
    },
    {
      "epoch": 0.027938342967244702,
      "grad_norm": 0.32456451654434204,
      "learning_rate": 0.0001,
      "loss": 1.6563,
      "step": 174
    },
    {
      "epoch": 0.028098908156711625,
      "grad_norm": 0.3301026225090027,
      "learning_rate": 0.0001,
      "loss": 1.6022,
      "step": 175
    },
    {
      "epoch": 0.028259473346178548,
      "grad_norm": 0.3518262505531311,
      "learning_rate": 0.0001,
      "loss": 1.6914,
      "step": 176
    },
    {
      "epoch": 0.02842003853564547,
      "grad_norm": 0.27261704206466675,
      "learning_rate": 0.0001,
      "loss": 1.5933,
      "step": 177
    },
    {
      "epoch": 0.028580603725112397,
      "grad_norm": 0.33590003848075867,
      "learning_rate": 0.0001,
      "loss": 1.6827,
      "step": 178
    },
    {
      "epoch": 0.02874116891457932,
      "grad_norm": 0.3073245882987976,
      "learning_rate": 0.0001,
      "loss": 1.6588,
      "step": 179
    },
    {
      "epoch": 0.028901734104046242,
      "grad_norm": 0.2971007525920868,
      "learning_rate": 0.0001,
      "loss": 1.641,
      "step": 180
    },
    {
      "epoch": 0.029062299293513165,
      "grad_norm": 0.29228779673576355,
      "learning_rate": 0.0001,
      "loss": 1.6388,
      "step": 181
    },
    {
      "epoch": 0.02922286448298009,
      "grad_norm": 0.3131601810455322,
      "learning_rate": 0.0001,
      "loss": 1.6233,
      "step": 182
    },
    {
      "epoch": 0.029383429672447014,
      "grad_norm": 0.30442118644714355,
      "learning_rate": 0.0001,
      "loss": 1.7528,
      "step": 183
    },
    {
      "epoch": 0.029543994861913937,
      "grad_norm": 0.2892727553844452,
      "learning_rate": 0.0001,
      "loss": 1.676,
      "step": 184
    },
    {
      "epoch": 0.02970456005138086,
      "grad_norm": 0.29788944125175476,
      "learning_rate": 0.0001,
      "loss": 1.5781,
      "step": 185
    },
    {
      "epoch": 0.029865125240847785,
      "grad_norm": 0.3052654266357422,
      "learning_rate": 0.0001,
      "loss": 1.6033,
      "step": 186
    },
    {
      "epoch": 0.030025690430314708,
      "grad_norm": 0.27221453189849854,
      "learning_rate": 0.0001,
      "loss": 1.6477,
      "step": 187
    },
    {
      "epoch": 0.03018625561978163,
      "grad_norm": 0.29384908080101013,
      "learning_rate": 0.0001,
      "loss": 1.6009,
      "step": 188
    },
    {
      "epoch": 0.030346820809248554,
      "grad_norm": 0.2946358919143677,
      "learning_rate": 0.0001,
      "loss": 1.6412,
      "step": 189
    },
    {
      "epoch": 0.03050738599871548,
      "grad_norm": 0.29987457394599915,
      "learning_rate": 0.0001,
      "loss": 1.6494,
      "step": 190
    },
    {
      "epoch": 0.030667951188182403,
      "grad_norm": 0.29672113060951233,
      "learning_rate": 0.0001,
      "loss": 1.675,
      "step": 191
    },
    {
      "epoch": 0.030828516377649325,
      "grad_norm": 0.2981245815753937,
      "learning_rate": 0.0001,
      "loss": 1.5827,
      "step": 192
    },
    {
      "epoch": 0.030989081567116248,
      "grad_norm": 0.30196821689605713,
      "learning_rate": 0.0001,
      "loss": 1.6423,
      "step": 193
    },
    {
      "epoch": 0.031149646756583174,
      "grad_norm": 0.3073696494102478,
      "learning_rate": 0.0001,
      "loss": 1.7611,
      "step": 194
    },
    {
      "epoch": 0.0313102119460501,
      "grad_norm": 0.3277694284915924,
      "learning_rate": 0.0001,
      "loss": 1.7411,
      "step": 195
    },
    {
      "epoch": 0.03147077713551702,
      "grad_norm": 0.3470703363418579,
      "learning_rate": 0.0001,
      "loss": 1.7407,
      "step": 196
    },
    {
      "epoch": 0.03163134232498394,
      "grad_norm": 0.2917807400226593,
      "learning_rate": 0.0001,
      "loss": 1.6177,
      "step": 197
    },
    {
      "epoch": 0.031791907514450865,
      "grad_norm": 0.3175638020038605,
      "learning_rate": 0.0001,
      "loss": 1.582,
      "step": 198
    },
    {
      "epoch": 0.03195247270391779,
      "grad_norm": 0.2842324674129486,
      "learning_rate": 0.0001,
      "loss": 1.6694,
      "step": 199
    },
    {
      "epoch": 0.03211303789338472,
      "grad_norm": 0.29128801822662354,
      "learning_rate": 0.0001,
      "loss": 1.6467,
      "step": 200
    },
    {
      "epoch": 0.03227360308285164,
      "grad_norm": 0.28581851720809937,
      "learning_rate": 0.0001,
      "loss": 1.6484,
      "step": 201
    },
    {
      "epoch": 0.03243416827231856,
      "grad_norm": 0.28554877638816833,
      "learning_rate": 0.0001,
      "loss": 1.6476,
      "step": 202
    },
    {
      "epoch": 0.032594733461785486,
      "grad_norm": 0.4070573151111603,
      "learning_rate": 0.0001,
      "loss": 1.6423,
      "step": 203
    },
    {
      "epoch": 0.03275529865125241,
      "grad_norm": 0.2945961654186249,
      "learning_rate": 0.0001,
      "loss": 1.7596,
      "step": 204
    },
    {
      "epoch": 0.03291586384071933,
      "grad_norm": 0.3131120502948761,
      "learning_rate": 0.0001,
      "loss": 1.649,
      "step": 205
    },
    {
      "epoch": 0.033076429030186254,
      "grad_norm": 0.29569193720817566,
      "learning_rate": 0.0001,
      "loss": 1.7106,
      "step": 206
    },
    {
      "epoch": 0.033236994219653176,
      "grad_norm": 0.30611786246299744,
      "learning_rate": 0.0001,
      "loss": 1.6625,
      "step": 207
    },
    {
      "epoch": 0.033397559409120106,
      "grad_norm": 0.27818742394447327,
      "learning_rate": 0.0001,
      "loss": 1.6802,
      "step": 208
    },
    {
      "epoch": 0.03355812459858703,
      "grad_norm": 0.2986738979816437,
      "learning_rate": 0.0001,
      "loss": 1.7492,
      "step": 209
    },
    {
      "epoch": 0.03371868978805395,
      "grad_norm": 0.2596496343612671,
      "learning_rate": 0.0001,
      "loss": 1.5599,
      "step": 210
    },
    {
      "epoch": 0.033879254977520874,
      "grad_norm": 0.27559900283813477,
      "learning_rate": 0.0001,
      "loss": 1.6069,
      "step": 211
    },
    {
      "epoch": 0.0340398201669878,
      "grad_norm": 0.2860763370990753,
      "learning_rate": 0.0001,
      "loss": 1.5767,
      "step": 212
    },
    {
      "epoch": 0.03420038535645472,
      "grad_norm": 0.2946445345878601,
      "learning_rate": 0.0001,
      "loss": 1.6497,
      "step": 213
    },
    {
      "epoch": 0.03436095054592164,
      "grad_norm": 0.3171544373035431,
      "learning_rate": 0.0001,
      "loss": 1.6091,
      "step": 214
    },
    {
      "epoch": 0.034521515735388565,
      "grad_norm": 0.2965378761291504,
      "learning_rate": 0.0001,
      "loss": 1.6472,
      "step": 215
    },
    {
      "epoch": 0.03468208092485549,
      "grad_norm": 0.27333325147628784,
      "learning_rate": 0.0001,
      "loss": 1.5804,
      "step": 216
    },
    {
      "epoch": 0.03484264611432242,
      "grad_norm": 0.29446908831596375,
      "learning_rate": 0.0001,
      "loss": 1.657,
      "step": 217
    },
    {
      "epoch": 0.03500321130378934,
      "grad_norm": 0.3440585732460022,
      "learning_rate": 0.0001,
      "loss": 1.6271,
      "step": 218
    },
    {
      "epoch": 0.03516377649325626,
      "grad_norm": 0.2965490520000458,
      "learning_rate": 0.0001,
      "loss": 1.6513,
      "step": 219
    },
    {
      "epoch": 0.035324341682723186,
      "grad_norm": 0.2846764624118805,
      "learning_rate": 0.0001,
      "loss": 1.6948,
      "step": 220
    },
    {
      "epoch": 0.03548490687219011,
      "grad_norm": 0.27987420558929443,
      "learning_rate": 0.0001,
      "loss": 1.6841,
      "step": 221
    },
    {
      "epoch": 0.03564547206165703,
      "grad_norm": 0.29974129796028137,
      "learning_rate": 0.0001,
      "loss": 1.6245,
      "step": 222
    },
    {
      "epoch": 0.035806037251123954,
      "grad_norm": 0.28039097785949707,
      "learning_rate": 0.0001,
      "loss": 1.6591,
      "step": 223
    },
    {
      "epoch": 0.03596660244059088,
      "grad_norm": 0.3027125597000122,
      "learning_rate": 0.0001,
      "loss": 1.664,
      "step": 224
    },
    {
      "epoch": 0.036127167630057806,
      "grad_norm": 0.308106392621994,
      "learning_rate": 0.0001,
      "loss": 1.7404,
      "step": 225
    },
    {
      "epoch": 0.03628773281952473,
      "grad_norm": 0.2665157914161682,
      "learning_rate": 0.0001,
      "loss": 1.638,
      "step": 226
    },
    {
      "epoch": 0.03644829800899165,
      "grad_norm": 0.2794928252696991,
      "learning_rate": 0.0001,
      "loss": 1.5889,
      "step": 227
    },
    {
      "epoch": 0.036608863198458574,
      "grad_norm": 0.33470234274864197,
      "learning_rate": 0.0001,
      "loss": 1.6182,
      "step": 228
    },
    {
      "epoch": 0.0367694283879255,
      "grad_norm": 0.27957892417907715,
      "learning_rate": 0.0001,
      "loss": 1.6607,
      "step": 229
    },
    {
      "epoch": 0.03692999357739242,
      "grad_norm": 0.287702739238739,
      "learning_rate": 0.0001,
      "loss": 1.592,
      "step": 230
    },
    {
      "epoch": 0.03709055876685934,
      "grad_norm": 0.27235883474349976,
      "learning_rate": 0.0001,
      "loss": 1.5924,
      "step": 231
    },
    {
      "epoch": 0.037251123956326265,
      "grad_norm": 0.24706554412841797,
      "learning_rate": 0.0001,
      "loss": 1.4878,
      "step": 232
    },
    {
      "epoch": 0.037411689145793195,
      "grad_norm": 0.28109508752822876,
      "learning_rate": 0.0001,
      "loss": 1.66,
      "step": 233
    },
    {
      "epoch": 0.03757225433526012,
      "grad_norm": 0.301008015871048,
      "learning_rate": 0.0001,
      "loss": 1.7872,
      "step": 234
    },
    {
      "epoch": 0.03773281952472704,
      "grad_norm": 0.2830420732498169,
      "learning_rate": 0.0001,
      "loss": 1.6013,
      "step": 235
    },
    {
      "epoch": 0.03789338471419396,
      "grad_norm": 0.26912617683410645,
      "learning_rate": 0.0001,
      "loss": 1.6176,
      "step": 236
    },
    {
      "epoch": 0.038053949903660886,
      "grad_norm": 0.2907160818576813,
      "learning_rate": 0.0001,
      "loss": 1.6932,
      "step": 237
    },
    {
      "epoch": 0.03821451509312781,
      "grad_norm": 0.3068007826805115,
      "learning_rate": 0.0001,
      "loss": 1.685,
      "step": 238
    },
    {
      "epoch": 0.03837508028259473,
      "grad_norm": 0.27787286043167114,
      "learning_rate": 0.0001,
      "loss": 1.7064,
      "step": 239
    },
    {
      "epoch": 0.038535645472061654,
      "grad_norm": 0.30287325382232666,
      "learning_rate": 0.0001,
      "loss": 1.6296,
      "step": 240
    },
    {
      "epoch": 0.038696210661528584,
      "grad_norm": 0.27720868587493896,
      "learning_rate": 0.0001,
      "loss": 1.6549,
      "step": 241
    },
    {
      "epoch": 0.038856775850995506,
      "grad_norm": 0.2869109809398651,
      "learning_rate": 0.0001,
      "loss": 1.7045,
      "step": 242
    },
    {
      "epoch": 0.03901734104046243,
      "grad_norm": 0.30262935161590576,
      "learning_rate": 0.0001,
      "loss": 1.6358,
      "step": 243
    },
    {
      "epoch": 0.03917790622992935,
      "grad_norm": 0.2900177538394928,
      "learning_rate": 0.0001,
      "loss": 1.6479,
      "step": 244
    },
    {
      "epoch": 0.039338471419396275,
      "grad_norm": 0.29951053857803345,
      "learning_rate": 0.0001,
      "loss": 1.6419,
      "step": 245
    },
    {
      "epoch": 0.0394990366088632,
      "grad_norm": 0.30172184109687805,
      "learning_rate": 0.0001,
      "loss": 1.7259,
      "step": 246
    },
    {
      "epoch": 0.03965960179833012,
      "grad_norm": 0.2781960070133209,
      "learning_rate": 0.0001,
      "loss": 1.6788,
      "step": 247
    },
    {
      "epoch": 0.03982016698779704,
      "grad_norm": 0.2786746025085449,
      "learning_rate": 0.0001,
      "loss": 1.6411,
      "step": 248
    },
    {
      "epoch": 0.03998073217726397,
      "grad_norm": 0.273116797208786,
      "learning_rate": 0.0001,
      "loss": 1.6447,
      "step": 249
    },
    {
      "epoch": 0.040141297366730895,
      "grad_norm": 0.28072962164878845,
      "learning_rate": 0.0001,
      "loss": 1.6603,
      "step": 250
    },
    {
      "epoch": 0.04030186255619782,
      "grad_norm": 0.28691452741622925,
      "learning_rate": 0.0001,
      "loss": 1.6517,
      "step": 251
    },
    {
      "epoch": 0.04046242774566474,
      "grad_norm": 0.280073344707489,
      "learning_rate": 0.0001,
      "loss": 1.6652,
      "step": 252
    },
    {
      "epoch": 0.04062299293513166,
      "grad_norm": 0.28682181239128113,
      "learning_rate": 0.0001,
      "loss": 1.6401,
      "step": 253
    },
    {
      "epoch": 0.040783558124598586,
      "grad_norm": 0.2880575656890869,
      "learning_rate": 0.0001,
      "loss": 1.6264,
      "step": 254
    },
    {
      "epoch": 0.04094412331406551,
      "grad_norm": 0.2657603919506073,
      "learning_rate": 0.0001,
      "loss": 1.6107,
      "step": 255
    },
    {
      "epoch": 0.04110468850353243,
      "grad_norm": 0.3001009523868561,
      "learning_rate": 0.0001,
      "loss": 1.7025,
      "step": 256
    },
    {
      "epoch": 0.04126525369299936,
      "grad_norm": 0.2876819670200348,
      "learning_rate": 0.0001,
      "loss": 1.6628,
      "step": 257
    },
    {
      "epoch": 0.041425818882466284,
      "grad_norm": 0.27692848443984985,
      "learning_rate": 0.0001,
      "loss": 1.6402,
      "step": 258
    },
    {
      "epoch": 0.041586384071933206,
      "grad_norm": 0.27744626998901367,
      "learning_rate": 0.0001,
      "loss": 1.6806,
      "step": 259
    },
    {
      "epoch": 0.04174694926140013,
      "grad_norm": 0.27917224168777466,
      "learning_rate": 0.0001,
      "loss": 1.612,
      "step": 260
    },
    {
      "epoch": 0.04190751445086705,
      "grad_norm": 0.3135955035686493,
      "learning_rate": 0.0001,
      "loss": 1.6123,
      "step": 261
    },
    {
      "epoch": 0.042068079640333975,
      "grad_norm": 0.28715232014656067,
      "learning_rate": 0.0001,
      "loss": 1.7021,
      "step": 262
    },
    {
      "epoch": 0.0422286448298009,
      "grad_norm": 0.2720044255256653,
      "learning_rate": 0.0001,
      "loss": 1.5852,
      "step": 263
    },
    {
      "epoch": 0.04238921001926782,
      "grad_norm": 0.2853310704231262,
      "learning_rate": 0.0001,
      "loss": 1.5508,
      "step": 264
    },
    {
      "epoch": 0.04254977520873475,
      "grad_norm": 0.27966177463531494,
      "learning_rate": 0.0001,
      "loss": 1.6642,
      "step": 265
    },
    {
      "epoch": 0.04271034039820167,
      "grad_norm": 0.2797499895095825,
      "learning_rate": 0.0001,
      "loss": 1.6073,
      "step": 266
    },
    {
      "epoch": 0.042870905587668595,
      "grad_norm": 0.27873483300209045,
      "learning_rate": 0.0001,
      "loss": 1.6507,
      "step": 267
    },
    {
      "epoch": 0.04303147077713552,
      "grad_norm": 0.28432849049568176,
      "learning_rate": 0.0001,
      "loss": 1.6058,
      "step": 268
    },
    {
      "epoch": 0.04319203596660244,
      "grad_norm": 0.26404669880867004,
      "learning_rate": 0.0001,
      "loss": 1.6179,
      "step": 269
    },
    {
      "epoch": 0.04335260115606936,
      "grad_norm": 0.2764602303504944,
      "learning_rate": 0.0001,
      "loss": 1.6378,
      "step": 270
    },
    {
      "epoch": 0.043513166345536286,
      "grad_norm": 0.34712472558021545,
      "learning_rate": 0.0001,
      "loss": 1.7007,
      "step": 271
    },
    {
      "epoch": 0.04367373153500321,
      "grad_norm": 0.26928162574768066,
      "learning_rate": 0.0001,
      "loss": 1.676,
      "step": 272
    },
    {
      "epoch": 0.04383429672447013,
      "grad_norm": 0.26622453331947327,
      "learning_rate": 0.0001,
      "loss": 1.5913,
      "step": 273
    },
    {
      "epoch": 0.04399486191393706,
      "grad_norm": 0.2929988205432892,
      "learning_rate": 0.0001,
      "loss": 1.5786,
      "step": 274
    },
    {
      "epoch": 0.044155427103403984,
      "grad_norm": 0.2612575590610504,
      "learning_rate": 0.0001,
      "loss": 1.609,
      "step": 275
    },
    {
      "epoch": 0.04431599229287091,
      "grad_norm": 0.27061378955841064,
      "learning_rate": 0.0001,
      "loss": 1.6165,
      "step": 276
    },
    {
      "epoch": 0.04447655748233783,
      "grad_norm": 0.30242618918418884,
      "learning_rate": 0.0001,
      "loss": 1.6786,
      "step": 277
    },
    {
      "epoch": 0.04463712267180475,
      "grad_norm": 0.25789663195610046,
      "learning_rate": 0.0001,
      "loss": 1.5475,
      "step": 278
    },
    {
      "epoch": 0.044797687861271675,
      "grad_norm": 0.2688475251197815,
      "learning_rate": 0.0001,
      "loss": 1.6198,
      "step": 279
    },
    {
      "epoch": 0.0449582530507386,
      "grad_norm": 0.2619808316230774,
      "learning_rate": 0.0001,
      "loss": 1.6332,
      "step": 280
    },
    {
      "epoch": 0.04511881824020552,
      "grad_norm": 0.2763867676258087,
      "learning_rate": 0.0001,
      "loss": 1.7022,
      "step": 281
    },
    {
      "epoch": 0.04527938342967245,
      "grad_norm": 0.29728758335113525,
      "learning_rate": 0.0001,
      "loss": 1.6367,
      "step": 282
    },
    {
      "epoch": 0.04543994861913937,
      "grad_norm": 0.27272191643714905,
      "learning_rate": 0.0001,
      "loss": 1.6244,
      "step": 283
    },
    {
      "epoch": 0.045600513808606295,
      "grad_norm": 0.28270816802978516,
      "learning_rate": 0.0001,
      "loss": 1.6786,
      "step": 284
    },
    {
      "epoch": 0.04576107899807322,
      "grad_norm": 0.25999021530151367,
      "learning_rate": 0.0001,
      "loss": 1.6132,
      "step": 285
    },
    {
      "epoch": 0.04592164418754014,
      "grad_norm": 0.27927297353744507,
      "learning_rate": 0.0001,
      "loss": 1.5973,
      "step": 286
    },
    {
      "epoch": 0.04608220937700706,
      "grad_norm": 0.29601696133613586,
      "learning_rate": 0.0001,
      "loss": 1.6679,
      "step": 287
    },
    {
      "epoch": 0.046242774566473986,
      "grad_norm": 0.31276965141296387,
      "learning_rate": 0.0001,
      "loss": 1.6493,
      "step": 288
    },
    {
      "epoch": 0.04640333975594091,
      "grad_norm": 0.2943185865879059,
      "learning_rate": 0.0001,
      "loss": 1.6838,
      "step": 289
    },
    {
      "epoch": 0.04656390494540784,
      "grad_norm": 0.2697433531284332,
      "learning_rate": 0.0001,
      "loss": 1.5281,
      "step": 290
    },
    {
      "epoch": 0.04672447013487476,
      "grad_norm": 0.28271886706352234,
      "learning_rate": 0.0001,
      "loss": 1.6035,
      "step": 291
    },
    {
      "epoch": 0.046885035324341684,
      "grad_norm": 0.2874354124069214,
      "learning_rate": 0.0001,
      "loss": 1.6086,
      "step": 292
    },
    {
      "epoch": 0.04704560051380861,
      "grad_norm": 0.29864948987960815,
      "learning_rate": 0.0001,
      "loss": 1.6596,
      "step": 293
    },
    {
      "epoch": 0.04720616570327553,
      "grad_norm": 0.3923736810684204,
      "learning_rate": 0.0001,
      "loss": 1.6065,
      "step": 294
    },
    {
      "epoch": 0.04736673089274245,
      "grad_norm": 0.2731698453426361,
      "learning_rate": 0.0001,
      "loss": 1.5518,
      "step": 295
    },
    {
      "epoch": 0.047527296082209375,
      "grad_norm": 0.2886182367801666,
      "learning_rate": 0.0001,
      "loss": 1.7004,
      "step": 296
    },
    {
      "epoch": 0.0476878612716763,
      "grad_norm": 0.26350656151771545,
      "learning_rate": 0.0001,
      "loss": 1.5923,
      "step": 297
    },
    {
      "epoch": 0.04784842646114323,
      "grad_norm": 0.263208270072937,
      "learning_rate": 0.0001,
      "loss": 1.6354,
      "step": 298
    },
    {
      "epoch": 0.04800899165061015,
      "grad_norm": 0.2772878408432007,
      "learning_rate": 0.0001,
      "loss": 1.7047,
      "step": 299
    },
    {
      "epoch": 0.04816955684007707,
      "grad_norm": 0.2817806899547577,
      "learning_rate": 0.0001,
      "loss": 1.7032,
      "step": 300
    },
    {
      "epoch": 0.048330122029543995,
      "grad_norm": 0.26130443811416626,
      "learning_rate": 0.0001,
      "loss": 1.5879,
      "step": 301
    },
    {
      "epoch": 0.04849068721901092,
      "grad_norm": 0.28344425559043884,
      "learning_rate": 0.0001,
      "loss": 1.6826,
      "step": 302
    },
    {
      "epoch": 0.04865125240847784,
      "grad_norm": 0.2728808522224426,
      "learning_rate": 0.0001,
      "loss": 1.6019,
      "step": 303
    },
    {
      "epoch": 0.048811817597944764,
      "grad_norm": 0.2807309031486511,
      "learning_rate": 0.0001,
      "loss": 1.5904,
      "step": 304
    },
    {
      "epoch": 0.048972382787411686,
      "grad_norm": 0.2673819959163666,
      "learning_rate": 0.0001,
      "loss": 1.6582,
      "step": 305
    },
    {
      "epoch": 0.049132947976878616,
      "grad_norm": 0.307579904794693,
      "learning_rate": 0.0001,
      "loss": 1.5884,
      "step": 306
    },
    {
      "epoch": 0.04929351316634554,
      "grad_norm": 0.2933973968029022,
      "learning_rate": 0.0001,
      "loss": 1.701,
      "step": 307
    },
    {
      "epoch": 0.04945407835581246,
      "grad_norm": 0.2675729990005493,
      "learning_rate": 0.0001,
      "loss": 1.6637,
      "step": 308
    },
    {
      "epoch": 0.049614643545279384,
      "grad_norm": 0.2698867917060852,
      "learning_rate": 0.0001,
      "loss": 1.6131,
      "step": 309
    },
    {
      "epoch": 0.04977520873474631,
      "grad_norm": 0.28606680035591125,
      "learning_rate": 0.0001,
      "loss": 1.6189,
      "step": 310
    },
    {
      "epoch": 0.04993577392421323,
      "grad_norm": 0.28758710622787476,
      "learning_rate": 0.0001,
      "loss": 1.7053,
      "step": 311
    },
    {
      "epoch": 0.05009633911368015,
      "grad_norm": 0.29196128249168396,
      "learning_rate": 0.0001,
      "loss": 1.6379,
      "step": 312
    },
    {
      "epoch": 0.050256904303147075,
      "grad_norm": 0.2753434479236603,
      "learning_rate": 0.0001,
      "loss": 1.6491,
      "step": 313
    },
    {
      "epoch": 0.050417469492614005,
      "grad_norm": 0.2925824224948883,
      "learning_rate": 0.0001,
      "loss": 1.6025,
      "step": 314
    },
    {
      "epoch": 0.05057803468208093,
      "grad_norm": 0.2825222313404083,
      "learning_rate": 0.0001,
      "loss": 1.6813,
      "step": 315
    },
    {
      "epoch": 0.05073859987154785,
      "grad_norm": 0.26458314061164856,
      "learning_rate": 0.0001,
      "loss": 1.6451,
      "step": 316
    },
    {
      "epoch": 0.05089916506101477,
      "grad_norm": 0.2626109719276428,
      "learning_rate": 0.0001,
      "loss": 1.6472,
      "step": 317
    },
    {
      "epoch": 0.051059730250481696,
      "grad_norm": 0.2748810052871704,
      "learning_rate": 0.0001,
      "loss": 1.6277,
      "step": 318
    },
    {
      "epoch": 0.05122029543994862,
      "grad_norm": 0.26177430152893066,
      "learning_rate": 0.0001,
      "loss": 1.6687,
      "step": 319
    },
    {
      "epoch": 0.05138086062941554,
      "grad_norm": 0.27573367953300476,
      "learning_rate": 0.0001,
      "loss": 1.6763,
      "step": 320
    },
    {
      "epoch": 0.051541425818882464,
      "grad_norm": 0.2771402895450592,
      "learning_rate": 0.0001,
      "loss": 1.6566,
      "step": 321
    },
    {
      "epoch": 0.05170199100834939,
      "grad_norm": 0.26614296436309814,
      "learning_rate": 0.0001,
      "loss": 1.6022,
      "step": 322
    },
    {
      "epoch": 0.051862556197816316,
      "grad_norm": 0.2790323495864868,
      "learning_rate": 0.0001,
      "loss": 1.5592,
      "step": 323
    },
    {
      "epoch": 0.05202312138728324,
      "grad_norm": 0.2867019474506378,
      "learning_rate": 0.0001,
      "loss": 1.5999,
      "step": 324
    },
    {
      "epoch": 0.05218368657675016,
      "grad_norm": 0.2921207845211029,
      "learning_rate": 0.0001,
      "loss": 1.598,
      "step": 325
    },
    {
      "epoch": 0.052344251766217084,
      "grad_norm": 0.2791954278945923,
      "learning_rate": 0.0001,
      "loss": 1.5413,
      "step": 326
    },
    {
      "epoch": 0.05250481695568401,
      "grad_norm": 0.29377323389053345,
      "learning_rate": 0.0001,
      "loss": 1.6703,
      "step": 327
    },
    {
      "epoch": 0.05266538214515093,
      "grad_norm": 0.2944749593734741,
      "learning_rate": 0.0001,
      "loss": 1.6845,
      "step": 328
    },
    {
      "epoch": 0.05282594733461785,
      "grad_norm": 0.2755807042121887,
      "learning_rate": 0.0001,
      "loss": 1.6008,
      "step": 329
    },
    {
      "epoch": 0.052986512524084775,
      "grad_norm": 0.26535558700561523,
      "learning_rate": 0.0001,
      "loss": 1.6392,
      "step": 330
    },
    {
      "epoch": 0.053147077713551705,
      "grad_norm": 0.2698773443698883,
      "learning_rate": 0.0001,
      "loss": 1.6352,
      "step": 331
    },
    {
      "epoch": 0.05330764290301863,
      "grad_norm": 0.2726115584373474,
      "learning_rate": 0.0001,
      "loss": 1.6676,
      "step": 332
    },
    {
      "epoch": 0.05346820809248555,
      "grad_norm": 1.7487785816192627,
      "learning_rate": 0.0001,
      "loss": 1.5882,
      "step": 333
    },
    {
      "epoch": 0.05362877328195247,
      "grad_norm": 0.29556626081466675,
      "learning_rate": 0.0001,
      "loss": 1.5875,
      "step": 334
    },
    {
      "epoch": 0.053789338471419396,
      "grad_norm": 0.28750649094581604,
      "learning_rate": 0.0001,
      "loss": 1.6948,
      "step": 335
    },
    {
      "epoch": 0.05394990366088632,
      "grad_norm": 0.26631468534469604,
      "learning_rate": 0.0001,
      "loss": 1.6341,
      "step": 336
    },
    {
      "epoch": 0.05411046885035324,
      "grad_norm": 0.262188583612442,
      "learning_rate": 0.0001,
      "loss": 1.6069,
      "step": 337
    },
    {
      "epoch": 0.054271034039820164,
      "grad_norm": 0.3178561329841614,
      "learning_rate": 0.0001,
      "loss": 1.6833,
      "step": 338
    },
    {
      "epoch": 0.05443159922928709,
      "grad_norm": 0.28462502360343933,
      "learning_rate": 0.0001,
      "loss": 1.6664,
      "step": 339
    },
    {
      "epoch": 0.054592164418754016,
      "grad_norm": 0.2982574701309204,
      "learning_rate": 0.0001,
      "loss": 1.5992,
      "step": 340
    },
    {
      "epoch": 0.05475272960822094,
      "grad_norm": 0.26741865277290344,
      "learning_rate": 0.0001,
      "loss": 1.5997,
      "step": 341
    },
    {
      "epoch": 0.05491329479768786,
      "grad_norm": 0.28718245029449463,
      "learning_rate": 0.0001,
      "loss": 1.5932,
      "step": 342
    },
    {
      "epoch": 0.055073859987154784,
      "grad_norm": 0.28356030583381653,
      "learning_rate": 0.0001,
      "loss": 1.6303,
      "step": 343
    },
    {
      "epoch": 0.05523442517662171,
      "grad_norm": 0.2817205786705017,
      "learning_rate": 0.0001,
      "loss": 1.7115,
      "step": 344
    },
    {
      "epoch": 0.05539499036608863,
      "grad_norm": 0.26058703660964966,
      "learning_rate": 0.0001,
      "loss": 1.5112,
      "step": 345
    },
    {
      "epoch": 0.05555555555555555,
      "grad_norm": 0.2597590386867523,
      "learning_rate": 0.0001,
      "loss": 1.6047,
      "step": 346
    },
    {
      "epoch": 0.05571612074502248,
      "grad_norm": 0.3029089868068695,
      "learning_rate": 0.0001,
      "loss": 1.6968,
      "step": 347
    },
    {
      "epoch": 0.055876685934489405,
      "grad_norm": 0.27384987473487854,
      "learning_rate": 0.0001,
      "loss": 1.5922,
      "step": 348
    },
    {
      "epoch": 0.05603725112395633,
      "grad_norm": 0.2676237225532532,
      "learning_rate": 0.0001,
      "loss": 1.6603,
      "step": 349
    },
    {
      "epoch": 0.05619781631342325,
      "grad_norm": 0.25814104080200195,
      "learning_rate": 0.0001,
      "loss": 1.6006,
      "step": 350
    },
    {
      "epoch": 0.05635838150289017,
      "grad_norm": 0.27472570538520813,
      "learning_rate": 0.0001,
      "loss": 1.6377,
      "step": 351
    },
    {
      "epoch": 0.056518946692357096,
      "grad_norm": 0.2804344892501831,
      "learning_rate": 0.0001,
      "loss": 1.6267,
      "step": 352
    },
    {
      "epoch": 0.05667951188182402,
      "grad_norm": 0.29046690464019775,
      "learning_rate": 0.0001,
      "loss": 1.6689,
      "step": 353
    },
    {
      "epoch": 0.05684007707129094,
      "grad_norm": 0.3239709436893463,
      "learning_rate": 0.0001,
      "loss": 1.6366,
      "step": 354
    },
    {
      "epoch": 0.05700064226075787,
      "grad_norm": 0.3020557463169098,
      "learning_rate": 0.0001,
      "loss": 1.5986,
      "step": 355
    },
    {
      "epoch": 0.057161207450224794,
      "grad_norm": 0.27851709723472595,
      "learning_rate": 0.0001,
      "loss": 1.587,
      "step": 356
    },
    {
      "epoch": 0.057321772639691716,
      "grad_norm": 0.2869919240474701,
      "learning_rate": 0.0001,
      "loss": 1.6423,
      "step": 357
    },
    {
      "epoch": 0.05748233782915864,
      "grad_norm": 0.3175436854362488,
      "learning_rate": 0.0001,
      "loss": 1.6393,
      "step": 358
    },
    {
      "epoch": 0.05764290301862556,
      "grad_norm": 0.2586315870285034,
      "learning_rate": 0.0001,
      "loss": 1.5907,
      "step": 359
    },
    {
      "epoch": 0.057803468208092484,
      "grad_norm": 0.28906723856925964,
      "learning_rate": 0.0001,
      "loss": 1.6697,
      "step": 360
    },
    {
      "epoch": 0.05796403339755941,
      "grad_norm": 0.27379968762397766,
      "learning_rate": 0.0001,
      "loss": 1.7089,
      "step": 361
    },
    {
      "epoch": 0.05812459858702633,
      "grad_norm": 0.2720244526863098,
      "learning_rate": 0.0001,
      "loss": 1.5185,
      "step": 362
    },
    {
      "epoch": 0.05828516377649326,
      "grad_norm": 0.26641982793807983,
      "learning_rate": 0.0001,
      "loss": 1.6971,
      "step": 363
    },
    {
      "epoch": 0.05844572896596018,
      "grad_norm": 0.29224488139152527,
      "learning_rate": 0.0001,
      "loss": 1.6498,
      "step": 364
    },
    {
      "epoch": 0.058606294155427105,
      "grad_norm": 0.27155160903930664,
      "learning_rate": 0.0001,
      "loss": 1.5972,
      "step": 365
    },
    {
      "epoch": 0.05876685934489403,
      "grad_norm": 0.2602423429489136,
      "learning_rate": 0.0001,
      "loss": 1.6307,
      "step": 366
    },
    {
      "epoch": 0.05892742453436095,
      "grad_norm": 0.24640707671642303,
      "learning_rate": 0.0001,
      "loss": 1.6352,
      "step": 367
    },
    {
      "epoch": 0.05908798972382787,
      "grad_norm": 0.28325408697128296,
      "learning_rate": 0.0001,
      "loss": 1.7033,
      "step": 368
    },
    {
      "epoch": 0.059248554913294796,
      "grad_norm": 0.2639716863632202,
      "learning_rate": 0.0001,
      "loss": 1.5925,
      "step": 369
    },
    {
      "epoch": 0.05940912010276172,
      "grad_norm": 0.287239134311676,
      "learning_rate": 0.0001,
      "loss": 1.6584,
      "step": 370
    },
    {
      "epoch": 0.05956968529222865,
      "grad_norm": 0.2728608548641205,
      "learning_rate": 0.0001,
      "loss": 1.6749,
      "step": 371
    },
    {
      "epoch": 0.05973025048169557,
      "grad_norm": 0.2690257132053375,
      "learning_rate": 0.0001,
      "loss": 1.688,
      "step": 372
    },
    {
      "epoch": 0.059890815671162494,
      "grad_norm": 0.26607444882392883,
      "learning_rate": 0.0001,
      "loss": 1.6221,
      "step": 373
    },
    {
      "epoch": 0.060051380860629416,
      "grad_norm": 0.25512611865997314,
      "learning_rate": 0.0001,
      "loss": 1.6003,
      "step": 374
    },
    {
      "epoch": 0.06021194605009634,
      "grad_norm": 0.2618836760520935,
      "learning_rate": 0.0001,
      "loss": 1.6317,
      "step": 375
    },
    {
      "epoch": 0.06037251123956326,
      "grad_norm": 0.25583288073539734,
      "learning_rate": 0.0001,
      "loss": 1.6169,
      "step": 376
    },
    {
      "epoch": 0.060533076429030185,
      "grad_norm": 0.278244286775589,
      "learning_rate": 0.0001,
      "loss": 1.6447,
      "step": 377
    },
    {
      "epoch": 0.06069364161849711,
      "grad_norm": 0.2562451660633087,
      "learning_rate": 0.0001,
      "loss": 1.6416,
      "step": 378
    },
    {
      "epoch": 0.06085420680796403,
      "grad_norm": 0.3050384223461151,
      "learning_rate": 0.0001,
      "loss": 1.6582,
      "step": 379
    },
    {
      "epoch": 0.06101477199743096,
      "grad_norm": 0.30344587564468384,
      "learning_rate": 0.0001,
      "loss": 1.5562,
      "step": 380
    },
    {
      "epoch": 0.06117533718689788,
      "grad_norm": 0.2712695002555847,
      "learning_rate": 0.0001,
      "loss": 1.59,
      "step": 381
    },
    {
      "epoch": 0.061335902376364805,
      "grad_norm": 0.27521881461143494,
      "learning_rate": 0.0001,
      "loss": 1.6101,
      "step": 382
    },
    {
      "epoch": 0.06149646756583173,
      "grad_norm": 0.26694977283477783,
      "learning_rate": 0.0001,
      "loss": 1.6551,
      "step": 383
    },
    {
      "epoch": 0.06165703275529865,
      "grad_norm": 0.3174509108066559,
      "learning_rate": 0.0001,
      "loss": 1.6294,
      "step": 384
    },
    {
      "epoch": 0.06181759794476557,
      "grad_norm": 0.26048147678375244,
      "learning_rate": 0.0001,
      "loss": 1.6333,
      "step": 385
    },
    {
      "epoch": 0.061978163134232496,
      "grad_norm": 0.2684613764286041,
      "learning_rate": 0.0001,
      "loss": 1.4829,
      "step": 386
    },
    {
      "epoch": 0.06213872832369942,
      "grad_norm": 0.2893260717391968,
      "learning_rate": 0.0001,
      "loss": 1.7229,
      "step": 387
    },
    {
      "epoch": 0.06229929351316635,
      "grad_norm": 0.2736869156360626,
      "learning_rate": 0.0001,
      "loss": 1.7301,
      "step": 388
    },
    {
      "epoch": 0.06245985870263327,
      "grad_norm": 0.2580438256263733,
      "learning_rate": 0.0001,
      "loss": 1.672,
      "step": 389
    },
    {
      "epoch": 0.0626204238921002,
      "grad_norm": 0.2694301903247833,
      "learning_rate": 0.0001,
      "loss": 1.6613,
      "step": 390
    },
    {
      "epoch": 0.06278098908156711,
      "grad_norm": 0.28036239743232727,
      "learning_rate": 0.0001,
      "loss": 1.639,
      "step": 391
    },
    {
      "epoch": 0.06294155427103404,
      "grad_norm": 0.273385226726532,
      "learning_rate": 0.0001,
      "loss": 1.616,
      "step": 392
    },
    {
      "epoch": 0.06310211946050097,
      "grad_norm": 0.27833032608032227,
      "learning_rate": 0.0001,
      "loss": 1.6453,
      "step": 393
    },
    {
      "epoch": 0.06326268464996788,
      "grad_norm": 0.27655139565467834,
      "learning_rate": 0.0001,
      "loss": 1.6777,
      "step": 394
    },
    {
      "epoch": 0.06342324983943481,
      "grad_norm": 0.24965526163578033,
      "learning_rate": 0.0001,
      "loss": 1.5135,
      "step": 395
    },
    {
      "epoch": 0.06358381502890173,
      "grad_norm": 0.28703129291534424,
      "learning_rate": 0.0001,
      "loss": 1.6204,
      "step": 396
    },
    {
      "epoch": 0.06374438021836866,
      "grad_norm": 0.2821877598762512,
      "learning_rate": 0.0001,
      "loss": 1.6637,
      "step": 397
    },
    {
      "epoch": 0.06390494540783558,
      "grad_norm": 0.26222217082977295,
      "learning_rate": 0.0001,
      "loss": 1.5248,
      "step": 398
    },
    {
      "epoch": 0.0640655105973025,
      "grad_norm": 0.2721595764160156,
      "learning_rate": 0.0001,
      "loss": 1.5602,
      "step": 399
    },
    {
      "epoch": 0.06422607578676943,
      "grad_norm": 0.270107626914978,
      "learning_rate": 0.0001,
      "loss": 1.6016,
      "step": 400
    },
    {
      "epoch": 0.06438664097623635,
      "grad_norm": 0.26544928550720215,
      "learning_rate": 0.0001,
      "loss": 1.5488,
      "step": 401
    },
    {
      "epoch": 0.06454720616570328,
      "grad_norm": 0.2908627390861511,
      "learning_rate": 0.0001,
      "loss": 1.6418,
      "step": 402
    },
    {
      "epoch": 0.0647077713551702,
      "grad_norm": 0.2766053080558777,
      "learning_rate": 0.0001,
      "loss": 1.6592,
      "step": 403
    },
    {
      "epoch": 0.06486833654463713,
      "grad_norm": 0.2724528908729553,
      "learning_rate": 0.0001,
      "loss": 1.5596,
      "step": 404
    },
    {
      "epoch": 0.06502890173410404,
      "grad_norm": 0.2631203830242157,
      "learning_rate": 0.0001,
      "loss": 1.5603,
      "step": 405
    },
    {
      "epoch": 0.06518946692357097,
      "grad_norm": 0.2576027512550354,
      "learning_rate": 0.0001,
      "loss": 1.5777,
      "step": 406
    },
    {
      "epoch": 0.06535003211303789,
      "grad_norm": 0.29033124446868896,
      "learning_rate": 0.0001,
      "loss": 1.6131,
      "step": 407
    },
    {
      "epoch": 0.06551059730250482,
      "grad_norm": 0.27214476466178894,
      "learning_rate": 0.0001,
      "loss": 1.6768,
      "step": 408
    },
    {
      "epoch": 0.06567116249197175,
      "grad_norm": 0.2475825399160385,
      "learning_rate": 0.0001,
      "loss": 1.5325,
      "step": 409
    },
    {
      "epoch": 0.06583172768143866,
      "grad_norm": 0.2732248902320862,
      "learning_rate": 0.0001,
      "loss": 1.6699,
      "step": 410
    },
    {
      "epoch": 0.06599229287090559,
      "grad_norm": 0.2896178364753723,
      "learning_rate": 0.0001,
      "loss": 1.7157,
      "step": 411
    },
    {
      "epoch": 0.06615285806037251,
      "grad_norm": 0.27780890464782715,
      "learning_rate": 0.0001,
      "loss": 1.6669,
      "step": 412
    },
    {
      "epoch": 0.06631342324983944,
      "grad_norm": 0.26925742626190186,
      "learning_rate": 0.0001,
      "loss": 1.7434,
      "step": 413
    },
    {
      "epoch": 0.06647398843930635,
      "grad_norm": 0.2978026866912842,
      "learning_rate": 0.0001,
      "loss": 1.6731,
      "step": 414
    },
    {
      "epoch": 0.06663455362877328,
      "grad_norm": 0.26644179224967957,
      "learning_rate": 0.0001,
      "loss": 1.6674,
      "step": 415
    },
    {
      "epoch": 0.06679511881824021,
      "grad_norm": 0.2594255208969116,
      "learning_rate": 0.0001,
      "loss": 1.6407,
      "step": 416
    },
    {
      "epoch": 0.06695568400770713,
      "grad_norm": 0.29270321130752563,
      "learning_rate": 0.0001,
      "loss": 1.6436,
      "step": 417
    },
    {
      "epoch": 0.06711624919717406,
      "grad_norm": 0.2654949426651001,
      "learning_rate": 0.0001,
      "loss": 1.6121,
      "step": 418
    },
    {
      "epoch": 0.06727681438664097,
      "grad_norm": 0.2827872633934021,
      "learning_rate": 0.0001,
      "loss": 1.6427,
      "step": 419
    },
    {
      "epoch": 0.0674373795761079,
      "grad_norm": 0.2690976560115814,
      "learning_rate": 0.0001,
      "loss": 1.569,
      "step": 420
    },
    {
      "epoch": 0.06759794476557482,
      "grad_norm": 0.269963800907135,
      "learning_rate": 0.0001,
      "loss": 1.5509,
      "step": 421
    },
    {
      "epoch": 0.06775850995504175,
      "grad_norm": 0.2637880742549896,
      "learning_rate": 0.0001,
      "loss": 1.6754,
      "step": 422
    },
    {
      "epoch": 0.06791907514450866,
      "grad_norm": 0.28625962138175964,
      "learning_rate": 0.0001,
      "loss": 1.6278,
      "step": 423
    },
    {
      "epoch": 0.0680796403339756,
      "grad_norm": 0.2607976496219635,
      "learning_rate": 0.0001,
      "loss": 1.5234,
      "step": 424
    },
    {
      "epoch": 0.06824020552344252,
      "grad_norm": 0.2630302608013153,
      "learning_rate": 0.0001,
      "loss": 1.6054,
      "step": 425
    },
    {
      "epoch": 0.06840077071290944,
      "grad_norm": 0.26833876967430115,
      "learning_rate": 0.0001,
      "loss": 1.6376,
      "step": 426
    },
    {
      "epoch": 0.06856133590237637,
      "grad_norm": 0.2762647867202759,
      "learning_rate": 0.0001,
      "loss": 1.6134,
      "step": 427
    },
    {
      "epoch": 0.06872190109184328,
      "grad_norm": 0.27359503507614136,
      "learning_rate": 0.0001,
      "loss": 1.5636,
      "step": 428
    },
    {
      "epoch": 0.06888246628131021,
      "grad_norm": 0.24629630148410797,
      "learning_rate": 0.0001,
      "loss": 1.5002,
      "step": 429
    },
    {
      "epoch": 0.06904303147077713,
      "grad_norm": 0.2808135151863098,
      "learning_rate": 0.0001,
      "loss": 1.6491,
      "step": 430
    },
    {
      "epoch": 0.06920359666024406,
      "grad_norm": 0.2696678936481476,
      "learning_rate": 0.0001,
      "loss": 1.6919,
      "step": 431
    },
    {
      "epoch": 0.06936416184971098,
      "grad_norm": 0.31605294346809387,
      "learning_rate": 0.0001,
      "loss": 1.6632,
      "step": 432
    },
    {
      "epoch": 0.0695247270391779,
      "grad_norm": 0.2688290476799011,
      "learning_rate": 0.0001,
      "loss": 1.5792,
      "step": 433
    },
    {
      "epoch": 0.06968529222864484,
      "grad_norm": 0.2849433124065399,
      "learning_rate": 0.0001,
      "loss": 1.5376,
      "step": 434
    },
    {
      "epoch": 0.06984585741811175,
      "grad_norm": 0.26482313871383667,
      "learning_rate": 0.0001,
      "loss": 1.6478,
      "step": 435
    },
    {
      "epoch": 0.07000642260757868,
      "grad_norm": 0.2657308876514435,
      "learning_rate": 0.0001,
      "loss": 1.59,
      "step": 436
    },
    {
      "epoch": 0.0701669877970456,
      "grad_norm": 0.28563132882118225,
      "learning_rate": 0.0001,
      "loss": 1.6054,
      "step": 437
    },
    {
      "epoch": 0.07032755298651253,
      "grad_norm": 0.27213236689567566,
      "learning_rate": 0.0001,
      "loss": 1.5857,
      "step": 438
    },
    {
      "epoch": 0.07048811817597944,
      "grad_norm": 0.26817741990089417,
      "learning_rate": 0.0001,
      "loss": 1.5646,
      "step": 439
    },
    {
      "epoch": 0.07064868336544637,
      "grad_norm": 0.28620007634162903,
      "learning_rate": 0.0001,
      "loss": 1.6234,
      "step": 440
    },
    {
      "epoch": 0.0708092485549133,
      "grad_norm": 0.2723342180252075,
      "learning_rate": 0.0001,
      "loss": 1.6453,
      "step": 441
    },
    {
      "epoch": 0.07096981374438022,
      "grad_norm": 0.30084824562072754,
      "learning_rate": 0.0001,
      "loss": 1.5685,
      "step": 442
    },
    {
      "epoch": 0.07113037893384715,
      "grad_norm": 0.2717418968677521,
      "learning_rate": 0.0001,
      "loss": 1.5707,
      "step": 443
    },
    {
      "epoch": 0.07129094412331406,
      "grad_norm": 0.2876720130443573,
      "learning_rate": 0.0001,
      "loss": 1.5972,
      "step": 444
    },
    {
      "epoch": 0.07145150931278099,
      "grad_norm": 0.2772391140460968,
      "learning_rate": 0.0001,
      "loss": 1.5141,
      "step": 445
    },
    {
      "epoch": 0.07161207450224791,
      "grad_norm": 0.3008630871772766,
      "learning_rate": 0.0001,
      "loss": 1.6756,
      "step": 446
    },
    {
      "epoch": 0.07177263969171484,
      "grad_norm": 0.29428958892822266,
      "learning_rate": 0.0001,
      "loss": 1.7047,
      "step": 447
    },
    {
      "epoch": 0.07193320488118175,
      "grad_norm": 0.28210246562957764,
      "learning_rate": 0.0001,
      "loss": 1.5358,
      "step": 448
    },
    {
      "epoch": 0.07209377007064868,
      "grad_norm": 0.26412245631217957,
      "learning_rate": 0.0001,
      "loss": 1.6833,
      "step": 449
    },
    {
      "epoch": 0.07225433526011561,
      "grad_norm": 0.322248637676239,
      "learning_rate": 0.0001,
      "loss": 1.6587,
      "step": 450
    },
    {
      "epoch": 0.07241490044958253,
      "grad_norm": 0.2715447247028351,
      "learning_rate": 0.0001,
      "loss": 1.651,
      "step": 451
    },
    {
      "epoch": 0.07257546563904946,
      "grad_norm": 0.28355979919433594,
      "learning_rate": 0.0001,
      "loss": 1.6032,
      "step": 452
    },
    {
      "epoch": 0.07273603082851637,
      "grad_norm": 0.28326234221458435,
      "learning_rate": 0.0001,
      "loss": 1.536,
      "step": 453
    },
    {
      "epoch": 0.0728965960179833,
      "grad_norm": 0.27278533577919006,
      "learning_rate": 0.0001,
      "loss": 1.597,
      "step": 454
    },
    {
      "epoch": 0.07305716120745022,
      "grad_norm": 1.5439703464508057,
      "learning_rate": 0.0001,
      "loss": 1.597,
      "step": 455
    },
    {
      "epoch": 0.07321772639691715,
      "grad_norm": 0.2848699986934662,
      "learning_rate": 0.0001,
      "loss": 1.6438,
      "step": 456
    },
    {
      "epoch": 0.07337829158638408,
      "grad_norm": 0.27211204171180725,
      "learning_rate": 0.0001,
      "loss": 1.6598,
      "step": 457
    },
    {
      "epoch": 0.073538856775851,
      "grad_norm": 0.2694942355155945,
      "learning_rate": 0.0001,
      "loss": 1.5787,
      "step": 458
    },
    {
      "epoch": 0.07369942196531792,
      "grad_norm": 0.24894948303699493,
      "learning_rate": 0.0001,
      "loss": 1.5849,
      "step": 459
    },
    {
      "epoch": 0.07385998715478484,
      "grad_norm": 0.2886010408401489,
      "learning_rate": 0.0001,
      "loss": 1.6779,
      "step": 460
    },
    {
      "epoch": 0.07402055234425177,
      "grad_norm": 0.25323522090911865,
      "learning_rate": 0.0001,
      "loss": 1.613,
      "step": 461
    },
    {
      "epoch": 0.07418111753371869,
      "grad_norm": 0.2417057752609253,
      "learning_rate": 0.0001,
      "loss": 1.5457,
      "step": 462
    },
    {
      "epoch": 0.07434168272318561,
      "grad_norm": 0.25295180082321167,
      "learning_rate": 0.0001,
      "loss": 1.5663,
      "step": 463
    },
    {
      "epoch": 0.07450224791265253,
      "grad_norm": 0.27051475644111633,
      "learning_rate": 0.0001,
      "loss": 1.5979,
      "step": 464
    },
    {
      "epoch": 0.07466281310211946,
      "grad_norm": 0.26878947019577026,
      "learning_rate": 0.0001,
      "loss": 1.6463,
      "step": 465
    },
    {
      "epoch": 0.07482337829158639,
      "grad_norm": 0.25430378317832947,
      "learning_rate": 0.0001,
      "loss": 1.5731,
      "step": 466
    },
    {
      "epoch": 0.0749839434810533,
      "grad_norm": 0.27556899189949036,
      "learning_rate": 0.0001,
      "loss": 1.6599,
      "step": 467
    },
    {
      "epoch": 0.07514450867052024,
      "grad_norm": 0.28547075390815735,
      "learning_rate": 0.0001,
      "loss": 1.6284,
      "step": 468
    },
    {
      "epoch": 0.07530507385998715,
      "grad_norm": 0.2858209013938904,
      "learning_rate": 0.0001,
      "loss": 1.6093,
      "step": 469
    },
    {
      "epoch": 0.07546563904945408,
      "grad_norm": 0.2524562180042267,
      "learning_rate": 0.0001,
      "loss": 1.5948,
      "step": 470
    },
    {
      "epoch": 0.075626204238921,
      "grad_norm": 0.26532092690467834,
      "learning_rate": 0.0001,
      "loss": 1.5673,
      "step": 471
    },
    {
      "epoch": 0.07578676942838793,
      "grad_norm": 0.2942146062850952,
      "learning_rate": 0.0001,
      "loss": 1.6223,
      "step": 472
    },
    {
      "epoch": 0.07594733461785486,
      "grad_norm": 0.26330721378326416,
      "learning_rate": 0.0001,
      "loss": 1.6887,
      "step": 473
    },
    {
      "epoch": 0.07610789980732177,
      "grad_norm": 0.29492637515068054,
      "learning_rate": 0.0001,
      "loss": 1.6618,
      "step": 474
    },
    {
      "epoch": 0.0762684649967887,
      "grad_norm": 0.28351297974586487,
      "learning_rate": 0.0001,
      "loss": 1.5956,
      "step": 475
    },
    {
      "epoch": 0.07642903018625562,
      "grad_norm": 0.2659725546836853,
      "learning_rate": 0.0001,
      "loss": 1.562,
      "step": 476
    },
    {
      "epoch": 0.07658959537572255,
      "grad_norm": 0.2990523874759674,
      "learning_rate": 0.0001,
      "loss": 1.6288,
      "step": 477
    },
    {
      "epoch": 0.07675016056518946,
      "grad_norm": 0.26317718625068665,
      "learning_rate": 0.0001,
      "loss": 1.6535,
      "step": 478
    },
    {
      "epoch": 0.07691072575465639,
      "grad_norm": 0.3027506470680237,
      "learning_rate": 0.0001,
      "loss": 1.6961,
      "step": 479
    },
    {
      "epoch": 0.07707129094412331,
      "grad_norm": 0.24868576228618622,
      "learning_rate": 0.0001,
      "loss": 1.5759,
      "step": 480
    },
    {
      "epoch": 0.07723185613359024,
      "grad_norm": 0.27065804600715637,
      "learning_rate": 0.0001,
      "loss": 1.6023,
      "step": 481
    },
    {
      "epoch": 0.07739242132305717,
      "grad_norm": 0.30835163593292236,
      "learning_rate": 0.0001,
      "loss": 1.6865,
      "step": 482
    },
    {
      "epoch": 0.07755298651252408,
      "grad_norm": 0.265004962682724,
      "learning_rate": 0.0001,
      "loss": 1.6617,
      "step": 483
    },
    {
      "epoch": 0.07771355170199101,
      "grad_norm": 0.2614341974258423,
      "learning_rate": 0.0001,
      "loss": 1.6596,
      "step": 484
    },
    {
      "epoch": 0.07787411689145793,
      "grad_norm": 0.25203412771224976,
      "learning_rate": 0.0001,
      "loss": 1.5894,
      "step": 485
    },
    {
      "epoch": 0.07803468208092486,
      "grad_norm": 0.28151461482048035,
      "learning_rate": 0.0001,
      "loss": 1.703,
      "step": 486
    },
    {
      "epoch": 0.07819524727039177,
      "grad_norm": 0.2849961817264557,
      "learning_rate": 0.0001,
      "loss": 1.633,
      "step": 487
    },
    {
      "epoch": 0.0783558124598587,
      "grad_norm": 0.25807830691337585,
      "learning_rate": 0.0001,
      "loss": 1.5597,
      "step": 488
    },
    {
      "epoch": 0.07851637764932562,
      "grad_norm": 0.26951682567596436,
      "learning_rate": 0.0001,
      "loss": 1.6252,
      "step": 489
    },
    {
      "epoch": 0.07867694283879255,
      "grad_norm": 0.27103808522224426,
      "learning_rate": 0.0001,
      "loss": 1.652,
      "step": 490
    },
    {
      "epoch": 0.07883750802825948,
      "grad_norm": 0.2701008915901184,
      "learning_rate": 0.0001,
      "loss": 1.6301,
      "step": 491
    },
    {
      "epoch": 0.0789980732177264,
      "grad_norm": 0.2753893733024597,
      "learning_rate": 0.0001,
      "loss": 1.564,
      "step": 492
    },
    {
      "epoch": 0.07915863840719332,
      "grad_norm": 0.27593153715133667,
      "learning_rate": 0.0001,
      "loss": 1.7073,
      "step": 493
    },
    {
      "epoch": 0.07931920359666024,
      "grad_norm": 0.25339603424072266,
      "learning_rate": 0.0001,
      "loss": 1.5699,
      "step": 494
    },
    {
      "epoch": 0.07947976878612717,
      "grad_norm": 0.2900751233100891,
      "learning_rate": 0.0001,
      "loss": 1.6138,
      "step": 495
    },
    {
      "epoch": 0.07964033397559409,
      "grad_norm": 0.2772659361362457,
      "learning_rate": 0.0001,
      "loss": 1.6653,
      "step": 496
    },
    {
      "epoch": 0.07980089916506101,
      "grad_norm": 0.28388726711273193,
      "learning_rate": 0.0001,
      "loss": 1.6259,
      "step": 497
    },
    {
      "epoch": 0.07996146435452794,
      "grad_norm": 0.2737922966480255,
      "learning_rate": 0.0001,
      "loss": 1.594,
      "step": 498
    },
    {
      "epoch": 0.08012202954399486,
      "grad_norm": 0.27199098467826843,
      "learning_rate": 0.0001,
      "loss": 1.5972,
      "step": 499
    },
    {
      "epoch": 0.08028259473346179,
      "grad_norm": 0.29927489161491394,
      "learning_rate": 0.0001,
      "loss": 1.5758,
      "step": 500
    },
    {
      "epoch": 0.0804431599229287,
      "grad_norm": 60.146236419677734,
      "learning_rate": 0.0001,
      "loss": 2.0838,
      "step": 501
    },
    {
      "epoch": 0.08060372511239564,
      "grad_norm": 0.28392574191093445,
      "learning_rate": 0.0001,
      "loss": 1.6233,
      "step": 502
    },
    {
      "epoch": 0.08076429030186255,
      "grad_norm": 0.2903461456298828,
      "learning_rate": 0.0001,
      "loss": 1.7171,
      "step": 503
    },
    {
      "epoch": 0.08092485549132948,
      "grad_norm": 0.263322651386261,
      "learning_rate": 0.0001,
      "loss": 1.5696,
      "step": 504
    },
    {
      "epoch": 0.0810854206807964,
      "grad_norm": 0.2876776158809662,
      "learning_rate": 0.0001,
      "loss": 1.6347,
      "step": 505
    },
    {
      "epoch": 0.08124598587026333,
      "grad_norm": 0.28394991159439087,
      "learning_rate": 0.0001,
      "loss": 1.6217,
      "step": 506
    },
    {
      "epoch": 0.08140655105973026,
      "grad_norm": 0.26244935393333435,
      "learning_rate": 0.0001,
      "loss": 1.5834,
      "step": 507
    },
    {
      "epoch": 0.08156711624919717,
      "grad_norm": 0.25742456316947937,
      "learning_rate": 0.0001,
      "loss": 1.5667,
      "step": 508
    },
    {
      "epoch": 0.0817276814386641,
      "grad_norm": 0.26109588146209717,
      "learning_rate": 0.0001,
      "loss": 1.6388,
      "step": 509
    },
    {
      "epoch": 0.08188824662813102,
      "grad_norm": 0.2430950254201889,
      "learning_rate": 0.0001,
      "loss": 1.5202,
      "step": 510
    },
    {
      "epoch": 0.08204881181759795,
      "grad_norm": 0.2901354730129242,
      "learning_rate": 0.0001,
      "loss": 1.6619,
      "step": 511
    },
    {
      "epoch": 0.08220937700706486,
      "grad_norm": 0.2899554669857025,
      "learning_rate": 0.0001,
      "loss": 1.5867,
      "step": 512
    },
    {
      "epoch": 0.08236994219653179,
      "grad_norm": 0.2571408748626709,
      "learning_rate": 0.0001,
      "loss": 1.619,
      "step": 513
    },
    {
      "epoch": 0.08253050738599872,
      "grad_norm": 0.28769806027412415,
      "learning_rate": 0.0001,
      "loss": 1.6135,
      "step": 514
    },
    {
      "epoch": 0.08269107257546564,
      "grad_norm": 0.2652043104171753,
      "learning_rate": 0.0001,
      "loss": 1.559,
      "step": 515
    },
    {
      "epoch": 0.08285163776493257,
      "grad_norm": 0.2747846245765686,
      "learning_rate": 0.0001,
      "loss": 1.5978,
      "step": 516
    },
    {
      "epoch": 0.08301220295439948,
      "grad_norm": 0.2706848680973053,
      "learning_rate": 0.0001,
      "loss": 1.6333,
      "step": 517
    },
    {
      "epoch": 0.08317276814386641,
      "grad_norm": 0.28645724058151245,
      "learning_rate": 0.0001,
      "loss": 1.6845,
      "step": 518
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.2895065248012543,
      "learning_rate": 0.0001,
      "loss": 1.6928,
      "step": 519
    },
    {
      "epoch": 0.08349389852280026,
      "grad_norm": 0.2539268434047699,
      "learning_rate": 0.0001,
      "loss": 1.5785,
      "step": 520
    },
    {
      "epoch": 0.08365446371226717,
      "grad_norm": 0.2594113051891327,
      "learning_rate": 0.0001,
      "loss": 1.6307,
      "step": 521
    },
    {
      "epoch": 0.0838150289017341,
      "grad_norm": 0.26727333664894104,
      "learning_rate": 0.0001,
      "loss": 1.5499,
      "step": 522
    },
    {
      "epoch": 0.08397559409120103,
      "grad_norm": 0.258354514837265,
      "learning_rate": 0.0001,
      "loss": 1.5713,
      "step": 523
    },
    {
      "epoch": 0.08413615928066795,
      "grad_norm": 0.2712153196334839,
      "learning_rate": 0.0001,
      "loss": 1.6509,
      "step": 524
    },
    {
      "epoch": 0.08429672447013488,
      "grad_norm": 0.2564713656902313,
      "learning_rate": 0.0001,
      "loss": 1.6349,
      "step": 525
    },
    {
      "epoch": 0.0844572896596018,
      "grad_norm": 0.26457738876342773,
      "learning_rate": 0.0001,
      "loss": 1.6169,
      "step": 526
    },
    {
      "epoch": 0.08461785484906872,
      "grad_norm": 0.25640881061553955,
      "learning_rate": 0.0001,
      "loss": 1.5795,
      "step": 527
    },
    {
      "epoch": 0.08477842003853564,
      "grad_norm": 0.25883811712265015,
      "learning_rate": 0.0001,
      "loss": 1.6041,
      "step": 528
    },
    {
      "epoch": 0.08493898522800257,
      "grad_norm": 0.26522043347358704,
      "learning_rate": 0.0001,
      "loss": 1.6413,
      "step": 529
    },
    {
      "epoch": 0.0850995504174695,
      "grad_norm": 0.27930450439453125,
      "learning_rate": 0.0001,
      "loss": 1.6209,
      "step": 530
    },
    {
      "epoch": 0.08526011560693642,
      "grad_norm": 0.2752113342285156,
      "learning_rate": 0.0001,
      "loss": 1.5664,
      "step": 531
    },
    {
      "epoch": 0.08542068079640334,
      "grad_norm": 0.2911113202571869,
      "learning_rate": 0.0001,
      "loss": 1.6011,
      "step": 532
    },
    {
      "epoch": 0.08558124598587026,
      "grad_norm": 0.2742812931537628,
      "learning_rate": 0.0001,
      "loss": 1.7042,
      "step": 533
    },
    {
      "epoch": 0.08574181117533719,
      "grad_norm": 0.3127116560935974,
      "learning_rate": 0.0001,
      "loss": 1.5496,
      "step": 534
    },
    {
      "epoch": 0.0859023763648041,
      "grad_norm": 0.3009921908378601,
      "learning_rate": 0.0001,
      "loss": 1.6395,
      "step": 535
    },
    {
      "epoch": 0.08606294155427104,
      "grad_norm": 0.2412799447774887,
      "learning_rate": 0.0001,
      "loss": 1.4681,
      "step": 536
    },
    {
      "epoch": 0.08622350674373795,
      "grad_norm": 0.2689095735549927,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 537
    },
    {
      "epoch": 0.08638407193320488,
      "grad_norm": 0.253313273191452,
      "learning_rate": 0.0001,
      "loss": 1.6089,
      "step": 538
    },
    {
      "epoch": 0.08654463712267181,
      "grad_norm": 0.2606997489929199,
      "learning_rate": 0.0001,
      "loss": 1.5802,
      "step": 539
    },
    {
      "epoch": 0.08670520231213873,
      "grad_norm": 0.2673456370830536,
      "learning_rate": 0.0001,
      "loss": 1.5876,
      "step": 540
    },
    {
      "epoch": 0.08686576750160566,
      "grad_norm": 0.28409141302108765,
      "learning_rate": 0.0001,
      "loss": 1.6098,
      "step": 541
    },
    {
      "epoch": 0.08702633269107257,
      "grad_norm": 0.2831609845161438,
      "learning_rate": 0.0001,
      "loss": 1.6459,
      "step": 542
    },
    {
      "epoch": 0.0871868978805395,
      "grad_norm": 0.27968084812164307,
      "learning_rate": 0.0001,
      "loss": 1.6794,
      "step": 543
    },
    {
      "epoch": 0.08734746307000642,
      "grad_norm": 0.3048776388168335,
      "learning_rate": 0.0001,
      "loss": 1.7141,
      "step": 544
    },
    {
      "epoch": 0.08750802825947335,
      "grad_norm": 0.24242153763771057,
      "learning_rate": 0.0001,
      "loss": 1.5411,
      "step": 545
    },
    {
      "epoch": 0.08766859344894026,
      "grad_norm": 0.2752143442630768,
      "learning_rate": 0.0001,
      "loss": 1.5587,
      "step": 546
    },
    {
      "epoch": 0.08782915863840719,
      "grad_norm": 0.25544866919517517,
      "learning_rate": 0.0001,
      "loss": 1.5802,
      "step": 547
    },
    {
      "epoch": 0.08798972382787412,
      "grad_norm": 0.2552921175956726,
      "learning_rate": 0.0001,
      "loss": 1.5834,
      "step": 548
    },
    {
      "epoch": 0.08815028901734104,
      "grad_norm": 0.28750985860824585,
      "learning_rate": 0.0001,
      "loss": 1.6387,
      "step": 549
    },
    {
      "epoch": 0.08831085420680797,
      "grad_norm": 0.26201680302619934,
      "learning_rate": 0.0001,
      "loss": 1.677,
      "step": 550
    },
    {
      "epoch": 0.08847141939627488,
      "grad_norm": 0.2960847020149231,
      "learning_rate": 0.0001,
      "loss": 1.6203,
      "step": 551
    },
    {
      "epoch": 0.08863198458574181,
      "grad_norm": 0.27084705233573914,
      "learning_rate": 0.0001,
      "loss": 1.6122,
      "step": 552
    },
    {
      "epoch": 0.08879254977520873,
      "grad_norm": 0.2630038261413574,
      "learning_rate": 0.0001,
      "loss": 1.6549,
      "step": 553
    },
    {
      "epoch": 0.08895311496467566,
      "grad_norm": 0.26962539553642273,
      "learning_rate": 0.0001,
      "loss": 1.6445,
      "step": 554
    },
    {
      "epoch": 0.08911368015414259,
      "grad_norm": 0.2721734642982483,
      "learning_rate": 0.0001,
      "loss": 1.6049,
      "step": 555
    },
    {
      "epoch": 0.0892742453436095,
      "grad_norm": 0.25505608320236206,
      "learning_rate": 0.0001,
      "loss": 1.5966,
      "step": 556
    },
    {
      "epoch": 0.08943481053307643,
      "grad_norm": 0.2604336142539978,
      "learning_rate": 0.0001,
      "loss": 1.6185,
      "step": 557
    },
    {
      "epoch": 0.08959537572254335,
      "grad_norm": 0.2705906629562378,
      "learning_rate": 0.0001,
      "loss": 1.5978,
      "step": 558
    },
    {
      "epoch": 0.08975594091201028,
      "grad_norm": 0.26772817969322205,
      "learning_rate": 0.0001,
      "loss": 1.5895,
      "step": 559
    },
    {
      "epoch": 0.0899165061014772,
      "grad_norm": 0.28454136848449707,
      "learning_rate": 0.0001,
      "loss": 1.6466,
      "step": 560
    },
    {
      "epoch": 0.09007707129094412,
      "grad_norm": 0.27886369824409485,
      "learning_rate": 0.0001,
      "loss": 1.6432,
      "step": 561
    },
    {
      "epoch": 0.09023763648041104,
      "grad_norm": 0.2522517144680023,
      "learning_rate": 0.0001,
      "loss": 1.5409,
      "step": 562
    },
    {
      "epoch": 0.09039820166987797,
      "grad_norm": 0.31142204999923706,
      "learning_rate": 0.0001,
      "loss": 1.6552,
      "step": 563
    },
    {
      "epoch": 0.0905587668593449,
      "grad_norm": 0.26340827345848083,
      "learning_rate": 0.0001,
      "loss": 1.7059,
      "step": 564
    },
    {
      "epoch": 0.09071933204881182,
      "grad_norm": 0.2656070291996002,
      "learning_rate": 0.0001,
      "loss": 1.5082,
      "step": 565
    },
    {
      "epoch": 0.09087989723827875,
      "grad_norm": 0.2772354483604431,
      "learning_rate": 0.0001,
      "loss": 1.5607,
      "step": 566
    },
    {
      "epoch": 0.09104046242774566,
      "grad_norm": 0.2598869502544403,
      "learning_rate": 0.0001,
      "loss": 1.6182,
      "step": 567
    },
    {
      "epoch": 0.09120102761721259,
      "grad_norm": 0.27644407749176025,
      "learning_rate": 0.0001,
      "loss": 1.6045,
      "step": 568
    },
    {
      "epoch": 0.0913615928066795,
      "grad_norm": 2.638380289077759,
      "learning_rate": 0.0001,
      "loss": 1.5733,
      "step": 569
    },
    {
      "epoch": 0.09152215799614644,
      "grad_norm": 0.2717907130718231,
      "learning_rate": 0.0001,
      "loss": 1.546,
      "step": 570
    },
    {
      "epoch": 0.09168272318561337,
      "grad_norm": 0.26251983642578125,
      "learning_rate": 0.0001,
      "loss": 1.6072,
      "step": 571
    },
    {
      "epoch": 0.09184328837508028,
      "grad_norm": 0.2579016089439392,
      "learning_rate": 0.0001,
      "loss": 1.5423,
      "step": 572
    },
    {
      "epoch": 0.09200385356454721,
      "grad_norm": 0.2622748017311096,
      "learning_rate": 0.0001,
      "loss": 1.6253,
      "step": 573
    },
    {
      "epoch": 0.09216441875401413,
      "grad_norm": 0.2696703374385834,
      "learning_rate": 0.0001,
      "loss": 1.6311,
      "step": 574
    },
    {
      "epoch": 0.09232498394348106,
      "grad_norm": 0.2684510052204132,
      "learning_rate": 0.0001,
      "loss": 1.5872,
      "step": 575
    },
    {
      "epoch": 0.09248554913294797,
      "grad_norm": 0.2625136375427246,
      "learning_rate": 0.0001,
      "loss": 1.5734,
      "step": 576
    },
    {
      "epoch": 0.0926461143224149,
      "grad_norm": 0.2556249499320984,
      "learning_rate": 0.0001,
      "loss": 1.5753,
      "step": 577
    },
    {
      "epoch": 0.09280667951188182,
      "grad_norm": 0.28377729654312134,
      "learning_rate": 0.0001,
      "loss": 1.6421,
      "step": 578
    },
    {
      "epoch": 0.09296724470134875,
      "grad_norm": 0.2608239948749542,
      "learning_rate": 0.0001,
      "loss": 1.6478,
      "step": 579
    },
    {
      "epoch": 0.09312780989081568,
      "grad_norm": 0.263975590467453,
      "learning_rate": 0.0001,
      "loss": 1.628,
      "step": 580
    },
    {
      "epoch": 0.09328837508028259,
      "grad_norm": 0.2537274956703186,
      "learning_rate": 0.0001,
      "loss": 1.6145,
      "step": 581
    },
    {
      "epoch": 0.09344894026974952,
      "grad_norm": 0.25347456336021423,
      "learning_rate": 0.0001,
      "loss": 1.5722,
      "step": 582
    },
    {
      "epoch": 0.09360950545921644,
      "grad_norm": 0.3010750710964203,
      "learning_rate": 0.0001,
      "loss": 1.5706,
      "step": 583
    },
    {
      "epoch": 0.09377007064868337,
      "grad_norm": 0.2591670751571655,
      "learning_rate": 0.0001,
      "loss": 1.5847,
      "step": 584
    },
    {
      "epoch": 0.09393063583815028,
      "grad_norm": 0.26924920082092285,
      "learning_rate": 0.0001,
      "loss": 1.577,
      "step": 585
    },
    {
      "epoch": 0.09409120102761721,
      "grad_norm": 0.275130033493042,
      "learning_rate": 0.0001,
      "loss": 1.6624,
      "step": 586
    },
    {
      "epoch": 0.09425176621708414,
      "grad_norm": 0.25509998202323914,
      "learning_rate": 0.0001,
      "loss": 1.6138,
      "step": 587
    },
    {
      "epoch": 0.09441233140655106,
      "grad_norm": 0.2571599781513214,
      "learning_rate": 0.0001,
      "loss": 1.583,
      "step": 588
    },
    {
      "epoch": 0.09457289659601799,
      "grad_norm": 0.25363945960998535,
      "learning_rate": 0.0001,
      "loss": 1.6196,
      "step": 589
    },
    {
      "epoch": 0.0947334617854849,
      "grad_norm": 0.2683764100074768,
      "learning_rate": 0.0001,
      "loss": 1.6649,
      "step": 590
    },
    {
      "epoch": 0.09489402697495183,
      "grad_norm": 0.27039188146591187,
      "learning_rate": 0.0001,
      "loss": 1.6218,
      "step": 591
    },
    {
      "epoch": 0.09505459216441875,
      "grad_norm": 0.29481080174446106,
      "learning_rate": 0.0001,
      "loss": 1.6723,
      "step": 592
    },
    {
      "epoch": 0.09521515735388568,
      "grad_norm": 0.2520233988761902,
      "learning_rate": 0.0001,
      "loss": 1.5608,
      "step": 593
    },
    {
      "epoch": 0.0953757225433526,
      "grad_norm": 0.286181777715683,
      "learning_rate": 0.0001,
      "loss": 1.6259,
      "step": 594
    },
    {
      "epoch": 0.09553628773281952,
      "grad_norm": 0.2637265920639038,
      "learning_rate": 0.0001,
      "loss": 1.6169,
      "step": 595
    },
    {
      "epoch": 0.09569685292228645,
      "grad_norm": 0.2822468876838684,
      "learning_rate": 0.0001,
      "loss": 1.6582,
      "step": 596
    },
    {
      "epoch": 0.09585741811175337,
      "grad_norm": 0.24705122411251068,
      "learning_rate": 0.0001,
      "loss": 1.6213,
      "step": 597
    },
    {
      "epoch": 0.0960179833012203,
      "grad_norm": 0.26160547137260437,
      "learning_rate": 0.0001,
      "loss": 1.6719,
      "step": 598
    },
    {
      "epoch": 0.09617854849068722,
      "grad_norm": 0.25721287727355957,
      "learning_rate": 0.0001,
      "loss": 1.63,
      "step": 599
    },
    {
      "epoch": 0.09633911368015415,
      "grad_norm": 0.26050350069999695,
      "learning_rate": 0.0001,
      "loss": 1.6177,
      "step": 600
    },
    {
      "epoch": 0.09649967886962106,
      "grad_norm": 0.27546998858451843,
      "learning_rate": 0.0001,
      "loss": 1.6334,
      "step": 601
    },
    {
      "epoch": 0.09666024405908799,
      "grad_norm": 0.256801962852478,
      "learning_rate": 0.0001,
      "loss": 1.5526,
      "step": 602
    },
    {
      "epoch": 0.0968208092485549,
      "grad_norm": 0.2656329870223999,
      "learning_rate": 0.0001,
      "loss": 1.6107,
      "step": 603
    },
    {
      "epoch": 0.09698137443802184,
      "grad_norm": 0.26477161049842834,
      "learning_rate": 0.0001,
      "loss": 1.6383,
      "step": 604
    },
    {
      "epoch": 0.09714193962748877,
      "grad_norm": 0.2643807530403137,
      "learning_rate": 0.0001,
      "loss": 1.6143,
      "step": 605
    },
    {
      "epoch": 0.09730250481695568,
      "grad_norm": 0.25192004442214966,
      "learning_rate": 0.0001,
      "loss": 1.5833,
      "step": 606
    },
    {
      "epoch": 0.09746307000642261,
      "grad_norm": 0.2630442678928375,
      "learning_rate": 0.0001,
      "loss": 1.6459,
      "step": 607
    },
    {
      "epoch": 0.09762363519588953,
      "grad_norm": 0.258197158575058,
      "learning_rate": 0.0001,
      "loss": 1.5996,
      "step": 608
    },
    {
      "epoch": 0.09778420038535646,
      "grad_norm": 0.26932215690612793,
      "learning_rate": 0.0001,
      "loss": 1.6144,
      "step": 609
    },
    {
      "epoch": 0.09794476557482337,
      "grad_norm": 0.28247031569480896,
      "learning_rate": 0.0001,
      "loss": 1.5781,
      "step": 610
    },
    {
      "epoch": 0.0981053307642903,
      "grad_norm": 0.268706738948822,
      "learning_rate": 0.0001,
      "loss": 1.618,
      "step": 611
    },
    {
      "epoch": 0.09826589595375723,
      "grad_norm": 0.2516535222530365,
      "learning_rate": 0.0001,
      "loss": 1.561,
      "step": 612
    },
    {
      "epoch": 0.09842646114322415,
      "grad_norm": 0.2579093873500824,
      "learning_rate": 0.0001,
      "loss": 1.6603,
      "step": 613
    },
    {
      "epoch": 0.09858702633269108,
      "grad_norm": 0.2705880105495453,
      "learning_rate": 0.0001,
      "loss": 1.5977,
      "step": 614
    },
    {
      "epoch": 0.098747591522158,
      "grad_norm": 0.28428101539611816,
      "learning_rate": 0.0001,
      "loss": 1.6044,
      "step": 615
    },
    {
      "epoch": 0.09890815671162492,
      "grad_norm": 0.26495474576950073,
      "learning_rate": 0.0001,
      "loss": 1.6648,
      "step": 616
    },
    {
      "epoch": 0.09906872190109184,
      "grad_norm": 0.2962563633918762,
      "learning_rate": 0.0001,
      "loss": 1.6269,
      "step": 617
    },
    {
      "epoch": 0.09922928709055877,
      "grad_norm": 0.26589927077293396,
      "learning_rate": 0.0001,
      "loss": 1.6277,
      "step": 618
    },
    {
      "epoch": 0.09938985228002568,
      "grad_norm": 0.2665221691131592,
      "learning_rate": 0.0001,
      "loss": 1.5782,
      "step": 619
    },
    {
      "epoch": 0.09955041746949261,
      "grad_norm": 0.24735838174819946,
      "learning_rate": 0.0001,
      "loss": 1.5841,
      "step": 620
    },
    {
      "epoch": 0.09971098265895954,
      "grad_norm": 0.25282835960388184,
      "learning_rate": 0.0001,
      "loss": 1.5395,
      "step": 621
    },
    {
      "epoch": 0.09987154784842646,
      "grad_norm": 0.27968940138816833,
      "learning_rate": 0.0001,
      "loss": 1.6733,
      "step": 622
    },
    {
      "epoch": 0.10003211303789339,
      "grad_norm": 0.25760510563850403,
      "learning_rate": 0.0001,
      "loss": 1.5227,
      "step": 623
    },
    {
      "epoch": 0.1001926782273603,
      "grad_norm": 0.2664651572704315,
      "learning_rate": 0.0001,
      "loss": 1.5838,
      "step": 624
    },
    {
      "epoch": 0.10035324341682723,
      "grad_norm": 0.2841992974281311,
      "learning_rate": 0.0001,
      "loss": 1.6479,
      "step": 625
    },
    {
      "epoch": 0.10051380860629415,
      "grad_norm": 0.2621322274208069,
      "learning_rate": 0.0001,
      "loss": 1.5879,
      "step": 626
    },
    {
      "epoch": 0.10067437379576108,
      "grad_norm": 0.27068841457366943,
      "learning_rate": 0.0001,
      "loss": 1.5974,
      "step": 627
    },
    {
      "epoch": 0.10083493898522801,
      "grad_norm": 0.30539098381996155,
      "learning_rate": 0.0001,
      "loss": 1.6288,
      "step": 628
    },
    {
      "epoch": 0.10099550417469493,
      "grad_norm": 0.24707889556884766,
      "learning_rate": 0.0001,
      "loss": 1.6518,
      "step": 629
    },
    {
      "epoch": 0.10115606936416185,
      "grad_norm": 0.26346689462661743,
      "learning_rate": 0.0001,
      "loss": 1.6062,
      "step": 630
    },
    {
      "epoch": 0.10131663455362877,
      "grad_norm": 0.2612040042877197,
      "learning_rate": 0.0001,
      "loss": 1.5447,
      "step": 631
    },
    {
      "epoch": 0.1014771997430957,
      "grad_norm": 0.2816939055919647,
      "learning_rate": 0.0001,
      "loss": 1.6229,
      "step": 632
    },
    {
      "epoch": 0.10163776493256262,
      "grad_norm": 0.27160507440567017,
      "learning_rate": 0.0001,
      "loss": 1.6842,
      "step": 633
    },
    {
      "epoch": 0.10179833012202955,
      "grad_norm": 0.25368353724479675,
      "learning_rate": 0.0001,
      "loss": 1.5702,
      "step": 634
    },
    {
      "epoch": 0.10195889531149646,
      "grad_norm": 0.26571977138519287,
      "learning_rate": 0.0001,
      "loss": 1.633,
      "step": 635
    },
    {
      "epoch": 0.10211946050096339,
      "grad_norm": 0.27272340655326843,
      "learning_rate": 0.0001,
      "loss": 1.6395,
      "step": 636
    },
    {
      "epoch": 0.10228002569043032,
      "grad_norm": 0.27077198028564453,
      "learning_rate": 0.0001,
      "loss": 1.6647,
      "step": 637
    },
    {
      "epoch": 0.10244059087989724,
      "grad_norm": 0.3076954185962677,
      "learning_rate": 0.0001,
      "loss": 1.5771,
      "step": 638
    },
    {
      "epoch": 0.10260115606936417,
      "grad_norm": 0.26889434456825256,
      "learning_rate": 0.0001,
      "loss": 1.6317,
      "step": 639
    },
    {
      "epoch": 0.10276172125883108,
      "grad_norm": 0.273711234331131,
      "learning_rate": 0.0001,
      "loss": 1.5547,
      "step": 640
    },
    {
      "epoch": 0.10292228644829801,
      "grad_norm": 0.25578954815864563,
      "learning_rate": 0.0001,
      "loss": 1.5703,
      "step": 641
    },
    {
      "epoch": 0.10308285163776493,
      "grad_norm": 0.2453918308019638,
      "learning_rate": 0.0001,
      "loss": 1.5894,
      "step": 642
    },
    {
      "epoch": 0.10324341682723186,
      "grad_norm": 0.2527920603752136,
      "learning_rate": 0.0001,
      "loss": 1.5275,
      "step": 643
    },
    {
      "epoch": 0.10340398201669879,
      "grad_norm": 0.2707948088645935,
      "learning_rate": 0.0001,
      "loss": 1.6847,
      "step": 644
    },
    {
      "epoch": 0.1035645472061657,
      "grad_norm": 0.2517828047275543,
      "learning_rate": 0.0001,
      "loss": 1.6775,
      "step": 645
    },
    {
      "epoch": 0.10372511239563263,
      "grad_norm": 0.2665092647075653,
      "learning_rate": 0.0001,
      "loss": 1.7126,
      "step": 646
    },
    {
      "epoch": 0.10388567758509955,
      "grad_norm": 0.2530919909477234,
      "learning_rate": 0.0001,
      "loss": 1.6325,
      "step": 647
    },
    {
      "epoch": 0.10404624277456648,
      "grad_norm": 0.25849777460098267,
      "learning_rate": 0.0001,
      "loss": 1.598,
      "step": 648
    },
    {
      "epoch": 0.1042068079640334,
      "grad_norm": 0.26646289229393005,
      "learning_rate": 0.0001,
      "loss": 1.6072,
      "step": 649
    },
    {
      "epoch": 0.10436737315350032,
      "grad_norm": 0.259625643491745,
      "learning_rate": 0.0001,
      "loss": 1.6249,
      "step": 650
    },
    {
      "epoch": 0.10452793834296724,
      "grad_norm": 0.2740899622440338,
      "learning_rate": 0.0001,
      "loss": 1.6599,
      "step": 651
    },
    {
      "epoch": 0.10468850353243417,
      "grad_norm": 0.24367207288742065,
      "learning_rate": 0.0001,
      "loss": 1.4812,
      "step": 652
    },
    {
      "epoch": 0.1048490687219011,
      "grad_norm": 0.260138601064682,
      "learning_rate": 0.0001,
      "loss": 1.6092,
      "step": 653
    },
    {
      "epoch": 0.10500963391136801,
      "grad_norm": 0.25125980377197266,
      "learning_rate": 0.0001,
      "loss": 1.5853,
      "step": 654
    },
    {
      "epoch": 0.10517019910083494,
      "grad_norm": 0.2736695110797882,
      "learning_rate": 0.0001,
      "loss": 1.593,
      "step": 655
    },
    {
      "epoch": 0.10533076429030186,
      "grad_norm": 1.329075813293457,
      "learning_rate": 0.0001,
      "loss": 1.6744,
      "step": 656
    },
    {
      "epoch": 0.10549132947976879,
      "grad_norm": 0.24662800133228302,
      "learning_rate": 0.0001,
      "loss": 1.5462,
      "step": 657
    },
    {
      "epoch": 0.1056518946692357,
      "grad_norm": 0.27666983008384705,
      "learning_rate": 0.0001,
      "loss": 1.6646,
      "step": 658
    },
    {
      "epoch": 0.10581245985870263,
      "grad_norm": 0.26717808842658997,
      "learning_rate": 0.0001,
      "loss": 1.7311,
      "step": 659
    },
    {
      "epoch": 0.10597302504816955,
      "grad_norm": 0.24117274582386017,
      "learning_rate": 0.0001,
      "loss": 1.4865,
      "step": 660
    },
    {
      "epoch": 0.10613359023763648,
      "grad_norm": 0.27742815017700195,
      "learning_rate": 0.0001,
      "loss": 1.5919,
      "step": 661
    },
    {
      "epoch": 0.10629415542710341,
      "grad_norm": 0.2660296559333801,
      "learning_rate": 0.0001,
      "loss": 1.6086,
      "step": 662
    },
    {
      "epoch": 0.10645472061657033,
      "grad_norm": 0.2646217346191406,
      "learning_rate": 0.0001,
      "loss": 1.7272,
      "step": 663
    },
    {
      "epoch": 0.10661528580603725,
      "grad_norm": 0.284201055765152,
      "learning_rate": 0.0001,
      "loss": 1.665,
      "step": 664
    },
    {
      "epoch": 0.10677585099550417,
      "grad_norm": 0.26237398386001587,
      "learning_rate": 0.0001,
      "loss": 1.6256,
      "step": 665
    },
    {
      "epoch": 0.1069364161849711,
      "grad_norm": 0.251699835062027,
      "learning_rate": 0.0001,
      "loss": 1.5611,
      "step": 666
    },
    {
      "epoch": 0.10709698137443802,
      "grad_norm": 0.26088041067123413,
      "learning_rate": 0.0001,
      "loss": 1.6255,
      "step": 667
    },
    {
      "epoch": 0.10725754656390495,
      "grad_norm": 0.2808876633644104,
      "learning_rate": 0.0001,
      "loss": 1.6421,
      "step": 668
    },
    {
      "epoch": 0.10741811175337188,
      "grad_norm": 0.25902828574180603,
      "learning_rate": 0.0001,
      "loss": 1.5471,
      "step": 669
    },
    {
      "epoch": 0.10757867694283879,
      "grad_norm": 0.27159273624420166,
      "learning_rate": 0.0001,
      "loss": 1.548,
      "step": 670
    },
    {
      "epoch": 0.10773924213230572,
      "grad_norm": 0.2766246199607849,
      "learning_rate": 0.0001,
      "loss": 1.5603,
      "step": 671
    },
    {
      "epoch": 0.10789980732177264,
      "grad_norm": 0.2636677622795105,
      "learning_rate": 0.0001,
      "loss": 1.663,
      "step": 672
    },
    {
      "epoch": 0.10806037251123957,
      "grad_norm": 0.2700072228908539,
      "learning_rate": 0.0001,
      "loss": 1.6586,
      "step": 673
    },
    {
      "epoch": 0.10822093770070648,
      "grad_norm": 0.26270678639411926,
      "learning_rate": 0.0001,
      "loss": 1.6112,
      "step": 674
    },
    {
      "epoch": 0.10838150289017341,
      "grad_norm": 0.2790539562702179,
      "learning_rate": 0.0001,
      "loss": 1.5652,
      "step": 675
    },
    {
      "epoch": 0.10854206807964033,
      "grad_norm": 0.2826235890388489,
      "learning_rate": 0.0001,
      "loss": 1.6552,
      "step": 676
    },
    {
      "epoch": 0.10870263326910726,
      "grad_norm": 0.2595786452293396,
      "learning_rate": 0.0001,
      "loss": 1.5313,
      "step": 677
    },
    {
      "epoch": 0.10886319845857419,
      "grad_norm": 0.25537219643592834,
      "learning_rate": 0.0001,
      "loss": 1.4861,
      "step": 678
    },
    {
      "epoch": 0.1090237636480411,
      "grad_norm": 0.26303696632385254,
      "learning_rate": 0.0001,
      "loss": 1.5731,
      "step": 679
    },
    {
      "epoch": 0.10918432883750803,
      "grad_norm": 0.5021303296089172,
      "learning_rate": 0.0001,
      "loss": 1.6219,
      "step": 680
    },
    {
      "epoch": 0.10934489402697495,
      "grad_norm": 0.2666873037815094,
      "learning_rate": 0.0001,
      "loss": 1.5963,
      "step": 681
    },
    {
      "epoch": 0.10950545921644188,
      "grad_norm": 0.30422693490982056,
      "learning_rate": 0.0001,
      "loss": 1.6347,
      "step": 682
    },
    {
      "epoch": 0.1096660244059088,
      "grad_norm": 0.26333922147750854,
      "learning_rate": 0.0001,
      "loss": 1.5852,
      "step": 683
    },
    {
      "epoch": 0.10982658959537572,
      "grad_norm": 0.24885666370391846,
      "learning_rate": 0.0001,
      "loss": 1.6133,
      "step": 684
    },
    {
      "epoch": 0.10998715478484265,
      "grad_norm": 0.2608605921268463,
      "learning_rate": 0.0001,
      "loss": 1.6446,
      "step": 685
    },
    {
      "epoch": 0.11014771997430957,
      "grad_norm": 0.29082754254341125,
      "learning_rate": 0.0001,
      "loss": 1.7178,
      "step": 686
    },
    {
      "epoch": 0.1103082851637765,
      "grad_norm": 0.2474960833787918,
      "learning_rate": 0.0001,
      "loss": 1.5778,
      "step": 687
    },
    {
      "epoch": 0.11046885035324341,
      "grad_norm": 0.2687987983226776,
      "learning_rate": 0.0001,
      "loss": 1.6162,
      "step": 688
    },
    {
      "epoch": 0.11062941554271034,
      "grad_norm": 0.28898414969444275,
      "learning_rate": 0.0001,
      "loss": 1.6355,
      "step": 689
    },
    {
      "epoch": 0.11078998073217726,
      "grad_norm": 0.28025883436203003,
      "learning_rate": 0.0001,
      "loss": 1.6369,
      "step": 690
    },
    {
      "epoch": 0.11095054592164419,
      "grad_norm": 0.29039886593818665,
      "learning_rate": 0.0001,
      "loss": 1.6717,
      "step": 691
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 0.2517712116241455,
      "learning_rate": 0.0001,
      "loss": 1.575,
      "step": 692
    },
    {
      "epoch": 0.11127167630057803,
      "grad_norm": 0.24026477336883545,
      "learning_rate": 0.0001,
      "loss": 1.5107,
      "step": 693
    },
    {
      "epoch": 0.11143224149004496,
      "grad_norm": 0.3061399459838867,
      "learning_rate": 0.0001,
      "loss": 1.5785,
      "step": 694
    },
    {
      "epoch": 0.11159280667951188,
      "grad_norm": 0.25361010432243347,
      "learning_rate": 0.0001,
      "loss": 1.5685,
      "step": 695
    },
    {
      "epoch": 0.11175337186897881,
      "grad_norm": 0.27122291922569275,
      "learning_rate": 0.0001,
      "loss": 1.547,
      "step": 696
    },
    {
      "epoch": 0.11191393705844573,
      "grad_norm": 0.2916577458381653,
      "learning_rate": 0.0001,
      "loss": 1.6699,
      "step": 697
    },
    {
      "epoch": 0.11207450224791266,
      "grad_norm": 0.2575892210006714,
      "learning_rate": 0.0001,
      "loss": 1.5471,
      "step": 698
    },
    {
      "epoch": 0.11223506743737957,
      "grad_norm": 0.27215683460235596,
      "learning_rate": 0.0001,
      "loss": 1.4966,
      "step": 699
    },
    {
      "epoch": 0.1123956326268465,
      "grad_norm": 0.2941299378871918,
      "learning_rate": 0.0001,
      "loss": 1.6237,
      "step": 700
    },
    {
      "epoch": 0.11255619781631342,
      "grad_norm": 0.26432883739471436,
      "learning_rate": 0.0001,
      "loss": 1.5561,
      "step": 701
    },
    {
      "epoch": 0.11271676300578035,
      "grad_norm": 0.2935466170310974,
      "learning_rate": 0.0001,
      "loss": 1.6628,
      "step": 702
    },
    {
      "epoch": 0.11287732819524728,
      "grad_norm": 0.283089280128479,
      "learning_rate": 0.0001,
      "loss": 1.6165,
      "step": 703
    },
    {
      "epoch": 0.11303789338471419,
      "grad_norm": 0.29502588510513306,
      "learning_rate": 0.0001,
      "loss": 1.6731,
      "step": 704
    },
    {
      "epoch": 0.11319845857418112,
      "grad_norm": 0.27801328897476196,
      "learning_rate": 0.0001,
      "loss": 1.6759,
      "step": 705
    },
    {
      "epoch": 0.11335902376364804,
      "grad_norm": 0.2630738317966461,
      "learning_rate": 0.0001,
      "loss": 1.5675,
      "step": 706
    },
    {
      "epoch": 0.11351958895311497,
      "grad_norm": 0.2872748076915741,
      "learning_rate": 0.0001,
      "loss": 1.6502,
      "step": 707
    },
    {
      "epoch": 0.11368015414258188,
      "grad_norm": 0.265083372592926,
      "learning_rate": 0.0001,
      "loss": 1.5443,
      "step": 708
    },
    {
      "epoch": 0.11384071933204881,
      "grad_norm": 0.2837506830692291,
      "learning_rate": 0.0001,
      "loss": 1.5847,
      "step": 709
    },
    {
      "epoch": 0.11400128452151574,
      "grad_norm": 0.29870733618736267,
      "learning_rate": 0.0001,
      "loss": 1.5679,
      "step": 710
    },
    {
      "epoch": 0.11416184971098266,
      "grad_norm": 0.27681732177734375,
      "learning_rate": 0.0001,
      "loss": 1.6262,
      "step": 711
    },
    {
      "epoch": 0.11432241490044959,
      "grad_norm": 0.2766120135784149,
      "learning_rate": 0.0001,
      "loss": 1.618,
      "step": 712
    },
    {
      "epoch": 0.1144829800899165,
      "grad_norm": 0.2983716130256653,
      "learning_rate": 0.0001,
      "loss": 1.6639,
      "step": 713
    },
    {
      "epoch": 0.11464354527938343,
      "grad_norm": 0.25736159086227417,
      "learning_rate": 0.0001,
      "loss": 1.5842,
      "step": 714
    },
    {
      "epoch": 0.11480411046885035,
      "grad_norm": 0.2921600043773651,
      "learning_rate": 0.0001,
      "loss": 1.6578,
      "step": 715
    },
    {
      "epoch": 0.11496467565831728,
      "grad_norm": 0.2478378266096115,
      "learning_rate": 0.0001,
      "loss": 1.5059,
      "step": 716
    },
    {
      "epoch": 0.1151252408477842,
      "grad_norm": 0.262857049703598,
      "learning_rate": 0.0001,
      "loss": 1.5717,
      "step": 717
    },
    {
      "epoch": 0.11528580603725112,
      "grad_norm": 0.2785440683364868,
      "learning_rate": 0.0001,
      "loss": 1.6867,
      "step": 718
    },
    {
      "epoch": 0.11544637122671805,
      "grad_norm": 0.2898574769496918,
      "learning_rate": 0.0001,
      "loss": 1.6168,
      "step": 719
    },
    {
      "epoch": 0.11560693641618497,
      "grad_norm": 0.2668989300727844,
      "learning_rate": 0.0001,
      "loss": 1.6193,
      "step": 720
    },
    {
      "epoch": 0.1157675016056519,
      "grad_norm": 0.2536221742630005,
      "learning_rate": 0.0001,
      "loss": 1.62,
      "step": 721
    },
    {
      "epoch": 0.11592806679511881,
      "grad_norm": 0.2669164836406708,
      "learning_rate": 0.0001,
      "loss": 1.6766,
      "step": 722
    },
    {
      "epoch": 0.11608863198458574,
      "grad_norm": 0.2606102526187897,
      "learning_rate": 0.0001,
      "loss": 1.6556,
      "step": 723
    },
    {
      "epoch": 0.11624919717405266,
      "grad_norm": 0.27430862188339233,
      "learning_rate": 0.0001,
      "loss": 1.6073,
      "step": 724
    },
    {
      "epoch": 0.11640976236351959,
      "grad_norm": 0.25970223546028137,
      "learning_rate": 0.0001,
      "loss": 1.6094,
      "step": 725
    },
    {
      "epoch": 0.11657032755298652,
      "grad_norm": 0.2728593349456787,
      "learning_rate": 0.0001,
      "loss": 1.6689,
      "step": 726
    },
    {
      "epoch": 0.11673089274245343,
      "grad_norm": 0.24974913895130157,
      "learning_rate": 0.0001,
      "loss": 1.5673,
      "step": 727
    },
    {
      "epoch": 0.11689145793192036,
      "grad_norm": 0.24918828904628754,
      "learning_rate": 0.0001,
      "loss": 1.566,
      "step": 728
    },
    {
      "epoch": 0.11705202312138728,
      "grad_norm": 0.2884836494922638,
      "learning_rate": 0.0001,
      "loss": 1.6147,
      "step": 729
    },
    {
      "epoch": 0.11721258831085421,
      "grad_norm": 0.33627474308013916,
      "learning_rate": 0.0001,
      "loss": 1.675,
      "step": 730
    },
    {
      "epoch": 0.11737315350032113,
      "grad_norm": 0.2641352415084839,
      "learning_rate": 0.0001,
      "loss": 1.6296,
      "step": 731
    },
    {
      "epoch": 0.11753371868978806,
      "grad_norm": 0.25515982508659363,
      "learning_rate": 0.0001,
      "loss": 1.6185,
      "step": 732
    },
    {
      "epoch": 0.11769428387925497,
      "grad_norm": 0.2719225585460663,
      "learning_rate": 0.0001,
      "loss": 1.6463,
      "step": 733
    },
    {
      "epoch": 0.1178548490687219,
      "grad_norm": 0.25415921211242676,
      "learning_rate": 0.0001,
      "loss": 1.606,
      "step": 734
    },
    {
      "epoch": 0.11801541425818883,
      "grad_norm": 0.24442121386528015,
      "learning_rate": 0.0001,
      "loss": 1.5864,
      "step": 735
    },
    {
      "epoch": 0.11817597944765575,
      "grad_norm": 0.25296634435653687,
      "learning_rate": 0.0001,
      "loss": 1.5527,
      "step": 736
    },
    {
      "epoch": 0.11833654463712268,
      "grad_norm": 0.2576388418674469,
      "learning_rate": 0.0001,
      "loss": 1.563,
      "step": 737
    },
    {
      "epoch": 0.11849710982658959,
      "grad_norm": 0.25560155510902405,
      "learning_rate": 0.0001,
      "loss": 1.5732,
      "step": 738
    },
    {
      "epoch": 0.11865767501605652,
      "grad_norm": 0.2664654552936554,
      "learning_rate": 0.0001,
      "loss": 1.654,
      "step": 739
    },
    {
      "epoch": 0.11881824020552344,
      "grad_norm": 0.2674829661846161,
      "learning_rate": 0.0001,
      "loss": 1.6087,
      "step": 740
    },
    {
      "epoch": 0.11897880539499037,
      "grad_norm": 0.25803428888320923,
      "learning_rate": 0.0001,
      "loss": 1.6313,
      "step": 741
    },
    {
      "epoch": 0.1191393705844573,
      "grad_norm": 0.26967284083366394,
      "learning_rate": 0.0001,
      "loss": 1.6864,
      "step": 742
    },
    {
      "epoch": 0.11929993577392421,
      "grad_norm": 0.2588898539543152,
      "learning_rate": 0.0001,
      "loss": 1.6192,
      "step": 743
    },
    {
      "epoch": 0.11946050096339114,
      "grad_norm": 0.25680235028266907,
      "learning_rate": 0.0001,
      "loss": 1.6432,
      "step": 744
    },
    {
      "epoch": 0.11962106615285806,
      "grad_norm": 0.2503674328327179,
      "learning_rate": 0.0001,
      "loss": 1.584,
      "step": 745
    },
    {
      "epoch": 0.11978163134232499,
      "grad_norm": 0.2691657245159149,
      "learning_rate": 0.0001,
      "loss": 1.6631,
      "step": 746
    },
    {
      "epoch": 0.1199421965317919,
      "grad_norm": 0.25727444887161255,
      "learning_rate": 0.0001,
      "loss": 1.611,
      "step": 747
    },
    {
      "epoch": 0.12010276172125883,
      "grad_norm": 0.2589010000228882,
      "learning_rate": 0.0001,
      "loss": 1.5458,
      "step": 748
    },
    {
      "epoch": 0.12026332691072575,
      "grad_norm": 0.2563769221305847,
      "learning_rate": 0.0001,
      "loss": 1.5711,
      "step": 749
    },
    {
      "epoch": 0.12042389210019268,
      "grad_norm": 0.27601945400238037,
      "learning_rate": 0.0001,
      "loss": 1.611,
      "step": 750
    },
    {
      "epoch": 0.12058445728965961,
      "grad_norm": 0.24667346477508545,
      "learning_rate": 0.0001,
      "loss": 1.5895,
      "step": 751
    },
    {
      "epoch": 0.12074502247912652,
      "grad_norm": 0.2612060010433197,
      "learning_rate": 0.0001,
      "loss": 1.5545,
      "step": 752
    },
    {
      "epoch": 0.12090558766859345,
      "grad_norm": 0.2722240686416626,
      "learning_rate": 0.0001,
      "loss": 1.5912,
      "step": 753
    },
    {
      "epoch": 0.12106615285806037,
      "grad_norm": 0.2581103444099426,
      "learning_rate": 0.0001,
      "loss": 1.6089,
      "step": 754
    },
    {
      "epoch": 0.1212267180475273,
      "grad_norm": 0.29264286160469055,
      "learning_rate": 0.0001,
      "loss": 1.5833,
      "step": 755
    },
    {
      "epoch": 0.12138728323699421,
      "grad_norm": 0.27083325386047363,
      "learning_rate": 0.0001,
      "loss": 1.6171,
      "step": 756
    },
    {
      "epoch": 0.12154784842646114,
      "grad_norm": 0.2839059829711914,
      "learning_rate": 0.0001,
      "loss": 1.6719,
      "step": 757
    },
    {
      "epoch": 0.12170841361592806,
      "grad_norm": 0.2692236602306366,
      "learning_rate": 0.0001,
      "loss": 1.5431,
      "step": 758
    },
    {
      "epoch": 0.12186897880539499,
      "grad_norm": 0.2881883680820465,
      "learning_rate": 0.0001,
      "loss": 1.677,
      "step": 759
    },
    {
      "epoch": 0.12202954399486192,
      "grad_norm": 0.25900858640670776,
      "learning_rate": 0.0001,
      "loss": 1.5977,
      "step": 760
    },
    {
      "epoch": 0.12219010918432884,
      "grad_norm": 0.2649819850921631,
      "learning_rate": 0.0001,
      "loss": 1.6206,
      "step": 761
    },
    {
      "epoch": 0.12235067437379576,
      "grad_norm": 0.26542365550994873,
      "learning_rate": 0.0001,
      "loss": 1.5952,
      "step": 762
    },
    {
      "epoch": 0.12251123956326268,
      "grad_norm": 0.25755128264427185,
      "learning_rate": 0.0001,
      "loss": 1.6292,
      "step": 763
    },
    {
      "epoch": 0.12267180475272961,
      "grad_norm": 0.25057458877563477,
      "learning_rate": 0.0001,
      "loss": 1.6348,
      "step": 764
    },
    {
      "epoch": 0.12283236994219653,
      "grad_norm": 0.2558939456939697,
      "learning_rate": 0.0001,
      "loss": 1.5879,
      "step": 765
    },
    {
      "epoch": 0.12299293513166346,
      "grad_norm": 0.2470790445804596,
      "learning_rate": 0.0001,
      "loss": 1.5863,
      "step": 766
    },
    {
      "epoch": 0.12315350032113039,
      "grad_norm": 0.2466525435447693,
      "learning_rate": 0.0001,
      "loss": 1.6173,
      "step": 767
    },
    {
      "epoch": 0.1233140655105973,
      "grad_norm": 0.24029389023780823,
      "learning_rate": 0.0001,
      "loss": 1.5194,
      "step": 768
    },
    {
      "epoch": 0.12347463070006423,
      "grad_norm": 0.25701624155044556,
      "learning_rate": 0.0001,
      "loss": 1.6094,
      "step": 769
    },
    {
      "epoch": 0.12363519588953115,
      "grad_norm": 0.2579065263271332,
      "learning_rate": 0.0001,
      "loss": 1.6371,
      "step": 770
    },
    {
      "epoch": 0.12379576107899808,
      "grad_norm": 0.25433599948883057,
      "learning_rate": 0.0001,
      "loss": 1.5433,
      "step": 771
    },
    {
      "epoch": 0.12395632626846499,
      "grad_norm": 0.25113752484321594,
      "learning_rate": 0.0001,
      "loss": 1.6899,
      "step": 772
    },
    {
      "epoch": 0.12411689145793192,
      "grad_norm": 0.25137612223625183,
      "learning_rate": 0.0001,
      "loss": 1.5753,
      "step": 773
    },
    {
      "epoch": 0.12427745664739884,
      "grad_norm": 0.25248846411705017,
      "learning_rate": 0.0001,
      "loss": 1.5925,
      "step": 774
    },
    {
      "epoch": 0.12443802183686577,
      "grad_norm": 0.25833266973495483,
      "learning_rate": 0.0001,
      "loss": 1.6114,
      "step": 775
    },
    {
      "epoch": 0.1245985870263327,
      "grad_norm": 0.26226380467414856,
      "learning_rate": 0.0001,
      "loss": 1.6383,
      "step": 776
    },
    {
      "epoch": 0.12475915221579961,
      "grad_norm": 0.27248066663742065,
      "learning_rate": 0.0001,
      "loss": 1.6382,
      "step": 777
    },
    {
      "epoch": 0.12491971740526654,
      "grad_norm": 0.2538253962993622,
      "learning_rate": 0.0001,
      "loss": 1.6168,
      "step": 778
    },
    {
      "epoch": 0.12508028259473347,
      "grad_norm": 0.2772725820541382,
      "learning_rate": 0.0001,
      "loss": 1.5979,
      "step": 779
    },
    {
      "epoch": 0.1252408477842004,
      "grad_norm": 0.2635462284088135,
      "learning_rate": 0.0001,
      "loss": 1.6585,
      "step": 780
    },
    {
      "epoch": 0.1254014129736673,
      "grad_norm": 0.24889424443244934,
      "learning_rate": 0.0001,
      "loss": 1.6109,
      "step": 781
    },
    {
      "epoch": 0.12556197816313422,
      "grad_norm": 0.2484816163778305,
      "learning_rate": 0.0001,
      "loss": 1.6207,
      "step": 782
    },
    {
      "epoch": 0.12572254335260116,
      "grad_norm": 0.2596292495727539,
      "learning_rate": 0.0001,
      "loss": 1.6653,
      "step": 783
    },
    {
      "epoch": 0.12588310854206808,
      "grad_norm": 0.26616355776786804,
      "learning_rate": 0.0001,
      "loss": 1.6816,
      "step": 784
    },
    {
      "epoch": 0.126043673731535,
      "grad_norm": 0.494791716337204,
      "learning_rate": 0.0001,
      "loss": 1.5498,
      "step": 785
    },
    {
      "epoch": 0.12620423892100194,
      "grad_norm": 0.2505752146244049,
      "learning_rate": 0.0001,
      "loss": 1.5923,
      "step": 786
    },
    {
      "epoch": 0.12636480411046885,
      "grad_norm": 0.26114824414253235,
      "learning_rate": 0.0001,
      "loss": 1.5349,
      "step": 787
    },
    {
      "epoch": 0.12652536929993577,
      "grad_norm": 0.2598491311073303,
      "learning_rate": 0.0001,
      "loss": 1.6301,
      "step": 788
    },
    {
      "epoch": 0.12668593448940269,
      "grad_norm": 0.2534463405609131,
      "learning_rate": 0.0001,
      "loss": 1.5631,
      "step": 789
    },
    {
      "epoch": 0.12684649967886963,
      "grad_norm": 0.27598607540130615,
      "learning_rate": 0.0001,
      "loss": 1.6233,
      "step": 790
    },
    {
      "epoch": 0.12700706486833654,
      "grad_norm": 0.2497626096010208,
      "learning_rate": 0.0001,
      "loss": 1.5669,
      "step": 791
    },
    {
      "epoch": 0.12716763005780346,
      "grad_norm": 0.2503911256790161,
      "learning_rate": 0.0001,
      "loss": 1.5778,
      "step": 792
    },
    {
      "epoch": 0.1273281952472704,
      "grad_norm": 0.2954845726490021,
      "learning_rate": 0.0001,
      "loss": 1.6333,
      "step": 793
    },
    {
      "epoch": 0.12748876043673732,
      "grad_norm": 0.2428663820028305,
      "learning_rate": 0.0001,
      "loss": 1.4572,
      "step": 794
    },
    {
      "epoch": 0.12764932562620424,
      "grad_norm": 0.2646285891532898,
      "learning_rate": 0.0001,
      "loss": 1.5702,
      "step": 795
    },
    {
      "epoch": 0.12780989081567115,
      "grad_norm": 0.25721773505210876,
      "learning_rate": 0.0001,
      "loss": 1.6594,
      "step": 796
    },
    {
      "epoch": 0.1279704560051381,
      "grad_norm": 0.2589038610458374,
      "learning_rate": 0.0001,
      "loss": 1.5998,
      "step": 797
    },
    {
      "epoch": 0.128131021194605,
      "grad_norm": 0.27920010685920715,
      "learning_rate": 0.0001,
      "loss": 1.6561,
      "step": 798
    },
    {
      "epoch": 0.12829158638407193,
      "grad_norm": 0.45557379722595215,
      "learning_rate": 0.0001,
      "loss": 1.5964,
      "step": 799
    },
    {
      "epoch": 0.12845215157353887,
      "grad_norm": 0.2536633610725403,
      "learning_rate": 0.0001,
      "loss": 1.5482,
      "step": 800
    },
    {
      "epoch": 0.12861271676300579,
      "grad_norm": 0.2707673907279968,
      "learning_rate": 0.0001,
      "loss": 1.6412,
      "step": 801
    },
    {
      "epoch": 0.1287732819524727,
      "grad_norm": 0.24133442342281342,
      "learning_rate": 0.0001,
      "loss": 1.589,
      "step": 802
    },
    {
      "epoch": 0.12893384714193962,
      "grad_norm": 0.2665124535560608,
      "learning_rate": 0.0001,
      "loss": 1.568,
      "step": 803
    },
    {
      "epoch": 0.12909441233140656,
      "grad_norm": 0.2640987038612366,
      "learning_rate": 0.0001,
      "loss": 1.565,
      "step": 804
    },
    {
      "epoch": 0.12925497752087348,
      "grad_norm": 0.2914842367172241,
      "learning_rate": 0.0001,
      "loss": 1.6363,
      "step": 805
    },
    {
      "epoch": 0.1294155427103404,
      "grad_norm": 0.25757482647895813,
      "learning_rate": 0.0001,
      "loss": 1.6507,
      "step": 806
    },
    {
      "epoch": 0.1295761078998073,
      "grad_norm": 0.24502956867218018,
      "learning_rate": 0.0001,
      "loss": 1.5228,
      "step": 807
    },
    {
      "epoch": 0.12973667308927425,
      "grad_norm": 0.25205034017562866,
      "learning_rate": 0.0001,
      "loss": 1.551,
      "step": 808
    },
    {
      "epoch": 0.12989723827874117,
      "grad_norm": 0.25903329253196716,
      "learning_rate": 0.0001,
      "loss": 1.566,
      "step": 809
    },
    {
      "epoch": 0.13005780346820808,
      "grad_norm": 0.24316738545894623,
      "learning_rate": 0.0001,
      "loss": 1.6612,
      "step": 810
    },
    {
      "epoch": 0.13021836865767503,
      "grad_norm": 0.24032652378082275,
      "learning_rate": 0.0001,
      "loss": 1.5407,
      "step": 811
    },
    {
      "epoch": 0.13037893384714194,
      "grad_norm": 0.2632538378238678,
      "learning_rate": 0.0001,
      "loss": 1.5716,
      "step": 812
    },
    {
      "epoch": 0.13053949903660886,
      "grad_norm": 0.27078595757484436,
      "learning_rate": 0.0001,
      "loss": 1.5277,
      "step": 813
    },
    {
      "epoch": 0.13070006422607577,
      "grad_norm": 0.2537182867527008,
      "learning_rate": 0.0001,
      "loss": 1.5279,
      "step": 814
    },
    {
      "epoch": 0.13086062941554272,
      "grad_norm": 0.2644798755645752,
      "learning_rate": 0.0001,
      "loss": 1.6266,
      "step": 815
    },
    {
      "epoch": 0.13102119460500963,
      "grad_norm": 0.24856843054294586,
      "learning_rate": 0.0001,
      "loss": 1.5094,
      "step": 816
    },
    {
      "epoch": 0.13118175979447655,
      "grad_norm": 0.24715793132781982,
      "learning_rate": 0.0001,
      "loss": 1.5693,
      "step": 817
    },
    {
      "epoch": 0.1313423249839435,
      "grad_norm": 0.2910032272338867,
      "learning_rate": 0.0001,
      "loss": 1.6246,
      "step": 818
    },
    {
      "epoch": 0.1315028901734104,
      "grad_norm": 0.2600892186164856,
      "learning_rate": 0.0001,
      "loss": 1.5608,
      "step": 819
    },
    {
      "epoch": 0.13166345536287732,
      "grad_norm": 0.28039804100990295,
      "learning_rate": 0.0001,
      "loss": 1.6317,
      "step": 820
    },
    {
      "epoch": 0.13182402055234424,
      "grad_norm": 0.25438904762268066,
      "learning_rate": 0.0001,
      "loss": 1.5446,
      "step": 821
    },
    {
      "epoch": 0.13198458574181118,
      "grad_norm": 0.24804502725601196,
      "learning_rate": 0.0001,
      "loss": 1.578,
      "step": 822
    },
    {
      "epoch": 0.1321451509312781,
      "grad_norm": 0.2627662122249603,
      "learning_rate": 0.0001,
      "loss": 1.6289,
      "step": 823
    },
    {
      "epoch": 0.13230571612074501,
      "grad_norm": 0.2525002062320709,
      "learning_rate": 0.0001,
      "loss": 1.6144,
      "step": 824
    },
    {
      "epoch": 0.13246628131021196,
      "grad_norm": 0.26678743958473206,
      "learning_rate": 0.0001,
      "loss": 1.5999,
      "step": 825
    },
    {
      "epoch": 0.13262684649967887,
      "grad_norm": 0.26153433322906494,
      "learning_rate": 0.0001,
      "loss": 1.6103,
      "step": 826
    },
    {
      "epoch": 0.1327874116891458,
      "grad_norm": 0.2502874732017517,
      "learning_rate": 0.0001,
      "loss": 1.5924,
      "step": 827
    },
    {
      "epoch": 0.1329479768786127,
      "grad_norm": 0.25310254096984863,
      "learning_rate": 0.0001,
      "loss": 1.6216,
      "step": 828
    },
    {
      "epoch": 0.13310854206807965,
      "grad_norm": 0.2579217851161957,
      "learning_rate": 0.0001,
      "loss": 1.6204,
      "step": 829
    },
    {
      "epoch": 0.13326910725754657,
      "grad_norm": 0.2544805109500885,
      "learning_rate": 0.0001,
      "loss": 1.5787,
      "step": 830
    },
    {
      "epoch": 0.13342967244701348,
      "grad_norm": 0.24990282952785492,
      "learning_rate": 0.0001,
      "loss": 1.5582,
      "step": 831
    },
    {
      "epoch": 0.13359023763648042,
      "grad_norm": 0.2465953677892685,
      "learning_rate": 0.0001,
      "loss": 1.5143,
      "step": 832
    },
    {
      "epoch": 0.13375080282594734,
      "grad_norm": 0.2525157928466797,
      "learning_rate": 0.0001,
      "loss": 1.6545,
      "step": 833
    },
    {
      "epoch": 0.13391136801541426,
      "grad_norm": 0.2696436643600464,
      "learning_rate": 0.0001,
      "loss": 1.5574,
      "step": 834
    },
    {
      "epoch": 0.13407193320488117,
      "grad_norm": 0.2418709248304367,
      "learning_rate": 0.0001,
      "loss": 1.5706,
      "step": 835
    },
    {
      "epoch": 0.13423249839434812,
      "grad_norm": 0.25983524322509766,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 836
    },
    {
      "epoch": 0.13439306358381503,
      "grad_norm": 0.27395179867744446,
      "learning_rate": 0.0001,
      "loss": 1.6555,
      "step": 837
    },
    {
      "epoch": 0.13455362877328195,
      "grad_norm": 0.2571142315864563,
      "learning_rate": 0.0001,
      "loss": 1.5881,
      "step": 838
    },
    {
      "epoch": 0.13471419396274886,
      "grad_norm": 0.26157963275909424,
      "learning_rate": 0.0001,
      "loss": 1.5938,
      "step": 839
    },
    {
      "epoch": 0.1348747591522158,
      "grad_norm": 0.2452707439661026,
      "learning_rate": 0.0001,
      "loss": 1.5732,
      "step": 840
    },
    {
      "epoch": 0.13503532434168272,
      "grad_norm": 0.2631083130836487,
      "learning_rate": 0.0001,
      "loss": 1.5928,
      "step": 841
    },
    {
      "epoch": 0.13519588953114964,
      "grad_norm": 0.2500780522823334,
      "learning_rate": 0.0001,
      "loss": 1.5071,
      "step": 842
    },
    {
      "epoch": 0.13535645472061658,
      "grad_norm": 0.25931090116500854,
      "learning_rate": 0.0001,
      "loss": 1.6517,
      "step": 843
    },
    {
      "epoch": 0.1355170199100835,
      "grad_norm": 0.2653353214263916,
      "learning_rate": 0.0001,
      "loss": 1.5604,
      "step": 844
    },
    {
      "epoch": 0.1356775850995504,
      "grad_norm": 0.2534460127353668,
      "learning_rate": 0.0001,
      "loss": 1.5848,
      "step": 845
    },
    {
      "epoch": 0.13583815028901733,
      "grad_norm": 0.25302770733833313,
      "learning_rate": 0.0001,
      "loss": 1.6274,
      "step": 846
    },
    {
      "epoch": 0.13599871547848427,
      "grad_norm": 0.28456246852874756,
      "learning_rate": 0.0001,
      "loss": 1.6039,
      "step": 847
    },
    {
      "epoch": 0.1361592806679512,
      "grad_norm": 0.2481805682182312,
      "learning_rate": 0.0001,
      "loss": 1.5726,
      "step": 848
    },
    {
      "epoch": 0.1363198458574181,
      "grad_norm": 0.2782268226146698,
      "learning_rate": 0.0001,
      "loss": 1.6275,
      "step": 849
    },
    {
      "epoch": 0.13648041104688505,
      "grad_norm": 0.24883927404880524,
      "learning_rate": 0.0001,
      "loss": 1.5345,
      "step": 850
    },
    {
      "epoch": 0.13664097623635196,
      "grad_norm": 0.26905950903892517,
      "learning_rate": 0.0001,
      "loss": 1.5555,
      "step": 851
    },
    {
      "epoch": 0.13680154142581888,
      "grad_norm": 0.278736412525177,
      "learning_rate": 0.0001,
      "loss": 1.6644,
      "step": 852
    },
    {
      "epoch": 0.1369621066152858,
      "grad_norm": 0.2647725045681,
      "learning_rate": 0.0001,
      "loss": 1.6271,
      "step": 853
    },
    {
      "epoch": 0.13712267180475274,
      "grad_norm": 0.26414796710014343,
      "learning_rate": 0.0001,
      "loss": 1.6707,
      "step": 854
    },
    {
      "epoch": 0.13728323699421965,
      "grad_norm": 0.26472604274749756,
      "learning_rate": 0.0001,
      "loss": 1.63,
      "step": 855
    },
    {
      "epoch": 0.13744380218368657,
      "grad_norm": 0.2509171664714813,
      "learning_rate": 0.0001,
      "loss": 1.6083,
      "step": 856
    },
    {
      "epoch": 0.1376043673731535,
      "grad_norm": 0.30091553926467896,
      "learning_rate": 0.0001,
      "loss": 1.6461,
      "step": 857
    },
    {
      "epoch": 0.13776493256262043,
      "grad_norm": 0.2546672523021698,
      "learning_rate": 0.0001,
      "loss": 1.6342,
      "step": 858
    },
    {
      "epoch": 0.13792549775208734,
      "grad_norm": 0.2552092671394348,
      "learning_rate": 0.0001,
      "loss": 1.4802,
      "step": 859
    },
    {
      "epoch": 0.13808606294155426,
      "grad_norm": 0.2652498781681061,
      "learning_rate": 0.0001,
      "loss": 1.5446,
      "step": 860
    },
    {
      "epoch": 0.1382466281310212,
      "grad_norm": 0.2531950771808624,
      "learning_rate": 0.0001,
      "loss": 1.625,
      "step": 861
    },
    {
      "epoch": 0.13840719332048812,
      "grad_norm": 0.2755614221096039,
      "learning_rate": 0.0001,
      "loss": 1.6873,
      "step": 862
    },
    {
      "epoch": 0.13856775850995504,
      "grad_norm": 0.269375741481781,
      "learning_rate": 0.0001,
      "loss": 1.5331,
      "step": 863
    },
    {
      "epoch": 0.13872832369942195,
      "grad_norm": 0.26859086751937866,
      "learning_rate": 0.0001,
      "loss": 1.6087,
      "step": 864
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 0.2632748782634735,
      "learning_rate": 0.0001,
      "loss": 1.5878,
      "step": 865
    },
    {
      "epoch": 0.1390494540783558,
      "grad_norm": 0.2613421678543091,
      "learning_rate": 0.0001,
      "loss": 1.6823,
      "step": 866
    },
    {
      "epoch": 0.13921001926782273,
      "grad_norm": 0.2982962727546692,
      "learning_rate": 0.0001,
      "loss": 1.644,
      "step": 867
    },
    {
      "epoch": 0.13937058445728967,
      "grad_norm": 0.2554149329662323,
      "learning_rate": 0.0001,
      "loss": 1.5431,
      "step": 868
    },
    {
      "epoch": 0.13953114964675659,
      "grad_norm": 0.2654079496860504,
      "learning_rate": 0.0001,
      "loss": 1.5379,
      "step": 869
    },
    {
      "epoch": 0.1396917148362235,
      "grad_norm": 0.2743433713912964,
      "learning_rate": 0.0001,
      "loss": 1.612,
      "step": 870
    },
    {
      "epoch": 0.13985228002569042,
      "grad_norm": 0.24886168539524078,
      "learning_rate": 0.0001,
      "loss": 1.6143,
      "step": 871
    },
    {
      "epoch": 0.14001284521515736,
      "grad_norm": 0.29661887884140015,
      "learning_rate": 0.0001,
      "loss": 1.69,
      "step": 872
    },
    {
      "epoch": 0.14017341040462428,
      "grad_norm": 0.25791892409324646,
      "learning_rate": 0.0001,
      "loss": 1.5737,
      "step": 873
    },
    {
      "epoch": 0.1403339755940912,
      "grad_norm": 0.25356629490852356,
      "learning_rate": 0.0001,
      "loss": 1.6322,
      "step": 874
    },
    {
      "epoch": 0.14049454078355814,
      "grad_norm": 0.2713717520236969,
      "learning_rate": 0.0001,
      "loss": 1.5806,
      "step": 875
    },
    {
      "epoch": 0.14065510597302505,
      "grad_norm": 0.25718873739242554,
      "learning_rate": 0.0001,
      "loss": 1.6829,
      "step": 876
    },
    {
      "epoch": 0.14081567116249197,
      "grad_norm": 0.24350662529468536,
      "learning_rate": 0.0001,
      "loss": 1.544,
      "step": 877
    },
    {
      "epoch": 0.14097623635195888,
      "grad_norm": 0.2600824236869812,
      "learning_rate": 0.0001,
      "loss": 1.584,
      "step": 878
    },
    {
      "epoch": 0.14113680154142583,
      "grad_norm": 0.26631075143814087,
      "learning_rate": 0.0001,
      "loss": 1.6089,
      "step": 879
    },
    {
      "epoch": 0.14129736673089274,
      "grad_norm": 0.2674790918827057,
      "learning_rate": 0.0001,
      "loss": 1.6576,
      "step": 880
    },
    {
      "epoch": 0.14145793192035966,
      "grad_norm": 0.27016696333885193,
      "learning_rate": 0.0001,
      "loss": 1.5664,
      "step": 881
    },
    {
      "epoch": 0.1416184971098266,
      "grad_norm": 0.2517258822917938,
      "learning_rate": 0.0001,
      "loss": 1.5919,
      "step": 882
    },
    {
      "epoch": 0.14177906229929352,
      "grad_norm": 0.25702401995658875,
      "learning_rate": 0.0001,
      "loss": 1.55,
      "step": 883
    },
    {
      "epoch": 0.14193962748876043,
      "grad_norm": 0.2697002589702606,
      "learning_rate": 0.0001,
      "loss": 1.6049,
      "step": 884
    },
    {
      "epoch": 0.14210019267822735,
      "grad_norm": 0.25588372349739075,
      "learning_rate": 0.0001,
      "loss": 1.5859,
      "step": 885
    },
    {
      "epoch": 0.1422607578676943,
      "grad_norm": 0.2595311105251312,
      "learning_rate": 0.0001,
      "loss": 1.5958,
      "step": 886
    },
    {
      "epoch": 0.1424213230571612,
      "grad_norm": 0.2488589882850647,
      "learning_rate": 0.0001,
      "loss": 1.6681,
      "step": 887
    },
    {
      "epoch": 0.14258188824662812,
      "grad_norm": 0.23933641612529755,
      "learning_rate": 0.0001,
      "loss": 1.5913,
      "step": 888
    },
    {
      "epoch": 0.14274245343609507,
      "grad_norm": 0.24618078768253326,
      "learning_rate": 0.0001,
      "loss": 1.5106,
      "step": 889
    },
    {
      "epoch": 0.14290301862556198,
      "grad_norm": 0.24883906543254852,
      "learning_rate": 0.0001,
      "loss": 1.582,
      "step": 890
    },
    {
      "epoch": 0.1430635838150289,
      "grad_norm": 0.27218297123908997,
      "learning_rate": 0.0001,
      "loss": 1.5886,
      "step": 891
    },
    {
      "epoch": 0.14322414900449582,
      "grad_norm": 0.2581946849822998,
      "learning_rate": 0.0001,
      "loss": 1.5981,
      "step": 892
    },
    {
      "epoch": 0.14338471419396276,
      "grad_norm": 0.24821656942367554,
      "learning_rate": 0.0001,
      "loss": 1.5871,
      "step": 893
    },
    {
      "epoch": 0.14354527938342967,
      "grad_norm": 0.2561013996601105,
      "learning_rate": 0.0001,
      "loss": 1.6149,
      "step": 894
    },
    {
      "epoch": 0.1437058445728966,
      "grad_norm": 0.2528505325317383,
      "learning_rate": 0.0001,
      "loss": 1.6011,
      "step": 895
    },
    {
      "epoch": 0.1438664097623635,
      "grad_norm": 0.2521515190601349,
      "learning_rate": 0.0001,
      "loss": 1.5481,
      "step": 896
    },
    {
      "epoch": 0.14402697495183045,
      "grad_norm": 0.262277215719223,
      "learning_rate": 0.0001,
      "loss": 1.5968,
      "step": 897
    },
    {
      "epoch": 0.14418754014129737,
      "grad_norm": 0.2619973421096802,
      "learning_rate": 0.0001,
      "loss": 1.5859,
      "step": 898
    },
    {
      "epoch": 0.14434810533076428,
      "grad_norm": 0.2554990351200104,
      "learning_rate": 0.0001,
      "loss": 1.6384,
      "step": 899
    },
    {
      "epoch": 0.14450867052023122,
      "grad_norm": 0.25106117129325867,
      "learning_rate": 0.0001,
      "loss": 1.5078,
      "step": 900
    },
    {
      "epoch": 0.14466923570969814,
      "grad_norm": 0.2812288701534271,
      "learning_rate": 0.0001,
      "loss": 1.697,
      "step": 901
    },
    {
      "epoch": 0.14482980089916506,
      "grad_norm": 0.24550653994083405,
      "learning_rate": 0.0001,
      "loss": 1.4961,
      "step": 902
    },
    {
      "epoch": 0.14499036608863197,
      "grad_norm": 0.2630595564842224,
      "learning_rate": 0.0001,
      "loss": 1.6954,
      "step": 903
    },
    {
      "epoch": 0.14515093127809892,
      "grad_norm": 0.25216740369796753,
      "learning_rate": 0.0001,
      "loss": 1.6342,
      "step": 904
    },
    {
      "epoch": 0.14531149646756583,
      "grad_norm": 0.2595326602458954,
      "learning_rate": 0.0001,
      "loss": 1.6311,
      "step": 905
    },
    {
      "epoch": 0.14547206165703275,
      "grad_norm": 0.24139900505542755,
      "learning_rate": 0.0001,
      "loss": 1.5731,
      "step": 906
    },
    {
      "epoch": 0.1456326268464997,
      "grad_norm": 0.2659880816936493,
      "learning_rate": 0.0001,
      "loss": 1.626,
      "step": 907
    },
    {
      "epoch": 0.1457931920359666,
      "grad_norm": 0.24561461806297302,
      "learning_rate": 0.0001,
      "loss": 1.5737,
      "step": 908
    },
    {
      "epoch": 0.14595375722543352,
      "grad_norm": 0.23585672676563263,
      "learning_rate": 0.0001,
      "loss": 1.5343,
      "step": 909
    },
    {
      "epoch": 0.14611432241490044,
      "grad_norm": 0.2718529999256134,
      "learning_rate": 0.0001,
      "loss": 1.5992,
      "step": 910
    },
    {
      "epoch": 0.14627488760436738,
      "grad_norm": 0.2534932494163513,
      "learning_rate": 0.0001,
      "loss": 1.6032,
      "step": 911
    },
    {
      "epoch": 0.1464354527938343,
      "grad_norm": 0.2514909505844116,
      "learning_rate": 0.0001,
      "loss": 1.5442,
      "step": 912
    },
    {
      "epoch": 0.1465960179833012,
      "grad_norm": 0.2621341943740845,
      "learning_rate": 0.0001,
      "loss": 1.6254,
      "step": 913
    },
    {
      "epoch": 0.14675658317276816,
      "grad_norm": 0.2527523636817932,
      "learning_rate": 0.0001,
      "loss": 1.5434,
      "step": 914
    },
    {
      "epoch": 0.14691714836223507,
      "grad_norm": 0.2540346086025238,
      "learning_rate": 0.0001,
      "loss": 1.5594,
      "step": 915
    },
    {
      "epoch": 0.147077713551702,
      "grad_norm": 0.2796066701412201,
      "learning_rate": 0.0001,
      "loss": 1.6697,
      "step": 916
    },
    {
      "epoch": 0.1472382787411689,
      "grad_norm": 0.2621716558933258,
      "learning_rate": 0.0001,
      "loss": 1.6149,
      "step": 917
    },
    {
      "epoch": 0.14739884393063585,
      "grad_norm": 0.2522697448730469,
      "learning_rate": 0.0001,
      "loss": 1.5868,
      "step": 918
    },
    {
      "epoch": 0.14755940912010276,
      "grad_norm": 0.2512277662754059,
      "learning_rate": 0.0001,
      "loss": 1.5457,
      "step": 919
    },
    {
      "epoch": 0.14771997430956968,
      "grad_norm": 0.2664959728717804,
      "learning_rate": 0.0001,
      "loss": 1.609,
      "step": 920
    },
    {
      "epoch": 0.1478805394990366,
      "grad_norm": 0.2599131166934967,
      "learning_rate": 0.0001,
      "loss": 1.6409,
      "step": 921
    },
    {
      "epoch": 0.14804110468850354,
      "grad_norm": 0.2691696584224701,
      "learning_rate": 0.0001,
      "loss": 1.6161,
      "step": 922
    },
    {
      "epoch": 0.14820166987797045,
      "grad_norm": 0.26327449083328247,
      "learning_rate": 0.0001,
      "loss": 1.5896,
      "step": 923
    },
    {
      "epoch": 0.14836223506743737,
      "grad_norm": 0.25307759642601013,
      "learning_rate": 0.0001,
      "loss": 1.5517,
      "step": 924
    },
    {
      "epoch": 0.1485228002569043,
      "grad_norm": 0.27031588554382324,
      "learning_rate": 0.0001,
      "loss": 1.6378,
      "step": 925
    },
    {
      "epoch": 0.14868336544637123,
      "grad_norm": 0.24228115379810333,
      "learning_rate": 0.0001,
      "loss": 1.4805,
      "step": 926
    },
    {
      "epoch": 0.14884393063583815,
      "grad_norm": 0.25565674901008606,
      "learning_rate": 0.0001,
      "loss": 1.6749,
      "step": 927
    },
    {
      "epoch": 0.14900449582530506,
      "grad_norm": 0.2539302706718445,
      "learning_rate": 0.0001,
      "loss": 1.5772,
      "step": 928
    },
    {
      "epoch": 0.149165061014772,
      "grad_norm": 0.2562558650970459,
      "learning_rate": 0.0001,
      "loss": 1.6548,
      "step": 929
    },
    {
      "epoch": 0.14932562620423892,
      "grad_norm": 0.26420676708221436,
      "learning_rate": 0.0001,
      "loss": 1.6085,
      "step": 930
    },
    {
      "epoch": 0.14948619139370584,
      "grad_norm": 0.28227153420448303,
      "learning_rate": 0.0001,
      "loss": 1.5889,
      "step": 931
    },
    {
      "epoch": 0.14964675658317278,
      "grad_norm": 0.2469702661037445,
      "learning_rate": 0.0001,
      "loss": 1.5479,
      "step": 932
    },
    {
      "epoch": 0.1498073217726397,
      "grad_norm": 0.25898122787475586,
      "learning_rate": 0.0001,
      "loss": 1.5962,
      "step": 933
    },
    {
      "epoch": 0.1499678869621066,
      "grad_norm": 0.2539020776748657,
      "learning_rate": 0.0001,
      "loss": 1.5412,
      "step": 934
    },
    {
      "epoch": 0.15012845215157353,
      "grad_norm": 0.2724524438381195,
      "learning_rate": 0.0001,
      "loss": 1.6697,
      "step": 935
    },
    {
      "epoch": 0.15028901734104047,
      "grad_norm": 2.3618476390838623,
      "learning_rate": 0.0001,
      "loss": 1.6803,
      "step": 936
    },
    {
      "epoch": 0.1504495825305074,
      "grad_norm": 0.2667105197906494,
      "learning_rate": 0.0001,
      "loss": 1.6604,
      "step": 937
    },
    {
      "epoch": 0.1506101477199743,
      "grad_norm": 0.2506130039691925,
      "learning_rate": 0.0001,
      "loss": 1.4764,
      "step": 938
    },
    {
      "epoch": 0.15077071290944125,
      "grad_norm": 0.250129759311676,
      "learning_rate": 0.0001,
      "loss": 1.546,
      "step": 939
    },
    {
      "epoch": 0.15093127809890816,
      "grad_norm": 0.2476504147052765,
      "learning_rate": 0.0001,
      "loss": 1.5858,
      "step": 940
    },
    {
      "epoch": 0.15109184328837508,
      "grad_norm": 0.2472328394651413,
      "learning_rate": 0.0001,
      "loss": 1.5386,
      "step": 941
    },
    {
      "epoch": 0.151252408477842,
      "grad_norm": 0.24910295009613037,
      "learning_rate": 0.0001,
      "loss": 1.6534,
      "step": 942
    },
    {
      "epoch": 0.15141297366730894,
      "grad_norm": 0.2372818887233734,
      "learning_rate": 0.0001,
      "loss": 1.5527,
      "step": 943
    },
    {
      "epoch": 0.15157353885677585,
      "grad_norm": 0.2683623135089874,
      "learning_rate": 0.0001,
      "loss": 1.6166,
      "step": 944
    },
    {
      "epoch": 0.15173410404624277,
      "grad_norm": 0.24610085785388947,
      "learning_rate": 0.0001,
      "loss": 1.5643,
      "step": 945
    },
    {
      "epoch": 0.1518946692357097,
      "grad_norm": 0.2642036974430084,
      "learning_rate": 0.0001,
      "loss": 1.6591,
      "step": 946
    },
    {
      "epoch": 0.15205523442517663,
      "grad_norm": 0.25564056634902954,
      "learning_rate": 0.0001,
      "loss": 1.5975,
      "step": 947
    },
    {
      "epoch": 0.15221579961464354,
      "grad_norm": 0.25094327330589294,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 948
    },
    {
      "epoch": 0.15237636480411046,
      "grad_norm": 0.27053749561309814,
      "learning_rate": 0.0001,
      "loss": 1.6405,
      "step": 949
    },
    {
      "epoch": 0.1525369299935774,
      "grad_norm": 0.25663068890571594,
      "learning_rate": 0.0001,
      "loss": 1.6203,
      "step": 950
    },
    {
      "epoch": 0.15269749518304432,
      "grad_norm": 0.2562296986579895,
      "learning_rate": 0.0001,
      "loss": 1.5238,
      "step": 951
    },
    {
      "epoch": 0.15285806037251123,
      "grad_norm": 0.45618748664855957,
      "learning_rate": 0.0001,
      "loss": 1.6316,
      "step": 952
    },
    {
      "epoch": 0.15301862556197815,
      "grad_norm": 0.2593305706977844,
      "learning_rate": 0.0001,
      "loss": 1.6199,
      "step": 953
    },
    {
      "epoch": 0.1531791907514451,
      "grad_norm": 0.2535325288772583,
      "learning_rate": 0.0001,
      "loss": 1.5393,
      "step": 954
    },
    {
      "epoch": 0.153339755940912,
      "grad_norm": 0.2570097744464874,
      "learning_rate": 0.0001,
      "loss": 1.5836,
      "step": 955
    },
    {
      "epoch": 0.15350032113037893,
      "grad_norm": 0.26658475399017334,
      "learning_rate": 0.0001,
      "loss": 1.5609,
      "step": 956
    },
    {
      "epoch": 0.15366088631984587,
      "grad_norm": 0.23367327451705933,
      "learning_rate": 0.0001,
      "loss": 1.4769,
      "step": 957
    },
    {
      "epoch": 0.15382145150931278,
      "grad_norm": 0.2894875109195709,
      "learning_rate": 0.0001,
      "loss": 1.6935,
      "step": 958
    },
    {
      "epoch": 0.1539820166987797,
      "grad_norm": 0.24367184937000275,
      "learning_rate": 0.0001,
      "loss": 1.5402,
      "step": 959
    },
    {
      "epoch": 0.15414258188824662,
      "grad_norm": 0.2570075988769531,
      "learning_rate": 0.0001,
      "loss": 1.5435,
      "step": 960
    },
    {
      "epoch": 0.15430314707771356,
      "grad_norm": 0.26234692335128784,
      "learning_rate": 0.0001,
      "loss": 1.612,
      "step": 961
    },
    {
      "epoch": 0.15446371226718048,
      "grad_norm": 0.2623891830444336,
      "learning_rate": 0.0001,
      "loss": 1.5721,
      "step": 962
    },
    {
      "epoch": 0.1546242774566474,
      "grad_norm": 0.26043277978897095,
      "learning_rate": 0.0001,
      "loss": 1.6106,
      "step": 963
    },
    {
      "epoch": 0.15478484264611433,
      "grad_norm": 0.2588985860347748,
      "learning_rate": 0.0001,
      "loss": 1.5302,
      "step": 964
    },
    {
      "epoch": 0.15494540783558125,
      "grad_norm": 0.25718897581100464,
      "learning_rate": 0.0001,
      "loss": 1.5629,
      "step": 965
    },
    {
      "epoch": 0.15510597302504817,
      "grad_norm": 0.2608674466609955,
      "learning_rate": 0.0001,
      "loss": 1.5503,
      "step": 966
    },
    {
      "epoch": 0.15526653821451508,
      "grad_norm": 0.26131555438041687,
      "learning_rate": 0.0001,
      "loss": 1.5558,
      "step": 967
    },
    {
      "epoch": 0.15542710340398203,
      "grad_norm": 0.2566625475883484,
      "learning_rate": 0.0001,
      "loss": 1.6243,
      "step": 968
    },
    {
      "epoch": 0.15558766859344894,
      "grad_norm": 0.27965059876441956,
      "learning_rate": 0.0001,
      "loss": 1.5893,
      "step": 969
    },
    {
      "epoch": 0.15574823378291586,
      "grad_norm": 0.25014302134513855,
      "learning_rate": 0.0001,
      "loss": 1.5958,
      "step": 970
    },
    {
      "epoch": 0.1559087989723828,
      "grad_norm": 0.2523461580276489,
      "learning_rate": 0.0001,
      "loss": 1.5695,
      "step": 971
    },
    {
      "epoch": 0.15606936416184972,
      "grad_norm": 0.2674293518066406,
      "learning_rate": 0.0001,
      "loss": 1.5749,
      "step": 972
    },
    {
      "epoch": 0.15622992935131663,
      "grad_norm": 0.2695954144001007,
      "learning_rate": 0.0001,
      "loss": 1.723,
      "step": 973
    },
    {
      "epoch": 0.15639049454078355,
      "grad_norm": 0.25040581822395325,
      "learning_rate": 0.0001,
      "loss": 1.6021,
      "step": 974
    },
    {
      "epoch": 0.1565510597302505,
      "grad_norm": 0.252522349357605,
      "learning_rate": 0.0001,
      "loss": 1.5453,
      "step": 975
    },
    {
      "epoch": 0.1567116249197174,
      "grad_norm": 0.263759583234787,
      "learning_rate": 0.0001,
      "loss": 1.5177,
      "step": 976
    },
    {
      "epoch": 0.15687219010918432,
      "grad_norm": 0.26149657368659973,
      "learning_rate": 0.0001,
      "loss": 1.6225,
      "step": 977
    },
    {
      "epoch": 0.15703275529865124,
      "grad_norm": 0.2430988848209381,
      "learning_rate": 0.0001,
      "loss": 1.5972,
      "step": 978
    },
    {
      "epoch": 0.15719332048811818,
      "grad_norm": 0.2938035726547241,
      "learning_rate": 0.0001,
      "loss": 1.5743,
      "step": 979
    },
    {
      "epoch": 0.1573538856775851,
      "grad_norm": 0.24420037865638733,
      "learning_rate": 0.0001,
      "loss": 1.6323,
      "step": 980
    },
    {
      "epoch": 0.157514450867052,
      "grad_norm": 0.269610196352005,
      "learning_rate": 0.0001,
      "loss": 1.6412,
      "step": 981
    },
    {
      "epoch": 0.15767501605651896,
      "grad_norm": 0.25811803340911865,
      "learning_rate": 0.0001,
      "loss": 1.6129,
      "step": 982
    },
    {
      "epoch": 0.15783558124598587,
      "grad_norm": 0.2406802475452423,
      "learning_rate": 0.0001,
      "loss": 1.5603,
      "step": 983
    },
    {
      "epoch": 0.1579961464354528,
      "grad_norm": 0.2595805525779724,
      "learning_rate": 0.0001,
      "loss": 1.5083,
      "step": 984
    },
    {
      "epoch": 0.1581567116249197,
      "grad_norm": 0.260524183511734,
      "learning_rate": 0.0001,
      "loss": 1.614,
      "step": 985
    },
    {
      "epoch": 0.15831727681438665,
      "grad_norm": 0.25833332538604736,
      "learning_rate": 0.0001,
      "loss": 1.5933,
      "step": 986
    },
    {
      "epoch": 0.15847784200385356,
      "grad_norm": 0.25654345750808716,
      "learning_rate": 0.0001,
      "loss": 1.6111,
      "step": 987
    },
    {
      "epoch": 0.15863840719332048,
      "grad_norm": 0.26671066880226135,
      "learning_rate": 0.0001,
      "loss": 1.5813,
      "step": 988
    },
    {
      "epoch": 0.15879897238278742,
      "grad_norm": 0.2486371546983719,
      "learning_rate": 0.0001,
      "loss": 1.5531,
      "step": 989
    },
    {
      "epoch": 0.15895953757225434,
      "grad_norm": 0.2565583288669586,
      "learning_rate": 0.0001,
      "loss": 1.5289,
      "step": 990
    },
    {
      "epoch": 0.15912010276172125,
      "grad_norm": 0.24416203796863556,
      "learning_rate": 0.0001,
      "loss": 1.5113,
      "step": 991
    },
    {
      "epoch": 0.15928066795118817,
      "grad_norm": 0.27367034554481506,
      "learning_rate": 0.0001,
      "loss": 1.6646,
      "step": 992
    },
    {
      "epoch": 0.15944123314065511,
      "grad_norm": 0.26712900400161743,
      "learning_rate": 0.0001,
      "loss": 1.5482,
      "step": 993
    },
    {
      "epoch": 0.15960179833012203,
      "grad_norm": 0.2603553533554077,
      "learning_rate": 0.0001,
      "loss": 1.6457,
      "step": 994
    },
    {
      "epoch": 0.15976236351958895,
      "grad_norm": 0.23969604074954987,
      "learning_rate": 0.0001,
      "loss": 1.5153,
      "step": 995
    },
    {
      "epoch": 0.1599229287090559,
      "grad_norm": 0.27515581250190735,
      "learning_rate": 0.0001,
      "loss": 1.5772,
      "step": 996
    },
    {
      "epoch": 0.1600834938985228,
      "grad_norm": 0.2648704946041107,
      "learning_rate": 0.0001,
      "loss": 1.587,
      "step": 997
    },
    {
      "epoch": 0.16024405908798972,
      "grad_norm": 0.27304160594940186,
      "learning_rate": 0.0001,
      "loss": 1.6008,
      "step": 998
    },
    {
      "epoch": 0.16040462427745664,
      "grad_norm": 0.25929275155067444,
      "learning_rate": 0.0001,
      "loss": 1.594,
      "step": 999
    },
    {
      "epoch": 0.16056518946692358,
      "grad_norm": 0.25871768593788147,
      "learning_rate": 0.0001,
      "loss": 1.5781,
      "step": 1000
    },
    {
      "epoch": 0.1607257546563905,
      "grad_norm": 0.2427658885717392,
      "learning_rate": 0.0001,
      "loss": 1.5846,
      "step": 1001
    },
    {
      "epoch": 0.1608863198458574,
      "grad_norm": 0.24906103312969208,
      "learning_rate": 0.0001,
      "loss": 1.535,
      "step": 1002
    },
    {
      "epoch": 0.16104688503532436,
      "grad_norm": 0.24545247852802277,
      "learning_rate": 0.0001,
      "loss": 1.5767,
      "step": 1003
    },
    {
      "epoch": 0.16120745022479127,
      "grad_norm": 0.23803065717220306,
      "learning_rate": 0.0001,
      "loss": 1.5541,
      "step": 1004
    },
    {
      "epoch": 0.1613680154142582,
      "grad_norm": 0.23469966650009155,
      "learning_rate": 0.0001,
      "loss": 1.5283,
      "step": 1005
    },
    {
      "epoch": 0.1615285806037251,
      "grad_norm": 0.24923951923847198,
      "learning_rate": 0.0001,
      "loss": 1.5867,
      "step": 1006
    },
    {
      "epoch": 0.16168914579319205,
      "grad_norm": 0.24812714755535126,
      "learning_rate": 0.0001,
      "loss": 1.6495,
      "step": 1007
    },
    {
      "epoch": 0.16184971098265896,
      "grad_norm": 0.2443220466375351,
      "learning_rate": 0.0001,
      "loss": 1.5867,
      "step": 1008
    },
    {
      "epoch": 0.16201027617212588,
      "grad_norm": 0.2561798095703125,
      "learning_rate": 0.0001,
      "loss": 1.5966,
      "step": 1009
    },
    {
      "epoch": 0.1621708413615928,
      "grad_norm": 0.2338484227657318,
      "learning_rate": 0.0001,
      "loss": 1.5094,
      "step": 1010
    },
    {
      "epoch": 0.16233140655105974,
      "grad_norm": 0.24996957182884216,
      "learning_rate": 0.0001,
      "loss": 1.6699,
      "step": 1011
    },
    {
      "epoch": 0.16249197174052665,
      "grad_norm": 0.231418639421463,
      "learning_rate": 0.0001,
      "loss": 1.4153,
      "step": 1012
    },
    {
      "epoch": 0.16265253692999357,
      "grad_norm": 0.24210630357265472,
      "learning_rate": 0.0001,
      "loss": 1.5622,
      "step": 1013
    },
    {
      "epoch": 0.1628131021194605,
      "grad_norm": 0.239084392786026,
      "learning_rate": 0.0001,
      "loss": 1.552,
      "step": 1014
    },
    {
      "epoch": 0.16297366730892743,
      "grad_norm": 0.2544187009334564,
      "learning_rate": 0.0001,
      "loss": 1.6438,
      "step": 1015
    },
    {
      "epoch": 0.16313423249839434,
      "grad_norm": 0.249491348862648,
      "learning_rate": 0.0001,
      "loss": 1.5598,
      "step": 1016
    },
    {
      "epoch": 0.16329479768786126,
      "grad_norm": 0.24316665530204773,
      "learning_rate": 0.0001,
      "loss": 1.5345,
      "step": 1017
    },
    {
      "epoch": 0.1634553628773282,
      "grad_norm": 0.2577227056026459,
      "learning_rate": 0.0001,
      "loss": 1.582,
      "step": 1018
    },
    {
      "epoch": 0.16361592806679512,
      "grad_norm": 0.2507282793521881,
      "learning_rate": 0.0001,
      "loss": 1.5339,
      "step": 1019
    },
    {
      "epoch": 0.16377649325626203,
      "grad_norm": 0.24784612655639648,
      "learning_rate": 0.0001,
      "loss": 1.5851,
      "step": 1020
    },
    {
      "epoch": 0.16393705844572898,
      "grad_norm": 0.263863205909729,
      "learning_rate": 0.0001,
      "loss": 1.618,
      "step": 1021
    },
    {
      "epoch": 0.1640976236351959,
      "grad_norm": 0.24095465242862701,
      "learning_rate": 0.0001,
      "loss": 1.5363,
      "step": 1022
    },
    {
      "epoch": 0.1642581888246628,
      "grad_norm": 0.2621472179889679,
      "learning_rate": 0.0001,
      "loss": 1.6758,
      "step": 1023
    },
    {
      "epoch": 0.16441875401412973,
      "grad_norm": 0.26015326380729675,
      "learning_rate": 0.0001,
      "loss": 1.6516,
      "step": 1024
    },
    {
      "epoch": 0.16457931920359667,
      "grad_norm": 0.24528542160987854,
      "learning_rate": 0.0001,
      "loss": 1.6122,
      "step": 1025
    },
    {
      "epoch": 0.16473988439306358,
      "grad_norm": 0.25606489181518555,
      "learning_rate": 0.0001,
      "loss": 1.6446,
      "step": 1026
    },
    {
      "epoch": 0.1649004495825305,
      "grad_norm": 0.2509962320327759,
      "learning_rate": 0.0001,
      "loss": 1.6436,
      "step": 1027
    },
    {
      "epoch": 0.16506101477199744,
      "grad_norm": 0.26206785440444946,
      "learning_rate": 0.0001,
      "loss": 1.6145,
      "step": 1028
    },
    {
      "epoch": 0.16522157996146436,
      "grad_norm": 0.2522035539150238,
      "learning_rate": 0.0001,
      "loss": 1.6218,
      "step": 1029
    },
    {
      "epoch": 0.16538214515093128,
      "grad_norm": 1.347176432609558,
      "learning_rate": 0.0001,
      "loss": 1.6483,
      "step": 1030
    },
    {
      "epoch": 0.1655427103403982,
      "grad_norm": 0.2490011751651764,
      "learning_rate": 0.0001,
      "loss": 1.612,
      "step": 1031
    },
    {
      "epoch": 0.16570327552986513,
      "grad_norm": 0.26856088638305664,
      "learning_rate": 0.0001,
      "loss": 1.6311,
      "step": 1032
    },
    {
      "epoch": 0.16586384071933205,
      "grad_norm": 0.2517376244068146,
      "learning_rate": 0.0001,
      "loss": 1.6378,
      "step": 1033
    },
    {
      "epoch": 0.16602440590879897,
      "grad_norm": 0.25993338227272034,
      "learning_rate": 0.0001,
      "loss": 1.5903,
      "step": 1034
    },
    {
      "epoch": 0.16618497109826588,
      "grad_norm": 0.2448819875717163,
      "learning_rate": 0.0001,
      "loss": 1.5829,
      "step": 1035
    },
    {
      "epoch": 0.16634553628773283,
      "grad_norm": 0.24984216690063477,
      "learning_rate": 0.0001,
      "loss": 1.5087,
      "step": 1036
    },
    {
      "epoch": 0.16650610147719974,
      "grad_norm": 0.2472984492778778,
      "learning_rate": 0.0001,
      "loss": 1.6002,
      "step": 1037
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.2638058364391327,
      "learning_rate": 0.0001,
      "loss": 1.4944,
      "step": 1038
    },
    {
      "epoch": 0.1668272318561336,
      "grad_norm": 0.2725336253643036,
      "learning_rate": 0.0001,
      "loss": 1.6597,
      "step": 1039
    },
    {
      "epoch": 0.16698779704560052,
      "grad_norm": 0.24750974774360657,
      "learning_rate": 0.0001,
      "loss": 1.4934,
      "step": 1040
    },
    {
      "epoch": 0.16714836223506743,
      "grad_norm": 0.2630518972873688,
      "learning_rate": 0.0001,
      "loss": 1.5621,
      "step": 1041
    },
    {
      "epoch": 0.16730892742453435,
      "grad_norm": 0.2586844861507416,
      "learning_rate": 0.0001,
      "loss": 1.6803,
      "step": 1042
    },
    {
      "epoch": 0.1674694926140013,
      "grad_norm": 0.2508007287979126,
      "learning_rate": 0.0001,
      "loss": 1.5383,
      "step": 1043
    },
    {
      "epoch": 0.1676300578034682,
      "grad_norm": 0.252737432718277,
      "learning_rate": 0.0001,
      "loss": 1.5275,
      "step": 1044
    },
    {
      "epoch": 0.16779062299293512,
      "grad_norm": 0.2449912428855896,
      "learning_rate": 0.0001,
      "loss": 1.5562,
      "step": 1045
    },
    {
      "epoch": 0.16795118818240207,
      "grad_norm": 0.27893170714378357,
      "learning_rate": 0.0001,
      "loss": 1.5714,
      "step": 1046
    },
    {
      "epoch": 0.16811175337186898,
      "grad_norm": 0.2636607587337494,
      "learning_rate": 0.0001,
      "loss": 1.6262,
      "step": 1047
    },
    {
      "epoch": 0.1682723185613359,
      "grad_norm": 0.2528233826160431,
      "learning_rate": 0.0001,
      "loss": 1.5882,
      "step": 1048
    },
    {
      "epoch": 0.16843288375080281,
      "grad_norm": 0.2734587788581848,
      "learning_rate": 0.0001,
      "loss": 1.5933,
      "step": 1049
    },
    {
      "epoch": 0.16859344894026976,
      "grad_norm": 0.26790207624435425,
      "learning_rate": 0.0001,
      "loss": 1.5537,
      "step": 1050
    },
    {
      "epoch": 0.16875401412973667,
      "grad_norm": 0.27698516845703125,
      "learning_rate": 0.0001,
      "loss": 1.6398,
      "step": 1051
    },
    {
      "epoch": 0.1689145793192036,
      "grad_norm": 0.26569440960884094,
      "learning_rate": 0.0001,
      "loss": 1.6539,
      "step": 1052
    },
    {
      "epoch": 0.16907514450867053,
      "grad_norm": 0.24960124492645264,
      "learning_rate": 0.0001,
      "loss": 1.5665,
      "step": 1053
    },
    {
      "epoch": 0.16923570969813745,
      "grad_norm": 0.26630985736846924,
      "learning_rate": 0.0001,
      "loss": 1.5947,
      "step": 1054
    },
    {
      "epoch": 0.16939627488760436,
      "grad_norm": 0.24836736917495728,
      "learning_rate": 0.0001,
      "loss": 1.5871,
      "step": 1055
    },
    {
      "epoch": 0.16955684007707128,
      "grad_norm": 0.2526874244213104,
      "learning_rate": 0.0001,
      "loss": 1.5957,
      "step": 1056
    },
    {
      "epoch": 0.16971740526653822,
      "grad_norm": 0.25865864753723145,
      "learning_rate": 0.0001,
      "loss": 1.5824,
      "step": 1057
    },
    {
      "epoch": 0.16987797045600514,
      "grad_norm": 0.24482880532741547,
      "learning_rate": 0.0001,
      "loss": 1.5549,
      "step": 1058
    },
    {
      "epoch": 0.17003853564547206,
      "grad_norm": 0.3854556381702423,
      "learning_rate": 0.0001,
      "loss": 1.6224,
      "step": 1059
    },
    {
      "epoch": 0.170199100834939,
      "grad_norm": 0.25107231736183167,
      "learning_rate": 0.0001,
      "loss": 1.6952,
      "step": 1060
    },
    {
      "epoch": 0.17035966602440591,
      "grad_norm": 0.2519952654838562,
      "learning_rate": 0.0001,
      "loss": 1.5618,
      "step": 1061
    },
    {
      "epoch": 0.17052023121387283,
      "grad_norm": 0.2582648694515228,
      "learning_rate": 0.0001,
      "loss": 1.6372,
      "step": 1062
    },
    {
      "epoch": 0.17068079640333975,
      "grad_norm": 0.2552756071090698,
      "learning_rate": 0.0001,
      "loss": 1.6483,
      "step": 1063
    },
    {
      "epoch": 0.1708413615928067,
      "grad_norm": 0.261065274477005,
      "learning_rate": 0.0001,
      "loss": 1.5369,
      "step": 1064
    },
    {
      "epoch": 0.1710019267822736,
      "grad_norm": 0.26455509662628174,
      "learning_rate": 0.0001,
      "loss": 1.6229,
      "step": 1065
    },
    {
      "epoch": 0.17116249197174052,
      "grad_norm": 0.26041489839553833,
      "learning_rate": 0.0001,
      "loss": 1.642,
      "step": 1066
    },
    {
      "epoch": 0.17132305716120744,
      "grad_norm": 0.2590771019458771,
      "learning_rate": 0.0001,
      "loss": 1.5824,
      "step": 1067
    },
    {
      "epoch": 0.17148362235067438,
      "grad_norm": 0.32040154933929443,
      "learning_rate": 0.0001,
      "loss": 1.5452,
      "step": 1068
    },
    {
      "epoch": 0.1716441875401413,
      "grad_norm": 44.333099365234375,
      "learning_rate": 0.0001,
      "loss": 1.575,
      "step": 1069
    },
    {
      "epoch": 0.1718047527296082,
      "grad_norm": 0.24626383185386658,
      "learning_rate": 0.0001,
      "loss": 1.5173,
      "step": 1070
    },
    {
      "epoch": 0.17196531791907516,
      "grad_norm": 0.2662312686443329,
      "learning_rate": 0.0001,
      "loss": 1.6004,
      "step": 1071
    },
    {
      "epoch": 0.17212588310854207,
      "grad_norm": 0.31177830696105957,
      "learning_rate": 0.0001,
      "loss": 1.6923,
      "step": 1072
    },
    {
      "epoch": 0.172286448298009,
      "grad_norm": 0.2966936230659485,
      "learning_rate": 0.0001,
      "loss": 1.5856,
      "step": 1073
    },
    {
      "epoch": 0.1724470134874759,
      "grad_norm": 0.2461671680212021,
      "learning_rate": 0.0001,
      "loss": 1.553,
      "step": 1074
    },
    {
      "epoch": 0.17260757867694285,
      "grad_norm": 0.24973897635936737,
      "learning_rate": 0.0001,
      "loss": 1.6427,
      "step": 1075
    },
    {
      "epoch": 0.17276814386640976,
      "grad_norm": 1.4923425912857056,
      "learning_rate": 0.0001,
      "loss": 1.627,
      "step": 1076
    },
    {
      "epoch": 0.17292870905587668,
      "grad_norm": 0.2629109025001526,
      "learning_rate": 0.0001,
      "loss": 1.5898,
      "step": 1077
    },
    {
      "epoch": 0.17308927424534362,
      "grad_norm": 0.2483101487159729,
      "learning_rate": 0.0001,
      "loss": 1.5575,
      "step": 1078
    },
    {
      "epoch": 0.17324983943481054,
      "grad_norm": 0.2540942430496216,
      "learning_rate": 0.0001,
      "loss": 1.5033,
      "step": 1079
    },
    {
      "epoch": 0.17341040462427745,
      "grad_norm": 0.2435886114835739,
      "learning_rate": 0.0001,
      "loss": 1.5284,
      "step": 1080
    },
    {
      "epoch": 0.17357096981374437,
      "grad_norm": 0.2666841447353363,
      "learning_rate": 0.0001,
      "loss": 1.6102,
      "step": 1081
    },
    {
      "epoch": 0.1737315350032113,
      "grad_norm": 0.24547235667705536,
      "learning_rate": 0.0001,
      "loss": 1.5151,
      "step": 1082
    },
    {
      "epoch": 0.17389210019267823,
      "grad_norm": 0.6907444000244141,
      "learning_rate": 0.0001,
      "loss": 1.5529,
      "step": 1083
    },
    {
      "epoch": 0.17405266538214514,
      "grad_norm": 0.8070302605628967,
      "learning_rate": 0.0001,
      "loss": 1.6539,
      "step": 1084
    },
    {
      "epoch": 0.1742132305716121,
      "grad_norm": 0.25271305441856384,
      "learning_rate": 0.0001,
      "loss": 1.6012,
      "step": 1085
    },
    {
      "epoch": 0.174373795761079,
      "grad_norm": 0.2487424910068512,
      "learning_rate": 0.0001,
      "loss": 1.4805,
      "step": 1086
    },
    {
      "epoch": 0.17453436095054592,
      "grad_norm": 0.2643222510814667,
      "learning_rate": 0.0001,
      "loss": 1.5592,
      "step": 1087
    },
    {
      "epoch": 0.17469492614001284,
      "grad_norm": 0.2576553225517273,
      "learning_rate": 0.0001,
      "loss": 1.6074,
      "step": 1088
    },
    {
      "epoch": 0.17485549132947978,
      "grad_norm": 0.2615315318107605,
      "learning_rate": 0.0001,
      "loss": 1.5465,
      "step": 1089
    },
    {
      "epoch": 0.1750160565189467,
      "grad_norm": 0.2399469017982483,
      "learning_rate": 0.0001,
      "loss": 1.5739,
      "step": 1090
    },
    {
      "epoch": 0.1751766217084136,
      "grad_norm": 0.2683163285255432,
      "learning_rate": 0.0001,
      "loss": 1.6383,
      "step": 1091
    },
    {
      "epoch": 0.17533718689788053,
      "grad_norm": 0.2417677938938141,
      "learning_rate": 0.0001,
      "loss": 1.4673,
      "step": 1092
    },
    {
      "epoch": 0.17549775208734747,
      "grad_norm": 0.2879345118999481,
      "learning_rate": 0.0001,
      "loss": 1.6886,
      "step": 1093
    },
    {
      "epoch": 0.17565831727681439,
      "grad_norm": 0.2529812157154083,
      "learning_rate": 0.0001,
      "loss": 1.5832,
      "step": 1094
    },
    {
      "epoch": 0.1758188824662813,
      "grad_norm": 0.24131765961647034,
      "learning_rate": 0.0001,
      "loss": 1.4863,
      "step": 1095
    },
    {
      "epoch": 0.17597944765574824,
      "grad_norm": 0.25778815150260925,
      "learning_rate": 0.0001,
      "loss": 1.5694,
      "step": 1096
    },
    {
      "epoch": 0.17614001284521516,
      "grad_norm": 0.24608175456523895,
      "learning_rate": 0.0001,
      "loss": 1.6426,
      "step": 1097
    },
    {
      "epoch": 0.17630057803468208,
      "grad_norm": 0.2669673562049866,
      "learning_rate": 0.0001,
      "loss": 1.621,
      "step": 1098
    },
    {
      "epoch": 0.176461143224149,
      "grad_norm": 0.25561201572418213,
      "learning_rate": 0.0001,
      "loss": 1.5283,
      "step": 1099
    },
    {
      "epoch": 0.17662170841361594,
      "grad_norm": 0.25698336958885193,
      "learning_rate": 0.0001,
      "loss": 1.6021,
      "step": 1100
    },
    {
      "epoch": 0.17678227360308285,
      "grad_norm": 0.2444210797548294,
      "learning_rate": 0.0001,
      "loss": 1.6462,
      "step": 1101
    },
    {
      "epoch": 0.17694283879254977,
      "grad_norm": 0.26224303245544434,
      "learning_rate": 0.0001,
      "loss": 1.582,
      "step": 1102
    },
    {
      "epoch": 0.1771034039820167,
      "grad_norm": 0.25329750776290894,
      "learning_rate": 0.0001,
      "loss": 1.5771,
      "step": 1103
    },
    {
      "epoch": 0.17726396917148363,
      "grad_norm": 0.2472059726715088,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 1104
    },
    {
      "epoch": 0.17742453436095054,
      "grad_norm": 0.25361812114715576,
      "learning_rate": 0.0001,
      "loss": 1.5983,
      "step": 1105
    },
    {
      "epoch": 0.17758509955041746,
      "grad_norm": 0.25120607018470764,
      "learning_rate": 0.0001,
      "loss": 1.5643,
      "step": 1106
    },
    {
      "epoch": 0.1777456647398844,
      "grad_norm": 0.2414521723985672,
      "learning_rate": 0.0001,
      "loss": 1.539,
      "step": 1107
    },
    {
      "epoch": 0.17790622992935132,
      "grad_norm": 0.25235995650291443,
      "learning_rate": 0.0001,
      "loss": 1.6059,
      "step": 1108
    },
    {
      "epoch": 0.17806679511881823,
      "grad_norm": 0.252722829580307,
      "learning_rate": 0.0001,
      "loss": 1.6828,
      "step": 1109
    },
    {
      "epoch": 0.17822736030828518,
      "grad_norm": 0.25269708037376404,
      "learning_rate": 0.0001,
      "loss": 1.4884,
      "step": 1110
    },
    {
      "epoch": 0.1783879254977521,
      "grad_norm": 0.25341561436653137,
      "learning_rate": 0.0001,
      "loss": 1.5377,
      "step": 1111
    },
    {
      "epoch": 0.178548490687219,
      "grad_norm": 0.2385052740573883,
      "learning_rate": 0.0001,
      "loss": 1.52,
      "step": 1112
    },
    {
      "epoch": 0.17870905587668592,
      "grad_norm": 0.255615234375,
      "learning_rate": 0.0001,
      "loss": 1.5391,
      "step": 1113
    },
    {
      "epoch": 0.17886962106615287,
      "grad_norm": 0.255342960357666,
      "learning_rate": 0.0001,
      "loss": 1.5574,
      "step": 1114
    },
    {
      "epoch": 0.17903018625561978,
      "grad_norm": 0.2434520423412323,
      "learning_rate": 0.0001,
      "loss": 1.5761,
      "step": 1115
    },
    {
      "epoch": 0.1791907514450867,
      "grad_norm": 0.2581034302711487,
      "learning_rate": 0.0001,
      "loss": 1.5969,
      "step": 1116
    },
    {
      "epoch": 0.17935131663455364,
      "grad_norm": 0.2473067194223404,
      "learning_rate": 0.0001,
      "loss": 1.561,
      "step": 1117
    },
    {
      "epoch": 0.17951188182402056,
      "grad_norm": 0.24469049274921417,
      "learning_rate": 0.0001,
      "loss": 1.5901,
      "step": 1118
    },
    {
      "epoch": 0.17967244701348747,
      "grad_norm": 0.25427860021591187,
      "learning_rate": 0.0001,
      "loss": 1.5611,
      "step": 1119
    },
    {
      "epoch": 0.1798330122029544,
      "grad_norm": 0.24965278804302216,
      "learning_rate": 0.0001,
      "loss": 1.5981,
      "step": 1120
    },
    {
      "epoch": 0.17999357739242133,
      "grad_norm": 0.24722306430339813,
      "learning_rate": 0.0001,
      "loss": 1.5614,
      "step": 1121
    },
    {
      "epoch": 0.18015414258188825,
      "grad_norm": 0.2540864050388336,
      "learning_rate": 0.0001,
      "loss": 1.5544,
      "step": 1122
    },
    {
      "epoch": 0.18031470777135516,
      "grad_norm": 0.2447497397661209,
      "learning_rate": 0.0001,
      "loss": 1.5944,
      "step": 1123
    },
    {
      "epoch": 0.18047527296082208,
      "grad_norm": 0.2738570272922516,
      "learning_rate": 0.0001,
      "loss": 1.6893,
      "step": 1124
    },
    {
      "epoch": 0.18063583815028902,
      "grad_norm": 0.2362639307975769,
      "learning_rate": 0.0001,
      "loss": 1.5471,
      "step": 1125
    },
    {
      "epoch": 0.18079640333975594,
      "grad_norm": 0.2534329295158386,
      "learning_rate": 0.0001,
      "loss": 1.616,
      "step": 1126
    },
    {
      "epoch": 0.18095696852922286,
      "grad_norm": 0.2539317011833191,
      "learning_rate": 0.0001,
      "loss": 1.4954,
      "step": 1127
    },
    {
      "epoch": 0.1811175337186898,
      "grad_norm": 0.24718168377876282,
      "learning_rate": 0.0001,
      "loss": 1.6153,
      "step": 1128
    },
    {
      "epoch": 0.18127809890815672,
      "grad_norm": 0.2623962461948395,
      "learning_rate": 0.0001,
      "loss": 1.6629,
      "step": 1129
    },
    {
      "epoch": 0.18143866409762363,
      "grad_norm": 0.26475292444229126,
      "learning_rate": 0.0001,
      "loss": 1.69,
      "step": 1130
    },
    {
      "epoch": 0.18159922928709055,
      "grad_norm": 0.276611864566803,
      "learning_rate": 0.0001,
      "loss": 1.6318,
      "step": 1131
    },
    {
      "epoch": 0.1817597944765575,
      "grad_norm": 0.24636362493038177,
      "learning_rate": 0.0001,
      "loss": 1.5763,
      "step": 1132
    },
    {
      "epoch": 0.1819203596660244,
      "grad_norm": 0.2588915526866913,
      "learning_rate": 0.0001,
      "loss": 1.6406,
      "step": 1133
    },
    {
      "epoch": 0.18208092485549132,
      "grad_norm": 0.2608543038368225,
      "learning_rate": 0.0001,
      "loss": 1.6187,
      "step": 1134
    },
    {
      "epoch": 0.18224149004495827,
      "grad_norm": 0.26090532541275024,
      "learning_rate": 0.0001,
      "loss": 1.5682,
      "step": 1135
    },
    {
      "epoch": 0.18240205523442518,
      "grad_norm": 0.25060582160949707,
      "learning_rate": 0.0001,
      "loss": 1.59,
      "step": 1136
    },
    {
      "epoch": 0.1825626204238921,
      "grad_norm": 0.2568562626838684,
      "learning_rate": 0.0001,
      "loss": 1.6352,
      "step": 1137
    },
    {
      "epoch": 0.182723185613359,
      "grad_norm": 0.27537429332733154,
      "learning_rate": 0.0001,
      "loss": 1.6926,
      "step": 1138
    },
    {
      "epoch": 0.18288375080282596,
      "grad_norm": 0.25080350041389465,
      "learning_rate": 0.0001,
      "loss": 1.614,
      "step": 1139
    },
    {
      "epoch": 0.18304431599229287,
      "grad_norm": 0.24731677770614624,
      "learning_rate": 0.0001,
      "loss": 1.633,
      "step": 1140
    },
    {
      "epoch": 0.1832048811817598,
      "grad_norm": 0.25230592489242554,
      "learning_rate": 0.0001,
      "loss": 1.5854,
      "step": 1141
    },
    {
      "epoch": 0.18336544637122673,
      "grad_norm": 0.25337666273117065,
      "learning_rate": 0.0001,
      "loss": 1.6147,
      "step": 1142
    },
    {
      "epoch": 0.18352601156069365,
      "grad_norm": 0.24595242738723755,
      "learning_rate": 0.0001,
      "loss": 1.5107,
      "step": 1143
    },
    {
      "epoch": 0.18368657675016056,
      "grad_norm": 0.2588496804237366,
      "learning_rate": 0.0001,
      "loss": 1.5937,
      "step": 1144
    },
    {
      "epoch": 0.18384714193962748,
      "grad_norm": 0.2558797299861908,
      "learning_rate": 0.0001,
      "loss": 1.5961,
      "step": 1145
    },
    {
      "epoch": 0.18400770712909442,
      "grad_norm": 0.25319406390190125,
      "learning_rate": 0.0001,
      "loss": 1.663,
      "step": 1146
    },
    {
      "epoch": 0.18416827231856134,
      "grad_norm": 0.2560921907424927,
      "learning_rate": 0.0001,
      "loss": 1.5537,
      "step": 1147
    },
    {
      "epoch": 0.18432883750802825,
      "grad_norm": 0.25958332419395447,
      "learning_rate": 0.0001,
      "loss": 1.6162,
      "step": 1148
    },
    {
      "epoch": 0.18448940269749517,
      "grad_norm": 0.24979592859745026,
      "learning_rate": 0.0001,
      "loss": 1.5961,
      "step": 1149
    },
    {
      "epoch": 0.1846499678869621,
      "grad_norm": 0.26627790927886963,
      "learning_rate": 0.0001,
      "loss": 1.4225,
      "step": 1150
    },
    {
      "epoch": 0.18481053307642903,
      "grad_norm": 0.25156256556510925,
      "learning_rate": 0.0001,
      "loss": 1.5706,
      "step": 1151
    },
    {
      "epoch": 0.18497109826589594,
      "grad_norm": 0.2600122392177582,
      "learning_rate": 0.0001,
      "loss": 1.5701,
      "step": 1152
    },
    {
      "epoch": 0.1851316634553629,
      "grad_norm": 0.25203970074653625,
      "learning_rate": 0.0001,
      "loss": 1.6395,
      "step": 1153
    },
    {
      "epoch": 0.1852922286448298,
      "grad_norm": 0.2482793778181076,
      "learning_rate": 0.0001,
      "loss": 1.5757,
      "step": 1154
    },
    {
      "epoch": 0.18545279383429672,
      "grad_norm": 0.2505303621292114,
      "learning_rate": 0.0001,
      "loss": 1.5482,
      "step": 1155
    },
    {
      "epoch": 0.18561335902376364,
      "grad_norm": 0.24639439582824707,
      "learning_rate": 0.0001,
      "loss": 1.5351,
      "step": 1156
    },
    {
      "epoch": 0.18577392421323058,
      "grad_norm": 0.2650209367275238,
      "learning_rate": 0.0001,
      "loss": 1.4979,
      "step": 1157
    },
    {
      "epoch": 0.1859344894026975,
      "grad_norm": 0.25111427903175354,
      "learning_rate": 0.0001,
      "loss": 1.5307,
      "step": 1158
    },
    {
      "epoch": 0.1860950545921644,
      "grad_norm": 0.2652316689491272,
      "learning_rate": 0.0001,
      "loss": 1.6746,
      "step": 1159
    },
    {
      "epoch": 0.18625561978163135,
      "grad_norm": 0.24217385053634644,
      "learning_rate": 0.0001,
      "loss": 1.5441,
      "step": 1160
    },
    {
      "epoch": 0.18641618497109827,
      "grad_norm": 0.2778951823711395,
      "learning_rate": 0.0001,
      "loss": 1.6431,
      "step": 1161
    },
    {
      "epoch": 0.18657675016056519,
      "grad_norm": 0.2490694373846054,
      "learning_rate": 0.0001,
      "loss": 1.4861,
      "step": 1162
    },
    {
      "epoch": 0.1867373153500321,
      "grad_norm": 0.2578921616077423,
      "learning_rate": 0.0001,
      "loss": 1.5689,
      "step": 1163
    },
    {
      "epoch": 0.18689788053949905,
      "grad_norm": 0.24337232112884521,
      "learning_rate": 0.0001,
      "loss": 1.5358,
      "step": 1164
    },
    {
      "epoch": 0.18705844572896596,
      "grad_norm": 0.256675660610199,
      "learning_rate": 0.0001,
      "loss": 1.5463,
      "step": 1165
    },
    {
      "epoch": 0.18721901091843288,
      "grad_norm": 0.26293712854385376,
      "learning_rate": 0.0001,
      "loss": 1.6198,
      "step": 1166
    },
    {
      "epoch": 0.18737957610789982,
      "grad_norm": 0.25028562545776367,
      "learning_rate": 0.0001,
      "loss": 1.525,
      "step": 1167
    },
    {
      "epoch": 0.18754014129736674,
      "grad_norm": 0.240972101688385,
      "learning_rate": 0.0001,
      "loss": 1.4788,
      "step": 1168
    },
    {
      "epoch": 0.18770070648683365,
      "grad_norm": 0.26054659485816956,
      "learning_rate": 0.0001,
      "loss": 1.6252,
      "step": 1169
    },
    {
      "epoch": 0.18786127167630057,
      "grad_norm": 0.263796865940094,
      "learning_rate": 0.0001,
      "loss": 1.5822,
      "step": 1170
    },
    {
      "epoch": 0.1880218368657675,
      "grad_norm": 0.2466707080602646,
      "learning_rate": 0.0001,
      "loss": 1.5136,
      "step": 1171
    },
    {
      "epoch": 0.18818240205523443,
      "grad_norm": 0.2484586089849472,
      "learning_rate": 0.0001,
      "loss": 1.6043,
      "step": 1172
    },
    {
      "epoch": 0.18834296724470134,
      "grad_norm": 0.253840833902359,
      "learning_rate": 0.0001,
      "loss": 1.6295,
      "step": 1173
    },
    {
      "epoch": 0.18850353243416829,
      "grad_norm": 0.2369823455810547,
      "learning_rate": 0.0001,
      "loss": 1.4692,
      "step": 1174
    },
    {
      "epoch": 0.1886640976236352,
      "grad_norm": 0.24843278527259827,
      "learning_rate": 0.0001,
      "loss": 1.575,
      "step": 1175
    },
    {
      "epoch": 0.18882466281310212,
      "grad_norm": 0.2453308403491974,
      "learning_rate": 0.0001,
      "loss": 1.5405,
      "step": 1176
    },
    {
      "epoch": 0.18898522800256903,
      "grad_norm": 0.25337234139442444,
      "learning_rate": 0.0001,
      "loss": 1.5851,
      "step": 1177
    },
    {
      "epoch": 0.18914579319203598,
      "grad_norm": 0.2726234197616577,
      "learning_rate": 0.0001,
      "loss": 1.6019,
      "step": 1178
    },
    {
      "epoch": 0.1893063583815029,
      "grad_norm": 0.24739961326122284,
      "learning_rate": 0.0001,
      "loss": 1.5996,
      "step": 1179
    },
    {
      "epoch": 0.1894669235709698,
      "grad_norm": 0.26011112332344055,
      "learning_rate": 0.0001,
      "loss": 1.5776,
      "step": 1180
    },
    {
      "epoch": 0.18962748876043672,
      "grad_norm": 0.267156183719635,
      "learning_rate": 0.0001,
      "loss": 1.622,
      "step": 1181
    },
    {
      "epoch": 0.18978805394990367,
      "grad_norm": 0.24337074160575867,
      "learning_rate": 0.0001,
      "loss": 1.5286,
      "step": 1182
    },
    {
      "epoch": 0.18994861913937058,
      "grad_norm": 0.2607194781303406,
      "learning_rate": 0.0001,
      "loss": 1.589,
      "step": 1183
    },
    {
      "epoch": 0.1901091843288375,
      "grad_norm": 0.25958162546157837,
      "learning_rate": 0.0001,
      "loss": 1.6048,
      "step": 1184
    },
    {
      "epoch": 0.19026974951830444,
      "grad_norm": 0.25642454624176025,
      "learning_rate": 0.0001,
      "loss": 1.618,
      "step": 1185
    },
    {
      "epoch": 0.19043031470777136,
      "grad_norm": 0.25708210468292236,
      "learning_rate": 0.0001,
      "loss": 1.6579,
      "step": 1186
    },
    {
      "epoch": 0.19059087989723827,
      "grad_norm": 0.3139567971229553,
      "learning_rate": 0.0001,
      "loss": 1.6066,
      "step": 1187
    },
    {
      "epoch": 0.1907514450867052,
      "grad_norm": 0.26446476578712463,
      "learning_rate": 0.0001,
      "loss": 1.6322,
      "step": 1188
    },
    {
      "epoch": 0.19091201027617213,
      "grad_norm": 0.25026997923851013,
      "learning_rate": 0.0001,
      "loss": 1.601,
      "step": 1189
    },
    {
      "epoch": 0.19107257546563905,
      "grad_norm": 0.2574523091316223,
      "learning_rate": 0.0001,
      "loss": 1.5596,
      "step": 1190
    },
    {
      "epoch": 0.19123314065510597,
      "grad_norm": 0.2602752149105072,
      "learning_rate": 0.0001,
      "loss": 1.5271,
      "step": 1191
    },
    {
      "epoch": 0.1913937058445729,
      "grad_norm": 0.24762839078903198,
      "learning_rate": 0.0001,
      "loss": 1.593,
      "step": 1192
    },
    {
      "epoch": 0.19155427103403982,
      "grad_norm": 0.2613926827907562,
      "learning_rate": 0.0001,
      "loss": 1.6103,
      "step": 1193
    },
    {
      "epoch": 0.19171483622350674,
      "grad_norm": 0.2612307667732239,
      "learning_rate": 0.0001,
      "loss": 1.6969,
      "step": 1194
    },
    {
      "epoch": 0.19187540141297366,
      "grad_norm": 0.25217312574386597,
      "learning_rate": 0.0001,
      "loss": 1.5758,
      "step": 1195
    },
    {
      "epoch": 0.1920359666024406,
      "grad_norm": 0.244678795337677,
      "learning_rate": 0.0001,
      "loss": 1.5781,
      "step": 1196
    },
    {
      "epoch": 0.19219653179190752,
      "grad_norm": 0.23493874073028564,
      "learning_rate": 0.0001,
      "loss": 1.5919,
      "step": 1197
    },
    {
      "epoch": 0.19235709698137443,
      "grad_norm": 0.2504349648952484,
      "learning_rate": 0.0001,
      "loss": 1.5639,
      "step": 1198
    },
    {
      "epoch": 0.19251766217084137,
      "grad_norm": 0.24955222010612488,
      "learning_rate": 0.0001,
      "loss": 1.5091,
      "step": 1199
    },
    {
      "epoch": 0.1926782273603083,
      "grad_norm": 0.23621833324432373,
      "learning_rate": 0.0001,
      "loss": 1.5365,
      "step": 1200
    },
    {
      "epoch": 0.1928387925497752,
      "grad_norm": 0.24546657502651215,
      "learning_rate": 0.0001,
      "loss": 1.5253,
      "step": 1201
    },
    {
      "epoch": 0.19299935773924212,
      "grad_norm": 0.24631303548812866,
      "learning_rate": 0.0001,
      "loss": 1.5587,
      "step": 1202
    },
    {
      "epoch": 0.19315992292870907,
      "grad_norm": 0.27041104435920715,
      "learning_rate": 0.0001,
      "loss": 1.7019,
      "step": 1203
    },
    {
      "epoch": 0.19332048811817598,
      "grad_norm": 0.2491196244955063,
      "learning_rate": 0.0001,
      "loss": 1.5807,
      "step": 1204
    },
    {
      "epoch": 0.1934810533076429,
      "grad_norm": 0.2602118253707886,
      "learning_rate": 0.0001,
      "loss": 1.5792,
      "step": 1205
    },
    {
      "epoch": 0.1936416184971098,
      "grad_norm": 0.26140204071998596,
      "learning_rate": 0.0001,
      "loss": 1.5863,
      "step": 1206
    },
    {
      "epoch": 0.19380218368657676,
      "grad_norm": 0.25287631154060364,
      "learning_rate": 0.0001,
      "loss": 1.6427,
      "step": 1207
    },
    {
      "epoch": 0.19396274887604367,
      "grad_norm": 0.24496833980083466,
      "learning_rate": 0.0001,
      "loss": 1.5007,
      "step": 1208
    },
    {
      "epoch": 0.1941233140655106,
      "grad_norm": 0.25292402505874634,
      "learning_rate": 0.0001,
      "loss": 1.5218,
      "step": 1209
    },
    {
      "epoch": 0.19428387925497753,
      "grad_norm": 0.2573296129703522,
      "learning_rate": 0.0001,
      "loss": 1.6401,
      "step": 1210
    },
    {
      "epoch": 0.19444444444444445,
      "grad_norm": 0.2588058114051819,
      "learning_rate": 0.0001,
      "loss": 1.6037,
      "step": 1211
    },
    {
      "epoch": 0.19460500963391136,
      "grad_norm": 0.2602480947971344,
      "learning_rate": 0.0001,
      "loss": 1.4993,
      "step": 1212
    },
    {
      "epoch": 0.19476557482337828,
      "grad_norm": 0.24529221653938293,
      "learning_rate": 0.0001,
      "loss": 1.606,
      "step": 1213
    },
    {
      "epoch": 0.19492614001284522,
      "grad_norm": 0.2493453323841095,
      "learning_rate": 0.0001,
      "loss": 1.581,
      "step": 1214
    },
    {
      "epoch": 0.19508670520231214,
      "grad_norm": 0.25749775767326355,
      "learning_rate": 0.0001,
      "loss": 1.648,
      "step": 1215
    },
    {
      "epoch": 0.19524727039177905,
      "grad_norm": 0.24454443156719208,
      "learning_rate": 0.0001,
      "loss": 1.5418,
      "step": 1216
    },
    {
      "epoch": 0.195407835581246,
      "grad_norm": 0.25549542903900146,
      "learning_rate": 0.0001,
      "loss": 1.5627,
      "step": 1217
    },
    {
      "epoch": 0.1955684007707129,
      "grad_norm": 0.25463199615478516,
      "learning_rate": 0.0001,
      "loss": 1.6265,
      "step": 1218
    },
    {
      "epoch": 0.19572896596017983,
      "grad_norm": 0.24837295711040497,
      "learning_rate": 0.0001,
      "loss": 1.515,
      "step": 1219
    },
    {
      "epoch": 0.19588953114964675,
      "grad_norm": 0.24428117275238037,
      "learning_rate": 0.0001,
      "loss": 1.6023,
      "step": 1220
    },
    {
      "epoch": 0.1960500963391137,
      "grad_norm": 0.24188461899757385,
      "learning_rate": 0.0001,
      "loss": 1.5239,
      "step": 1221
    },
    {
      "epoch": 0.1962106615285806,
      "grad_norm": 0.28273800015449524,
      "learning_rate": 0.0001,
      "loss": 1.628,
      "step": 1222
    },
    {
      "epoch": 0.19637122671804752,
      "grad_norm": 0.24625249207019806,
      "learning_rate": 0.0001,
      "loss": 1.5282,
      "step": 1223
    },
    {
      "epoch": 0.19653179190751446,
      "grad_norm": 0.2541029751300812,
      "learning_rate": 0.0001,
      "loss": 1.5845,
      "step": 1224
    },
    {
      "epoch": 0.19669235709698138,
      "grad_norm": 0.23067116737365723,
      "learning_rate": 0.0001,
      "loss": 1.4668,
      "step": 1225
    },
    {
      "epoch": 0.1968529222864483,
      "grad_norm": 0.25950169563293457,
      "learning_rate": 0.0001,
      "loss": 1.7031,
      "step": 1226
    },
    {
      "epoch": 0.1970134874759152,
      "grad_norm": 0.266619473695755,
      "learning_rate": 0.0001,
      "loss": 1.6598,
      "step": 1227
    },
    {
      "epoch": 0.19717405266538215,
      "grad_norm": 0.2611779272556305,
      "learning_rate": 0.0001,
      "loss": 1.4975,
      "step": 1228
    },
    {
      "epoch": 0.19733461785484907,
      "grad_norm": 0.24724836647510529,
      "learning_rate": 0.0001,
      "loss": 1.5402,
      "step": 1229
    },
    {
      "epoch": 0.197495183044316,
      "grad_norm": 0.2501607835292816,
      "learning_rate": 0.0001,
      "loss": 1.5692,
      "step": 1230
    },
    {
      "epoch": 0.19765574823378293,
      "grad_norm": 0.25537505745887756,
      "learning_rate": 0.0001,
      "loss": 1.5311,
      "step": 1231
    },
    {
      "epoch": 0.19781631342324985,
      "grad_norm": 0.265559583902359,
      "learning_rate": 0.0001,
      "loss": 1.5564,
      "step": 1232
    },
    {
      "epoch": 0.19797687861271676,
      "grad_norm": 0.25961583852767944,
      "learning_rate": 0.0001,
      "loss": 1.5156,
      "step": 1233
    },
    {
      "epoch": 0.19813744380218368,
      "grad_norm": 0.2424548864364624,
      "learning_rate": 0.0001,
      "loss": 1.4394,
      "step": 1234
    },
    {
      "epoch": 0.19829800899165062,
      "grad_norm": 0.25589460134506226,
      "learning_rate": 0.0001,
      "loss": 1.5529,
      "step": 1235
    },
    {
      "epoch": 0.19845857418111754,
      "grad_norm": 0.41593801975250244,
      "learning_rate": 0.0001,
      "loss": 1.6208,
      "step": 1236
    },
    {
      "epoch": 0.19861913937058445,
      "grad_norm": 0.2542087733745575,
      "learning_rate": 0.0001,
      "loss": 1.5614,
      "step": 1237
    },
    {
      "epoch": 0.19877970456005137,
      "grad_norm": 0.2820034623146057,
      "learning_rate": 0.0001,
      "loss": 1.6148,
      "step": 1238
    },
    {
      "epoch": 0.1989402697495183,
      "grad_norm": 0.23934297263622284,
      "learning_rate": 0.0001,
      "loss": 1.5318,
      "step": 1239
    },
    {
      "epoch": 0.19910083493898523,
      "grad_norm": 0.2777564227581024,
      "learning_rate": 0.0001,
      "loss": 1.555,
      "step": 1240
    },
    {
      "epoch": 0.19926140012845214,
      "grad_norm": 0.24646690487861633,
      "learning_rate": 0.0001,
      "loss": 1.5614,
      "step": 1241
    },
    {
      "epoch": 0.1994219653179191,
      "grad_norm": 1.28658127784729,
      "learning_rate": 0.0001,
      "loss": 1.6237,
      "step": 1242
    },
    {
      "epoch": 0.199582530507386,
      "grad_norm": 0.2608278691768646,
      "learning_rate": 0.0001,
      "loss": 1.5288,
      "step": 1243
    },
    {
      "epoch": 0.19974309569685292,
      "grad_norm": 0.25941115617752075,
      "learning_rate": 0.0001,
      "loss": 1.5484,
      "step": 1244
    },
    {
      "epoch": 0.19990366088631983,
      "grad_norm": 0.26302090287208557,
      "learning_rate": 0.0001,
      "loss": 1.6686,
      "step": 1245
    },
    {
      "epoch": 0.20006422607578678,
      "grad_norm": 0.28321418166160583,
      "learning_rate": 0.0001,
      "loss": 1.6189,
      "step": 1246
    },
    {
      "epoch": 0.2002247912652537,
      "grad_norm": 0.2480607032775879,
      "learning_rate": 0.0001,
      "loss": 1.5462,
      "step": 1247
    },
    {
      "epoch": 0.2003853564547206,
      "grad_norm": 0.2750486433506012,
      "learning_rate": 0.0001,
      "loss": 1.5818,
      "step": 1248
    },
    {
      "epoch": 0.20054592164418755,
      "grad_norm": 0.25892406702041626,
      "learning_rate": 0.0001,
      "loss": 1.4851,
      "step": 1249
    },
    {
      "epoch": 0.20070648683365447,
      "grad_norm": 0.2393769770860672,
      "learning_rate": 0.0001,
      "loss": 1.5491,
      "step": 1250
    },
    {
      "epoch": 0.20086705202312138,
      "grad_norm": 0.2535555362701416,
      "learning_rate": 0.0001,
      "loss": 1.5317,
      "step": 1251
    },
    {
      "epoch": 0.2010276172125883,
      "grad_norm": 0.25529196858406067,
      "learning_rate": 0.0001,
      "loss": 1.5794,
      "step": 1252
    },
    {
      "epoch": 0.20118818240205524,
      "grad_norm": 0.25667497515678406,
      "learning_rate": 0.0001,
      "loss": 1.6444,
      "step": 1253
    },
    {
      "epoch": 0.20134874759152216,
      "grad_norm": 0.2479962259531021,
      "learning_rate": 0.0001,
      "loss": 1.5517,
      "step": 1254
    },
    {
      "epoch": 0.20150931278098907,
      "grad_norm": 0.28191882371902466,
      "learning_rate": 0.0001,
      "loss": 1.6213,
      "step": 1255
    },
    {
      "epoch": 0.20166987797045602,
      "grad_norm": 0.27171066403388977,
      "learning_rate": 0.0001,
      "loss": 1.6614,
      "step": 1256
    },
    {
      "epoch": 0.20183044315992293,
      "grad_norm": 0.2494252473115921,
      "learning_rate": 0.0001,
      "loss": 1.5451,
      "step": 1257
    },
    {
      "epoch": 0.20199100834938985,
      "grad_norm": 0.24599884450435638,
      "learning_rate": 0.0001,
      "loss": 1.5817,
      "step": 1258
    },
    {
      "epoch": 0.20215157353885677,
      "grad_norm": 0.2512761950492859,
      "learning_rate": 0.0001,
      "loss": 1.5855,
      "step": 1259
    },
    {
      "epoch": 0.2023121387283237,
      "grad_norm": 0.24004778265953064,
      "learning_rate": 0.0001,
      "loss": 1.5994,
      "step": 1260
    },
    {
      "epoch": 0.20247270391779063,
      "grad_norm": 0.25274306535720825,
      "learning_rate": 0.0001,
      "loss": 1.5803,
      "step": 1261
    },
    {
      "epoch": 0.20263326910725754,
      "grad_norm": 0.2604442238807678,
      "learning_rate": 0.0001,
      "loss": 1.6233,
      "step": 1262
    },
    {
      "epoch": 0.20279383429672446,
      "grad_norm": 0.2535599172115326,
      "learning_rate": 0.0001,
      "loss": 1.6,
      "step": 1263
    },
    {
      "epoch": 0.2029543994861914,
      "grad_norm": 0.27341902256011963,
      "learning_rate": 0.0001,
      "loss": 1.5431,
      "step": 1264
    },
    {
      "epoch": 0.20311496467565832,
      "grad_norm": 0.24905379116535187,
      "learning_rate": 0.0001,
      "loss": 1.5125,
      "step": 1265
    },
    {
      "epoch": 0.20327552986512523,
      "grad_norm": 0.24562866985797882,
      "learning_rate": 0.0001,
      "loss": 1.5394,
      "step": 1266
    },
    {
      "epoch": 0.20343609505459218,
      "grad_norm": 0.26494455337524414,
      "learning_rate": 0.0001,
      "loss": 1.5749,
      "step": 1267
    },
    {
      "epoch": 0.2035966602440591,
      "grad_norm": 0.24077832698822021,
      "learning_rate": 0.0001,
      "loss": 1.6049,
      "step": 1268
    },
    {
      "epoch": 0.203757225433526,
      "grad_norm": 0.2567788362503052,
      "learning_rate": 0.0001,
      "loss": 1.5985,
      "step": 1269
    },
    {
      "epoch": 0.20391779062299292,
      "grad_norm": 0.23835213482379913,
      "learning_rate": 0.0001,
      "loss": 1.5971,
      "step": 1270
    },
    {
      "epoch": 0.20407835581245987,
      "grad_norm": 0.2541971802711487,
      "learning_rate": 0.0001,
      "loss": 1.5794,
      "step": 1271
    },
    {
      "epoch": 0.20423892100192678,
      "grad_norm": 0.24991245567798615,
      "learning_rate": 0.0001,
      "loss": 1.5234,
      "step": 1272
    },
    {
      "epoch": 0.2043994861913937,
      "grad_norm": 0.2575179636478424,
      "learning_rate": 0.0001,
      "loss": 1.6103,
      "step": 1273
    },
    {
      "epoch": 0.20456005138086064,
      "grad_norm": 0.2621893584728241,
      "learning_rate": 0.0001,
      "loss": 1.6919,
      "step": 1274
    },
    {
      "epoch": 0.20472061657032756,
      "grad_norm": 0.2681543827056885,
      "learning_rate": 0.0001,
      "loss": 1.5648,
      "step": 1275
    },
    {
      "epoch": 0.20488118175979447,
      "grad_norm": 0.25278764963150024,
      "learning_rate": 0.0001,
      "loss": 1.5622,
      "step": 1276
    },
    {
      "epoch": 0.2050417469492614,
      "grad_norm": 0.24915283918380737,
      "learning_rate": 0.0001,
      "loss": 1.6034,
      "step": 1277
    },
    {
      "epoch": 0.20520231213872833,
      "grad_norm": 0.2452990710735321,
      "learning_rate": 0.0001,
      "loss": 1.557,
      "step": 1278
    },
    {
      "epoch": 0.20536287732819525,
      "grad_norm": 0.2584128677845001,
      "learning_rate": 0.0001,
      "loss": 1.5505,
      "step": 1279
    },
    {
      "epoch": 0.20552344251766216,
      "grad_norm": 0.2779355049133301,
      "learning_rate": 0.0001,
      "loss": 1.6823,
      "step": 1280
    },
    {
      "epoch": 0.2056840077071291,
      "grad_norm": 0.2508663535118103,
      "learning_rate": 0.0001,
      "loss": 1.5681,
      "step": 1281
    },
    {
      "epoch": 0.20584457289659602,
      "grad_norm": 0.2578074336051941,
      "learning_rate": 0.0001,
      "loss": 1.5981,
      "step": 1282
    },
    {
      "epoch": 0.20600513808606294,
      "grad_norm": 0.2573000192642212,
      "learning_rate": 0.0001,
      "loss": 1.5426,
      "step": 1283
    },
    {
      "epoch": 0.20616570327552985,
      "grad_norm": 0.24457018077373505,
      "learning_rate": 0.0001,
      "loss": 1.5179,
      "step": 1284
    },
    {
      "epoch": 0.2063262684649968,
      "grad_norm": 0.24787230789661407,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 1285
    },
    {
      "epoch": 0.20648683365446371,
      "grad_norm": 0.2664269804954529,
      "learning_rate": 0.0001,
      "loss": 1.6439,
      "step": 1286
    },
    {
      "epoch": 0.20664739884393063,
      "grad_norm": 0.2581881284713745,
      "learning_rate": 0.0001,
      "loss": 1.5825,
      "step": 1287
    },
    {
      "epoch": 0.20680796403339757,
      "grad_norm": 0.3322311043739319,
      "learning_rate": 0.0001,
      "loss": 1.5597,
      "step": 1288
    },
    {
      "epoch": 0.2069685292228645,
      "grad_norm": 0.26221877336502075,
      "learning_rate": 0.0001,
      "loss": 1.6434,
      "step": 1289
    },
    {
      "epoch": 0.2071290944123314,
      "grad_norm": 0.23221048712730408,
      "learning_rate": 0.0001,
      "loss": 1.5598,
      "step": 1290
    },
    {
      "epoch": 0.20728965960179832,
      "grad_norm": 0.2560206353664398,
      "learning_rate": 0.0001,
      "loss": 1.5869,
      "step": 1291
    },
    {
      "epoch": 0.20745022479126526,
      "grad_norm": 0.26476046442985535,
      "learning_rate": 0.0001,
      "loss": 1.5432,
      "step": 1292
    },
    {
      "epoch": 0.20761078998073218,
      "grad_norm": 0.24677133560180664,
      "learning_rate": 0.0001,
      "loss": 1.5761,
      "step": 1293
    },
    {
      "epoch": 0.2077713551701991,
      "grad_norm": 0.2672979533672333,
      "learning_rate": 0.0001,
      "loss": 1.6243,
      "step": 1294
    },
    {
      "epoch": 0.207931920359666,
      "grad_norm": 0.25498655438423157,
      "learning_rate": 0.0001,
      "loss": 1.5831,
      "step": 1295
    },
    {
      "epoch": 0.20809248554913296,
      "grad_norm": 0.27964186668395996,
      "learning_rate": 0.0001,
      "loss": 1.6878,
      "step": 1296
    },
    {
      "epoch": 0.20825305073859987,
      "grad_norm": 0.26787513494491577,
      "learning_rate": 0.0001,
      "loss": 1.601,
      "step": 1297
    },
    {
      "epoch": 0.2084136159280668,
      "grad_norm": 0.2635231912136078,
      "learning_rate": 0.0001,
      "loss": 1.6095,
      "step": 1298
    },
    {
      "epoch": 0.20857418111753373,
      "grad_norm": 0.2523629367351532,
      "learning_rate": 0.0001,
      "loss": 1.568,
      "step": 1299
    },
    {
      "epoch": 0.20873474630700065,
      "grad_norm": 0.2625834047794342,
      "learning_rate": 0.0001,
      "loss": 1.5783,
      "step": 1300
    },
    {
      "epoch": 0.20889531149646756,
      "grad_norm": 0.2797095775604248,
      "learning_rate": 0.0001,
      "loss": 1.6573,
      "step": 1301
    },
    {
      "epoch": 0.20905587668593448,
      "grad_norm": 0.25999319553375244,
      "learning_rate": 0.0001,
      "loss": 1.5702,
      "step": 1302
    },
    {
      "epoch": 0.20921644187540142,
      "grad_norm": 0.2916666865348816,
      "learning_rate": 0.0001,
      "loss": 1.5995,
      "step": 1303
    },
    {
      "epoch": 0.20937700706486834,
      "grad_norm": 0.28023776412010193,
      "learning_rate": 0.0001,
      "loss": 1.5945,
      "step": 1304
    },
    {
      "epoch": 0.20953757225433525,
      "grad_norm": 0.2535582184791565,
      "learning_rate": 0.0001,
      "loss": 1.5452,
      "step": 1305
    },
    {
      "epoch": 0.2096981374438022,
      "grad_norm": 0.2611648142337799,
      "learning_rate": 0.0001,
      "loss": 1.5214,
      "step": 1306
    },
    {
      "epoch": 0.2098587026332691,
      "grad_norm": 0.27532505989074707,
      "learning_rate": 0.0001,
      "loss": 1.6111,
      "step": 1307
    },
    {
      "epoch": 0.21001926782273603,
      "grad_norm": 0.2723951041698456,
      "learning_rate": 0.0001,
      "loss": 1.636,
      "step": 1308
    },
    {
      "epoch": 0.21017983301220294,
      "grad_norm": 0.24837726354599,
      "learning_rate": 0.0001,
      "loss": 1.6595,
      "step": 1309
    },
    {
      "epoch": 0.2103403982016699,
      "grad_norm": 0.26135581731796265,
      "learning_rate": 0.0001,
      "loss": 1.6177,
      "step": 1310
    },
    {
      "epoch": 0.2105009633911368,
      "grad_norm": 0.27552881836891174,
      "learning_rate": 0.0001,
      "loss": 1.6029,
      "step": 1311
    },
    {
      "epoch": 0.21066152858060372,
      "grad_norm": 0.7178780436515808,
      "learning_rate": 0.0001,
      "loss": 1.5322,
      "step": 1312
    },
    {
      "epoch": 0.21082209377007066,
      "grad_norm": 0.2589340806007385,
      "learning_rate": 0.0001,
      "loss": 1.6379,
      "step": 1313
    },
    {
      "epoch": 0.21098265895953758,
      "grad_norm": 0.26356926560401917,
      "learning_rate": 0.0001,
      "loss": 1.6699,
      "step": 1314
    },
    {
      "epoch": 0.2111432241490045,
      "grad_norm": 0.2632535994052887,
      "learning_rate": 0.0001,
      "loss": 1.611,
      "step": 1315
    },
    {
      "epoch": 0.2113037893384714,
      "grad_norm": 0.2623354494571686,
      "learning_rate": 0.0001,
      "loss": 1.55,
      "step": 1316
    },
    {
      "epoch": 0.21146435452793835,
      "grad_norm": 0.2578772008419037,
      "learning_rate": 0.0001,
      "loss": 1.5774,
      "step": 1317
    },
    {
      "epoch": 0.21162491971740527,
      "grad_norm": 0.26975879073143005,
      "learning_rate": 0.0001,
      "loss": 1.6416,
      "step": 1318
    },
    {
      "epoch": 0.21178548490687218,
      "grad_norm": 0.28247830271720886,
      "learning_rate": 0.0001,
      "loss": 1.6189,
      "step": 1319
    },
    {
      "epoch": 0.2119460500963391,
      "grad_norm": 0.24312029778957367,
      "learning_rate": 0.0001,
      "loss": 1.5905,
      "step": 1320
    },
    {
      "epoch": 0.21210661528580604,
      "grad_norm": 0.26209938526153564,
      "learning_rate": 0.0001,
      "loss": 1.6204,
      "step": 1321
    },
    {
      "epoch": 0.21226718047527296,
      "grad_norm": 0.262643039226532,
      "learning_rate": 0.0001,
      "loss": 1.6566,
      "step": 1322
    },
    {
      "epoch": 0.21242774566473988,
      "grad_norm": 0.25319334864616394,
      "learning_rate": 0.0001,
      "loss": 1.5506,
      "step": 1323
    },
    {
      "epoch": 0.21258831085420682,
      "grad_norm": 0.2527086138725281,
      "learning_rate": 0.0001,
      "loss": 1.5746,
      "step": 1324
    },
    {
      "epoch": 0.21274887604367373,
      "grad_norm": 0.24679316580295563,
      "learning_rate": 0.0001,
      "loss": 1.5769,
      "step": 1325
    },
    {
      "epoch": 0.21290944123314065,
      "grad_norm": 0.6807023882865906,
      "learning_rate": 0.0001,
      "loss": 1.5954,
      "step": 1326
    },
    {
      "epoch": 0.21307000642260757,
      "grad_norm": 0.25651562213897705,
      "learning_rate": 0.0001,
      "loss": 1.5962,
      "step": 1327
    },
    {
      "epoch": 0.2132305716120745,
      "grad_norm": 0.27083685994148254,
      "learning_rate": 0.0001,
      "loss": 1.5687,
      "step": 1328
    },
    {
      "epoch": 0.21339113680154143,
      "grad_norm": 0.26416775584220886,
      "learning_rate": 0.0001,
      "loss": 1.6133,
      "step": 1329
    },
    {
      "epoch": 0.21355170199100834,
      "grad_norm": 0.2596018612384796,
      "learning_rate": 0.0001,
      "loss": 1.6272,
      "step": 1330
    },
    {
      "epoch": 0.21371226718047528,
      "grad_norm": 0.2947336733341217,
      "learning_rate": 0.0001,
      "loss": 1.5521,
      "step": 1331
    },
    {
      "epoch": 0.2138728323699422,
      "grad_norm": 0.2585946023464203,
      "learning_rate": 0.0001,
      "loss": 1.5276,
      "step": 1332
    },
    {
      "epoch": 0.21403339755940912,
      "grad_norm": 0.24610991775989532,
      "learning_rate": 0.0001,
      "loss": 1.5619,
      "step": 1333
    },
    {
      "epoch": 0.21419396274887603,
      "grad_norm": 0.2546312212944031,
      "learning_rate": 0.0001,
      "loss": 1.5592,
      "step": 1334
    },
    {
      "epoch": 0.21435452793834298,
      "grad_norm": 0.25936055183410645,
      "learning_rate": 0.0001,
      "loss": 1.5805,
      "step": 1335
    },
    {
      "epoch": 0.2145150931278099,
      "grad_norm": 0.26311784982681274,
      "learning_rate": 0.0001,
      "loss": 1.6176,
      "step": 1336
    },
    {
      "epoch": 0.2146756583172768,
      "grad_norm": 0.2564358115196228,
      "learning_rate": 0.0001,
      "loss": 1.6126,
      "step": 1337
    },
    {
      "epoch": 0.21483622350674375,
      "grad_norm": 0.25600630044937134,
      "learning_rate": 0.0001,
      "loss": 1.5878,
      "step": 1338
    },
    {
      "epoch": 0.21499678869621067,
      "grad_norm": 0.24495278298854828,
      "learning_rate": 0.0001,
      "loss": 1.5802,
      "step": 1339
    },
    {
      "epoch": 0.21515735388567758,
      "grad_norm": 0.24395935237407684,
      "learning_rate": 0.0001,
      "loss": 1.578,
      "step": 1340
    },
    {
      "epoch": 0.2153179190751445,
      "grad_norm": 0.25721320509910583,
      "learning_rate": 0.0001,
      "loss": 1.6087,
      "step": 1341
    },
    {
      "epoch": 0.21547848426461144,
      "grad_norm": 0.2649543583393097,
      "learning_rate": 0.0001,
      "loss": 1.5431,
      "step": 1342
    },
    {
      "epoch": 0.21563904945407836,
      "grad_norm": 0.2546660304069519,
      "learning_rate": 0.0001,
      "loss": 1.5543,
      "step": 1343
    },
    {
      "epoch": 0.21579961464354527,
      "grad_norm": 0.26365596055984497,
      "learning_rate": 0.0001,
      "loss": 1.5639,
      "step": 1344
    },
    {
      "epoch": 0.2159601798330122,
      "grad_norm": 0.24007868766784668,
      "learning_rate": 0.0001,
      "loss": 1.5436,
      "step": 1345
    },
    {
      "epoch": 0.21612074502247913,
      "grad_norm": 0.24529314041137695,
      "learning_rate": 0.0001,
      "loss": 1.5986,
      "step": 1346
    },
    {
      "epoch": 0.21628131021194605,
      "grad_norm": 0.27236080169677734,
      "learning_rate": 0.0001,
      "loss": 1.6892,
      "step": 1347
    },
    {
      "epoch": 0.21644187540141296,
      "grad_norm": 0.26769787073135376,
      "learning_rate": 0.0001,
      "loss": 1.5531,
      "step": 1348
    },
    {
      "epoch": 0.2166024405908799,
      "grad_norm": 0.23053252696990967,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 1349
    },
    {
      "epoch": 0.21676300578034682,
      "grad_norm": 0.3684787452220917,
      "learning_rate": 0.0001,
      "loss": 1.5096,
      "step": 1350
    },
    {
      "epoch": 0.21692357096981374,
      "grad_norm": 0.2830764055252075,
      "learning_rate": 0.0001,
      "loss": 1.6556,
      "step": 1351
    },
    {
      "epoch": 0.21708413615928066,
      "grad_norm": 0.2534649670124054,
      "learning_rate": 0.0001,
      "loss": 1.4971,
      "step": 1352
    },
    {
      "epoch": 0.2172447013487476,
      "grad_norm": 0.2649475336074829,
      "learning_rate": 0.0001,
      "loss": 1.5548,
      "step": 1353
    },
    {
      "epoch": 0.21740526653821451,
      "grad_norm": 0.2695215046405792,
      "learning_rate": 0.0001,
      "loss": 1.5673,
      "step": 1354
    },
    {
      "epoch": 0.21756583172768143,
      "grad_norm": 0.24033614993095398,
      "learning_rate": 0.0001,
      "loss": 1.5446,
      "step": 1355
    },
    {
      "epoch": 0.21772639691714837,
      "grad_norm": 0.2498067319393158,
      "learning_rate": 0.0001,
      "loss": 1.5808,
      "step": 1356
    },
    {
      "epoch": 0.2178869621066153,
      "grad_norm": 0.2555305063724518,
      "learning_rate": 0.0001,
      "loss": 1.5075,
      "step": 1357
    },
    {
      "epoch": 0.2180475272960822,
      "grad_norm": 0.24671803414821625,
      "learning_rate": 0.0001,
      "loss": 1.5241,
      "step": 1358
    },
    {
      "epoch": 0.21820809248554912,
      "grad_norm": 0.25955885648727417,
      "learning_rate": 0.0001,
      "loss": 1.6222,
      "step": 1359
    },
    {
      "epoch": 0.21836865767501606,
      "grad_norm": 0.2679576873779297,
      "learning_rate": 0.0001,
      "loss": 1.6631,
      "step": 1360
    },
    {
      "epoch": 0.21852922286448298,
      "grad_norm": 0.26426801085472107,
      "learning_rate": 0.0001,
      "loss": 1.516,
      "step": 1361
    },
    {
      "epoch": 0.2186897880539499,
      "grad_norm": 0.25338250398635864,
      "learning_rate": 0.0001,
      "loss": 1.6029,
      "step": 1362
    },
    {
      "epoch": 0.21885035324341684,
      "grad_norm": 0.27184170484542847,
      "learning_rate": 0.0001,
      "loss": 1.6391,
      "step": 1363
    },
    {
      "epoch": 0.21901091843288376,
      "grad_norm": 0.25309568643569946,
      "learning_rate": 0.0001,
      "loss": 1.4754,
      "step": 1364
    },
    {
      "epoch": 0.21917148362235067,
      "grad_norm": 0.38762274384498596,
      "learning_rate": 0.0001,
      "loss": 1.5477,
      "step": 1365
    },
    {
      "epoch": 0.2193320488118176,
      "grad_norm": 0.2547190189361572,
      "learning_rate": 0.0001,
      "loss": 1.5877,
      "step": 1366
    },
    {
      "epoch": 0.21949261400128453,
      "grad_norm": 0.24988453090190887,
      "learning_rate": 0.0001,
      "loss": 1.5625,
      "step": 1367
    },
    {
      "epoch": 0.21965317919075145,
      "grad_norm": 0.25075602531433105,
      "learning_rate": 0.0001,
      "loss": 1.4408,
      "step": 1368
    },
    {
      "epoch": 0.21981374438021836,
      "grad_norm": 0.25572770833969116,
      "learning_rate": 0.0001,
      "loss": 1.6669,
      "step": 1369
    },
    {
      "epoch": 0.2199743095696853,
      "grad_norm": 0.24106429517269135,
      "learning_rate": 0.0001,
      "loss": 1.4878,
      "step": 1370
    },
    {
      "epoch": 0.22013487475915222,
      "grad_norm": 0.25900036096572876,
      "learning_rate": 0.0001,
      "loss": 1.5423,
      "step": 1371
    },
    {
      "epoch": 0.22029543994861914,
      "grad_norm": 0.25564947724342346,
      "learning_rate": 0.0001,
      "loss": 1.6369,
      "step": 1372
    },
    {
      "epoch": 0.22045600513808605,
      "grad_norm": 0.24209898710250854,
      "learning_rate": 0.0001,
      "loss": 1.5761,
      "step": 1373
    },
    {
      "epoch": 0.220616570327553,
      "grad_norm": 0.2538258135318756,
      "learning_rate": 0.0001,
      "loss": 1.5493,
      "step": 1374
    },
    {
      "epoch": 0.2207771355170199,
      "grad_norm": 0.25893691182136536,
      "learning_rate": 0.0001,
      "loss": 1.5965,
      "step": 1375
    },
    {
      "epoch": 0.22093770070648683,
      "grad_norm": 0.2606137990951538,
      "learning_rate": 0.0001,
      "loss": 1.57,
      "step": 1376
    },
    {
      "epoch": 0.22109826589595374,
      "grad_norm": 0.24876049160957336,
      "learning_rate": 0.0001,
      "loss": 1.5434,
      "step": 1377
    },
    {
      "epoch": 0.2212588310854207,
      "grad_norm": 0.24675314128398895,
      "learning_rate": 0.0001,
      "loss": 1.5311,
      "step": 1378
    },
    {
      "epoch": 0.2214193962748876,
      "grad_norm": 0.2675689160823822,
      "learning_rate": 0.0001,
      "loss": 1.6377,
      "step": 1379
    },
    {
      "epoch": 0.22157996146435452,
      "grad_norm": 0.25446370244026184,
      "learning_rate": 0.0001,
      "loss": 1.5623,
      "step": 1380
    },
    {
      "epoch": 0.22174052665382146,
      "grad_norm": 0.35192978382110596,
      "learning_rate": 0.0001,
      "loss": 1.5927,
      "step": 1381
    },
    {
      "epoch": 0.22190109184328838,
      "grad_norm": 0.26166224479675293,
      "learning_rate": 0.0001,
      "loss": 1.7022,
      "step": 1382
    },
    {
      "epoch": 0.2220616570327553,
      "grad_norm": 0.24951639771461487,
      "learning_rate": 0.0001,
      "loss": 1.6084,
      "step": 1383
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.2615184783935547,
      "learning_rate": 0.0001,
      "loss": 1.5301,
      "step": 1384
    },
    {
      "epoch": 0.22238278741168915,
      "grad_norm": 0.2507842481136322,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 1385
    },
    {
      "epoch": 0.22254335260115607,
      "grad_norm": 0.24396218359470367,
      "learning_rate": 0.0001,
      "loss": 1.5362,
      "step": 1386
    },
    {
      "epoch": 0.22270391779062298,
      "grad_norm": 0.25023460388183594,
      "learning_rate": 0.0001,
      "loss": 1.5818,
      "step": 1387
    },
    {
      "epoch": 0.22286448298008993,
      "grad_norm": 0.26676830649375916,
      "learning_rate": 0.0001,
      "loss": 1.5556,
      "step": 1388
    },
    {
      "epoch": 0.22302504816955684,
      "grad_norm": 0.2628074586391449,
      "learning_rate": 0.0001,
      "loss": 1.584,
      "step": 1389
    },
    {
      "epoch": 0.22318561335902376,
      "grad_norm": 0.24453000724315643,
      "learning_rate": 0.0001,
      "loss": 1.5694,
      "step": 1390
    },
    {
      "epoch": 0.22334617854849068,
      "grad_norm": 0.24808208644390106,
      "learning_rate": 0.0001,
      "loss": 1.5199,
      "step": 1391
    },
    {
      "epoch": 0.22350674373795762,
      "grad_norm": 0.25541916489601135,
      "learning_rate": 0.0001,
      "loss": 1.5991,
      "step": 1392
    },
    {
      "epoch": 0.22366730892742454,
      "grad_norm": 0.25001993775367737,
      "learning_rate": 0.0001,
      "loss": 1.5395,
      "step": 1393
    },
    {
      "epoch": 0.22382787411689145,
      "grad_norm": 0.24925601482391357,
      "learning_rate": 0.0001,
      "loss": 1.5594,
      "step": 1394
    },
    {
      "epoch": 0.2239884393063584,
      "grad_norm": 0.24612492322921753,
      "learning_rate": 0.0001,
      "loss": 1.5818,
      "step": 1395
    },
    {
      "epoch": 0.2241490044958253,
      "grad_norm": 0.25647103786468506,
      "learning_rate": 0.0001,
      "loss": 1.6169,
      "step": 1396
    },
    {
      "epoch": 0.22430956968529223,
      "grad_norm": 0.25585880875587463,
      "learning_rate": 0.0001,
      "loss": 1.5923,
      "step": 1397
    },
    {
      "epoch": 0.22447013487475914,
      "grad_norm": 0.25814953446388245,
      "learning_rate": 0.0001,
      "loss": 1.6253,
      "step": 1398
    },
    {
      "epoch": 0.22463070006422609,
      "grad_norm": 0.2706148624420166,
      "learning_rate": 0.0001,
      "loss": 1.6775,
      "step": 1399
    },
    {
      "epoch": 0.224791265253693,
      "grad_norm": 0.24915863573551178,
      "learning_rate": 0.0001,
      "loss": 1.4988,
      "step": 1400
    },
    {
      "epoch": 0.22495183044315992,
      "grad_norm": 0.25413191318511963,
      "learning_rate": 0.0001,
      "loss": 1.5438,
      "step": 1401
    },
    {
      "epoch": 0.22511239563262683,
      "grad_norm": 0.25795644521713257,
      "learning_rate": 0.0001,
      "loss": 1.6046,
      "step": 1402
    },
    {
      "epoch": 0.22527296082209378,
      "grad_norm": 0.24239666759967804,
      "learning_rate": 0.0001,
      "loss": 1.5229,
      "step": 1403
    },
    {
      "epoch": 0.2254335260115607,
      "grad_norm": 0.2521439492702484,
      "learning_rate": 0.0001,
      "loss": 1.5751,
      "step": 1404
    },
    {
      "epoch": 0.2255940912010276,
      "grad_norm": 0.25757378339767456,
      "learning_rate": 0.0001,
      "loss": 1.5828,
      "step": 1405
    },
    {
      "epoch": 0.22575465639049455,
      "grad_norm": 0.24728292226791382,
      "learning_rate": 0.0001,
      "loss": 1.5471,
      "step": 1406
    },
    {
      "epoch": 0.22591522157996147,
      "grad_norm": 0.2591513693332672,
      "learning_rate": 0.0001,
      "loss": 1.6014,
      "step": 1407
    },
    {
      "epoch": 0.22607578676942838,
      "grad_norm": 0.2557453513145447,
      "learning_rate": 0.0001,
      "loss": 1.5125,
      "step": 1408
    },
    {
      "epoch": 0.2262363519588953,
      "grad_norm": 0.24972213804721832,
      "learning_rate": 0.0001,
      "loss": 1.56,
      "step": 1409
    },
    {
      "epoch": 0.22639691714836224,
      "grad_norm": 0.2518574297428131,
      "learning_rate": 0.0001,
      "loss": 1.5866,
      "step": 1410
    },
    {
      "epoch": 0.22655748233782916,
      "grad_norm": 0.2644212245941162,
      "learning_rate": 0.0001,
      "loss": 1.505,
      "step": 1411
    },
    {
      "epoch": 0.22671804752729607,
      "grad_norm": 0.25966617465019226,
      "learning_rate": 0.0001,
      "loss": 1.6544,
      "step": 1412
    },
    {
      "epoch": 0.22687861271676302,
      "grad_norm": 0.25631988048553467,
      "learning_rate": 0.0001,
      "loss": 1.5058,
      "step": 1413
    },
    {
      "epoch": 0.22703917790622993,
      "grad_norm": 0.2735203504562378,
      "learning_rate": 0.0001,
      "loss": 1.5849,
      "step": 1414
    },
    {
      "epoch": 0.22719974309569685,
      "grad_norm": 0.24585656821727753,
      "learning_rate": 0.0001,
      "loss": 1.6229,
      "step": 1415
    },
    {
      "epoch": 0.22736030828516376,
      "grad_norm": 0.2850990891456604,
      "learning_rate": 0.0001,
      "loss": 1.6063,
      "step": 1416
    },
    {
      "epoch": 0.2275208734746307,
      "grad_norm": 0.23922717571258545,
      "learning_rate": 0.0001,
      "loss": 1.4942,
      "step": 1417
    },
    {
      "epoch": 0.22768143866409762,
      "grad_norm": 0.24667245149612427,
      "learning_rate": 0.0001,
      "loss": 1.5835,
      "step": 1418
    },
    {
      "epoch": 0.22784200385356454,
      "grad_norm": 0.2509988248348236,
      "learning_rate": 0.0001,
      "loss": 1.5977,
      "step": 1419
    },
    {
      "epoch": 0.22800256904303148,
      "grad_norm": 0.24224527180194855,
      "learning_rate": 0.0001,
      "loss": 1.5217,
      "step": 1420
    },
    {
      "epoch": 0.2281631342324984,
      "grad_norm": 0.25608593225479126,
      "learning_rate": 0.0001,
      "loss": 1.6166,
      "step": 1421
    },
    {
      "epoch": 0.22832369942196531,
      "grad_norm": 0.2535424828529358,
      "learning_rate": 0.0001,
      "loss": 1.6606,
      "step": 1422
    },
    {
      "epoch": 0.22848426461143223,
      "grad_norm": 3.90877628326416,
      "learning_rate": 0.0001,
      "loss": 1.6422,
      "step": 1423
    },
    {
      "epoch": 0.22864482980089917,
      "grad_norm": 0.24865517020225525,
      "learning_rate": 0.0001,
      "loss": 1.6414,
      "step": 1424
    },
    {
      "epoch": 0.2288053949903661,
      "grad_norm": 0.2509647607803345,
      "learning_rate": 0.0001,
      "loss": 1.5995,
      "step": 1425
    },
    {
      "epoch": 0.228965960179833,
      "grad_norm": 0.25810936093330383,
      "learning_rate": 0.0001,
      "loss": 1.6074,
      "step": 1426
    },
    {
      "epoch": 0.22912652536929995,
      "grad_norm": 0.2706359326839447,
      "learning_rate": 0.0001,
      "loss": 1.5623,
      "step": 1427
    },
    {
      "epoch": 0.22928709055876687,
      "grad_norm": 0.2858353853225708,
      "learning_rate": 0.0001,
      "loss": 1.6383,
      "step": 1428
    },
    {
      "epoch": 0.22944765574823378,
      "grad_norm": 0.26110613346099854,
      "learning_rate": 0.0001,
      "loss": 1.6075,
      "step": 1429
    },
    {
      "epoch": 0.2296082209377007,
      "grad_norm": 0.25887665152549744,
      "learning_rate": 0.0001,
      "loss": 1.5545,
      "step": 1430
    },
    {
      "epoch": 0.22976878612716764,
      "grad_norm": 0.25360432267189026,
      "learning_rate": 0.0001,
      "loss": 1.6354,
      "step": 1431
    },
    {
      "epoch": 0.22992935131663456,
      "grad_norm": 0.24859270453453064,
      "learning_rate": 0.0001,
      "loss": 1.5574,
      "step": 1432
    },
    {
      "epoch": 0.23008991650610147,
      "grad_norm": 0.24973681569099426,
      "learning_rate": 0.0001,
      "loss": 1.5883,
      "step": 1433
    },
    {
      "epoch": 0.2302504816955684,
      "grad_norm": 0.2463744878768921,
      "learning_rate": 0.0001,
      "loss": 1.5352,
      "step": 1434
    },
    {
      "epoch": 0.23041104688503533,
      "grad_norm": 0.2605551779270172,
      "learning_rate": 0.0001,
      "loss": 1.5923,
      "step": 1435
    },
    {
      "epoch": 0.23057161207450225,
      "grad_norm": 0.2524442970752716,
      "learning_rate": 0.0001,
      "loss": 1.5501,
      "step": 1436
    },
    {
      "epoch": 0.23073217726396916,
      "grad_norm": 0.26319846510887146,
      "learning_rate": 0.0001,
      "loss": 1.6097,
      "step": 1437
    },
    {
      "epoch": 0.2308927424534361,
      "grad_norm": 0.2604261636734009,
      "learning_rate": 0.0001,
      "loss": 1.6013,
      "step": 1438
    },
    {
      "epoch": 0.23105330764290302,
      "grad_norm": 0.2449166178703308,
      "learning_rate": 0.0001,
      "loss": 1.4734,
      "step": 1439
    },
    {
      "epoch": 0.23121387283236994,
      "grad_norm": 0.2389906495809555,
      "learning_rate": 0.0001,
      "loss": 1.5959,
      "step": 1440
    },
    {
      "epoch": 0.23137443802183685,
      "grad_norm": 0.26003342866897583,
      "learning_rate": 0.0001,
      "loss": 1.6533,
      "step": 1441
    },
    {
      "epoch": 0.2315350032113038,
      "grad_norm": 0.2461751103401184,
      "learning_rate": 0.0001,
      "loss": 1.5297,
      "step": 1442
    },
    {
      "epoch": 0.2316955684007707,
      "grad_norm": 0.25976449251174927,
      "learning_rate": 0.0001,
      "loss": 1.5842,
      "step": 1443
    },
    {
      "epoch": 0.23185613359023763,
      "grad_norm": 0.24590584635734558,
      "learning_rate": 0.0001,
      "loss": 1.4882,
      "step": 1444
    },
    {
      "epoch": 0.23201669877970457,
      "grad_norm": 0.25528788566589355,
      "learning_rate": 0.0001,
      "loss": 1.5965,
      "step": 1445
    },
    {
      "epoch": 0.2321772639691715,
      "grad_norm": 0.2836610674858093,
      "learning_rate": 0.0001,
      "loss": 1.6253,
      "step": 1446
    },
    {
      "epoch": 0.2323378291586384,
      "grad_norm": 0.24190182983875275,
      "learning_rate": 0.0001,
      "loss": 1.5365,
      "step": 1447
    },
    {
      "epoch": 0.23249839434810532,
      "grad_norm": 0.2515237033367157,
      "learning_rate": 0.0001,
      "loss": 1.6431,
      "step": 1448
    },
    {
      "epoch": 0.23265895953757226,
      "grad_norm": 0.24769434332847595,
      "learning_rate": 0.0001,
      "loss": 1.5132,
      "step": 1449
    },
    {
      "epoch": 0.23281952472703918,
      "grad_norm": 0.24459420144557953,
      "learning_rate": 0.0001,
      "loss": 1.5509,
      "step": 1450
    },
    {
      "epoch": 0.2329800899165061,
      "grad_norm": 0.2621827721595764,
      "learning_rate": 0.0001,
      "loss": 1.5563,
      "step": 1451
    },
    {
      "epoch": 0.23314065510597304,
      "grad_norm": 0.27197253704071045,
      "learning_rate": 0.0001,
      "loss": 1.5884,
      "step": 1452
    },
    {
      "epoch": 0.23330122029543995,
      "grad_norm": 0.2468864619731903,
      "learning_rate": 0.0001,
      "loss": 1.594,
      "step": 1453
    },
    {
      "epoch": 0.23346178548490687,
      "grad_norm": 0.25369641184806824,
      "learning_rate": 0.0001,
      "loss": 1.5619,
      "step": 1454
    },
    {
      "epoch": 0.23362235067437379,
      "grad_norm": 0.250773549079895,
      "learning_rate": 0.0001,
      "loss": 1.5888,
      "step": 1455
    },
    {
      "epoch": 0.23378291586384073,
      "grad_norm": 0.23992547392845154,
      "learning_rate": 0.0001,
      "loss": 1.5831,
      "step": 1456
    },
    {
      "epoch": 0.23394348105330764,
      "grad_norm": 0.2697206139564514,
      "learning_rate": 0.0001,
      "loss": 1.5004,
      "step": 1457
    },
    {
      "epoch": 0.23410404624277456,
      "grad_norm": 0.25312143564224243,
      "learning_rate": 0.0001,
      "loss": 1.578,
      "step": 1458
    },
    {
      "epoch": 0.23426461143224148,
      "grad_norm": 0.256674587726593,
      "learning_rate": 0.0001,
      "loss": 1.5219,
      "step": 1459
    },
    {
      "epoch": 0.23442517662170842,
      "grad_norm": 0.25566184520721436,
      "learning_rate": 0.0001,
      "loss": 1.5939,
      "step": 1460
    },
    {
      "epoch": 0.23458574181117534,
      "grad_norm": 0.24389982223510742,
      "learning_rate": 0.0001,
      "loss": 1.5649,
      "step": 1461
    },
    {
      "epoch": 0.23474630700064225,
      "grad_norm": 0.24940608441829681,
      "learning_rate": 0.0001,
      "loss": 1.5485,
      "step": 1462
    },
    {
      "epoch": 0.2349068721901092,
      "grad_norm": 0.24859265983104706,
      "learning_rate": 0.0001,
      "loss": 1.5356,
      "step": 1463
    },
    {
      "epoch": 0.2350674373795761,
      "grad_norm": 0.2573135793209076,
      "learning_rate": 0.0001,
      "loss": 1.6227,
      "step": 1464
    },
    {
      "epoch": 0.23522800256904303,
      "grad_norm": 0.24468062818050385,
      "learning_rate": 0.0001,
      "loss": 1.593,
      "step": 1465
    },
    {
      "epoch": 0.23538856775850994,
      "grad_norm": 0.27289363741874695,
      "learning_rate": 0.0001,
      "loss": 1.5822,
      "step": 1466
    },
    {
      "epoch": 0.23554913294797689,
      "grad_norm": 0.2553497850894928,
      "learning_rate": 0.0001,
      "loss": 1.5125,
      "step": 1467
    },
    {
      "epoch": 0.2357096981374438,
      "grad_norm": 0.24384398758411407,
      "learning_rate": 0.0001,
      "loss": 1.5465,
      "step": 1468
    },
    {
      "epoch": 0.23587026332691072,
      "grad_norm": 0.2581036686897278,
      "learning_rate": 0.0001,
      "loss": 1.519,
      "step": 1469
    },
    {
      "epoch": 0.23603082851637766,
      "grad_norm": 0.246010884642601,
      "learning_rate": 0.0001,
      "loss": 1.5899,
      "step": 1470
    },
    {
      "epoch": 0.23619139370584458,
      "grad_norm": 0.25303733348846436,
      "learning_rate": 0.0001,
      "loss": 1.5814,
      "step": 1471
    },
    {
      "epoch": 0.2363519588953115,
      "grad_norm": 0.25244733691215515,
      "learning_rate": 0.0001,
      "loss": 1.5473,
      "step": 1472
    },
    {
      "epoch": 0.2365125240847784,
      "grad_norm": 0.2596549391746521,
      "learning_rate": 0.0001,
      "loss": 1.5539,
      "step": 1473
    },
    {
      "epoch": 0.23667308927424535,
      "grad_norm": 0.24361246824264526,
      "learning_rate": 0.0001,
      "loss": 1.5328,
      "step": 1474
    },
    {
      "epoch": 0.23683365446371227,
      "grad_norm": 0.2471035271883011,
      "learning_rate": 0.0001,
      "loss": 1.5761,
      "step": 1475
    },
    {
      "epoch": 0.23699421965317918,
      "grad_norm": 0.25243648886680603,
      "learning_rate": 0.0001,
      "loss": 1.6034,
      "step": 1476
    },
    {
      "epoch": 0.23715478484264613,
      "grad_norm": 0.25919926166534424,
      "learning_rate": 0.0001,
      "loss": 1.5426,
      "step": 1477
    },
    {
      "epoch": 0.23731535003211304,
      "grad_norm": 0.260970801115036,
      "learning_rate": 0.0001,
      "loss": 1.531,
      "step": 1478
    },
    {
      "epoch": 0.23747591522157996,
      "grad_norm": 0.24548634886741638,
      "learning_rate": 0.0001,
      "loss": 1.5057,
      "step": 1479
    },
    {
      "epoch": 0.23763648041104687,
      "grad_norm": 0.2470322996377945,
      "learning_rate": 0.0001,
      "loss": 1.5392,
      "step": 1480
    },
    {
      "epoch": 0.23779704560051382,
      "grad_norm": 0.2650474011898041,
      "learning_rate": 0.0001,
      "loss": 1.588,
      "step": 1481
    },
    {
      "epoch": 0.23795761078998073,
      "grad_norm": 0.2452380210161209,
      "learning_rate": 0.0001,
      "loss": 1.575,
      "step": 1482
    },
    {
      "epoch": 0.23811817597944765,
      "grad_norm": 0.22939066588878632,
      "learning_rate": 0.0001,
      "loss": 1.4756,
      "step": 1483
    },
    {
      "epoch": 0.2382787411689146,
      "grad_norm": 0.2571943998336792,
      "learning_rate": 0.0001,
      "loss": 1.645,
      "step": 1484
    },
    {
      "epoch": 0.2384393063583815,
      "grad_norm": 0.2676110863685608,
      "learning_rate": 0.0001,
      "loss": 1.5951,
      "step": 1485
    },
    {
      "epoch": 0.23859987154784842,
      "grad_norm": 0.26024967432022095,
      "learning_rate": 0.0001,
      "loss": 1.5544,
      "step": 1486
    },
    {
      "epoch": 0.23876043673731534,
      "grad_norm": 0.279931902885437,
      "learning_rate": 0.0001,
      "loss": 1.5804,
      "step": 1487
    },
    {
      "epoch": 0.23892100192678228,
      "grad_norm": 0.24358607828617096,
      "learning_rate": 0.0001,
      "loss": 1.5338,
      "step": 1488
    },
    {
      "epoch": 0.2390815671162492,
      "grad_norm": 0.25573456287384033,
      "learning_rate": 0.0001,
      "loss": 1.6381,
      "step": 1489
    },
    {
      "epoch": 0.23924213230571612,
      "grad_norm": 0.24336667358875275,
      "learning_rate": 0.0001,
      "loss": 1.5297,
      "step": 1490
    },
    {
      "epoch": 0.23940269749518303,
      "grad_norm": 0.2546198070049286,
      "learning_rate": 0.0001,
      "loss": 1.5391,
      "step": 1491
    },
    {
      "epoch": 0.23956326268464997,
      "grad_norm": 0.24233075976371765,
      "learning_rate": 0.0001,
      "loss": 1.5365,
      "step": 1492
    },
    {
      "epoch": 0.2397238278741169,
      "grad_norm": 0.26909855008125305,
      "learning_rate": 0.0001,
      "loss": 1.601,
      "step": 1493
    },
    {
      "epoch": 0.2398843930635838,
      "grad_norm": 0.2723981440067291,
      "learning_rate": 0.0001,
      "loss": 1.6736,
      "step": 1494
    },
    {
      "epoch": 0.24004495825305075,
      "grad_norm": 0.26914265751838684,
      "learning_rate": 0.0001,
      "loss": 1.6543,
      "step": 1495
    },
    {
      "epoch": 0.24020552344251767,
      "grad_norm": 0.2784527838230133,
      "learning_rate": 0.0001,
      "loss": 1.6494,
      "step": 1496
    },
    {
      "epoch": 0.24036608863198458,
      "grad_norm": 0.2566381096839905,
      "learning_rate": 0.0001,
      "loss": 1.5029,
      "step": 1497
    },
    {
      "epoch": 0.2405266538214515,
      "grad_norm": 0.24081818759441376,
      "learning_rate": 0.0001,
      "loss": 1.5187,
      "step": 1498
    },
    {
      "epoch": 0.24068721901091844,
      "grad_norm": 0.25618815422058105,
      "learning_rate": 0.0001,
      "loss": 1.6156,
      "step": 1499
    },
    {
      "epoch": 0.24084778420038536,
      "grad_norm": 0.24754783511161804,
      "learning_rate": 0.0001,
      "loss": 1.5812,
      "step": 1500
    },
    {
      "epoch": 0.24100834938985227,
      "grad_norm": 0.26266035437583923,
      "learning_rate": 0.0001,
      "loss": 1.6404,
      "step": 1501
    },
    {
      "epoch": 0.24116891457931922,
      "grad_norm": 0.24704232811927795,
      "learning_rate": 0.0001,
      "loss": 1.5442,
      "step": 1502
    },
    {
      "epoch": 0.24132947976878613,
      "grad_norm": 0.25548985600471497,
      "learning_rate": 0.0001,
      "loss": 1.5383,
      "step": 1503
    },
    {
      "epoch": 0.24149004495825305,
      "grad_norm": 0.2660962641239166,
      "learning_rate": 0.0001,
      "loss": 1.5722,
      "step": 1504
    },
    {
      "epoch": 0.24165061014771996,
      "grad_norm": 0.24749712646007538,
      "learning_rate": 0.0001,
      "loss": 1.4693,
      "step": 1505
    },
    {
      "epoch": 0.2418111753371869,
      "grad_norm": 0.24952630698680878,
      "learning_rate": 0.0001,
      "loss": 1.5219,
      "step": 1506
    },
    {
      "epoch": 0.24197174052665382,
      "grad_norm": 0.2579048275947571,
      "learning_rate": 0.0001,
      "loss": 1.6029,
      "step": 1507
    },
    {
      "epoch": 0.24213230571612074,
      "grad_norm": 0.25550124049186707,
      "learning_rate": 0.0001,
      "loss": 1.6096,
      "step": 1508
    },
    {
      "epoch": 0.24229287090558768,
      "grad_norm": 0.25661399960517883,
      "learning_rate": 0.0001,
      "loss": 1.5581,
      "step": 1509
    },
    {
      "epoch": 0.2424534360950546,
      "grad_norm": 0.2509407103061676,
      "learning_rate": 0.0001,
      "loss": 1.5554,
      "step": 1510
    },
    {
      "epoch": 0.2426140012845215,
      "grad_norm": 0.25561586022377014,
      "learning_rate": 0.0001,
      "loss": 1.519,
      "step": 1511
    },
    {
      "epoch": 0.24277456647398843,
      "grad_norm": 0.2742331027984619,
      "learning_rate": 0.0001,
      "loss": 1.6084,
      "step": 1512
    },
    {
      "epoch": 0.24293513166345537,
      "grad_norm": 0.2725457549095154,
      "learning_rate": 0.0001,
      "loss": 1.5215,
      "step": 1513
    },
    {
      "epoch": 0.2430956968529223,
      "grad_norm": 0.25804951786994934,
      "learning_rate": 0.0001,
      "loss": 1.5492,
      "step": 1514
    },
    {
      "epoch": 0.2432562620423892,
      "grad_norm": 0.2545168697834015,
      "learning_rate": 0.0001,
      "loss": 1.5754,
      "step": 1515
    },
    {
      "epoch": 0.24341682723185612,
      "grad_norm": 0.25674569606781006,
      "learning_rate": 0.0001,
      "loss": 1.5144,
      "step": 1516
    },
    {
      "epoch": 0.24357739242132306,
      "grad_norm": 0.2617294192314148,
      "learning_rate": 0.0001,
      "loss": 1.6112,
      "step": 1517
    },
    {
      "epoch": 0.24373795761078998,
      "grad_norm": 0.24985378980636597,
      "learning_rate": 0.0001,
      "loss": 1.5346,
      "step": 1518
    },
    {
      "epoch": 0.2438985228002569,
      "grad_norm": 0.25484099984169006,
      "learning_rate": 0.0001,
      "loss": 1.4858,
      "step": 1519
    },
    {
      "epoch": 0.24405908798972384,
      "grad_norm": 0.2507209777832031,
      "learning_rate": 0.0001,
      "loss": 1.5587,
      "step": 1520
    },
    {
      "epoch": 0.24421965317919075,
      "grad_norm": 0.2766501009464264,
      "learning_rate": 0.0001,
      "loss": 1.6381,
      "step": 1521
    },
    {
      "epoch": 0.24438021836865767,
      "grad_norm": 0.25130683183670044,
      "learning_rate": 0.0001,
      "loss": 1.494,
      "step": 1522
    },
    {
      "epoch": 0.24454078355812459,
      "grad_norm": 0.24670766294002533,
      "learning_rate": 0.0001,
      "loss": 1.5293,
      "step": 1523
    },
    {
      "epoch": 0.24470134874759153,
      "grad_norm": 0.25187036395072937,
      "learning_rate": 0.0001,
      "loss": 1.5031,
      "step": 1524
    },
    {
      "epoch": 0.24486191393705845,
      "grad_norm": 0.24657708406448364,
      "learning_rate": 0.0001,
      "loss": 1.5075,
      "step": 1525
    },
    {
      "epoch": 0.24502247912652536,
      "grad_norm": 0.24846403300762177,
      "learning_rate": 0.0001,
      "loss": 1.5942,
      "step": 1526
    },
    {
      "epoch": 0.2451830443159923,
      "grad_norm": 0.24523480236530304,
      "learning_rate": 0.0001,
      "loss": 1.5384,
      "step": 1527
    },
    {
      "epoch": 0.24534360950545922,
      "grad_norm": 0.25702452659606934,
      "learning_rate": 0.0001,
      "loss": 1.5396,
      "step": 1528
    },
    {
      "epoch": 0.24550417469492614,
      "grad_norm": 0.2500796318054199,
      "learning_rate": 0.0001,
      "loss": 1.4692,
      "step": 1529
    },
    {
      "epoch": 0.24566473988439305,
      "grad_norm": 0.24745675921440125,
      "learning_rate": 0.0001,
      "loss": 1.5358,
      "step": 1530
    },
    {
      "epoch": 0.24582530507386,
      "grad_norm": 0.26025524735450745,
      "learning_rate": 0.0001,
      "loss": 1.5659,
      "step": 1531
    },
    {
      "epoch": 0.2459858702633269,
      "grad_norm": 0.2700679302215576,
      "learning_rate": 0.0001,
      "loss": 1.6448,
      "step": 1532
    },
    {
      "epoch": 0.24614643545279383,
      "grad_norm": 0.24460889399051666,
      "learning_rate": 0.0001,
      "loss": 1.6364,
      "step": 1533
    },
    {
      "epoch": 0.24630700064226077,
      "grad_norm": 0.24033024907112122,
      "learning_rate": 0.0001,
      "loss": 1.5412,
      "step": 1534
    },
    {
      "epoch": 0.2464675658317277,
      "grad_norm": 0.25126370787620544,
      "learning_rate": 0.0001,
      "loss": 1.6012,
      "step": 1535
    },
    {
      "epoch": 0.2466281310211946,
      "grad_norm": 0.2466874122619629,
      "learning_rate": 0.0001,
      "loss": 1.6127,
      "step": 1536
    },
    {
      "epoch": 0.24678869621066152,
      "grad_norm": 0.24531720578670502,
      "learning_rate": 0.0001,
      "loss": 1.5165,
      "step": 1537
    },
    {
      "epoch": 0.24694926140012846,
      "grad_norm": 0.2571285367012024,
      "learning_rate": 0.0001,
      "loss": 1.6267,
      "step": 1538
    },
    {
      "epoch": 0.24710982658959538,
      "grad_norm": 0.25678715109825134,
      "learning_rate": 0.0001,
      "loss": 1.5682,
      "step": 1539
    },
    {
      "epoch": 0.2472703917790623,
      "grad_norm": 0.2479630708694458,
      "learning_rate": 0.0001,
      "loss": 1.5792,
      "step": 1540
    },
    {
      "epoch": 0.24743095696852924,
      "grad_norm": 0.2584831118583679,
      "learning_rate": 0.0001,
      "loss": 1.5203,
      "step": 1541
    },
    {
      "epoch": 0.24759152215799615,
      "grad_norm": 0.2596479654312134,
      "learning_rate": 0.0001,
      "loss": 1.6176,
      "step": 1542
    },
    {
      "epoch": 0.24775208734746307,
      "grad_norm": 0.2657060921192169,
      "learning_rate": 0.0001,
      "loss": 1.6722,
      "step": 1543
    },
    {
      "epoch": 0.24791265253692998,
      "grad_norm": 0.25417783856391907,
      "learning_rate": 0.0001,
      "loss": 1.5195,
      "step": 1544
    },
    {
      "epoch": 0.24807321772639693,
      "grad_norm": 0.25777631998062134,
      "learning_rate": 0.0001,
      "loss": 1.566,
      "step": 1545
    },
    {
      "epoch": 0.24823378291586384,
      "grad_norm": 0.24382346868515015,
      "learning_rate": 0.0001,
      "loss": 1.499,
      "step": 1546
    },
    {
      "epoch": 0.24839434810533076,
      "grad_norm": 0.26882243156433105,
      "learning_rate": 0.0001,
      "loss": 1.6298,
      "step": 1547
    },
    {
      "epoch": 0.24855491329479767,
      "grad_norm": 0.25480836629867554,
      "learning_rate": 0.0001,
      "loss": 1.5445,
      "step": 1548
    },
    {
      "epoch": 0.24871547848426462,
      "grad_norm": 0.269992858171463,
      "learning_rate": 0.0001,
      "loss": 1.5671,
      "step": 1549
    },
    {
      "epoch": 0.24887604367373153,
      "grad_norm": 0.27160781621932983,
      "learning_rate": 0.0001,
      "loss": 1.563,
      "step": 1550
    },
    {
      "epoch": 0.24903660886319845,
      "grad_norm": 0.24489404261112213,
      "learning_rate": 0.0001,
      "loss": 1.5698,
      "step": 1551
    },
    {
      "epoch": 0.2491971740526654,
      "grad_norm": 0.2606049180030823,
      "learning_rate": 0.0001,
      "loss": 1.5125,
      "step": 1552
    },
    {
      "epoch": 0.2493577392421323,
      "grad_norm": 0.2516750395298004,
      "learning_rate": 0.0001,
      "loss": 1.621,
      "step": 1553
    },
    {
      "epoch": 0.24951830443159922,
      "grad_norm": 0.2516483962535858,
      "learning_rate": 0.0001,
      "loss": 1.6886,
      "step": 1554
    },
    {
      "epoch": 0.24967886962106614,
      "grad_norm": 0.24207676947116852,
      "learning_rate": 0.0001,
      "loss": 1.5208,
      "step": 1555
    },
    {
      "epoch": 0.24983943481053308,
      "grad_norm": 0.2646651864051819,
      "learning_rate": 0.0001,
      "loss": 1.5327,
      "step": 1556
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.2527795433998108,
      "learning_rate": 0.0001,
      "loss": 1.6105,
      "step": 1557
    },
    {
      "epoch": 0.25016056518946694,
      "grad_norm": 0.2506919801235199,
      "learning_rate": 0.0001,
      "loss": 1.4942,
      "step": 1558
    },
    {
      "epoch": 0.25032113037893383,
      "grad_norm": 0.24822531640529633,
      "learning_rate": 0.0001,
      "loss": 1.5199,
      "step": 1559
    },
    {
      "epoch": 0.2504816955684008,
      "grad_norm": 0.2495066374540329,
      "learning_rate": 0.0001,
      "loss": 1.5926,
      "step": 1560
    },
    {
      "epoch": 0.2506422607578677,
      "grad_norm": 0.26969271898269653,
      "learning_rate": 0.0001,
      "loss": 1.6055,
      "step": 1561
    },
    {
      "epoch": 0.2508028259473346,
      "grad_norm": 0.2782869040966034,
      "learning_rate": 0.0001,
      "loss": 1.5934,
      "step": 1562
    },
    {
      "epoch": 0.25096339113680155,
      "grad_norm": 0.24785871803760529,
      "learning_rate": 0.0001,
      "loss": 1.5717,
      "step": 1563
    },
    {
      "epoch": 0.25112395632626844,
      "grad_norm": 0.257266104221344,
      "learning_rate": 0.0001,
      "loss": 1.5615,
      "step": 1564
    },
    {
      "epoch": 0.2512845215157354,
      "grad_norm": 0.24169054627418518,
      "learning_rate": 0.0001,
      "loss": 1.5768,
      "step": 1565
    },
    {
      "epoch": 0.2514450867052023,
      "grad_norm": 0.2433304488658905,
      "learning_rate": 0.0001,
      "loss": 1.5674,
      "step": 1566
    },
    {
      "epoch": 0.2516056518946692,
      "grad_norm": 0.26393190026283264,
      "learning_rate": 0.0001,
      "loss": 1.6549,
      "step": 1567
    },
    {
      "epoch": 0.25176621708413616,
      "grad_norm": 0.2528260350227356,
      "learning_rate": 0.0001,
      "loss": 1.5326,
      "step": 1568
    },
    {
      "epoch": 0.2519267822736031,
      "grad_norm": 0.2481459379196167,
      "learning_rate": 0.0001,
      "loss": 1.5466,
      "step": 1569
    },
    {
      "epoch": 0.25208734746307,
      "grad_norm": 0.2505651116371155,
      "learning_rate": 0.0001,
      "loss": 1.4812,
      "step": 1570
    },
    {
      "epoch": 0.25224791265253693,
      "grad_norm": 0.24493616819381714,
      "learning_rate": 0.0001,
      "loss": 1.548,
      "step": 1571
    },
    {
      "epoch": 0.2524084778420039,
      "grad_norm": 0.2680916488170624,
      "learning_rate": 0.0001,
      "loss": 1.5595,
      "step": 1572
    },
    {
      "epoch": 0.25256904303147076,
      "grad_norm": 0.239388108253479,
      "learning_rate": 0.0001,
      "loss": 1.5187,
      "step": 1573
    },
    {
      "epoch": 0.2527296082209377,
      "grad_norm": 0.25344353914260864,
      "learning_rate": 0.0001,
      "loss": 1.5244,
      "step": 1574
    },
    {
      "epoch": 0.25289017341040465,
      "grad_norm": 0.23965075612068176,
      "learning_rate": 0.0001,
      "loss": 1.4892,
      "step": 1575
    },
    {
      "epoch": 0.25305073859987154,
      "grad_norm": 0.2578268349170685,
      "learning_rate": 0.0001,
      "loss": 1.5644,
      "step": 1576
    },
    {
      "epoch": 0.2532113037893385,
      "grad_norm": 0.25776371359825134,
      "learning_rate": 0.0001,
      "loss": 1.5863,
      "step": 1577
    },
    {
      "epoch": 0.25337186897880537,
      "grad_norm": 0.2552551031112671,
      "learning_rate": 0.0001,
      "loss": 1.5657,
      "step": 1578
    },
    {
      "epoch": 0.2535324341682723,
      "grad_norm": 0.2545454204082489,
      "learning_rate": 0.0001,
      "loss": 1.5294,
      "step": 1579
    },
    {
      "epoch": 0.25369299935773926,
      "grad_norm": 0.24245326220989227,
      "learning_rate": 0.0001,
      "loss": 1.5041,
      "step": 1580
    },
    {
      "epoch": 0.25385356454720615,
      "grad_norm": 0.2521633207798004,
      "learning_rate": 0.0001,
      "loss": 1.5243,
      "step": 1581
    },
    {
      "epoch": 0.2540141297366731,
      "grad_norm": 0.2599114775657654,
      "learning_rate": 0.0001,
      "loss": 1.5851,
      "step": 1582
    },
    {
      "epoch": 0.25417469492614003,
      "grad_norm": 0.43263909220695496,
      "learning_rate": 0.0001,
      "loss": 1.6567,
      "step": 1583
    },
    {
      "epoch": 0.2543352601156069,
      "grad_norm": 0.24669121205806732,
      "learning_rate": 0.0001,
      "loss": 1.6013,
      "step": 1584
    },
    {
      "epoch": 0.25449582530507386,
      "grad_norm": 0.25480395555496216,
      "learning_rate": 0.0001,
      "loss": 1.5368,
      "step": 1585
    },
    {
      "epoch": 0.2546563904945408,
      "grad_norm": 0.23804716765880585,
      "learning_rate": 0.0001,
      "loss": 1.5947,
      "step": 1586
    },
    {
      "epoch": 0.2548169556840077,
      "grad_norm": 0.25204139947891235,
      "learning_rate": 0.0001,
      "loss": 1.6157,
      "step": 1587
    },
    {
      "epoch": 0.25497752087347464,
      "grad_norm": 0.25400885939598083,
      "learning_rate": 0.0001,
      "loss": 1.5967,
      "step": 1588
    },
    {
      "epoch": 0.2551380860629415,
      "grad_norm": 0.25997045636177063,
      "learning_rate": 0.0001,
      "loss": 1.5763,
      "step": 1589
    },
    {
      "epoch": 0.25529865125240847,
      "grad_norm": 0.24753816425800323,
      "learning_rate": 0.0001,
      "loss": 1.6301,
      "step": 1590
    },
    {
      "epoch": 0.2554592164418754,
      "grad_norm": 0.24728982150554657,
      "learning_rate": 0.0001,
      "loss": 1.5954,
      "step": 1591
    },
    {
      "epoch": 0.2556197816313423,
      "grad_norm": 0.25942108035087585,
      "learning_rate": 0.0001,
      "loss": 1.6172,
      "step": 1592
    },
    {
      "epoch": 0.25578034682080925,
      "grad_norm": 0.23737384378910065,
      "learning_rate": 0.0001,
      "loss": 1.5592,
      "step": 1593
    },
    {
      "epoch": 0.2559409120102762,
      "grad_norm": 0.26012396812438965,
      "learning_rate": 0.0001,
      "loss": 1.4835,
      "step": 1594
    },
    {
      "epoch": 0.2561014771997431,
      "grad_norm": 0.24980948865413666,
      "learning_rate": 0.0001,
      "loss": 1.5963,
      "step": 1595
    },
    {
      "epoch": 0.25626204238921,
      "grad_norm": 0.25194302201271057,
      "learning_rate": 0.0001,
      "loss": 1.4944,
      "step": 1596
    },
    {
      "epoch": 0.25642260757867696,
      "grad_norm": 0.3093951344490051,
      "learning_rate": 0.0001,
      "loss": 1.561,
      "step": 1597
    },
    {
      "epoch": 0.25658317276814385,
      "grad_norm": 0.25668489933013916,
      "learning_rate": 0.0001,
      "loss": 1.598,
      "step": 1598
    },
    {
      "epoch": 0.2567437379576108,
      "grad_norm": 0.25375503301620483,
      "learning_rate": 0.0001,
      "loss": 1.5625,
      "step": 1599
    },
    {
      "epoch": 0.25690430314707774,
      "grad_norm": 0.2650846540927887,
      "learning_rate": 0.0001,
      "loss": 1.5933,
      "step": 1600
    },
    {
      "epoch": 0.2570648683365446,
      "grad_norm": 0.24364729225635529,
      "learning_rate": 0.0001,
      "loss": 1.5039,
      "step": 1601
    },
    {
      "epoch": 0.25722543352601157,
      "grad_norm": 0.241117924451828,
      "learning_rate": 0.0001,
      "loss": 1.549,
      "step": 1602
    },
    {
      "epoch": 0.25738599871547846,
      "grad_norm": 0.2372046262025833,
      "learning_rate": 0.0001,
      "loss": 1.5241,
      "step": 1603
    },
    {
      "epoch": 0.2575465639049454,
      "grad_norm": 0.27673447132110596,
      "learning_rate": 0.0001,
      "loss": 1.6402,
      "step": 1604
    },
    {
      "epoch": 0.25770712909441235,
      "grad_norm": 0.24358339607715607,
      "learning_rate": 0.0001,
      "loss": 1.4948,
      "step": 1605
    },
    {
      "epoch": 0.25786769428387923,
      "grad_norm": 0.2623170018196106,
      "learning_rate": 0.0001,
      "loss": 1.6064,
      "step": 1606
    },
    {
      "epoch": 0.2580282594733462,
      "grad_norm": 0.2529672086238861,
      "learning_rate": 0.0001,
      "loss": 1.5286,
      "step": 1607
    },
    {
      "epoch": 0.2581888246628131,
      "grad_norm": 0.2579810619354248,
      "learning_rate": 0.0001,
      "loss": 1.5211,
      "step": 1608
    },
    {
      "epoch": 0.25834938985228,
      "grad_norm": 26.601112365722656,
      "learning_rate": 0.0001,
      "loss": 1.6601,
      "step": 1609
    },
    {
      "epoch": 0.25850995504174695,
      "grad_norm": 0.2715809643268585,
      "learning_rate": 0.0001,
      "loss": 1.511,
      "step": 1610
    },
    {
      "epoch": 0.2586705202312139,
      "grad_norm": 0.2676183581352234,
      "learning_rate": 0.0001,
      "loss": 1.6073,
      "step": 1611
    },
    {
      "epoch": 0.2588310854206808,
      "grad_norm": 0.3018186688423157,
      "learning_rate": 0.0001,
      "loss": 1.5862,
      "step": 1612
    },
    {
      "epoch": 0.25899165061014773,
      "grad_norm": 0.34301313757896423,
      "learning_rate": 0.0001,
      "loss": 1.5543,
      "step": 1613
    },
    {
      "epoch": 0.2591522157996146,
      "grad_norm": 0.30930280685424805,
      "learning_rate": 0.0001,
      "loss": 1.6061,
      "step": 1614
    },
    {
      "epoch": 0.25931278098908156,
      "grad_norm": 0.29000356793403625,
      "learning_rate": 0.0001,
      "loss": 1.5435,
      "step": 1615
    },
    {
      "epoch": 0.2594733461785485,
      "grad_norm": 0.2772104740142822,
      "learning_rate": 0.0001,
      "loss": 1.6181,
      "step": 1616
    },
    {
      "epoch": 0.2596339113680154,
      "grad_norm": 0.26202407479286194,
      "learning_rate": 0.0001,
      "loss": 1.6247,
      "step": 1617
    },
    {
      "epoch": 0.25979447655748233,
      "grad_norm": 0.2536071240901947,
      "learning_rate": 0.0001,
      "loss": 1.5941,
      "step": 1618
    },
    {
      "epoch": 0.2599550417469493,
      "grad_norm": 0.2397879958152771,
      "learning_rate": 0.0001,
      "loss": 1.5219,
      "step": 1619
    },
    {
      "epoch": 0.26011560693641617,
      "grad_norm": 0.2490512728691101,
      "learning_rate": 0.0001,
      "loss": 1.5562,
      "step": 1620
    },
    {
      "epoch": 0.2602761721258831,
      "grad_norm": 0.24812810122966766,
      "learning_rate": 0.0001,
      "loss": 1.5705,
      "step": 1621
    },
    {
      "epoch": 0.26043673731535005,
      "grad_norm": 0.2508559226989746,
      "learning_rate": 0.0001,
      "loss": 1.6012,
      "step": 1622
    },
    {
      "epoch": 0.26059730250481694,
      "grad_norm": 0.24309803545475006,
      "learning_rate": 0.0001,
      "loss": 1.5945,
      "step": 1623
    },
    {
      "epoch": 0.2607578676942839,
      "grad_norm": 0.2553725242614746,
      "learning_rate": 0.0001,
      "loss": 1.5472,
      "step": 1624
    },
    {
      "epoch": 0.26091843288375083,
      "grad_norm": 0.26383575797080994,
      "learning_rate": 0.0001,
      "loss": 1.5708,
      "step": 1625
    },
    {
      "epoch": 0.2610789980732177,
      "grad_norm": 0.26289352774620056,
      "learning_rate": 0.0001,
      "loss": 1.5839,
      "step": 1626
    },
    {
      "epoch": 0.26123956326268466,
      "grad_norm": 0.24623575806617737,
      "learning_rate": 0.0001,
      "loss": 1.5777,
      "step": 1627
    },
    {
      "epoch": 0.26140012845215155,
      "grad_norm": 0.2563294470310211,
      "learning_rate": 0.0001,
      "loss": 1.5699,
      "step": 1628
    },
    {
      "epoch": 0.2615606936416185,
      "grad_norm": 0.2564566135406494,
      "learning_rate": 0.0001,
      "loss": 1.5754,
      "step": 1629
    },
    {
      "epoch": 0.26172125883108543,
      "grad_norm": 0.26774367690086365,
      "learning_rate": 0.0001,
      "loss": 1.5544,
      "step": 1630
    },
    {
      "epoch": 0.2618818240205523,
      "grad_norm": 0.2636314928531647,
      "learning_rate": 0.0001,
      "loss": 1.642,
      "step": 1631
    },
    {
      "epoch": 0.26204238921001927,
      "grad_norm": 0.2636924386024475,
      "learning_rate": 0.0001,
      "loss": 1.578,
      "step": 1632
    },
    {
      "epoch": 0.2622029543994862,
      "grad_norm": 0.29126644134521484,
      "learning_rate": 0.0001,
      "loss": 1.6165,
      "step": 1633
    },
    {
      "epoch": 0.2623635195889531,
      "grad_norm": 0.24962428212165833,
      "learning_rate": 0.0001,
      "loss": 1.5517,
      "step": 1634
    },
    {
      "epoch": 0.26252408477842004,
      "grad_norm": 0.2517024278640747,
      "learning_rate": 0.0001,
      "loss": 1.5816,
      "step": 1635
    },
    {
      "epoch": 0.262684649967887,
      "grad_norm": 0.25700587034225464,
      "learning_rate": 0.0001,
      "loss": 1.5736,
      "step": 1636
    },
    {
      "epoch": 0.2628452151573539,
      "grad_norm": 0.25166627764701843,
      "learning_rate": 0.0001,
      "loss": 1.5726,
      "step": 1637
    },
    {
      "epoch": 0.2630057803468208,
      "grad_norm": 0.249893918633461,
      "learning_rate": 0.0001,
      "loss": 1.5654,
      "step": 1638
    },
    {
      "epoch": 0.26316634553628776,
      "grad_norm": 0.2570607662200928,
      "learning_rate": 0.0001,
      "loss": 1.5542,
      "step": 1639
    },
    {
      "epoch": 0.26332691072575465,
      "grad_norm": 0.26825419068336487,
      "learning_rate": 0.0001,
      "loss": 1.5636,
      "step": 1640
    },
    {
      "epoch": 0.2634874759152216,
      "grad_norm": 0.25213783979415894,
      "learning_rate": 0.0001,
      "loss": 1.4873,
      "step": 1641
    },
    {
      "epoch": 0.2636480411046885,
      "grad_norm": 0.25579217076301575,
      "learning_rate": 0.0001,
      "loss": 1.5429,
      "step": 1642
    },
    {
      "epoch": 0.2638086062941554,
      "grad_norm": 0.2491072118282318,
      "learning_rate": 0.0001,
      "loss": 1.5657,
      "step": 1643
    },
    {
      "epoch": 0.26396917148362237,
      "grad_norm": 0.24955804646015167,
      "learning_rate": 0.0001,
      "loss": 1.5761,
      "step": 1644
    },
    {
      "epoch": 0.26412973667308925,
      "grad_norm": 0.272416889667511,
      "learning_rate": 0.0001,
      "loss": 1.525,
      "step": 1645
    },
    {
      "epoch": 0.2642903018625562,
      "grad_norm": 1.6191034317016602,
      "learning_rate": 0.0001,
      "loss": 1.6203,
      "step": 1646
    },
    {
      "epoch": 0.26445086705202314,
      "grad_norm": 0.24718856811523438,
      "learning_rate": 0.0001,
      "loss": 1.5627,
      "step": 1647
    },
    {
      "epoch": 0.26461143224149003,
      "grad_norm": 0.23944930732250214,
      "learning_rate": 0.0001,
      "loss": 1.5695,
      "step": 1648
    },
    {
      "epoch": 0.264771997430957,
      "grad_norm": 0.250887006521225,
      "learning_rate": 0.0001,
      "loss": 1.6176,
      "step": 1649
    },
    {
      "epoch": 0.2649325626204239,
      "grad_norm": 0.2513943016529083,
      "learning_rate": 0.0001,
      "loss": 1.5467,
      "step": 1650
    },
    {
      "epoch": 0.2650931278098908,
      "grad_norm": 0.24110578000545502,
      "learning_rate": 0.0001,
      "loss": 1.5811,
      "step": 1651
    },
    {
      "epoch": 0.26525369299935775,
      "grad_norm": 0.24858005344867706,
      "learning_rate": 0.0001,
      "loss": 1.5724,
      "step": 1652
    },
    {
      "epoch": 0.26541425818882464,
      "grad_norm": 0.7188080549240112,
      "learning_rate": 0.0001,
      "loss": 1.5834,
      "step": 1653
    },
    {
      "epoch": 0.2655748233782916,
      "grad_norm": 0.24724483489990234,
      "learning_rate": 0.0001,
      "loss": 1.5319,
      "step": 1654
    },
    {
      "epoch": 0.2657353885677585,
      "grad_norm": 0.310080349445343,
      "learning_rate": 0.0001,
      "loss": 1.5385,
      "step": 1655
    },
    {
      "epoch": 0.2658959537572254,
      "grad_norm": 0.26888203620910645,
      "learning_rate": 0.0001,
      "loss": 1.5919,
      "step": 1656
    },
    {
      "epoch": 0.26605651894669236,
      "grad_norm": 0.2484586238861084,
      "learning_rate": 0.0001,
      "loss": 1.5613,
      "step": 1657
    },
    {
      "epoch": 0.2662170841361593,
      "grad_norm": 0.25118693709373474,
      "learning_rate": 0.0001,
      "loss": 1.5215,
      "step": 1658
    },
    {
      "epoch": 0.2663776493256262,
      "grad_norm": 0.25257745385169983,
      "learning_rate": 0.0001,
      "loss": 1.5993,
      "step": 1659
    },
    {
      "epoch": 0.26653821451509313,
      "grad_norm": 0.2566640079021454,
      "learning_rate": 0.0001,
      "loss": 1.5679,
      "step": 1660
    },
    {
      "epoch": 0.2666987797045601,
      "grad_norm": 0.26186603307724,
      "learning_rate": 0.0001,
      "loss": 1.6129,
      "step": 1661
    },
    {
      "epoch": 0.26685934489402696,
      "grad_norm": 1.3143200874328613,
      "learning_rate": 0.0001,
      "loss": 1.5932,
      "step": 1662
    },
    {
      "epoch": 0.2670199100834939,
      "grad_norm": 0.25016048550605774,
      "learning_rate": 0.0001,
      "loss": 1.6306,
      "step": 1663
    },
    {
      "epoch": 0.26718047527296085,
      "grad_norm": 0.2505226731300354,
      "learning_rate": 0.0001,
      "loss": 1.6446,
      "step": 1664
    },
    {
      "epoch": 0.26734104046242774,
      "grad_norm": 0.2522873878479004,
      "learning_rate": 0.0001,
      "loss": 1.551,
      "step": 1665
    },
    {
      "epoch": 0.2675016056518947,
      "grad_norm": 0.26987871527671814,
      "learning_rate": 0.0001,
      "loss": 1.6022,
      "step": 1666
    },
    {
      "epoch": 0.26766217084136157,
      "grad_norm": 0.25615522265434265,
      "learning_rate": 0.0001,
      "loss": 1.601,
      "step": 1667
    },
    {
      "epoch": 0.2678227360308285,
      "grad_norm": 0.25392067432403564,
      "learning_rate": 0.0001,
      "loss": 1.5457,
      "step": 1668
    },
    {
      "epoch": 0.26798330122029546,
      "grad_norm": 0.2800738513469696,
      "learning_rate": 0.0001,
      "loss": 1.5882,
      "step": 1669
    },
    {
      "epoch": 0.26814386640976234,
      "grad_norm": 0.2578084468841553,
      "learning_rate": 0.0001,
      "loss": 1.5746,
      "step": 1670
    },
    {
      "epoch": 0.2683044315992293,
      "grad_norm": 0.25805479288101196,
      "learning_rate": 0.0001,
      "loss": 1.6026,
      "step": 1671
    },
    {
      "epoch": 0.26846499678869623,
      "grad_norm": 0.24522823095321655,
      "learning_rate": 0.0001,
      "loss": 1.5259,
      "step": 1672
    },
    {
      "epoch": 0.2686255619781631,
      "grad_norm": 0.244820237159729,
      "learning_rate": 0.0001,
      "loss": 1.5385,
      "step": 1673
    },
    {
      "epoch": 0.26878612716763006,
      "grad_norm": 0.26194414496421814,
      "learning_rate": 0.0001,
      "loss": 1.5749,
      "step": 1674
    },
    {
      "epoch": 0.268946692357097,
      "grad_norm": 0.23983000218868256,
      "learning_rate": 0.0001,
      "loss": 1.5383,
      "step": 1675
    },
    {
      "epoch": 0.2691072575465639,
      "grad_norm": 0.24566598236560822,
      "learning_rate": 0.0001,
      "loss": 1.5422,
      "step": 1676
    },
    {
      "epoch": 0.26926782273603084,
      "grad_norm": 0.2699583172798157,
      "learning_rate": 0.0001,
      "loss": 1.567,
      "step": 1677
    },
    {
      "epoch": 0.2694283879254977,
      "grad_norm": 0.2636047899723053,
      "learning_rate": 0.0001,
      "loss": 1.5411,
      "step": 1678
    },
    {
      "epoch": 0.26958895311496467,
      "grad_norm": 0.2641831636428833,
      "learning_rate": 0.0001,
      "loss": 1.6095,
      "step": 1679
    },
    {
      "epoch": 0.2697495183044316,
      "grad_norm": 0.24597400426864624,
      "learning_rate": 0.0001,
      "loss": 1.5763,
      "step": 1680
    },
    {
      "epoch": 0.2699100834938985,
      "grad_norm": 0.24628034234046936,
      "learning_rate": 0.0001,
      "loss": 1.5545,
      "step": 1681
    },
    {
      "epoch": 0.27007064868336544,
      "grad_norm": 0.2375197559595108,
      "learning_rate": 0.0001,
      "loss": 1.4873,
      "step": 1682
    },
    {
      "epoch": 0.2702312138728324,
      "grad_norm": 0.23666463792324066,
      "learning_rate": 0.0001,
      "loss": 1.5143,
      "step": 1683
    },
    {
      "epoch": 0.2703917790622993,
      "grad_norm": 0.24433469772338867,
      "learning_rate": 0.0001,
      "loss": 1.4781,
      "step": 1684
    },
    {
      "epoch": 0.2705523442517662,
      "grad_norm": 0.2497817575931549,
      "learning_rate": 0.0001,
      "loss": 1.5663,
      "step": 1685
    },
    {
      "epoch": 0.27071290944123316,
      "grad_norm": 0.25570669770240784,
      "learning_rate": 0.0001,
      "loss": 1.505,
      "step": 1686
    },
    {
      "epoch": 0.27087347463070005,
      "grad_norm": 0.2590802311897278,
      "learning_rate": 0.0001,
      "loss": 1.5872,
      "step": 1687
    },
    {
      "epoch": 0.271034039820167,
      "grad_norm": 0.25787225365638733,
      "learning_rate": 0.0001,
      "loss": 1.5975,
      "step": 1688
    },
    {
      "epoch": 0.27119460500963394,
      "grad_norm": 0.2446824163198471,
      "learning_rate": 0.0001,
      "loss": 1.5587,
      "step": 1689
    },
    {
      "epoch": 0.2713551701991008,
      "grad_norm": 0.25209805369377136,
      "learning_rate": 0.0001,
      "loss": 1.6,
      "step": 1690
    },
    {
      "epoch": 0.27151573538856777,
      "grad_norm": 0.2606227397918701,
      "learning_rate": 0.0001,
      "loss": 1.5605,
      "step": 1691
    },
    {
      "epoch": 0.27167630057803466,
      "grad_norm": 0.24890798330307007,
      "learning_rate": 0.0001,
      "loss": 1.5679,
      "step": 1692
    },
    {
      "epoch": 0.2718368657675016,
      "grad_norm": 0.247292622923851,
      "learning_rate": 0.0001,
      "loss": 1.591,
      "step": 1693
    },
    {
      "epoch": 0.27199743095696854,
      "grad_norm": 0.2650787830352783,
      "learning_rate": 0.0001,
      "loss": 1.576,
      "step": 1694
    },
    {
      "epoch": 0.27215799614643543,
      "grad_norm": 0.238000750541687,
      "learning_rate": 0.0001,
      "loss": 1.4601,
      "step": 1695
    },
    {
      "epoch": 0.2723185613359024,
      "grad_norm": 0.2528303861618042,
      "learning_rate": 0.0001,
      "loss": 1.552,
      "step": 1696
    },
    {
      "epoch": 0.2724791265253693,
      "grad_norm": 0.28379806876182556,
      "learning_rate": 0.0001,
      "loss": 1.5382,
      "step": 1697
    },
    {
      "epoch": 0.2726396917148362,
      "grad_norm": 0.2615777850151062,
      "learning_rate": 0.0001,
      "loss": 1.5511,
      "step": 1698
    },
    {
      "epoch": 0.27280025690430315,
      "grad_norm": 0.6315487623214722,
      "learning_rate": 0.0001,
      "loss": 1.5045,
      "step": 1699
    },
    {
      "epoch": 0.2729608220937701,
      "grad_norm": 0.2547401487827301,
      "learning_rate": 0.0001,
      "loss": 1.5649,
      "step": 1700
    },
    {
      "epoch": 0.273121387283237,
      "grad_norm": 0.2653486132621765,
      "learning_rate": 0.0001,
      "loss": 1.5767,
      "step": 1701
    },
    {
      "epoch": 0.2732819524727039,
      "grad_norm": 0.25836852192878723,
      "learning_rate": 0.0001,
      "loss": 1.5331,
      "step": 1702
    },
    {
      "epoch": 0.2734425176621708,
      "grad_norm": 0.27013924717903137,
      "learning_rate": 0.0001,
      "loss": 1.5654,
      "step": 1703
    },
    {
      "epoch": 0.27360308285163776,
      "grad_norm": 0.2533862590789795,
      "learning_rate": 0.0001,
      "loss": 1.5766,
      "step": 1704
    },
    {
      "epoch": 0.2737636480411047,
      "grad_norm": 0.2571921646595001,
      "learning_rate": 0.0001,
      "loss": 1.4723,
      "step": 1705
    },
    {
      "epoch": 0.2739242132305716,
      "grad_norm": 0.27870211005210876,
      "learning_rate": 0.0001,
      "loss": 1.5889,
      "step": 1706
    },
    {
      "epoch": 0.27408477842003853,
      "grad_norm": 0.27164745330810547,
      "learning_rate": 0.0001,
      "loss": 1.5685,
      "step": 1707
    },
    {
      "epoch": 0.2742453436095055,
      "grad_norm": 0.2703072130680084,
      "learning_rate": 0.0001,
      "loss": 1.5222,
      "step": 1708
    },
    {
      "epoch": 0.27440590879897236,
      "grad_norm": 0.280581533908844,
      "learning_rate": 0.0001,
      "loss": 1.5795,
      "step": 1709
    },
    {
      "epoch": 0.2745664739884393,
      "grad_norm": 0.25726816058158875,
      "learning_rate": 0.0001,
      "loss": 1.5816,
      "step": 1710
    },
    {
      "epoch": 0.27472703917790625,
      "grad_norm": 0.258029043674469,
      "learning_rate": 0.0001,
      "loss": 1.5867,
      "step": 1711
    },
    {
      "epoch": 0.27488760436737314,
      "grad_norm": 0.28576207160949707,
      "learning_rate": 0.0001,
      "loss": 1.586,
      "step": 1712
    },
    {
      "epoch": 0.2750481695568401,
      "grad_norm": 0.24984319508075714,
      "learning_rate": 0.0001,
      "loss": 1.6312,
      "step": 1713
    },
    {
      "epoch": 0.275208734746307,
      "grad_norm": 0.251816987991333,
      "learning_rate": 0.0001,
      "loss": 1.548,
      "step": 1714
    },
    {
      "epoch": 0.2753692999357739,
      "grad_norm": 0.2735465168952942,
      "learning_rate": 0.0001,
      "loss": 1.6175,
      "step": 1715
    },
    {
      "epoch": 0.27552986512524086,
      "grad_norm": 0.27161598205566406,
      "learning_rate": 0.0001,
      "loss": 1.5866,
      "step": 1716
    },
    {
      "epoch": 0.27569043031470775,
      "grad_norm": 0.254117488861084,
      "learning_rate": 0.0001,
      "loss": 1.5676,
      "step": 1717
    },
    {
      "epoch": 0.2758509955041747,
      "grad_norm": 0.2731083929538727,
      "learning_rate": 0.0001,
      "loss": 1.5807,
      "step": 1718
    },
    {
      "epoch": 0.27601156069364163,
      "grad_norm": 0.25791260600090027,
      "learning_rate": 0.0001,
      "loss": 1.5614,
      "step": 1719
    },
    {
      "epoch": 0.2761721258831085,
      "grad_norm": 0.2510983347892761,
      "learning_rate": 0.0001,
      "loss": 1.5552,
      "step": 1720
    },
    {
      "epoch": 0.27633269107257546,
      "grad_norm": 0.2803308367729187,
      "learning_rate": 0.0001,
      "loss": 1.6478,
      "step": 1721
    },
    {
      "epoch": 0.2764932562620424,
      "grad_norm": 0.2563945949077606,
      "learning_rate": 0.0001,
      "loss": 1.6291,
      "step": 1722
    },
    {
      "epoch": 0.2766538214515093,
      "grad_norm": 0.24843338131904602,
      "learning_rate": 0.0001,
      "loss": 1.5558,
      "step": 1723
    },
    {
      "epoch": 0.27681438664097624,
      "grad_norm": 0.2584153115749359,
      "learning_rate": 0.0001,
      "loss": 1.5683,
      "step": 1724
    },
    {
      "epoch": 0.2769749518304432,
      "grad_norm": 0.24279606342315674,
      "learning_rate": 0.0001,
      "loss": 1.5223,
      "step": 1725
    },
    {
      "epoch": 0.27713551701991007,
      "grad_norm": 0.24455226957798004,
      "learning_rate": 0.0001,
      "loss": 1.5448,
      "step": 1726
    },
    {
      "epoch": 0.277296082209377,
      "grad_norm": 0.2676351070404053,
      "learning_rate": 0.0001,
      "loss": 1.5099,
      "step": 1727
    },
    {
      "epoch": 0.2774566473988439,
      "grad_norm": 0.26381057500839233,
      "learning_rate": 0.0001,
      "loss": 1.5533,
      "step": 1728
    },
    {
      "epoch": 0.27761721258831085,
      "grad_norm": 0.23981605470180511,
      "learning_rate": 0.0001,
      "loss": 1.5411,
      "step": 1729
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.27142682671546936,
      "learning_rate": 0.0001,
      "loss": 1.5469,
      "step": 1730
    },
    {
      "epoch": 0.2779383429672447,
      "grad_norm": 0.2526998519897461,
      "learning_rate": 0.0001,
      "loss": 1.573,
      "step": 1731
    },
    {
      "epoch": 0.2780989081567116,
      "grad_norm": 0.2599260210990906,
      "learning_rate": 0.0001,
      "loss": 1.564,
      "step": 1732
    },
    {
      "epoch": 0.27825947334617857,
      "grad_norm": 0.23529094457626343,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 1733
    },
    {
      "epoch": 0.27842003853564545,
      "grad_norm": 0.2527669370174408,
      "learning_rate": 0.0001,
      "loss": 1.5827,
      "step": 1734
    },
    {
      "epoch": 0.2785806037251124,
      "grad_norm": 0.24101081490516663,
      "learning_rate": 0.0001,
      "loss": 1.526,
      "step": 1735
    },
    {
      "epoch": 0.27874116891457934,
      "grad_norm": 0.25223127007484436,
      "learning_rate": 0.0001,
      "loss": 1.5617,
      "step": 1736
    },
    {
      "epoch": 0.27890173410404623,
      "grad_norm": 0.26755383610725403,
      "learning_rate": 0.0001,
      "loss": 1.6618,
      "step": 1737
    },
    {
      "epoch": 0.27906229929351317,
      "grad_norm": 0.24168036878108978,
      "learning_rate": 0.0001,
      "loss": 1.5689,
      "step": 1738
    },
    {
      "epoch": 0.2792228644829801,
      "grad_norm": 0.2531331479549408,
      "learning_rate": 0.0001,
      "loss": 1.5838,
      "step": 1739
    },
    {
      "epoch": 0.279383429672447,
      "grad_norm": 0.2503527104854584,
      "learning_rate": 0.0001,
      "loss": 1.5343,
      "step": 1740
    },
    {
      "epoch": 0.27954399486191395,
      "grad_norm": 0.2647632956504822,
      "learning_rate": 0.0001,
      "loss": 1.5654,
      "step": 1741
    },
    {
      "epoch": 0.27970456005138083,
      "grad_norm": 0.27943092584609985,
      "learning_rate": 0.0001,
      "loss": 1.6067,
      "step": 1742
    },
    {
      "epoch": 0.2798651252408478,
      "grad_norm": 0.25607460737228394,
      "learning_rate": 0.0001,
      "loss": 1.548,
      "step": 1743
    },
    {
      "epoch": 0.2800256904303147,
      "grad_norm": 0.23967596888542175,
      "learning_rate": 0.0001,
      "loss": 1.4828,
      "step": 1744
    },
    {
      "epoch": 0.2801862556197816,
      "grad_norm": 0.2720800042152405,
      "learning_rate": 0.0001,
      "loss": 1.5305,
      "step": 1745
    },
    {
      "epoch": 0.28034682080924855,
      "grad_norm": 0.23610429465770721,
      "learning_rate": 0.0001,
      "loss": 1.5107,
      "step": 1746
    },
    {
      "epoch": 0.2805073859987155,
      "grad_norm": 0.2505277395248413,
      "learning_rate": 0.0001,
      "loss": 1.525,
      "step": 1747
    },
    {
      "epoch": 0.2806679511881824,
      "grad_norm": 0.26155003905296326,
      "learning_rate": 0.0001,
      "loss": 1.5811,
      "step": 1748
    },
    {
      "epoch": 0.28082851637764933,
      "grad_norm": 0.2556462585926056,
      "learning_rate": 0.0001,
      "loss": 1.5804,
      "step": 1749
    },
    {
      "epoch": 0.2809890815671163,
      "grad_norm": 0.2547653019428253,
      "learning_rate": 0.0001,
      "loss": 1.5572,
      "step": 1750
    },
    {
      "epoch": 0.28114964675658316,
      "grad_norm": 0.2455403059720993,
      "learning_rate": 0.0001,
      "loss": 1.5749,
      "step": 1751
    },
    {
      "epoch": 0.2813102119460501,
      "grad_norm": 0.264738529920578,
      "learning_rate": 0.0001,
      "loss": 1.6097,
      "step": 1752
    },
    {
      "epoch": 0.28147077713551705,
      "grad_norm": 0.23003803193569183,
      "learning_rate": 0.0001,
      "loss": 1.4107,
      "step": 1753
    },
    {
      "epoch": 0.28163134232498394,
      "grad_norm": 0.24369120597839355,
      "learning_rate": 0.0001,
      "loss": 1.5842,
      "step": 1754
    },
    {
      "epoch": 0.2817919075144509,
      "grad_norm": 0.24253031611442566,
      "learning_rate": 0.0001,
      "loss": 1.5105,
      "step": 1755
    },
    {
      "epoch": 0.28195247270391777,
      "grad_norm": 0.2415713667869568,
      "learning_rate": 0.0001,
      "loss": 1.5624,
      "step": 1756
    },
    {
      "epoch": 0.2821130378933847,
      "grad_norm": 0.25316810607910156,
      "learning_rate": 0.0001,
      "loss": 1.5913,
      "step": 1757
    },
    {
      "epoch": 0.28227360308285165,
      "grad_norm": 0.25151219964027405,
      "learning_rate": 0.0001,
      "loss": 1.5772,
      "step": 1758
    },
    {
      "epoch": 0.28243416827231854,
      "grad_norm": 0.2585172951221466,
      "learning_rate": 0.0001,
      "loss": 1.6321,
      "step": 1759
    },
    {
      "epoch": 0.2825947334617855,
      "grad_norm": 0.2640420198440552,
      "learning_rate": 0.0001,
      "loss": 1.6788,
      "step": 1760
    },
    {
      "epoch": 0.28275529865125243,
      "grad_norm": 0.2329113632440567,
      "learning_rate": 0.0001,
      "loss": 1.5101,
      "step": 1761
    },
    {
      "epoch": 0.2829158638407193,
      "grad_norm": 0.2561537027359009,
      "learning_rate": 0.0001,
      "loss": 1.5514,
      "step": 1762
    },
    {
      "epoch": 0.28307642903018626,
      "grad_norm": 0.25183650851249695,
      "learning_rate": 0.0001,
      "loss": 1.6311,
      "step": 1763
    },
    {
      "epoch": 0.2832369942196532,
      "grad_norm": 0.23833124339580536,
      "learning_rate": 0.0001,
      "loss": 1.58,
      "step": 1764
    },
    {
      "epoch": 0.2833975594091201,
      "grad_norm": 0.24028009176254272,
      "learning_rate": 0.0001,
      "loss": 1.5065,
      "step": 1765
    },
    {
      "epoch": 0.28355812459858704,
      "grad_norm": 0.2601412534713745,
      "learning_rate": 0.0001,
      "loss": 1.5283,
      "step": 1766
    },
    {
      "epoch": 0.2837186897880539,
      "grad_norm": 0.24666845798492432,
      "learning_rate": 0.0001,
      "loss": 1.4702,
      "step": 1767
    },
    {
      "epoch": 0.28387925497752087,
      "grad_norm": 0.24733489751815796,
      "learning_rate": 0.0001,
      "loss": 1.5611,
      "step": 1768
    },
    {
      "epoch": 0.2840398201669878,
      "grad_norm": 0.24647156894207,
      "learning_rate": 0.0001,
      "loss": 1.5644,
      "step": 1769
    },
    {
      "epoch": 0.2842003853564547,
      "grad_norm": 0.25269049406051636,
      "learning_rate": 0.0001,
      "loss": 1.5607,
      "step": 1770
    },
    {
      "epoch": 0.28436095054592164,
      "grad_norm": 0.26135051250457764,
      "learning_rate": 0.0001,
      "loss": 1.618,
      "step": 1771
    },
    {
      "epoch": 0.2845215157353886,
      "grad_norm": 0.2525368928909302,
      "learning_rate": 0.0001,
      "loss": 1.4101,
      "step": 1772
    },
    {
      "epoch": 0.2846820809248555,
      "grad_norm": 0.2654610872268677,
      "learning_rate": 0.0001,
      "loss": 1.492,
      "step": 1773
    },
    {
      "epoch": 0.2848426461143224,
      "grad_norm": 0.24932320415973663,
      "learning_rate": 0.0001,
      "loss": 1.4902,
      "step": 1774
    },
    {
      "epoch": 0.28500321130378936,
      "grad_norm": 0.26075899600982666,
      "learning_rate": 0.0001,
      "loss": 1.5023,
      "step": 1775
    },
    {
      "epoch": 0.28516377649325625,
      "grad_norm": 0.2603805363178253,
      "learning_rate": 0.0001,
      "loss": 1.5619,
      "step": 1776
    },
    {
      "epoch": 0.2853243416827232,
      "grad_norm": 0.26587820053100586,
      "learning_rate": 0.0001,
      "loss": 1.5643,
      "step": 1777
    },
    {
      "epoch": 0.28548490687219014,
      "grad_norm": 0.2653835415840149,
      "learning_rate": 0.0001,
      "loss": 1.5833,
      "step": 1778
    },
    {
      "epoch": 0.285645472061657,
      "grad_norm": 0.4035285711288452,
      "learning_rate": 0.0001,
      "loss": 1.5592,
      "step": 1779
    },
    {
      "epoch": 0.28580603725112397,
      "grad_norm": 0.27328649163246155,
      "learning_rate": 0.0001,
      "loss": 1.5187,
      "step": 1780
    },
    {
      "epoch": 0.28596660244059086,
      "grad_norm": 0.2567174732685089,
      "learning_rate": 0.0001,
      "loss": 1.6116,
      "step": 1781
    },
    {
      "epoch": 0.2861271676300578,
      "grad_norm": 0.2559148967266083,
      "learning_rate": 0.0001,
      "loss": 1.5958,
      "step": 1782
    },
    {
      "epoch": 0.28628773281952474,
      "grad_norm": 0.26825404167175293,
      "learning_rate": 0.0001,
      "loss": 1.535,
      "step": 1783
    },
    {
      "epoch": 0.28644829800899163,
      "grad_norm": 0.24423590302467346,
      "learning_rate": 0.0001,
      "loss": 1.5303,
      "step": 1784
    },
    {
      "epoch": 0.2866088631984586,
      "grad_norm": 0.24644210934638977,
      "learning_rate": 0.0001,
      "loss": 1.4775,
      "step": 1785
    },
    {
      "epoch": 0.2867694283879255,
      "grad_norm": 0.2514526844024658,
      "learning_rate": 0.0001,
      "loss": 1.5425,
      "step": 1786
    },
    {
      "epoch": 0.2869299935773924,
      "grad_norm": 0.2657131254673004,
      "learning_rate": 0.0001,
      "loss": 1.5442,
      "step": 1787
    },
    {
      "epoch": 0.28709055876685935,
      "grad_norm": 0.24807108938694,
      "learning_rate": 0.0001,
      "loss": 1.5325,
      "step": 1788
    },
    {
      "epoch": 0.2872511239563263,
      "grad_norm": 0.24604395031929016,
      "learning_rate": 0.0001,
      "loss": 1.6047,
      "step": 1789
    },
    {
      "epoch": 0.2874116891457932,
      "grad_norm": 0.25090017914772034,
      "learning_rate": 0.0001,
      "loss": 1.5903,
      "step": 1790
    },
    {
      "epoch": 0.2875722543352601,
      "grad_norm": 0.2556399703025818,
      "learning_rate": 0.0001,
      "loss": 1.6333,
      "step": 1791
    },
    {
      "epoch": 0.287732819524727,
      "grad_norm": 0.250573068857193,
      "learning_rate": 0.0001,
      "loss": 1.6146,
      "step": 1792
    },
    {
      "epoch": 0.28789338471419396,
      "grad_norm": 0.25975993275642395,
      "learning_rate": 0.0001,
      "loss": 1.6196,
      "step": 1793
    },
    {
      "epoch": 0.2880539499036609,
      "grad_norm": 0.23993436992168427,
      "learning_rate": 0.0001,
      "loss": 1.5714,
      "step": 1794
    },
    {
      "epoch": 0.2882145150931278,
      "grad_norm": 0.2708066701889038,
      "learning_rate": 0.0001,
      "loss": 1.5882,
      "step": 1795
    },
    {
      "epoch": 0.28837508028259473,
      "grad_norm": 0.2533797025680542,
      "learning_rate": 0.0001,
      "loss": 1.6494,
      "step": 1796
    },
    {
      "epoch": 0.2885356454720617,
      "grad_norm": 0.24018733203411102,
      "learning_rate": 0.0001,
      "loss": 1.4955,
      "step": 1797
    },
    {
      "epoch": 0.28869621066152856,
      "grad_norm": 0.25984418392181396,
      "learning_rate": 0.0001,
      "loss": 1.5455,
      "step": 1798
    },
    {
      "epoch": 0.2888567758509955,
      "grad_norm": 0.23806238174438477,
      "learning_rate": 0.0001,
      "loss": 1.5164,
      "step": 1799
    },
    {
      "epoch": 0.28901734104046245,
      "grad_norm": 0.26575228571891785,
      "learning_rate": 0.0001,
      "loss": 1.547,
      "step": 1800
    },
    {
      "epoch": 0.28917790622992934,
      "grad_norm": 0.26070842146873474,
      "learning_rate": 0.0001,
      "loss": 1.6211,
      "step": 1801
    },
    {
      "epoch": 0.2893384714193963,
      "grad_norm": 0.24455119669437408,
      "learning_rate": 0.0001,
      "loss": 1.5171,
      "step": 1802
    },
    {
      "epoch": 0.2894990366088632,
      "grad_norm": 0.25976407527923584,
      "learning_rate": 0.0001,
      "loss": 1.6557,
      "step": 1803
    },
    {
      "epoch": 0.2896596017983301,
      "grad_norm": 0.26272279024124146,
      "learning_rate": 0.0001,
      "loss": 1.497,
      "step": 1804
    },
    {
      "epoch": 0.28982016698779706,
      "grad_norm": 0.25596457719802856,
      "learning_rate": 0.0001,
      "loss": 1.6153,
      "step": 1805
    },
    {
      "epoch": 0.28998073217726394,
      "grad_norm": 0.2553982436656952,
      "learning_rate": 0.0001,
      "loss": 1.5958,
      "step": 1806
    },
    {
      "epoch": 0.2901412973667309,
      "grad_norm": 0.24437467753887177,
      "learning_rate": 0.0001,
      "loss": 1.5436,
      "step": 1807
    },
    {
      "epoch": 0.29030186255619783,
      "grad_norm": 0.25055429339408875,
      "learning_rate": 0.0001,
      "loss": 1.5484,
      "step": 1808
    },
    {
      "epoch": 0.2904624277456647,
      "grad_norm": 0.25160881876945496,
      "learning_rate": 0.0001,
      "loss": 1.5617,
      "step": 1809
    },
    {
      "epoch": 0.29062299293513166,
      "grad_norm": 0.26008692383766174,
      "learning_rate": 0.0001,
      "loss": 1.5801,
      "step": 1810
    },
    {
      "epoch": 0.2907835581245986,
      "grad_norm": 0.2523382902145386,
      "learning_rate": 0.0001,
      "loss": 1.5315,
      "step": 1811
    },
    {
      "epoch": 0.2909441233140655,
      "grad_norm": 0.269346684217453,
      "learning_rate": 0.0001,
      "loss": 1.5889,
      "step": 1812
    },
    {
      "epoch": 0.29110468850353244,
      "grad_norm": 0.24388159811496735,
      "learning_rate": 0.0001,
      "loss": 1.4948,
      "step": 1813
    },
    {
      "epoch": 0.2912652536929994,
      "grad_norm": 0.25000280141830444,
      "learning_rate": 0.0001,
      "loss": 1.5402,
      "step": 1814
    },
    {
      "epoch": 0.29142581888246627,
      "grad_norm": 0.2693331241607666,
      "learning_rate": 0.0001,
      "loss": 1.5676,
      "step": 1815
    },
    {
      "epoch": 0.2915863840719332,
      "grad_norm": 0.25427648425102234,
      "learning_rate": 0.0001,
      "loss": 1.6353,
      "step": 1816
    },
    {
      "epoch": 0.2917469492614001,
      "grad_norm": 0.24013175070285797,
      "learning_rate": 0.0001,
      "loss": 1.5296,
      "step": 1817
    },
    {
      "epoch": 0.29190751445086704,
      "grad_norm": 0.27462080121040344,
      "learning_rate": 0.0001,
      "loss": 1.5787,
      "step": 1818
    },
    {
      "epoch": 0.292068079640334,
      "grad_norm": 0.27167797088623047,
      "learning_rate": 0.0001,
      "loss": 1.5974,
      "step": 1819
    },
    {
      "epoch": 0.2922286448298009,
      "grad_norm": 0.263680100440979,
      "learning_rate": 0.0001,
      "loss": 1.6008,
      "step": 1820
    },
    {
      "epoch": 0.2923892100192678,
      "grad_norm": 0.24112461507320404,
      "learning_rate": 0.0001,
      "loss": 1.5226,
      "step": 1821
    },
    {
      "epoch": 0.29254977520873476,
      "grad_norm": 0.2383682131767273,
      "learning_rate": 0.0001,
      "loss": 1.5073,
      "step": 1822
    },
    {
      "epoch": 0.29271034039820165,
      "grad_norm": 0.23662635684013367,
      "learning_rate": 0.0001,
      "loss": 1.4506,
      "step": 1823
    },
    {
      "epoch": 0.2928709055876686,
      "grad_norm": 0.2591504752635956,
      "learning_rate": 0.0001,
      "loss": 1.5848,
      "step": 1824
    },
    {
      "epoch": 0.29303147077713554,
      "grad_norm": 0.26476243138313293,
      "learning_rate": 0.0001,
      "loss": 1.5316,
      "step": 1825
    },
    {
      "epoch": 0.2931920359666024,
      "grad_norm": 0.2454085797071457,
      "learning_rate": 0.0001,
      "loss": 1.5379,
      "step": 1826
    },
    {
      "epoch": 0.29335260115606937,
      "grad_norm": 0.2538471817970276,
      "learning_rate": 0.0001,
      "loss": 1.6104,
      "step": 1827
    },
    {
      "epoch": 0.2935131663455363,
      "grad_norm": 0.2743769586086273,
      "learning_rate": 0.0001,
      "loss": 1.6508,
      "step": 1828
    },
    {
      "epoch": 0.2936737315350032,
      "grad_norm": 0.25146836042404175,
      "learning_rate": 0.0001,
      "loss": 1.5778,
      "step": 1829
    },
    {
      "epoch": 0.29383429672447015,
      "grad_norm": 0.2563507556915283,
      "learning_rate": 0.0001,
      "loss": 1.5977,
      "step": 1830
    },
    {
      "epoch": 0.29399486191393703,
      "grad_norm": 0.2630648612976074,
      "learning_rate": 0.0001,
      "loss": 1.5948,
      "step": 1831
    },
    {
      "epoch": 0.294155427103404,
      "grad_norm": 0.26110172271728516,
      "learning_rate": 0.0001,
      "loss": 1.5926,
      "step": 1832
    },
    {
      "epoch": 0.2943159922928709,
      "grad_norm": 0.2517088055610657,
      "learning_rate": 0.0001,
      "loss": 1.5772,
      "step": 1833
    },
    {
      "epoch": 0.2944765574823378,
      "grad_norm": 0.24928121268749237,
      "learning_rate": 0.0001,
      "loss": 1.6135,
      "step": 1834
    },
    {
      "epoch": 0.29463712267180475,
      "grad_norm": 0.23833926022052765,
      "learning_rate": 0.0001,
      "loss": 1.5628,
      "step": 1835
    },
    {
      "epoch": 0.2947976878612717,
      "grad_norm": 0.24982048571109772,
      "learning_rate": 0.0001,
      "loss": 1.491,
      "step": 1836
    },
    {
      "epoch": 0.2949582530507386,
      "grad_norm": 0.3186301290988922,
      "learning_rate": 0.0001,
      "loss": 1.6437,
      "step": 1837
    },
    {
      "epoch": 0.2951188182402055,
      "grad_norm": 0.2510720491409302,
      "learning_rate": 0.0001,
      "loss": 1.5703,
      "step": 1838
    },
    {
      "epoch": 0.29527938342967247,
      "grad_norm": 0.2615893483161926,
      "learning_rate": 0.0001,
      "loss": 1.591,
      "step": 1839
    },
    {
      "epoch": 0.29543994861913936,
      "grad_norm": 0.25472602248191833,
      "learning_rate": 0.0001,
      "loss": 1.5373,
      "step": 1840
    },
    {
      "epoch": 0.2956005138086063,
      "grad_norm": 0.2813533842563629,
      "learning_rate": 0.0001,
      "loss": 1.508,
      "step": 1841
    },
    {
      "epoch": 0.2957610789980732,
      "grad_norm": 0.25984275341033936,
      "learning_rate": 0.0001,
      "loss": 1.5794,
      "step": 1842
    },
    {
      "epoch": 0.29592164418754013,
      "grad_norm": 0.25177472829818726,
      "learning_rate": 0.0001,
      "loss": 1.5174,
      "step": 1843
    },
    {
      "epoch": 0.2960822093770071,
      "grad_norm": 0.24683226644992828,
      "learning_rate": 0.0001,
      "loss": 1.4671,
      "step": 1844
    },
    {
      "epoch": 0.29624277456647397,
      "grad_norm": 0.26041093468666077,
      "learning_rate": 0.0001,
      "loss": 1.6242,
      "step": 1845
    },
    {
      "epoch": 0.2964033397559409,
      "grad_norm": 0.24409382045269012,
      "learning_rate": 0.0001,
      "loss": 1.5268,
      "step": 1846
    },
    {
      "epoch": 0.29656390494540785,
      "grad_norm": 0.25203752517700195,
      "learning_rate": 0.0001,
      "loss": 1.5728,
      "step": 1847
    },
    {
      "epoch": 0.29672447013487474,
      "grad_norm": 0.27876532077789307,
      "learning_rate": 0.0001,
      "loss": 1.5732,
      "step": 1848
    },
    {
      "epoch": 0.2968850353243417,
      "grad_norm": 0.28675970435142517,
      "learning_rate": 0.0001,
      "loss": 1.5728,
      "step": 1849
    },
    {
      "epoch": 0.2970456005138086,
      "grad_norm": 0.2579909563064575,
      "learning_rate": 0.0001,
      "loss": 1.5657,
      "step": 1850
    },
    {
      "epoch": 0.2972061657032755,
      "grad_norm": 0.26867666840553284,
      "learning_rate": 0.0001,
      "loss": 1.5831,
      "step": 1851
    },
    {
      "epoch": 0.29736673089274246,
      "grad_norm": 0.24972176551818848,
      "learning_rate": 0.0001,
      "loss": 1.5103,
      "step": 1852
    },
    {
      "epoch": 0.2975272960822094,
      "grad_norm": 0.2798141539096832,
      "learning_rate": 0.0001,
      "loss": 1.6262,
      "step": 1853
    },
    {
      "epoch": 0.2976878612716763,
      "grad_norm": 0.27341991662979126,
      "learning_rate": 0.0001,
      "loss": 1.6127,
      "step": 1854
    },
    {
      "epoch": 0.29784842646114323,
      "grad_norm": 0.28879499435424805,
      "learning_rate": 0.0001,
      "loss": 1.5381,
      "step": 1855
    },
    {
      "epoch": 0.2980089916506101,
      "grad_norm": 0.2521066963672638,
      "learning_rate": 0.0001,
      "loss": 1.5628,
      "step": 1856
    },
    {
      "epoch": 0.29816955684007707,
      "grad_norm": 0.24854888021945953,
      "learning_rate": 0.0001,
      "loss": 1.548,
      "step": 1857
    },
    {
      "epoch": 0.298330122029544,
      "grad_norm": 0.2579093873500824,
      "learning_rate": 0.0001,
      "loss": 1.5704,
      "step": 1858
    },
    {
      "epoch": 0.2984906872190109,
      "grad_norm": 0.26755714416503906,
      "learning_rate": 0.0001,
      "loss": 1.6673,
      "step": 1859
    },
    {
      "epoch": 0.29865125240847784,
      "grad_norm": 0.2619021236896515,
      "learning_rate": 0.0001,
      "loss": 1.5102,
      "step": 1860
    },
    {
      "epoch": 0.2988118175979448,
      "grad_norm": 0.2696104347705841,
      "learning_rate": 0.0001,
      "loss": 1.4678,
      "step": 1861
    },
    {
      "epoch": 0.2989723827874117,
      "grad_norm": 0.2595376670360565,
      "learning_rate": 0.0001,
      "loss": 1.5342,
      "step": 1862
    },
    {
      "epoch": 0.2991329479768786,
      "grad_norm": 0.2659249007701874,
      "learning_rate": 0.0001,
      "loss": 1.5667,
      "step": 1863
    },
    {
      "epoch": 0.29929351316634556,
      "grad_norm": 0.27641892433166504,
      "learning_rate": 0.0001,
      "loss": 1.7052,
      "step": 1864
    },
    {
      "epoch": 0.29945407835581245,
      "grad_norm": 0.24762192368507385,
      "learning_rate": 0.0001,
      "loss": 1.5204,
      "step": 1865
    },
    {
      "epoch": 0.2996146435452794,
      "grad_norm": 0.25400760769844055,
      "learning_rate": 0.0001,
      "loss": 1.5343,
      "step": 1866
    },
    {
      "epoch": 0.29977520873474633,
      "grad_norm": 0.24175044894218445,
      "learning_rate": 0.0001,
      "loss": 1.6044,
      "step": 1867
    },
    {
      "epoch": 0.2999357739242132,
      "grad_norm": 0.24022619426250458,
      "learning_rate": 0.0001,
      "loss": 1.5563,
      "step": 1868
    },
    {
      "epoch": 0.30009633911368017,
      "grad_norm": 0.25152572989463806,
      "learning_rate": 0.0001,
      "loss": 1.5692,
      "step": 1869
    },
    {
      "epoch": 0.30025690430314705,
      "grad_norm": 0.26082396507263184,
      "learning_rate": 0.0001,
      "loss": 1.5929,
      "step": 1870
    },
    {
      "epoch": 0.300417469492614,
      "grad_norm": 0.28792721033096313,
      "learning_rate": 0.0001,
      "loss": 1.5404,
      "step": 1871
    },
    {
      "epoch": 0.30057803468208094,
      "grad_norm": 0.25064149498939514,
      "learning_rate": 0.0001,
      "loss": 1.6066,
      "step": 1872
    },
    {
      "epoch": 0.30073859987154783,
      "grad_norm": 0.2479681670665741,
      "learning_rate": 0.0001,
      "loss": 1.5462,
      "step": 1873
    },
    {
      "epoch": 0.3008991650610148,
      "grad_norm": 0.6160130500793457,
      "learning_rate": 0.0001,
      "loss": 1.5941,
      "step": 1874
    },
    {
      "epoch": 0.3010597302504817,
      "grad_norm": 0.2493082731962204,
      "learning_rate": 0.0001,
      "loss": 1.5703,
      "step": 1875
    },
    {
      "epoch": 0.3012202954399486,
      "grad_norm": 0.23921550810337067,
      "learning_rate": 0.0001,
      "loss": 1.4552,
      "step": 1876
    },
    {
      "epoch": 0.30138086062941555,
      "grad_norm": 0.2838565409183502,
      "learning_rate": 0.0001,
      "loss": 1.5857,
      "step": 1877
    },
    {
      "epoch": 0.3015414258188825,
      "grad_norm": 0.268409788608551,
      "learning_rate": 0.0001,
      "loss": 1.5909,
      "step": 1878
    },
    {
      "epoch": 0.3017019910083494,
      "grad_norm": 0.27479058504104614,
      "learning_rate": 0.0001,
      "loss": 1.5766,
      "step": 1879
    },
    {
      "epoch": 0.3018625561978163,
      "grad_norm": 0.24375787377357483,
      "learning_rate": 0.0001,
      "loss": 1.514,
      "step": 1880
    },
    {
      "epoch": 0.3020231213872832,
      "grad_norm": 0.24530580639839172,
      "learning_rate": 0.0001,
      "loss": 1.5464,
      "step": 1881
    },
    {
      "epoch": 0.30218368657675015,
      "grad_norm": 0.2485753893852234,
      "learning_rate": 0.0001,
      "loss": 1.5329,
      "step": 1882
    },
    {
      "epoch": 0.3023442517662171,
      "grad_norm": 0.24819721281528473,
      "learning_rate": 0.0001,
      "loss": 1.5531,
      "step": 1883
    },
    {
      "epoch": 0.302504816955684,
      "grad_norm": 0.27060645818710327,
      "learning_rate": 0.0001,
      "loss": 1.6018,
      "step": 1884
    },
    {
      "epoch": 0.30266538214515093,
      "grad_norm": 0.2594221532344818,
      "learning_rate": 0.0001,
      "loss": 1.4978,
      "step": 1885
    },
    {
      "epoch": 0.3028259473346179,
      "grad_norm": 0.23679520189762115,
      "learning_rate": 0.0001,
      "loss": 1.486,
      "step": 1886
    },
    {
      "epoch": 0.30298651252408476,
      "grad_norm": 0.24668680131435394,
      "learning_rate": 0.0001,
      "loss": 1.4878,
      "step": 1887
    },
    {
      "epoch": 0.3031470777135517,
      "grad_norm": 0.24873335659503937,
      "learning_rate": 0.0001,
      "loss": 1.4877,
      "step": 1888
    },
    {
      "epoch": 0.30330764290301865,
      "grad_norm": 0.25774168968200684,
      "learning_rate": 0.0001,
      "loss": 1.5324,
      "step": 1889
    },
    {
      "epoch": 0.30346820809248554,
      "grad_norm": 0.25117889046669006,
      "learning_rate": 0.0001,
      "loss": 1.5499,
      "step": 1890
    },
    {
      "epoch": 0.3036287732819525,
      "grad_norm": 0.2548123896121979,
      "learning_rate": 0.0001,
      "loss": 1.5137,
      "step": 1891
    },
    {
      "epoch": 0.3037893384714194,
      "grad_norm": 0.23779501020908356,
      "learning_rate": 0.0001,
      "loss": 1.5505,
      "step": 1892
    },
    {
      "epoch": 0.3039499036608863,
      "grad_norm": 0.24799437820911407,
      "learning_rate": 0.0001,
      "loss": 1.5272,
      "step": 1893
    },
    {
      "epoch": 0.30411046885035325,
      "grad_norm": 0.2357570379972458,
      "learning_rate": 0.0001,
      "loss": 1.5371,
      "step": 1894
    },
    {
      "epoch": 0.30427103403982014,
      "grad_norm": 0.24633687734603882,
      "learning_rate": 0.0001,
      "loss": 1.5216,
      "step": 1895
    },
    {
      "epoch": 0.3044315992292871,
      "grad_norm": 0.2573365271091461,
      "learning_rate": 0.0001,
      "loss": 1.608,
      "step": 1896
    },
    {
      "epoch": 0.30459216441875403,
      "grad_norm": 0.24100179970264435,
      "learning_rate": 0.0001,
      "loss": 1.4889,
      "step": 1897
    },
    {
      "epoch": 0.3047527296082209,
      "grad_norm": 0.24734078347682953,
      "learning_rate": 0.0001,
      "loss": 1.455,
      "step": 1898
    },
    {
      "epoch": 0.30491329479768786,
      "grad_norm": 0.25962188839912415,
      "learning_rate": 0.0001,
      "loss": 1.6056,
      "step": 1899
    },
    {
      "epoch": 0.3050738599871548,
      "grad_norm": 0.2501012682914734,
      "learning_rate": 0.0001,
      "loss": 1.538,
      "step": 1900
    },
    {
      "epoch": 0.3052344251766217,
      "grad_norm": 0.25188666582107544,
      "learning_rate": 0.0001,
      "loss": 1.612,
      "step": 1901
    },
    {
      "epoch": 0.30539499036608864,
      "grad_norm": 0.25298765301704407,
      "learning_rate": 0.0001,
      "loss": 1.6036,
      "step": 1902
    },
    {
      "epoch": 0.3055555555555556,
      "grad_norm": 0.268204003572464,
      "learning_rate": 0.0001,
      "loss": 1.6405,
      "step": 1903
    },
    {
      "epoch": 0.30571612074502247,
      "grad_norm": 1.586869239807129,
      "learning_rate": 0.0001,
      "loss": 1.573,
      "step": 1904
    },
    {
      "epoch": 0.3058766859344894,
      "grad_norm": 0.24489398300647736,
      "learning_rate": 0.0001,
      "loss": 1.5984,
      "step": 1905
    },
    {
      "epoch": 0.3060372511239563,
      "grad_norm": 0.2484893947839737,
      "learning_rate": 0.0001,
      "loss": 1.543,
      "step": 1906
    },
    {
      "epoch": 0.30619781631342324,
      "grad_norm": 0.23659148812294006,
      "learning_rate": 0.0001,
      "loss": 1.4638,
      "step": 1907
    },
    {
      "epoch": 0.3063583815028902,
      "grad_norm": 0.238989919424057,
      "learning_rate": 0.0001,
      "loss": 1.4976,
      "step": 1908
    },
    {
      "epoch": 0.3065189466923571,
      "grad_norm": 0.24608288705348969,
      "learning_rate": 0.0001,
      "loss": 1.6556,
      "step": 1909
    },
    {
      "epoch": 0.306679511881824,
      "grad_norm": 0.24990175664424896,
      "learning_rate": 0.0001,
      "loss": 1.5045,
      "step": 1910
    },
    {
      "epoch": 0.30684007707129096,
      "grad_norm": 0.2556162178516388,
      "learning_rate": 0.0001,
      "loss": 1.6116,
      "step": 1911
    },
    {
      "epoch": 0.30700064226075785,
      "grad_norm": 0.24361644685268402,
      "learning_rate": 0.0001,
      "loss": 1.5561,
      "step": 1912
    },
    {
      "epoch": 0.3071612074502248,
      "grad_norm": 0.2393457591533661,
      "learning_rate": 0.0001,
      "loss": 1.5945,
      "step": 1913
    },
    {
      "epoch": 0.30732177263969174,
      "grad_norm": 0.2425204962491989,
      "learning_rate": 0.0001,
      "loss": 1.5432,
      "step": 1914
    },
    {
      "epoch": 0.3074823378291586,
      "grad_norm": 0.25019174814224243,
      "learning_rate": 0.0001,
      "loss": 1.6036,
      "step": 1915
    },
    {
      "epoch": 0.30764290301862557,
      "grad_norm": 0.24608489871025085,
      "learning_rate": 0.0001,
      "loss": 1.5638,
      "step": 1916
    },
    {
      "epoch": 0.3078034682080925,
      "grad_norm": 0.25412517786026,
      "learning_rate": 0.0001,
      "loss": 1.5885,
      "step": 1917
    },
    {
      "epoch": 0.3079640333975594,
      "grad_norm": 0.247113436460495,
      "learning_rate": 0.0001,
      "loss": 1.4763,
      "step": 1918
    },
    {
      "epoch": 0.30812459858702634,
      "grad_norm": 0.24102932214736938,
      "learning_rate": 0.0001,
      "loss": 1.5049,
      "step": 1919
    },
    {
      "epoch": 0.30828516377649323,
      "grad_norm": 1.6683555841445923,
      "learning_rate": 0.0001,
      "loss": 1.5579,
      "step": 1920
    },
    {
      "epoch": 0.3084457289659602,
      "grad_norm": 0.2455155998468399,
      "learning_rate": 0.0001,
      "loss": 1.5239,
      "step": 1921
    },
    {
      "epoch": 0.3086062941554271,
      "grad_norm": 0.26336005330085754,
      "learning_rate": 0.0001,
      "loss": 1.4679,
      "step": 1922
    },
    {
      "epoch": 0.308766859344894,
      "grad_norm": 0.26241132616996765,
      "learning_rate": 0.0001,
      "loss": 1.5808,
      "step": 1923
    },
    {
      "epoch": 0.30892742453436095,
      "grad_norm": 0.2679413855075836,
      "learning_rate": 0.0001,
      "loss": 1.5523,
      "step": 1924
    },
    {
      "epoch": 0.3090879897238279,
      "grad_norm": 0.2441592514514923,
      "learning_rate": 0.0001,
      "loss": 1.5253,
      "step": 1925
    },
    {
      "epoch": 0.3092485549132948,
      "grad_norm": 0.2611231505870819,
      "learning_rate": 0.0001,
      "loss": 1.5624,
      "step": 1926
    },
    {
      "epoch": 0.3094091201027617,
      "grad_norm": 0.26747897267341614,
      "learning_rate": 0.0001,
      "loss": 1.5377,
      "step": 1927
    },
    {
      "epoch": 0.30956968529222867,
      "grad_norm": 0.28684693574905396,
      "learning_rate": 0.0001,
      "loss": 1.5811,
      "step": 1928
    },
    {
      "epoch": 0.30973025048169556,
      "grad_norm": 0.2699681222438812,
      "learning_rate": 0.0001,
      "loss": 1.5805,
      "step": 1929
    },
    {
      "epoch": 0.3098908156711625,
      "grad_norm": 0.2753625810146332,
      "learning_rate": 0.0001,
      "loss": 1.5864,
      "step": 1930
    },
    {
      "epoch": 0.3100513808606294,
      "grad_norm": 0.2829086482524872,
      "learning_rate": 0.0001,
      "loss": 1.6383,
      "step": 1931
    },
    {
      "epoch": 0.31021194605009633,
      "grad_norm": 0.27932077646255493,
      "learning_rate": 0.0001,
      "loss": 1.5027,
      "step": 1932
    },
    {
      "epoch": 0.3103725112395633,
      "grad_norm": 0.2526090145111084,
      "learning_rate": 0.0001,
      "loss": 1.4831,
      "step": 1933
    },
    {
      "epoch": 0.31053307642903016,
      "grad_norm": 0.27693843841552734,
      "learning_rate": 0.0001,
      "loss": 1.6333,
      "step": 1934
    },
    {
      "epoch": 0.3106936416184971,
      "grad_norm": 0.2676289677619934,
      "learning_rate": 0.0001,
      "loss": 1.5865,
      "step": 1935
    },
    {
      "epoch": 0.31085420680796405,
      "grad_norm": 0.2501024603843689,
      "learning_rate": 0.0001,
      "loss": 1.6406,
      "step": 1936
    },
    {
      "epoch": 0.31101477199743094,
      "grad_norm": 0.31236693263053894,
      "learning_rate": 0.0001,
      "loss": 1.5404,
      "step": 1937
    },
    {
      "epoch": 0.3111753371868979,
      "grad_norm": 0.2553372085094452,
      "learning_rate": 0.0001,
      "loss": 1.5008,
      "step": 1938
    },
    {
      "epoch": 0.3113359023763648,
      "grad_norm": 0.247478649020195,
      "learning_rate": 0.0001,
      "loss": 1.5338,
      "step": 1939
    },
    {
      "epoch": 0.3114964675658317,
      "grad_norm": 0.2544630467891693,
      "learning_rate": 0.0001,
      "loss": 1.5287,
      "step": 1940
    },
    {
      "epoch": 0.31165703275529866,
      "grad_norm": 0.2486809939146042,
      "learning_rate": 0.0001,
      "loss": 1.5386,
      "step": 1941
    },
    {
      "epoch": 0.3118175979447656,
      "grad_norm": 0.26548662781715393,
      "learning_rate": 0.0001,
      "loss": 1.4916,
      "step": 1942
    },
    {
      "epoch": 0.3119781631342325,
      "grad_norm": 0.26302793622016907,
      "learning_rate": 0.0001,
      "loss": 1.6307,
      "step": 1943
    },
    {
      "epoch": 0.31213872832369943,
      "grad_norm": 0.2523314952850342,
      "learning_rate": 0.0001,
      "loss": 1.5184,
      "step": 1944
    },
    {
      "epoch": 0.3122992935131663,
      "grad_norm": 0.2504449188709259,
      "learning_rate": 0.0001,
      "loss": 1.4592,
      "step": 1945
    },
    {
      "epoch": 0.31245985870263326,
      "grad_norm": 0.2648429870605469,
      "learning_rate": 0.0001,
      "loss": 1.6783,
      "step": 1946
    },
    {
      "epoch": 0.3126204238921002,
      "grad_norm": 0.27570006251335144,
      "learning_rate": 0.0001,
      "loss": 1.5943,
      "step": 1947
    },
    {
      "epoch": 0.3127809890815671,
      "grad_norm": 0.2519308924674988,
      "learning_rate": 0.0001,
      "loss": 1.544,
      "step": 1948
    },
    {
      "epoch": 0.31294155427103404,
      "grad_norm": 0.26972678303718567,
      "learning_rate": 0.0001,
      "loss": 1.5713,
      "step": 1949
    },
    {
      "epoch": 0.313102119460501,
      "grad_norm": 0.2651619017124176,
      "learning_rate": 0.0001,
      "loss": 1.4749,
      "step": 1950
    },
    {
      "epoch": 0.31326268464996787,
      "grad_norm": 0.2575755715370178,
      "learning_rate": 0.0001,
      "loss": 1.5378,
      "step": 1951
    },
    {
      "epoch": 0.3134232498394348,
      "grad_norm": 0.2738771140575409,
      "learning_rate": 0.0001,
      "loss": 1.5779,
      "step": 1952
    },
    {
      "epoch": 0.31358381502890176,
      "grad_norm": 0.2623549699783325,
      "learning_rate": 0.0001,
      "loss": 1.5892,
      "step": 1953
    },
    {
      "epoch": 0.31374438021836865,
      "grad_norm": 0.26148900389671326,
      "learning_rate": 0.0001,
      "loss": 1.4955,
      "step": 1954
    },
    {
      "epoch": 0.3139049454078356,
      "grad_norm": 0.26656344532966614,
      "learning_rate": 0.0001,
      "loss": 1.5875,
      "step": 1955
    },
    {
      "epoch": 0.3140655105973025,
      "grad_norm": 0.2636772394180298,
      "learning_rate": 0.0001,
      "loss": 1.5877,
      "step": 1956
    },
    {
      "epoch": 0.3142260757867694,
      "grad_norm": 0.26439574360847473,
      "learning_rate": 0.0001,
      "loss": 1.5937,
      "step": 1957
    },
    {
      "epoch": 0.31438664097623636,
      "grad_norm": 0.3177263140678406,
      "learning_rate": 0.0001,
      "loss": 1.4055,
      "step": 1958
    },
    {
      "epoch": 0.31454720616570325,
      "grad_norm": 0.31999045610427856,
      "learning_rate": 0.0001,
      "loss": 1.5173,
      "step": 1959
    },
    {
      "epoch": 0.3147077713551702,
      "grad_norm": 0.28947094082832336,
      "learning_rate": 0.0001,
      "loss": 1.6825,
      "step": 1960
    },
    {
      "epoch": 0.31486833654463714,
      "grad_norm": 0.26300951838493347,
      "learning_rate": 0.0001,
      "loss": 1.5699,
      "step": 1961
    },
    {
      "epoch": 0.315028901734104,
      "grad_norm": 0.31268611550331116,
      "learning_rate": 0.0001,
      "loss": 1.6053,
      "step": 1962
    },
    {
      "epoch": 0.31518946692357097,
      "grad_norm": 0.31239306926727295,
      "learning_rate": 0.0001,
      "loss": 1.4967,
      "step": 1963
    },
    {
      "epoch": 0.3153500321130379,
      "grad_norm": 0.4538760483264923,
      "learning_rate": 0.0001,
      "loss": 1.5428,
      "step": 1964
    },
    {
      "epoch": 0.3155105973025048,
      "grad_norm": 0.2723877429962158,
      "learning_rate": 0.0001,
      "loss": 1.5532,
      "step": 1965
    },
    {
      "epoch": 0.31567116249197175,
      "grad_norm": 0.273652583360672,
      "learning_rate": 0.0001,
      "loss": 1.5211,
      "step": 1966
    },
    {
      "epoch": 0.3158317276814387,
      "grad_norm": 0.2829582691192627,
      "learning_rate": 0.0001,
      "loss": 1.5397,
      "step": 1967
    },
    {
      "epoch": 0.3159922928709056,
      "grad_norm": 0.28165462613105774,
      "learning_rate": 0.0001,
      "loss": 1.6203,
      "step": 1968
    },
    {
      "epoch": 0.3161528580603725,
      "grad_norm": 0.27137845754623413,
      "learning_rate": 0.0001,
      "loss": 1.5079,
      "step": 1969
    },
    {
      "epoch": 0.3163134232498394,
      "grad_norm": 0.28451213240623474,
      "learning_rate": 0.0001,
      "loss": 1.5272,
      "step": 1970
    },
    {
      "epoch": 0.31647398843930635,
      "grad_norm": 0.2487514317035675,
      "learning_rate": 0.0001,
      "loss": 1.5149,
      "step": 1971
    },
    {
      "epoch": 0.3166345536287733,
      "grad_norm": 0.2665540277957916,
      "learning_rate": 0.0001,
      "loss": 1.584,
      "step": 1972
    },
    {
      "epoch": 0.3167951188182402,
      "grad_norm": 0.29046347737312317,
      "learning_rate": 0.0001,
      "loss": 1.579,
      "step": 1973
    },
    {
      "epoch": 0.31695568400770713,
      "grad_norm": 0.25909698009490967,
      "learning_rate": 0.0001,
      "loss": 1.6118,
      "step": 1974
    },
    {
      "epoch": 0.31711624919717407,
      "grad_norm": 0.24843347072601318,
      "learning_rate": 0.0001,
      "loss": 1.5208,
      "step": 1975
    },
    {
      "epoch": 0.31727681438664096,
      "grad_norm": 0.24475157260894775,
      "learning_rate": 0.0001,
      "loss": 1.5602,
      "step": 1976
    },
    {
      "epoch": 0.3174373795761079,
      "grad_norm": 0.2558642327785492,
      "learning_rate": 0.0001,
      "loss": 1.5627,
      "step": 1977
    },
    {
      "epoch": 0.31759794476557485,
      "grad_norm": 0.26439404487609863,
      "learning_rate": 0.0001,
      "loss": 1.5986,
      "step": 1978
    },
    {
      "epoch": 0.31775850995504173,
      "grad_norm": 0.24407227337360382,
      "learning_rate": 0.0001,
      "loss": 1.5732,
      "step": 1979
    },
    {
      "epoch": 0.3179190751445087,
      "grad_norm": 0.28595051169395447,
      "learning_rate": 0.0001,
      "loss": 1.6268,
      "step": 1980
    },
    {
      "epoch": 0.3180796403339756,
      "grad_norm": 0.26790329813957214,
      "learning_rate": 0.0001,
      "loss": 1.6188,
      "step": 1981
    },
    {
      "epoch": 0.3182402055234425,
      "grad_norm": 0.33726316690444946,
      "learning_rate": 0.0001,
      "loss": 1.5407,
      "step": 1982
    },
    {
      "epoch": 0.31840077071290945,
      "grad_norm": 0.2506847083568573,
      "learning_rate": 0.0001,
      "loss": 1.5907,
      "step": 1983
    },
    {
      "epoch": 0.31856133590237634,
      "grad_norm": 0.26663774251937866,
      "learning_rate": 0.0001,
      "loss": 1.5612,
      "step": 1984
    },
    {
      "epoch": 0.3187219010918433,
      "grad_norm": 0.2664209008216858,
      "learning_rate": 0.0001,
      "loss": 1.5583,
      "step": 1985
    },
    {
      "epoch": 0.31888246628131023,
      "grad_norm": 0.2599117159843445,
      "learning_rate": 0.0001,
      "loss": 1.5069,
      "step": 1986
    },
    {
      "epoch": 0.3190430314707771,
      "grad_norm": 0.25136682391166687,
      "learning_rate": 0.0001,
      "loss": 1.5309,
      "step": 1987
    },
    {
      "epoch": 0.31920359666024406,
      "grad_norm": 0.2607286870479584,
      "learning_rate": 0.0001,
      "loss": 1.5507,
      "step": 1988
    },
    {
      "epoch": 0.319364161849711,
      "grad_norm": 0.25644397735595703,
      "learning_rate": 0.0001,
      "loss": 1.5686,
      "step": 1989
    },
    {
      "epoch": 0.3195247270391779,
      "grad_norm": 0.25758540630340576,
      "learning_rate": 0.0001,
      "loss": 1.4948,
      "step": 1990
    },
    {
      "epoch": 0.31968529222864484,
      "grad_norm": 0.24424684047698975,
      "learning_rate": 0.0001,
      "loss": 1.4676,
      "step": 1991
    },
    {
      "epoch": 0.3198458574181118,
      "grad_norm": 2.369795560836792,
      "learning_rate": 0.0001,
      "loss": 1.6539,
      "step": 1992
    },
    {
      "epoch": 0.32000642260757867,
      "grad_norm": 0.2553209364414215,
      "learning_rate": 0.0001,
      "loss": 1.5653,
      "step": 1993
    },
    {
      "epoch": 0.3201669877970456,
      "grad_norm": 0.25574493408203125,
      "learning_rate": 0.0001,
      "loss": 1.5519,
      "step": 1994
    },
    {
      "epoch": 0.3203275529865125,
      "grad_norm": 0.24983198940753937,
      "learning_rate": 0.0001,
      "loss": 1.638,
      "step": 1995
    },
    {
      "epoch": 0.32048811817597944,
      "grad_norm": 0.24426138401031494,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 1996
    },
    {
      "epoch": 0.3206486833654464,
      "grad_norm": 0.2396833598613739,
      "learning_rate": 0.0001,
      "loss": 1.5439,
      "step": 1997
    },
    {
      "epoch": 0.3208092485549133,
      "grad_norm": 0.2616952359676361,
      "learning_rate": 0.0001,
      "loss": 1.532,
      "step": 1998
    },
    {
      "epoch": 0.3209698137443802,
      "grad_norm": 0.2405225783586502,
      "learning_rate": 0.0001,
      "loss": 1.5721,
      "step": 1999
    },
    {
      "epoch": 0.32113037893384716,
      "grad_norm": 0.24469563364982605,
      "learning_rate": 0.0001,
      "loss": 1.5501,
      "step": 2000
    },
    {
      "epoch": 0.32129094412331405,
      "grad_norm": 0.24045726656913757,
      "learning_rate": 0.0001,
      "loss": 1.4864,
      "step": 2001
    },
    {
      "epoch": 0.321451509312781,
      "grad_norm": 0.23491191864013672,
      "learning_rate": 0.0001,
      "loss": 1.5456,
      "step": 2002
    },
    {
      "epoch": 0.32161207450224794,
      "grad_norm": 0.25913017988204956,
      "learning_rate": 0.0001,
      "loss": 1.5277,
      "step": 2003
    },
    {
      "epoch": 0.3217726396917148,
      "grad_norm": 0.24957261979579926,
      "learning_rate": 0.0001,
      "loss": 1.5388,
      "step": 2004
    },
    {
      "epoch": 0.32193320488118177,
      "grad_norm": 0.26871004700660706,
      "learning_rate": 0.0001,
      "loss": 1.5986,
      "step": 2005
    },
    {
      "epoch": 0.3220937700706487,
      "grad_norm": 0.25158387422561646,
      "learning_rate": 0.0001,
      "loss": 1.5362,
      "step": 2006
    },
    {
      "epoch": 0.3222543352601156,
      "grad_norm": 0.25004780292510986,
      "learning_rate": 0.0001,
      "loss": 1.5961,
      "step": 2007
    },
    {
      "epoch": 0.32241490044958254,
      "grad_norm": 0.2508402466773987,
      "learning_rate": 0.0001,
      "loss": 1.5096,
      "step": 2008
    },
    {
      "epoch": 0.32257546563904943,
      "grad_norm": 0.2602085471153259,
      "learning_rate": 0.0001,
      "loss": 1.5506,
      "step": 2009
    },
    {
      "epoch": 0.3227360308285164,
      "grad_norm": 0.24283818900585175,
      "learning_rate": 0.0001,
      "loss": 1.4731,
      "step": 2010
    },
    {
      "epoch": 0.3228965960179833,
      "grad_norm": 0.2539307177066803,
      "learning_rate": 0.0001,
      "loss": 1.551,
      "step": 2011
    },
    {
      "epoch": 0.3230571612074502,
      "grad_norm": 0.2527778148651123,
      "learning_rate": 0.0001,
      "loss": 1.6136,
      "step": 2012
    },
    {
      "epoch": 0.32321772639691715,
      "grad_norm": 0.2855991721153259,
      "learning_rate": 0.0001,
      "loss": 1.66,
      "step": 2013
    },
    {
      "epoch": 0.3233782915863841,
      "grad_norm": 0.2550317645072937,
      "learning_rate": 0.0001,
      "loss": 1.6189,
      "step": 2014
    },
    {
      "epoch": 0.323538856775851,
      "grad_norm": 0.2562001049518585,
      "learning_rate": 0.0001,
      "loss": 1.5835,
      "step": 2015
    },
    {
      "epoch": 0.3236994219653179,
      "grad_norm": 0.2569708526134491,
      "learning_rate": 0.0001,
      "loss": 1.5579,
      "step": 2016
    },
    {
      "epoch": 0.32385998715478487,
      "grad_norm": 0.25206586718559265,
      "learning_rate": 0.0001,
      "loss": 1.4494,
      "step": 2017
    },
    {
      "epoch": 0.32402055234425176,
      "grad_norm": 0.260757178068161,
      "learning_rate": 0.0001,
      "loss": 1.5939,
      "step": 2018
    },
    {
      "epoch": 0.3241811175337187,
      "grad_norm": 0.24333572387695312,
      "learning_rate": 0.0001,
      "loss": 1.5233,
      "step": 2019
    },
    {
      "epoch": 0.3243416827231856,
      "grad_norm": 0.2474576234817505,
      "learning_rate": 0.0001,
      "loss": 1.531,
      "step": 2020
    },
    {
      "epoch": 0.32450224791265253,
      "grad_norm": 0.24919936060905457,
      "learning_rate": 0.0001,
      "loss": 1.5104,
      "step": 2021
    },
    {
      "epoch": 0.3246628131021195,
      "grad_norm": 0.27407601475715637,
      "learning_rate": 0.0001,
      "loss": 1.5779,
      "step": 2022
    },
    {
      "epoch": 0.32482337829158636,
      "grad_norm": 0.2472204566001892,
      "learning_rate": 0.0001,
      "loss": 1.548,
      "step": 2023
    },
    {
      "epoch": 0.3249839434810533,
      "grad_norm": 0.2495601326227188,
      "learning_rate": 0.0001,
      "loss": 1.5391,
      "step": 2024
    },
    {
      "epoch": 0.32514450867052025,
      "grad_norm": 0.34387463331222534,
      "learning_rate": 0.0001,
      "loss": 1.477,
      "step": 2025
    },
    {
      "epoch": 0.32530507385998714,
      "grad_norm": 0.2577751874923706,
      "learning_rate": 0.0001,
      "loss": 1.5778,
      "step": 2026
    },
    {
      "epoch": 0.3254656390494541,
      "grad_norm": 0.2544448971748352,
      "learning_rate": 0.0001,
      "loss": 1.4909,
      "step": 2027
    },
    {
      "epoch": 0.325626204238921,
      "grad_norm": 0.26194360852241516,
      "learning_rate": 0.0001,
      "loss": 1.4971,
      "step": 2028
    },
    {
      "epoch": 0.3257867694283879,
      "grad_norm": 0.2775893211364746,
      "learning_rate": 0.0001,
      "loss": 1.57,
      "step": 2029
    },
    {
      "epoch": 0.32594733461785486,
      "grad_norm": 0.24274244904518127,
      "learning_rate": 0.0001,
      "loss": 1.4804,
      "step": 2030
    },
    {
      "epoch": 0.3261078998073218,
      "grad_norm": 0.25908997654914856,
      "learning_rate": 0.0001,
      "loss": 1.5899,
      "step": 2031
    },
    {
      "epoch": 0.3262684649967887,
      "grad_norm": 0.24661113321781158,
      "learning_rate": 0.0001,
      "loss": 1.533,
      "step": 2032
    },
    {
      "epoch": 0.32642903018625563,
      "grad_norm": 0.2571931481361389,
      "learning_rate": 0.0001,
      "loss": 1.5652,
      "step": 2033
    },
    {
      "epoch": 0.3265895953757225,
      "grad_norm": 0.27937066555023193,
      "learning_rate": 0.0001,
      "loss": 1.5215,
      "step": 2034
    },
    {
      "epoch": 0.32675016056518946,
      "grad_norm": 0.25084108114242554,
      "learning_rate": 0.0001,
      "loss": 1.5691,
      "step": 2035
    },
    {
      "epoch": 0.3269107257546564,
      "grad_norm": 0.2674304246902466,
      "learning_rate": 0.0001,
      "loss": 1.6143,
      "step": 2036
    },
    {
      "epoch": 0.3270712909441233,
      "grad_norm": 0.2760143578052521,
      "learning_rate": 0.0001,
      "loss": 1.6586,
      "step": 2037
    },
    {
      "epoch": 0.32723185613359024,
      "grad_norm": 0.2676118016242981,
      "learning_rate": 0.0001,
      "loss": 1.5496,
      "step": 2038
    },
    {
      "epoch": 0.3273924213230572,
      "grad_norm": 0.2591191232204437,
      "learning_rate": 0.0001,
      "loss": 1.5747,
      "step": 2039
    },
    {
      "epoch": 0.32755298651252407,
      "grad_norm": 0.24934707581996918,
      "learning_rate": 0.0001,
      "loss": 1.5478,
      "step": 2040
    },
    {
      "epoch": 0.327713551701991,
      "grad_norm": 0.24062830209732056,
      "learning_rate": 0.0001,
      "loss": 1.5283,
      "step": 2041
    },
    {
      "epoch": 0.32787411689145796,
      "grad_norm": 0.2677200138568878,
      "learning_rate": 0.0001,
      "loss": 1.5722,
      "step": 2042
    },
    {
      "epoch": 0.32803468208092484,
      "grad_norm": 0.25619253516197205,
      "learning_rate": 0.0001,
      "loss": 1.5749,
      "step": 2043
    },
    {
      "epoch": 0.3281952472703918,
      "grad_norm": 0.25234487652778625,
      "learning_rate": 0.0001,
      "loss": 1.5149,
      "step": 2044
    },
    {
      "epoch": 0.3283558124598587,
      "grad_norm": 0.24501624703407288,
      "learning_rate": 0.0001,
      "loss": 1.5608,
      "step": 2045
    },
    {
      "epoch": 0.3285163776493256,
      "grad_norm": 0.2601943910121918,
      "learning_rate": 0.0001,
      "loss": 1.5929,
      "step": 2046
    },
    {
      "epoch": 0.32867694283879256,
      "grad_norm": 0.2691039443016052,
      "learning_rate": 0.0001,
      "loss": 1.6367,
      "step": 2047
    },
    {
      "epoch": 0.32883750802825945,
      "grad_norm": 0.2641131281852722,
      "learning_rate": 0.0001,
      "loss": 1.5494,
      "step": 2048
    },
    {
      "epoch": 0.3289980732177264,
      "grad_norm": 0.24564731121063232,
      "learning_rate": 0.0001,
      "loss": 1.5653,
      "step": 2049
    },
    {
      "epoch": 0.32915863840719334,
      "grad_norm": 0.24584510922431946,
      "learning_rate": 0.0001,
      "loss": 1.6153,
      "step": 2050
    },
    {
      "epoch": 0.3293192035966602,
      "grad_norm": 0.25819748640060425,
      "learning_rate": 0.0001,
      "loss": 1.4515,
      "step": 2051
    },
    {
      "epoch": 0.32947976878612717,
      "grad_norm": 0.25013402104377747,
      "learning_rate": 0.0001,
      "loss": 1.5023,
      "step": 2052
    },
    {
      "epoch": 0.3296403339755941,
      "grad_norm": 0.2627883553504944,
      "learning_rate": 0.0001,
      "loss": 1.5698,
      "step": 2053
    },
    {
      "epoch": 0.329800899165061,
      "grad_norm": 0.2595508098602295,
      "learning_rate": 0.0001,
      "loss": 1.5887,
      "step": 2054
    },
    {
      "epoch": 0.32996146435452794,
      "grad_norm": 0.2511318027973175,
      "learning_rate": 0.0001,
      "loss": 1.5807,
      "step": 2055
    },
    {
      "epoch": 0.3301220295439949,
      "grad_norm": 0.28229302167892456,
      "learning_rate": 0.0001,
      "loss": 1.5293,
      "step": 2056
    },
    {
      "epoch": 0.3302825947334618,
      "grad_norm": 0.2682824730873108,
      "learning_rate": 0.0001,
      "loss": 1.5524,
      "step": 2057
    },
    {
      "epoch": 0.3304431599229287,
      "grad_norm": 0.2473607063293457,
      "learning_rate": 0.0001,
      "loss": 1.5423,
      "step": 2058
    },
    {
      "epoch": 0.3306037251123956,
      "grad_norm": 0.26974809169769287,
      "learning_rate": 0.0001,
      "loss": 1.6149,
      "step": 2059
    },
    {
      "epoch": 0.33076429030186255,
      "grad_norm": 0.26088446378707886,
      "learning_rate": 0.0001,
      "loss": 1.5512,
      "step": 2060
    },
    {
      "epoch": 0.3309248554913295,
      "grad_norm": 0.42597144842147827,
      "learning_rate": 0.0001,
      "loss": 1.5373,
      "step": 2061
    },
    {
      "epoch": 0.3310854206807964,
      "grad_norm": 0.25390321016311646,
      "learning_rate": 0.0001,
      "loss": 1.5602,
      "step": 2062
    },
    {
      "epoch": 0.3312459858702633,
      "grad_norm": 0.2506279945373535,
      "learning_rate": 0.0001,
      "loss": 1.4774,
      "step": 2063
    },
    {
      "epoch": 0.33140655105973027,
      "grad_norm": 0.2644400894641876,
      "learning_rate": 0.0001,
      "loss": 1.5258,
      "step": 2064
    },
    {
      "epoch": 0.33156711624919716,
      "grad_norm": 0.27889135479927063,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 2065
    },
    {
      "epoch": 0.3317276814386641,
      "grad_norm": 0.2611287832260132,
      "learning_rate": 0.0001,
      "loss": 1.5797,
      "step": 2066
    },
    {
      "epoch": 0.33188824662813105,
      "grad_norm": 0.26141950488090515,
      "learning_rate": 0.0001,
      "loss": 1.5697,
      "step": 2067
    },
    {
      "epoch": 0.33204881181759793,
      "grad_norm": 0.2811838984489441,
      "learning_rate": 0.0001,
      "loss": 1.5889,
      "step": 2068
    },
    {
      "epoch": 0.3322093770070649,
      "grad_norm": 0.2627982497215271,
      "learning_rate": 0.0001,
      "loss": 1.598,
      "step": 2069
    },
    {
      "epoch": 0.33236994219653176,
      "grad_norm": 0.24998469650745392,
      "learning_rate": 0.0001,
      "loss": 1.5854,
      "step": 2070
    },
    {
      "epoch": 0.3325305073859987,
      "grad_norm": 0.260353147983551,
      "learning_rate": 0.0001,
      "loss": 1.618,
      "step": 2071
    },
    {
      "epoch": 0.33269107257546565,
      "grad_norm": 0.250887393951416,
      "learning_rate": 0.0001,
      "loss": 1.4957,
      "step": 2072
    },
    {
      "epoch": 0.33285163776493254,
      "grad_norm": 0.2872818112373352,
      "learning_rate": 0.0001,
      "loss": 1.5865,
      "step": 2073
    },
    {
      "epoch": 0.3330122029543995,
      "grad_norm": 0.24831518530845642,
      "learning_rate": 0.0001,
      "loss": 1.4169,
      "step": 2074
    },
    {
      "epoch": 0.3331727681438664,
      "grad_norm": 0.2671929597854614,
      "learning_rate": 0.0001,
      "loss": 1.5842,
      "step": 2075
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.2549666166305542,
      "learning_rate": 0.0001,
      "loss": 1.4667,
      "step": 2076
    },
    {
      "epoch": 0.33349389852280026,
      "grad_norm": 0.25841936469078064,
      "learning_rate": 0.0001,
      "loss": 1.5821,
      "step": 2077
    },
    {
      "epoch": 0.3336544637122672,
      "grad_norm": 0.2738978862762451,
      "learning_rate": 0.0001,
      "loss": 1.6261,
      "step": 2078
    },
    {
      "epoch": 0.3338150289017341,
      "grad_norm": 0.2430199831724167,
      "learning_rate": 0.0001,
      "loss": 1.5608,
      "step": 2079
    },
    {
      "epoch": 0.33397559409120103,
      "grad_norm": 0.24916157126426697,
      "learning_rate": 0.0001,
      "loss": 1.5206,
      "step": 2080
    },
    {
      "epoch": 0.334136159280668,
      "grad_norm": 0.2479800283908844,
      "learning_rate": 0.0001,
      "loss": 1.6286,
      "step": 2081
    },
    {
      "epoch": 0.33429672447013487,
      "grad_norm": 0.25704678893089294,
      "learning_rate": 0.0001,
      "loss": 1.5186,
      "step": 2082
    },
    {
      "epoch": 0.3344572896596018,
      "grad_norm": 0.24137870967388153,
      "learning_rate": 0.0001,
      "loss": 1.6058,
      "step": 2083
    },
    {
      "epoch": 0.3346178548490687,
      "grad_norm": 0.26002371311187744,
      "learning_rate": 0.0001,
      "loss": 1.6203,
      "step": 2084
    },
    {
      "epoch": 0.33477842003853564,
      "grad_norm": 0.2356281876564026,
      "learning_rate": 0.0001,
      "loss": 1.5415,
      "step": 2085
    },
    {
      "epoch": 0.3349389852280026,
      "grad_norm": 0.26551684737205505,
      "learning_rate": 0.0001,
      "loss": 1.6309,
      "step": 2086
    },
    {
      "epoch": 0.33509955041746947,
      "grad_norm": 0.250840425491333,
      "learning_rate": 0.0001,
      "loss": 1.5145,
      "step": 2087
    },
    {
      "epoch": 0.3352601156069364,
      "grad_norm": 0.25762513279914856,
      "learning_rate": 0.0001,
      "loss": 1.5706,
      "step": 2088
    },
    {
      "epoch": 0.33542068079640336,
      "grad_norm": 0.24373485147953033,
      "learning_rate": 0.0001,
      "loss": 1.5207,
      "step": 2089
    },
    {
      "epoch": 0.33558124598587025,
      "grad_norm": 0.3432878851890564,
      "learning_rate": 0.0001,
      "loss": 1.6484,
      "step": 2090
    },
    {
      "epoch": 0.3357418111753372,
      "grad_norm": 0.25250113010406494,
      "learning_rate": 0.0001,
      "loss": 1.6051,
      "step": 2091
    },
    {
      "epoch": 0.33590237636480413,
      "grad_norm": 0.24866685271263123,
      "learning_rate": 0.0001,
      "loss": 1.5427,
      "step": 2092
    },
    {
      "epoch": 0.336062941554271,
      "grad_norm": 0.2524840235710144,
      "learning_rate": 0.0001,
      "loss": 1.5751,
      "step": 2093
    },
    {
      "epoch": 0.33622350674373797,
      "grad_norm": 0.2584114670753479,
      "learning_rate": 0.0001,
      "loss": 1.6,
      "step": 2094
    },
    {
      "epoch": 0.33638407193320485,
      "grad_norm": 0.2565653920173645,
      "learning_rate": 0.0001,
      "loss": 1.5776,
      "step": 2095
    },
    {
      "epoch": 0.3365446371226718,
      "grad_norm": 0.26194992661476135,
      "learning_rate": 0.0001,
      "loss": 1.5842,
      "step": 2096
    },
    {
      "epoch": 0.33670520231213874,
      "grad_norm": 0.25302010774612427,
      "learning_rate": 0.0001,
      "loss": 1.5818,
      "step": 2097
    },
    {
      "epoch": 0.33686576750160563,
      "grad_norm": 0.2598724365234375,
      "learning_rate": 0.0001,
      "loss": 1.6093,
      "step": 2098
    },
    {
      "epoch": 0.33702633269107257,
      "grad_norm": 0.2545692026615143,
      "learning_rate": 0.0001,
      "loss": 1.4404,
      "step": 2099
    },
    {
      "epoch": 0.3371868978805395,
      "grad_norm": 0.24554964900016785,
      "learning_rate": 0.0001,
      "loss": 1.4851,
      "step": 2100
    },
    {
      "epoch": 0.3373474630700064,
      "grad_norm": 0.25128766894340515,
      "learning_rate": 0.0001,
      "loss": 1.5647,
      "step": 2101
    },
    {
      "epoch": 0.33750802825947335,
      "grad_norm": 0.2542661428451538,
      "learning_rate": 0.0001,
      "loss": 1.5352,
      "step": 2102
    },
    {
      "epoch": 0.3376685934489403,
      "grad_norm": 0.26056137681007385,
      "learning_rate": 0.0001,
      "loss": 1.6635,
      "step": 2103
    },
    {
      "epoch": 0.3378291586384072,
      "grad_norm": 0.2550312280654907,
      "learning_rate": 0.0001,
      "loss": 1.5557,
      "step": 2104
    },
    {
      "epoch": 0.3379897238278741,
      "grad_norm": 0.2502883970737457,
      "learning_rate": 0.0001,
      "loss": 1.5526,
      "step": 2105
    },
    {
      "epoch": 0.33815028901734107,
      "grad_norm": 0.25496160984039307,
      "learning_rate": 0.0001,
      "loss": 1.5514,
      "step": 2106
    },
    {
      "epoch": 0.33831085420680795,
      "grad_norm": 0.2540586292743683,
      "learning_rate": 0.0001,
      "loss": 1.5898,
      "step": 2107
    },
    {
      "epoch": 0.3384714193962749,
      "grad_norm": 0.25860700011253357,
      "learning_rate": 0.0001,
      "loss": 1.5184,
      "step": 2108
    },
    {
      "epoch": 0.3386319845857418,
      "grad_norm": 0.2346867173910141,
      "learning_rate": 0.0001,
      "loss": 1.4983,
      "step": 2109
    },
    {
      "epoch": 0.33879254977520873,
      "grad_norm": 0.2642350196838379,
      "learning_rate": 0.0001,
      "loss": 1.6796,
      "step": 2110
    },
    {
      "epoch": 0.3389531149646757,
      "grad_norm": 0.23771946132183075,
      "learning_rate": 0.0001,
      "loss": 1.5507,
      "step": 2111
    },
    {
      "epoch": 0.33911368015414256,
      "grad_norm": 0.24657626450061798,
      "learning_rate": 0.0001,
      "loss": 1.5134,
      "step": 2112
    },
    {
      "epoch": 0.3392742453436095,
      "grad_norm": 0.27198708057403564,
      "learning_rate": 0.0001,
      "loss": 1.5508,
      "step": 2113
    },
    {
      "epoch": 0.33943481053307645,
      "grad_norm": 0.2621020972728729,
      "learning_rate": 0.0001,
      "loss": 1.6011,
      "step": 2114
    },
    {
      "epoch": 0.33959537572254334,
      "grad_norm": 0.27582138776779175,
      "learning_rate": 0.0001,
      "loss": 1.6013,
      "step": 2115
    },
    {
      "epoch": 0.3397559409120103,
      "grad_norm": 0.2563203275203705,
      "learning_rate": 0.0001,
      "loss": 1.4894,
      "step": 2116
    },
    {
      "epoch": 0.3399165061014772,
      "grad_norm": 0.2562795877456665,
      "learning_rate": 0.0001,
      "loss": 1.581,
      "step": 2117
    },
    {
      "epoch": 0.3400770712909441,
      "grad_norm": 0.26509010791778564,
      "learning_rate": 0.0001,
      "loss": 1.5237,
      "step": 2118
    },
    {
      "epoch": 0.34023763648041105,
      "grad_norm": 0.25600680708885193,
      "learning_rate": 0.0001,
      "loss": 1.6093,
      "step": 2119
    },
    {
      "epoch": 0.340398201669878,
      "grad_norm": 0.2511206269264221,
      "learning_rate": 0.0001,
      "loss": 1.5343,
      "step": 2120
    },
    {
      "epoch": 0.3405587668593449,
      "grad_norm": 0.24978840351104736,
      "learning_rate": 0.0001,
      "loss": 1.5208,
      "step": 2121
    },
    {
      "epoch": 0.34071933204881183,
      "grad_norm": 0.25395306944847107,
      "learning_rate": 0.0001,
      "loss": 1.6241,
      "step": 2122
    },
    {
      "epoch": 0.3408798972382787,
      "grad_norm": 0.26245689392089844,
      "learning_rate": 0.0001,
      "loss": 1.5836,
      "step": 2123
    },
    {
      "epoch": 0.34104046242774566,
      "grad_norm": 0.25019681453704834,
      "learning_rate": 0.0001,
      "loss": 1.5727,
      "step": 2124
    },
    {
      "epoch": 0.3412010276172126,
      "grad_norm": 0.2606225609779358,
      "learning_rate": 0.0001,
      "loss": 1.5617,
      "step": 2125
    },
    {
      "epoch": 0.3413615928066795,
      "grad_norm": 0.24390485882759094,
      "learning_rate": 0.0001,
      "loss": 1.5801,
      "step": 2126
    },
    {
      "epoch": 0.34152215799614644,
      "grad_norm": 0.24285078048706055,
      "learning_rate": 0.0001,
      "loss": 1.5468,
      "step": 2127
    },
    {
      "epoch": 0.3416827231856134,
      "grad_norm": 0.5630396008491516,
      "learning_rate": 0.0001,
      "loss": 1.5057,
      "step": 2128
    },
    {
      "epoch": 0.34184328837508027,
      "grad_norm": 0.2508193552494049,
      "learning_rate": 0.0001,
      "loss": 1.5735,
      "step": 2129
    },
    {
      "epoch": 0.3420038535645472,
      "grad_norm": 0.24452221393585205,
      "learning_rate": 0.0001,
      "loss": 1.5448,
      "step": 2130
    },
    {
      "epoch": 0.34216441875401415,
      "grad_norm": 0.24207033216953278,
      "learning_rate": 0.0001,
      "loss": 1.5064,
      "step": 2131
    },
    {
      "epoch": 0.34232498394348104,
      "grad_norm": 0.2509983777999878,
      "learning_rate": 0.0001,
      "loss": 1.5166,
      "step": 2132
    },
    {
      "epoch": 0.342485549132948,
      "grad_norm": 0.33120840787887573,
      "learning_rate": 0.0001,
      "loss": 1.5876,
      "step": 2133
    },
    {
      "epoch": 0.3426461143224149,
      "grad_norm": 0.26178523898124695,
      "learning_rate": 0.0001,
      "loss": 1.6726,
      "step": 2134
    },
    {
      "epoch": 0.3428066795118818,
      "grad_norm": 0.2442718893289566,
      "learning_rate": 0.0001,
      "loss": 1.5507,
      "step": 2135
    },
    {
      "epoch": 0.34296724470134876,
      "grad_norm": 0.2566917836666107,
      "learning_rate": 0.0001,
      "loss": 1.5711,
      "step": 2136
    },
    {
      "epoch": 0.34312780989081565,
      "grad_norm": 0.26134663820266724,
      "learning_rate": 0.0001,
      "loss": 1.5366,
      "step": 2137
    },
    {
      "epoch": 0.3432883750802826,
      "grad_norm": 0.2464238554239273,
      "learning_rate": 0.0001,
      "loss": 1.5535,
      "step": 2138
    },
    {
      "epoch": 0.34344894026974954,
      "grad_norm": 0.26522767543792725,
      "learning_rate": 0.0001,
      "loss": 1.519,
      "step": 2139
    },
    {
      "epoch": 0.3436095054592164,
      "grad_norm": 0.25250810384750366,
      "learning_rate": 0.0001,
      "loss": 1.5288,
      "step": 2140
    },
    {
      "epoch": 0.34377007064868337,
      "grad_norm": 0.2520618736743927,
      "learning_rate": 0.0001,
      "loss": 1.5671,
      "step": 2141
    },
    {
      "epoch": 0.3439306358381503,
      "grad_norm": 0.24800851941108704,
      "learning_rate": 0.0001,
      "loss": 1.5362,
      "step": 2142
    },
    {
      "epoch": 0.3440912010276172,
      "grad_norm": 0.24773597717285156,
      "learning_rate": 0.0001,
      "loss": 1.4873,
      "step": 2143
    },
    {
      "epoch": 0.34425176621708414,
      "grad_norm": 0.25473371148109436,
      "learning_rate": 0.0001,
      "loss": 1.5516,
      "step": 2144
    },
    {
      "epoch": 0.3444123314065511,
      "grad_norm": 0.24132798612117767,
      "learning_rate": 0.0001,
      "loss": 1.507,
      "step": 2145
    },
    {
      "epoch": 0.344572896596018,
      "grad_norm": 0.24645288288593292,
      "learning_rate": 0.0001,
      "loss": 1.6099,
      "step": 2146
    },
    {
      "epoch": 0.3447334617854849,
      "grad_norm": 0.24814775586128235,
      "learning_rate": 0.0001,
      "loss": 1.5343,
      "step": 2147
    },
    {
      "epoch": 0.3448940269749518,
      "grad_norm": 0.25247395038604736,
      "learning_rate": 0.0001,
      "loss": 1.5251,
      "step": 2148
    },
    {
      "epoch": 0.34505459216441875,
      "grad_norm": 0.26199835538864136,
      "learning_rate": 0.0001,
      "loss": 1.5467,
      "step": 2149
    },
    {
      "epoch": 0.3452151573538857,
      "grad_norm": 0.2571370601654053,
      "learning_rate": 0.0001,
      "loss": 1.564,
      "step": 2150
    },
    {
      "epoch": 0.3453757225433526,
      "grad_norm": 0.27830779552459717,
      "learning_rate": 0.0001,
      "loss": 1.5887,
      "step": 2151
    },
    {
      "epoch": 0.3455362877328195,
      "grad_norm": 0.2565908432006836,
      "learning_rate": 0.0001,
      "loss": 1.574,
      "step": 2152
    },
    {
      "epoch": 0.34569685292228647,
      "grad_norm": 0.2541201412677765,
      "learning_rate": 0.0001,
      "loss": 1.4921,
      "step": 2153
    },
    {
      "epoch": 0.34585741811175336,
      "grad_norm": 0.25625211000442505,
      "learning_rate": 0.0001,
      "loss": 1.5355,
      "step": 2154
    },
    {
      "epoch": 0.3460179833012203,
      "grad_norm": 0.24762476980686188,
      "learning_rate": 0.0001,
      "loss": 1.5887,
      "step": 2155
    },
    {
      "epoch": 0.34617854849068724,
      "grad_norm": 0.25672975182533264,
      "learning_rate": 0.0001,
      "loss": 1.5681,
      "step": 2156
    },
    {
      "epoch": 0.34633911368015413,
      "grad_norm": 0.24759668111801147,
      "learning_rate": 0.0001,
      "loss": 1.5086,
      "step": 2157
    },
    {
      "epoch": 0.3464996788696211,
      "grad_norm": 0.2625550627708435,
      "learning_rate": 0.0001,
      "loss": 1.5708,
      "step": 2158
    },
    {
      "epoch": 0.34666024405908796,
      "grad_norm": 0.25980886816978455,
      "learning_rate": 0.0001,
      "loss": 1.5546,
      "step": 2159
    },
    {
      "epoch": 0.3468208092485549,
      "grad_norm": 0.2516760230064392,
      "learning_rate": 0.0001,
      "loss": 1.5672,
      "step": 2160
    },
    {
      "epoch": 0.34698137443802185,
      "grad_norm": 0.2552073299884796,
      "learning_rate": 0.0001,
      "loss": 1.5915,
      "step": 2161
    },
    {
      "epoch": 0.34714193962748874,
      "grad_norm": 0.26119092106819153,
      "learning_rate": 0.0001,
      "loss": 1.5021,
      "step": 2162
    },
    {
      "epoch": 0.3473025048169557,
      "grad_norm": 0.2548002302646637,
      "learning_rate": 0.0001,
      "loss": 1.5178,
      "step": 2163
    },
    {
      "epoch": 0.3474630700064226,
      "grad_norm": 0.24852673709392548,
      "learning_rate": 0.0001,
      "loss": 1.5639,
      "step": 2164
    },
    {
      "epoch": 0.3476236351958895,
      "grad_norm": 0.24551789462566376,
      "learning_rate": 0.0001,
      "loss": 1.6229,
      "step": 2165
    },
    {
      "epoch": 0.34778420038535646,
      "grad_norm": 0.2583712935447693,
      "learning_rate": 0.0001,
      "loss": 1.5728,
      "step": 2166
    },
    {
      "epoch": 0.3479447655748234,
      "grad_norm": 0.26149067282676697,
      "learning_rate": 0.0001,
      "loss": 1.4892,
      "step": 2167
    },
    {
      "epoch": 0.3481053307642903,
      "grad_norm": 0.2641371786594391,
      "learning_rate": 0.0001,
      "loss": 1.5411,
      "step": 2168
    },
    {
      "epoch": 0.34826589595375723,
      "grad_norm": 0.25248417258262634,
      "learning_rate": 0.0001,
      "loss": 1.5704,
      "step": 2169
    },
    {
      "epoch": 0.3484264611432242,
      "grad_norm": 0.250577837228775,
      "learning_rate": 0.0001,
      "loss": 1.5639,
      "step": 2170
    },
    {
      "epoch": 0.34858702633269106,
      "grad_norm": 0.2572253346443176,
      "learning_rate": 0.0001,
      "loss": 1.5327,
      "step": 2171
    },
    {
      "epoch": 0.348747591522158,
      "grad_norm": 0.26197028160095215,
      "learning_rate": 0.0001,
      "loss": 1.594,
      "step": 2172
    },
    {
      "epoch": 0.3489081567116249,
      "grad_norm": 0.2460302859544754,
      "learning_rate": 0.0001,
      "loss": 1.5568,
      "step": 2173
    },
    {
      "epoch": 0.34906872190109184,
      "grad_norm": 0.24514052271842957,
      "learning_rate": 0.0001,
      "loss": 1.4846,
      "step": 2174
    },
    {
      "epoch": 0.3492292870905588,
      "grad_norm": 0.2499423325061798,
      "learning_rate": 0.0001,
      "loss": 1.4849,
      "step": 2175
    },
    {
      "epoch": 0.34938985228002567,
      "grad_norm": 0.2529575526714325,
      "learning_rate": 0.0001,
      "loss": 1.5793,
      "step": 2176
    },
    {
      "epoch": 0.3495504174694926,
      "grad_norm": 0.2539237439632416,
      "learning_rate": 0.0001,
      "loss": 1.4962,
      "step": 2177
    },
    {
      "epoch": 0.34971098265895956,
      "grad_norm": 0.2468840479850769,
      "learning_rate": 0.0001,
      "loss": 1.5231,
      "step": 2178
    },
    {
      "epoch": 0.34987154784842645,
      "grad_norm": 0.2504441738128662,
      "learning_rate": 0.0001,
      "loss": 1.589,
      "step": 2179
    },
    {
      "epoch": 0.3500321130378934,
      "grad_norm": 0.2503768503665924,
      "learning_rate": 0.0001,
      "loss": 1.5372,
      "step": 2180
    },
    {
      "epoch": 0.35019267822736033,
      "grad_norm": 0.2639516294002533,
      "learning_rate": 0.0001,
      "loss": 1.4467,
      "step": 2181
    },
    {
      "epoch": 0.3503532434168272,
      "grad_norm": 0.27165842056274414,
      "learning_rate": 0.0001,
      "loss": 1.5509,
      "step": 2182
    },
    {
      "epoch": 0.35051380860629416,
      "grad_norm": 0.25772613286972046,
      "learning_rate": 0.0001,
      "loss": 1.472,
      "step": 2183
    },
    {
      "epoch": 0.35067437379576105,
      "grad_norm": 0.25208979845046997,
      "learning_rate": 0.0001,
      "loss": 1.5356,
      "step": 2184
    },
    {
      "epoch": 0.350834938985228,
      "grad_norm": 0.2547012269496918,
      "learning_rate": 0.0001,
      "loss": 1.5456,
      "step": 2185
    },
    {
      "epoch": 0.35099550417469494,
      "grad_norm": 0.27544358372688293,
      "learning_rate": 0.0001,
      "loss": 1.6191,
      "step": 2186
    },
    {
      "epoch": 0.3511560693641618,
      "grad_norm": 0.2565270960330963,
      "learning_rate": 0.0001,
      "loss": 1.5449,
      "step": 2187
    },
    {
      "epoch": 0.35131663455362877,
      "grad_norm": 0.2597314119338989,
      "learning_rate": 0.0001,
      "loss": 1.6189,
      "step": 2188
    },
    {
      "epoch": 0.3514771997430957,
      "grad_norm": 0.2584989666938782,
      "learning_rate": 0.0001,
      "loss": 1.5808,
      "step": 2189
    },
    {
      "epoch": 0.3516377649325626,
      "grad_norm": 0.2660825252532959,
      "learning_rate": 0.0001,
      "loss": 1.659,
      "step": 2190
    },
    {
      "epoch": 0.35179833012202955,
      "grad_norm": 0.25640127062797546,
      "learning_rate": 0.0001,
      "loss": 1.5751,
      "step": 2191
    },
    {
      "epoch": 0.3519588953114965,
      "grad_norm": 1.5304124355316162,
      "learning_rate": 0.0001,
      "loss": 1.562,
      "step": 2192
    },
    {
      "epoch": 0.3521194605009634,
      "grad_norm": 0.26782211661338806,
      "learning_rate": 0.0001,
      "loss": 1.688,
      "step": 2193
    },
    {
      "epoch": 0.3522800256904303,
      "grad_norm": 0.238906130194664,
      "learning_rate": 0.0001,
      "loss": 1.4125,
      "step": 2194
    },
    {
      "epoch": 0.35244059087989726,
      "grad_norm": 0.2628589868545532,
      "learning_rate": 0.0001,
      "loss": 1.5951,
      "step": 2195
    },
    {
      "epoch": 0.35260115606936415,
      "grad_norm": 0.24343790113925934,
      "learning_rate": 0.0001,
      "loss": 1.4637,
      "step": 2196
    },
    {
      "epoch": 0.3527617212588311,
      "grad_norm": 0.2564511299133301,
      "learning_rate": 0.0001,
      "loss": 1.5038,
      "step": 2197
    },
    {
      "epoch": 0.352922286448298,
      "grad_norm": 0.24965737760066986,
      "learning_rate": 0.0001,
      "loss": 1.5078,
      "step": 2198
    },
    {
      "epoch": 0.3530828516377649,
      "grad_norm": 0.288682222366333,
      "learning_rate": 0.0001,
      "loss": 1.5988,
      "step": 2199
    },
    {
      "epoch": 0.35324341682723187,
      "grad_norm": 0.23998339474201202,
      "learning_rate": 0.0001,
      "loss": 1.5456,
      "step": 2200
    },
    {
      "epoch": 0.35340398201669876,
      "grad_norm": 0.25241750478744507,
      "learning_rate": 0.0001,
      "loss": 1.5322,
      "step": 2201
    },
    {
      "epoch": 0.3535645472061657,
      "grad_norm": 0.2497698813676834,
      "learning_rate": 0.0001,
      "loss": 1.5637,
      "step": 2202
    },
    {
      "epoch": 0.35372511239563265,
      "grad_norm": 0.25241801142692566,
      "learning_rate": 0.0001,
      "loss": 1.5509,
      "step": 2203
    },
    {
      "epoch": 0.35388567758509953,
      "grad_norm": 0.2644217908382416,
      "learning_rate": 0.0001,
      "loss": 1.6523,
      "step": 2204
    },
    {
      "epoch": 0.3540462427745665,
      "grad_norm": 0.25793546438217163,
      "learning_rate": 0.0001,
      "loss": 1.5697,
      "step": 2205
    },
    {
      "epoch": 0.3542068079640334,
      "grad_norm": 0.24560660123825073,
      "learning_rate": 0.0001,
      "loss": 1.4703,
      "step": 2206
    },
    {
      "epoch": 0.3543673731535003,
      "grad_norm": 0.26065948605537415,
      "learning_rate": 0.0001,
      "loss": 1.5204,
      "step": 2207
    },
    {
      "epoch": 0.35452793834296725,
      "grad_norm": 0.250878244638443,
      "learning_rate": 0.0001,
      "loss": 1.5404,
      "step": 2208
    },
    {
      "epoch": 0.35468850353243414,
      "grad_norm": 0.2677420675754547,
      "learning_rate": 0.0001,
      "loss": 1.6749,
      "step": 2209
    },
    {
      "epoch": 0.3548490687219011,
      "grad_norm": 0.2563774883747101,
      "learning_rate": 0.0001,
      "loss": 1.5647,
      "step": 2210
    },
    {
      "epoch": 0.355009633911368,
      "grad_norm": 0.2509433925151825,
      "learning_rate": 0.0001,
      "loss": 1.5338,
      "step": 2211
    },
    {
      "epoch": 0.3551701991008349,
      "grad_norm": 0.24012061953544617,
      "learning_rate": 0.0001,
      "loss": 1.5578,
      "step": 2212
    },
    {
      "epoch": 0.35533076429030186,
      "grad_norm": 0.23800039291381836,
      "learning_rate": 0.0001,
      "loss": 1.4734,
      "step": 2213
    },
    {
      "epoch": 0.3554913294797688,
      "grad_norm": 0.24657626450061798,
      "learning_rate": 0.0001,
      "loss": 1.5262,
      "step": 2214
    },
    {
      "epoch": 0.3556518946692357,
      "grad_norm": 0.24189968407154083,
      "learning_rate": 0.0001,
      "loss": 1.5417,
      "step": 2215
    },
    {
      "epoch": 0.35581245985870263,
      "grad_norm": 0.25544416904449463,
      "learning_rate": 0.0001,
      "loss": 1.6089,
      "step": 2216
    },
    {
      "epoch": 0.3559730250481696,
      "grad_norm": 0.266414999961853,
      "learning_rate": 0.0001,
      "loss": 1.5198,
      "step": 2217
    },
    {
      "epoch": 0.35613359023763647,
      "grad_norm": 0.2565564811229706,
      "learning_rate": 0.0001,
      "loss": 1.5381,
      "step": 2218
    },
    {
      "epoch": 0.3562941554271034,
      "grad_norm": 0.26174500584602356,
      "learning_rate": 0.0001,
      "loss": 1.596,
      "step": 2219
    },
    {
      "epoch": 0.35645472061657035,
      "grad_norm": 0.24598649144172668,
      "learning_rate": 0.0001,
      "loss": 1.502,
      "step": 2220
    },
    {
      "epoch": 0.35661528580603724,
      "grad_norm": 0.25279855728149414,
      "learning_rate": 0.0001,
      "loss": 1.5968,
      "step": 2221
    },
    {
      "epoch": 0.3567758509955042,
      "grad_norm": 0.25305691361427307,
      "learning_rate": 0.0001,
      "loss": 1.5453,
      "step": 2222
    },
    {
      "epoch": 0.3569364161849711,
      "grad_norm": 0.25044116377830505,
      "learning_rate": 0.0001,
      "loss": 1.5798,
      "step": 2223
    },
    {
      "epoch": 0.357096981374438,
      "grad_norm": 0.25510168075561523,
      "learning_rate": 0.0001,
      "loss": 1.5435,
      "step": 2224
    },
    {
      "epoch": 0.35725754656390496,
      "grad_norm": 0.25853925943374634,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 2225
    },
    {
      "epoch": 0.35741811175337185,
      "grad_norm": 0.278916597366333,
      "learning_rate": 0.0001,
      "loss": 1.5436,
      "step": 2226
    },
    {
      "epoch": 0.3575786769428388,
      "grad_norm": 0.26227280497550964,
      "learning_rate": 0.0001,
      "loss": 1.5987,
      "step": 2227
    },
    {
      "epoch": 0.35773924213230573,
      "grad_norm": 0.2559322416782379,
      "learning_rate": 0.0001,
      "loss": 1.5715,
      "step": 2228
    },
    {
      "epoch": 0.3578998073217726,
      "grad_norm": 0.26765576004981995,
      "learning_rate": 0.0001,
      "loss": 1.5298,
      "step": 2229
    },
    {
      "epoch": 0.35806037251123957,
      "grad_norm": 0.25036296248435974,
      "learning_rate": 0.0001,
      "loss": 1.5902,
      "step": 2230
    },
    {
      "epoch": 0.3582209377007065,
      "grad_norm": 0.2417023926973343,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 2231
    },
    {
      "epoch": 0.3583815028901734,
      "grad_norm": 0.2664749026298523,
      "learning_rate": 0.0001,
      "loss": 1.521,
      "step": 2232
    },
    {
      "epoch": 0.35854206807964034,
      "grad_norm": 0.2519276738166809,
      "learning_rate": 0.0001,
      "loss": 1.5354,
      "step": 2233
    },
    {
      "epoch": 0.3587026332691073,
      "grad_norm": 0.27917492389678955,
      "learning_rate": 0.0001,
      "loss": 1.6605,
      "step": 2234
    },
    {
      "epoch": 0.3588631984585742,
      "grad_norm": 0.25916269421577454,
      "learning_rate": 0.0001,
      "loss": 1.6011,
      "step": 2235
    },
    {
      "epoch": 0.3590237636480411,
      "grad_norm": 0.27903980016708374,
      "learning_rate": 0.0001,
      "loss": 1.5656,
      "step": 2236
    },
    {
      "epoch": 0.359184328837508,
      "grad_norm": 0.26287680864334106,
      "learning_rate": 0.0001,
      "loss": 1.5848,
      "step": 2237
    },
    {
      "epoch": 0.35934489402697495,
      "grad_norm": 0.255217969417572,
      "learning_rate": 0.0001,
      "loss": 1.6421,
      "step": 2238
    },
    {
      "epoch": 0.3595054592164419,
      "grad_norm": 0.2615929841995239,
      "learning_rate": 0.0001,
      "loss": 1.5013,
      "step": 2239
    },
    {
      "epoch": 0.3596660244059088,
      "grad_norm": 0.25425776839256287,
      "learning_rate": 0.0001,
      "loss": 1.4706,
      "step": 2240
    },
    {
      "epoch": 0.3598265895953757,
      "grad_norm": 0.26095038652420044,
      "learning_rate": 0.0001,
      "loss": 1.5401,
      "step": 2241
    },
    {
      "epoch": 0.35998715478484267,
      "grad_norm": 0.2528751790523529,
      "learning_rate": 0.0001,
      "loss": 1.5341,
      "step": 2242
    },
    {
      "epoch": 0.36014771997430955,
      "grad_norm": 0.2808654308319092,
      "learning_rate": 0.0001,
      "loss": 1.5915,
      "step": 2243
    },
    {
      "epoch": 0.3603082851637765,
      "grad_norm": 0.27228716015815735,
      "learning_rate": 0.0001,
      "loss": 1.5292,
      "step": 2244
    },
    {
      "epoch": 0.36046885035324344,
      "grad_norm": 0.24799329042434692,
      "learning_rate": 0.0001,
      "loss": 1.4773,
      "step": 2245
    },
    {
      "epoch": 0.36062941554271033,
      "grad_norm": 0.26087716221809387,
      "learning_rate": 0.0001,
      "loss": 1.4816,
      "step": 2246
    },
    {
      "epoch": 0.3607899807321773,
      "grad_norm": 0.29650628566741943,
      "learning_rate": 0.0001,
      "loss": 1.5467,
      "step": 2247
    },
    {
      "epoch": 0.36095054592164416,
      "grad_norm": 0.27107101678848267,
      "learning_rate": 0.0001,
      "loss": 1.6258,
      "step": 2248
    },
    {
      "epoch": 0.3611111111111111,
      "grad_norm": 0.2658062279224396,
      "learning_rate": 0.0001,
      "loss": 1.5173,
      "step": 2249
    },
    {
      "epoch": 0.36127167630057805,
      "grad_norm": 0.2656562030315399,
      "learning_rate": 0.0001,
      "loss": 1.5613,
      "step": 2250
    },
    {
      "epoch": 0.36143224149004494,
      "grad_norm": 0.254713773727417,
      "learning_rate": 0.0001,
      "loss": 1.59,
      "step": 2251
    },
    {
      "epoch": 0.3615928066795119,
      "grad_norm": 0.2695602476596832,
      "learning_rate": 0.0001,
      "loss": 1.5835,
      "step": 2252
    },
    {
      "epoch": 0.3617533718689788,
      "grad_norm": 0.2540896534919739,
      "learning_rate": 0.0001,
      "loss": 1.6294,
      "step": 2253
    },
    {
      "epoch": 0.3619139370584457,
      "grad_norm": 0.24553243815898895,
      "learning_rate": 0.0001,
      "loss": 1.4728,
      "step": 2254
    },
    {
      "epoch": 0.36207450224791266,
      "grad_norm": 0.25581425428390503,
      "learning_rate": 0.0001,
      "loss": 1.5225,
      "step": 2255
    },
    {
      "epoch": 0.3622350674373796,
      "grad_norm": 0.25077083706855774,
      "learning_rate": 0.0001,
      "loss": 1.4769,
      "step": 2256
    },
    {
      "epoch": 0.3623956326268465,
      "grad_norm": 0.2618795931339264,
      "learning_rate": 0.0001,
      "loss": 1.4968,
      "step": 2257
    },
    {
      "epoch": 0.36255619781631343,
      "grad_norm": 0.24708011746406555,
      "learning_rate": 0.0001,
      "loss": 1.4932,
      "step": 2258
    },
    {
      "epoch": 0.3627167630057804,
      "grad_norm": 0.26081109046936035,
      "learning_rate": 0.0001,
      "loss": 1.5161,
      "step": 2259
    },
    {
      "epoch": 0.36287732819524726,
      "grad_norm": 0.26388704776763916,
      "learning_rate": 0.0001,
      "loss": 1.5432,
      "step": 2260
    },
    {
      "epoch": 0.3630378933847142,
      "grad_norm": 0.2524031698703766,
      "learning_rate": 0.0001,
      "loss": 1.5497,
      "step": 2261
    },
    {
      "epoch": 0.3631984585741811,
      "grad_norm": 0.25277817249298096,
      "learning_rate": 0.0001,
      "loss": 1.4921,
      "step": 2262
    },
    {
      "epoch": 0.36335902376364804,
      "grad_norm": 0.24862059950828552,
      "learning_rate": 0.0001,
      "loss": 1.5437,
      "step": 2263
    },
    {
      "epoch": 0.363519588953115,
      "grad_norm": 0.26767322421073914,
      "learning_rate": 0.0001,
      "loss": 1.5406,
      "step": 2264
    },
    {
      "epoch": 0.36368015414258187,
      "grad_norm": 0.2560683488845825,
      "learning_rate": 0.0001,
      "loss": 1.5773,
      "step": 2265
    },
    {
      "epoch": 0.3638407193320488,
      "grad_norm": 0.27597951889038086,
      "learning_rate": 0.0001,
      "loss": 1.609,
      "step": 2266
    },
    {
      "epoch": 0.36400128452151576,
      "grad_norm": 0.2604365348815918,
      "learning_rate": 0.0001,
      "loss": 1.5812,
      "step": 2267
    },
    {
      "epoch": 0.36416184971098264,
      "grad_norm": 0.268728643655777,
      "learning_rate": 0.0001,
      "loss": 1.5985,
      "step": 2268
    },
    {
      "epoch": 0.3643224149004496,
      "grad_norm": 0.2544846832752228,
      "learning_rate": 0.0001,
      "loss": 1.5059,
      "step": 2269
    },
    {
      "epoch": 0.36448298008991653,
      "grad_norm": 0.2506101727485657,
      "learning_rate": 0.0001,
      "loss": 1.5233,
      "step": 2270
    },
    {
      "epoch": 0.3646435452793834,
      "grad_norm": 0.2539709508419037,
      "learning_rate": 0.0001,
      "loss": 1.5356,
      "step": 2271
    },
    {
      "epoch": 0.36480411046885036,
      "grad_norm": 0.2580517530441284,
      "learning_rate": 0.0001,
      "loss": 1.4624,
      "step": 2272
    },
    {
      "epoch": 0.36496467565831725,
      "grad_norm": 0.24341896176338196,
      "learning_rate": 0.0001,
      "loss": 1.4838,
      "step": 2273
    },
    {
      "epoch": 0.3651252408477842,
      "grad_norm": 0.3055890202522278,
      "learning_rate": 0.0001,
      "loss": 1.5315,
      "step": 2274
    },
    {
      "epoch": 0.36528580603725114,
      "grad_norm": 0.2585761547088623,
      "learning_rate": 0.0001,
      "loss": 1.5925,
      "step": 2275
    },
    {
      "epoch": 0.365446371226718,
      "grad_norm": 0.255572646856308,
      "learning_rate": 0.0001,
      "loss": 1.487,
      "step": 2276
    },
    {
      "epoch": 0.36560693641618497,
      "grad_norm": 0.2550049424171448,
      "learning_rate": 0.0001,
      "loss": 1.549,
      "step": 2277
    },
    {
      "epoch": 0.3657675016056519,
      "grad_norm": 0.2758522927761078,
      "learning_rate": 0.0001,
      "loss": 1.5406,
      "step": 2278
    },
    {
      "epoch": 0.3659280667951188,
      "grad_norm": 0.27849021553993225,
      "learning_rate": 0.0001,
      "loss": 1.659,
      "step": 2279
    },
    {
      "epoch": 0.36608863198458574,
      "grad_norm": 0.2738516330718994,
      "learning_rate": 0.0001,
      "loss": 1.5025,
      "step": 2280
    },
    {
      "epoch": 0.3662491971740527,
      "grad_norm": 0.27679434418678284,
      "learning_rate": 0.0001,
      "loss": 1.5515,
      "step": 2281
    },
    {
      "epoch": 0.3664097623635196,
      "grad_norm": 0.24797074496746063,
      "learning_rate": 0.0001,
      "loss": 1.5051,
      "step": 2282
    },
    {
      "epoch": 0.3665703275529865,
      "grad_norm": 0.2453446090221405,
      "learning_rate": 0.0001,
      "loss": 1.6077,
      "step": 2283
    },
    {
      "epoch": 0.36673089274245346,
      "grad_norm": 0.2575076222419739,
      "learning_rate": 0.0001,
      "loss": 1.569,
      "step": 2284
    },
    {
      "epoch": 0.36689145793192035,
      "grad_norm": 0.2688547372817993,
      "learning_rate": 0.0001,
      "loss": 1.5406,
      "step": 2285
    },
    {
      "epoch": 0.3670520231213873,
      "grad_norm": 0.2662636637687683,
      "learning_rate": 0.0001,
      "loss": 1.6246,
      "step": 2286
    },
    {
      "epoch": 0.3672125883108542,
      "grad_norm": 0.2721993923187256,
      "learning_rate": 0.0001,
      "loss": 1.5124,
      "step": 2287
    },
    {
      "epoch": 0.3673731535003211,
      "grad_norm": 0.26898816227912903,
      "learning_rate": 0.0001,
      "loss": 1.5695,
      "step": 2288
    },
    {
      "epoch": 0.36753371868978807,
      "grad_norm": 0.2678947150707245,
      "learning_rate": 0.0001,
      "loss": 1.5957,
      "step": 2289
    },
    {
      "epoch": 0.36769428387925496,
      "grad_norm": 0.2912648618221283,
      "learning_rate": 0.0001,
      "loss": 1.5751,
      "step": 2290
    },
    {
      "epoch": 0.3678548490687219,
      "grad_norm": 0.2662580907344818,
      "learning_rate": 0.0001,
      "loss": 1.6528,
      "step": 2291
    },
    {
      "epoch": 0.36801541425818884,
      "grad_norm": 0.24180811643600464,
      "learning_rate": 0.0001,
      "loss": 1.493,
      "step": 2292
    },
    {
      "epoch": 0.36817597944765573,
      "grad_norm": 0.26568982005119324,
      "learning_rate": 0.0001,
      "loss": 1.4992,
      "step": 2293
    },
    {
      "epoch": 0.3683365446371227,
      "grad_norm": 0.2533663511276245,
      "learning_rate": 0.0001,
      "loss": 1.5462,
      "step": 2294
    },
    {
      "epoch": 0.3684971098265896,
      "grad_norm": 0.2525925040245056,
      "learning_rate": 0.0001,
      "loss": 1.5829,
      "step": 2295
    },
    {
      "epoch": 0.3686576750160565,
      "grad_norm": 0.2568332552909851,
      "learning_rate": 0.0001,
      "loss": 1.5286,
      "step": 2296
    },
    {
      "epoch": 0.36881824020552345,
      "grad_norm": 0.2500928044319153,
      "learning_rate": 0.0001,
      "loss": 1.4974,
      "step": 2297
    },
    {
      "epoch": 0.36897880539499034,
      "grad_norm": 0.24780963361263275,
      "learning_rate": 0.0001,
      "loss": 1.5155,
      "step": 2298
    },
    {
      "epoch": 0.3691393705844573,
      "grad_norm": 0.2531459927558899,
      "learning_rate": 0.0001,
      "loss": 1.5436,
      "step": 2299
    },
    {
      "epoch": 0.3692999357739242,
      "grad_norm": 0.2483982890844345,
      "learning_rate": 0.0001,
      "loss": 1.4811,
      "step": 2300
    },
    {
      "epoch": 0.3694605009633911,
      "grad_norm": 0.24589164555072784,
      "learning_rate": 0.0001,
      "loss": 1.5705,
      "step": 2301
    },
    {
      "epoch": 0.36962106615285806,
      "grad_norm": 0.24893982708454132,
      "learning_rate": 0.0001,
      "loss": 1.5017,
      "step": 2302
    },
    {
      "epoch": 0.369781631342325,
      "grad_norm": 0.24600011110305786,
      "learning_rate": 0.0001,
      "loss": 1.4987,
      "step": 2303
    },
    {
      "epoch": 0.3699421965317919,
      "grad_norm": 0.24123330414295197,
      "learning_rate": 0.0001,
      "loss": 1.5392,
      "step": 2304
    },
    {
      "epoch": 0.37010276172125883,
      "grad_norm": 0.26374539732933044,
      "learning_rate": 0.0001,
      "loss": 1.5464,
      "step": 2305
    },
    {
      "epoch": 0.3702633269107258,
      "grad_norm": 0.238440603017807,
      "learning_rate": 0.0001,
      "loss": 1.4842,
      "step": 2306
    },
    {
      "epoch": 0.37042389210019266,
      "grad_norm": 0.25389161705970764,
      "learning_rate": 0.0001,
      "loss": 1.5679,
      "step": 2307
    },
    {
      "epoch": 0.3705844572896596,
      "grad_norm": 0.24614334106445312,
      "learning_rate": 0.0001,
      "loss": 1.4467,
      "step": 2308
    },
    {
      "epoch": 0.37074502247912655,
      "grad_norm": 0.2794525623321533,
      "learning_rate": 0.0001,
      "loss": 1.4848,
      "step": 2309
    },
    {
      "epoch": 0.37090558766859344,
      "grad_norm": 0.2474374771118164,
      "learning_rate": 0.0001,
      "loss": 1.5284,
      "step": 2310
    },
    {
      "epoch": 0.3710661528580604,
      "grad_norm": 0.25746452808380127,
      "learning_rate": 0.0001,
      "loss": 1.6253,
      "step": 2311
    },
    {
      "epoch": 0.37122671804752727,
      "grad_norm": 0.26093387603759766,
      "learning_rate": 0.0001,
      "loss": 1.5352,
      "step": 2312
    },
    {
      "epoch": 0.3713872832369942,
      "grad_norm": 1.198448657989502,
      "learning_rate": 0.0001,
      "loss": 1.4913,
      "step": 2313
    },
    {
      "epoch": 0.37154784842646116,
      "grad_norm": 0.24500393867492676,
      "learning_rate": 0.0001,
      "loss": 1.4841,
      "step": 2314
    },
    {
      "epoch": 0.37170841361592805,
      "grad_norm": 0.2583453953266144,
      "learning_rate": 0.0001,
      "loss": 1.5428,
      "step": 2315
    },
    {
      "epoch": 0.371868978805395,
      "grad_norm": 0.2561831474304199,
      "learning_rate": 0.0001,
      "loss": 1.5289,
      "step": 2316
    },
    {
      "epoch": 0.37202954399486193,
      "grad_norm": 0.29225724935531616,
      "learning_rate": 0.0001,
      "loss": 1.5268,
      "step": 2317
    },
    {
      "epoch": 0.3721901091843288,
      "grad_norm": 0.26211458444595337,
      "learning_rate": 0.0001,
      "loss": 1.5352,
      "step": 2318
    },
    {
      "epoch": 0.37235067437379576,
      "grad_norm": 0.2648836672306061,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 2319
    },
    {
      "epoch": 0.3725112395632627,
      "grad_norm": 0.2531552016735077,
      "learning_rate": 0.0001,
      "loss": 1.518,
      "step": 2320
    },
    {
      "epoch": 0.3726718047527296,
      "grad_norm": 0.2425978034734726,
      "learning_rate": 0.0001,
      "loss": 1.4766,
      "step": 2321
    },
    {
      "epoch": 0.37283236994219654,
      "grad_norm": 0.25794047117233276,
      "learning_rate": 0.0001,
      "loss": 1.5844,
      "step": 2322
    },
    {
      "epoch": 0.37299293513166343,
      "grad_norm": 0.2569916546344757,
      "learning_rate": 0.0001,
      "loss": 1.5871,
      "step": 2323
    },
    {
      "epoch": 0.37315350032113037,
      "grad_norm": 0.2629062235355377,
      "learning_rate": 0.0001,
      "loss": 1.4939,
      "step": 2324
    },
    {
      "epoch": 0.3733140655105973,
      "grad_norm": 0.24914056062698364,
      "learning_rate": 0.0001,
      "loss": 1.5539,
      "step": 2325
    },
    {
      "epoch": 0.3734746307000642,
      "grad_norm": 0.255267471075058,
      "learning_rate": 0.0001,
      "loss": 1.4701,
      "step": 2326
    },
    {
      "epoch": 0.37363519588953115,
      "grad_norm": 0.24833789467811584,
      "learning_rate": 0.0001,
      "loss": 1.5282,
      "step": 2327
    },
    {
      "epoch": 0.3737957610789981,
      "grad_norm": 0.24636401236057281,
      "learning_rate": 0.0001,
      "loss": 1.5551,
      "step": 2328
    },
    {
      "epoch": 0.373956326268465,
      "grad_norm": 0.24946527183055878,
      "learning_rate": 0.0001,
      "loss": 1.5193,
      "step": 2329
    },
    {
      "epoch": 0.3741168914579319,
      "grad_norm": 0.253368616104126,
      "learning_rate": 0.0001,
      "loss": 1.4774,
      "step": 2330
    },
    {
      "epoch": 0.37427745664739887,
      "grad_norm": 0.2617979943752289,
      "learning_rate": 0.0001,
      "loss": 1.5071,
      "step": 2331
    },
    {
      "epoch": 0.37443802183686575,
      "grad_norm": 0.264935165643692,
      "learning_rate": 0.0001,
      "loss": 1.5944,
      "step": 2332
    },
    {
      "epoch": 0.3745985870263327,
      "grad_norm": 0.2448675036430359,
      "learning_rate": 0.0001,
      "loss": 1.5315,
      "step": 2333
    },
    {
      "epoch": 0.37475915221579964,
      "grad_norm": 0.2725590169429779,
      "learning_rate": 0.0001,
      "loss": 1.5681,
      "step": 2334
    },
    {
      "epoch": 0.37491971740526653,
      "grad_norm": 0.25366848707199097,
      "learning_rate": 0.0001,
      "loss": 1.3628,
      "step": 2335
    },
    {
      "epoch": 0.37508028259473347,
      "grad_norm": 0.2598473131656647,
      "learning_rate": 0.0001,
      "loss": 1.482,
      "step": 2336
    },
    {
      "epoch": 0.37524084778420036,
      "grad_norm": 0.2570243775844574,
      "learning_rate": 0.0001,
      "loss": 1.5799,
      "step": 2337
    },
    {
      "epoch": 0.3754014129736673,
      "grad_norm": 0.2450980842113495,
      "learning_rate": 0.0001,
      "loss": 1.5059,
      "step": 2338
    },
    {
      "epoch": 0.37556197816313425,
      "grad_norm": 0.24217240512371063,
      "learning_rate": 0.0001,
      "loss": 1.5302,
      "step": 2339
    },
    {
      "epoch": 0.37572254335260113,
      "grad_norm": 0.2534436583518982,
      "learning_rate": 0.0001,
      "loss": 1.5605,
      "step": 2340
    },
    {
      "epoch": 0.3758831085420681,
      "grad_norm": 0.24209095537662506,
      "learning_rate": 0.0001,
      "loss": 1.4772,
      "step": 2341
    },
    {
      "epoch": 0.376043673731535,
      "grad_norm": 0.25760748982429504,
      "learning_rate": 0.0001,
      "loss": 1.521,
      "step": 2342
    },
    {
      "epoch": 0.3762042389210019,
      "grad_norm": 0.2506316900253296,
      "learning_rate": 0.0001,
      "loss": 1.5185,
      "step": 2343
    },
    {
      "epoch": 0.37636480411046885,
      "grad_norm": 0.2690396308898926,
      "learning_rate": 0.0001,
      "loss": 1.5056,
      "step": 2344
    },
    {
      "epoch": 0.3765253692999358,
      "grad_norm": 0.2650214731693268,
      "learning_rate": 0.0001,
      "loss": 1.5884,
      "step": 2345
    },
    {
      "epoch": 0.3766859344894027,
      "grad_norm": 0.2703990638256073,
      "learning_rate": 0.0001,
      "loss": 1.6113,
      "step": 2346
    },
    {
      "epoch": 0.37684649967886963,
      "grad_norm": 0.24998964369297028,
      "learning_rate": 0.0001,
      "loss": 1.5366,
      "step": 2347
    },
    {
      "epoch": 0.37700706486833657,
      "grad_norm": 0.2509383261203766,
      "learning_rate": 0.0001,
      "loss": 1.5678,
      "step": 2348
    },
    {
      "epoch": 0.37716763005780346,
      "grad_norm": 0.2529858648777008,
      "learning_rate": 0.0001,
      "loss": 1.4946,
      "step": 2349
    },
    {
      "epoch": 0.3773281952472704,
      "grad_norm": 0.26368749141693115,
      "learning_rate": 0.0001,
      "loss": 1.5901,
      "step": 2350
    },
    {
      "epoch": 0.3774887604367373,
      "grad_norm": 0.25796085596084595,
      "learning_rate": 0.0001,
      "loss": 1.5278,
      "step": 2351
    },
    {
      "epoch": 0.37764932562620424,
      "grad_norm": 0.24968422949314117,
      "learning_rate": 0.0001,
      "loss": 1.4562,
      "step": 2352
    },
    {
      "epoch": 0.3778098908156712,
      "grad_norm": 0.24780969321727753,
      "learning_rate": 0.0001,
      "loss": 1.5084,
      "step": 2353
    },
    {
      "epoch": 0.37797045600513807,
      "grad_norm": 0.2629697620868683,
      "learning_rate": 0.0001,
      "loss": 1.583,
      "step": 2354
    },
    {
      "epoch": 0.378131021194605,
      "grad_norm": 0.2600143551826477,
      "learning_rate": 0.0001,
      "loss": 1.5817,
      "step": 2355
    },
    {
      "epoch": 0.37829158638407195,
      "grad_norm": 0.3038232624530792,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 2356
    },
    {
      "epoch": 0.37845215157353884,
      "grad_norm": 0.25345587730407715,
      "learning_rate": 0.0001,
      "loss": 1.6067,
      "step": 2357
    },
    {
      "epoch": 0.3786127167630058,
      "grad_norm": 0.26497361063957214,
      "learning_rate": 0.0001,
      "loss": 1.5539,
      "step": 2358
    },
    {
      "epoch": 0.37877328195247273,
      "grad_norm": 0.26825860142707825,
      "learning_rate": 0.0001,
      "loss": 1.5312,
      "step": 2359
    },
    {
      "epoch": 0.3789338471419396,
      "grad_norm": 0.2717742323875427,
      "learning_rate": 0.0001,
      "loss": 1.509,
      "step": 2360
    },
    {
      "epoch": 0.37909441233140656,
      "grad_norm": 0.24727971851825714,
      "learning_rate": 0.0001,
      "loss": 1.5388,
      "step": 2361
    },
    {
      "epoch": 0.37925497752087345,
      "grad_norm": 0.29400545358657837,
      "learning_rate": 0.0001,
      "loss": 1.5218,
      "step": 2362
    },
    {
      "epoch": 0.3794155427103404,
      "grad_norm": 0.2577107548713684,
      "learning_rate": 0.0001,
      "loss": 1.6147,
      "step": 2363
    },
    {
      "epoch": 0.37957610789980734,
      "grad_norm": 0.27200621366500854,
      "learning_rate": 0.0001,
      "loss": 1.6134,
      "step": 2364
    },
    {
      "epoch": 0.3797366730892742,
      "grad_norm": 0.2652324438095093,
      "learning_rate": 0.0001,
      "loss": 1.5501,
      "step": 2365
    },
    {
      "epoch": 0.37989723827874117,
      "grad_norm": 0.2527463436126709,
      "learning_rate": 0.0001,
      "loss": 1.5035,
      "step": 2366
    },
    {
      "epoch": 0.3800578034682081,
      "grad_norm": 0.24923215806484222,
      "learning_rate": 0.0001,
      "loss": 1.4147,
      "step": 2367
    },
    {
      "epoch": 0.380218368657675,
      "grad_norm": 0.2674259543418884,
      "learning_rate": 0.0001,
      "loss": 1.5684,
      "step": 2368
    },
    {
      "epoch": 0.38037893384714194,
      "grad_norm": 0.25354260206222534,
      "learning_rate": 0.0001,
      "loss": 1.6004,
      "step": 2369
    },
    {
      "epoch": 0.3805394990366089,
      "grad_norm": 0.24319404363632202,
      "learning_rate": 0.0001,
      "loss": 1.4466,
      "step": 2370
    },
    {
      "epoch": 0.3807000642260758,
      "grad_norm": 0.24842718243598938,
      "learning_rate": 0.0001,
      "loss": 1.4414,
      "step": 2371
    },
    {
      "epoch": 0.3808606294155427,
      "grad_norm": 0.25003471970558167,
      "learning_rate": 0.0001,
      "loss": 1.5588,
      "step": 2372
    },
    {
      "epoch": 0.38102119460500966,
      "grad_norm": 0.26346343755722046,
      "learning_rate": 0.0001,
      "loss": 1.5043,
      "step": 2373
    },
    {
      "epoch": 0.38118175979447655,
      "grad_norm": 0.2605985105037689,
      "learning_rate": 0.0001,
      "loss": 1.4751,
      "step": 2374
    },
    {
      "epoch": 0.3813423249839435,
      "grad_norm": 0.2635630965232849,
      "learning_rate": 0.0001,
      "loss": 1.5325,
      "step": 2375
    },
    {
      "epoch": 0.3815028901734104,
      "grad_norm": 0.2483413815498352,
      "learning_rate": 0.0001,
      "loss": 1.5835,
      "step": 2376
    },
    {
      "epoch": 0.3816634553628773,
      "grad_norm": 0.2556459307670593,
      "learning_rate": 0.0001,
      "loss": 1.5943,
      "step": 2377
    },
    {
      "epoch": 0.38182402055234427,
      "grad_norm": 0.28714779019355774,
      "learning_rate": 0.0001,
      "loss": 1.5772,
      "step": 2378
    },
    {
      "epoch": 0.38198458574181116,
      "grad_norm": 0.24681034684181213,
      "learning_rate": 0.0001,
      "loss": 1.5301,
      "step": 2379
    },
    {
      "epoch": 0.3821451509312781,
      "grad_norm": 0.24625767767429352,
      "learning_rate": 0.0001,
      "loss": 1.5651,
      "step": 2380
    },
    {
      "epoch": 0.38230571612074504,
      "grad_norm": 0.2521355450153351,
      "learning_rate": 0.0001,
      "loss": 1.5356,
      "step": 2381
    },
    {
      "epoch": 0.38246628131021193,
      "grad_norm": 0.2519856095314026,
      "learning_rate": 0.0001,
      "loss": 1.5701,
      "step": 2382
    },
    {
      "epoch": 0.3826268464996789,
      "grad_norm": 0.2626766562461853,
      "learning_rate": 0.0001,
      "loss": 1.5292,
      "step": 2383
    },
    {
      "epoch": 0.3827874116891458,
      "grad_norm": 0.25843772292137146,
      "learning_rate": 0.0001,
      "loss": 1.5953,
      "step": 2384
    },
    {
      "epoch": 0.3829479768786127,
      "grad_norm": 0.25508567690849304,
      "learning_rate": 0.0001,
      "loss": 1.5657,
      "step": 2385
    },
    {
      "epoch": 0.38310854206807965,
      "grad_norm": 0.2541292607784271,
      "learning_rate": 0.0001,
      "loss": 1.5553,
      "step": 2386
    },
    {
      "epoch": 0.38326910725754654,
      "grad_norm": 0.26325348019599915,
      "learning_rate": 0.0001,
      "loss": 1.5455,
      "step": 2387
    },
    {
      "epoch": 0.3834296724470135,
      "grad_norm": 0.259388267993927,
      "learning_rate": 0.0001,
      "loss": 1.5504,
      "step": 2388
    },
    {
      "epoch": 0.3835902376364804,
      "grad_norm": 0.2743922472000122,
      "learning_rate": 0.0001,
      "loss": 1.5492,
      "step": 2389
    },
    {
      "epoch": 0.3837508028259473,
      "grad_norm": 0.261371374130249,
      "learning_rate": 0.0001,
      "loss": 1.5414,
      "step": 2390
    },
    {
      "epoch": 0.38391136801541426,
      "grad_norm": 0.24848809838294983,
      "learning_rate": 0.0001,
      "loss": 1.4902,
      "step": 2391
    },
    {
      "epoch": 0.3840719332048812,
      "grad_norm": 0.2744009494781494,
      "learning_rate": 0.0001,
      "loss": 1.4709,
      "step": 2392
    },
    {
      "epoch": 0.3842324983943481,
      "grad_norm": 0.2709813714027405,
      "learning_rate": 0.0001,
      "loss": 1.5535,
      "step": 2393
    },
    {
      "epoch": 0.38439306358381503,
      "grad_norm": 0.24891509115695953,
      "learning_rate": 0.0001,
      "loss": 1.5542,
      "step": 2394
    },
    {
      "epoch": 0.384553628773282,
      "grad_norm": 0.2544114589691162,
      "learning_rate": 0.0001,
      "loss": 1.5439,
      "step": 2395
    },
    {
      "epoch": 0.38471419396274886,
      "grad_norm": 0.2500731348991394,
      "learning_rate": 0.0001,
      "loss": 1.5878,
      "step": 2396
    },
    {
      "epoch": 0.3848747591522158,
      "grad_norm": 0.2746654748916626,
      "learning_rate": 0.0001,
      "loss": 1.5404,
      "step": 2397
    },
    {
      "epoch": 0.38503532434168275,
      "grad_norm": 0.25532686710357666,
      "learning_rate": 0.0001,
      "loss": 1.6097,
      "step": 2398
    },
    {
      "epoch": 0.38519588953114964,
      "grad_norm": 0.2480587363243103,
      "learning_rate": 0.0001,
      "loss": 1.599,
      "step": 2399
    },
    {
      "epoch": 0.3853564547206166,
      "grad_norm": 0.2571102976799011,
      "learning_rate": 0.0001,
      "loss": 1.5527,
      "step": 2400
    },
    {
      "epoch": 0.38551701991008347,
      "grad_norm": 0.24474580585956573,
      "learning_rate": 0.0001,
      "loss": 1.6144,
      "step": 2401
    },
    {
      "epoch": 0.3856775850995504,
      "grad_norm": 0.2899627089500427,
      "learning_rate": 0.0001,
      "loss": 1.5268,
      "step": 2402
    },
    {
      "epoch": 0.38583815028901736,
      "grad_norm": 0.2516287863254547,
      "learning_rate": 0.0001,
      "loss": 1.527,
      "step": 2403
    },
    {
      "epoch": 0.38599871547848424,
      "grad_norm": 0.2463321089744568,
      "learning_rate": 0.0001,
      "loss": 1.5011,
      "step": 2404
    },
    {
      "epoch": 0.3861592806679512,
      "grad_norm": 0.26257675886154175,
      "learning_rate": 0.0001,
      "loss": 1.5527,
      "step": 2405
    },
    {
      "epoch": 0.38631984585741813,
      "grad_norm": 0.2730129063129425,
      "learning_rate": 0.0001,
      "loss": 1.5943,
      "step": 2406
    },
    {
      "epoch": 0.386480411046885,
      "grad_norm": 0.2407064586877823,
      "learning_rate": 0.0001,
      "loss": 1.4716,
      "step": 2407
    },
    {
      "epoch": 0.38664097623635196,
      "grad_norm": 0.25162702798843384,
      "learning_rate": 0.0001,
      "loss": 1.5346,
      "step": 2408
    },
    {
      "epoch": 0.3868015414258189,
      "grad_norm": 0.26848462224006653,
      "learning_rate": 0.0001,
      "loss": 1.5723,
      "step": 2409
    },
    {
      "epoch": 0.3869621066152858,
      "grad_norm": 0.34027060866355896,
      "learning_rate": 0.0001,
      "loss": 1.4739,
      "step": 2410
    },
    {
      "epoch": 0.38712267180475274,
      "grad_norm": 0.26073455810546875,
      "learning_rate": 0.0001,
      "loss": 1.5585,
      "step": 2411
    },
    {
      "epoch": 0.3872832369942196,
      "grad_norm": 0.25451356172561646,
      "learning_rate": 0.0001,
      "loss": 1.5711,
      "step": 2412
    },
    {
      "epoch": 0.38744380218368657,
      "grad_norm": 0.2457560896873474,
      "learning_rate": 0.0001,
      "loss": 1.5013,
      "step": 2413
    },
    {
      "epoch": 0.3876043673731535,
      "grad_norm": 0.2544652819633484,
      "learning_rate": 0.0001,
      "loss": 1.6145,
      "step": 2414
    },
    {
      "epoch": 0.3877649325626204,
      "grad_norm": 0.2672406733036041,
      "learning_rate": 0.0001,
      "loss": 1.5614,
      "step": 2415
    },
    {
      "epoch": 0.38792549775208734,
      "grad_norm": 0.2651098072528839,
      "learning_rate": 0.0001,
      "loss": 1.5902,
      "step": 2416
    },
    {
      "epoch": 0.3880860629415543,
      "grad_norm": 0.3771659731864929,
      "learning_rate": 0.0001,
      "loss": 1.5871,
      "step": 2417
    },
    {
      "epoch": 0.3882466281310212,
      "grad_norm": 0.24673216044902802,
      "learning_rate": 0.0001,
      "loss": 1.4673,
      "step": 2418
    },
    {
      "epoch": 0.3884071933204881,
      "grad_norm": 0.25624924898147583,
      "learning_rate": 0.0001,
      "loss": 1.6001,
      "step": 2419
    },
    {
      "epoch": 0.38856775850995506,
      "grad_norm": 0.2547285556793213,
      "learning_rate": 0.0001,
      "loss": 1.4866,
      "step": 2420
    },
    {
      "epoch": 0.38872832369942195,
      "grad_norm": 0.5639746785163879,
      "learning_rate": 0.0001,
      "loss": 1.4773,
      "step": 2421
    },
    {
      "epoch": 0.3888888888888889,
      "grad_norm": 0.2534834146499634,
      "learning_rate": 0.0001,
      "loss": 1.5176,
      "step": 2422
    },
    {
      "epoch": 0.38904945407835584,
      "grad_norm": 0.26092657446861267,
      "learning_rate": 0.0001,
      "loss": 1.6083,
      "step": 2423
    },
    {
      "epoch": 0.3892100192678227,
      "grad_norm": 0.2513125240802765,
      "learning_rate": 0.0001,
      "loss": 1.5747,
      "step": 2424
    },
    {
      "epoch": 0.38937058445728967,
      "grad_norm": 0.2617529034614563,
      "learning_rate": 0.0001,
      "loss": 1.5298,
      "step": 2425
    },
    {
      "epoch": 0.38953114964675656,
      "grad_norm": 0.26412928104400635,
      "learning_rate": 0.0001,
      "loss": 1.5612,
      "step": 2426
    },
    {
      "epoch": 0.3896917148362235,
      "grad_norm": 0.2545403242111206,
      "learning_rate": 0.0001,
      "loss": 1.5585,
      "step": 2427
    },
    {
      "epoch": 0.38985228002569045,
      "grad_norm": 0.2545302212238312,
      "learning_rate": 0.0001,
      "loss": 1.5066,
      "step": 2428
    },
    {
      "epoch": 0.39001284521515733,
      "grad_norm": 0.25672441720962524,
      "learning_rate": 0.0001,
      "loss": 1.5193,
      "step": 2429
    },
    {
      "epoch": 0.3901734104046243,
      "grad_norm": 0.26069778203964233,
      "learning_rate": 0.0001,
      "loss": 1.5505,
      "step": 2430
    },
    {
      "epoch": 0.3903339755940912,
      "grad_norm": 0.2753305733203888,
      "learning_rate": 0.0001,
      "loss": 1.5965,
      "step": 2431
    },
    {
      "epoch": 0.3904945407835581,
      "grad_norm": 0.2471272051334381,
      "learning_rate": 0.0001,
      "loss": 1.449,
      "step": 2432
    },
    {
      "epoch": 0.39065510597302505,
      "grad_norm": 0.27465078234672546,
      "learning_rate": 0.0001,
      "loss": 1.5436,
      "step": 2433
    },
    {
      "epoch": 0.390815671162492,
      "grad_norm": 0.2718292772769928,
      "learning_rate": 0.0001,
      "loss": 1.528,
      "step": 2434
    },
    {
      "epoch": 0.3909762363519589,
      "grad_norm": 0.2727717161178589,
      "learning_rate": 0.0001,
      "loss": 1.5999,
      "step": 2435
    },
    {
      "epoch": 0.3911368015414258,
      "grad_norm": 0.2534559965133667,
      "learning_rate": 0.0001,
      "loss": 1.6012,
      "step": 2436
    },
    {
      "epoch": 0.3912973667308927,
      "grad_norm": 0.24747711420059204,
      "learning_rate": 0.0001,
      "loss": 1.5206,
      "step": 2437
    },
    {
      "epoch": 0.39145793192035966,
      "grad_norm": 0.26092955470085144,
      "learning_rate": 0.0001,
      "loss": 1.5382,
      "step": 2438
    },
    {
      "epoch": 0.3916184971098266,
      "grad_norm": 0.25769567489624023,
      "learning_rate": 0.0001,
      "loss": 1.5279,
      "step": 2439
    },
    {
      "epoch": 0.3917790622992935,
      "grad_norm": 0.25086402893066406,
      "learning_rate": 0.0001,
      "loss": 1.5302,
      "step": 2440
    },
    {
      "epoch": 0.39193962748876043,
      "grad_norm": 0.26384875178337097,
      "learning_rate": 0.0001,
      "loss": 1.458,
      "step": 2441
    },
    {
      "epoch": 0.3921001926782274,
      "grad_norm": 0.27071601152420044,
      "learning_rate": 0.0001,
      "loss": 1.5745,
      "step": 2442
    },
    {
      "epoch": 0.39226075786769427,
      "grad_norm": 0.24402453005313873,
      "learning_rate": 0.0001,
      "loss": 1.5489,
      "step": 2443
    },
    {
      "epoch": 0.3924213230571612,
      "grad_norm": 0.25266405940055847,
      "learning_rate": 0.0001,
      "loss": 1.5293,
      "step": 2444
    },
    {
      "epoch": 0.39258188824662815,
      "grad_norm": 0.2891020178794861,
      "learning_rate": 0.0001,
      "loss": 1.5921,
      "step": 2445
    },
    {
      "epoch": 0.39274245343609504,
      "grad_norm": 0.2697277069091797,
      "learning_rate": 0.0001,
      "loss": 1.5985,
      "step": 2446
    },
    {
      "epoch": 0.392903018625562,
      "grad_norm": 0.2564629316329956,
      "learning_rate": 0.0001,
      "loss": 1.5235,
      "step": 2447
    },
    {
      "epoch": 0.3930635838150289,
      "grad_norm": 0.27220186591148376,
      "learning_rate": 0.0001,
      "loss": 1.5511,
      "step": 2448
    },
    {
      "epoch": 0.3932241490044958,
      "grad_norm": 0.2559594213962555,
      "learning_rate": 0.0001,
      "loss": 1.5893,
      "step": 2449
    },
    {
      "epoch": 0.39338471419396276,
      "grad_norm": 0.2827349305152893,
      "learning_rate": 0.0001,
      "loss": 1.4945,
      "step": 2450
    },
    {
      "epoch": 0.39354527938342965,
      "grad_norm": 0.2591390311717987,
      "learning_rate": 0.0001,
      "loss": 1.5236,
      "step": 2451
    },
    {
      "epoch": 0.3937058445728966,
      "grad_norm": 0.2526845932006836,
      "learning_rate": 0.0001,
      "loss": 1.5154,
      "step": 2452
    },
    {
      "epoch": 0.39386640976236353,
      "grad_norm": 0.2741335928440094,
      "learning_rate": 0.0001,
      "loss": 1.5649,
      "step": 2453
    },
    {
      "epoch": 0.3940269749518304,
      "grad_norm": 0.26153987646102905,
      "learning_rate": 0.0001,
      "loss": 1.5225,
      "step": 2454
    },
    {
      "epoch": 0.39418754014129737,
      "grad_norm": 0.2582678496837616,
      "learning_rate": 0.0001,
      "loss": 1.5703,
      "step": 2455
    },
    {
      "epoch": 0.3943481053307643,
      "grad_norm": 0.25584661960601807,
      "learning_rate": 0.0001,
      "loss": 1.4649,
      "step": 2456
    },
    {
      "epoch": 0.3945086705202312,
      "grad_norm": 0.2764172852039337,
      "learning_rate": 0.0001,
      "loss": 1.58,
      "step": 2457
    },
    {
      "epoch": 0.39466923570969814,
      "grad_norm": 0.2539321184158325,
      "learning_rate": 0.0001,
      "loss": 1.5412,
      "step": 2458
    },
    {
      "epoch": 0.3948298008991651,
      "grad_norm": 0.2499200999736786,
      "learning_rate": 0.0001,
      "loss": 1.5115,
      "step": 2459
    },
    {
      "epoch": 0.394990366088632,
      "grad_norm": 0.27872729301452637,
      "learning_rate": 0.0001,
      "loss": 1.542,
      "step": 2460
    },
    {
      "epoch": 0.3951509312780989,
      "grad_norm": 0.2624555826187134,
      "learning_rate": 0.0001,
      "loss": 1.6081,
      "step": 2461
    },
    {
      "epoch": 0.39531149646756586,
      "grad_norm": 0.24843905866146088,
      "learning_rate": 0.0001,
      "loss": 1.5184,
      "step": 2462
    },
    {
      "epoch": 0.39547206165703275,
      "grad_norm": 2.4374990463256836,
      "learning_rate": 0.0001,
      "loss": 1.5083,
      "step": 2463
    },
    {
      "epoch": 0.3956326268464997,
      "grad_norm": 0.27892616391181946,
      "learning_rate": 0.0001,
      "loss": 1.4696,
      "step": 2464
    },
    {
      "epoch": 0.3957931920359666,
      "grad_norm": 0.2818267345428467,
      "learning_rate": 0.0001,
      "loss": 1.4486,
      "step": 2465
    },
    {
      "epoch": 0.3959537572254335,
      "grad_norm": 0.3038250207901001,
      "learning_rate": 0.0001,
      "loss": 1.6259,
      "step": 2466
    },
    {
      "epoch": 0.39611432241490047,
      "grad_norm": 0.31338241696357727,
      "learning_rate": 0.0001,
      "loss": 1.5498,
      "step": 2467
    },
    {
      "epoch": 0.39627488760436735,
      "grad_norm": 0.3117619752883911,
      "learning_rate": 0.0001,
      "loss": 1.6265,
      "step": 2468
    },
    {
      "epoch": 0.3964354527938343,
      "grad_norm": 0.3030302822589874,
      "learning_rate": 0.0001,
      "loss": 1.4984,
      "step": 2469
    },
    {
      "epoch": 0.39659601798330124,
      "grad_norm": 0.30077430605888367,
      "learning_rate": 0.0001,
      "loss": 1.5468,
      "step": 2470
    },
    {
      "epoch": 0.39675658317276813,
      "grad_norm": 0.2581309676170349,
      "learning_rate": 0.0001,
      "loss": 1.4842,
      "step": 2471
    },
    {
      "epoch": 0.3969171483622351,
      "grad_norm": 0.26717257499694824,
      "learning_rate": 0.0001,
      "loss": 1.4888,
      "step": 2472
    },
    {
      "epoch": 0.397077713551702,
      "grad_norm": 0.274960458278656,
      "learning_rate": 0.0001,
      "loss": 1.6078,
      "step": 2473
    },
    {
      "epoch": 0.3972382787411689,
      "grad_norm": 0.2550688683986664,
      "learning_rate": 0.0001,
      "loss": 1.5199,
      "step": 2474
    },
    {
      "epoch": 0.39739884393063585,
      "grad_norm": 0.26114949584007263,
      "learning_rate": 0.0001,
      "loss": 1.5089,
      "step": 2475
    },
    {
      "epoch": 0.39755940912010274,
      "grad_norm": 0.2530492842197418,
      "learning_rate": 0.0001,
      "loss": 1.4758,
      "step": 2476
    },
    {
      "epoch": 0.3977199743095697,
      "grad_norm": 0.2503330409526825,
      "learning_rate": 0.0001,
      "loss": 1.4575,
      "step": 2477
    },
    {
      "epoch": 0.3978805394990366,
      "grad_norm": 0.2625015079975128,
      "learning_rate": 0.0001,
      "loss": 1.5725,
      "step": 2478
    },
    {
      "epoch": 0.3980411046885035,
      "grad_norm": 0.25151559710502625,
      "learning_rate": 0.0001,
      "loss": 1.524,
      "step": 2479
    },
    {
      "epoch": 0.39820166987797045,
      "grad_norm": 0.2679506838321686,
      "learning_rate": 0.0001,
      "loss": 1.5126,
      "step": 2480
    },
    {
      "epoch": 0.3983622350674374,
      "grad_norm": 0.2590963840484619,
      "learning_rate": 0.0001,
      "loss": 1.5943,
      "step": 2481
    },
    {
      "epoch": 0.3985228002569043,
      "grad_norm": 0.24560026824474335,
      "learning_rate": 0.0001,
      "loss": 1.5301,
      "step": 2482
    },
    {
      "epoch": 0.39868336544637123,
      "grad_norm": 0.2637335956096649,
      "learning_rate": 0.0001,
      "loss": 1.5418,
      "step": 2483
    },
    {
      "epoch": 0.3988439306358382,
      "grad_norm": 0.2577708661556244,
      "learning_rate": 0.0001,
      "loss": 1.543,
      "step": 2484
    },
    {
      "epoch": 0.39900449582530506,
      "grad_norm": 0.2670186460018158,
      "learning_rate": 0.0001,
      "loss": 1.5635,
      "step": 2485
    },
    {
      "epoch": 0.399165061014772,
      "grad_norm": 0.2714400887489319,
      "learning_rate": 0.0001,
      "loss": 1.5687,
      "step": 2486
    },
    {
      "epoch": 0.39932562620423895,
      "grad_norm": 0.26190483570098877,
      "learning_rate": 0.0001,
      "loss": 1.5756,
      "step": 2487
    },
    {
      "epoch": 0.39948619139370584,
      "grad_norm": 0.24080729484558105,
      "learning_rate": 0.0001,
      "loss": 1.4703,
      "step": 2488
    },
    {
      "epoch": 0.3996467565831728,
      "grad_norm": 0.25955361127853394,
      "learning_rate": 0.0001,
      "loss": 1.6049,
      "step": 2489
    },
    {
      "epoch": 0.39980732177263967,
      "grad_norm": 0.2651372253894806,
      "learning_rate": 0.0001,
      "loss": 1.6211,
      "step": 2490
    },
    {
      "epoch": 0.3999678869621066,
      "grad_norm": 0.24466051161289215,
      "learning_rate": 0.0001,
      "loss": 1.5628,
      "step": 2491
    },
    {
      "epoch": 0.40012845215157355,
      "grad_norm": 0.25570055842399597,
      "learning_rate": 0.0001,
      "loss": 1.5573,
      "step": 2492
    },
    {
      "epoch": 0.40028901734104044,
      "grad_norm": 0.2497667521238327,
      "learning_rate": 0.0001,
      "loss": 1.5025,
      "step": 2493
    },
    {
      "epoch": 0.4004495825305074,
      "grad_norm": 0.26511573791503906,
      "learning_rate": 0.0001,
      "loss": 1.5379,
      "step": 2494
    },
    {
      "epoch": 0.40061014771997433,
      "grad_norm": 0.2591298222541809,
      "learning_rate": 0.0001,
      "loss": 1.4809,
      "step": 2495
    },
    {
      "epoch": 0.4007707129094412,
      "grad_norm": 0.2511709928512573,
      "learning_rate": 0.0001,
      "loss": 1.4905,
      "step": 2496
    },
    {
      "epoch": 0.40093127809890816,
      "grad_norm": 0.2514796555042267,
      "learning_rate": 0.0001,
      "loss": 1.5073,
      "step": 2497
    },
    {
      "epoch": 0.4010918432883751,
      "grad_norm": 0.2433425784111023,
      "learning_rate": 0.0001,
      "loss": 1.5256,
      "step": 2498
    },
    {
      "epoch": 0.401252408477842,
      "grad_norm": 0.25819021463394165,
      "learning_rate": 0.0001,
      "loss": 1.5117,
      "step": 2499
    },
    {
      "epoch": 0.40141297366730894,
      "grad_norm": 0.2464933842420578,
      "learning_rate": 0.0001,
      "loss": 1.5366,
      "step": 2500
    },
    {
      "epoch": 0.4015735388567758,
      "grad_norm": 0.30560988187789917,
      "learning_rate": 0.0001,
      "loss": 1.5413,
      "step": 2501
    },
    {
      "epoch": 0.40173410404624277,
      "grad_norm": 0.2743026316165924,
      "learning_rate": 0.0001,
      "loss": 1.5325,
      "step": 2502
    },
    {
      "epoch": 0.4018946692357097,
      "grad_norm": 0.24112781882286072,
      "learning_rate": 0.0001,
      "loss": 1.5483,
      "step": 2503
    },
    {
      "epoch": 0.4020552344251766,
      "grad_norm": 0.26679348945617676,
      "learning_rate": 0.0001,
      "loss": 1.5957,
      "step": 2504
    },
    {
      "epoch": 0.40221579961464354,
      "grad_norm": 0.2718173563480377,
      "learning_rate": 0.0001,
      "loss": 1.5665,
      "step": 2505
    },
    {
      "epoch": 0.4023763648041105,
      "grad_norm": 0.2669573426246643,
      "learning_rate": 0.0001,
      "loss": 1.5665,
      "step": 2506
    },
    {
      "epoch": 0.4025369299935774,
      "grad_norm": 0.24345548450946808,
      "learning_rate": 0.0001,
      "loss": 1.5037,
      "step": 2507
    },
    {
      "epoch": 0.4026974951830443,
      "grad_norm": 0.2644270360469818,
      "learning_rate": 0.0001,
      "loss": 1.5585,
      "step": 2508
    },
    {
      "epoch": 0.40285806037251126,
      "grad_norm": 0.2604409456253052,
      "learning_rate": 0.0001,
      "loss": 1.4758,
      "step": 2509
    },
    {
      "epoch": 0.40301862556197815,
      "grad_norm": 0.24744658172130585,
      "learning_rate": 0.0001,
      "loss": 1.5103,
      "step": 2510
    },
    {
      "epoch": 0.4031791907514451,
      "grad_norm": 0.25284886360168457,
      "learning_rate": 0.0001,
      "loss": 1.5241,
      "step": 2511
    },
    {
      "epoch": 0.40333975594091204,
      "grad_norm": 0.2449856996536255,
      "learning_rate": 0.0001,
      "loss": 1.5397,
      "step": 2512
    },
    {
      "epoch": 0.4035003211303789,
      "grad_norm": 0.24185772240161896,
      "learning_rate": 0.0001,
      "loss": 1.4647,
      "step": 2513
    },
    {
      "epoch": 0.40366088631984587,
      "grad_norm": 0.24948136508464813,
      "learning_rate": 0.0001,
      "loss": 1.4474,
      "step": 2514
    },
    {
      "epoch": 0.40382145150931276,
      "grad_norm": 0.23663552105426788,
      "learning_rate": 0.0001,
      "loss": 1.5678,
      "step": 2515
    },
    {
      "epoch": 0.4039820166987797,
      "grad_norm": 0.2513238489627838,
      "learning_rate": 0.0001,
      "loss": 1.5092,
      "step": 2516
    },
    {
      "epoch": 0.40414258188824664,
      "grad_norm": 0.2486760914325714,
      "learning_rate": 0.0001,
      "loss": 1.4202,
      "step": 2517
    },
    {
      "epoch": 0.40430314707771353,
      "grad_norm": 0.27109813690185547,
      "learning_rate": 0.0001,
      "loss": 1.5239,
      "step": 2518
    },
    {
      "epoch": 0.4044637122671805,
      "grad_norm": 0.23884856700897217,
      "learning_rate": 0.0001,
      "loss": 1.4775,
      "step": 2519
    },
    {
      "epoch": 0.4046242774566474,
      "grad_norm": 0.25711917877197266,
      "learning_rate": 0.0001,
      "loss": 1.5313,
      "step": 2520
    },
    {
      "epoch": 0.4047848426461143,
      "grad_norm": 0.2518276572227478,
      "learning_rate": 0.0001,
      "loss": 1.5513,
      "step": 2521
    },
    {
      "epoch": 0.40494540783558125,
      "grad_norm": 0.2606884241104126,
      "learning_rate": 0.0001,
      "loss": 1.5232,
      "step": 2522
    },
    {
      "epoch": 0.4051059730250482,
      "grad_norm": 0.25613394379615784,
      "learning_rate": 0.0001,
      "loss": 1.5433,
      "step": 2523
    },
    {
      "epoch": 0.4052665382145151,
      "grad_norm": 0.24913008511066437,
      "learning_rate": 0.0001,
      "loss": 1.462,
      "step": 2524
    },
    {
      "epoch": 0.405427103403982,
      "grad_norm": 0.25905099511146545,
      "learning_rate": 0.0001,
      "loss": 1.6206,
      "step": 2525
    },
    {
      "epoch": 0.4055876685934489,
      "grad_norm": 0.24587513506412506,
      "learning_rate": 0.0001,
      "loss": 1.4632,
      "step": 2526
    },
    {
      "epoch": 0.40574823378291586,
      "grad_norm": 0.23414070904254913,
      "learning_rate": 0.0001,
      "loss": 1.4336,
      "step": 2527
    },
    {
      "epoch": 0.4059087989723828,
      "grad_norm": 0.24958279728889465,
      "learning_rate": 0.0001,
      "loss": 1.5619,
      "step": 2528
    },
    {
      "epoch": 0.4060693641618497,
      "grad_norm": 0.26070019602775574,
      "learning_rate": 0.0001,
      "loss": 1.5057,
      "step": 2529
    },
    {
      "epoch": 0.40622992935131663,
      "grad_norm": 0.24881942570209503,
      "learning_rate": 0.0001,
      "loss": 1.591,
      "step": 2530
    },
    {
      "epoch": 0.4063904945407836,
      "grad_norm": 0.26656574010849,
      "learning_rate": 0.0001,
      "loss": 1.5512,
      "step": 2531
    },
    {
      "epoch": 0.40655105973025046,
      "grad_norm": 0.2782878875732422,
      "learning_rate": 0.0001,
      "loss": 1.5695,
      "step": 2532
    },
    {
      "epoch": 0.4067116249197174,
      "grad_norm": 0.25589877367019653,
      "learning_rate": 0.0001,
      "loss": 1.5471,
      "step": 2533
    },
    {
      "epoch": 0.40687219010918435,
      "grad_norm": 0.26112449169158936,
      "learning_rate": 0.0001,
      "loss": 1.5348,
      "step": 2534
    },
    {
      "epoch": 0.40703275529865124,
      "grad_norm": 0.2865256667137146,
      "learning_rate": 0.0001,
      "loss": 1.5444,
      "step": 2535
    },
    {
      "epoch": 0.4071933204881182,
      "grad_norm": 0.2590349316596985,
      "learning_rate": 0.0001,
      "loss": 1.4691,
      "step": 2536
    },
    {
      "epoch": 0.4073538856775851,
      "grad_norm": 0.25301647186279297,
      "learning_rate": 0.0001,
      "loss": 1.5163,
      "step": 2537
    },
    {
      "epoch": 0.407514450867052,
      "grad_norm": 0.24199478328227997,
      "learning_rate": 0.0001,
      "loss": 1.4866,
      "step": 2538
    },
    {
      "epoch": 0.40767501605651896,
      "grad_norm": 0.26721832156181335,
      "learning_rate": 0.0001,
      "loss": 1.5946,
      "step": 2539
    },
    {
      "epoch": 0.40783558124598585,
      "grad_norm": 0.2533133327960968,
      "learning_rate": 0.0001,
      "loss": 1.5306,
      "step": 2540
    },
    {
      "epoch": 0.4079961464354528,
      "grad_norm": 0.2587657570838928,
      "learning_rate": 0.0001,
      "loss": 1.5496,
      "step": 2541
    },
    {
      "epoch": 0.40815671162491973,
      "grad_norm": 0.2488706260919571,
      "learning_rate": 0.0001,
      "loss": 1.5212,
      "step": 2542
    },
    {
      "epoch": 0.4083172768143866,
      "grad_norm": 0.26686549186706543,
      "learning_rate": 0.0001,
      "loss": 1.5308,
      "step": 2543
    },
    {
      "epoch": 0.40847784200385356,
      "grad_norm": 0.27136510610580444,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 2544
    },
    {
      "epoch": 0.4086384071933205,
      "grad_norm": 0.2445228099822998,
      "learning_rate": 0.0001,
      "loss": 1.5685,
      "step": 2545
    },
    {
      "epoch": 0.4087989723827874,
      "grad_norm": 0.24528813362121582,
      "learning_rate": 0.0001,
      "loss": 1.5103,
      "step": 2546
    },
    {
      "epoch": 0.40895953757225434,
      "grad_norm": 0.2633833885192871,
      "learning_rate": 0.0001,
      "loss": 1.5417,
      "step": 2547
    },
    {
      "epoch": 0.4091201027617213,
      "grad_norm": 0.25173237919807434,
      "learning_rate": 0.0001,
      "loss": 1.5045,
      "step": 2548
    },
    {
      "epoch": 0.40928066795118817,
      "grad_norm": 0.2596408426761627,
      "learning_rate": 0.0001,
      "loss": 1.5999,
      "step": 2549
    },
    {
      "epoch": 0.4094412331406551,
      "grad_norm": 0.27274206280708313,
      "learning_rate": 0.0001,
      "loss": 1.5617,
      "step": 2550
    },
    {
      "epoch": 0.409601798330122,
      "grad_norm": 0.2476007491350174,
      "learning_rate": 0.0001,
      "loss": 1.5232,
      "step": 2551
    },
    {
      "epoch": 0.40976236351958895,
      "grad_norm": 0.25796622037887573,
      "learning_rate": 0.0001,
      "loss": 1.5114,
      "step": 2552
    },
    {
      "epoch": 0.4099229287090559,
      "grad_norm": 0.24175529181957245,
      "learning_rate": 0.0001,
      "loss": 1.4924,
      "step": 2553
    },
    {
      "epoch": 0.4100834938985228,
      "grad_norm": 0.2576192617416382,
      "learning_rate": 0.0001,
      "loss": 1.5433,
      "step": 2554
    },
    {
      "epoch": 0.4102440590879897,
      "grad_norm": 0.2633851170539856,
      "learning_rate": 0.0001,
      "loss": 1.6007,
      "step": 2555
    },
    {
      "epoch": 0.41040462427745666,
      "grad_norm": 0.2631693482398987,
      "learning_rate": 0.0001,
      "loss": 1.527,
      "step": 2556
    },
    {
      "epoch": 0.41056518946692355,
      "grad_norm": 0.26312610507011414,
      "learning_rate": 0.0001,
      "loss": 1.6563,
      "step": 2557
    },
    {
      "epoch": 0.4107257546563905,
      "grad_norm": 0.25023066997528076,
      "learning_rate": 0.0001,
      "loss": 1.5059,
      "step": 2558
    },
    {
      "epoch": 0.41088631984585744,
      "grad_norm": 0.2595348060131073,
      "learning_rate": 0.0001,
      "loss": 1.6379,
      "step": 2559
    },
    {
      "epoch": 0.4110468850353243,
      "grad_norm": 0.2501184642314911,
      "learning_rate": 0.0001,
      "loss": 1.5306,
      "step": 2560
    },
    {
      "epoch": 0.41120745022479127,
      "grad_norm": 0.3003421425819397,
      "learning_rate": 0.0001,
      "loss": 1.5155,
      "step": 2561
    },
    {
      "epoch": 0.4113680154142582,
      "grad_norm": 0.2603636384010315,
      "learning_rate": 0.0001,
      "loss": 1.5076,
      "step": 2562
    },
    {
      "epoch": 0.4115285806037251,
      "grad_norm": 0.2558700442314148,
      "learning_rate": 0.0001,
      "loss": 1.5355,
      "step": 2563
    },
    {
      "epoch": 0.41168914579319205,
      "grad_norm": 0.24567385017871857,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 2564
    },
    {
      "epoch": 0.41184971098265893,
      "grad_norm": 0.26885974407196045,
      "learning_rate": 0.0001,
      "loss": 1.5623,
      "step": 2565
    },
    {
      "epoch": 0.4120102761721259,
      "grad_norm": 0.2442621886730194,
      "learning_rate": 0.0001,
      "loss": 1.5341,
      "step": 2566
    },
    {
      "epoch": 0.4121708413615928,
      "grad_norm": 0.2417394071817398,
      "learning_rate": 0.0001,
      "loss": 1.4942,
      "step": 2567
    },
    {
      "epoch": 0.4123314065510597,
      "grad_norm": 0.2529677450656891,
      "learning_rate": 0.0001,
      "loss": 1.5544,
      "step": 2568
    },
    {
      "epoch": 0.41249197174052665,
      "grad_norm": 0.26576799154281616,
      "learning_rate": 0.0001,
      "loss": 1.5618,
      "step": 2569
    },
    {
      "epoch": 0.4126525369299936,
      "grad_norm": 0.25639644265174866,
      "learning_rate": 0.0001,
      "loss": 1.6553,
      "step": 2570
    },
    {
      "epoch": 0.4128131021194605,
      "grad_norm": 0.24996112287044525,
      "learning_rate": 0.0001,
      "loss": 1.5138,
      "step": 2571
    },
    {
      "epoch": 0.41297366730892743,
      "grad_norm": 0.25390946865081787,
      "learning_rate": 0.0001,
      "loss": 1.605,
      "step": 2572
    },
    {
      "epoch": 0.41313423249839437,
      "grad_norm": 0.23816630244255066,
      "learning_rate": 0.0001,
      "loss": 1.5376,
      "step": 2573
    },
    {
      "epoch": 0.41329479768786126,
      "grad_norm": 0.2712462842464447,
      "learning_rate": 0.0001,
      "loss": 1.5503,
      "step": 2574
    },
    {
      "epoch": 0.4134553628773282,
      "grad_norm": 0.24352729320526123,
      "learning_rate": 0.0001,
      "loss": 1.5102,
      "step": 2575
    },
    {
      "epoch": 0.41361592806679515,
      "grad_norm": 0.2582276463508606,
      "learning_rate": 0.0001,
      "loss": 1.5741,
      "step": 2576
    },
    {
      "epoch": 0.41377649325626203,
      "grad_norm": 0.2590008080005646,
      "learning_rate": 0.0001,
      "loss": 1.4903,
      "step": 2577
    },
    {
      "epoch": 0.413937058445729,
      "grad_norm": 0.2693990468978882,
      "learning_rate": 0.0001,
      "loss": 1.5606,
      "step": 2578
    },
    {
      "epoch": 0.41409762363519587,
      "grad_norm": 0.24303196370601654,
      "learning_rate": 0.0001,
      "loss": 1.5016,
      "step": 2579
    },
    {
      "epoch": 0.4142581888246628,
      "grad_norm": 0.25497758388519287,
      "learning_rate": 0.0001,
      "loss": 1.5817,
      "step": 2580
    },
    {
      "epoch": 0.41441875401412975,
      "grad_norm": 0.2549613118171692,
      "learning_rate": 0.0001,
      "loss": 1.6054,
      "step": 2581
    },
    {
      "epoch": 0.41457931920359664,
      "grad_norm": 0.24005141854286194,
      "learning_rate": 0.0001,
      "loss": 1.5124,
      "step": 2582
    },
    {
      "epoch": 0.4147398843930636,
      "grad_norm": 0.2643565833568573,
      "learning_rate": 0.0001,
      "loss": 1.6373,
      "step": 2583
    },
    {
      "epoch": 0.41490044958253053,
      "grad_norm": 0.25584590435028076,
      "learning_rate": 0.0001,
      "loss": 1.6176,
      "step": 2584
    },
    {
      "epoch": 0.4150610147719974,
      "grad_norm": 0.25806379318237305,
      "learning_rate": 0.0001,
      "loss": 1.5022,
      "step": 2585
    },
    {
      "epoch": 0.41522157996146436,
      "grad_norm": 0.2439253330230713,
      "learning_rate": 0.0001,
      "loss": 1.572,
      "step": 2586
    },
    {
      "epoch": 0.4153821451509313,
      "grad_norm": 0.24600747227668762,
      "learning_rate": 0.0001,
      "loss": 1.5545,
      "step": 2587
    },
    {
      "epoch": 0.4155427103403982,
      "grad_norm": 0.23955826461315155,
      "learning_rate": 0.0001,
      "loss": 1.5198,
      "step": 2588
    },
    {
      "epoch": 0.41570327552986513,
      "grad_norm": 0.25872573256492615,
      "learning_rate": 0.0001,
      "loss": 1.607,
      "step": 2589
    },
    {
      "epoch": 0.415863840719332,
      "grad_norm": 0.25654417276382446,
      "learning_rate": 0.0001,
      "loss": 1.5204,
      "step": 2590
    },
    {
      "epoch": 0.41602440590879897,
      "grad_norm": 0.25154048204421997,
      "learning_rate": 0.0001,
      "loss": 1.4452,
      "step": 2591
    },
    {
      "epoch": 0.4161849710982659,
      "grad_norm": 0.25746750831604004,
      "learning_rate": 0.0001,
      "loss": 1.5712,
      "step": 2592
    },
    {
      "epoch": 0.4163455362877328,
      "grad_norm": 0.26046082377433777,
      "learning_rate": 0.0001,
      "loss": 1.5369,
      "step": 2593
    },
    {
      "epoch": 0.41650610147719974,
      "grad_norm": 0.2493816763162613,
      "learning_rate": 0.0001,
      "loss": 1.4769,
      "step": 2594
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.28193199634552,
      "learning_rate": 0.0001,
      "loss": 1.5421,
      "step": 2595
    },
    {
      "epoch": 0.4168272318561336,
      "grad_norm": 0.24330681562423706,
      "learning_rate": 0.0001,
      "loss": 1.489,
      "step": 2596
    },
    {
      "epoch": 0.4169877970456005,
      "grad_norm": 0.2504135072231293,
      "learning_rate": 0.0001,
      "loss": 1.5631,
      "step": 2597
    },
    {
      "epoch": 0.41714836223506746,
      "grad_norm": 0.2549284100532532,
      "learning_rate": 0.0001,
      "loss": 1.5078,
      "step": 2598
    },
    {
      "epoch": 0.41730892742453435,
      "grad_norm": 0.2654156982898712,
      "learning_rate": 0.0001,
      "loss": 1.4762,
      "step": 2599
    },
    {
      "epoch": 0.4174694926140013,
      "grad_norm": 0.23926936089992523,
      "learning_rate": 0.0001,
      "loss": 1.5948,
      "step": 2600
    },
    {
      "epoch": 0.41763005780346824,
      "grad_norm": 0.25596749782562256,
      "learning_rate": 0.0001,
      "loss": 1.5494,
      "step": 2601
    },
    {
      "epoch": 0.4177906229929351,
      "grad_norm": 0.2619558870792389,
      "learning_rate": 0.0001,
      "loss": 1.4978,
      "step": 2602
    },
    {
      "epoch": 0.41795118818240207,
      "grad_norm": 0.26040348410606384,
      "learning_rate": 0.0001,
      "loss": 1.5119,
      "step": 2603
    },
    {
      "epoch": 0.41811175337186895,
      "grad_norm": 0.25314196944236755,
      "learning_rate": 0.0001,
      "loss": 1.4858,
      "step": 2604
    },
    {
      "epoch": 0.4182723185613359,
      "grad_norm": 0.24116294085979462,
      "learning_rate": 0.0001,
      "loss": 1.4512,
      "step": 2605
    },
    {
      "epoch": 0.41843288375080284,
      "grad_norm": 0.26536011695861816,
      "learning_rate": 0.0001,
      "loss": 1.4884,
      "step": 2606
    },
    {
      "epoch": 0.41859344894026973,
      "grad_norm": 0.26319968700408936,
      "learning_rate": 0.0001,
      "loss": 1.5866,
      "step": 2607
    },
    {
      "epoch": 0.4187540141297367,
      "grad_norm": 0.2550050616264343,
      "learning_rate": 0.0001,
      "loss": 1.5778,
      "step": 2608
    },
    {
      "epoch": 0.4189145793192036,
      "grad_norm": 0.2522720992565155,
      "learning_rate": 0.0001,
      "loss": 1.485,
      "step": 2609
    },
    {
      "epoch": 0.4190751445086705,
      "grad_norm": 0.2624843418598175,
      "learning_rate": 0.0001,
      "loss": 1.5883,
      "step": 2610
    },
    {
      "epoch": 0.41923570969813745,
      "grad_norm": 0.26206061244010925,
      "learning_rate": 0.0001,
      "loss": 1.5232,
      "step": 2611
    },
    {
      "epoch": 0.4193962748876044,
      "grad_norm": 0.2567644417285919,
      "learning_rate": 0.0001,
      "loss": 1.4272,
      "step": 2612
    },
    {
      "epoch": 0.4195568400770713,
      "grad_norm": 0.29025423526763916,
      "learning_rate": 0.0001,
      "loss": 1.5642,
      "step": 2613
    },
    {
      "epoch": 0.4197174052665382,
      "grad_norm": 0.2984831929206848,
      "learning_rate": 0.0001,
      "loss": 1.5887,
      "step": 2614
    },
    {
      "epoch": 0.4198779704560051,
      "grad_norm": 0.24703791737556458,
      "learning_rate": 0.0001,
      "loss": 1.5293,
      "step": 2615
    },
    {
      "epoch": 0.42003853564547206,
      "grad_norm": 0.26482588052749634,
      "learning_rate": 0.0001,
      "loss": 1.5698,
      "step": 2616
    },
    {
      "epoch": 0.420199100834939,
      "grad_norm": 0.26263001561164856,
      "learning_rate": 0.0001,
      "loss": 1.6209,
      "step": 2617
    },
    {
      "epoch": 0.4203596660244059,
      "grad_norm": 0.24634204804897308,
      "learning_rate": 0.0001,
      "loss": 1.5267,
      "step": 2618
    },
    {
      "epoch": 0.42052023121387283,
      "grad_norm": 0.2785063087940216,
      "learning_rate": 0.0001,
      "loss": 1.5986,
      "step": 2619
    },
    {
      "epoch": 0.4206807964033398,
      "grad_norm": 0.26866859197616577,
      "learning_rate": 0.0001,
      "loss": 1.4961,
      "step": 2620
    },
    {
      "epoch": 0.42084136159280666,
      "grad_norm": 0.272344708442688,
      "learning_rate": 0.0001,
      "loss": 1.5206,
      "step": 2621
    },
    {
      "epoch": 0.4210019267822736,
      "grad_norm": 0.27001920342445374,
      "learning_rate": 0.0001,
      "loss": 1.6008,
      "step": 2622
    },
    {
      "epoch": 0.42116249197174055,
      "grad_norm": 0.2562403380870819,
      "learning_rate": 0.0001,
      "loss": 1.5091,
      "step": 2623
    },
    {
      "epoch": 0.42132305716120744,
      "grad_norm": 0.2639353573322296,
      "learning_rate": 0.0001,
      "loss": 1.5608,
      "step": 2624
    },
    {
      "epoch": 0.4214836223506744,
      "grad_norm": 0.28386738896369934,
      "learning_rate": 0.0001,
      "loss": 1.6232,
      "step": 2625
    },
    {
      "epoch": 0.4216441875401413,
      "grad_norm": 0.2597806453704834,
      "learning_rate": 0.0001,
      "loss": 1.5924,
      "step": 2626
    },
    {
      "epoch": 0.4218047527296082,
      "grad_norm": 0.26284340023994446,
      "learning_rate": 0.0001,
      "loss": 1.5389,
      "step": 2627
    },
    {
      "epoch": 0.42196531791907516,
      "grad_norm": 0.24359674751758575,
      "learning_rate": 0.0001,
      "loss": 1.4012,
      "step": 2628
    },
    {
      "epoch": 0.42212588310854204,
      "grad_norm": 0.25141337513923645,
      "learning_rate": 0.0001,
      "loss": 1.5518,
      "step": 2629
    },
    {
      "epoch": 0.422286448298009,
      "grad_norm": 0.2479681670665741,
      "learning_rate": 0.0001,
      "loss": 1.5493,
      "step": 2630
    },
    {
      "epoch": 0.42244701348747593,
      "grad_norm": 0.24880971014499664,
      "learning_rate": 0.0001,
      "loss": 1.5023,
      "step": 2631
    },
    {
      "epoch": 0.4226075786769428,
      "grad_norm": 0.25287386775016785,
      "learning_rate": 0.0001,
      "loss": 1.5166,
      "step": 2632
    },
    {
      "epoch": 0.42276814386640976,
      "grad_norm": 0.2639290690422058,
      "learning_rate": 0.0001,
      "loss": 1.5861,
      "step": 2633
    },
    {
      "epoch": 0.4229287090558767,
      "grad_norm": 0.2502775490283966,
      "learning_rate": 0.0001,
      "loss": 1.5268,
      "step": 2634
    },
    {
      "epoch": 0.4230892742453436,
      "grad_norm": 0.24484850466251373,
      "learning_rate": 0.0001,
      "loss": 1.4464,
      "step": 2635
    },
    {
      "epoch": 0.42324983943481054,
      "grad_norm": 0.2664753198623657,
      "learning_rate": 0.0001,
      "loss": 1.594,
      "step": 2636
    },
    {
      "epoch": 0.4234104046242775,
      "grad_norm": 0.2515271306037903,
      "learning_rate": 0.0001,
      "loss": 1.5276,
      "step": 2637
    },
    {
      "epoch": 0.42357096981374437,
      "grad_norm": 0.25124526023864746,
      "learning_rate": 0.0001,
      "loss": 1.52,
      "step": 2638
    },
    {
      "epoch": 0.4237315350032113,
      "grad_norm": 0.2585409879684448,
      "learning_rate": 0.0001,
      "loss": 1.5634,
      "step": 2639
    },
    {
      "epoch": 0.4238921001926782,
      "grad_norm": 0.25445839762687683,
      "learning_rate": 0.0001,
      "loss": 1.5547,
      "step": 2640
    },
    {
      "epoch": 0.42405266538214514,
      "grad_norm": 0.26580333709716797,
      "learning_rate": 0.0001,
      "loss": 1.4956,
      "step": 2641
    },
    {
      "epoch": 0.4242132305716121,
      "grad_norm": 0.25759628415107727,
      "learning_rate": 0.0001,
      "loss": 1.5288,
      "step": 2642
    },
    {
      "epoch": 0.424373795761079,
      "grad_norm": 0.25335782766342163,
      "learning_rate": 0.0001,
      "loss": 1.4992,
      "step": 2643
    },
    {
      "epoch": 0.4245343609505459,
      "grad_norm": 0.26285186409950256,
      "learning_rate": 0.0001,
      "loss": 1.5565,
      "step": 2644
    },
    {
      "epoch": 0.42469492614001286,
      "grad_norm": 0.2520076632499695,
      "learning_rate": 0.0001,
      "loss": 1.5347,
      "step": 2645
    },
    {
      "epoch": 0.42485549132947975,
      "grad_norm": 0.2650809586048126,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 2646
    },
    {
      "epoch": 0.4250160565189467,
      "grad_norm": 0.2627808451652527,
      "learning_rate": 0.0001,
      "loss": 1.4876,
      "step": 2647
    },
    {
      "epoch": 0.42517662170841364,
      "grad_norm": 0.2433909773826599,
      "learning_rate": 0.0001,
      "loss": 1.5079,
      "step": 2648
    },
    {
      "epoch": 0.4253371868978805,
      "grad_norm": 0.25884318351745605,
      "learning_rate": 0.0001,
      "loss": 1.583,
      "step": 2649
    },
    {
      "epoch": 0.42549775208734747,
      "grad_norm": 0.2662484347820282,
      "learning_rate": 0.0001,
      "loss": 1.5267,
      "step": 2650
    },
    {
      "epoch": 0.4256583172768144,
      "grad_norm": 0.2550729215145111,
      "learning_rate": 0.0001,
      "loss": 1.6167,
      "step": 2651
    },
    {
      "epoch": 0.4258188824662813,
      "grad_norm": 0.25947099924087524,
      "learning_rate": 0.0001,
      "loss": 1.5475,
      "step": 2652
    },
    {
      "epoch": 0.42597944765574824,
      "grad_norm": 0.35818055272102356,
      "learning_rate": 0.0001,
      "loss": 1.577,
      "step": 2653
    },
    {
      "epoch": 0.42614001284521513,
      "grad_norm": 0.27459511160850525,
      "learning_rate": 0.0001,
      "loss": 1.5577,
      "step": 2654
    },
    {
      "epoch": 0.4263005780346821,
      "grad_norm": 0.25434303283691406,
      "learning_rate": 0.0001,
      "loss": 1.5585,
      "step": 2655
    },
    {
      "epoch": 0.426461143224149,
      "grad_norm": 0.2576950490474701,
      "learning_rate": 0.0001,
      "loss": 1.5593,
      "step": 2656
    },
    {
      "epoch": 0.4266217084136159,
      "grad_norm": 0.2581554055213928,
      "learning_rate": 0.0001,
      "loss": 1.5572,
      "step": 2657
    },
    {
      "epoch": 0.42678227360308285,
      "grad_norm": 0.25048211216926575,
      "learning_rate": 0.0001,
      "loss": 1.4973,
      "step": 2658
    },
    {
      "epoch": 0.4269428387925498,
      "grad_norm": 0.246945321559906,
      "learning_rate": 0.0001,
      "loss": 1.5544,
      "step": 2659
    },
    {
      "epoch": 0.4271034039820167,
      "grad_norm": 0.2460125982761383,
      "learning_rate": 0.0001,
      "loss": 1.5221,
      "step": 2660
    },
    {
      "epoch": 0.4272639691714836,
      "grad_norm": 0.25332316756248474,
      "learning_rate": 0.0001,
      "loss": 1.4938,
      "step": 2661
    },
    {
      "epoch": 0.42742453436095057,
      "grad_norm": 0.2497895061969757,
      "learning_rate": 0.0001,
      "loss": 1.494,
      "step": 2662
    },
    {
      "epoch": 0.42758509955041746,
      "grad_norm": 0.2499072551727295,
      "learning_rate": 0.0001,
      "loss": 1.5508,
      "step": 2663
    },
    {
      "epoch": 0.4277456647398844,
      "grad_norm": 0.2507265508174896,
      "learning_rate": 0.0001,
      "loss": 1.4885,
      "step": 2664
    },
    {
      "epoch": 0.4279062299293513,
      "grad_norm": 0.25569257140159607,
      "learning_rate": 0.0001,
      "loss": 1.5684,
      "step": 2665
    },
    {
      "epoch": 0.42806679511881823,
      "grad_norm": 2.385692596435547,
      "learning_rate": 0.0001,
      "loss": 1.4908,
      "step": 2666
    },
    {
      "epoch": 0.4282273603082852,
      "grad_norm": 0.2744636535644531,
      "learning_rate": 0.0001,
      "loss": 1.6281,
      "step": 2667
    },
    {
      "epoch": 0.42838792549775206,
      "grad_norm": 0.2834700644016266,
      "learning_rate": 0.0001,
      "loss": 1.5577,
      "step": 2668
    },
    {
      "epoch": 0.428548490687219,
      "grad_norm": 0.2911142110824585,
      "learning_rate": 0.0001,
      "loss": 1.5719,
      "step": 2669
    },
    {
      "epoch": 0.42870905587668595,
      "grad_norm": 0.25528252124786377,
      "learning_rate": 0.0001,
      "loss": 1.4831,
      "step": 2670
    },
    {
      "epoch": 0.42886962106615284,
      "grad_norm": 0.2466645985841751,
      "learning_rate": 0.0001,
      "loss": 1.5427,
      "step": 2671
    },
    {
      "epoch": 0.4290301862556198,
      "grad_norm": 0.27905088663101196,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 2672
    },
    {
      "epoch": 0.4291907514450867,
      "grad_norm": 0.25568363070487976,
      "learning_rate": 0.0001,
      "loss": 1.4708,
      "step": 2673
    },
    {
      "epoch": 0.4293513166345536,
      "grad_norm": 0.25385648012161255,
      "learning_rate": 0.0001,
      "loss": 1.5574,
      "step": 2674
    },
    {
      "epoch": 0.42951188182402056,
      "grad_norm": 0.2757319509983063,
      "learning_rate": 0.0001,
      "loss": 1.5304,
      "step": 2675
    },
    {
      "epoch": 0.4296724470134875,
      "grad_norm": 0.2604694366455078,
      "learning_rate": 0.0001,
      "loss": 1.4783,
      "step": 2676
    },
    {
      "epoch": 0.4298330122029544,
      "grad_norm": 0.27133870124816895,
      "learning_rate": 0.0001,
      "loss": 1.503,
      "step": 2677
    },
    {
      "epoch": 0.42999357739242133,
      "grad_norm": 0.25154414772987366,
      "learning_rate": 0.0001,
      "loss": 1.4647,
      "step": 2678
    },
    {
      "epoch": 0.4301541425818882,
      "grad_norm": 0.24950821697711945,
      "learning_rate": 0.0001,
      "loss": 1.5182,
      "step": 2679
    },
    {
      "epoch": 0.43031470777135516,
      "grad_norm": 0.2645113468170166,
      "learning_rate": 0.0001,
      "loss": 1.5368,
      "step": 2680
    },
    {
      "epoch": 0.4304752729608221,
      "grad_norm": 0.25628626346588135,
      "learning_rate": 0.0001,
      "loss": 1.5494,
      "step": 2681
    },
    {
      "epoch": 0.430635838150289,
      "grad_norm": 0.25732356309890747,
      "learning_rate": 0.0001,
      "loss": 1.5515,
      "step": 2682
    },
    {
      "epoch": 0.43079640333975594,
      "grad_norm": 0.2506778836250305,
      "learning_rate": 0.0001,
      "loss": 1.6141,
      "step": 2683
    },
    {
      "epoch": 0.4309569685292229,
      "grad_norm": 0.2596699893474579,
      "learning_rate": 0.0001,
      "loss": 1.5307,
      "step": 2684
    },
    {
      "epoch": 0.43111753371868977,
      "grad_norm": 0.24951109290122986,
      "learning_rate": 0.0001,
      "loss": 1.5074,
      "step": 2685
    },
    {
      "epoch": 0.4312780989081567,
      "grad_norm": 0.26219576597213745,
      "learning_rate": 0.0001,
      "loss": 1.4332,
      "step": 2686
    },
    {
      "epoch": 0.43143866409762366,
      "grad_norm": 0.254175066947937,
      "learning_rate": 0.0001,
      "loss": 1.6227,
      "step": 2687
    },
    {
      "epoch": 0.43159922928709055,
      "grad_norm": 0.25088366866111755,
      "learning_rate": 0.0001,
      "loss": 1.5545,
      "step": 2688
    },
    {
      "epoch": 0.4317597944765575,
      "grad_norm": 0.2951388657093048,
      "learning_rate": 0.0001,
      "loss": 1.6021,
      "step": 2689
    },
    {
      "epoch": 0.4319203596660244,
      "grad_norm": 0.28665709495544434,
      "learning_rate": 0.0001,
      "loss": 1.5506,
      "step": 2690
    },
    {
      "epoch": 0.4320809248554913,
      "grad_norm": 0.25719329714775085,
      "learning_rate": 0.0001,
      "loss": 1.5669,
      "step": 2691
    },
    {
      "epoch": 0.43224149004495827,
      "grad_norm": 0.27239304780960083,
      "learning_rate": 0.0001,
      "loss": 1.5778,
      "step": 2692
    },
    {
      "epoch": 0.43240205523442515,
      "grad_norm": 0.24096353352069855,
      "learning_rate": 0.0001,
      "loss": 1.4177,
      "step": 2693
    },
    {
      "epoch": 0.4325626204238921,
      "grad_norm": 0.25921630859375,
      "learning_rate": 0.0001,
      "loss": 1.5777,
      "step": 2694
    },
    {
      "epoch": 0.43272318561335904,
      "grad_norm": 0.2616073191165924,
      "learning_rate": 0.0001,
      "loss": 1.5244,
      "step": 2695
    },
    {
      "epoch": 0.43288375080282593,
      "grad_norm": 0.24035929143428802,
      "learning_rate": 0.0001,
      "loss": 1.4818,
      "step": 2696
    },
    {
      "epoch": 0.43304431599229287,
      "grad_norm": 0.2576081454753876,
      "learning_rate": 0.0001,
      "loss": 1.6338,
      "step": 2697
    },
    {
      "epoch": 0.4332048811817598,
      "grad_norm": 0.26582103967666626,
      "learning_rate": 0.0001,
      "loss": 1.5621,
      "step": 2698
    },
    {
      "epoch": 0.4333654463712267,
      "grad_norm": 0.2568143308162689,
      "learning_rate": 0.0001,
      "loss": 1.554,
      "step": 2699
    },
    {
      "epoch": 0.43352601156069365,
      "grad_norm": 0.2558077573776245,
      "learning_rate": 0.0001,
      "loss": 1.5883,
      "step": 2700
    },
    {
      "epoch": 0.4336865767501606,
      "grad_norm": 0.2493770569562912,
      "learning_rate": 0.0001,
      "loss": 1.5059,
      "step": 2701
    },
    {
      "epoch": 0.4338471419396275,
      "grad_norm": 0.24986302852630615,
      "learning_rate": 0.0001,
      "loss": 1.4514,
      "step": 2702
    },
    {
      "epoch": 0.4340077071290944,
      "grad_norm": 0.2509508728981018,
      "learning_rate": 0.0001,
      "loss": 1.6067,
      "step": 2703
    },
    {
      "epoch": 0.4341682723185613,
      "grad_norm": 0.2587379217147827,
      "learning_rate": 0.0001,
      "loss": 1.5094,
      "step": 2704
    },
    {
      "epoch": 0.43432883750802825,
      "grad_norm": 0.25891369581222534,
      "learning_rate": 0.0001,
      "loss": 1.5213,
      "step": 2705
    },
    {
      "epoch": 0.4344894026974952,
      "grad_norm": 0.2458716183900833,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 2706
    },
    {
      "epoch": 0.4346499678869621,
      "grad_norm": 0.24901600182056427,
      "learning_rate": 0.0001,
      "loss": 1.5352,
      "step": 2707
    },
    {
      "epoch": 0.43481053307642903,
      "grad_norm": 0.24522405862808228,
      "learning_rate": 0.0001,
      "loss": 1.5502,
      "step": 2708
    },
    {
      "epoch": 0.434971098265896,
      "grad_norm": 0.25697633624076843,
      "learning_rate": 0.0001,
      "loss": 1.5078,
      "step": 2709
    },
    {
      "epoch": 0.43513166345536286,
      "grad_norm": 0.25953882932662964,
      "learning_rate": 0.0001,
      "loss": 1.4886,
      "step": 2710
    },
    {
      "epoch": 0.4352922286448298,
      "grad_norm": 0.26224443316459656,
      "learning_rate": 0.0001,
      "loss": 1.5709,
      "step": 2711
    },
    {
      "epoch": 0.43545279383429675,
      "grad_norm": 0.2708311378955841,
      "learning_rate": 0.0001,
      "loss": 1.5595,
      "step": 2712
    },
    {
      "epoch": 0.43561335902376364,
      "grad_norm": 0.24821874499320984,
      "learning_rate": 0.0001,
      "loss": 1.499,
      "step": 2713
    },
    {
      "epoch": 0.4357739242132306,
      "grad_norm": 0.35547691583633423,
      "learning_rate": 0.0001,
      "loss": 1.5083,
      "step": 2714
    },
    {
      "epoch": 0.4359344894026975,
      "grad_norm": 0.24064689874649048,
      "learning_rate": 0.0001,
      "loss": 1.5219,
      "step": 2715
    },
    {
      "epoch": 0.4360950545921644,
      "grad_norm": 0.26905253529548645,
      "learning_rate": 0.0001,
      "loss": 1.6141,
      "step": 2716
    },
    {
      "epoch": 0.43625561978163135,
      "grad_norm": 0.2711796462535858,
      "learning_rate": 0.0001,
      "loss": 1.573,
      "step": 2717
    },
    {
      "epoch": 0.43641618497109824,
      "grad_norm": 0.2584587037563324,
      "learning_rate": 0.0001,
      "loss": 1.5267,
      "step": 2718
    },
    {
      "epoch": 0.4365767501605652,
      "grad_norm": 0.25472304224967957,
      "learning_rate": 0.0001,
      "loss": 1.5016,
      "step": 2719
    },
    {
      "epoch": 0.43673731535003213,
      "grad_norm": 0.257657527923584,
      "learning_rate": 0.0001,
      "loss": 1.5757,
      "step": 2720
    },
    {
      "epoch": 0.436897880539499,
      "grad_norm": 0.25497204065322876,
      "learning_rate": 0.0001,
      "loss": 1.4792,
      "step": 2721
    },
    {
      "epoch": 0.43705844572896596,
      "grad_norm": 0.2629372775554657,
      "learning_rate": 0.0001,
      "loss": 1.5797,
      "step": 2722
    },
    {
      "epoch": 0.4372190109184329,
      "grad_norm": 0.40733203291893005,
      "learning_rate": 0.0001,
      "loss": 1.5442,
      "step": 2723
    },
    {
      "epoch": 0.4373795761078998,
      "grad_norm": 0.2550099194049835,
      "learning_rate": 0.0001,
      "loss": 1.5965,
      "step": 2724
    },
    {
      "epoch": 0.43754014129736674,
      "grad_norm": 0.2658746540546417,
      "learning_rate": 0.0001,
      "loss": 1.4625,
      "step": 2725
    },
    {
      "epoch": 0.4377007064868337,
      "grad_norm": 0.25851795077323914,
      "learning_rate": 0.0001,
      "loss": 1.6022,
      "step": 2726
    },
    {
      "epoch": 0.43786127167630057,
      "grad_norm": 0.27039971947669983,
      "learning_rate": 0.0001,
      "loss": 1.5828,
      "step": 2727
    },
    {
      "epoch": 0.4380218368657675,
      "grad_norm": 0.25005993247032166,
      "learning_rate": 0.0001,
      "loss": 1.4932,
      "step": 2728
    },
    {
      "epoch": 0.4381824020552344,
      "grad_norm": 0.24069570004940033,
      "learning_rate": 0.0001,
      "loss": 1.5304,
      "step": 2729
    },
    {
      "epoch": 0.43834296724470134,
      "grad_norm": 0.23995202779769897,
      "learning_rate": 0.0001,
      "loss": 1.5175,
      "step": 2730
    },
    {
      "epoch": 0.4385035324341683,
      "grad_norm": 0.26718127727508545,
      "learning_rate": 0.0001,
      "loss": 1.534,
      "step": 2731
    },
    {
      "epoch": 0.4386640976236352,
      "grad_norm": 0.2617626190185547,
      "learning_rate": 0.0001,
      "loss": 1.6382,
      "step": 2732
    },
    {
      "epoch": 0.4388246628131021,
      "grad_norm": 0.2636631429195404,
      "learning_rate": 0.0001,
      "loss": 1.5521,
      "step": 2733
    },
    {
      "epoch": 0.43898522800256906,
      "grad_norm": 0.26319217681884766,
      "learning_rate": 0.0001,
      "loss": 1.5235,
      "step": 2734
    },
    {
      "epoch": 0.43914579319203595,
      "grad_norm": 0.2425624132156372,
      "learning_rate": 0.0001,
      "loss": 1.5166,
      "step": 2735
    },
    {
      "epoch": 0.4393063583815029,
      "grad_norm": 0.2700788080692291,
      "learning_rate": 0.0001,
      "loss": 1.6358,
      "step": 2736
    },
    {
      "epoch": 0.43946692357096984,
      "grad_norm": 0.25637128949165344,
      "learning_rate": 0.0001,
      "loss": 1.5037,
      "step": 2737
    },
    {
      "epoch": 0.4396274887604367,
      "grad_norm": 0.27559953927993774,
      "learning_rate": 0.0001,
      "loss": 1.5561,
      "step": 2738
    },
    {
      "epoch": 0.43978805394990367,
      "grad_norm": 0.25348690152168274,
      "learning_rate": 0.0001,
      "loss": 1.4787,
      "step": 2739
    },
    {
      "epoch": 0.4399486191393706,
      "grad_norm": 0.2682728171348572,
      "learning_rate": 0.0001,
      "loss": 1.476,
      "step": 2740
    },
    {
      "epoch": 0.4401091843288375,
      "grad_norm": 0.2554301619529724,
      "learning_rate": 0.0001,
      "loss": 1.514,
      "step": 2741
    },
    {
      "epoch": 0.44026974951830444,
      "grad_norm": 0.2548758387565613,
      "learning_rate": 0.0001,
      "loss": 1.5879,
      "step": 2742
    },
    {
      "epoch": 0.44043031470777133,
      "grad_norm": 0.2555524408817291,
      "learning_rate": 0.0001,
      "loss": 1.6307,
      "step": 2743
    },
    {
      "epoch": 0.4405908798972383,
      "grad_norm": 0.2641182243824005,
      "learning_rate": 0.0001,
      "loss": 1.5381,
      "step": 2744
    },
    {
      "epoch": 0.4407514450867052,
      "grad_norm": 0.25536489486694336,
      "learning_rate": 0.0001,
      "loss": 1.4782,
      "step": 2745
    },
    {
      "epoch": 0.4409120102761721,
      "grad_norm": 0.26611775159835815,
      "learning_rate": 0.0001,
      "loss": 1.5721,
      "step": 2746
    },
    {
      "epoch": 0.44107257546563905,
      "grad_norm": 0.25221577286720276,
      "learning_rate": 0.0001,
      "loss": 1.5091,
      "step": 2747
    },
    {
      "epoch": 0.441233140655106,
      "grad_norm": 0.25327831506729126,
      "learning_rate": 0.0001,
      "loss": 1.5007,
      "step": 2748
    },
    {
      "epoch": 0.4413937058445729,
      "grad_norm": 0.26875635981559753,
      "learning_rate": 0.0001,
      "loss": 1.558,
      "step": 2749
    },
    {
      "epoch": 0.4415542710340398,
      "grad_norm": 0.2573522925376892,
      "learning_rate": 0.0001,
      "loss": 1.4633,
      "step": 2750
    },
    {
      "epoch": 0.44171483622350677,
      "grad_norm": 0.2467297464609146,
      "learning_rate": 0.0001,
      "loss": 1.4585,
      "step": 2751
    },
    {
      "epoch": 0.44187540141297366,
      "grad_norm": 0.26516193151474,
      "learning_rate": 0.0001,
      "loss": 1.5107,
      "step": 2752
    },
    {
      "epoch": 0.4420359666024406,
      "grad_norm": 0.24327106773853302,
      "learning_rate": 0.0001,
      "loss": 1.4813,
      "step": 2753
    },
    {
      "epoch": 0.4421965317919075,
      "grad_norm": 0.24396276473999023,
      "learning_rate": 0.0001,
      "loss": 1.4599,
      "step": 2754
    },
    {
      "epoch": 0.44235709698137443,
      "grad_norm": 0.2524784505367279,
      "learning_rate": 0.0001,
      "loss": 1.4842,
      "step": 2755
    },
    {
      "epoch": 0.4425176621708414,
      "grad_norm": 0.2601543962955475,
      "learning_rate": 0.0001,
      "loss": 1.5943,
      "step": 2756
    },
    {
      "epoch": 0.44267822736030826,
      "grad_norm": 0.28155869245529175,
      "learning_rate": 0.0001,
      "loss": 1.4808,
      "step": 2757
    },
    {
      "epoch": 0.4428387925497752,
      "grad_norm": 0.2583642899990082,
      "learning_rate": 0.0001,
      "loss": 1.486,
      "step": 2758
    },
    {
      "epoch": 0.44299935773924215,
      "grad_norm": 0.2680513858795166,
      "learning_rate": 0.0001,
      "loss": 1.6168,
      "step": 2759
    },
    {
      "epoch": 0.44315992292870904,
      "grad_norm": 0.24765296280384064,
      "learning_rate": 0.0001,
      "loss": 1.5884,
      "step": 2760
    },
    {
      "epoch": 0.443320488118176,
      "grad_norm": 0.24888968467712402,
      "learning_rate": 0.0001,
      "loss": 1.4023,
      "step": 2761
    },
    {
      "epoch": 0.4434810533076429,
      "grad_norm": 0.2736065685749054,
      "learning_rate": 0.0001,
      "loss": 1.5299,
      "step": 2762
    },
    {
      "epoch": 0.4436416184971098,
      "grad_norm": 0.26058515906333923,
      "learning_rate": 0.0001,
      "loss": 1.5805,
      "step": 2763
    },
    {
      "epoch": 0.44380218368657676,
      "grad_norm": 0.2461216151714325,
      "learning_rate": 0.0001,
      "loss": 1.5148,
      "step": 2764
    },
    {
      "epoch": 0.4439627488760437,
      "grad_norm": 0.27771174907684326,
      "learning_rate": 0.0001,
      "loss": 1.5703,
      "step": 2765
    },
    {
      "epoch": 0.4441233140655106,
      "grad_norm": 0.24980218708515167,
      "learning_rate": 0.0001,
      "loss": 1.4626,
      "step": 2766
    },
    {
      "epoch": 0.44428387925497753,
      "grad_norm": 0.24767200648784637,
      "learning_rate": 0.0001,
      "loss": 1.4348,
      "step": 2767
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.26564276218414307,
      "learning_rate": 0.0001,
      "loss": 1.5237,
      "step": 2768
    },
    {
      "epoch": 0.44460500963391136,
      "grad_norm": 0.2671497166156769,
      "learning_rate": 0.0001,
      "loss": 1.5847,
      "step": 2769
    },
    {
      "epoch": 0.4447655748233783,
      "grad_norm": 0.2535061240196228,
      "learning_rate": 0.0001,
      "loss": 1.5423,
      "step": 2770
    },
    {
      "epoch": 0.4449261400128452,
      "grad_norm": 0.2804764211177826,
      "learning_rate": 0.0001,
      "loss": 1.5995,
      "step": 2771
    },
    {
      "epoch": 0.44508670520231214,
      "grad_norm": 0.2646520137786865,
      "learning_rate": 0.0001,
      "loss": 1.557,
      "step": 2772
    },
    {
      "epoch": 0.4452472703917791,
      "grad_norm": 0.2630246877670288,
      "learning_rate": 0.0001,
      "loss": 1.5388,
      "step": 2773
    },
    {
      "epoch": 0.44540783558124597,
      "grad_norm": 0.25226303935050964,
      "learning_rate": 0.0001,
      "loss": 1.5313,
      "step": 2774
    },
    {
      "epoch": 0.4455684007707129,
      "grad_norm": 0.2842601239681244,
      "learning_rate": 0.0001,
      "loss": 1.556,
      "step": 2775
    },
    {
      "epoch": 0.44572896596017986,
      "grad_norm": 0.2518823742866516,
      "learning_rate": 0.0001,
      "loss": 1.5384,
      "step": 2776
    },
    {
      "epoch": 0.44588953114964675,
      "grad_norm": 0.27817302942276,
      "learning_rate": 0.0001,
      "loss": 1.6704,
      "step": 2777
    },
    {
      "epoch": 0.4460500963391137,
      "grad_norm": 0.26197052001953125,
      "learning_rate": 0.0001,
      "loss": 1.6203,
      "step": 2778
    },
    {
      "epoch": 0.4462106615285806,
      "grad_norm": 0.24848100543022156,
      "learning_rate": 0.0001,
      "loss": 1.502,
      "step": 2779
    },
    {
      "epoch": 0.4463712267180475,
      "grad_norm": 0.2504656910896301,
      "learning_rate": 0.0001,
      "loss": 1.5106,
      "step": 2780
    },
    {
      "epoch": 0.44653179190751446,
      "grad_norm": 0.24837298691272736,
      "learning_rate": 0.0001,
      "loss": 1.5564,
      "step": 2781
    },
    {
      "epoch": 0.44669235709698135,
      "grad_norm": 0.2552327811717987,
      "learning_rate": 0.0001,
      "loss": 1.5696,
      "step": 2782
    },
    {
      "epoch": 0.4468529222864483,
      "grad_norm": 0.26277005672454834,
      "learning_rate": 0.0001,
      "loss": 1.5916,
      "step": 2783
    },
    {
      "epoch": 0.44701348747591524,
      "grad_norm": 0.24569441378116608,
      "learning_rate": 0.0001,
      "loss": 1.4613,
      "step": 2784
    },
    {
      "epoch": 0.4471740526653821,
      "grad_norm": 0.24393264949321747,
      "learning_rate": 0.0001,
      "loss": 1.5542,
      "step": 2785
    },
    {
      "epoch": 0.44733461785484907,
      "grad_norm": 0.2537849247455597,
      "learning_rate": 0.0001,
      "loss": 1.5466,
      "step": 2786
    },
    {
      "epoch": 0.447495183044316,
      "grad_norm": 0.245890811085701,
      "learning_rate": 0.0001,
      "loss": 1.5347,
      "step": 2787
    },
    {
      "epoch": 0.4476557482337829,
      "grad_norm": 0.2544203996658325,
      "learning_rate": 0.0001,
      "loss": 1.4941,
      "step": 2788
    },
    {
      "epoch": 0.44781631342324985,
      "grad_norm": 0.2503323256969452,
      "learning_rate": 0.0001,
      "loss": 1.5464,
      "step": 2789
    },
    {
      "epoch": 0.4479768786127168,
      "grad_norm": 0.5950185656547546,
      "learning_rate": 0.0001,
      "loss": 1.5325,
      "step": 2790
    },
    {
      "epoch": 0.4481374438021837,
      "grad_norm": 0.25244295597076416,
      "learning_rate": 0.0001,
      "loss": 1.5653,
      "step": 2791
    },
    {
      "epoch": 0.4482980089916506,
      "grad_norm": 0.2557685673236847,
      "learning_rate": 0.0001,
      "loss": 1.4407,
      "step": 2792
    },
    {
      "epoch": 0.4484585741811175,
      "grad_norm": 0.2686178684234619,
      "learning_rate": 0.0001,
      "loss": 1.6164,
      "step": 2793
    },
    {
      "epoch": 0.44861913937058445,
      "grad_norm": 0.2816196382045746,
      "learning_rate": 0.0001,
      "loss": 1.5574,
      "step": 2794
    },
    {
      "epoch": 0.4487797045600514,
      "grad_norm": 0.27569353580474854,
      "learning_rate": 0.0001,
      "loss": 1.5461,
      "step": 2795
    },
    {
      "epoch": 0.4489402697495183,
      "grad_norm": 0.27293121814727783,
      "learning_rate": 0.0001,
      "loss": 1.5787,
      "step": 2796
    },
    {
      "epoch": 0.4491008349389852,
      "grad_norm": 0.24752502143383026,
      "learning_rate": 0.0001,
      "loss": 1.5157,
      "step": 2797
    },
    {
      "epoch": 0.44926140012845217,
      "grad_norm": 0.2575712203979492,
      "learning_rate": 0.0001,
      "loss": 1.5448,
      "step": 2798
    },
    {
      "epoch": 0.44942196531791906,
      "grad_norm": 0.2574111223220825,
      "learning_rate": 0.0001,
      "loss": 1.4557,
      "step": 2799
    },
    {
      "epoch": 0.449582530507386,
      "grad_norm": 0.6211435198783875,
      "learning_rate": 0.0001,
      "loss": 1.5214,
      "step": 2800
    },
    {
      "epoch": 0.44974309569685295,
      "grad_norm": 0.2662184536457062,
      "learning_rate": 0.0001,
      "loss": 1.6333,
      "step": 2801
    },
    {
      "epoch": 0.44990366088631983,
      "grad_norm": 0.2594231069087982,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 2802
    },
    {
      "epoch": 0.4500642260757868,
      "grad_norm": 0.33451321721076965,
      "learning_rate": 0.0001,
      "loss": 1.5315,
      "step": 2803
    },
    {
      "epoch": 0.45022479126525367,
      "grad_norm": 0.26403456926345825,
      "learning_rate": 0.0001,
      "loss": 1.5518,
      "step": 2804
    },
    {
      "epoch": 0.4503853564547206,
      "grad_norm": 0.24283012747764587,
      "learning_rate": 0.0001,
      "loss": 1.5375,
      "step": 2805
    },
    {
      "epoch": 0.45054592164418755,
      "grad_norm": 0.28998836874961853,
      "learning_rate": 0.0001,
      "loss": 1.5779,
      "step": 2806
    },
    {
      "epoch": 0.45070648683365444,
      "grad_norm": 0.2587253451347351,
      "learning_rate": 0.0001,
      "loss": 1.4139,
      "step": 2807
    },
    {
      "epoch": 0.4508670520231214,
      "grad_norm": 0.24802348017692566,
      "learning_rate": 0.0001,
      "loss": 1.5288,
      "step": 2808
    },
    {
      "epoch": 0.4510276172125883,
      "grad_norm": 0.2477966547012329,
      "learning_rate": 0.0001,
      "loss": 1.5519,
      "step": 2809
    },
    {
      "epoch": 0.4511881824020552,
      "grad_norm": 0.33115845918655396,
      "learning_rate": 0.0001,
      "loss": 1.4968,
      "step": 2810
    },
    {
      "epoch": 0.45134874759152216,
      "grad_norm": 0.2585131824016571,
      "learning_rate": 0.0001,
      "loss": 1.4975,
      "step": 2811
    },
    {
      "epoch": 0.4515093127809891,
      "grad_norm": 0.2601938843727112,
      "learning_rate": 0.0001,
      "loss": 1.4967,
      "step": 2812
    },
    {
      "epoch": 0.451669877970456,
      "grad_norm": 0.2451581507921219,
      "learning_rate": 0.0001,
      "loss": 1.4733,
      "step": 2813
    },
    {
      "epoch": 0.45183044315992293,
      "grad_norm": 0.2767625153064728,
      "learning_rate": 0.0001,
      "loss": 1.5122,
      "step": 2814
    },
    {
      "epoch": 0.4519910083493899,
      "grad_norm": 0.2543317973613739,
      "learning_rate": 0.0001,
      "loss": 1.5517,
      "step": 2815
    },
    {
      "epoch": 0.45215157353885677,
      "grad_norm": 0.26612985134124756,
      "learning_rate": 0.0001,
      "loss": 1.5807,
      "step": 2816
    },
    {
      "epoch": 0.4523121387283237,
      "grad_norm": 0.25472018122673035,
      "learning_rate": 0.0001,
      "loss": 1.5078,
      "step": 2817
    },
    {
      "epoch": 0.4524727039177906,
      "grad_norm": 0.24495795369148254,
      "learning_rate": 0.0001,
      "loss": 1.5455,
      "step": 2818
    },
    {
      "epoch": 0.45263326910725754,
      "grad_norm": 0.24734511971473694,
      "learning_rate": 0.0001,
      "loss": 1.464,
      "step": 2819
    },
    {
      "epoch": 0.4527938342967245,
      "grad_norm": 0.25817036628723145,
      "learning_rate": 0.0001,
      "loss": 1.5298,
      "step": 2820
    },
    {
      "epoch": 0.4529543994861914,
      "grad_norm": 0.5361108779907227,
      "learning_rate": 0.0001,
      "loss": 1.4777,
      "step": 2821
    },
    {
      "epoch": 0.4531149646756583,
      "grad_norm": 0.24240370094776154,
      "learning_rate": 0.0001,
      "loss": 1.5158,
      "step": 2822
    },
    {
      "epoch": 0.45327552986512526,
      "grad_norm": 0.24924057722091675,
      "learning_rate": 0.0001,
      "loss": 1.5477,
      "step": 2823
    },
    {
      "epoch": 0.45343609505459215,
      "grad_norm": 0.2728411853313446,
      "learning_rate": 0.0001,
      "loss": 1.6117,
      "step": 2824
    },
    {
      "epoch": 0.4535966602440591,
      "grad_norm": 0.2432524412870407,
      "learning_rate": 0.0001,
      "loss": 1.5073,
      "step": 2825
    },
    {
      "epoch": 0.45375722543352603,
      "grad_norm": 0.2530824840068817,
      "learning_rate": 0.0001,
      "loss": 1.5505,
      "step": 2826
    },
    {
      "epoch": 0.4539177906229929,
      "grad_norm": 0.24826349318027496,
      "learning_rate": 0.0001,
      "loss": 1.5805,
      "step": 2827
    },
    {
      "epoch": 0.45407835581245987,
      "grad_norm": 0.26144877076148987,
      "learning_rate": 0.0001,
      "loss": 1.5111,
      "step": 2828
    },
    {
      "epoch": 0.4542389210019268,
      "grad_norm": 0.2444932460784912,
      "learning_rate": 0.0001,
      "loss": 1.5571,
      "step": 2829
    },
    {
      "epoch": 0.4543994861913937,
      "grad_norm": 0.25554490089416504,
      "learning_rate": 0.0001,
      "loss": 1.5486,
      "step": 2830
    },
    {
      "epoch": 0.45456005138086064,
      "grad_norm": 0.25208333134651184,
      "learning_rate": 0.0001,
      "loss": 1.4875,
      "step": 2831
    },
    {
      "epoch": 0.45472061657032753,
      "grad_norm": 0.25124403834342957,
      "learning_rate": 0.0001,
      "loss": 1.4813,
      "step": 2832
    },
    {
      "epoch": 0.4548811817597945,
      "grad_norm": 0.26822763681411743,
      "learning_rate": 0.0001,
      "loss": 1.5502,
      "step": 2833
    },
    {
      "epoch": 0.4550417469492614,
      "grad_norm": 0.24606886506080627,
      "learning_rate": 0.0001,
      "loss": 1.5009,
      "step": 2834
    },
    {
      "epoch": 0.4552023121387283,
      "grad_norm": 1.08027982711792,
      "learning_rate": 0.0001,
      "loss": 1.4697,
      "step": 2835
    },
    {
      "epoch": 0.45536287732819525,
      "grad_norm": 0.2689116597175598,
      "learning_rate": 0.0001,
      "loss": 1.5069,
      "step": 2836
    },
    {
      "epoch": 0.4555234425176622,
      "grad_norm": 0.24950948357582092,
      "learning_rate": 0.0001,
      "loss": 1.4455,
      "step": 2837
    },
    {
      "epoch": 0.4556840077071291,
      "grad_norm": 0.2667357921600342,
      "learning_rate": 0.0001,
      "loss": 1.4858,
      "step": 2838
    },
    {
      "epoch": 0.455844572896596,
      "grad_norm": 0.2685289978981018,
      "learning_rate": 0.0001,
      "loss": 1.4995,
      "step": 2839
    },
    {
      "epoch": 0.45600513808606297,
      "grad_norm": 0.25204527378082275,
      "learning_rate": 0.0001,
      "loss": 1.5202,
      "step": 2840
    },
    {
      "epoch": 0.45616570327552985,
      "grad_norm": 0.24358847737312317,
      "learning_rate": 0.0001,
      "loss": 1.4787,
      "step": 2841
    },
    {
      "epoch": 0.4563262684649968,
      "grad_norm": 0.2502761781215668,
      "learning_rate": 0.0001,
      "loss": 1.5346,
      "step": 2842
    },
    {
      "epoch": 0.4564868336544637,
      "grad_norm": 0.25408902764320374,
      "learning_rate": 0.0001,
      "loss": 1.5318,
      "step": 2843
    },
    {
      "epoch": 0.45664739884393063,
      "grad_norm": 0.24858152866363525,
      "learning_rate": 0.0001,
      "loss": 1.4884,
      "step": 2844
    },
    {
      "epoch": 0.4568079640333976,
      "grad_norm": 0.2565762996673584,
      "learning_rate": 0.0001,
      "loss": 1.493,
      "step": 2845
    },
    {
      "epoch": 0.45696852922286446,
      "grad_norm": 0.25907644629478455,
      "learning_rate": 0.0001,
      "loss": 1.521,
      "step": 2846
    },
    {
      "epoch": 0.4571290944123314,
      "grad_norm": 0.278012216091156,
      "learning_rate": 0.0001,
      "loss": 1.6365,
      "step": 2847
    },
    {
      "epoch": 0.45728965960179835,
      "grad_norm": 0.25443723797798157,
      "learning_rate": 0.0001,
      "loss": 1.4585,
      "step": 2848
    },
    {
      "epoch": 0.45745022479126524,
      "grad_norm": 0.2635557949542999,
      "learning_rate": 0.0001,
      "loss": 1.516,
      "step": 2849
    },
    {
      "epoch": 0.4576107899807322,
      "grad_norm": 0.2510010898113251,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 2850
    },
    {
      "epoch": 0.4577713551701991,
      "grad_norm": 0.2531384527683258,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 2851
    },
    {
      "epoch": 0.457931920359666,
      "grad_norm": 0.2632436454296112,
      "learning_rate": 0.0001,
      "loss": 1.6984,
      "step": 2852
    },
    {
      "epoch": 0.45809248554913296,
      "grad_norm": 0.25483858585357666,
      "learning_rate": 0.0001,
      "loss": 1.4538,
      "step": 2853
    },
    {
      "epoch": 0.4582530507385999,
      "grad_norm": 0.25285083055496216,
      "learning_rate": 0.0001,
      "loss": 1.5308,
      "step": 2854
    },
    {
      "epoch": 0.4584136159280668,
      "grad_norm": 0.26279085874557495,
      "learning_rate": 0.0001,
      "loss": 1.5755,
      "step": 2855
    },
    {
      "epoch": 0.45857418111753373,
      "grad_norm": 0.24838216602802277,
      "learning_rate": 0.0001,
      "loss": 1.5077,
      "step": 2856
    },
    {
      "epoch": 0.4587347463070006,
      "grad_norm": 0.2317473590373993,
      "learning_rate": 0.0001,
      "loss": 1.4682,
      "step": 2857
    },
    {
      "epoch": 0.45889531149646756,
      "grad_norm": 0.24912071228027344,
      "learning_rate": 0.0001,
      "loss": 1.528,
      "step": 2858
    },
    {
      "epoch": 0.4590558766859345,
      "grad_norm": 0.25990068912506104,
      "learning_rate": 0.0001,
      "loss": 1.5177,
      "step": 2859
    },
    {
      "epoch": 0.4592164418754014,
      "grad_norm": 0.26524385809898376,
      "learning_rate": 0.0001,
      "loss": 1.6028,
      "step": 2860
    },
    {
      "epoch": 0.45937700706486834,
      "grad_norm": 0.2510650157928467,
      "learning_rate": 0.0001,
      "loss": 1.4484,
      "step": 2861
    },
    {
      "epoch": 0.4595375722543353,
      "grad_norm": 0.24385692179203033,
      "learning_rate": 0.0001,
      "loss": 1.5092,
      "step": 2862
    },
    {
      "epoch": 0.45969813744380217,
      "grad_norm": 0.26305243372917175,
      "learning_rate": 0.0001,
      "loss": 1.4875,
      "step": 2863
    },
    {
      "epoch": 0.4598587026332691,
      "grad_norm": 0.24691642820835114,
      "learning_rate": 0.0001,
      "loss": 1.5286,
      "step": 2864
    },
    {
      "epoch": 0.46001926782273606,
      "grad_norm": 0.24307402968406677,
      "learning_rate": 0.0001,
      "loss": 1.4736,
      "step": 2865
    },
    {
      "epoch": 0.46017983301220294,
      "grad_norm": 0.2772844433784485,
      "learning_rate": 0.0001,
      "loss": 1.6063,
      "step": 2866
    },
    {
      "epoch": 0.4603403982016699,
      "grad_norm": 0.2490212619304657,
      "learning_rate": 0.0001,
      "loss": 1.4737,
      "step": 2867
    },
    {
      "epoch": 0.4605009633911368,
      "grad_norm": 0.27187564969062805,
      "learning_rate": 0.0001,
      "loss": 1.5311,
      "step": 2868
    },
    {
      "epoch": 0.4606615285806037,
      "grad_norm": 0.25558462738990784,
      "learning_rate": 0.0001,
      "loss": 1.4869,
      "step": 2869
    },
    {
      "epoch": 0.46082209377007066,
      "grad_norm": 0.2559339702129364,
      "learning_rate": 0.0001,
      "loss": 1.5277,
      "step": 2870
    },
    {
      "epoch": 0.46098265895953755,
      "grad_norm": 0.2592107951641083,
      "learning_rate": 0.0001,
      "loss": 1.4847,
      "step": 2871
    },
    {
      "epoch": 0.4611432241490045,
      "grad_norm": 0.2574782073497772,
      "learning_rate": 0.0001,
      "loss": 1.5467,
      "step": 2872
    },
    {
      "epoch": 0.46130378933847144,
      "grad_norm": 0.25029680132865906,
      "learning_rate": 0.0001,
      "loss": 1.5419,
      "step": 2873
    },
    {
      "epoch": 0.4614643545279383,
      "grad_norm": 0.9509021639823914,
      "learning_rate": 0.0001,
      "loss": 1.5162,
      "step": 2874
    },
    {
      "epoch": 0.46162491971740527,
      "grad_norm": 0.2531764805316925,
      "learning_rate": 0.0001,
      "loss": 1.5323,
      "step": 2875
    },
    {
      "epoch": 0.4617854849068722,
      "grad_norm": 0.245522141456604,
      "learning_rate": 0.0001,
      "loss": 1.5567,
      "step": 2876
    },
    {
      "epoch": 0.4619460500963391,
      "grad_norm": 0.26436519622802734,
      "learning_rate": 0.0001,
      "loss": 1.5695,
      "step": 2877
    },
    {
      "epoch": 0.46210661528580604,
      "grad_norm": 0.2703956663608551,
      "learning_rate": 0.0001,
      "loss": 1.5229,
      "step": 2878
    },
    {
      "epoch": 0.462267180475273,
      "grad_norm": 0.26570844650268555,
      "learning_rate": 0.0001,
      "loss": 1.5541,
      "step": 2879
    },
    {
      "epoch": 0.4624277456647399,
      "grad_norm": 0.26499056816101074,
      "learning_rate": 0.0001,
      "loss": 1.5506,
      "step": 2880
    },
    {
      "epoch": 0.4625883108542068,
      "grad_norm": 0.27163347601890564,
      "learning_rate": 0.0001,
      "loss": 1.5466,
      "step": 2881
    },
    {
      "epoch": 0.4627488760436737,
      "grad_norm": 0.26002606749534607,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 2882
    },
    {
      "epoch": 0.46290944123314065,
      "grad_norm": 0.24343660473823547,
      "learning_rate": 0.0001,
      "loss": 1.514,
      "step": 2883
    },
    {
      "epoch": 0.4630700064226076,
      "grad_norm": 0.2543664574623108,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 2884
    },
    {
      "epoch": 0.4632305716120745,
      "grad_norm": 0.2580462396144867,
      "learning_rate": 0.0001,
      "loss": 1.5423,
      "step": 2885
    },
    {
      "epoch": 0.4633911368015414,
      "grad_norm": 0.259204626083374,
      "learning_rate": 0.0001,
      "loss": 1.5541,
      "step": 2886
    },
    {
      "epoch": 0.46355170199100837,
      "grad_norm": 0.26318368315696716,
      "learning_rate": 0.0001,
      "loss": 1.5519,
      "step": 2887
    },
    {
      "epoch": 0.46371226718047526,
      "grad_norm": 0.26448196172714233,
      "learning_rate": 0.0001,
      "loss": 1.5723,
      "step": 2888
    },
    {
      "epoch": 0.4638728323699422,
      "grad_norm": 0.25672727823257446,
      "learning_rate": 0.0001,
      "loss": 1.5213,
      "step": 2889
    },
    {
      "epoch": 0.46403339755940914,
      "grad_norm": 0.2626258134841919,
      "learning_rate": 0.0001,
      "loss": 1.6652,
      "step": 2890
    },
    {
      "epoch": 0.46419396274887603,
      "grad_norm": 0.255079060792923,
      "learning_rate": 0.0001,
      "loss": 1.5671,
      "step": 2891
    },
    {
      "epoch": 0.464354527938343,
      "grad_norm": 0.2678382694721222,
      "learning_rate": 0.0001,
      "loss": 1.5555,
      "step": 2892
    },
    {
      "epoch": 0.46451509312780986,
      "grad_norm": 0.2653570771217346,
      "learning_rate": 0.0001,
      "loss": 1.5375,
      "step": 2893
    },
    {
      "epoch": 0.4646756583172768,
      "grad_norm": 0.2625160813331604,
      "learning_rate": 0.0001,
      "loss": 1.5035,
      "step": 2894
    },
    {
      "epoch": 0.46483622350674375,
      "grad_norm": 0.2426673173904419,
      "learning_rate": 0.0001,
      "loss": 1.4554,
      "step": 2895
    },
    {
      "epoch": 0.46499678869621064,
      "grad_norm": 0.25273504853248596,
      "learning_rate": 0.0001,
      "loss": 1.4972,
      "step": 2896
    },
    {
      "epoch": 0.4651573538856776,
      "grad_norm": 0.2501536011695862,
      "learning_rate": 0.0001,
      "loss": 1.4182,
      "step": 2897
    },
    {
      "epoch": 0.4653179190751445,
      "grad_norm": 0.25146669149398804,
      "learning_rate": 0.0001,
      "loss": 1.5331,
      "step": 2898
    },
    {
      "epoch": 0.4654784842646114,
      "grad_norm": 0.25322243571281433,
      "learning_rate": 0.0001,
      "loss": 1.5151,
      "step": 2899
    },
    {
      "epoch": 0.46563904945407836,
      "grad_norm": 0.26858463883399963,
      "learning_rate": 0.0001,
      "loss": 1.6399,
      "step": 2900
    },
    {
      "epoch": 0.4657996146435453,
      "grad_norm": 0.2575323283672333,
      "learning_rate": 0.0001,
      "loss": 1.5164,
      "step": 2901
    },
    {
      "epoch": 0.4659601798330122,
      "grad_norm": 0.2752355635166168,
      "learning_rate": 0.0001,
      "loss": 1.5762,
      "step": 2902
    },
    {
      "epoch": 0.46612074502247913,
      "grad_norm": 0.2613184452056885,
      "learning_rate": 0.0001,
      "loss": 1.585,
      "step": 2903
    },
    {
      "epoch": 0.4662813102119461,
      "grad_norm": 0.2515537142753601,
      "learning_rate": 0.0001,
      "loss": 1.5035,
      "step": 2904
    },
    {
      "epoch": 0.46644187540141296,
      "grad_norm": 0.2640692889690399,
      "learning_rate": 0.0001,
      "loss": 1.6099,
      "step": 2905
    },
    {
      "epoch": 0.4666024405908799,
      "grad_norm": 0.26171159744262695,
      "learning_rate": 0.0001,
      "loss": 1.5364,
      "step": 2906
    },
    {
      "epoch": 0.4667630057803468,
      "grad_norm": 0.25064030289649963,
      "learning_rate": 0.0001,
      "loss": 1.5213,
      "step": 2907
    },
    {
      "epoch": 0.46692357096981374,
      "grad_norm": 0.26227661967277527,
      "learning_rate": 0.0001,
      "loss": 1.6114,
      "step": 2908
    },
    {
      "epoch": 0.4670841361592807,
      "grad_norm": 0.24718691408634186,
      "learning_rate": 0.0001,
      "loss": 1.5096,
      "step": 2909
    },
    {
      "epoch": 0.46724470134874757,
      "grad_norm": 0.251468300819397,
      "learning_rate": 0.0001,
      "loss": 1.5939,
      "step": 2910
    },
    {
      "epoch": 0.4674052665382145,
      "grad_norm": 0.24605131149291992,
      "learning_rate": 0.0001,
      "loss": 1.448,
      "step": 2911
    },
    {
      "epoch": 0.46756583172768146,
      "grad_norm": 0.25557634234428406,
      "learning_rate": 0.0001,
      "loss": 1.4951,
      "step": 2912
    },
    {
      "epoch": 0.46772639691714835,
      "grad_norm": 0.2542295455932617,
      "learning_rate": 0.0001,
      "loss": 1.4564,
      "step": 2913
    },
    {
      "epoch": 0.4678869621066153,
      "grad_norm": 0.25903230905532837,
      "learning_rate": 0.0001,
      "loss": 1.4722,
      "step": 2914
    },
    {
      "epoch": 0.46804752729608223,
      "grad_norm": 0.2592344880104065,
      "learning_rate": 0.0001,
      "loss": 1.5616,
      "step": 2915
    },
    {
      "epoch": 0.4682080924855491,
      "grad_norm": 0.25056353211402893,
      "learning_rate": 0.0001,
      "loss": 1.4715,
      "step": 2916
    },
    {
      "epoch": 0.46836865767501606,
      "grad_norm": 0.26791608333587646,
      "learning_rate": 0.0001,
      "loss": 1.623,
      "step": 2917
    },
    {
      "epoch": 0.46852922286448295,
      "grad_norm": 0.25067391991615295,
      "learning_rate": 0.0001,
      "loss": 1.5268,
      "step": 2918
    },
    {
      "epoch": 0.4686897880539499,
      "grad_norm": 0.25362953543663025,
      "learning_rate": 0.0001,
      "loss": 1.5774,
      "step": 2919
    },
    {
      "epoch": 0.46885035324341684,
      "grad_norm": 0.24526727199554443,
      "learning_rate": 0.0001,
      "loss": 1.5023,
      "step": 2920
    },
    {
      "epoch": 0.4690109184328837,
      "grad_norm": 0.2462223917245865,
      "learning_rate": 0.0001,
      "loss": 1.4341,
      "step": 2921
    },
    {
      "epoch": 0.46917148362235067,
      "grad_norm": 0.24873018264770508,
      "learning_rate": 0.0001,
      "loss": 1.5341,
      "step": 2922
    },
    {
      "epoch": 0.4693320488118176,
      "grad_norm": 0.24852591753005981,
      "learning_rate": 0.0001,
      "loss": 1.538,
      "step": 2923
    },
    {
      "epoch": 0.4694926140012845,
      "grad_norm": 0.24487623572349548,
      "learning_rate": 0.0001,
      "loss": 1.4673,
      "step": 2924
    },
    {
      "epoch": 0.46965317919075145,
      "grad_norm": 0.25096890330314636,
      "learning_rate": 0.0001,
      "loss": 1.4821,
      "step": 2925
    },
    {
      "epoch": 0.4698137443802184,
      "grad_norm": 0.2560618817806244,
      "learning_rate": 0.0001,
      "loss": 1.5734,
      "step": 2926
    },
    {
      "epoch": 0.4699743095696853,
      "grad_norm": 0.24424079060554504,
      "learning_rate": 0.0001,
      "loss": 1.5163,
      "step": 2927
    },
    {
      "epoch": 0.4701348747591522,
      "grad_norm": 0.2547134459018707,
      "learning_rate": 0.0001,
      "loss": 1.5492,
      "step": 2928
    },
    {
      "epoch": 0.47029543994861917,
      "grad_norm": 0.25469645857810974,
      "learning_rate": 0.0001,
      "loss": 1.6081,
      "step": 2929
    },
    {
      "epoch": 0.47045600513808605,
      "grad_norm": 0.2547605335712433,
      "learning_rate": 0.0001,
      "loss": 1.5611,
      "step": 2930
    },
    {
      "epoch": 0.470616570327553,
      "grad_norm": 0.2461174577474594,
      "learning_rate": 0.0001,
      "loss": 1.5565,
      "step": 2931
    },
    {
      "epoch": 0.4707771355170199,
      "grad_norm": 0.26163119077682495,
      "learning_rate": 0.0001,
      "loss": 1.5466,
      "step": 2932
    },
    {
      "epoch": 0.47093770070648683,
      "grad_norm": 0.24534977972507477,
      "learning_rate": 0.0001,
      "loss": 1.4996,
      "step": 2933
    },
    {
      "epoch": 0.47109826589595377,
      "grad_norm": 0.25218233466148376,
      "learning_rate": 0.0001,
      "loss": 1.4699,
      "step": 2934
    },
    {
      "epoch": 0.47125883108542066,
      "grad_norm": 0.24794015288352966,
      "learning_rate": 0.0001,
      "loss": 1.4439,
      "step": 2935
    },
    {
      "epoch": 0.4714193962748876,
      "grad_norm": 0.2545740008354187,
      "learning_rate": 0.0001,
      "loss": 1.4699,
      "step": 2936
    },
    {
      "epoch": 0.47157996146435455,
      "grad_norm": 0.2666550874710083,
      "learning_rate": 0.0001,
      "loss": 1.634,
      "step": 2937
    },
    {
      "epoch": 0.47174052665382143,
      "grad_norm": 0.2550573945045471,
      "learning_rate": 0.0001,
      "loss": 1.5672,
      "step": 2938
    },
    {
      "epoch": 0.4719010918432884,
      "grad_norm": 0.2551001310348511,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 2939
    },
    {
      "epoch": 0.4720616570327553,
      "grad_norm": 0.24900071322917938,
      "learning_rate": 0.0001,
      "loss": 1.4897,
      "step": 2940
    },
    {
      "epoch": 0.4722222222222222,
      "grad_norm": 0.2513026297092438,
      "learning_rate": 0.0001,
      "loss": 1.5128,
      "step": 2941
    },
    {
      "epoch": 0.47238278741168915,
      "grad_norm": 0.2582988739013672,
      "learning_rate": 0.0001,
      "loss": 1.5086,
      "step": 2942
    },
    {
      "epoch": 0.4725433526011561,
      "grad_norm": 0.24734970927238464,
      "learning_rate": 0.0001,
      "loss": 1.5136,
      "step": 2943
    },
    {
      "epoch": 0.472703917790623,
      "grad_norm": 1.768375277519226,
      "learning_rate": 0.0001,
      "loss": 1.4875,
      "step": 2944
    },
    {
      "epoch": 0.47286448298008993,
      "grad_norm": 0.25934040546417236,
      "learning_rate": 0.0001,
      "loss": 1.5262,
      "step": 2945
    },
    {
      "epoch": 0.4730250481695568,
      "grad_norm": 0.2954626977443695,
      "learning_rate": 0.0001,
      "loss": 1.5517,
      "step": 2946
    },
    {
      "epoch": 0.47318561335902376,
      "grad_norm": 0.29073503613471985,
      "learning_rate": 0.0001,
      "loss": 1.5996,
      "step": 2947
    },
    {
      "epoch": 0.4733461785484907,
      "grad_norm": 0.2698410153388977,
      "learning_rate": 0.0001,
      "loss": 1.5954,
      "step": 2948
    },
    {
      "epoch": 0.4735067437379576,
      "grad_norm": 0.24963271617889404,
      "learning_rate": 0.0001,
      "loss": 1.4816,
      "step": 2949
    },
    {
      "epoch": 0.47366730892742454,
      "grad_norm": 0.2844211161136627,
      "learning_rate": 0.0001,
      "loss": 1.564,
      "step": 2950
    },
    {
      "epoch": 0.4738278741168915,
      "grad_norm": 0.2843918800354004,
      "learning_rate": 0.0001,
      "loss": 1.5153,
      "step": 2951
    },
    {
      "epoch": 0.47398843930635837,
      "grad_norm": 0.27777576446533203,
      "learning_rate": 0.0001,
      "loss": 1.5159,
      "step": 2952
    },
    {
      "epoch": 0.4741490044958253,
      "grad_norm": 0.3000965118408203,
      "learning_rate": 0.0001,
      "loss": 1.5341,
      "step": 2953
    },
    {
      "epoch": 0.47430956968529225,
      "grad_norm": 0.3191080093383789,
      "learning_rate": 0.0001,
      "loss": 1.5554,
      "step": 2954
    },
    {
      "epoch": 0.47447013487475914,
      "grad_norm": 0.2943609058856964,
      "learning_rate": 0.0001,
      "loss": 1.5191,
      "step": 2955
    },
    {
      "epoch": 0.4746307000642261,
      "grad_norm": 0.267706036567688,
      "learning_rate": 0.0001,
      "loss": 1.5717,
      "step": 2956
    },
    {
      "epoch": 0.474791265253693,
      "grad_norm": 0.2766854763031006,
      "learning_rate": 0.0001,
      "loss": 1.5028,
      "step": 2957
    },
    {
      "epoch": 0.4749518304431599,
      "grad_norm": 0.2598714232444763,
      "learning_rate": 0.0001,
      "loss": 1.4789,
      "step": 2958
    },
    {
      "epoch": 0.47511239563262686,
      "grad_norm": 0.31471943855285645,
      "learning_rate": 0.0001,
      "loss": 1.4731,
      "step": 2959
    },
    {
      "epoch": 0.47527296082209375,
      "grad_norm": 0.26923859119415283,
      "learning_rate": 0.0001,
      "loss": 1.5163,
      "step": 2960
    },
    {
      "epoch": 0.4754335260115607,
      "grad_norm": 0.3559715151786804,
      "learning_rate": 0.0001,
      "loss": 1.5698,
      "step": 2961
    },
    {
      "epoch": 0.47559409120102764,
      "grad_norm": 0.32424938678741455,
      "learning_rate": 0.0001,
      "loss": 1.5003,
      "step": 2962
    },
    {
      "epoch": 0.4757546563904945,
      "grad_norm": 0.2876241207122803,
      "learning_rate": 0.0001,
      "loss": 1.5818,
      "step": 2963
    },
    {
      "epoch": 0.47591522157996147,
      "grad_norm": 0.2469119131565094,
      "learning_rate": 0.0001,
      "loss": 1.5285,
      "step": 2964
    },
    {
      "epoch": 0.4760757867694284,
      "grad_norm": 0.26568013429641724,
      "learning_rate": 0.0001,
      "loss": 1.5193,
      "step": 2965
    },
    {
      "epoch": 0.4762363519588953,
      "grad_norm": 0.2633431553840637,
      "learning_rate": 0.0001,
      "loss": 1.5456,
      "step": 2966
    },
    {
      "epoch": 0.47639691714836224,
      "grad_norm": 0.2920028269290924,
      "learning_rate": 0.0001,
      "loss": 1.5056,
      "step": 2967
    },
    {
      "epoch": 0.4765574823378292,
      "grad_norm": 0.2750384211540222,
      "learning_rate": 0.0001,
      "loss": 1.4879,
      "step": 2968
    },
    {
      "epoch": 0.4767180475272961,
      "grad_norm": 0.25811445713043213,
      "learning_rate": 0.0001,
      "loss": 1.5274,
      "step": 2969
    },
    {
      "epoch": 0.476878612716763,
      "grad_norm": 0.2579619586467743,
      "learning_rate": 0.0001,
      "loss": 1.5256,
      "step": 2970
    },
    {
      "epoch": 0.4770391779062299,
      "grad_norm": 0.2626616060733795,
      "learning_rate": 0.0001,
      "loss": 1.4944,
      "step": 2971
    },
    {
      "epoch": 0.47719974309569685,
      "grad_norm": 0.28014877438545227,
      "learning_rate": 0.0001,
      "loss": 1.5417,
      "step": 2972
    },
    {
      "epoch": 0.4773603082851638,
      "grad_norm": 0.25855451822280884,
      "learning_rate": 0.0001,
      "loss": 1.4902,
      "step": 2973
    },
    {
      "epoch": 0.4775208734746307,
      "grad_norm": 0.26190412044525146,
      "learning_rate": 0.0001,
      "loss": 1.5015,
      "step": 2974
    },
    {
      "epoch": 0.4776814386640976,
      "grad_norm": 0.264337420463562,
      "learning_rate": 0.0001,
      "loss": 1.4599,
      "step": 2975
    },
    {
      "epoch": 0.47784200385356457,
      "grad_norm": 0.25439706444740295,
      "learning_rate": 0.0001,
      "loss": 1.489,
      "step": 2976
    },
    {
      "epoch": 0.47800256904303146,
      "grad_norm": 0.2698615491390228,
      "learning_rate": 0.0001,
      "loss": 1.6834,
      "step": 2977
    },
    {
      "epoch": 0.4781631342324984,
      "grad_norm": 0.25165149569511414,
      "learning_rate": 0.0001,
      "loss": 1.4425,
      "step": 2978
    },
    {
      "epoch": 0.47832369942196534,
      "grad_norm": 0.2794475853443146,
      "learning_rate": 0.0001,
      "loss": 1.6159,
      "step": 2979
    },
    {
      "epoch": 0.47848426461143223,
      "grad_norm": 0.23459677398204803,
      "learning_rate": 0.0001,
      "loss": 1.4629,
      "step": 2980
    },
    {
      "epoch": 0.4786448298008992,
      "grad_norm": 0.24268727004528046,
      "learning_rate": 0.0001,
      "loss": 1.5017,
      "step": 2981
    },
    {
      "epoch": 0.47880539499036606,
      "grad_norm": 0.2444038689136505,
      "learning_rate": 0.0001,
      "loss": 1.4956,
      "step": 2982
    },
    {
      "epoch": 0.478965960179833,
      "grad_norm": 0.263205885887146,
      "learning_rate": 0.0001,
      "loss": 1.57,
      "step": 2983
    },
    {
      "epoch": 0.47912652536929995,
      "grad_norm": 0.26336947083473206,
      "learning_rate": 0.0001,
      "loss": 1.5825,
      "step": 2984
    },
    {
      "epoch": 0.47928709055876684,
      "grad_norm": 0.25302061438560486,
      "learning_rate": 0.0001,
      "loss": 1.4827,
      "step": 2985
    },
    {
      "epoch": 0.4794476557482338,
      "grad_norm": 0.26255595684051514,
      "learning_rate": 0.0001,
      "loss": 1.5722,
      "step": 2986
    },
    {
      "epoch": 0.4796082209377007,
      "grad_norm": 0.26336580514907837,
      "learning_rate": 0.0001,
      "loss": 1.5097,
      "step": 2987
    },
    {
      "epoch": 0.4797687861271676,
      "grad_norm": 0.24177514016628265,
      "learning_rate": 0.0001,
      "loss": 1.409,
      "step": 2988
    },
    {
      "epoch": 0.47992935131663456,
      "grad_norm": 0.24892064929008484,
      "learning_rate": 0.0001,
      "loss": 1.4953,
      "step": 2989
    },
    {
      "epoch": 0.4800899165061015,
      "grad_norm": 0.2624559700489044,
      "learning_rate": 0.0001,
      "loss": 1.542,
      "step": 2990
    },
    {
      "epoch": 0.4802504816955684,
      "grad_norm": 0.27438074350357056,
      "learning_rate": 0.0001,
      "loss": 1.6751,
      "step": 2991
    },
    {
      "epoch": 0.48041104688503533,
      "grad_norm": 0.26733386516571045,
      "learning_rate": 0.0001,
      "loss": 1.6132,
      "step": 2992
    },
    {
      "epoch": 0.4805716120745023,
      "grad_norm": 0.2562370002269745,
      "learning_rate": 0.0001,
      "loss": 1.5572,
      "step": 2993
    },
    {
      "epoch": 0.48073217726396916,
      "grad_norm": 0.25975048542022705,
      "learning_rate": 0.0001,
      "loss": 1.6142,
      "step": 2994
    },
    {
      "epoch": 0.4808927424534361,
      "grad_norm": 0.2624019384384155,
      "learning_rate": 0.0001,
      "loss": 1.5079,
      "step": 2995
    },
    {
      "epoch": 0.481053307642903,
      "grad_norm": 0.25996801257133484,
      "learning_rate": 0.0001,
      "loss": 1.5772,
      "step": 2996
    },
    {
      "epoch": 0.48121387283236994,
      "grad_norm": 0.2643839418888092,
      "learning_rate": 0.0001,
      "loss": 1.57,
      "step": 2997
    },
    {
      "epoch": 0.4813744380218369,
      "grad_norm": 0.25774019956588745,
      "learning_rate": 0.0001,
      "loss": 1.5286,
      "step": 2998
    },
    {
      "epoch": 0.48153500321130377,
      "grad_norm": 0.2576081156730652,
      "learning_rate": 0.0001,
      "loss": 1.5145,
      "step": 2999
    },
    {
      "epoch": 0.4816955684007707,
      "grad_norm": 0.25839635729789734,
      "learning_rate": 0.0001,
      "loss": 1.5477,
      "step": 3000
    },
    {
      "epoch": 0.48185613359023766,
      "grad_norm": 0.2544023096561432,
      "learning_rate": 0.0001,
      "loss": 1.556,
      "step": 3001
    },
    {
      "epoch": 0.48201669877970454,
      "grad_norm": 0.2556130588054657,
      "learning_rate": 0.0001,
      "loss": 1.5974,
      "step": 3002
    },
    {
      "epoch": 0.4821772639691715,
      "grad_norm": 0.2694602608680725,
      "learning_rate": 0.0001,
      "loss": 1.4956,
      "step": 3003
    },
    {
      "epoch": 0.48233782915863843,
      "grad_norm": 0.24750924110412598,
      "learning_rate": 0.0001,
      "loss": 1.4942,
      "step": 3004
    },
    {
      "epoch": 0.4824983943481053,
      "grad_norm": 0.25541990995407104,
      "learning_rate": 0.0001,
      "loss": 1.4511,
      "step": 3005
    },
    {
      "epoch": 0.48265895953757226,
      "grad_norm": 0.2663719058036804,
      "learning_rate": 0.0001,
      "loss": 1.578,
      "step": 3006
    },
    {
      "epoch": 0.48281952472703915,
      "grad_norm": 0.26437413692474365,
      "learning_rate": 0.0001,
      "loss": 1.5265,
      "step": 3007
    },
    {
      "epoch": 0.4829800899165061,
      "grad_norm": 0.24333199858665466,
      "learning_rate": 0.0001,
      "loss": 1.4902,
      "step": 3008
    },
    {
      "epoch": 0.48314065510597304,
      "grad_norm": 0.2512253224849701,
      "learning_rate": 0.0001,
      "loss": 1.4467,
      "step": 3009
    },
    {
      "epoch": 0.4833012202954399,
      "grad_norm": 0.2544972002506256,
      "learning_rate": 0.0001,
      "loss": 1.547,
      "step": 3010
    },
    {
      "epoch": 0.48346178548490687,
      "grad_norm": 0.2563701868057251,
      "learning_rate": 0.0001,
      "loss": 1.5568,
      "step": 3011
    },
    {
      "epoch": 0.4836223506743738,
      "grad_norm": 0.2565440535545349,
      "learning_rate": 0.0001,
      "loss": 1.456,
      "step": 3012
    },
    {
      "epoch": 0.4837829158638407,
      "grad_norm": 0.27084702253341675,
      "learning_rate": 0.0001,
      "loss": 1.542,
      "step": 3013
    },
    {
      "epoch": 0.48394348105330764,
      "grad_norm": 0.27865681052207947,
      "learning_rate": 0.0001,
      "loss": 1.5371,
      "step": 3014
    },
    {
      "epoch": 0.4841040462427746,
      "grad_norm": 0.25021904706954956,
      "learning_rate": 0.0001,
      "loss": 1.5203,
      "step": 3015
    },
    {
      "epoch": 0.4842646114322415,
      "grad_norm": 0.27153530716896057,
      "learning_rate": 0.0001,
      "loss": 1.5202,
      "step": 3016
    },
    {
      "epoch": 0.4844251766217084,
      "grad_norm": 0.28725340962409973,
      "learning_rate": 0.0001,
      "loss": 1.534,
      "step": 3017
    },
    {
      "epoch": 0.48458574181117536,
      "grad_norm": 0.2531295716762543,
      "learning_rate": 0.0001,
      "loss": 1.5382,
      "step": 3018
    },
    {
      "epoch": 0.48474630700064225,
      "grad_norm": 0.26707106828689575,
      "learning_rate": 0.0001,
      "loss": 1.481,
      "step": 3019
    },
    {
      "epoch": 0.4849068721901092,
      "grad_norm": 0.2601153254508972,
      "learning_rate": 0.0001,
      "loss": 1.4935,
      "step": 3020
    },
    {
      "epoch": 0.4850674373795761,
      "grad_norm": 0.23785848915576935,
      "learning_rate": 0.0001,
      "loss": 1.4438,
      "step": 3021
    },
    {
      "epoch": 0.485228002569043,
      "grad_norm": 0.2714046239852905,
      "learning_rate": 0.0001,
      "loss": 1.5395,
      "step": 3022
    },
    {
      "epoch": 0.48538856775850997,
      "grad_norm": 0.2642495334148407,
      "learning_rate": 0.0001,
      "loss": 1.5503,
      "step": 3023
    },
    {
      "epoch": 0.48554913294797686,
      "grad_norm": 0.2607245147228241,
      "learning_rate": 0.0001,
      "loss": 1.4963,
      "step": 3024
    },
    {
      "epoch": 0.4857096981374438,
      "grad_norm": 0.26534876227378845,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 3025
    },
    {
      "epoch": 0.48587026332691075,
      "grad_norm": 0.2559542655944824,
      "learning_rate": 0.0001,
      "loss": 1.4475,
      "step": 3026
    },
    {
      "epoch": 0.48603082851637763,
      "grad_norm": 0.2548435628414154,
      "learning_rate": 0.0001,
      "loss": 1.6015,
      "step": 3027
    },
    {
      "epoch": 0.4861913937058446,
      "grad_norm": 0.2658431828022003,
      "learning_rate": 0.0001,
      "loss": 1.5076,
      "step": 3028
    },
    {
      "epoch": 0.4863519588953115,
      "grad_norm": 0.28509068489074707,
      "learning_rate": 0.0001,
      "loss": 1.505,
      "step": 3029
    },
    {
      "epoch": 0.4865125240847784,
      "grad_norm": 0.2549109160900116,
      "learning_rate": 0.0001,
      "loss": 1.5436,
      "step": 3030
    },
    {
      "epoch": 0.48667308927424535,
      "grad_norm": 0.41223037242889404,
      "learning_rate": 0.0001,
      "loss": 1.5205,
      "step": 3031
    },
    {
      "epoch": 0.48683365446371224,
      "grad_norm": 0.2568729519844055,
      "learning_rate": 0.0001,
      "loss": 1.4396,
      "step": 3032
    },
    {
      "epoch": 0.4869942196531792,
      "grad_norm": 0.24964162707328796,
      "learning_rate": 0.0001,
      "loss": 1.4121,
      "step": 3033
    },
    {
      "epoch": 0.4871547848426461,
      "grad_norm": 0.2588632106781006,
      "learning_rate": 0.0001,
      "loss": 1.5287,
      "step": 3034
    },
    {
      "epoch": 0.487315350032113,
      "grad_norm": 0.25589656829833984,
      "learning_rate": 0.0001,
      "loss": 1.5095,
      "step": 3035
    },
    {
      "epoch": 0.48747591522157996,
      "grad_norm": 0.2549636960029602,
      "learning_rate": 0.0001,
      "loss": 1.5844,
      "step": 3036
    },
    {
      "epoch": 0.4876364804110469,
      "grad_norm": 0.24072504043579102,
      "learning_rate": 0.0001,
      "loss": 1.4383,
      "step": 3037
    },
    {
      "epoch": 0.4877970456005138,
      "grad_norm": 0.24526798725128174,
      "learning_rate": 0.0001,
      "loss": 1.4131,
      "step": 3038
    },
    {
      "epoch": 0.48795761078998073,
      "grad_norm": 0.2538144290447235,
      "learning_rate": 0.0001,
      "loss": 1.5171,
      "step": 3039
    },
    {
      "epoch": 0.4881181759794477,
      "grad_norm": 0.2606353163719177,
      "learning_rate": 0.0001,
      "loss": 1.5911,
      "step": 3040
    },
    {
      "epoch": 0.48827874116891457,
      "grad_norm": 0.25958704948425293,
      "learning_rate": 0.0001,
      "loss": 1.5363,
      "step": 3041
    },
    {
      "epoch": 0.4884393063583815,
      "grad_norm": 0.2852924168109894,
      "learning_rate": 0.0001,
      "loss": 1.5482,
      "step": 3042
    },
    {
      "epoch": 0.48859987154784845,
      "grad_norm": 0.25527581572532654,
      "learning_rate": 0.0001,
      "loss": 1.5255,
      "step": 3043
    },
    {
      "epoch": 0.48876043673731534,
      "grad_norm": 0.2534925937652588,
      "learning_rate": 0.0001,
      "loss": 1.4663,
      "step": 3044
    },
    {
      "epoch": 0.4889210019267823,
      "grad_norm": 0.27186769247055054,
      "learning_rate": 0.0001,
      "loss": 1.4537,
      "step": 3045
    },
    {
      "epoch": 0.48908156711624917,
      "grad_norm": 0.28862160444259644,
      "learning_rate": 0.0001,
      "loss": 1.5088,
      "step": 3046
    },
    {
      "epoch": 0.4892421323057161,
      "grad_norm": 0.270006388425827,
      "learning_rate": 0.0001,
      "loss": 1.5399,
      "step": 3047
    },
    {
      "epoch": 0.48940269749518306,
      "grad_norm": 0.26035580039024353,
      "learning_rate": 0.0001,
      "loss": 1.5279,
      "step": 3048
    },
    {
      "epoch": 0.48956326268464995,
      "grad_norm": 0.3059391677379608,
      "learning_rate": 0.0001,
      "loss": 1.6101,
      "step": 3049
    },
    {
      "epoch": 0.4897238278741169,
      "grad_norm": 0.2891250252723694,
      "learning_rate": 0.0001,
      "loss": 1.524,
      "step": 3050
    },
    {
      "epoch": 0.48988439306358383,
      "grad_norm": 0.2657817304134369,
      "learning_rate": 0.0001,
      "loss": 1.473,
      "step": 3051
    },
    {
      "epoch": 0.4900449582530507,
      "grad_norm": 0.27785712480545044,
      "learning_rate": 0.0001,
      "loss": 1.5539,
      "step": 3052
    },
    {
      "epoch": 0.49020552344251767,
      "grad_norm": 0.23967069387435913,
      "learning_rate": 0.0001,
      "loss": 1.5034,
      "step": 3053
    },
    {
      "epoch": 0.4903660886319846,
      "grad_norm": 0.2672518491744995,
      "learning_rate": 0.0001,
      "loss": 1.4786,
      "step": 3054
    },
    {
      "epoch": 0.4905266538214515,
      "grad_norm": 0.25132396817207336,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 3055
    },
    {
      "epoch": 0.49068721901091844,
      "grad_norm": 0.2554181218147278,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 3056
    },
    {
      "epoch": 0.4908477842003854,
      "grad_norm": 0.24807313084602356,
      "learning_rate": 0.0001,
      "loss": 1.4963,
      "step": 3057
    },
    {
      "epoch": 0.49100834938985227,
      "grad_norm": 0.2602437138557434,
      "learning_rate": 0.0001,
      "loss": 1.5575,
      "step": 3058
    },
    {
      "epoch": 0.4911689145793192,
      "grad_norm": 0.26603472232818604,
      "learning_rate": 0.0001,
      "loss": 1.5486,
      "step": 3059
    },
    {
      "epoch": 0.4913294797687861,
      "grad_norm": 0.2473047375679016,
      "learning_rate": 0.0001,
      "loss": 1.5221,
      "step": 3060
    },
    {
      "epoch": 0.49149004495825305,
      "grad_norm": 0.2520563006401062,
      "learning_rate": 0.0001,
      "loss": 1.4736,
      "step": 3061
    },
    {
      "epoch": 0.49165061014772,
      "grad_norm": 0.24585667252540588,
      "learning_rate": 0.0001,
      "loss": 1.5442,
      "step": 3062
    },
    {
      "epoch": 0.4918111753371869,
      "grad_norm": 0.29285928606987,
      "learning_rate": 0.0001,
      "loss": 1.6057,
      "step": 3063
    },
    {
      "epoch": 0.4919717405266538,
      "grad_norm": 0.24709568917751312,
      "learning_rate": 0.0001,
      "loss": 1.5296,
      "step": 3064
    },
    {
      "epoch": 0.49213230571612077,
      "grad_norm": 0.2599603235721588,
      "learning_rate": 0.0001,
      "loss": 1.5214,
      "step": 3065
    },
    {
      "epoch": 0.49229287090558765,
      "grad_norm": 0.25023606419563293,
      "learning_rate": 0.0001,
      "loss": 1.5966,
      "step": 3066
    },
    {
      "epoch": 0.4924534360950546,
      "grad_norm": 0.26718711853027344,
      "learning_rate": 0.0001,
      "loss": 1.554,
      "step": 3067
    },
    {
      "epoch": 0.49261400128452154,
      "grad_norm": 0.2554135322570801,
      "learning_rate": 0.0001,
      "loss": 1.4141,
      "step": 3068
    },
    {
      "epoch": 0.49277456647398843,
      "grad_norm": 0.2615140974521637,
      "learning_rate": 0.0001,
      "loss": 1.5525,
      "step": 3069
    },
    {
      "epoch": 0.4929351316634554,
      "grad_norm": 0.25347504019737244,
      "learning_rate": 0.0001,
      "loss": 1.5742,
      "step": 3070
    },
    {
      "epoch": 0.49309569685292226,
      "grad_norm": 0.2668316662311554,
      "learning_rate": 0.0001,
      "loss": 1.5778,
      "step": 3071
    },
    {
      "epoch": 0.4932562620423892,
      "grad_norm": 0.2649231553077698,
      "learning_rate": 0.0001,
      "loss": 1.4718,
      "step": 3072
    },
    {
      "epoch": 0.49341682723185615,
      "grad_norm": 0.2498873621225357,
      "learning_rate": 0.0001,
      "loss": 1.4198,
      "step": 3073
    },
    {
      "epoch": 0.49357739242132304,
      "grad_norm": 0.2470768839120865,
      "learning_rate": 0.0001,
      "loss": 1.567,
      "step": 3074
    },
    {
      "epoch": 0.49373795761079,
      "grad_norm": 0.2564491331577301,
      "learning_rate": 0.0001,
      "loss": 1.4942,
      "step": 3075
    },
    {
      "epoch": 0.4938985228002569,
      "grad_norm": 0.2566014528274536,
      "learning_rate": 0.0001,
      "loss": 1.5559,
      "step": 3076
    },
    {
      "epoch": 0.4940590879897238,
      "grad_norm": 0.24887849390506744,
      "learning_rate": 0.0001,
      "loss": 1.4746,
      "step": 3077
    },
    {
      "epoch": 0.49421965317919075,
      "grad_norm": 0.25308987498283386,
      "learning_rate": 0.0001,
      "loss": 1.4983,
      "step": 3078
    },
    {
      "epoch": 0.4943802183686577,
      "grad_norm": 0.2783001661300659,
      "learning_rate": 0.0001,
      "loss": 1.579,
      "step": 3079
    },
    {
      "epoch": 0.4945407835581246,
      "grad_norm": 0.25055360794067383,
      "learning_rate": 0.0001,
      "loss": 1.5075,
      "step": 3080
    },
    {
      "epoch": 0.49470134874759153,
      "grad_norm": 0.27468374371528625,
      "learning_rate": 0.0001,
      "loss": 1.4941,
      "step": 3081
    },
    {
      "epoch": 0.4948619139370585,
      "grad_norm": 0.2519631087779999,
      "learning_rate": 0.0001,
      "loss": 1.5096,
      "step": 3082
    },
    {
      "epoch": 0.49502247912652536,
      "grad_norm": 0.271618127822876,
      "learning_rate": 0.0001,
      "loss": 1.4984,
      "step": 3083
    },
    {
      "epoch": 0.4951830443159923,
      "grad_norm": 0.24738885462284088,
      "learning_rate": 0.0001,
      "loss": 1.559,
      "step": 3084
    },
    {
      "epoch": 0.4953436095054592,
      "grad_norm": 0.24938714504241943,
      "learning_rate": 0.0001,
      "loss": 1.5353,
      "step": 3085
    },
    {
      "epoch": 0.49550417469492614,
      "grad_norm": 0.24343474209308624,
      "learning_rate": 0.0001,
      "loss": 1.5173,
      "step": 3086
    },
    {
      "epoch": 0.4956647398843931,
      "grad_norm": 0.2473086416721344,
      "learning_rate": 0.0001,
      "loss": 1.5159,
      "step": 3087
    },
    {
      "epoch": 0.49582530507385997,
      "grad_norm": 0.2625449597835541,
      "learning_rate": 0.0001,
      "loss": 1.5636,
      "step": 3088
    },
    {
      "epoch": 0.4959858702633269,
      "grad_norm": 0.2572121322154999,
      "learning_rate": 0.0001,
      "loss": 1.5225,
      "step": 3089
    },
    {
      "epoch": 0.49614643545279385,
      "grad_norm": 0.25577324628829956,
      "learning_rate": 0.0001,
      "loss": 1.52,
      "step": 3090
    },
    {
      "epoch": 0.49630700064226074,
      "grad_norm": 0.250839501619339,
      "learning_rate": 0.0001,
      "loss": 1.4678,
      "step": 3091
    },
    {
      "epoch": 0.4964675658317277,
      "grad_norm": 0.25855231285095215,
      "learning_rate": 0.0001,
      "loss": 1.5609,
      "step": 3092
    },
    {
      "epoch": 0.49662813102119463,
      "grad_norm": 0.26504799723625183,
      "learning_rate": 0.0001,
      "loss": 1.4408,
      "step": 3093
    },
    {
      "epoch": 0.4967886962106615,
      "grad_norm": 0.2471844106912613,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 3094
    },
    {
      "epoch": 0.49694926140012846,
      "grad_norm": 0.24830299615859985,
      "learning_rate": 0.0001,
      "loss": 1.4682,
      "step": 3095
    },
    {
      "epoch": 0.49710982658959535,
      "grad_norm": 0.25124719738960266,
      "learning_rate": 0.0001,
      "loss": 1.5363,
      "step": 3096
    },
    {
      "epoch": 0.4972703917790623,
      "grad_norm": 0.26362356543540955,
      "learning_rate": 0.0001,
      "loss": 1.553,
      "step": 3097
    },
    {
      "epoch": 0.49743095696852924,
      "grad_norm": 0.2732217609882355,
      "learning_rate": 0.0001,
      "loss": 1.4665,
      "step": 3098
    },
    {
      "epoch": 0.4975915221579961,
      "grad_norm": 0.25540000200271606,
      "learning_rate": 0.0001,
      "loss": 1.5296,
      "step": 3099
    },
    {
      "epoch": 0.49775208734746307,
      "grad_norm": 0.25769975781440735,
      "learning_rate": 0.0001,
      "loss": 1.5072,
      "step": 3100
    },
    {
      "epoch": 0.49791265253693,
      "grad_norm": 0.266641229391098,
      "learning_rate": 0.0001,
      "loss": 1.5302,
      "step": 3101
    },
    {
      "epoch": 0.4980732177263969,
      "grad_norm": 0.2833288013935089,
      "learning_rate": 0.0001,
      "loss": 1.4988,
      "step": 3102
    },
    {
      "epoch": 0.49823378291586384,
      "grad_norm": 0.252909779548645,
      "learning_rate": 0.0001,
      "loss": 1.4831,
      "step": 3103
    },
    {
      "epoch": 0.4983943481053308,
      "grad_norm": 0.2659388780593872,
      "learning_rate": 0.0001,
      "loss": 1.5238,
      "step": 3104
    },
    {
      "epoch": 0.4985549132947977,
      "grad_norm": 0.2778969407081604,
      "learning_rate": 0.0001,
      "loss": 1.4491,
      "step": 3105
    },
    {
      "epoch": 0.4987154784842646,
      "grad_norm": 0.24570304155349731,
      "learning_rate": 0.0001,
      "loss": 1.5025,
      "step": 3106
    },
    {
      "epoch": 0.49887604367373156,
      "grad_norm": 0.24787116050720215,
      "learning_rate": 0.0001,
      "loss": 1.5433,
      "step": 3107
    },
    {
      "epoch": 0.49903660886319845,
      "grad_norm": 0.28059741854667664,
      "learning_rate": 0.0001,
      "loss": 1.4792,
      "step": 3108
    },
    {
      "epoch": 0.4991971740526654,
      "grad_norm": 0.24759264290332794,
      "learning_rate": 0.0001,
      "loss": 1.5283,
      "step": 3109
    },
    {
      "epoch": 0.4993577392421323,
      "grad_norm": 0.2525692582130432,
      "learning_rate": 0.0001,
      "loss": 1.5304,
      "step": 3110
    },
    {
      "epoch": 0.4995183044315992,
      "grad_norm": 0.2865564525127411,
      "learning_rate": 0.0001,
      "loss": 1.5287,
      "step": 3111
    },
    {
      "epoch": 0.49967886962106617,
      "grad_norm": 0.2567136585712433,
      "learning_rate": 0.0001,
      "loss": 1.5395,
      "step": 3112
    },
    {
      "epoch": 0.49983943481053306,
      "grad_norm": 0.24678722023963928,
      "learning_rate": 0.0001,
      "loss": 1.481,
      "step": 3113
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2739090919494629,
      "learning_rate": 0.0001,
      "loss": 1.5642,
      "step": 3114
    },
    {
      "epoch": 0.5001605651894669,
      "grad_norm": 0.2589796781539917,
      "learning_rate": 0.0001,
      "loss": 1.4718,
      "step": 3115
    },
    {
      "epoch": 0.5003211303789339,
      "grad_norm": 0.2633964419364929,
      "learning_rate": 0.0001,
      "loss": 1.4272,
      "step": 3116
    },
    {
      "epoch": 0.5004816955684007,
      "grad_norm": 0.2695305049419403,
      "learning_rate": 0.0001,
      "loss": 1.593,
      "step": 3117
    },
    {
      "epoch": 0.5006422607578677,
      "grad_norm": 0.29884904623031616,
      "learning_rate": 0.0001,
      "loss": 1.5662,
      "step": 3118
    },
    {
      "epoch": 0.5008028259473346,
      "grad_norm": 0.26133275032043457,
      "learning_rate": 0.0001,
      "loss": 1.5149,
      "step": 3119
    },
    {
      "epoch": 0.5009633911368016,
      "grad_norm": 0.25415998697280884,
      "learning_rate": 0.0001,
      "loss": 1.5085,
      "step": 3120
    },
    {
      "epoch": 0.5011239563262685,
      "grad_norm": 0.23989804089069366,
      "learning_rate": 0.0001,
      "loss": 1.4415,
      "step": 3121
    },
    {
      "epoch": 0.5012845215157354,
      "grad_norm": 0.25352632999420166,
      "learning_rate": 0.0001,
      "loss": 1.518,
      "step": 3122
    },
    {
      "epoch": 0.5014450867052023,
      "grad_norm": 0.245565265417099,
      "learning_rate": 0.0001,
      "loss": 1.4381,
      "step": 3123
    },
    {
      "epoch": 0.5016056518946692,
      "grad_norm": 0.25735533237457275,
      "learning_rate": 0.0001,
      "loss": 1.5262,
      "step": 3124
    },
    {
      "epoch": 0.5017662170841362,
      "grad_norm": 0.2700367867946625,
      "learning_rate": 0.0001,
      "loss": 1.5335,
      "step": 3125
    },
    {
      "epoch": 0.5019267822736031,
      "grad_norm": 0.2597363293170929,
      "learning_rate": 0.0001,
      "loss": 1.5915,
      "step": 3126
    },
    {
      "epoch": 0.50208734746307,
      "grad_norm": 0.246708482503891,
      "learning_rate": 0.0001,
      "loss": 1.566,
      "step": 3127
    },
    {
      "epoch": 0.5022479126525369,
      "grad_norm": 0.2523314356803894,
      "learning_rate": 0.0001,
      "loss": 1.5981,
      "step": 3128
    },
    {
      "epoch": 0.5024084778420038,
      "grad_norm": 0.26154765486717224,
      "learning_rate": 0.0001,
      "loss": 1.5659,
      "step": 3129
    },
    {
      "epoch": 0.5025690430314708,
      "grad_norm": 0.27075454592704773,
      "learning_rate": 0.0001,
      "loss": 1.5471,
      "step": 3130
    },
    {
      "epoch": 0.5027296082209377,
      "grad_norm": 0.2411806732416153,
      "learning_rate": 0.0001,
      "loss": 1.4735,
      "step": 3131
    },
    {
      "epoch": 0.5028901734104047,
      "grad_norm": 0.2575620710849762,
      "learning_rate": 0.0001,
      "loss": 1.5578,
      "step": 3132
    },
    {
      "epoch": 0.5030507385998716,
      "grad_norm": 0.2657145857810974,
      "learning_rate": 0.0001,
      "loss": 1.5552,
      "step": 3133
    },
    {
      "epoch": 0.5032113037893384,
      "grad_norm": 0.2549566626548767,
      "learning_rate": 0.0001,
      "loss": 1.5077,
      "step": 3134
    },
    {
      "epoch": 0.5033718689788054,
      "grad_norm": 0.2419261783361435,
      "learning_rate": 0.0001,
      "loss": 1.4125,
      "step": 3135
    },
    {
      "epoch": 0.5035324341682723,
      "grad_norm": 0.2513786256313324,
      "learning_rate": 0.0001,
      "loss": 1.5263,
      "step": 3136
    },
    {
      "epoch": 0.5036929993577393,
      "grad_norm": 0.47951874136924744,
      "learning_rate": 0.0001,
      "loss": 1.5759,
      "step": 3137
    },
    {
      "epoch": 0.5038535645472062,
      "grad_norm": 0.2813974916934967,
      "learning_rate": 0.0001,
      "loss": 1.5249,
      "step": 3138
    },
    {
      "epoch": 0.5040141297366731,
      "grad_norm": 0.2584213614463806,
      "learning_rate": 0.0001,
      "loss": 1.4307,
      "step": 3139
    },
    {
      "epoch": 0.50417469492614,
      "grad_norm": 0.2642471194267273,
      "learning_rate": 0.0001,
      "loss": 1.4891,
      "step": 3140
    },
    {
      "epoch": 0.5043352601156069,
      "grad_norm": 0.2796146869659424,
      "learning_rate": 0.0001,
      "loss": 1.5579,
      "step": 3141
    },
    {
      "epoch": 0.5044958253050739,
      "grad_norm": 0.27717068791389465,
      "learning_rate": 0.0001,
      "loss": 1.5323,
      "step": 3142
    },
    {
      "epoch": 0.5046563904945408,
      "grad_norm": 0.26738211512565613,
      "learning_rate": 0.0001,
      "loss": 1.5566,
      "step": 3143
    },
    {
      "epoch": 0.5048169556840078,
      "grad_norm": 0.2563300132751465,
      "learning_rate": 0.0001,
      "loss": 1.5415,
      "step": 3144
    },
    {
      "epoch": 0.5049775208734746,
      "grad_norm": 0.261578232049942,
      "learning_rate": 0.0001,
      "loss": 1.5141,
      "step": 3145
    },
    {
      "epoch": 0.5051380860629415,
      "grad_norm": 0.24926215410232544,
      "learning_rate": 0.0001,
      "loss": 1.4775,
      "step": 3146
    },
    {
      "epoch": 0.5052986512524085,
      "grad_norm": 0.26072317361831665,
      "learning_rate": 0.0001,
      "loss": 1.5386,
      "step": 3147
    },
    {
      "epoch": 0.5054592164418754,
      "grad_norm": 0.26383572816848755,
      "learning_rate": 0.0001,
      "loss": 1.5988,
      "step": 3148
    },
    {
      "epoch": 0.5056197816313424,
      "grad_norm": 0.2610394358634949,
      "learning_rate": 0.0001,
      "loss": 1.5487,
      "step": 3149
    },
    {
      "epoch": 0.5057803468208093,
      "grad_norm": 0.25946468114852905,
      "learning_rate": 0.0001,
      "loss": 1.6254,
      "step": 3150
    },
    {
      "epoch": 0.5059409120102761,
      "grad_norm": 0.2548661530017853,
      "learning_rate": 0.0001,
      "loss": 1.5542,
      "step": 3151
    },
    {
      "epoch": 0.5061014771997431,
      "grad_norm": 0.2439573109149933,
      "learning_rate": 0.0001,
      "loss": 1.499,
      "step": 3152
    },
    {
      "epoch": 0.50626204238921,
      "grad_norm": 0.2605626583099365,
      "learning_rate": 0.0001,
      "loss": 1.4585,
      "step": 3153
    },
    {
      "epoch": 0.506422607578677,
      "grad_norm": 0.25361889600753784,
      "learning_rate": 0.0001,
      "loss": 1.5259,
      "step": 3154
    },
    {
      "epoch": 0.5065831727681439,
      "grad_norm": 0.26460692286491394,
      "learning_rate": 0.0001,
      "loss": 1.5612,
      "step": 3155
    },
    {
      "epoch": 0.5067437379576107,
      "grad_norm": 0.26998046040534973,
      "learning_rate": 0.0001,
      "loss": 1.5473,
      "step": 3156
    },
    {
      "epoch": 0.5069043031470777,
      "grad_norm": 0.2783830165863037,
      "learning_rate": 0.0001,
      "loss": 1.5556,
      "step": 3157
    },
    {
      "epoch": 0.5070648683365446,
      "grad_norm": 0.2519178092479706,
      "learning_rate": 0.0001,
      "loss": 1.5053,
      "step": 3158
    },
    {
      "epoch": 0.5072254335260116,
      "grad_norm": 0.2751830816268921,
      "learning_rate": 0.0001,
      "loss": 1.6179,
      "step": 3159
    },
    {
      "epoch": 0.5073859987154785,
      "grad_norm": 0.2519207000732422,
      "learning_rate": 0.0001,
      "loss": 1.4305,
      "step": 3160
    },
    {
      "epoch": 0.5075465639049455,
      "grad_norm": 0.2452932894229889,
      "learning_rate": 0.0001,
      "loss": 1.4406,
      "step": 3161
    },
    {
      "epoch": 0.5077071290944123,
      "grad_norm": 0.24579176306724548,
      "learning_rate": 0.0001,
      "loss": 1.4543,
      "step": 3162
    },
    {
      "epoch": 0.5078676942838792,
      "grad_norm": 0.24985000491142273,
      "learning_rate": 0.0001,
      "loss": 1.5073,
      "step": 3163
    },
    {
      "epoch": 0.5080282594733462,
      "grad_norm": 0.26153042912483215,
      "learning_rate": 0.0001,
      "loss": 1.4945,
      "step": 3164
    },
    {
      "epoch": 0.5081888246628131,
      "grad_norm": 0.2678162753582001,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 3165
    },
    {
      "epoch": 0.5083493898522801,
      "grad_norm": 0.2512088716030121,
      "learning_rate": 0.0001,
      "loss": 1.5458,
      "step": 3166
    },
    {
      "epoch": 0.5085099550417469,
      "grad_norm": 0.25522008538246155,
      "learning_rate": 0.0001,
      "loss": 1.5105,
      "step": 3167
    },
    {
      "epoch": 0.5086705202312138,
      "grad_norm": 0.26836958527565,
      "learning_rate": 0.0001,
      "loss": 1.5729,
      "step": 3168
    },
    {
      "epoch": 0.5088310854206808,
      "grad_norm": 0.2670387625694275,
      "learning_rate": 0.0001,
      "loss": 1.5368,
      "step": 3169
    },
    {
      "epoch": 0.5089916506101477,
      "grad_norm": 0.2495080679655075,
      "learning_rate": 0.0001,
      "loss": 1.5713,
      "step": 3170
    },
    {
      "epoch": 0.5091522157996147,
      "grad_norm": 0.26266956329345703,
      "learning_rate": 0.0001,
      "loss": 1.5738,
      "step": 3171
    },
    {
      "epoch": 0.5093127809890816,
      "grad_norm": 0.2562830448150635,
      "learning_rate": 0.0001,
      "loss": 1.5904,
      "step": 3172
    },
    {
      "epoch": 0.5094733461785484,
      "grad_norm": 0.24958689510822296,
      "learning_rate": 0.0001,
      "loss": 1.4577,
      "step": 3173
    },
    {
      "epoch": 0.5096339113680154,
      "grad_norm": 0.24641390144824982,
      "learning_rate": 0.0001,
      "loss": 1.5625,
      "step": 3174
    },
    {
      "epoch": 0.5097944765574823,
      "grad_norm": 0.2717821002006531,
      "learning_rate": 0.0001,
      "loss": 1.499,
      "step": 3175
    },
    {
      "epoch": 0.5099550417469493,
      "grad_norm": 0.28674250841140747,
      "learning_rate": 0.0001,
      "loss": 1.545,
      "step": 3176
    },
    {
      "epoch": 0.5101156069364162,
      "grad_norm": 0.2587761878967285,
      "learning_rate": 0.0001,
      "loss": 1.5364,
      "step": 3177
    },
    {
      "epoch": 0.510276172125883,
      "grad_norm": 0.27704596519470215,
      "learning_rate": 0.0001,
      "loss": 1.5522,
      "step": 3178
    },
    {
      "epoch": 0.51043673731535,
      "grad_norm": 0.26586246490478516,
      "learning_rate": 0.0001,
      "loss": 1.5379,
      "step": 3179
    },
    {
      "epoch": 0.5105973025048169,
      "grad_norm": 0.2559185028076172,
      "learning_rate": 0.0001,
      "loss": 1.5399,
      "step": 3180
    },
    {
      "epoch": 0.5107578676942839,
      "grad_norm": 0.26495662331581116,
      "learning_rate": 0.0001,
      "loss": 1.5068,
      "step": 3181
    },
    {
      "epoch": 0.5109184328837508,
      "grad_norm": 0.2635836899280548,
      "learning_rate": 0.0001,
      "loss": 1.5815,
      "step": 3182
    },
    {
      "epoch": 0.5110789980732178,
      "grad_norm": 0.27640485763549805,
      "learning_rate": 0.0001,
      "loss": 1.5914,
      "step": 3183
    },
    {
      "epoch": 0.5112395632626846,
      "grad_norm": 0.25381994247436523,
      "learning_rate": 0.0001,
      "loss": 1.5241,
      "step": 3184
    },
    {
      "epoch": 0.5114001284521515,
      "grad_norm": 0.2735559046268463,
      "learning_rate": 0.0001,
      "loss": 1.6167,
      "step": 3185
    },
    {
      "epoch": 0.5115606936416185,
      "grad_norm": 0.2589644193649292,
      "learning_rate": 0.0001,
      "loss": 1.5265,
      "step": 3186
    },
    {
      "epoch": 0.5117212588310854,
      "grad_norm": 0.2958916425704956,
      "learning_rate": 0.0001,
      "loss": 1.5123,
      "step": 3187
    },
    {
      "epoch": 0.5118818240205524,
      "grad_norm": 0.2542179524898529,
      "learning_rate": 0.0001,
      "loss": 1.5355,
      "step": 3188
    },
    {
      "epoch": 0.5120423892100193,
      "grad_norm": 0.26672956347465515,
      "learning_rate": 0.0001,
      "loss": 1.5225,
      "step": 3189
    },
    {
      "epoch": 0.5122029543994862,
      "grad_norm": 0.2564774453639984,
      "learning_rate": 0.0001,
      "loss": 1.4778,
      "step": 3190
    },
    {
      "epoch": 0.5123635195889531,
      "grad_norm": 0.2652626633644104,
      "learning_rate": 0.0001,
      "loss": 1.525,
      "step": 3191
    },
    {
      "epoch": 0.51252408477842,
      "grad_norm": 0.24786992371082306,
      "learning_rate": 0.0001,
      "loss": 1.4826,
      "step": 3192
    },
    {
      "epoch": 0.512684649967887,
      "grad_norm": 0.2644241154193878,
      "learning_rate": 0.0001,
      "loss": 1.47,
      "step": 3193
    },
    {
      "epoch": 0.5128452151573539,
      "grad_norm": 0.2515641152858734,
      "learning_rate": 0.0001,
      "loss": 1.4372,
      "step": 3194
    },
    {
      "epoch": 0.5130057803468208,
      "grad_norm": 0.25682637095451355,
      "learning_rate": 0.0001,
      "loss": 1.449,
      "step": 3195
    },
    {
      "epoch": 0.5131663455362877,
      "grad_norm": 0.2886198163032532,
      "learning_rate": 0.0001,
      "loss": 1.6067,
      "step": 3196
    },
    {
      "epoch": 0.5133269107257546,
      "grad_norm": 0.2510821521282196,
      "learning_rate": 0.0001,
      "loss": 1.4527,
      "step": 3197
    },
    {
      "epoch": 0.5134874759152216,
      "grad_norm": 0.25297778844833374,
      "learning_rate": 0.0001,
      "loss": 1.5021,
      "step": 3198
    },
    {
      "epoch": 0.5136480411046885,
      "grad_norm": 0.2658636271953583,
      "learning_rate": 0.0001,
      "loss": 1.5454,
      "step": 3199
    },
    {
      "epoch": 0.5138086062941555,
      "grad_norm": 0.24878738820552826,
      "learning_rate": 0.0001,
      "loss": 1.491,
      "step": 3200
    },
    {
      "epoch": 0.5139691714836223,
      "grad_norm": 0.27713683247566223,
      "learning_rate": 0.0001,
      "loss": 1.4908,
      "step": 3201
    },
    {
      "epoch": 0.5141297366730893,
      "grad_norm": 0.26166701316833496,
      "learning_rate": 0.0001,
      "loss": 1.5481,
      "step": 3202
    },
    {
      "epoch": 0.5142903018625562,
      "grad_norm": 0.2589658796787262,
      "learning_rate": 0.0001,
      "loss": 1.5556,
      "step": 3203
    },
    {
      "epoch": 0.5144508670520231,
      "grad_norm": 0.2689337432384491,
      "learning_rate": 0.0001,
      "loss": 1.5803,
      "step": 3204
    },
    {
      "epoch": 0.5146114322414901,
      "grad_norm": 0.2718493342399597,
      "learning_rate": 0.0001,
      "loss": 1.5135,
      "step": 3205
    },
    {
      "epoch": 0.5147719974309569,
      "grad_norm": 0.2672770619392395,
      "learning_rate": 0.0001,
      "loss": 1.534,
      "step": 3206
    },
    {
      "epoch": 0.5149325626204239,
      "grad_norm": 0.25379711389541626,
      "learning_rate": 0.0001,
      "loss": 1.5178,
      "step": 3207
    },
    {
      "epoch": 0.5150931278098908,
      "grad_norm": 0.24569812417030334,
      "learning_rate": 0.0001,
      "loss": 1.4503,
      "step": 3208
    },
    {
      "epoch": 0.5152536929993577,
      "grad_norm": 0.259088933467865,
      "learning_rate": 0.0001,
      "loss": 1.5196,
      "step": 3209
    },
    {
      "epoch": 0.5154142581888247,
      "grad_norm": 0.2550090551376343,
      "learning_rate": 0.0001,
      "loss": 1.4543,
      "step": 3210
    },
    {
      "epoch": 0.5155748233782916,
      "grad_norm": 0.2598012089729309,
      "learning_rate": 0.0001,
      "loss": 1.4943,
      "step": 3211
    },
    {
      "epoch": 0.5157353885677585,
      "grad_norm": 0.2577979266643524,
      "learning_rate": 0.0001,
      "loss": 1.4927,
      "step": 3212
    },
    {
      "epoch": 0.5158959537572254,
      "grad_norm": 0.25564226508140564,
      "learning_rate": 0.0001,
      "loss": 1.4854,
      "step": 3213
    },
    {
      "epoch": 0.5160565189466924,
      "grad_norm": 0.25783076882362366,
      "learning_rate": 0.0001,
      "loss": 1.535,
      "step": 3214
    },
    {
      "epoch": 0.5162170841361593,
      "grad_norm": 0.23981228470802307,
      "learning_rate": 0.0001,
      "loss": 1.4536,
      "step": 3215
    },
    {
      "epoch": 0.5163776493256262,
      "grad_norm": 0.28006672859191895,
      "learning_rate": 0.0001,
      "loss": 1.489,
      "step": 3216
    },
    {
      "epoch": 0.5165382145150931,
      "grad_norm": 0.2745996415615082,
      "learning_rate": 0.0001,
      "loss": 1.5246,
      "step": 3217
    },
    {
      "epoch": 0.51669877970456,
      "grad_norm": 0.24766501784324646,
      "learning_rate": 0.0001,
      "loss": 1.4865,
      "step": 3218
    },
    {
      "epoch": 0.516859344894027,
      "grad_norm": 0.24522282183170319,
      "learning_rate": 0.0001,
      "loss": 1.4773,
      "step": 3219
    },
    {
      "epoch": 0.5170199100834939,
      "grad_norm": 0.26986628770828247,
      "learning_rate": 0.0001,
      "loss": 1.5138,
      "step": 3220
    },
    {
      "epoch": 0.5171804752729608,
      "grad_norm": 0.2706376910209656,
      "learning_rate": 0.0001,
      "loss": 1.5568,
      "step": 3221
    },
    {
      "epoch": 0.5173410404624278,
      "grad_norm": 0.27984869480133057,
      "learning_rate": 0.0001,
      "loss": 1.5308,
      "step": 3222
    },
    {
      "epoch": 0.5175016056518946,
      "grad_norm": 0.2790790796279907,
      "learning_rate": 0.0001,
      "loss": 1.5512,
      "step": 3223
    },
    {
      "epoch": 0.5176621708413616,
      "grad_norm": 0.2629150450229645,
      "learning_rate": 0.0001,
      "loss": 1.5213,
      "step": 3224
    },
    {
      "epoch": 0.5178227360308285,
      "grad_norm": 0.2627519965171814,
      "learning_rate": 0.0001,
      "loss": 1.4908,
      "step": 3225
    },
    {
      "epoch": 0.5179833012202955,
      "grad_norm": 0.2655399441719055,
      "learning_rate": 0.0001,
      "loss": 1.5327,
      "step": 3226
    },
    {
      "epoch": 0.5181438664097624,
      "grad_norm": 0.25235995650291443,
      "learning_rate": 0.0001,
      "loss": 1.4728,
      "step": 3227
    },
    {
      "epoch": 0.5183044315992292,
      "grad_norm": 0.2675730586051941,
      "learning_rate": 0.0001,
      "loss": 1.5385,
      "step": 3228
    },
    {
      "epoch": 0.5184649967886962,
      "grad_norm": 0.2632604241371155,
      "learning_rate": 0.0001,
      "loss": 1.4902,
      "step": 3229
    },
    {
      "epoch": 0.5186255619781631,
      "grad_norm": 0.2656300663948059,
      "learning_rate": 0.0001,
      "loss": 1.5251,
      "step": 3230
    },
    {
      "epoch": 0.5187861271676301,
      "grad_norm": 0.2696046531200409,
      "learning_rate": 0.0001,
      "loss": 1.533,
      "step": 3231
    },
    {
      "epoch": 0.518946692357097,
      "grad_norm": 0.2575431168079376,
      "learning_rate": 0.0001,
      "loss": 1.4346,
      "step": 3232
    },
    {
      "epoch": 0.519107257546564,
      "grad_norm": 0.2561548352241516,
      "learning_rate": 0.0001,
      "loss": 1.5137,
      "step": 3233
    },
    {
      "epoch": 0.5192678227360308,
      "grad_norm": 0.25344690680503845,
      "learning_rate": 0.0001,
      "loss": 1.473,
      "step": 3234
    },
    {
      "epoch": 0.5194283879254977,
      "grad_norm": 0.2702086269855499,
      "learning_rate": 0.0001,
      "loss": 1.5989,
      "step": 3235
    },
    {
      "epoch": 0.5195889531149647,
      "grad_norm": 0.26273682713508606,
      "learning_rate": 0.0001,
      "loss": 1.4993,
      "step": 3236
    },
    {
      "epoch": 0.5197495183044316,
      "grad_norm": 0.2502880096435547,
      "learning_rate": 0.0001,
      "loss": 1.5172,
      "step": 3237
    },
    {
      "epoch": 0.5199100834938986,
      "grad_norm": 0.2783799171447754,
      "learning_rate": 0.0001,
      "loss": 1.5978,
      "step": 3238
    },
    {
      "epoch": 0.5200706486833655,
      "grad_norm": 0.26468271017074585,
      "learning_rate": 0.0001,
      "loss": 1.478,
      "step": 3239
    },
    {
      "epoch": 0.5202312138728323,
      "grad_norm": 0.24852678179740906,
      "learning_rate": 0.0001,
      "loss": 1.5157,
      "step": 3240
    },
    {
      "epoch": 0.5203917790622993,
      "grad_norm": 0.254946231842041,
      "learning_rate": 0.0001,
      "loss": 1.4678,
      "step": 3241
    },
    {
      "epoch": 0.5205523442517662,
      "grad_norm": 0.24661694467067719,
      "learning_rate": 0.0001,
      "loss": 1.4429,
      "step": 3242
    },
    {
      "epoch": 0.5207129094412332,
      "grad_norm": 0.2673511803150177,
      "learning_rate": 0.0001,
      "loss": 1.5049,
      "step": 3243
    },
    {
      "epoch": 0.5208734746307001,
      "grad_norm": 0.2501184940338135,
      "learning_rate": 0.0001,
      "loss": 1.5195,
      "step": 3244
    },
    {
      "epoch": 0.5210340398201669,
      "grad_norm": 0.32538872957229614,
      "learning_rate": 0.0001,
      "loss": 1.6076,
      "step": 3245
    },
    {
      "epoch": 0.5211946050096339,
      "grad_norm": 0.25452202558517456,
      "learning_rate": 0.0001,
      "loss": 1.5015,
      "step": 3246
    },
    {
      "epoch": 0.5213551701991008,
      "grad_norm": 0.24114397168159485,
      "learning_rate": 0.0001,
      "loss": 1.4725,
      "step": 3247
    },
    {
      "epoch": 0.5215157353885678,
      "grad_norm": 0.2677294611930847,
      "learning_rate": 0.0001,
      "loss": 1.4785,
      "step": 3248
    },
    {
      "epoch": 0.5216763005780347,
      "grad_norm": 0.2708767354488373,
      "learning_rate": 0.0001,
      "loss": 1.5621,
      "step": 3249
    },
    {
      "epoch": 0.5218368657675017,
      "grad_norm": 0.2655431032180786,
      "learning_rate": 0.0001,
      "loss": 1.5123,
      "step": 3250
    },
    {
      "epoch": 0.5219974309569685,
      "grad_norm": 0.25391721725463867,
      "learning_rate": 0.0001,
      "loss": 1.4559,
      "step": 3251
    },
    {
      "epoch": 0.5221579961464354,
      "grad_norm": 0.2556287944316864,
      "learning_rate": 0.0001,
      "loss": 1.4503,
      "step": 3252
    },
    {
      "epoch": 0.5223185613359024,
      "grad_norm": 0.281301349401474,
      "learning_rate": 0.0001,
      "loss": 1.5309,
      "step": 3253
    },
    {
      "epoch": 0.5224791265253693,
      "grad_norm": 0.26665544509887695,
      "learning_rate": 0.0001,
      "loss": 1.5339,
      "step": 3254
    },
    {
      "epoch": 0.5226396917148363,
      "grad_norm": 0.24834078550338745,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 3255
    },
    {
      "epoch": 0.5228002569043031,
      "grad_norm": 0.2413569688796997,
      "learning_rate": 0.0001,
      "loss": 1.3731,
      "step": 3256
    },
    {
      "epoch": 0.52296082209377,
      "grad_norm": 0.2564075291156769,
      "learning_rate": 0.0001,
      "loss": 1.4544,
      "step": 3257
    },
    {
      "epoch": 0.523121387283237,
      "grad_norm": 0.24661500751972198,
      "learning_rate": 0.0001,
      "loss": 1.4722,
      "step": 3258
    },
    {
      "epoch": 0.5232819524727039,
      "grad_norm": 0.24206063151359558,
      "learning_rate": 0.0001,
      "loss": 1.4261,
      "step": 3259
    },
    {
      "epoch": 0.5234425176621709,
      "grad_norm": 0.27224233746528625,
      "learning_rate": 0.0001,
      "loss": 1.5751,
      "step": 3260
    },
    {
      "epoch": 0.5236030828516378,
      "grad_norm": 0.2451929748058319,
      "learning_rate": 0.0001,
      "loss": 1.4378,
      "step": 3261
    },
    {
      "epoch": 0.5237636480411046,
      "grad_norm": 0.26787272095680237,
      "learning_rate": 0.0001,
      "loss": 1.5243,
      "step": 3262
    },
    {
      "epoch": 0.5239242132305716,
      "grad_norm": 0.25678402185440063,
      "learning_rate": 0.0001,
      "loss": 1.5373,
      "step": 3263
    },
    {
      "epoch": 0.5240847784200385,
      "grad_norm": 0.2668248414993286,
      "learning_rate": 0.0001,
      "loss": 1.5171,
      "step": 3264
    },
    {
      "epoch": 0.5242453436095055,
      "grad_norm": 0.2636510133743286,
      "learning_rate": 0.0001,
      "loss": 1.5734,
      "step": 3265
    },
    {
      "epoch": 0.5244059087989724,
      "grad_norm": 0.2673064172267914,
      "learning_rate": 0.0001,
      "loss": 1.5351,
      "step": 3266
    },
    {
      "epoch": 0.5245664739884393,
      "grad_norm": 0.24744990468025208,
      "learning_rate": 0.0001,
      "loss": 1.4201,
      "step": 3267
    },
    {
      "epoch": 0.5247270391779062,
      "grad_norm": 0.2554832100868225,
      "learning_rate": 0.0001,
      "loss": 1.5381,
      "step": 3268
    },
    {
      "epoch": 0.5248876043673731,
      "grad_norm": 0.25067007541656494,
      "learning_rate": 0.0001,
      "loss": 1.4835,
      "step": 3269
    },
    {
      "epoch": 0.5250481695568401,
      "grad_norm": 0.2619760036468506,
      "learning_rate": 0.0001,
      "loss": 1.63,
      "step": 3270
    },
    {
      "epoch": 0.525208734746307,
      "grad_norm": 0.2762417793273926,
      "learning_rate": 0.0001,
      "loss": 1.5615,
      "step": 3271
    },
    {
      "epoch": 0.525369299935774,
      "grad_norm": 0.2523985207080841,
      "learning_rate": 0.0001,
      "loss": 1.4261,
      "step": 3272
    },
    {
      "epoch": 0.5255298651252408,
      "grad_norm": 0.25158926844596863,
      "learning_rate": 0.0001,
      "loss": 1.4379,
      "step": 3273
    },
    {
      "epoch": 0.5256904303147077,
      "grad_norm": 0.2579212486743927,
      "learning_rate": 0.0001,
      "loss": 1.5615,
      "step": 3274
    },
    {
      "epoch": 0.5258509955041747,
      "grad_norm": 0.25403761863708496,
      "learning_rate": 0.0001,
      "loss": 1.5598,
      "step": 3275
    },
    {
      "epoch": 0.5260115606936416,
      "grad_norm": 0.2513054311275482,
      "learning_rate": 0.0001,
      "loss": 1.5315,
      "step": 3276
    },
    {
      "epoch": 0.5261721258831086,
      "grad_norm": 0.24935093522071838,
      "learning_rate": 0.0001,
      "loss": 1.5597,
      "step": 3277
    },
    {
      "epoch": 0.5263326910725755,
      "grad_norm": 0.2529204189777374,
      "learning_rate": 0.0001,
      "loss": 1.5671,
      "step": 3278
    },
    {
      "epoch": 0.5264932562620424,
      "grad_norm": 0.253094345331192,
      "learning_rate": 0.0001,
      "loss": 1.4921,
      "step": 3279
    },
    {
      "epoch": 0.5266538214515093,
      "grad_norm": 0.26417168974876404,
      "learning_rate": 0.0001,
      "loss": 1.5414,
      "step": 3280
    },
    {
      "epoch": 0.5268143866409762,
      "grad_norm": 0.24993251264095306,
      "learning_rate": 0.0001,
      "loss": 1.4082,
      "step": 3281
    },
    {
      "epoch": 0.5269749518304432,
      "grad_norm": 0.2519819438457489,
      "learning_rate": 0.0001,
      "loss": 1.5673,
      "step": 3282
    },
    {
      "epoch": 0.5271355170199101,
      "grad_norm": 0.27065905928611755,
      "learning_rate": 0.0001,
      "loss": 1.5241,
      "step": 3283
    },
    {
      "epoch": 0.527296082209377,
      "grad_norm": 0.2681165039539337,
      "learning_rate": 0.0001,
      "loss": 1.5656,
      "step": 3284
    },
    {
      "epoch": 0.5274566473988439,
      "grad_norm": 0.26704514026641846,
      "learning_rate": 0.0001,
      "loss": 1.4909,
      "step": 3285
    },
    {
      "epoch": 0.5276172125883108,
      "grad_norm": 0.2553468346595764,
      "learning_rate": 0.0001,
      "loss": 1.5596,
      "step": 3286
    },
    {
      "epoch": 0.5277777777777778,
      "grad_norm": 0.2722179591655731,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 3287
    },
    {
      "epoch": 0.5279383429672447,
      "grad_norm": 0.26517441868782043,
      "learning_rate": 0.0001,
      "loss": 1.5728,
      "step": 3288
    },
    {
      "epoch": 0.5280989081567117,
      "grad_norm": 0.3017769455909729,
      "learning_rate": 0.0001,
      "loss": 1.5826,
      "step": 3289
    },
    {
      "epoch": 0.5282594733461785,
      "grad_norm": 0.25480830669403076,
      "learning_rate": 0.0001,
      "loss": 1.5278,
      "step": 3290
    },
    {
      "epoch": 0.5284200385356455,
      "grad_norm": 0.25403088331222534,
      "learning_rate": 0.0001,
      "loss": 1.5433,
      "step": 3291
    },
    {
      "epoch": 0.5285806037251124,
      "grad_norm": 0.2752486765384674,
      "learning_rate": 0.0001,
      "loss": 1.533,
      "step": 3292
    },
    {
      "epoch": 0.5287411689145793,
      "grad_norm": 0.2497430443763733,
      "learning_rate": 0.0001,
      "loss": 1.4959,
      "step": 3293
    },
    {
      "epoch": 0.5289017341040463,
      "grad_norm": 0.2565641701221466,
      "learning_rate": 0.0001,
      "loss": 1.5025,
      "step": 3294
    },
    {
      "epoch": 0.5290622992935131,
      "grad_norm": 0.2425820529460907,
      "learning_rate": 0.0001,
      "loss": 1.5128,
      "step": 3295
    },
    {
      "epoch": 0.5292228644829801,
      "grad_norm": 0.23783937096595764,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 3296
    },
    {
      "epoch": 0.529383429672447,
      "grad_norm": 0.2435399293899536,
      "learning_rate": 0.0001,
      "loss": 1.4828,
      "step": 3297
    },
    {
      "epoch": 0.529543994861914,
      "grad_norm": 0.2555021047592163,
      "learning_rate": 0.0001,
      "loss": 1.4554,
      "step": 3298
    },
    {
      "epoch": 0.5297045600513809,
      "grad_norm": 0.2563108801841736,
      "learning_rate": 0.0001,
      "loss": 1.5174,
      "step": 3299
    },
    {
      "epoch": 0.5298651252408478,
      "grad_norm": 0.24869240820407867,
      "learning_rate": 0.0001,
      "loss": 1.5288,
      "step": 3300
    },
    {
      "epoch": 0.5300256904303147,
      "grad_norm": 0.2502351999282837,
      "learning_rate": 0.0001,
      "loss": 1.5042,
      "step": 3301
    },
    {
      "epoch": 0.5301862556197816,
      "grad_norm": 0.2555674910545349,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 3302
    },
    {
      "epoch": 0.5303468208092486,
      "grad_norm": 0.2545374929904938,
      "learning_rate": 0.0001,
      "loss": 1.5459,
      "step": 3303
    },
    {
      "epoch": 0.5305073859987155,
      "grad_norm": 0.2690221667289734,
      "learning_rate": 0.0001,
      "loss": 1.5089,
      "step": 3304
    },
    {
      "epoch": 0.5306679511881824,
      "grad_norm": 0.24777981638908386,
      "learning_rate": 0.0001,
      "loss": 1.487,
      "step": 3305
    },
    {
      "epoch": 0.5308285163776493,
      "grad_norm": 0.2689094841480255,
      "learning_rate": 0.0001,
      "loss": 1.5838,
      "step": 3306
    },
    {
      "epoch": 0.5309890815671162,
      "grad_norm": 0.41664373874664307,
      "learning_rate": 0.0001,
      "loss": 1.4959,
      "step": 3307
    },
    {
      "epoch": 0.5311496467565832,
      "grad_norm": 0.25525885820388794,
      "learning_rate": 0.0001,
      "loss": 1.5469,
      "step": 3308
    },
    {
      "epoch": 0.5313102119460501,
      "grad_norm": 0.2596135139465332,
      "learning_rate": 0.0001,
      "loss": 1.506,
      "step": 3309
    },
    {
      "epoch": 0.531470777135517,
      "grad_norm": 0.25523683428764343,
      "learning_rate": 0.0001,
      "loss": 1.525,
      "step": 3310
    },
    {
      "epoch": 0.531631342324984,
      "grad_norm": 0.23948585987091064,
      "learning_rate": 0.0001,
      "loss": 1.3896,
      "step": 3311
    },
    {
      "epoch": 0.5317919075144508,
      "grad_norm": 0.25522205233573914,
      "learning_rate": 0.0001,
      "loss": 1.4894,
      "step": 3312
    },
    {
      "epoch": 0.5319524727039178,
      "grad_norm": 0.2610253691673279,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 3313
    },
    {
      "epoch": 0.5321130378933847,
      "grad_norm": 0.2702656090259552,
      "learning_rate": 0.0001,
      "loss": 1.5766,
      "step": 3314
    },
    {
      "epoch": 0.5322736030828517,
      "grad_norm": 0.25961098074913025,
      "learning_rate": 0.0001,
      "loss": 1.5311,
      "step": 3315
    },
    {
      "epoch": 0.5324341682723186,
      "grad_norm": 0.2641583979129791,
      "learning_rate": 0.0001,
      "loss": 1.4961,
      "step": 3316
    },
    {
      "epoch": 0.5325947334617854,
      "grad_norm": 0.25887444615364075,
      "learning_rate": 0.0001,
      "loss": 1.4627,
      "step": 3317
    },
    {
      "epoch": 0.5327552986512524,
      "grad_norm": 0.2602037489414215,
      "learning_rate": 0.0001,
      "loss": 1.5018,
      "step": 3318
    },
    {
      "epoch": 0.5329158638407193,
      "grad_norm": 0.2509191930294037,
      "learning_rate": 0.0001,
      "loss": 1.5104,
      "step": 3319
    },
    {
      "epoch": 0.5330764290301863,
      "grad_norm": 0.25888267159461975,
      "learning_rate": 0.0001,
      "loss": 1.5678,
      "step": 3320
    },
    {
      "epoch": 0.5332369942196532,
      "grad_norm": 0.27837073802948,
      "learning_rate": 0.0001,
      "loss": 1.4706,
      "step": 3321
    },
    {
      "epoch": 0.5333975594091201,
      "grad_norm": 0.28042373061180115,
      "learning_rate": 0.0001,
      "loss": 1.589,
      "step": 3322
    },
    {
      "epoch": 0.533558124598587,
      "grad_norm": 0.2434910088777542,
      "learning_rate": 0.0001,
      "loss": 1.5362,
      "step": 3323
    },
    {
      "epoch": 0.5337186897880539,
      "grad_norm": 0.2693328857421875,
      "learning_rate": 0.0001,
      "loss": 1.4727,
      "step": 3324
    },
    {
      "epoch": 0.5338792549775209,
      "grad_norm": 0.25696882605552673,
      "learning_rate": 0.0001,
      "loss": 1.487,
      "step": 3325
    },
    {
      "epoch": 0.5340398201669878,
      "grad_norm": 0.25650927424430847,
      "learning_rate": 0.0001,
      "loss": 1.4711,
      "step": 3326
    },
    {
      "epoch": 0.5342003853564548,
      "grad_norm": 0.2530691623687744,
      "learning_rate": 0.0001,
      "loss": 1.5132,
      "step": 3327
    },
    {
      "epoch": 0.5343609505459217,
      "grad_norm": 0.25183364748954773,
      "learning_rate": 0.0001,
      "loss": 1.5159,
      "step": 3328
    },
    {
      "epoch": 0.5345215157353885,
      "grad_norm": 0.24335777759552002,
      "learning_rate": 0.0001,
      "loss": 1.4891,
      "step": 3329
    },
    {
      "epoch": 0.5346820809248555,
      "grad_norm": 0.25641438364982605,
      "learning_rate": 0.0001,
      "loss": 1.4974,
      "step": 3330
    },
    {
      "epoch": 0.5348426461143224,
      "grad_norm": 0.25543588399887085,
      "learning_rate": 0.0001,
      "loss": 1.614,
      "step": 3331
    },
    {
      "epoch": 0.5350032113037894,
      "grad_norm": 0.23903098702430725,
      "learning_rate": 0.0001,
      "loss": 1.4576,
      "step": 3332
    },
    {
      "epoch": 0.5351637764932563,
      "grad_norm": 0.25110888481140137,
      "learning_rate": 0.0001,
      "loss": 1.5597,
      "step": 3333
    },
    {
      "epoch": 0.5353243416827231,
      "grad_norm": 0.24475426971912384,
      "learning_rate": 0.0001,
      "loss": 1.4571,
      "step": 3334
    },
    {
      "epoch": 0.5354849068721901,
      "grad_norm": 0.2556982934474945,
      "learning_rate": 0.0001,
      "loss": 1.5103,
      "step": 3335
    },
    {
      "epoch": 0.535645472061657,
      "grad_norm": 0.2632392942905426,
      "learning_rate": 0.0001,
      "loss": 1.5196,
      "step": 3336
    },
    {
      "epoch": 0.535806037251124,
      "grad_norm": 0.27056071162223816,
      "learning_rate": 0.0001,
      "loss": 1.5133,
      "step": 3337
    },
    {
      "epoch": 0.5359666024405909,
      "grad_norm": 0.2630089223384857,
      "learning_rate": 0.0001,
      "loss": 1.5428,
      "step": 3338
    },
    {
      "epoch": 0.5361271676300579,
      "grad_norm": 0.2578960061073303,
      "learning_rate": 0.0001,
      "loss": 1.4556,
      "step": 3339
    },
    {
      "epoch": 0.5362877328195247,
      "grad_norm": 0.2513989806175232,
      "learning_rate": 0.0001,
      "loss": 1.4638,
      "step": 3340
    },
    {
      "epoch": 0.5364482980089916,
      "grad_norm": 0.26084649562835693,
      "learning_rate": 0.0001,
      "loss": 1.5662,
      "step": 3341
    },
    {
      "epoch": 0.5366088631984586,
      "grad_norm": 0.25264763832092285,
      "learning_rate": 0.0001,
      "loss": 1.4682,
      "step": 3342
    },
    {
      "epoch": 0.5367694283879255,
      "grad_norm": 0.2598627209663391,
      "learning_rate": 0.0001,
      "loss": 1.4693,
      "step": 3343
    },
    {
      "epoch": 0.5369299935773925,
      "grad_norm": 0.24570077657699585,
      "learning_rate": 0.0001,
      "loss": 1.5695,
      "step": 3344
    },
    {
      "epoch": 0.5370905587668593,
      "grad_norm": 0.2574644982814789,
      "learning_rate": 0.0001,
      "loss": 1.5303,
      "step": 3345
    },
    {
      "epoch": 0.5372511239563262,
      "grad_norm": 0.27043068408966064,
      "learning_rate": 0.0001,
      "loss": 1.5588,
      "step": 3346
    },
    {
      "epoch": 0.5374116891457932,
      "grad_norm": 0.2574154734611511,
      "learning_rate": 0.0001,
      "loss": 1.5076,
      "step": 3347
    },
    {
      "epoch": 0.5375722543352601,
      "grad_norm": 0.24641282856464386,
      "learning_rate": 0.0001,
      "loss": 1.5372,
      "step": 3348
    },
    {
      "epoch": 0.5377328195247271,
      "grad_norm": 0.2506469488143921,
      "learning_rate": 0.0001,
      "loss": 1.3844,
      "step": 3349
    },
    {
      "epoch": 0.537893384714194,
      "grad_norm": 0.26251789927482605,
      "learning_rate": 0.0001,
      "loss": 1.451,
      "step": 3350
    },
    {
      "epoch": 0.5380539499036608,
      "grad_norm": 0.2515432834625244,
      "learning_rate": 0.0001,
      "loss": 1.405,
      "step": 3351
    },
    {
      "epoch": 0.5382145150931278,
      "grad_norm": 0.2543710768222809,
      "learning_rate": 0.0001,
      "loss": 1.4897,
      "step": 3352
    },
    {
      "epoch": 0.5383750802825947,
      "grad_norm": 0.2588856816291809,
      "learning_rate": 0.0001,
      "loss": 1.5267,
      "step": 3353
    },
    {
      "epoch": 0.5385356454720617,
      "grad_norm": 0.2631353735923767,
      "learning_rate": 0.0001,
      "loss": 1.4741,
      "step": 3354
    },
    {
      "epoch": 0.5386962106615286,
      "grad_norm": 0.25693097710609436,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 3355
    },
    {
      "epoch": 0.5388567758509955,
      "grad_norm": 0.25500455498695374,
      "learning_rate": 0.0001,
      "loss": 1.4464,
      "step": 3356
    },
    {
      "epoch": 0.5390173410404624,
      "grad_norm": 0.26337122917175293,
      "learning_rate": 0.0001,
      "loss": 1.5926,
      "step": 3357
    },
    {
      "epoch": 0.5391779062299293,
      "grad_norm": 0.2575652599334717,
      "learning_rate": 0.0001,
      "loss": 1.5151,
      "step": 3358
    },
    {
      "epoch": 0.5393384714193963,
      "grad_norm": 0.25726866722106934,
      "learning_rate": 0.0001,
      "loss": 1.456,
      "step": 3359
    },
    {
      "epoch": 0.5394990366088632,
      "grad_norm": 0.25708848237991333,
      "learning_rate": 0.0001,
      "loss": 1.4669,
      "step": 3360
    },
    {
      "epoch": 0.5396596017983302,
      "grad_norm": 0.2644226551055908,
      "learning_rate": 0.0001,
      "loss": 1.5217,
      "step": 3361
    },
    {
      "epoch": 0.539820166987797,
      "grad_norm": 0.26429903507232666,
      "learning_rate": 0.0001,
      "loss": 1.5444,
      "step": 3362
    },
    {
      "epoch": 0.539980732177264,
      "grad_norm": 0.25670334696769714,
      "learning_rate": 0.0001,
      "loss": 1.5248,
      "step": 3363
    },
    {
      "epoch": 0.5401412973667309,
      "grad_norm": 0.2515736520290375,
      "learning_rate": 0.0001,
      "loss": 1.5046,
      "step": 3364
    },
    {
      "epoch": 0.5403018625561978,
      "grad_norm": 0.26216983795166016,
      "learning_rate": 0.0001,
      "loss": 1.6001,
      "step": 3365
    },
    {
      "epoch": 0.5404624277456648,
      "grad_norm": 0.25578591227531433,
      "learning_rate": 0.0001,
      "loss": 1.5838,
      "step": 3366
    },
    {
      "epoch": 0.5406229929351317,
      "grad_norm": 0.25333574414253235,
      "learning_rate": 0.0001,
      "loss": 1.6201,
      "step": 3367
    },
    {
      "epoch": 0.5407835581245986,
      "grad_norm": 0.25093314051628113,
      "learning_rate": 0.0001,
      "loss": 1.5405,
      "step": 3368
    },
    {
      "epoch": 0.5409441233140655,
      "grad_norm": 0.26136481761932373,
      "learning_rate": 0.0001,
      "loss": 1.5464,
      "step": 3369
    },
    {
      "epoch": 0.5411046885035324,
      "grad_norm": 0.2559725344181061,
      "learning_rate": 0.0001,
      "loss": 1.4596,
      "step": 3370
    },
    {
      "epoch": 0.5412652536929994,
      "grad_norm": 0.24507717788219452,
      "learning_rate": 0.0001,
      "loss": 1.5425,
      "step": 3371
    },
    {
      "epoch": 0.5414258188824663,
      "grad_norm": 0.2500024139881134,
      "learning_rate": 0.0001,
      "loss": 1.5271,
      "step": 3372
    },
    {
      "epoch": 0.5415863840719332,
      "grad_norm": 0.26483073830604553,
      "learning_rate": 0.0001,
      "loss": 1.5551,
      "step": 3373
    },
    {
      "epoch": 0.5417469492614001,
      "grad_norm": 0.2631481885910034,
      "learning_rate": 0.0001,
      "loss": 1.5341,
      "step": 3374
    },
    {
      "epoch": 0.541907514450867,
      "grad_norm": 0.25673601031303406,
      "learning_rate": 0.0001,
      "loss": 1.5737,
      "step": 3375
    },
    {
      "epoch": 0.542068079640334,
      "grad_norm": 0.2587435841560364,
      "learning_rate": 0.0001,
      "loss": 1.5832,
      "step": 3376
    },
    {
      "epoch": 0.5422286448298009,
      "grad_norm": 0.28487128019332886,
      "learning_rate": 0.0001,
      "loss": 1.5452,
      "step": 3377
    },
    {
      "epoch": 0.5423892100192679,
      "grad_norm": 0.24453434348106384,
      "learning_rate": 0.0001,
      "loss": 1.4557,
      "step": 3378
    },
    {
      "epoch": 0.5425497752087347,
      "grad_norm": 0.2701871395111084,
      "learning_rate": 0.0001,
      "loss": 1.54,
      "step": 3379
    },
    {
      "epoch": 0.5427103403982017,
      "grad_norm": 0.25957581400871277,
      "learning_rate": 0.0001,
      "loss": 1.5388,
      "step": 3380
    },
    {
      "epoch": 0.5428709055876686,
      "grad_norm": 0.27372631430625916,
      "learning_rate": 0.0001,
      "loss": 1.5081,
      "step": 3381
    },
    {
      "epoch": 0.5430314707771355,
      "grad_norm": 0.2520945072174072,
      "learning_rate": 0.0001,
      "loss": 1.5461,
      "step": 3382
    },
    {
      "epoch": 0.5431920359666025,
      "grad_norm": 0.25508081912994385,
      "learning_rate": 0.0001,
      "loss": 1.4849,
      "step": 3383
    },
    {
      "epoch": 0.5433526011560693,
      "grad_norm": 0.25025564432144165,
      "learning_rate": 0.0001,
      "loss": 1.5892,
      "step": 3384
    },
    {
      "epoch": 0.5435131663455363,
      "grad_norm": 0.26551491022109985,
      "learning_rate": 0.0001,
      "loss": 1.5201,
      "step": 3385
    },
    {
      "epoch": 0.5436737315350032,
      "grad_norm": 0.2558436393737793,
      "learning_rate": 0.0001,
      "loss": 1.4543,
      "step": 3386
    },
    {
      "epoch": 0.5438342967244701,
      "grad_norm": 0.25744447112083435,
      "learning_rate": 0.0001,
      "loss": 1.4921,
      "step": 3387
    },
    {
      "epoch": 0.5439948619139371,
      "grad_norm": 0.2569776177406311,
      "learning_rate": 0.0001,
      "loss": 1.4524,
      "step": 3388
    },
    {
      "epoch": 0.544155427103404,
      "grad_norm": 0.26785245537757874,
      "learning_rate": 0.0001,
      "loss": 1.5942,
      "step": 3389
    },
    {
      "epoch": 0.5443159922928709,
      "grad_norm": 0.26799994707107544,
      "learning_rate": 0.0001,
      "loss": 1.6123,
      "step": 3390
    },
    {
      "epoch": 0.5444765574823378,
      "grad_norm": 0.27217844128608704,
      "learning_rate": 0.0001,
      "loss": 1.5506,
      "step": 3391
    },
    {
      "epoch": 0.5446371226718048,
      "grad_norm": 0.24782425165176392,
      "learning_rate": 0.0001,
      "loss": 1.4203,
      "step": 3392
    },
    {
      "epoch": 0.5447976878612717,
      "grad_norm": 0.25451624393463135,
      "learning_rate": 0.0001,
      "loss": 1.5459,
      "step": 3393
    },
    {
      "epoch": 0.5449582530507386,
      "grad_norm": 0.3037620186805725,
      "learning_rate": 0.0001,
      "loss": 1.5618,
      "step": 3394
    },
    {
      "epoch": 0.5451188182402055,
      "grad_norm": 0.263279527425766,
      "learning_rate": 0.0001,
      "loss": 1.531,
      "step": 3395
    },
    {
      "epoch": 0.5452793834296724,
      "grad_norm": 0.27302125096321106,
      "learning_rate": 0.0001,
      "loss": 1.5634,
      "step": 3396
    },
    {
      "epoch": 0.5454399486191394,
      "grad_norm": 0.25892695784568787,
      "learning_rate": 0.0001,
      "loss": 1.5196,
      "step": 3397
    },
    {
      "epoch": 0.5456005138086063,
      "grad_norm": 0.25457724928855896,
      "learning_rate": 0.0001,
      "loss": 1.4876,
      "step": 3398
    },
    {
      "epoch": 0.5457610789980732,
      "grad_norm": 0.26115864515304565,
      "learning_rate": 0.0001,
      "loss": 1.5297,
      "step": 3399
    },
    {
      "epoch": 0.5459216441875402,
      "grad_norm": 0.24479591846466064,
      "learning_rate": 0.0001,
      "loss": 1.4309,
      "step": 3400
    },
    {
      "epoch": 0.546082209377007,
      "grad_norm": 0.2485695779323578,
      "learning_rate": 0.0001,
      "loss": 1.5039,
      "step": 3401
    },
    {
      "epoch": 0.546242774566474,
      "grad_norm": 0.25134915113449097,
      "learning_rate": 0.0001,
      "loss": 1.5014,
      "step": 3402
    },
    {
      "epoch": 0.5464033397559409,
      "grad_norm": 0.24269607663154602,
      "learning_rate": 0.0001,
      "loss": 1.5118,
      "step": 3403
    },
    {
      "epoch": 0.5465639049454079,
      "grad_norm": 0.2524523138999939,
      "learning_rate": 0.0001,
      "loss": 1.5586,
      "step": 3404
    },
    {
      "epoch": 0.5467244701348748,
      "grad_norm": 0.256293386220932,
      "learning_rate": 0.0001,
      "loss": 1.5002,
      "step": 3405
    },
    {
      "epoch": 0.5468850353243416,
      "grad_norm": 0.2573462128639221,
      "learning_rate": 0.0001,
      "loss": 1.5701,
      "step": 3406
    },
    {
      "epoch": 0.5470456005138086,
      "grad_norm": 0.2703557014465332,
      "learning_rate": 0.0001,
      "loss": 1.5144,
      "step": 3407
    },
    {
      "epoch": 0.5472061657032755,
      "grad_norm": 0.2504549026489258,
      "learning_rate": 0.0001,
      "loss": 1.4945,
      "step": 3408
    },
    {
      "epoch": 0.5473667308927425,
      "grad_norm": 0.25042614340782166,
      "learning_rate": 0.0001,
      "loss": 1.5555,
      "step": 3409
    },
    {
      "epoch": 0.5475272960822094,
      "grad_norm": 0.282416969537735,
      "learning_rate": 0.0001,
      "loss": 1.5334,
      "step": 3410
    },
    {
      "epoch": 0.5476878612716763,
      "grad_norm": 0.27234387397766113,
      "learning_rate": 0.0001,
      "loss": 1.4861,
      "step": 3411
    },
    {
      "epoch": 0.5478484264611432,
      "grad_norm": 0.25599533319473267,
      "learning_rate": 0.0001,
      "loss": 1.571,
      "step": 3412
    },
    {
      "epoch": 0.5480089916506101,
      "grad_norm": 0.26688531041145325,
      "learning_rate": 0.0001,
      "loss": 1.453,
      "step": 3413
    },
    {
      "epoch": 0.5481695568400771,
      "grad_norm": 0.2721960246562958,
      "learning_rate": 0.0001,
      "loss": 1.5289,
      "step": 3414
    },
    {
      "epoch": 0.548330122029544,
      "grad_norm": 0.24172380566596985,
      "learning_rate": 0.0001,
      "loss": 1.4282,
      "step": 3415
    },
    {
      "epoch": 0.548490687219011,
      "grad_norm": 0.25997138023376465,
      "learning_rate": 0.0001,
      "loss": 1.5497,
      "step": 3416
    },
    {
      "epoch": 0.5486512524084779,
      "grad_norm": 0.2492331862449646,
      "learning_rate": 0.0001,
      "loss": 1.5111,
      "step": 3417
    },
    {
      "epoch": 0.5488118175979447,
      "grad_norm": 0.26237958669662476,
      "learning_rate": 0.0001,
      "loss": 1.5827,
      "step": 3418
    },
    {
      "epoch": 0.5489723827874117,
      "grad_norm": 0.25716063380241394,
      "learning_rate": 0.0001,
      "loss": 1.4674,
      "step": 3419
    },
    {
      "epoch": 0.5491329479768786,
      "grad_norm": 0.25188320875167847,
      "learning_rate": 0.0001,
      "loss": 1.522,
      "step": 3420
    },
    {
      "epoch": 0.5492935131663456,
      "grad_norm": 0.2570692002773285,
      "learning_rate": 0.0001,
      "loss": 1.5166,
      "step": 3421
    },
    {
      "epoch": 0.5494540783558125,
      "grad_norm": 0.2607956826686859,
      "learning_rate": 0.0001,
      "loss": 1.5785,
      "step": 3422
    },
    {
      "epoch": 0.5496146435452793,
      "grad_norm": 0.2597101628780365,
      "learning_rate": 0.0001,
      "loss": 1.5615,
      "step": 3423
    },
    {
      "epoch": 0.5497752087347463,
      "grad_norm": 0.26716485619544983,
      "learning_rate": 0.0001,
      "loss": 1.5188,
      "step": 3424
    },
    {
      "epoch": 0.5499357739242132,
      "grad_norm": 0.25290799140930176,
      "learning_rate": 0.0001,
      "loss": 1.5207,
      "step": 3425
    },
    {
      "epoch": 0.5500963391136802,
      "grad_norm": 0.24632811546325684,
      "learning_rate": 0.0001,
      "loss": 1.3944,
      "step": 3426
    },
    {
      "epoch": 0.5502569043031471,
      "grad_norm": 0.27936041355133057,
      "learning_rate": 0.0001,
      "loss": 1.4912,
      "step": 3427
    },
    {
      "epoch": 0.550417469492614,
      "grad_norm": 0.2731547951698303,
      "learning_rate": 0.0001,
      "loss": 1.5761,
      "step": 3428
    },
    {
      "epoch": 0.5505780346820809,
      "grad_norm": 0.26861652731895447,
      "learning_rate": 0.0001,
      "loss": 1.4822,
      "step": 3429
    },
    {
      "epoch": 0.5507385998715478,
      "grad_norm": 0.2558331787586212,
      "learning_rate": 0.0001,
      "loss": 1.5092,
      "step": 3430
    },
    {
      "epoch": 0.5508991650610148,
      "grad_norm": 0.2508247196674347,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 3431
    },
    {
      "epoch": 0.5510597302504817,
      "grad_norm": 0.24907223880290985,
      "learning_rate": 0.0001,
      "loss": 1.4887,
      "step": 3432
    },
    {
      "epoch": 0.5512202954399487,
      "grad_norm": 0.2866308391094208,
      "learning_rate": 0.0001,
      "loss": 1.5711,
      "step": 3433
    },
    {
      "epoch": 0.5513808606294155,
      "grad_norm": 0.27295202016830444,
      "learning_rate": 0.0001,
      "loss": 1.5576,
      "step": 3434
    },
    {
      "epoch": 0.5515414258188824,
      "grad_norm": 0.26600003242492676,
      "learning_rate": 0.0001,
      "loss": 1.5331,
      "step": 3435
    },
    {
      "epoch": 0.5517019910083494,
      "grad_norm": 0.26780498027801514,
      "learning_rate": 0.0001,
      "loss": 1.527,
      "step": 3436
    },
    {
      "epoch": 0.5518625561978163,
      "grad_norm": 0.26983973383903503,
      "learning_rate": 0.0001,
      "loss": 1.5242,
      "step": 3437
    },
    {
      "epoch": 0.5520231213872833,
      "grad_norm": 0.2618176341056824,
      "learning_rate": 0.0001,
      "loss": 1.6295,
      "step": 3438
    },
    {
      "epoch": 0.5521836865767502,
      "grad_norm": 0.24899901449680328,
      "learning_rate": 0.0001,
      "loss": 1.5081,
      "step": 3439
    },
    {
      "epoch": 0.552344251766217,
      "grad_norm": 0.2552049160003662,
      "learning_rate": 0.0001,
      "loss": 1.516,
      "step": 3440
    },
    {
      "epoch": 0.552504816955684,
      "grad_norm": 0.25595688819885254,
      "learning_rate": 0.0001,
      "loss": 1.5933,
      "step": 3441
    },
    {
      "epoch": 0.5526653821451509,
      "grad_norm": 0.2632392346858978,
      "learning_rate": 0.0001,
      "loss": 1.5124,
      "step": 3442
    },
    {
      "epoch": 0.5528259473346179,
      "grad_norm": 0.26812928915023804,
      "learning_rate": 0.0001,
      "loss": 1.5507,
      "step": 3443
    },
    {
      "epoch": 0.5529865125240848,
      "grad_norm": 3.5970778465270996,
      "learning_rate": 0.0001,
      "loss": 1.5527,
      "step": 3444
    },
    {
      "epoch": 0.5531470777135516,
      "grad_norm": 0.249466210603714,
      "learning_rate": 0.0001,
      "loss": 1.4624,
      "step": 3445
    },
    {
      "epoch": 0.5533076429030186,
      "grad_norm": 0.2638276517391205,
      "learning_rate": 0.0001,
      "loss": 1.4815,
      "step": 3446
    },
    {
      "epoch": 0.5534682080924855,
      "grad_norm": 0.23985153436660767,
      "learning_rate": 0.0001,
      "loss": 1.4399,
      "step": 3447
    },
    {
      "epoch": 0.5536287732819525,
      "grad_norm": 0.25133323669433594,
      "learning_rate": 0.0001,
      "loss": 1.473,
      "step": 3448
    },
    {
      "epoch": 0.5537893384714194,
      "grad_norm": 0.25573378801345825,
      "learning_rate": 0.0001,
      "loss": 1.4564,
      "step": 3449
    },
    {
      "epoch": 0.5539499036608864,
      "grad_norm": 0.2694820463657379,
      "learning_rate": 0.0001,
      "loss": 1.5474,
      "step": 3450
    },
    {
      "epoch": 0.5541104688503532,
      "grad_norm": 0.2726207375526428,
      "learning_rate": 0.0001,
      "loss": 1.6114,
      "step": 3451
    },
    {
      "epoch": 0.5542710340398201,
      "grad_norm": 0.2577306032180786,
      "learning_rate": 0.0001,
      "loss": 1.4994,
      "step": 3452
    },
    {
      "epoch": 0.5544315992292871,
      "grad_norm": 0.2638789117336273,
      "learning_rate": 0.0001,
      "loss": 1.5959,
      "step": 3453
    },
    {
      "epoch": 0.554592164418754,
      "grad_norm": 0.2736872434616089,
      "learning_rate": 0.0001,
      "loss": 1.5324,
      "step": 3454
    },
    {
      "epoch": 0.554752729608221,
      "grad_norm": 0.24926017224788666,
      "learning_rate": 0.0001,
      "loss": 1.4686,
      "step": 3455
    },
    {
      "epoch": 0.5549132947976878,
      "grad_norm": 0.23989923298358917,
      "learning_rate": 0.0001,
      "loss": 1.5231,
      "step": 3456
    },
    {
      "epoch": 0.5550738599871547,
      "grad_norm": 0.26691344380378723,
      "learning_rate": 0.0001,
      "loss": 1.4484,
      "step": 3457
    },
    {
      "epoch": 0.5552344251766217,
      "grad_norm": 0.251097708940506,
      "learning_rate": 0.0001,
      "loss": 1.4741,
      "step": 3458
    },
    {
      "epoch": 0.5553949903660886,
      "grad_norm": 0.25313010811805725,
      "learning_rate": 0.0001,
      "loss": 1.5321,
      "step": 3459
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.2603638172149658,
      "learning_rate": 0.0001,
      "loss": 1.565,
      "step": 3460
    },
    {
      "epoch": 0.5557161207450225,
      "grad_norm": 0.2619098126888275,
      "learning_rate": 0.0001,
      "loss": 1.5085,
      "step": 3461
    },
    {
      "epoch": 0.5558766859344894,
      "grad_norm": 0.26237279176712036,
      "learning_rate": 0.0001,
      "loss": 1.5434,
      "step": 3462
    },
    {
      "epoch": 0.5560372511239563,
      "grad_norm": 0.2590477764606476,
      "learning_rate": 0.0001,
      "loss": 1.5638,
      "step": 3463
    },
    {
      "epoch": 0.5561978163134232,
      "grad_norm": 0.2514772117137909,
      "learning_rate": 0.0001,
      "loss": 1.4436,
      "step": 3464
    },
    {
      "epoch": 0.5563583815028902,
      "grad_norm": 0.27150002121925354,
      "learning_rate": 0.0001,
      "loss": 1.5903,
      "step": 3465
    },
    {
      "epoch": 0.5565189466923571,
      "grad_norm": 0.2557409405708313,
      "learning_rate": 0.0001,
      "loss": 1.5477,
      "step": 3466
    },
    {
      "epoch": 0.5566795118818241,
      "grad_norm": 0.2467564195394516,
      "learning_rate": 0.0001,
      "loss": 1.452,
      "step": 3467
    },
    {
      "epoch": 0.5568400770712909,
      "grad_norm": 0.24928201735019684,
      "learning_rate": 0.0001,
      "loss": 1.492,
      "step": 3468
    },
    {
      "epoch": 0.5570006422607579,
      "grad_norm": 0.26520484685897827,
      "learning_rate": 0.0001,
      "loss": 1.5598,
      "step": 3469
    },
    {
      "epoch": 0.5571612074502248,
      "grad_norm": 0.2560626268386841,
      "learning_rate": 0.0001,
      "loss": 1.5434,
      "step": 3470
    },
    {
      "epoch": 0.5573217726396917,
      "grad_norm": 0.25910642743110657,
      "learning_rate": 0.0001,
      "loss": 1.6065,
      "step": 3471
    },
    {
      "epoch": 0.5574823378291587,
      "grad_norm": 0.2590223252773285,
      "learning_rate": 0.0001,
      "loss": 1.5274,
      "step": 3472
    },
    {
      "epoch": 0.5576429030186255,
      "grad_norm": 0.26514342427253723,
      "learning_rate": 0.0001,
      "loss": 1.5426,
      "step": 3473
    },
    {
      "epoch": 0.5578034682080925,
      "grad_norm": 0.253199964761734,
      "learning_rate": 0.0001,
      "loss": 1.4766,
      "step": 3474
    },
    {
      "epoch": 0.5579640333975594,
      "grad_norm": 0.24293287098407745,
      "learning_rate": 0.0001,
      "loss": 1.4573,
      "step": 3475
    },
    {
      "epoch": 0.5581245985870263,
      "grad_norm": 0.262442946434021,
      "learning_rate": 0.0001,
      "loss": 1.5558,
      "step": 3476
    },
    {
      "epoch": 0.5582851637764933,
      "grad_norm": 0.2672119736671448,
      "learning_rate": 0.0001,
      "loss": 1.5199,
      "step": 3477
    },
    {
      "epoch": 0.5584457289659602,
      "grad_norm": 0.2569417953491211,
      "learning_rate": 0.0001,
      "loss": 1.4778,
      "step": 3478
    },
    {
      "epoch": 0.5586062941554271,
      "grad_norm": 0.2576897144317627,
      "learning_rate": 0.0001,
      "loss": 1.4639,
      "step": 3479
    },
    {
      "epoch": 0.558766859344894,
      "grad_norm": 0.2620057463645935,
      "learning_rate": 0.0001,
      "loss": 1.5513,
      "step": 3480
    },
    {
      "epoch": 0.558927424534361,
      "grad_norm": 0.267472505569458,
      "learning_rate": 0.0001,
      "loss": 1.5409,
      "step": 3481
    },
    {
      "epoch": 0.5590879897238279,
      "grad_norm": 0.2704296410083771,
      "learning_rate": 0.0001,
      "loss": 1.5411,
      "step": 3482
    },
    {
      "epoch": 0.5592485549132948,
      "grad_norm": 0.2500203549861908,
      "learning_rate": 0.0001,
      "loss": 1.5201,
      "step": 3483
    },
    {
      "epoch": 0.5594091201027617,
      "grad_norm": 0.24683894217014313,
      "learning_rate": 0.0001,
      "loss": 1.4956,
      "step": 3484
    },
    {
      "epoch": 0.5595696852922286,
      "grad_norm": 0.2554965913295746,
      "learning_rate": 0.0001,
      "loss": 1.468,
      "step": 3485
    },
    {
      "epoch": 0.5597302504816956,
      "grad_norm": 0.26204511523246765,
      "learning_rate": 0.0001,
      "loss": 1.4225,
      "step": 3486
    },
    {
      "epoch": 0.5598908156711625,
      "grad_norm": 0.25450706481933594,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 3487
    },
    {
      "epoch": 0.5600513808606294,
      "grad_norm": 0.2704123556613922,
      "learning_rate": 0.0001,
      "loss": 1.5706,
      "step": 3488
    },
    {
      "epoch": 0.5602119460500964,
      "grad_norm": 0.2498333752155304,
      "learning_rate": 0.0001,
      "loss": 1.5064,
      "step": 3489
    },
    {
      "epoch": 0.5603725112395632,
      "grad_norm": 0.2551574409008026,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 3490
    },
    {
      "epoch": 0.5605330764290302,
      "grad_norm": 0.25675302743911743,
      "learning_rate": 0.0001,
      "loss": 1.5415,
      "step": 3491
    },
    {
      "epoch": 0.5606936416184971,
      "grad_norm": 0.2501094341278076,
      "learning_rate": 0.0001,
      "loss": 1.4979,
      "step": 3492
    },
    {
      "epoch": 0.560854206807964,
      "grad_norm": 0.24903394281864166,
      "learning_rate": 0.0001,
      "loss": 1.5462,
      "step": 3493
    },
    {
      "epoch": 0.561014771997431,
      "grad_norm": 0.24453727900981903,
      "learning_rate": 0.0001,
      "loss": 1.4863,
      "step": 3494
    },
    {
      "epoch": 0.5611753371868978,
      "grad_norm": 0.26887333393096924,
      "learning_rate": 0.0001,
      "loss": 1.449,
      "step": 3495
    },
    {
      "epoch": 0.5613359023763648,
      "grad_norm": 0.2653504014015198,
      "learning_rate": 0.0001,
      "loss": 1.6348,
      "step": 3496
    },
    {
      "epoch": 0.5614964675658317,
      "grad_norm": 0.2897603511810303,
      "learning_rate": 0.0001,
      "loss": 1.5136,
      "step": 3497
    },
    {
      "epoch": 0.5616570327552987,
      "grad_norm": 0.2508017420768738,
      "learning_rate": 0.0001,
      "loss": 1.4904,
      "step": 3498
    },
    {
      "epoch": 0.5618175979447656,
      "grad_norm": 0.2683207094669342,
      "learning_rate": 0.0001,
      "loss": 1.5295,
      "step": 3499
    },
    {
      "epoch": 0.5619781631342325,
      "grad_norm": 0.2680720388889313,
      "learning_rate": 0.0001,
      "loss": 1.5504,
      "step": 3500
    },
    {
      "epoch": 0.5621387283236994,
      "grad_norm": 0.24851787090301514,
      "learning_rate": 0.0001,
      "loss": 1.4618,
      "step": 3501
    },
    {
      "epoch": 0.5622992935131663,
      "grad_norm": 0.27713721990585327,
      "learning_rate": 0.0001,
      "loss": 1.5878,
      "step": 3502
    },
    {
      "epoch": 0.5624598587026333,
      "grad_norm": 0.2540768086910248,
      "learning_rate": 0.0001,
      "loss": 1.4579,
      "step": 3503
    },
    {
      "epoch": 0.5626204238921002,
      "grad_norm": 0.26961416006088257,
      "learning_rate": 0.0001,
      "loss": 1.5374,
      "step": 3504
    },
    {
      "epoch": 0.5627809890815672,
      "grad_norm": 0.2740367650985718,
      "learning_rate": 0.0001,
      "loss": 1.5216,
      "step": 3505
    },
    {
      "epoch": 0.5629415542710341,
      "grad_norm": 0.26434916257858276,
      "learning_rate": 0.0001,
      "loss": 1.593,
      "step": 3506
    },
    {
      "epoch": 0.5631021194605009,
      "grad_norm": 0.25455477833747864,
      "learning_rate": 0.0001,
      "loss": 1.473,
      "step": 3507
    },
    {
      "epoch": 0.5632626846499679,
      "grad_norm": 0.27761974930763245,
      "learning_rate": 0.0001,
      "loss": 1.4984,
      "step": 3508
    },
    {
      "epoch": 0.5634232498394348,
      "grad_norm": 0.2448902279138565,
      "learning_rate": 0.0001,
      "loss": 1.4101,
      "step": 3509
    },
    {
      "epoch": 0.5635838150289018,
      "grad_norm": 0.3218258023262024,
      "learning_rate": 0.0001,
      "loss": 1.3785,
      "step": 3510
    },
    {
      "epoch": 0.5637443802183687,
      "grad_norm": 0.2583906650543213,
      "learning_rate": 0.0001,
      "loss": 1.5542,
      "step": 3511
    },
    {
      "epoch": 0.5639049454078355,
      "grad_norm": 0.2744387686252594,
      "learning_rate": 0.0001,
      "loss": 1.5306,
      "step": 3512
    },
    {
      "epoch": 0.5640655105973025,
      "grad_norm": 0.2542562484741211,
      "learning_rate": 0.0001,
      "loss": 1.5573,
      "step": 3513
    },
    {
      "epoch": 0.5642260757867694,
      "grad_norm": 0.27191752195358276,
      "learning_rate": 0.0001,
      "loss": 1.5455,
      "step": 3514
    },
    {
      "epoch": 0.5643866409762364,
      "grad_norm": 0.2717880308628082,
      "learning_rate": 0.0001,
      "loss": 1.4949,
      "step": 3515
    },
    {
      "epoch": 0.5645472061657033,
      "grad_norm": 0.2628335654735565,
      "learning_rate": 0.0001,
      "loss": 1.4825,
      "step": 3516
    },
    {
      "epoch": 0.5647077713551703,
      "grad_norm": 0.2621069550514221,
      "learning_rate": 0.0001,
      "loss": 1.6146,
      "step": 3517
    },
    {
      "epoch": 0.5648683365446371,
      "grad_norm": 0.28826311230659485,
      "learning_rate": 0.0001,
      "loss": 1.5447,
      "step": 3518
    },
    {
      "epoch": 0.565028901734104,
      "grad_norm": 0.2646731436252594,
      "learning_rate": 0.0001,
      "loss": 1.5986,
      "step": 3519
    },
    {
      "epoch": 0.565189466923571,
      "grad_norm": 0.2607734501361847,
      "learning_rate": 0.0001,
      "loss": 1.4956,
      "step": 3520
    },
    {
      "epoch": 0.5653500321130379,
      "grad_norm": 0.25392746925354004,
      "learning_rate": 0.0001,
      "loss": 1.4504,
      "step": 3521
    },
    {
      "epoch": 0.5655105973025049,
      "grad_norm": 0.2602335810661316,
      "learning_rate": 0.0001,
      "loss": 1.4503,
      "step": 3522
    },
    {
      "epoch": 0.5656711624919717,
      "grad_norm": 0.2631988525390625,
      "learning_rate": 0.0001,
      "loss": 1.4503,
      "step": 3523
    },
    {
      "epoch": 0.5658317276814386,
      "grad_norm": 0.27291157841682434,
      "learning_rate": 0.0001,
      "loss": 1.5659,
      "step": 3524
    },
    {
      "epoch": 0.5659922928709056,
      "grad_norm": 0.25426504015922546,
      "learning_rate": 0.0001,
      "loss": 1.5112,
      "step": 3525
    },
    {
      "epoch": 0.5661528580603725,
      "grad_norm": 0.2615494430065155,
      "learning_rate": 0.0001,
      "loss": 1.4765,
      "step": 3526
    },
    {
      "epoch": 0.5663134232498395,
      "grad_norm": 0.25024959444999695,
      "learning_rate": 0.0001,
      "loss": 1.4416,
      "step": 3527
    },
    {
      "epoch": 0.5664739884393064,
      "grad_norm": 0.2591741979122162,
      "learning_rate": 0.0001,
      "loss": 1.4866,
      "step": 3528
    },
    {
      "epoch": 0.5666345536287732,
      "grad_norm": 0.26453763246536255,
      "learning_rate": 0.0001,
      "loss": 1.5254,
      "step": 3529
    },
    {
      "epoch": 0.5667951188182402,
      "grad_norm": 0.24958562850952148,
      "learning_rate": 0.0001,
      "loss": 1.4718,
      "step": 3530
    },
    {
      "epoch": 0.5669556840077071,
      "grad_norm": 0.2647601068019867,
      "learning_rate": 0.0001,
      "loss": 1.4356,
      "step": 3531
    },
    {
      "epoch": 0.5671162491971741,
      "grad_norm": 0.28071776032447815,
      "learning_rate": 0.0001,
      "loss": 1.4949,
      "step": 3532
    },
    {
      "epoch": 0.567276814386641,
      "grad_norm": 0.25910112261772156,
      "learning_rate": 0.0001,
      "loss": 1.4674,
      "step": 3533
    },
    {
      "epoch": 0.5674373795761078,
      "grad_norm": 0.2666882276535034,
      "learning_rate": 0.0001,
      "loss": 1.5594,
      "step": 3534
    },
    {
      "epoch": 0.5675979447655748,
      "grad_norm": 0.27435198426246643,
      "learning_rate": 0.0001,
      "loss": 1.4877,
      "step": 3535
    },
    {
      "epoch": 0.5677585099550417,
      "grad_norm": 0.2550223469734192,
      "learning_rate": 0.0001,
      "loss": 1.4289,
      "step": 3536
    },
    {
      "epoch": 0.5679190751445087,
      "grad_norm": 0.2811153531074524,
      "learning_rate": 0.0001,
      "loss": 1.5937,
      "step": 3537
    },
    {
      "epoch": 0.5680796403339756,
      "grad_norm": 0.27584466338157654,
      "learning_rate": 0.0001,
      "loss": 1.508,
      "step": 3538
    },
    {
      "epoch": 0.5682402055234426,
      "grad_norm": 0.2671346962451935,
      "learning_rate": 0.0001,
      "loss": 1.5133,
      "step": 3539
    },
    {
      "epoch": 0.5684007707129094,
      "grad_norm": 0.26532262563705444,
      "learning_rate": 0.0001,
      "loss": 1.5286,
      "step": 3540
    },
    {
      "epoch": 0.5685613359023763,
      "grad_norm": 0.2532363831996918,
      "learning_rate": 0.0001,
      "loss": 1.5986,
      "step": 3541
    },
    {
      "epoch": 0.5687219010918433,
      "grad_norm": 0.2529861032962799,
      "learning_rate": 0.0001,
      "loss": 1.5459,
      "step": 3542
    },
    {
      "epoch": 0.5688824662813102,
      "grad_norm": 0.26175493001937866,
      "learning_rate": 0.0001,
      "loss": 1.6012,
      "step": 3543
    },
    {
      "epoch": 0.5690430314707772,
      "grad_norm": 0.25633907318115234,
      "learning_rate": 0.0001,
      "loss": 1.5074,
      "step": 3544
    },
    {
      "epoch": 0.569203596660244,
      "grad_norm": 0.2499510496854782,
      "learning_rate": 0.0001,
      "loss": 1.4577,
      "step": 3545
    },
    {
      "epoch": 0.569364161849711,
      "grad_norm": 0.269633024930954,
      "learning_rate": 0.0001,
      "loss": 1.5499,
      "step": 3546
    },
    {
      "epoch": 0.5695247270391779,
      "grad_norm": 25.48210906982422,
      "learning_rate": 0.0001,
      "loss": 1.6888,
      "step": 3547
    },
    {
      "epoch": 0.5696852922286448,
      "grad_norm": 0.29812732338905334,
      "learning_rate": 0.0001,
      "loss": 1.4714,
      "step": 3548
    },
    {
      "epoch": 0.5698458574181118,
      "grad_norm": 0.30050480365753174,
      "learning_rate": 0.0001,
      "loss": 1.5308,
      "step": 3549
    },
    {
      "epoch": 0.5700064226075787,
      "grad_norm": 0.30164986848831177,
      "learning_rate": 0.0001,
      "loss": 1.5135,
      "step": 3550
    },
    {
      "epoch": 0.5701669877970456,
      "grad_norm": 0.29812827706336975,
      "learning_rate": 0.0001,
      "loss": 1.6057,
      "step": 3551
    },
    {
      "epoch": 0.5703275529865125,
      "grad_norm": 0.3121807873249054,
      "learning_rate": 0.0001,
      "loss": 1.4885,
      "step": 3552
    },
    {
      "epoch": 0.5704881181759794,
      "grad_norm": 0.28077369928359985,
      "learning_rate": 0.0001,
      "loss": 1.4681,
      "step": 3553
    },
    {
      "epoch": 0.5706486833654464,
      "grad_norm": 0.27972835302352905,
      "learning_rate": 0.0001,
      "loss": 1.4775,
      "step": 3554
    },
    {
      "epoch": 0.5708092485549133,
      "grad_norm": 0.2922600507736206,
      "learning_rate": 0.0001,
      "loss": 1.446,
      "step": 3555
    },
    {
      "epoch": 0.5709698137443803,
      "grad_norm": 0.27353715896606445,
      "learning_rate": 0.0001,
      "loss": 1.5218,
      "step": 3556
    },
    {
      "epoch": 0.5711303789338471,
      "grad_norm": 0.2593901455402374,
      "learning_rate": 0.0001,
      "loss": 1.4743,
      "step": 3557
    },
    {
      "epoch": 0.571290944123314,
      "grad_norm": 0.2923367917537689,
      "learning_rate": 0.0001,
      "loss": 1.5646,
      "step": 3558
    },
    {
      "epoch": 0.571451509312781,
      "grad_norm": 0.34674420952796936,
      "learning_rate": 0.0001,
      "loss": 1.5701,
      "step": 3559
    },
    {
      "epoch": 0.5716120745022479,
      "grad_norm": 0.3112606108188629,
      "learning_rate": 0.0001,
      "loss": 1.4606,
      "step": 3560
    },
    {
      "epoch": 0.5717726396917149,
      "grad_norm": 0.2782287299633026,
      "learning_rate": 0.0001,
      "loss": 1.4974,
      "step": 3561
    },
    {
      "epoch": 0.5719332048811817,
      "grad_norm": 0.2519095838069916,
      "learning_rate": 0.0001,
      "loss": 1.5007,
      "step": 3562
    },
    {
      "epoch": 0.5720937700706487,
      "grad_norm": 0.24846331775188446,
      "learning_rate": 0.0001,
      "loss": 1.4191,
      "step": 3563
    },
    {
      "epoch": 0.5722543352601156,
      "grad_norm": 0.2899773418903351,
      "learning_rate": 0.0001,
      "loss": 1.5408,
      "step": 3564
    },
    {
      "epoch": 0.5724149004495825,
      "grad_norm": 0.25818878412246704,
      "learning_rate": 0.0001,
      "loss": 1.5233,
      "step": 3565
    },
    {
      "epoch": 0.5725754656390495,
      "grad_norm": 0.30965518951416016,
      "learning_rate": 0.0001,
      "loss": 1.48,
      "step": 3566
    },
    {
      "epoch": 0.5727360308285164,
      "grad_norm": 0.26094964146614075,
      "learning_rate": 0.0001,
      "loss": 1.536,
      "step": 3567
    },
    {
      "epoch": 0.5728965960179833,
      "grad_norm": 0.2866499722003937,
      "learning_rate": 0.0001,
      "loss": 1.5216,
      "step": 3568
    },
    {
      "epoch": 0.5730571612074502,
      "grad_norm": 0.28775933384895325,
      "learning_rate": 0.0001,
      "loss": 1.5449,
      "step": 3569
    },
    {
      "epoch": 0.5732177263969171,
      "grad_norm": 0.2771734893321991,
      "learning_rate": 0.0001,
      "loss": 1.4515,
      "step": 3570
    },
    {
      "epoch": 0.5733782915863841,
      "grad_norm": 0.2711351215839386,
      "learning_rate": 0.0001,
      "loss": 1.4363,
      "step": 3571
    },
    {
      "epoch": 0.573538856775851,
      "grad_norm": 0.2581285238265991,
      "learning_rate": 0.0001,
      "loss": 1.4205,
      "step": 3572
    },
    {
      "epoch": 0.5736994219653179,
      "grad_norm": 0.27857813239097595,
      "learning_rate": 0.0001,
      "loss": 1.5103,
      "step": 3573
    },
    {
      "epoch": 0.5738599871547848,
      "grad_norm": 0.2940189242362976,
      "learning_rate": 0.0001,
      "loss": 1.4527,
      "step": 3574
    },
    {
      "epoch": 0.5740205523442518,
      "grad_norm": 0.2706471085548401,
      "learning_rate": 0.0001,
      "loss": 1.5351,
      "step": 3575
    },
    {
      "epoch": 0.5741811175337187,
      "grad_norm": 0.26701080799102783,
      "learning_rate": 0.0001,
      "loss": 1.4732,
      "step": 3576
    },
    {
      "epoch": 0.5743416827231856,
      "grad_norm": 0.27279427647590637,
      "learning_rate": 0.0001,
      "loss": 1.4732,
      "step": 3577
    },
    {
      "epoch": 0.5745022479126526,
      "grad_norm": 0.3203839659690857,
      "learning_rate": 0.0001,
      "loss": 1.4648,
      "step": 3578
    },
    {
      "epoch": 0.5746628131021194,
      "grad_norm": 0.25967317819595337,
      "learning_rate": 0.0001,
      "loss": 1.5407,
      "step": 3579
    },
    {
      "epoch": 0.5748233782915864,
      "grad_norm": 0.2614003121852875,
      "learning_rate": 0.0001,
      "loss": 1.5071,
      "step": 3580
    },
    {
      "epoch": 0.5749839434810533,
      "grad_norm": 0.2658310830593109,
      "learning_rate": 0.0001,
      "loss": 1.4692,
      "step": 3581
    },
    {
      "epoch": 0.5751445086705202,
      "grad_norm": 0.2863481938838959,
      "learning_rate": 0.0001,
      "loss": 1.4733,
      "step": 3582
    },
    {
      "epoch": 0.5753050738599872,
      "grad_norm": 0.2631807327270508,
      "learning_rate": 0.0001,
      "loss": 1.471,
      "step": 3583
    },
    {
      "epoch": 0.575465639049454,
      "grad_norm": 0.24985304474830627,
      "learning_rate": 0.0001,
      "loss": 1.5507,
      "step": 3584
    },
    {
      "epoch": 0.575626204238921,
      "grad_norm": 0.26440948247909546,
      "learning_rate": 0.0001,
      "loss": 1.5232,
      "step": 3585
    },
    {
      "epoch": 0.5757867694283879,
      "grad_norm": 0.2813529372215271,
      "learning_rate": 0.0001,
      "loss": 1.6416,
      "step": 3586
    },
    {
      "epoch": 0.5759473346178549,
      "grad_norm": 0.27158108353614807,
      "learning_rate": 0.0001,
      "loss": 1.4755,
      "step": 3587
    },
    {
      "epoch": 0.5761078998073218,
      "grad_norm": 0.26029500365257263,
      "learning_rate": 0.0001,
      "loss": 1.5513,
      "step": 3588
    },
    {
      "epoch": 0.5762684649967887,
      "grad_norm": 0.2807159125804901,
      "learning_rate": 0.0001,
      "loss": 1.5234,
      "step": 3589
    },
    {
      "epoch": 0.5764290301862556,
      "grad_norm": 0.25395241379737854,
      "learning_rate": 0.0001,
      "loss": 1.4321,
      "step": 3590
    },
    {
      "epoch": 0.5765895953757225,
      "grad_norm": 0.2527233362197876,
      "learning_rate": 0.0001,
      "loss": 1.5193,
      "step": 3591
    },
    {
      "epoch": 0.5767501605651895,
      "grad_norm": 0.27952840924263,
      "learning_rate": 0.0001,
      "loss": 1.4721,
      "step": 3592
    },
    {
      "epoch": 0.5769107257546564,
      "grad_norm": 0.26070594787597656,
      "learning_rate": 0.0001,
      "loss": 1.5733,
      "step": 3593
    },
    {
      "epoch": 0.5770712909441233,
      "grad_norm": 0.25981488823890686,
      "learning_rate": 0.0001,
      "loss": 1.5009,
      "step": 3594
    },
    {
      "epoch": 0.5772318561335903,
      "grad_norm": 0.2579481303691864,
      "learning_rate": 0.0001,
      "loss": 1.5066,
      "step": 3595
    },
    {
      "epoch": 0.5773924213230571,
      "grad_norm": 0.2575138807296753,
      "learning_rate": 0.0001,
      "loss": 1.548,
      "step": 3596
    },
    {
      "epoch": 0.5775529865125241,
      "grad_norm": 0.2819560766220093,
      "learning_rate": 0.0001,
      "loss": 1.4838,
      "step": 3597
    },
    {
      "epoch": 0.577713551701991,
      "grad_norm": 0.28816813230514526,
      "learning_rate": 0.0001,
      "loss": 1.5963,
      "step": 3598
    },
    {
      "epoch": 0.577874116891458,
      "grad_norm": 0.2554091215133667,
      "learning_rate": 0.0001,
      "loss": 1.4762,
      "step": 3599
    },
    {
      "epoch": 0.5780346820809249,
      "grad_norm": 0.2506262958049774,
      "learning_rate": 0.0001,
      "loss": 1.4941,
      "step": 3600
    },
    {
      "epoch": 0.5781952472703917,
      "grad_norm": 0.27438631653785706,
      "learning_rate": 0.0001,
      "loss": 1.4969,
      "step": 3601
    },
    {
      "epoch": 0.5783558124598587,
      "grad_norm": 0.30739980936050415,
      "learning_rate": 0.0001,
      "loss": 1.5721,
      "step": 3602
    },
    {
      "epoch": 0.5785163776493256,
      "grad_norm": 0.2560059726238251,
      "learning_rate": 0.0001,
      "loss": 1.6132,
      "step": 3603
    },
    {
      "epoch": 0.5786769428387926,
      "grad_norm": 0.26068899035453796,
      "learning_rate": 0.0001,
      "loss": 1.429,
      "step": 3604
    },
    {
      "epoch": 0.5788375080282595,
      "grad_norm": 0.26843971014022827,
      "learning_rate": 0.0001,
      "loss": 1.4722,
      "step": 3605
    },
    {
      "epoch": 0.5789980732177264,
      "grad_norm": 0.2547934353351593,
      "learning_rate": 0.0001,
      "loss": 1.4931,
      "step": 3606
    },
    {
      "epoch": 0.5791586384071933,
      "grad_norm": 0.2655860185623169,
      "learning_rate": 0.0001,
      "loss": 1.5597,
      "step": 3607
    },
    {
      "epoch": 0.5793192035966602,
      "grad_norm": 0.2529540956020355,
      "learning_rate": 0.0001,
      "loss": 1.4938,
      "step": 3608
    },
    {
      "epoch": 0.5794797687861272,
      "grad_norm": 0.2818930447101593,
      "learning_rate": 0.0001,
      "loss": 1.554,
      "step": 3609
    },
    {
      "epoch": 0.5796403339755941,
      "grad_norm": 0.25783148407936096,
      "learning_rate": 0.0001,
      "loss": 1.4835,
      "step": 3610
    },
    {
      "epoch": 0.5798008991650611,
      "grad_norm": 0.26304492354393005,
      "learning_rate": 0.0001,
      "loss": 1.4554,
      "step": 3611
    },
    {
      "epoch": 0.5799614643545279,
      "grad_norm": 0.3086816668510437,
      "learning_rate": 0.0001,
      "loss": 1.4979,
      "step": 3612
    },
    {
      "epoch": 0.5801220295439948,
      "grad_norm": 0.2533327341079712,
      "learning_rate": 0.0001,
      "loss": 1.5208,
      "step": 3613
    },
    {
      "epoch": 0.5802825947334618,
      "grad_norm": 0.2684653103351593,
      "learning_rate": 0.0001,
      "loss": 1.6161,
      "step": 3614
    },
    {
      "epoch": 0.5804431599229287,
      "grad_norm": 0.2483733743429184,
      "learning_rate": 0.0001,
      "loss": 1.49,
      "step": 3615
    },
    {
      "epoch": 0.5806037251123957,
      "grad_norm": 0.26590898633003235,
      "learning_rate": 0.0001,
      "loss": 1.4978,
      "step": 3616
    },
    {
      "epoch": 0.5807642903018626,
      "grad_norm": 0.25449588894844055,
      "learning_rate": 0.0001,
      "loss": 1.5236,
      "step": 3617
    },
    {
      "epoch": 0.5809248554913294,
      "grad_norm": 0.25207141041755676,
      "learning_rate": 0.0001,
      "loss": 1.4599,
      "step": 3618
    },
    {
      "epoch": 0.5810854206807964,
      "grad_norm": 0.2689812481403351,
      "learning_rate": 0.0001,
      "loss": 1.5096,
      "step": 3619
    },
    {
      "epoch": 0.5812459858702633,
      "grad_norm": 0.2550770044326782,
      "learning_rate": 0.0001,
      "loss": 1.5115,
      "step": 3620
    },
    {
      "epoch": 0.5814065510597303,
      "grad_norm": 0.2674384117126465,
      "learning_rate": 0.0001,
      "loss": 1.5326,
      "step": 3621
    },
    {
      "epoch": 0.5815671162491972,
      "grad_norm": 0.2624462842941284,
      "learning_rate": 0.0001,
      "loss": 1.5497,
      "step": 3622
    },
    {
      "epoch": 0.581727681438664,
      "grad_norm": 0.2592432200908661,
      "learning_rate": 0.0001,
      "loss": 1.5431,
      "step": 3623
    },
    {
      "epoch": 0.581888246628131,
      "grad_norm": 0.25049206614494324,
      "learning_rate": 0.0001,
      "loss": 1.4347,
      "step": 3624
    },
    {
      "epoch": 0.5820488118175979,
      "grad_norm": 0.2545790374279022,
      "learning_rate": 0.0001,
      "loss": 1.538,
      "step": 3625
    },
    {
      "epoch": 0.5822093770070649,
      "grad_norm": 0.2624324560165405,
      "learning_rate": 0.0001,
      "loss": 1.5579,
      "step": 3626
    },
    {
      "epoch": 0.5823699421965318,
      "grad_norm": 0.26768431067466736,
      "learning_rate": 0.0001,
      "loss": 1.526,
      "step": 3627
    },
    {
      "epoch": 0.5825305073859988,
      "grad_norm": 0.25618383288383484,
      "learning_rate": 0.0001,
      "loss": 1.5025,
      "step": 3628
    },
    {
      "epoch": 0.5826910725754656,
      "grad_norm": 0.2567412257194519,
      "learning_rate": 0.0001,
      "loss": 1.5235,
      "step": 3629
    },
    {
      "epoch": 0.5828516377649325,
      "grad_norm": 0.2692475914955139,
      "learning_rate": 0.0001,
      "loss": 1.5209,
      "step": 3630
    },
    {
      "epoch": 0.5830122029543995,
      "grad_norm": 0.2622545659542084,
      "learning_rate": 0.0001,
      "loss": 1.4982,
      "step": 3631
    },
    {
      "epoch": 0.5831727681438664,
      "grad_norm": 0.27413806319236755,
      "learning_rate": 0.0001,
      "loss": 1.5426,
      "step": 3632
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.2643025815486908,
      "learning_rate": 0.0001,
      "loss": 1.5348,
      "step": 3633
    },
    {
      "epoch": 0.5834938985228002,
      "grad_norm": 0.2692074477672577,
      "learning_rate": 0.0001,
      "loss": 1.5558,
      "step": 3634
    },
    {
      "epoch": 0.5836544637122671,
      "grad_norm": 0.25885412096977234,
      "learning_rate": 0.0001,
      "loss": 1.5358,
      "step": 3635
    },
    {
      "epoch": 0.5838150289017341,
      "grad_norm": 0.25478604435920715,
      "learning_rate": 0.0001,
      "loss": 1.5737,
      "step": 3636
    },
    {
      "epoch": 0.583975594091201,
      "grad_norm": 0.24968627095222473,
      "learning_rate": 0.0001,
      "loss": 1.471,
      "step": 3637
    },
    {
      "epoch": 0.584136159280668,
      "grad_norm": 0.24788379669189453,
      "learning_rate": 0.0001,
      "loss": 1.5014,
      "step": 3638
    },
    {
      "epoch": 0.5842967244701349,
      "grad_norm": 0.2518269121646881,
      "learning_rate": 0.0001,
      "loss": 1.5161,
      "step": 3639
    },
    {
      "epoch": 0.5844572896596018,
      "grad_norm": 0.25715354084968567,
      "learning_rate": 0.0001,
      "loss": 1.3978,
      "step": 3640
    },
    {
      "epoch": 0.5846178548490687,
      "grad_norm": 0.2593884766101837,
      "learning_rate": 0.0001,
      "loss": 1.4308,
      "step": 3641
    },
    {
      "epoch": 0.5847784200385356,
      "grad_norm": 0.27028846740722656,
      "learning_rate": 0.0001,
      "loss": 1.5397,
      "step": 3642
    },
    {
      "epoch": 0.5849389852280026,
      "grad_norm": 0.2675887942314148,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 3643
    },
    {
      "epoch": 0.5850995504174695,
      "grad_norm": 0.262739360332489,
      "learning_rate": 0.0001,
      "loss": 1.4573,
      "step": 3644
    },
    {
      "epoch": 0.5852601156069365,
      "grad_norm": 0.2562864422798157,
      "learning_rate": 0.0001,
      "loss": 1.5833,
      "step": 3645
    },
    {
      "epoch": 0.5854206807964033,
      "grad_norm": 0.2761216163635254,
      "learning_rate": 0.0001,
      "loss": 1.6049,
      "step": 3646
    },
    {
      "epoch": 0.5855812459858702,
      "grad_norm": 0.26393917202949524,
      "learning_rate": 0.0001,
      "loss": 1.5133,
      "step": 3647
    },
    {
      "epoch": 0.5857418111753372,
      "grad_norm": 0.26334771513938904,
      "learning_rate": 0.0001,
      "loss": 1.5213,
      "step": 3648
    },
    {
      "epoch": 0.5859023763648041,
      "grad_norm": 0.24718475341796875,
      "learning_rate": 0.0001,
      "loss": 1.4571,
      "step": 3649
    },
    {
      "epoch": 0.5860629415542711,
      "grad_norm": 0.27074918150901794,
      "learning_rate": 0.0001,
      "loss": 1.5857,
      "step": 3650
    },
    {
      "epoch": 0.5862235067437379,
      "grad_norm": 0.25103697180747986,
      "learning_rate": 0.0001,
      "loss": 1.5143,
      "step": 3651
    },
    {
      "epoch": 0.5863840719332049,
      "grad_norm": 0.2570626139640808,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 3652
    },
    {
      "epoch": 0.5865446371226718,
      "grad_norm": 0.2511909306049347,
      "learning_rate": 0.0001,
      "loss": 1.5365,
      "step": 3653
    },
    {
      "epoch": 0.5867052023121387,
      "grad_norm": 0.27300527691841125,
      "learning_rate": 0.0001,
      "loss": 1.4084,
      "step": 3654
    },
    {
      "epoch": 0.5868657675016057,
      "grad_norm": 0.28094717860221863,
      "learning_rate": 0.0001,
      "loss": 1.4748,
      "step": 3655
    },
    {
      "epoch": 0.5870263326910726,
      "grad_norm": 0.2525530159473419,
      "learning_rate": 0.0001,
      "loss": 1.4885,
      "step": 3656
    },
    {
      "epoch": 0.5871868978805395,
      "grad_norm": 0.25379708409309387,
      "learning_rate": 0.0001,
      "loss": 1.4142,
      "step": 3657
    },
    {
      "epoch": 0.5873474630700064,
      "grad_norm": 0.2683212161064148,
      "learning_rate": 0.0001,
      "loss": 1.5485,
      "step": 3658
    },
    {
      "epoch": 0.5875080282594733,
      "grad_norm": 0.25514858961105347,
      "learning_rate": 0.0001,
      "loss": 1.524,
      "step": 3659
    },
    {
      "epoch": 0.5876685934489403,
      "grad_norm": 0.2825925350189209,
      "learning_rate": 0.0001,
      "loss": 1.5755,
      "step": 3660
    },
    {
      "epoch": 0.5878291586384072,
      "grad_norm": 0.26648077368736267,
      "learning_rate": 0.0001,
      "loss": 1.5302,
      "step": 3661
    },
    {
      "epoch": 0.5879897238278741,
      "grad_norm": 0.3023388385772705,
      "learning_rate": 0.0001,
      "loss": 1.5183,
      "step": 3662
    },
    {
      "epoch": 0.588150289017341,
      "grad_norm": 0.2562459707260132,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 3663
    },
    {
      "epoch": 0.588310854206808,
      "grad_norm": 0.27271050214767456,
      "learning_rate": 0.0001,
      "loss": 1.4787,
      "step": 3664
    },
    {
      "epoch": 0.5884714193962749,
      "grad_norm": 0.2668638825416565,
      "learning_rate": 0.0001,
      "loss": 1.5099,
      "step": 3665
    },
    {
      "epoch": 0.5886319845857418,
      "grad_norm": 0.2641138732433319,
      "learning_rate": 0.0001,
      "loss": 1.5078,
      "step": 3666
    },
    {
      "epoch": 0.5887925497752088,
      "grad_norm": 0.2596166133880615,
      "learning_rate": 0.0001,
      "loss": 1.5392,
      "step": 3667
    },
    {
      "epoch": 0.5889531149646756,
      "grad_norm": 0.2693968117237091,
      "learning_rate": 0.0001,
      "loss": 1.5343,
      "step": 3668
    },
    {
      "epoch": 0.5891136801541426,
      "grad_norm": 0.2611267864704132,
      "learning_rate": 0.0001,
      "loss": 1.4792,
      "step": 3669
    },
    {
      "epoch": 0.5892742453436095,
      "grad_norm": 0.3592695891857147,
      "learning_rate": 0.0001,
      "loss": 1.526,
      "step": 3670
    },
    {
      "epoch": 0.5894348105330764,
      "grad_norm": 0.2588188648223877,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 3671
    },
    {
      "epoch": 0.5895953757225434,
      "grad_norm": 0.3175065517425537,
      "learning_rate": 0.0001,
      "loss": 1.5471,
      "step": 3672
    },
    {
      "epoch": 0.5897559409120102,
      "grad_norm": 0.2606745660305023,
      "learning_rate": 0.0001,
      "loss": 1.5067,
      "step": 3673
    },
    {
      "epoch": 0.5899165061014772,
      "grad_norm": 0.2673262655735016,
      "learning_rate": 0.0001,
      "loss": 1.5013,
      "step": 3674
    },
    {
      "epoch": 0.5900770712909441,
      "grad_norm": 0.24372580647468567,
      "learning_rate": 0.0001,
      "loss": 1.4876,
      "step": 3675
    },
    {
      "epoch": 0.590237636480411,
      "grad_norm": 0.25752654671669006,
      "learning_rate": 0.0001,
      "loss": 1.4871,
      "step": 3676
    },
    {
      "epoch": 0.590398201669878,
      "grad_norm": 0.25043997168540955,
      "learning_rate": 0.0001,
      "loss": 1.4646,
      "step": 3677
    },
    {
      "epoch": 0.5905587668593449,
      "grad_norm": 0.2570090889930725,
      "learning_rate": 0.0001,
      "loss": 1.5276,
      "step": 3678
    },
    {
      "epoch": 0.5907193320488118,
      "grad_norm": 0.2523142993450165,
      "learning_rate": 0.0001,
      "loss": 1.5158,
      "step": 3679
    },
    {
      "epoch": 0.5908798972382787,
      "grad_norm": 0.27723002433776855,
      "learning_rate": 0.0001,
      "loss": 1.5198,
      "step": 3680
    },
    {
      "epoch": 0.5910404624277457,
      "grad_norm": 0.2671026587486267,
      "learning_rate": 0.0001,
      "loss": 1.5323,
      "step": 3681
    },
    {
      "epoch": 0.5912010276172126,
      "grad_norm": 0.2731771171092987,
      "learning_rate": 0.0001,
      "loss": 1.5666,
      "step": 3682
    },
    {
      "epoch": 0.5913615928066795,
      "grad_norm": 0.2718448042869568,
      "learning_rate": 0.0001,
      "loss": 1.5432,
      "step": 3683
    },
    {
      "epoch": 0.5915221579961464,
      "grad_norm": 0.26460906863212585,
      "learning_rate": 0.0001,
      "loss": 1.4593,
      "step": 3684
    },
    {
      "epoch": 0.5916827231856133,
      "grad_norm": 0.248262420296669,
      "learning_rate": 0.0001,
      "loss": 1.4438,
      "step": 3685
    },
    {
      "epoch": 0.5918432883750803,
      "grad_norm": 0.3607558310031891,
      "learning_rate": 0.0001,
      "loss": 1.5388,
      "step": 3686
    },
    {
      "epoch": 0.5920038535645472,
      "grad_norm": 0.27468305826187134,
      "learning_rate": 0.0001,
      "loss": 1.5937,
      "step": 3687
    },
    {
      "epoch": 0.5921644187540142,
      "grad_norm": 0.26270800828933716,
      "learning_rate": 0.0001,
      "loss": 1.5876,
      "step": 3688
    },
    {
      "epoch": 0.5923249839434811,
      "grad_norm": 0.261650413274765,
      "learning_rate": 0.0001,
      "loss": 1.4658,
      "step": 3689
    },
    {
      "epoch": 0.5924855491329479,
      "grad_norm": 0.2564741373062134,
      "learning_rate": 0.0001,
      "loss": 1.5001,
      "step": 3690
    },
    {
      "epoch": 0.5926461143224149,
      "grad_norm": 0.2611472010612488,
      "learning_rate": 0.0001,
      "loss": 1.449,
      "step": 3691
    },
    {
      "epoch": 0.5928066795118818,
      "grad_norm": 0.25770461559295654,
      "learning_rate": 0.0001,
      "loss": 1.4628,
      "step": 3692
    },
    {
      "epoch": 0.5929672447013488,
      "grad_norm": 0.28395816683769226,
      "learning_rate": 0.0001,
      "loss": 1.4928,
      "step": 3693
    },
    {
      "epoch": 0.5931278098908157,
      "grad_norm": 0.2604933977127075,
      "learning_rate": 0.0001,
      "loss": 1.5971,
      "step": 3694
    },
    {
      "epoch": 0.5932883750802826,
      "grad_norm": 0.2781031131744385,
      "learning_rate": 0.0001,
      "loss": 1.5112,
      "step": 3695
    },
    {
      "epoch": 0.5934489402697495,
      "grad_norm": 0.2766863703727722,
      "learning_rate": 0.0001,
      "loss": 1.5517,
      "step": 3696
    },
    {
      "epoch": 0.5936095054592164,
      "grad_norm": 0.24513736367225647,
      "learning_rate": 0.0001,
      "loss": 1.4801,
      "step": 3697
    },
    {
      "epoch": 0.5937700706486834,
      "grad_norm": 0.2590049207210541,
      "learning_rate": 0.0001,
      "loss": 1.5126,
      "step": 3698
    },
    {
      "epoch": 0.5939306358381503,
      "grad_norm": 2.0267324447631836,
      "learning_rate": 0.0001,
      "loss": 1.4867,
      "step": 3699
    },
    {
      "epoch": 0.5940912010276173,
      "grad_norm": 0.27024027705192566,
      "learning_rate": 0.0001,
      "loss": 1.5394,
      "step": 3700
    },
    {
      "epoch": 0.5942517662170841,
      "grad_norm": 0.2798246443271637,
      "learning_rate": 0.0001,
      "loss": 1.5858,
      "step": 3701
    },
    {
      "epoch": 0.594412331406551,
      "grad_norm": 0.25654903054237366,
      "learning_rate": 0.0001,
      "loss": 1.5108,
      "step": 3702
    },
    {
      "epoch": 0.594572896596018,
      "grad_norm": 0.2551288902759552,
      "learning_rate": 0.0001,
      "loss": 1.4992,
      "step": 3703
    },
    {
      "epoch": 0.5947334617854849,
      "grad_norm": 0.24642886221408844,
      "learning_rate": 0.0001,
      "loss": 1.5106,
      "step": 3704
    },
    {
      "epoch": 0.5948940269749519,
      "grad_norm": 0.2912113070487976,
      "learning_rate": 0.0001,
      "loss": 1.509,
      "step": 3705
    },
    {
      "epoch": 0.5950545921644188,
      "grad_norm": 0.2692941725254059,
      "learning_rate": 0.0001,
      "loss": 1.4771,
      "step": 3706
    },
    {
      "epoch": 0.5952151573538856,
      "grad_norm": 0.267769992351532,
      "learning_rate": 0.0001,
      "loss": 1.5248,
      "step": 3707
    },
    {
      "epoch": 0.5953757225433526,
      "grad_norm": 0.2733960449695587,
      "learning_rate": 0.0001,
      "loss": 1.5027,
      "step": 3708
    },
    {
      "epoch": 0.5955362877328195,
      "grad_norm": 0.2550022304058075,
      "learning_rate": 0.0001,
      "loss": 1.5546,
      "step": 3709
    },
    {
      "epoch": 0.5956968529222865,
      "grad_norm": 0.2601030170917511,
      "learning_rate": 0.0001,
      "loss": 1.5325,
      "step": 3710
    },
    {
      "epoch": 0.5958574181117534,
      "grad_norm": 0.252185195684433,
      "learning_rate": 0.0001,
      "loss": 1.5049,
      "step": 3711
    },
    {
      "epoch": 0.5960179833012202,
      "grad_norm": 0.2696683704853058,
      "learning_rate": 0.0001,
      "loss": 1.4869,
      "step": 3712
    },
    {
      "epoch": 0.5961785484906872,
      "grad_norm": 0.25885897874832153,
      "learning_rate": 0.0001,
      "loss": 1.4866,
      "step": 3713
    },
    {
      "epoch": 0.5963391136801541,
      "grad_norm": 0.26752132177352905,
      "learning_rate": 0.0001,
      "loss": 1.5446,
      "step": 3714
    },
    {
      "epoch": 0.5964996788696211,
      "grad_norm": 0.2513367831707001,
      "learning_rate": 0.0001,
      "loss": 1.4674,
      "step": 3715
    },
    {
      "epoch": 0.596660244059088,
      "grad_norm": 0.25783655047416687,
      "learning_rate": 0.0001,
      "loss": 1.4558,
      "step": 3716
    },
    {
      "epoch": 0.596820809248555,
      "grad_norm": 0.2807951271533966,
      "learning_rate": 0.0001,
      "loss": 1.4932,
      "step": 3717
    },
    {
      "epoch": 0.5969813744380218,
      "grad_norm": 0.25926467776298523,
      "learning_rate": 0.0001,
      "loss": 1.572,
      "step": 3718
    },
    {
      "epoch": 0.5971419396274887,
      "grad_norm": 0.2520737648010254,
      "learning_rate": 0.0001,
      "loss": 1.5898,
      "step": 3719
    },
    {
      "epoch": 0.5973025048169557,
      "grad_norm": 0.24815239012241364,
      "learning_rate": 0.0001,
      "loss": 1.5617,
      "step": 3720
    },
    {
      "epoch": 0.5974630700064226,
      "grad_norm": 0.2779528498649597,
      "learning_rate": 0.0001,
      "loss": 1.5554,
      "step": 3721
    },
    {
      "epoch": 0.5976236351958896,
      "grad_norm": 0.26077648997306824,
      "learning_rate": 0.0001,
      "loss": 1.5922,
      "step": 3722
    },
    {
      "epoch": 0.5977842003853564,
      "grad_norm": 0.2544933557510376,
      "learning_rate": 0.0001,
      "loss": 1.5108,
      "step": 3723
    },
    {
      "epoch": 0.5979447655748233,
      "grad_norm": 0.2520194947719574,
      "learning_rate": 0.0001,
      "loss": 1.4948,
      "step": 3724
    },
    {
      "epoch": 0.5981053307642903,
      "grad_norm": 0.264365553855896,
      "learning_rate": 0.0001,
      "loss": 1.4446,
      "step": 3725
    },
    {
      "epoch": 0.5982658959537572,
      "grad_norm": 0.2670818269252777,
      "learning_rate": 0.0001,
      "loss": 1.4515,
      "step": 3726
    },
    {
      "epoch": 0.5984264611432242,
      "grad_norm": 0.2703360319137573,
      "learning_rate": 0.0001,
      "loss": 1.4897,
      "step": 3727
    },
    {
      "epoch": 0.5985870263326911,
      "grad_norm": 0.27654823660850525,
      "learning_rate": 0.0001,
      "loss": 1.4311,
      "step": 3728
    },
    {
      "epoch": 0.598747591522158,
      "grad_norm": 0.26018691062927246,
      "learning_rate": 0.0001,
      "loss": 1.5499,
      "step": 3729
    },
    {
      "epoch": 0.5989081567116249,
      "grad_norm": 0.26952722668647766,
      "learning_rate": 0.0001,
      "loss": 1.5381,
      "step": 3730
    },
    {
      "epoch": 0.5990687219010918,
      "grad_norm": 0.26965275406837463,
      "learning_rate": 0.0001,
      "loss": 1.52,
      "step": 3731
    },
    {
      "epoch": 0.5992292870905588,
      "grad_norm": 0.2653871178627014,
      "learning_rate": 0.0001,
      "loss": 1.5566,
      "step": 3732
    },
    {
      "epoch": 0.5993898522800257,
      "grad_norm": 0.2541314959526062,
      "learning_rate": 0.0001,
      "loss": 1.5139,
      "step": 3733
    },
    {
      "epoch": 0.5995504174694927,
      "grad_norm": 0.2657106816768646,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 3734
    },
    {
      "epoch": 0.5997109826589595,
      "grad_norm": 0.25480684638023376,
      "learning_rate": 0.0001,
      "loss": 1.5435,
      "step": 3735
    },
    {
      "epoch": 0.5998715478484264,
      "grad_norm": 0.2721002995967865,
      "learning_rate": 0.0001,
      "loss": 1.5409,
      "step": 3736
    },
    {
      "epoch": 0.6000321130378934,
      "grad_norm": 0.2616802155971527,
      "learning_rate": 0.0001,
      "loss": 1.5084,
      "step": 3737
    },
    {
      "epoch": 0.6001926782273603,
      "grad_norm": 0.26311367750167847,
      "learning_rate": 0.0001,
      "loss": 1.5424,
      "step": 3738
    },
    {
      "epoch": 0.6003532434168273,
      "grad_norm": 0.2507343888282776,
      "learning_rate": 0.0001,
      "loss": 1.4231,
      "step": 3739
    },
    {
      "epoch": 0.6005138086062941,
      "grad_norm": 0.26450052857398987,
      "learning_rate": 0.0001,
      "loss": 1.4702,
      "step": 3740
    },
    {
      "epoch": 0.600674373795761,
      "grad_norm": 0.2669855058193207,
      "learning_rate": 0.0001,
      "loss": 1.4722,
      "step": 3741
    },
    {
      "epoch": 0.600834938985228,
      "grad_norm": 0.2633204460144043,
      "learning_rate": 0.0001,
      "loss": 1.5088,
      "step": 3742
    },
    {
      "epoch": 0.6009955041746949,
      "grad_norm": 0.2775106728076935,
      "learning_rate": 0.0001,
      "loss": 1.4919,
      "step": 3743
    },
    {
      "epoch": 0.6011560693641619,
      "grad_norm": 0.25679922103881836,
      "learning_rate": 0.0001,
      "loss": 1.4259,
      "step": 3744
    },
    {
      "epoch": 0.6013166345536288,
      "grad_norm": 0.2631145119667053,
      "learning_rate": 0.0001,
      "loss": 1.5751,
      "step": 3745
    },
    {
      "epoch": 0.6014771997430957,
      "grad_norm": 0.28631600737571716,
      "learning_rate": 0.0001,
      "loss": 1.5594,
      "step": 3746
    },
    {
      "epoch": 0.6016377649325626,
      "grad_norm": 0.27606093883514404,
      "learning_rate": 0.0001,
      "loss": 1.4945,
      "step": 3747
    },
    {
      "epoch": 0.6017983301220295,
      "grad_norm": 0.28245851397514343,
      "learning_rate": 0.0001,
      "loss": 1.4792,
      "step": 3748
    },
    {
      "epoch": 0.6019588953114965,
      "grad_norm": 0.26298055052757263,
      "learning_rate": 0.0001,
      "loss": 1.5375,
      "step": 3749
    },
    {
      "epoch": 0.6021194605009634,
      "grad_norm": 0.27240198850631714,
      "learning_rate": 0.0001,
      "loss": 1.5082,
      "step": 3750
    },
    {
      "epoch": 0.6022800256904303,
      "grad_norm": 0.25478267669677734,
      "learning_rate": 0.0001,
      "loss": 1.5485,
      "step": 3751
    },
    {
      "epoch": 0.6024405908798972,
      "grad_norm": 0.2596980929374695,
      "learning_rate": 0.0001,
      "loss": 1.5212,
      "step": 3752
    },
    {
      "epoch": 0.6026011560693642,
      "grad_norm": 0.2807830572128296,
      "learning_rate": 0.0001,
      "loss": 1.4283,
      "step": 3753
    },
    {
      "epoch": 0.6027617212588311,
      "grad_norm": 0.2505226135253906,
      "learning_rate": 0.0001,
      "loss": 1.4848,
      "step": 3754
    },
    {
      "epoch": 0.602922286448298,
      "grad_norm": 0.2531312108039856,
      "learning_rate": 0.0001,
      "loss": 1.4272,
      "step": 3755
    },
    {
      "epoch": 0.603082851637765,
      "grad_norm": 0.27030253410339355,
      "learning_rate": 0.0001,
      "loss": 1.4415,
      "step": 3756
    },
    {
      "epoch": 0.6032434168272318,
      "grad_norm": 0.26890572905540466,
      "learning_rate": 0.0001,
      "loss": 1.5336,
      "step": 3757
    },
    {
      "epoch": 0.6034039820166988,
      "grad_norm": 0.28463757038116455,
      "learning_rate": 0.0001,
      "loss": 1.4643,
      "step": 3758
    },
    {
      "epoch": 0.6035645472061657,
      "grad_norm": 0.28352299332618713,
      "learning_rate": 0.0001,
      "loss": 1.5305,
      "step": 3759
    },
    {
      "epoch": 0.6037251123956326,
      "grad_norm": 0.27299073338508606,
      "learning_rate": 0.0001,
      "loss": 1.4717,
      "step": 3760
    },
    {
      "epoch": 0.6038856775850996,
      "grad_norm": 0.28827208280563354,
      "learning_rate": 0.0001,
      "loss": 1.4754,
      "step": 3761
    },
    {
      "epoch": 0.6040462427745664,
      "grad_norm": 0.2631355822086334,
      "learning_rate": 0.0001,
      "loss": 1.5013,
      "step": 3762
    },
    {
      "epoch": 0.6042068079640334,
      "grad_norm": 0.2724611163139343,
      "learning_rate": 0.0001,
      "loss": 1.514,
      "step": 3763
    },
    {
      "epoch": 0.6043673731535003,
      "grad_norm": 0.26698440313339233,
      "learning_rate": 0.0001,
      "loss": 1.5561,
      "step": 3764
    },
    {
      "epoch": 0.6045279383429673,
      "grad_norm": 0.2523908019065857,
      "learning_rate": 0.0001,
      "loss": 1.5108,
      "step": 3765
    },
    {
      "epoch": 0.6046885035324342,
      "grad_norm": 0.28086769580841064,
      "learning_rate": 0.0001,
      "loss": 1.554,
      "step": 3766
    },
    {
      "epoch": 0.6048490687219011,
      "grad_norm": 0.255831778049469,
      "learning_rate": 0.0001,
      "loss": 1.4552,
      "step": 3767
    },
    {
      "epoch": 0.605009633911368,
      "grad_norm": 0.24505458772182465,
      "learning_rate": 0.0001,
      "loss": 1.5219,
      "step": 3768
    },
    {
      "epoch": 0.6051701991008349,
      "grad_norm": 0.2664952278137207,
      "learning_rate": 0.0001,
      "loss": 1.5656,
      "step": 3769
    },
    {
      "epoch": 0.6053307642903019,
      "grad_norm": 0.24774408340454102,
      "learning_rate": 0.0001,
      "loss": 1.5415,
      "step": 3770
    },
    {
      "epoch": 0.6054913294797688,
      "grad_norm": 0.2702226936817169,
      "learning_rate": 0.0001,
      "loss": 1.4686,
      "step": 3771
    },
    {
      "epoch": 0.6056518946692357,
      "grad_norm": 0.25096145272254944,
      "learning_rate": 0.0001,
      "loss": 1.4664,
      "step": 3772
    },
    {
      "epoch": 0.6058124598587026,
      "grad_norm": 0.25611796975135803,
      "learning_rate": 0.0001,
      "loss": 1.5748,
      "step": 3773
    },
    {
      "epoch": 0.6059730250481695,
      "grad_norm": 0.2698650062084198,
      "learning_rate": 0.0001,
      "loss": 1.5064,
      "step": 3774
    },
    {
      "epoch": 0.6061335902376365,
      "grad_norm": 0.26493504643440247,
      "learning_rate": 0.0001,
      "loss": 1.5543,
      "step": 3775
    },
    {
      "epoch": 0.6062941554271034,
      "grad_norm": 0.2588040232658386,
      "learning_rate": 0.0001,
      "loss": 1.5167,
      "step": 3776
    },
    {
      "epoch": 0.6064547206165704,
      "grad_norm": 0.27098220586776733,
      "learning_rate": 0.0001,
      "loss": 1.456,
      "step": 3777
    },
    {
      "epoch": 0.6066152858060373,
      "grad_norm": 0.26216307282447815,
      "learning_rate": 0.0001,
      "loss": 1.454,
      "step": 3778
    },
    {
      "epoch": 0.6067758509955041,
      "grad_norm": 0.2592635452747345,
      "learning_rate": 0.0001,
      "loss": 1.4926,
      "step": 3779
    },
    {
      "epoch": 0.6069364161849711,
      "grad_norm": 0.2716052830219269,
      "learning_rate": 0.0001,
      "loss": 1.6326,
      "step": 3780
    },
    {
      "epoch": 0.607096981374438,
      "grad_norm": 0.2565079629421234,
      "learning_rate": 0.0001,
      "loss": 1.444,
      "step": 3781
    },
    {
      "epoch": 0.607257546563905,
      "grad_norm": 0.2609427869319916,
      "learning_rate": 0.0001,
      "loss": 1.5621,
      "step": 3782
    },
    {
      "epoch": 0.6074181117533719,
      "grad_norm": 0.2507987916469574,
      "learning_rate": 0.0001,
      "loss": 1.3649,
      "step": 3783
    },
    {
      "epoch": 0.6075786769428388,
      "grad_norm": 0.23893386125564575,
      "learning_rate": 0.0001,
      "loss": 1.4768,
      "step": 3784
    },
    {
      "epoch": 0.6077392421323057,
      "grad_norm": 0.2619777321815491,
      "learning_rate": 0.0001,
      "loss": 1.5338,
      "step": 3785
    },
    {
      "epoch": 0.6078998073217726,
      "grad_norm": 0.26731613278388977,
      "learning_rate": 0.0001,
      "loss": 1.5302,
      "step": 3786
    },
    {
      "epoch": 0.6080603725112396,
      "grad_norm": 0.25718340277671814,
      "learning_rate": 0.0001,
      "loss": 1.4965,
      "step": 3787
    },
    {
      "epoch": 0.6082209377007065,
      "grad_norm": 0.25129830837249756,
      "learning_rate": 0.0001,
      "loss": 1.4393,
      "step": 3788
    },
    {
      "epoch": 0.6083815028901735,
      "grad_norm": 0.27007824182510376,
      "learning_rate": 0.0001,
      "loss": 1.5635,
      "step": 3789
    },
    {
      "epoch": 0.6085420680796403,
      "grad_norm": 0.26350027322769165,
      "learning_rate": 0.0001,
      "loss": 1.4737,
      "step": 3790
    },
    {
      "epoch": 0.6087026332691072,
      "grad_norm": 0.2519645690917969,
      "learning_rate": 0.0001,
      "loss": 1.4655,
      "step": 3791
    },
    {
      "epoch": 0.6088631984585742,
      "grad_norm": 0.25365468859672546,
      "learning_rate": 0.0001,
      "loss": 1.4729,
      "step": 3792
    },
    {
      "epoch": 0.6090237636480411,
      "grad_norm": 0.26774725317955017,
      "learning_rate": 0.0001,
      "loss": 1.6337,
      "step": 3793
    },
    {
      "epoch": 0.6091843288375081,
      "grad_norm": 0.25014322996139526,
      "learning_rate": 0.0001,
      "loss": 1.4466,
      "step": 3794
    },
    {
      "epoch": 0.609344894026975,
      "grad_norm": 0.26495805382728577,
      "learning_rate": 0.0001,
      "loss": 1.5456,
      "step": 3795
    },
    {
      "epoch": 0.6095054592164418,
      "grad_norm": 0.2657485604286194,
      "learning_rate": 0.0001,
      "loss": 1.5112,
      "step": 3796
    },
    {
      "epoch": 0.6096660244059088,
      "grad_norm": 0.25781241059303284,
      "learning_rate": 0.0001,
      "loss": 1.5289,
      "step": 3797
    },
    {
      "epoch": 0.6098265895953757,
      "grad_norm": 0.2627398371696472,
      "learning_rate": 0.0001,
      "loss": 1.542,
      "step": 3798
    },
    {
      "epoch": 0.6099871547848427,
      "grad_norm": 0.26378151774406433,
      "learning_rate": 0.0001,
      "loss": 1.3881,
      "step": 3799
    },
    {
      "epoch": 0.6101477199743096,
      "grad_norm": 0.26991933584213257,
      "learning_rate": 0.0001,
      "loss": 1.5015,
      "step": 3800
    },
    {
      "epoch": 0.6103082851637764,
      "grad_norm": 0.2511008083820343,
      "learning_rate": 0.0001,
      "loss": 1.4836,
      "step": 3801
    },
    {
      "epoch": 0.6104688503532434,
      "grad_norm": 0.2760944068431854,
      "learning_rate": 0.0001,
      "loss": 1.4911,
      "step": 3802
    },
    {
      "epoch": 0.6106294155427103,
      "grad_norm": 0.2546599507331848,
      "learning_rate": 0.0001,
      "loss": 1.4023,
      "step": 3803
    },
    {
      "epoch": 0.6107899807321773,
      "grad_norm": 0.2641827464103699,
      "learning_rate": 0.0001,
      "loss": 1.5429,
      "step": 3804
    },
    {
      "epoch": 0.6109505459216442,
      "grad_norm": 0.30749645829200745,
      "learning_rate": 0.0001,
      "loss": 1.4374,
      "step": 3805
    },
    {
      "epoch": 0.6111111111111112,
      "grad_norm": 0.2694147527217865,
      "learning_rate": 0.0001,
      "loss": 1.5099,
      "step": 3806
    },
    {
      "epoch": 0.611271676300578,
      "grad_norm": 0.26679447293281555,
      "learning_rate": 0.0001,
      "loss": 1.5684,
      "step": 3807
    },
    {
      "epoch": 0.6114322414900449,
      "grad_norm": 0.2818480134010315,
      "learning_rate": 0.0001,
      "loss": 1.497,
      "step": 3808
    },
    {
      "epoch": 0.6115928066795119,
      "grad_norm": 0.28491002321243286,
      "learning_rate": 0.0001,
      "loss": 1.5879,
      "step": 3809
    },
    {
      "epoch": 0.6117533718689788,
      "grad_norm": 0.25981539487838745,
      "learning_rate": 0.0001,
      "loss": 1.5756,
      "step": 3810
    },
    {
      "epoch": 0.6119139370584458,
      "grad_norm": 0.26610076427459717,
      "learning_rate": 0.0001,
      "loss": 1.5041,
      "step": 3811
    },
    {
      "epoch": 0.6120745022479126,
      "grad_norm": 0.25697603821754456,
      "learning_rate": 0.0001,
      "loss": 1.428,
      "step": 3812
    },
    {
      "epoch": 0.6122350674373795,
      "grad_norm": 0.26291266083717346,
      "learning_rate": 0.0001,
      "loss": 1.5169,
      "step": 3813
    },
    {
      "epoch": 0.6123956326268465,
      "grad_norm": 0.26881441473960876,
      "learning_rate": 0.0001,
      "loss": 1.6022,
      "step": 3814
    },
    {
      "epoch": 0.6125561978163134,
      "grad_norm": 0.25986117124557495,
      "learning_rate": 0.0001,
      "loss": 1.5053,
      "step": 3815
    },
    {
      "epoch": 0.6127167630057804,
      "grad_norm": 0.2613845467567444,
      "learning_rate": 0.0001,
      "loss": 1.5652,
      "step": 3816
    },
    {
      "epoch": 0.6128773281952473,
      "grad_norm": 0.24931935966014862,
      "learning_rate": 0.0001,
      "loss": 1.4748,
      "step": 3817
    },
    {
      "epoch": 0.6130378933847141,
      "grad_norm": 0.2685846984386444,
      "learning_rate": 0.0001,
      "loss": 1.5155,
      "step": 3818
    },
    {
      "epoch": 0.6131984585741811,
      "grad_norm": 0.3231850564479828,
      "learning_rate": 0.0001,
      "loss": 1.5149,
      "step": 3819
    },
    {
      "epoch": 0.613359023763648,
      "grad_norm": 0.24862191081047058,
      "learning_rate": 0.0001,
      "loss": 1.4452,
      "step": 3820
    },
    {
      "epoch": 0.613519588953115,
      "grad_norm": 0.26581764221191406,
      "learning_rate": 0.0001,
      "loss": 1.4571,
      "step": 3821
    },
    {
      "epoch": 0.6136801541425819,
      "grad_norm": 0.27062392234802246,
      "learning_rate": 0.0001,
      "loss": 1.475,
      "step": 3822
    },
    {
      "epoch": 0.6138407193320488,
      "grad_norm": 0.27251020073890686,
      "learning_rate": 0.0001,
      "loss": 1.5969,
      "step": 3823
    },
    {
      "epoch": 0.6140012845215157,
      "grad_norm": 0.262570321559906,
      "learning_rate": 0.0001,
      "loss": 1.4885,
      "step": 3824
    },
    {
      "epoch": 0.6141618497109826,
      "grad_norm": 0.262735515832901,
      "learning_rate": 0.0001,
      "loss": 1.4949,
      "step": 3825
    },
    {
      "epoch": 0.6143224149004496,
      "grad_norm": 0.26099810004234314,
      "learning_rate": 0.0001,
      "loss": 1.567,
      "step": 3826
    },
    {
      "epoch": 0.6144829800899165,
      "grad_norm": 0.25535207986831665,
      "learning_rate": 0.0001,
      "loss": 1.5153,
      "step": 3827
    },
    {
      "epoch": 0.6146435452793835,
      "grad_norm": 0.2533794045448303,
      "learning_rate": 0.0001,
      "loss": 1.4156,
      "step": 3828
    },
    {
      "epoch": 0.6148041104688503,
      "grad_norm": 0.26601144671440125,
      "learning_rate": 0.0001,
      "loss": 1.488,
      "step": 3829
    },
    {
      "epoch": 0.6149646756583173,
      "grad_norm": 0.2557826340198517,
      "learning_rate": 0.0001,
      "loss": 1.4828,
      "step": 3830
    },
    {
      "epoch": 0.6151252408477842,
      "grad_norm": 0.28637704253196716,
      "learning_rate": 0.0001,
      "loss": 1.5266,
      "step": 3831
    },
    {
      "epoch": 0.6152858060372511,
      "grad_norm": 0.2602996826171875,
      "learning_rate": 0.0001,
      "loss": 1.49,
      "step": 3832
    },
    {
      "epoch": 0.6154463712267181,
      "grad_norm": 0.2608982026576996,
      "learning_rate": 0.0001,
      "loss": 1.5477,
      "step": 3833
    },
    {
      "epoch": 0.615606936416185,
      "grad_norm": 0.2602182626724243,
      "learning_rate": 0.0001,
      "loss": 1.5409,
      "step": 3834
    },
    {
      "epoch": 0.6157675016056519,
      "grad_norm": 0.2599102258682251,
      "learning_rate": 0.0001,
      "loss": 1.5139,
      "step": 3835
    },
    {
      "epoch": 0.6159280667951188,
      "grad_norm": 0.24739331007003784,
      "learning_rate": 0.0001,
      "loss": 1.4262,
      "step": 3836
    },
    {
      "epoch": 0.6160886319845857,
      "grad_norm": 0.25765302777290344,
      "learning_rate": 0.0001,
      "loss": 1.5392,
      "step": 3837
    },
    {
      "epoch": 0.6162491971740527,
      "grad_norm": 0.27270618081092834,
      "learning_rate": 0.0001,
      "loss": 1.5475,
      "step": 3838
    },
    {
      "epoch": 0.6164097623635196,
      "grad_norm": 0.25871002674102783,
      "learning_rate": 0.0001,
      "loss": 1.4366,
      "step": 3839
    },
    {
      "epoch": 0.6165703275529865,
      "grad_norm": 0.24694572389125824,
      "learning_rate": 0.0001,
      "loss": 1.4313,
      "step": 3840
    },
    {
      "epoch": 0.6167308927424534,
      "grad_norm": 0.27282777428627014,
      "learning_rate": 0.0001,
      "loss": 1.5298,
      "step": 3841
    },
    {
      "epoch": 0.6168914579319204,
      "grad_norm": 0.25624415278434753,
      "learning_rate": 0.0001,
      "loss": 1.4922,
      "step": 3842
    },
    {
      "epoch": 0.6170520231213873,
      "grad_norm": 0.25390180945396423,
      "learning_rate": 0.0001,
      "loss": 1.5232,
      "step": 3843
    },
    {
      "epoch": 0.6172125883108542,
      "grad_norm": 0.26919737458229065,
      "learning_rate": 0.0001,
      "loss": 1.444,
      "step": 3844
    },
    {
      "epoch": 0.6173731535003212,
      "grad_norm": 0.2601894438266754,
      "learning_rate": 0.0001,
      "loss": 1.5319,
      "step": 3845
    },
    {
      "epoch": 0.617533718689788,
      "grad_norm": 0.25561174750328064,
      "learning_rate": 0.0001,
      "loss": 1.5097,
      "step": 3846
    },
    {
      "epoch": 0.617694283879255,
      "grad_norm": 0.24631574749946594,
      "learning_rate": 0.0001,
      "loss": 1.4197,
      "step": 3847
    },
    {
      "epoch": 0.6178548490687219,
      "grad_norm": 0.25146305561065674,
      "learning_rate": 0.0001,
      "loss": 1.5236,
      "step": 3848
    },
    {
      "epoch": 0.6180154142581888,
      "grad_norm": 0.2760447561740875,
      "learning_rate": 0.0001,
      "loss": 1.5684,
      "step": 3849
    },
    {
      "epoch": 0.6181759794476558,
      "grad_norm": 0.25651994347572327,
      "learning_rate": 0.0001,
      "loss": 1.5364,
      "step": 3850
    },
    {
      "epoch": 0.6183365446371226,
      "grad_norm": 0.2498057633638382,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 3851
    },
    {
      "epoch": 0.6184971098265896,
      "grad_norm": 0.2638412415981293,
      "learning_rate": 0.0001,
      "loss": 1.5045,
      "step": 3852
    },
    {
      "epoch": 0.6186576750160565,
      "grad_norm": 0.24950148165225983,
      "learning_rate": 0.0001,
      "loss": 1.4867,
      "step": 3853
    },
    {
      "epoch": 0.6188182402055235,
      "grad_norm": 0.24246224761009216,
      "learning_rate": 0.0001,
      "loss": 1.4704,
      "step": 3854
    },
    {
      "epoch": 0.6189788053949904,
      "grad_norm": 0.24784599244594574,
      "learning_rate": 0.0001,
      "loss": 1.4354,
      "step": 3855
    },
    {
      "epoch": 0.6191393705844573,
      "grad_norm": 0.2517155110836029,
      "learning_rate": 0.0001,
      "loss": 1.5132,
      "step": 3856
    },
    {
      "epoch": 0.6192999357739242,
      "grad_norm": 0.2677137553691864,
      "learning_rate": 0.0001,
      "loss": 1.5903,
      "step": 3857
    },
    {
      "epoch": 0.6194605009633911,
      "grad_norm": 0.2763158082962036,
      "learning_rate": 0.0001,
      "loss": 1.5194,
      "step": 3858
    },
    {
      "epoch": 0.6196210661528581,
      "grad_norm": 0.28667938709259033,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 3859
    },
    {
      "epoch": 0.619781631342325,
      "grad_norm": 0.2558166980743408,
      "learning_rate": 0.0001,
      "loss": 1.5101,
      "step": 3860
    },
    {
      "epoch": 0.619942196531792,
      "grad_norm": 0.25873079895973206,
      "learning_rate": 0.0001,
      "loss": 1.473,
      "step": 3861
    },
    {
      "epoch": 0.6201027617212588,
      "grad_norm": 0.28501585125923157,
      "learning_rate": 0.0001,
      "loss": 1.5765,
      "step": 3862
    },
    {
      "epoch": 0.6202633269107257,
      "grad_norm": 0.2709580957889557,
      "learning_rate": 0.0001,
      "loss": 1.5434,
      "step": 3863
    },
    {
      "epoch": 0.6204238921001927,
      "grad_norm": 0.31828853487968445,
      "learning_rate": 0.0001,
      "loss": 1.4774,
      "step": 3864
    },
    {
      "epoch": 0.6205844572896596,
      "grad_norm": 0.2685384154319763,
      "learning_rate": 0.0001,
      "loss": 1.5567,
      "step": 3865
    },
    {
      "epoch": 0.6207450224791266,
      "grad_norm": 0.24856871366500854,
      "learning_rate": 0.0001,
      "loss": 1.4992,
      "step": 3866
    },
    {
      "epoch": 0.6209055876685935,
      "grad_norm": 0.2948998212814331,
      "learning_rate": 0.0001,
      "loss": 1.5635,
      "step": 3867
    },
    {
      "epoch": 0.6210661528580603,
      "grad_norm": 0.27943724393844604,
      "learning_rate": 0.0001,
      "loss": 1.544,
      "step": 3868
    },
    {
      "epoch": 0.6212267180475273,
      "grad_norm": 0.2594679892063141,
      "learning_rate": 0.0001,
      "loss": 1.5091,
      "step": 3869
    },
    {
      "epoch": 0.6213872832369942,
      "grad_norm": 0.26492151618003845,
      "learning_rate": 0.0001,
      "loss": 1.5173,
      "step": 3870
    },
    {
      "epoch": 0.6215478484264612,
      "grad_norm": 0.26779136061668396,
      "learning_rate": 0.0001,
      "loss": 1.4597,
      "step": 3871
    },
    {
      "epoch": 0.6217084136159281,
      "grad_norm": 0.2853734791278839,
      "learning_rate": 0.0001,
      "loss": 1.5688,
      "step": 3872
    },
    {
      "epoch": 0.621868978805395,
      "grad_norm": 0.2937188744544983,
      "learning_rate": 0.0001,
      "loss": 1.518,
      "step": 3873
    },
    {
      "epoch": 0.6220295439948619,
      "grad_norm": 0.27518853545188904,
      "learning_rate": 0.0001,
      "loss": 1.4543,
      "step": 3874
    },
    {
      "epoch": 0.6221901091843288,
      "grad_norm": 0.2555117905139923,
      "learning_rate": 0.0001,
      "loss": 1.4583,
      "step": 3875
    },
    {
      "epoch": 0.6223506743737958,
      "grad_norm": 0.2831689119338989,
      "learning_rate": 0.0001,
      "loss": 1.4717,
      "step": 3876
    },
    {
      "epoch": 0.6225112395632627,
      "grad_norm": 0.27565649151802063,
      "learning_rate": 0.0001,
      "loss": 1.5472,
      "step": 3877
    },
    {
      "epoch": 0.6226718047527297,
      "grad_norm": 0.28797581791877747,
      "learning_rate": 0.0001,
      "loss": 1.4364,
      "step": 3878
    },
    {
      "epoch": 0.6228323699421965,
      "grad_norm": 0.255044162273407,
      "learning_rate": 0.0001,
      "loss": 1.4476,
      "step": 3879
    },
    {
      "epoch": 0.6229929351316634,
      "grad_norm": 0.27069318294525146,
      "learning_rate": 0.0001,
      "loss": 1.5265,
      "step": 3880
    },
    {
      "epoch": 0.6231535003211304,
      "grad_norm": 0.26840701699256897,
      "learning_rate": 0.0001,
      "loss": 1.4822,
      "step": 3881
    },
    {
      "epoch": 0.6233140655105973,
      "grad_norm": 0.2857365012168884,
      "learning_rate": 0.0001,
      "loss": 1.5621,
      "step": 3882
    },
    {
      "epoch": 0.6234746307000643,
      "grad_norm": 0.26349303126335144,
      "learning_rate": 0.0001,
      "loss": 1.4404,
      "step": 3883
    },
    {
      "epoch": 0.6236351958895312,
      "grad_norm": 0.25878918170928955,
      "learning_rate": 0.0001,
      "loss": 1.4413,
      "step": 3884
    },
    {
      "epoch": 0.623795761078998,
      "grad_norm": 0.2577318847179413,
      "learning_rate": 0.0001,
      "loss": 1.4586,
      "step": 3885
    },
    {
      "epoch": 0.623956326268465,
      "grad_norm": 0.2663488984107971,
      "learning_rate": 0.0001,
      "loss": 1.4783,
      "step": 3886
    },
    {
      "epoch": 0.6241168914579319,
      "grad_norm": 0.38730761408805847,
      "learning_rate": 0.0001,
      "loss": 1.5137,
      "step": 3887
    },
    {
      "epoch": 0.6242774566473989,
      "grad_norm": 0.2595405578613281,
      "learning_rate": 0.0001,
      "loss": 1.5058,
      "step": 3888
    },
    {
      "epoch": 0.6244380218368658,
      "grad_norm": 0.25100287795066833,
      "learning_rate": 0.0001,
      "loss": 1.5188,
      "step": 3889
    },
    {
      "epoch": 0.6245985870263326,
      "grad_norm": 0.39545443654060364,
      "learning_rate": 0.0001,
      "loss": 1.5598,
      "step": 3890
    },
    {
      "epoch": 0.6247591522157996,
      "grad_norm": 0.2625770568847656,
      "learning_rate": 0.0001,
      "loss": 1.4418,
      "step": 3891
    },
    {
      "epoch": 0.6249197174052665,
      "grad_norm": 0.25303763151168823,
      "learning_rate": 0.0001,
      "loss": 1.3447,
      "step": 3892
    },
    {
      "epoch": 0.6250802825947335,
      "grad_norm": 0.26952648162841797,
      "learning_rate": 0.0001,
      "loss": 1.5335,
      "step": 3893
    },
    {
      "epoch": 0.6252408477842004,
      "grad_norm": 0.2523767650127411,
      "learning_rate": 0.0001,
      "loss": 1.4401,
      "step": 3894
    },
    {
      "epoch": 0.6254014129736674,
      "grad_norm": 0.2689189314842224,
      "learning_rate": 0.0001,
      "loss": 1.4875,
      "step": 3895
    },
    {
      "epoch": 0.6255619781631342,
      "grad_norm": 0.2631898522377014,
      "learning_rate": 0.0001,
      "loss": 1.4369,
      "step": 3896
    },
    {
      "epoch": 0.6257225433526011,
      "grad_norm": 0.28323471546173096,
      "learning_rate": 0.0001,
      "loss": 1.5534,
      "step": 3897
    },
    {
      "epoch": 0.6258831085420681,
      "grad_norm": 0.4033949673175812,
      "learning_rate": 0.0001,
      "loss": 1.4326,
      "step": 3898
    },
    {
      "epoch": 0.626043673731535,
      "grad_norm": 0.24338938295841217,
      "learning_rate": 0.0001,
      "loss": 1.3999,
      "step": 3899
    },
    {
      "epoch": 0.626204238921002,
      "grad_norm": 0.2539018988609314,
      "learning_rate": 0.0001,
      "loss": 1.5075,
      "step": 3900
    },
    {
      "epoch": 0.6263648041104688,
      "grad_norm": 0.2713395357131958,
      "learning_rate": 0.0001,
      "loss": 1.493,
      "step": 3901
    },
    {
      "epoch": 0.6265253692999357,
      "grad_norm": 0.26039496064186096,
      "learning_rate": 0.0001,
      "loss": 1.5543,
      "step": 3902
    },
    {
      "epoch": 0.6266859344894027,
      "grad_norm": 0.25155508518218994,
      "learning_rate": 0.0001,
      "loss": 1.4714,
      "step": 3903
    },
    {
      "epoch": 0.6268464996788696,
      "grad_norm": 0.26837146282196045,
      "learning_rate": 0.0001,
      "loss": 1.5416,
      "step": 3904
    },
    {
      "epoch": 0.6270070648683366,
      "grad_norm": 0.26188722252845764,
      "learning_rate": 0.0001,
      "loss": 1.5835,
      "step": 3905
    },
    {
      "epoch": 0.6271676300578035,
      "grad_norm": 0.27130991220474243,
      "learning_rate": 0.0001,
      "loss": 1.525,
      "step": 3906
    },
    {
      "epoch": 0.6273281952472703,
      "grad_norm": 0.25837934017181396,
      "learning_rate": 0.0001,
      "loss": 1.4969,
      "step": 3907
    },
    {
      "epoch": 0.6274887604367373,
      "grad_norm": 0.26127636432647705,
      "learning_rate": 0.0001,
      "loss": 1.6004,
      "step": 3908
    },
    {
      "epoch": 0.6276493256262042,
      "grad_norm": 0.2639945447444916,
      "learning_rate": 0.0001,
      "loss": 1.4771,
      "step": 3909
    },
    {
      "epoch": 0.6278098908156712,
      "grad_norm": 0.2688761353492737,
      "learning_rate": 0.0001,
      "loss": 1.4476,
      "step": 3910
    },
    {
      "epoch": 0.6279704560051381,
      "grad_norm": 0.25851279497146606,
      "learning_rate": 0.0001,
      "loss": 1.4896,
      "step": 3911
    },
    {
      "epoch": 0.628131021194605,
      "grad_norm": 0.2678278684616089,
      "learning_rate": 0.0001,
      "loss": 1.4605,
      "step": 3912
    },
    {
      "epoch": 0.6282915863840719,
      "grad_norm": 0.2609288692474365,
      "learning_rate": 0.0001,
      "loss": 1.4632,
      "step": 3913
    },
    {
      "epoch": 0.6284521515735388,
      "grad_norm": 0.24951615929603577,
      "learning_rate": 0.0001,
      "loss": 1.4165,
      "step": 3914
    },
    {
      "epoch": 0.6286127167630058,
      "grad_norm": 0.2530744671821594,
      "learning_rate": 0.0001,
      "loss": 1.5278,
      "step": 3915
    },
    {
      "epoch": 0.6287732819524727,
      "grad_norm": 0.2548890709877014,
      "learning_rate": 0.0001,
      "loss": 1.5422,
      "step": 3916
    },
    {
      "epoch": 0.6289338471419397,
      "grad_norm": 0.2568867802619934,
      "learning_rate": 0.0001,
      "loss": 1.4971,
      "step": 3917
    },
    {
      "epoch": 0.6290944123314065,
      "grad_norm": 0.26970985531806946,
      "learning_rate": 0.0001,
      "loss": 1.5194,
      "step": 3918
    },
    {
      "epoch": 0.6292549775208734,
      "grad_norm": 0.2785082161426544,
      "learning_rate": 0.0001,
      "loss": 1.4597,
      "step": 3919
    },
    {
      "epoch": 0.6294155427103404,
      "grad_norm": 0.2592335343360901,
      "learning_rate": 0.0001,
      "loss": 1.5084,
      "step": 3920
    },
    {
      "epoch": 0.6295761078998073,
      "grad_norm": 0.29501408338546753,
      "learning_rate": 0.0001,
      "loss": 1.526,
      "step": 3921
    },
    {
      "epoch": 0.6297366730892743,
      "grad_norm": 0.27205076813697815,
      "learning_rate": 0.0001,
      "loss": 1.5303,
      "step": 3922
    },
    {
      "epoch": 0.6298972382787412,
      "grad_norm": 0.27818426489830017,
      "learning_rate": 0.0001,
      "loss": 1.4993,
      "step": 3923
    },
    {
      "epoch": 0.630057803468208,
      "grad_norm": 0.25199609994888306,
      "learning_rate": 0.0001,
      "loss": 1.4921,
      "step": 3924
    },
    {
      "epoch": 0.630218368657675,
      "grad_norm": 0.259134441614151,
      "learning_rate": 0.0001,
      "loss": 1.5125,
      "step": 3925
    },
    {
      "epoch": 0.6303789338471419,
      "grad_norm": 0.2574685215950012,
      "learning_rate": 0.0001,
      "loss": 1.4471,
      "step": 3926
    },
    {
      "epoch": 0.6305394990366089,
      "grad_norm": 0.2670515477657318,
      "learning_rate": 0.0001,
      "loss": 1.5431,
      "step": 3927
    },
    {
      "epoch": 0.6307000642260758,
      "grad_norm": 0.2595072388648987,
      "learning_rate": 0.0001,
      "loss": 1.4905,
      "step": 3928
    },
    {
      "epoch": 0.6308606294155427,
      "grad_norm": 0.2716316282749176,
      "learning_rate": 0.0001,
      "loss": 1.4877,
      "step": 3929
    },
    {
      "epoch": 0.6310211946050096,
      "grad_norm": 0.26863327622413635,
      "learning_rate": 0.0001,
      "loss": 1.4627,
      "step": 3930
    },
    {
      "epoch": 0.6311817597944765,
      "grad_norm": 0.25725260376930237,
      "learning_rate": 0.0001,
      "loss": 1.4557,
      "step": 3931
    },
    {
      "epoch": 0.6313423249839435,
      "grad_norm": 0.2594561278820038,
      "learning_rate": 0.0001,
      "loss": 1.5729,
      "step": 3932
    },
    {
      "epoch": 0.6315028901734104,
      "grad_norm": 0.2631740868091583,
      "learning_rate": 0.0001,
      "loss": 1.4892,
      "step": 3933
    },
    {
      "epoch": 0.6316634553628774,
      "grad_norm": 0.270300030708313,
      "learning_rate": 0.0001,
      "loss": 1.5202,
      "step": 3934
    },
    {
      "epoch": 0.6318240205523442,
      "grad_norm": 0.2681209146976471,
      "learning_rate": 0.0001,
      "loss": 1.5378,
      "step": 3935
    },
    {
      "epoch": 0.6319845857418112,
      "grad_norm": 0.2748728394508362,
      "learning_rate": 0.0001,
      "loss": 1.6086,
      "step": 3936
    },
    {
      "epoch": 0.6321451509312781,
      "grad_norm": 0.28140565752983093,
      "learning_rate": 0.0001,
      "loss": 1.532,
      "step": 3937
    },
    {
      "epoch": 0.632305716120745,
      "grad_norm": 0.25414136052131653,
      "learning_rate": 0.0001,
      "loss": 1.5435,
      "step": 3938
    },
    {
      "epoch": 0.632466281310212,
      "grad_norm": 0.535088300704956,
      "learning_rate": 0.0001,
      "loss": 1.5236,
      "step": 3939
    },
    {
      "epoch": 0.6326268464996788,
      "grad_norm": 0.25747591257095337,
      "learning_rate": 0.0001,
      "loss": 1.406,
      "step": 3940
    },
    {
      "epoch": 0.6327874116891458,
      "grad_norm": 0.2462492734193802,
      "learning_rate": 0.0001,
      "loss": 1.4842,
      "step": 3941
    },
    {
      "epoch": 0.6329479768786127,
      "grad_norm": 0.2716800570487976,
      "learning_rate": 0.0001,
      "loss": 1.465,
      "step": 3942
    },
    {
      "epoch": 0.6331085420680796,
      "grad_norm": 0.26844802498817444,
      "learning_rate": 0.0001,
      "loss": 1.5626,
      "step": 3943
    },
    {
      "epoch": 0.6332691072575466,
      "grad_norm": 0.25362297892570496,
      "learning_rate": 0.0001,
      "loss": 1.534,
      "step": 3944
    },
    {
      "epoch": 0.6334296724470135,
      "grad_norm": 0.2911030650138855,
      "learning_rate": 0.0001,
      "loss": 1.5393,
      "step": 3945
    },
    {
      "epoch": 0.6335902376364804,
      "grad_norm": 0.24949724972248077,
      "learning_rate": 0.0001,
      "loss": 1.5277,
      "step": 3946
    },
    {
      "epoch": 0.6337508028259473,
      "grad_norm": 0.25194117426872253,
      "learning_rate": 0.0001,
      "loss": 1.5009,
      "step": 3947
    },
    {
      "epoch": 0.6339113680154143,
      "grad_norm": 0.25425615906715393,
      "learning_rate": 0.0001,
      "loss": 1.5279,
      "step": 3948
    },
    {
      "epoch": 0.6340719332048812,
      "grad_norm": 0.26466798782348633,
      "learning_rate": 0.0001,
      "loss": 1.5686,
      "step": 3949
    },
    {
      "epoch": 0.6342324983943481,
      "grad_norm": 0.3454229533672333,
      "learning_rate": 0.0001,
      "loss": 1.5514,
      "step": 3950
    },
    {
      "epoch": 0.634393063583815,
      "grad_norm": 0.24445679783821106,
      "learning_rate": 0.0001,
      "loss": 1.455,
      "step": 3951
    },
    {
      "epoch": 0.6345536287732819,
      "grad_norm": 0.27483898401260376,
      "learning_rate": 0.0001,
      "loss": 1.511,
      "step": 3952
    },
    {
      "epoch": 0.6347141939627489,
      "grad_norm": 0.25458669662475586,
      "learning_rate": 0.0001,
      "loss": 1.5065,
      "step": 3953
    },
    {
      "epoch": 0.6348747591522158,
      "grad_norm": 0.258373886346817,
      "learning_rate": 0.0001,
      "loss": 1.5748,
      "step": 3954
    },
    {
      "epoch": 0.6350353243416827,
      "grad_norm": 0.2767835557460785,
      "learning_rate": 0.0001,
      "loss": 1.5348,
      "step": 3955
    },
    {
      "epoch": 0.6351958895311497,
      "grad_norm": 0.25013065338134766,
      "learning_rate": 0.0001,
      "loss": 1.5059,
      "step": 3956
    },
    {
      "epoch": 0.6353564547206165,
      "grad_norm": 0.2646697461605072,
      "learning_rate": 0.0001,
      "loss": 1.5392,
      "step": 3957
    },
    {
      "epoch": 0.6355170199100835,
      "grad_norm": 0.280038058757782,
      "learning_rate": 0.0001,
      "loss": 1.5206,
      "step": 3958
    },
    {
      "epoch": 0.6356775850995504,
      "grad_norm": 0.2392078936100006,
      "learning_rate": 0.0001,
      "loss": 1.3745,
      "step": 3959
    },
    {
      "epoch": 0.6358381502890174,
      "grad_norm": 0.27248793840408325,
      "learning_rate": 0.0001,
      "loss": 1.4943,
      "step": 3960
    },
    {
      "epoch": 0.6359987154784843,
      "grad_norm": 0.2702663242816925,
      "learning_rate": 0.0001,
      "loss": 1.4945,
      "step": 3961
    },
    {
      "epoch": 0.6361592806679512,
      "grad_norm": 0.26339977979660034,
      "learning_rate": 0.0001,
      "loss": 1.5032,
      "step": 3962
    },
    {
      "epoch": 0.6363198458574181,
      "grad_norm": 0.27651360630989075,
      "learning_rate": 0.0001,
      "loss": 1.5513,
      "step": 3963
    },
    {
      "epoch": 0.636480411046885,
      "grad_norm": 0.25699666142463684,
      "learning_rate": 0.0001,
      "loss": 1.368,
      "step": 3964
    },
    {
      "epoch": 0.636640976236352,
      "grad_norm": 0.2467186003923416,
      "learning_rate": 0.0001,
      "loss": 1.5225,
      "step": 3965
    },
    {
      "epoch": 0.6368015414258189,
      "grad_norm": 0.26869550347328186,
      "learning_rate": 0.0001,
      "loss": 1.4593,
      "step": 3966
    },
    {
      "epoch": 0.6369621066152859,
      "grad_norm": 0.2678242027759552,
      "learning_rate": 0.0001,
      "loss": 1.5605,
      "step": 3967
    },
    {
      "epoch": 0.6371226718047527,
      "grad_norm": 0.2536470890045166,
      "learning_rate": 0.0001,
      "loss": 1.4875,
      "step": 3968
    },
    {
      "epoch": 0.6372832369942196,
      "grad_norm": 0.25990384817123413,
      "learning_rate": 0.0001,
      "loss": 1.4988,
      "step": 3969
    },
    {
      "epoch": 0.6374438021836866,
      "grad_norm": 0.24920576810836792,
      "learning_rate": 0.0001,
      "loss": 1.4493,
      "step": 3970
    },
    {
      "epoch": 0.6376043673731535,
      "grad_norm": 0.25677353143692017,
      "learning_rate": 0.0001,
      "loss": 1.5054,
      "step": 3971
    },
    {
      "epoch": 0.6377649325626205,
      "grad_norm": 0.2507385015487671,
      "learning_rate": 0.0001,
      "loss": 1.492,
      "step": 3972
    },
    {
      "epoch": 0.6379254977520874,
      "grad_norm": 0.2684652805328369,
      "learning_rate": 0.0001,
      "loss": 1.544,
      "step": 3973
    },
    {
      "epoch": 0.6380860629415542,
      "grad_norm": 0.24495548009872437,
      "learning_rate": 0.0001,
      "loss": 1.5038,
      "step": 3974
    },
    {
      "epoch": 0.6382466281310212,
      "grad_norm": 0.2707032561302185,
      "learning_rate": 0.0001,
      "loss": 1.5115,
      "step": 3975
    },
    {
      "epoch": 0.6384071933204881,
      "grad_norm": 0.27481770515441895,
      "learning_rate": 0.0001,
      "loss": 1.5197,
      "step": 3976
    },
    {
      "epoch": 0.6385677585099551,
      "grad_norm": 0.26256996393203735,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 3977
    },
    {
      "epoch": 0.638728323699422,
      "grad_norm": 0.28677448630332947,
      "learning_rate": 0.0001,
      "loss": 1.5337,
      "step": 3978
    },
    {
      "epoch": 0.6388888888888888,
      "grad_norm": 0.2571727931499481,
      "learning_rate": 0.0001,
      "loss": 1.4849,
      "step": 3979
    },
    {
      "epoch": 0.6390494540783558,
      "grad_norm": 0.25264930725097656,
      "learning_rate": 0.0001,
      "loss": 1.4524,
      "step": 3980
    },
    {
      "epoch": 0.6392100192678227,
      "grad_norm": 0.2521830201148987,
      "learning_rate": 0.0001,
      "loss": 1.521,
      "step": 3981
    },
    {
      "epoch": 0.6393705844572897,
      "grad_norm": 0.27169087529182434,
      "learning_rate": 0.0001,
      "loss": 1.5291,
      "step": 3982
    },
    {
      "epoch": 0.6395311496467566,
      "grad_norm": 0.2885984480381012,
      "learning_rate": 0.0001,
      "loss": 1.4407,
      "step": 3983
    },
    {
      "epoch": 0.6396917148362236,
      "grad_norm": 0.2694244980812073,
      "learning_rate": 0.0001,
      "loss": 1.4951,
      "step": 3984
    },
    {
      "epoch": 0.6398522800256904,
      "grad_norm": 0.26871436834335327,
      "learning_rate": 0.0001,
      "loss": 1.54,
      "step": 3985
    },
    {
      "epoch": 0.6400128452151573,
      "grad_norm": 0.27839118242263794,
      "learning_rate": 0.0001,
      "loss": 1.4836,
      "step": 3986
    },
    {
      "epoch": 0.6401734104046243,
      "grad_norm": 0.2824634909629822,
      "learning_rate": 0.0001,
      "loss": 1.4963,
      "step": 3987
    },
    {
      "epoch": 0.6403339755940912,
      "grad_norm": 0.2501073479652405,
      "learning_rate": 0.0001,
      "loss": 1.5297,
      "step": 3988
    },
    {
      "epoch": 0.6404945407835582,
      "grad_norm": 0.2556857764720917,
      "learning_rate": 0.0001,
      "loss": 1.5161,
      "step": 3989
    },
    {
      "epoch": 0.640655105973025,
      "grad_norm": 0.24514375627040863,
      "learning_rate": 0.0001,
      "loss": 1.3465,
      "step": 3990
    },
    {
      "epoch": 0.6408156711624919,
      "grad_norm": 0.2614043354988098,
      "learning_rate": 0.0001,
      "loss": 1.5163,
      "step": 3991
    },
    {
      "epoch": 0.6409762363519589,
      "grad_norm": 0.27069714665412903,
      "learning_rate": 0.0001,
      "loss": 1.5546,
      "step": 3992
    },
    {
      "epoch": 0.6411368015414258,
      "grad_norm": 0.2559158205986023,
      "learning_rate": 0.0001,
      "loss": 1.43,
      "step": 3993
    },
    {
      "epoch": 0.6412973667308928,
      "grad_norm": 0.26587527990341187,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 3994
    },
    {
      "epoch": 0.6414579319203597,
      "grad_norm": 0.2705190181732178,
      "learning_rate": 0.0001,
      "loss": 1.5092,
      "step": 3995
    },
    {
      "epoch": 0.6416184971098265,
      "grad_norm": 0.2716139256954193,
      "learning_rate": 0.0001,
      "loss": 1.5611,
      "step": 3996
    },
    {
      "epoch": 0.6417790622992935,
      "grad_norm": 0.30606609582901,
      "learning_rate": 0.0001,
      "loss": 1.5154,
      "step": 3997
    },
    {
      "epoch": 0.6419396274887604,
      "grad_norm": 0.2739136219024658,
      "learning_rate": 0.0001,
      "loss": 1.5407,
      "step": 3998
    },
    {
      "epoch": 0.6421001926782274,
      "grad_norm": 0.2726215422153473,
      "learning_rate": 0.0001,
      "loss": 1.5111,
      "step": 3999
    },
    {
      "epoch": 0.6422607578676943,
      "grad_norm": 0.2681017816066742,
      "learning_rate": 0.0001,
      "loss": 1.4121,
      "step": 4000
    },
    {
      "epoch": 0.6424213230571612,
      "grad_norm": 0.2679270803928375,
      "learning_rate": 0.0001,
      "loss": 1.5198,
      "step": 4001
    },
    {
      "epoch": 0.6425818882466281,
      "grad_norm": 0.2869030237197876,
      "learning_rate": 0.0001,
      "loss": 1.512,
      "step": 4002
    },
    {
      "epoch": 0.642742453436095,
      "grad_norm": 0.2672863304615021,
      "learning_rate": 0.0001,
      "loss": 1.5422,
      "step": 4003
    },
    {
      "epoch": 0.642903018625562,
      "grad_norm": 0.2606004476547241,
      "learning_rate": 0.0001,
      "loss": 1.4536,
      "step": 4004
    },
    {
      "epoch": 0.6430635838150289,
      "grad_norm": 0.26094046235084534,
      "learning_rate": 0.0001,
      "loss": 1.424,
      "step": 4005
    },
    {
      "epoch": 0.6432241490044959,
      "grad_norm": 0.2781563401222229,
      "learning_rate": 0.0001,
      "loss": 1.4617,
      "step": 4006
    },
    {
      "epoch": 0.6433847141939627,
      "grad_norm": 0.25793081521987915,
      "learning_rate": 0.0001,
      "loss": 1.4748,
      "step": 4007
    },
    {
      "epoch": 0.6435452793834296,
      "grad_norm": 0.2908867299556732,
      "learning_rate": 0.0001,
      "loss": 1.5213,
      "step": 4008
    },
    {
      "epoch": 0.6437058445728966,
      "grad_norm": 0.26451578736305237,
      "learning_rate": 0.0001,
      "loss": 1.5057,
      "step": 4009
    },
    {
      "epoch": 0.6438664097623635,
      "grad_norm": 0.2684488594532013,
      "learning_rate": 0.0001,
      "loss": 1.4973,
      "step": 4010
    },
    {
      "epoch": 0.6440269749518305,
      "grad_norm": 0.2536855638027191,
      "learning_rate": 0.0001,
      "loss": 1.4982,
      "step": 4011
    },
    {
      "epoch": 0.6441875401412974,
      "grad_norm": 0.27478665113449097,
      "learning_rate": 0.0001,
      "loss": 1.5806,
      "step": 4012
    },
    {
      "epoch": 0.6443481053307643,
      "grad_norm": 0.2817869186401367,
      "learning_rate": 0.0001,
      "loss": 1.551,
      "step": 4013
    },
    {
      "epoch": 0.6445086705202312,
      "grad_norm": 0.2698647677898407,
      "learning_rate": 0.0001,
      "loss": 1.4447,
      "step": 4014
    },
    {
      "epoch": 0.6446692357096981,
      "grad_norm": 0.26216360926628113,
      "learning_rate": 0.0001,
      "loss": 1.5068,
      "step": 4015
    },
    {
      "epoch": 0.6448298008991651,
      "grad_norm": 0.2640087306499481,
      "learning_rate": 0.0001,
      "loss": 1.4907,
      "step": 4016
    },
    {
      "epoch": 0.644990366088632,
      "grad_norm": 0.2662026286125183,
      "learning_rate": 0.0001,
      "loss": 1.4667,
      "step": 4017
    },
    {
      "epoch": 0.6451509312780989,
      "grad_norm": 0.25925353169441223,
      "learning_rate": 0.0001,
      "loss": 1.4734,
      "step": 4018
    },
    {
      "epoch": 0.6453114964675658,
      "grad_norm": 0.25807052850723267,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 4019
    },
    {
      "epoch": 0.6454720616570327,
      "grad_norm": 0.25807955861091614,
      "learning_rate": 0.0001,
      "loss": 1.4413,
      "step": 4020
    },
    {
      "epoch": 0.6456326268464997,
      "grad_norm": 0.25325387716293335,
      "learning_rate": 0.0001,
      "loss": 1.4866,
      "step": 4021
    },
    {
      "epoch": 0.6457931920359666,
      "grad_norm": 0.27402958273887634,
      "learning_rate": 0.0001,
      "loss": 1.489,
      "step": 4022
    },
    {
      "epoch": 0.6459537572254336,
      "grad_norm": 0.25539112091064453,
      "learning_rate": 0.0001,
      "loss": 1.4931,
      "step": 4023
    },
    {
      "epoch": 0.6461143224149004,
      "grad_norm": 0.2613573372364044,
      "learning_rate": 0.0001,
      "loss": 1.5178,
      "step": 4024
    },
    {
      "epoch": 0.6462748876043674,
      "grad_norm": 0.2629983425140381,
      "learning_rate": 0.0001,
      "loss": 1.4594,
      "step": 4025
    },
    {
      "epoch": 0.6464354527938343,
      "grad_norm": 0.27503475546836853,
      "learning_rate": 0.0001,
      "loss": 1.4597,
      "step": 4026
    },
    {
      "epoch": 0.6465960179833012,
      "grad_norm": 0.254550576210022,
      "learning_rate": 0.0001,
      "loss": 1.4578,
      "step": 4027
    },
    {
      "epoch": 0.6467565831727682,
      "grad_norm": 0.2574964761734009,
      "learning_rate": 0.0001,
      "loss": 1.3582,
      "step": 4028
    },
    {
      "epoch": 0.646917148362235,
      "grad_norm": 0.2585909366607666,
      "learning_rate": 0.0001,
      "loss": 1.4848,
      "step": 4029
    },
    {
      "epoch": 0.647077713551702,
      "grad_norm": 0.2635752558708191,
      "learning_rate": 0.0001,
      "loss": 1.5383,
      "step": 4030
    },
    {
      "epoch": 0.6472382787411689,
      "grad_norm": 0.2808038592338562,
      "learning_rate": 0.0001,
      "loss": 1.5712,
      "step": 4031
    },
    {
      "epoch": 0.6473988439306358,
      "grad_norm": 0.27354416251182556,
      "learning_rate": 0.0001,
      "loss": 1.5538,
      "step": 4032
    },
    {
      "epoch": 0.6475594091201028,
      "grad_norm": 0.2562684416770935,
      "learning_rate": 0.0001,
      "loss": 1.4289,
      "step": 4033
    },
    {
      "epoch": 0.6477199743095697,
      "grad_norm": 0.27846622467041016,
      "learning_rate": 0.0001,
      "loss": 1.5365,
      "step": 4034
    },
    {
      "epoch": 0.6478805394990366,
      "grad_norm": 0.26496443152427673,
      "learning_rate": 0.0001,
      "loss": 1.5764,
      "step": 4035
    },
    {
      "epoch": 0.6480411046885035,
      "grad_norm": 0.25747424364089966,
      "learning_rate": 0.0001,
      "loss": 1.5748,
      "step": 4036
    },
    {
      "epoch": 0.6482016698779705,
      "grad_norm": 0.2745968699455261,
      "learning_rate": 0.0001,
      "loss": 1.5252,
      "step": 4037
    },
    {
      "epoch": 0.6483622350674374,
      "grad_norm": 0.2700631320476532,
      "learning_rate": 0.0001,
      "loss": 1.506,
      "step": 4038
    },
    {
      "epoch": 0.6485228002569043,
      "grad_norm": 0.26957178115844727,
      "learning_rate": 0.0001,
      "loss": 1.5065,
      "step": 4039
    },
    {
      "epoch": 0.6486833654463712,
      "grad_norm": 0.25552108883857727,
      "learning_rate": 0.0001,
      "loss": 1.4456,
      "step": 4040
    },
    {
      "epoch": 0.6488439306358381,
      "grad_norm": 0.2736758589744568,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 4041
    },
    {
      "epoch": 0.6490044958253051,
      "grad_norm": 0.2582840621471405,
      "learning_rate": 0.0001,
      "loss": 1.4804,
      "step": 4042
    },
    {
      "epoch": 0.649165061014772,
      "grad_norm": 0.25374460220336914,
      "learning_rate": 0.0001,
      "loss": 1.5333,
      "step": 4043
    },
    {
      "epoch": 0.649325626204239,
      "grad_norm": 0.2682127058506012,
      "learning_rate": 0.0001,
      "loss": 1.4528,
      "step": 4044
    },
    {
      "epoch": 0.6494861913937059,
      "grad_norm": 0.26875752210617065,
      "learning_rate": 0.0001,
      "loss": 1.5214,
      "step": 4045
    },
    {
      "epoch": 0.6496467565831727,
      "grad_norm": 0.2548571825027466,
      "learning_rate": 0.0001,
      "loss": 1.5197,
      "step": 4046
    },
    {
      "epoch": 0.6498073217726397,
      "grad_norm": 0.24765118956565857,
      "learning_rate": 0.0001,
      "loss": 1.4662,
      "step": 4047
    },
    {
      "epoch": 0.6499678869621066,
      "grad_norm": 0.26132071018218994,
      "learning_rate": 0.0001,
      "loss": 1.4996,
      "step": 4048
    },
    {
      "epoch": 0.6501284521515736,
      "grad_norm": 0.25638461112976074,
      "learning_rate": 0.0001,
      "loss": 1.4963,
      "step": 4049
    },
    {
      "epoch": 0.6502890173410405,
      "grad_norm": 0.25048828125,
      "learning_rate": 0.0001,
      "loss": 1.3668,
      "step": 4050
    },
    {
      "epoch": 0.6504495825305073,
      "grad_norm": 0.2597469091415405,
      "learning_rate": 0.0001,
      "loss": 1.4992,
      "step": 4051
    },
    {
      "epoch": 0.6506101477199743,
      "grad_norm": 0.25522956252098083,
      "learning_rate": 0.0001,
      "loss": 1.5644,
      "step": 4052
    },
    {
      "epoch": 0.6507707129094412,
      "grad_norm": 0.351707398891449,
      "learning_rate": 0.0001,
      "loss": 1.4995,
      "step": 4053
    },
    {
      "epoch": 0.6509312780989082,
      "grad_norm": 0.2668572664260864,
      "learning_rate": 0.0001,
      "loss": 1.5145,
      "step": 4054
    },
    {
      "epoch": 0.6510918432883751,
      "grad_norm": 0.24314342439174652,
      "learning_rate": 0.0001,
      "loss": 1.4478,
      "step": 4055
    },
    {
      "epoch": 0.651252408477842,
      "grad_norm": 0.2775131165981293,
      "learning_rate": 0.0001,
      "loss": 1.5582,
      "step": 4056
    },
    {
      "epoch": 0.6514129736673089,
      "grad_norm": 0.24780496954917908,
      "learning_rate": 0.0001,
      "loss": 1.4818,
      "step": 4057
    },
    {
      "epoch": 0.6515735388567758,
      "grad_norm": 0.2634061574935913,
      "learning_rate": 0.0001,
      "loss": 1.5423,
      "step": 4058
    },
    {
      "epoch": 0.6517341040462428,
      "grad_norm": 0.2591491639614105,
      "learning_rate": 0.0001,
      "loss": 1.5061,
      "step": 4059
    },
    {
      "epoch": 0.6518946692357097,
      "grad_norm": 0.2679141163825989,
      "learning_rate": 0.0001,
      "loss": 1.5486,
      "step": 4060
    },
    {
      "epoch": 0.6520552344251767,
      "grad_norm": 0.27372705936431885,
      "learning_rate": 0.0001,
      "loss": 1.6082,
      "step": 4061
    },
    {
      "epoch": 0.6522157996146436,
      "grad_norm": 0.2574971914291382,
      "learning_rate": 0.0001,
      "loss": 1.643,
      "step": 4062
    },
    {
      "epoch": 0.6523763648041104,
      "grad_norm": 0.26333364844322205,
      "learning_rate": 0.0001,
      "loss": 1.5374,
      "step": 4063
    },
    {
      "epoch": 0.6525369299935774,
      "grad_norm": 0.2890571057796478,
      "learning_rate": 0.0001,
      "loss": 1.5404,
      "step": 4064
    },
    {
      "epoch": 0.6526974951830443,
      "grad_norm": 0.26847440004348755,
      "learning_rate": 0.0001,
      "loss": 1.4936,
      "step": 4065
    },
    {
      "epoch": 0.6528580603725113,
      "grad_norm": 0.2642154097557068,
      "learning_rate": 0.0001,
      "loss": 1.538,
      "step": 4066
    },
    {
      "epoch": 0.6530186255619782,
      "grad_norm": 0.29670676589012146,
      "learning_rate": 0.0001,
      "loss": 1.5223,
      "step": 4067
    },
    {
      "epoch": 0.653179190751445,
      "grad_norm": 0.2784339487552643,
      "learning_rate": 0.0001,
      "loss": 1.553,
      "step": 4068
    },
    {
      "epoch": 0.653339755940912,
      "grad_norm": 0.2552869915962219,
      "learning_rate": 0.0001,
      "loss": 1.3798,
      "step": 4069
    },
    {
      "epoch": 0.6535003211303789,
      "grad_norm": 0.26420438289642334,
      "learning_rate": 0.0001,
      "loss": 1.4712,
      "step": 4070
    },
    {
      "epoch": 0.6536608863198459,
      "grad_norm": 0.26188531517982483,
      "learning_rate": 0.0001,
      "loss": 1.5179,
      "step": 4071
    },
    {
      "epoch": 0.6538214515093128,
      "grad_norm": 0.25568708777427673,
      "learning_rate": 0.0001,
      "loss": 1.4788,
      "step": 4072
    },
    {
      "epoch": 0.6539820166987798,
      "grad_norm": 0.2482534795999527,
      "learning_rate": 0.0001,
      "loss": 1.4883,
      "step": 4073
    },
    {
      "epoch": 0.6541425818882466,
      "grad_norm": 0.269436240196228,
      "learning_rate": 0.0001,
      "loss": 1.4757,
      "step": 4074
    },
    {
      "epoch": 0.6543031470777135,
      "grad_norm": 0.2498292624950409,
      "learning_rate": 0.0001,
      "loss": 1.4839,
      "step": 4075
    },
    {
      "epoch": 0.6544637122671805,
      "grad_norm": 0.2830914258956909,
      "learning_rate": 0.0001,
      "loss": 1.4721,
      "step": 4076
    },
    {
      "epoch": 0.6546242774566474,
      "grad_norm": 0.28456372022628784,
      "learning_rate": 0.0001,
      "loss": 1.5467,
      "step": 4077
    },
    {
      "epoch": 0.6547848426461144,
      "grad_norm": 0.27140548825263977,
      "learning_rate": 0.0001,
      "loss": 1.6219,
      "step": 4078
    },
    {
      "epoch": 0.6549454078355812,
      "grad_norm": 0.26949694752693176,
      "learning_rate": 0.0001,
      "loss": 1.5581,
      "step": 4079
    },
    {
      "epoch": 0.6551059730250481,
      "grad_norm": 0.27604708075523376,
      "learning_rate": 0.0001,
      "loss": 1.5329,
      "step": 4080
    },
    {
      "epoch": 0.6552665382145151,
      "grad_norm": 0.2518842816352844,
      "learning_rate": 0.0001,
      "loss": 1.4373,
      "step": 4081
    },
    {
      "epoch": 0.655427103403982,
      "grad_norm": 0.26696524024009705,
      "learning_rate": 0.0001,
      "loss": 1.3639,
      "step": 4082
    },
    {
      "epoch": 0.655587668593449,
      "grad_norm": 0.2709606885910034,
      "learning_rate": 0.0001,
      "loss": 1.445,
      "step": 4083
    },
    {
      "epoch": 0.6557482337829159,
      "grad_norm": 0.2692444324493408,
      "learning_rate": 0.0001,
      "loss": 1.4507,
      "step": 4084
    },
    {
      "epoch": 0.6559087989723827,
      "grad_norm": 0.2680546045303345,
      "learning_rate": 0.0001,
      "loss": 1.5252,
      "step": 4085
    },
    {
      "epoch": 0.6560693641618497,
      "grad_norm": 0.278359055519104,
      "learning_rate": 0.0001,
      "loss": 1.4262,
      "step": 4086
    },
    {
      "epoch": 0.6562299293513166,
      "grad_norm": 0.27649831771850586,
      "learning_rate": 0.0001,
      "loss": 1.5494,
      "step": 4087
    },
    {
      "epoch": 0.6563904945407836,
      "grad_norm": 0.2656194865703583,
      "learning_rate": 0.0001,
      "loss": 1.5439,
      "step": 4088
    },
    {
      "epoch": 0.6565510597302505,
      "grad_norm": 0.26811110973358154,
      "learning_rate": 0.0001,
      "loss": 1.4896,
      "step": 4089
    },
    {
      "epoch": 0.6567116249197174,
      "grad_norm": 0.2669956386089325,
      "learning_rate": 0.0001,
      "loss": 1.4648,
      "step": 4090
    },
    {
      "epoch": 0.6568721901091843,
      "grad_norm": 0.2634719908237457,
      "learning_rate": 0.0001,
      "loss": 1.4921,
      "step": 4091
    },
    {
      "epoch": 0.6570327552986512,
      "grad_norm": 0.2633553445339203,
      "learning_rate": 0.0001,
      "loss": 1.5851,
      "step": 4092
    },
    {
      "epoch": 0.6571933204881182,
      "grad_norm": 0.25439122319221497,
      "learning_rate": 0.0001,
      "loss": 1.471,
      "step": 4093
    },
    {
      "epoch": 0.6573538856775851,
      "grad_norm": 0.28257036209106445,
      "learning_rate": 0.0001,
      "loss": 1.5814,
      "step": 4094
    },
    {
      "epoch": 0.6575144508670521,
      "grad_norm": 0.26080578565597534,
      "learning_rate": 0.0001,
      "loss": 1.484,
      "step": 4095
    },
    {
      "epoch": 0.6576750160565189,
      "grad_norm": 0.28057047724723816,
      "learning_rate": 0.0001,
      "loss": 1.5782,
      "step": 4096
    },
    {
      "epoch": 0.6578355812459858,
      "grad_norm": 0.25944429636001587,
      "learning_rate": 0.0001,
      "loss": 1.5081,
      "step": 4097
    },
    {
      "epoch": 0.6579961464354528,
      "grad_norm": 0.2557280361652374,
      "learning_rate": 0.0001,
      "loss": 1.5102,
      "step": 4098
    },
    {
      "epoch": 0.6581567116249197,
      "grad_norm": 0.2652299702167511,
      "learning_rate": 0.0001,
      "loss": 1.5274,
      "step": 4099
    },
    {
      "epoch": 0.6583172768143867,
      "grad_norm": 0.26031339168548584,
      "learning_rate": 0.0001,
      "loss": 1.5459,
      "step": 4100
    },
    {
      "epoch": 0.6584778420038536,
      "grad_norm": 0.2873890995979309,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 4101
    },
    {
      "epoch": 0.6586384071933205,
      "grad_norm": 0.2638516128063202,
      "learning_rate": 0.0001,
      "loss": 1.5402,
      "step": 4102
    },
    {
      "epoch": 0.6587989723827874,
      "grad_norm": 0.2758546471595764,
      "learning_rate": 0.0001,
      "loss": 1.5576,
      "step": 4103
    },
    {
      "epoch": 0.6589595375722543,
      "grad_norm": 0.25617489218711853,
      "learning_rate": 0.0001,
      "loss": 1.5061,
      "step": 4104
    },
    {
      "epoch": 0.6591201027617213,
      "grad_norm": 0.27788490056991577,
      "learning_rate": 0.0001,
      "loss": 1.5287,
      "step": 4105
    },
    {
      "epoch": 0.6592806679511882,
      "grad_norm": 0.27403321862220764,
      "learning_rate": 0.0001,
      "loss": 1.5544,
      "step": 4106
    },
    {
      "epoch": 0.6594412331406551,
      "grad_norm": 0.26027989387512207,
      "learning_rate": 0.0001,
      "loss": 1.61,
      "step": 4107
    },
    {
      "epoch": 0.659601798330122,
      "grad_norm": 0.266972154378891,
      "learning_rate": 0.0001,
      "loss": 1.5504,
      "step": 4108
    },
    {
      "epoch": 0.659762363519589,
      "grad_norm": 0.2704896628856659,
      "learning_rate": 0.0001,
      "loss": 1.547,
      "step": 4109
    },
    {
      "epoch": 0.6599229287090559,
      "grad_norm": 0.2757876217365265,
      "learning_rate": 0.0001,
      "loss": 1.4804,
      "step": 4110
    },
    {
      "epoch": 0.6600834938985228,
      "grad_norm": 0.2564221918582916,
      "learning_rate": 0.0001,
      "loss": 1.5409,
      "step": 4111
    },
    {
      "epoch": 0.6602440590879898,
      "grad_norm": 0.27693742513656616,
      "learning_rate": 0.0001,
      "loss": 1.4758,
      "step": 4112
    },
    {
      "epoch": 0.6604046242774566,
      "grad_norm": 0.2597733736038208,
      "learning_rate": 0.0001,
      "loss": 1.4343,
      "step": 4113
    },
    {
      "epoch": 0.6605651894669236,
      "grad_norm": 0.26988744735717773,
      "learning_rate": 0.0001,
      "loss": 1.522,
      "step": 4114
    },
    {
      "epoch": 0.6607257546563905,
      "grad_norm": 0.2563435137271881,
      "learning_rate": 0.0001,
      "loss": 1.451,
      "step": 4115
    },
    {
      "epoch": 0.6608863198458574,
      "grad_norm": 0.2942071557044983,
      "learning_rate": 0.0001,
      "loss": 1.5037,
      "step": 4116
    },
    {
      "epoch": 0.6610468850353244,
      "grad_norm": 0.2602013945579529,
      "learning_rate": 0.0001,
      "loss": 1.4699,
      "step": 4117
    },
    {
      "epoch": 0.6612074502247912,
      "grad_norm": 0.28853535652160645,
      "learning_rate": 0.0001,
      "loss": 1.417,
      "step": 4118
    },
    {
      "epoch": 0.6613680154142582,
      "grad_norm": 0.2661252021789551,
      "learning_rate": 0.0001,
      "loss": 1.4368,
      "step": 4119
    },
    {
      "epoch": 0.6615285806037251,
      "grad_norm": 0.27435213327407837,
      "learning_rate": 0.0001,
      "loss": 1.4582,
      "step": 4120
    },
    {
      "epoch": 0.661689145793192,
      "grad_norm": 0.2826497554779053,
      "learning_rate": 0.0001,
      "loss": 1.593,
      "step": 4121
    },
    {
      "epoch": 0.661849710982659,
      "grad_norm": 0.25823888182640076,
      "learning_rate": 0.0001,
      "loss": 1.4628,
      "step": 4122
    },
    {
      "epoch": 0.6620102761721259,
      "grad_norm": 0.27879124879837036,
      "learning_rate": 0.0001,
      "loss": 1.4943,
      "step": 4123
    },
    {
      "epoch": 0.6621708413615928,
      "grad_norm": 0.25891903042793274,
      "learning_rate": 0.0001,
      "loss": 1.4956,
      "step": 4124
    },
    {
      "epoch": 0.6623314065510597,
      "grad_norm": 0.261674165725708,
      "learning_rate": 0.0001,
      "loss": 1.4694,
      "step": 4125
    },
    {
      "epoch": 0.6624919717405267,
      "grad_norm": 0.2857045531272888,
      "learning_rate": 0.0001,
      "loss": 1.4331,
      "step": 4126
    },
    {
      "epoch": 0.6626525369299936,
      "grad_norm": 0.273318350315094,
      "learning_rate": 0.0001,
      "loss": 1.4926,
      "step": 4127
    },
    {
      "epoch": 0.6628131021194605,
      "grad_norm": 0.2702943682670593,
      "learning_rate": 0.0001,
      "loss": 1.5398,
      "step": 4128
    },
    {
      "epoch": 0.6629736673089274,
      "grad_norm": 0.2519720196723938,
      "learning_rate": 0.0001,
      "loss": 1.4458,
      "step": 4129
    },
    {
      "epoch": 0.6631342324983943,
      "grad_norm": 0.2591875493526459,
      "learning_rate": 0.0001,
      "loss": 1.4549,
      "step": 4130
    },
    {
      "epoch": 0.6632947976878613,
      "grad_norm": 0.263841837644577,
      "learning_rate": 0.0001,
      "loss": 1.5494,
      "step": 4131
    },
    {
      "epoch": 0.6634553628773282,
      "grad_norm": 0.26048779487609863,
      "learning_rate": 0.0001,
      "loss": 1.5098,
      "step": 4132
    },
    {
      "epoch": 0.6636159280667951,
      "grad_norm": 0.26579147577285767,
      "learning_rate": 0.0001,
      "loss": 1.5549,
      "step": 4133
    },
    {
      "epoch": 0.6637764932562621,
      "grad_norm": 0.262335866689682,
      "learning_rate": 0.0001,
      "loss": 1.5588,
      "step": 4134
    },
    {
      "epoch": 0.6639370584457289,
      "grad_norm": 0.26614880561828613,
      "learning_rate": 0.0001,
      "loss": 1.4807,
      "step": 4135
    },
    {
      "epoch": 0.6640976236351959,
      "grad_norm": 0.2592613101005554,
      "learning_rate": 0.0001,
      "loss": 1.4901,
      "step": 4136
    },
    {
      "epoch": 0.6642581888246628,
      "grad_norm": 0.25582200288772583,
      "learning_rate": 0.0001,
      "loss": 1.4766,
      "step": 4137
    },
    {
      "epoch": 0.6644187540141298,
      "grad_norm": 0.26614314317703247,
      "learning_rate": 0.0001,
      "loss": 1.5446,
      "step": 4138
    },
    {
      "epoch": 0.6645793192035967,
      "grad_norm": 0.2539105713367462,
      "learning_rate": 0.0001,
      "loss": 1.5006,
      "step": 4139
    },
    {
      "epoch": 0.6647398843930635,
      "grad_norm": 0.2665300667285919,
      "learning_rate": 0.0001,
      "loss": 1.5861,
      "step": 4140
    },
    {
      "epoch": 0.6649004495825305,
      "grad_norm": 0.2647922933101654,
      "learning_rate": 0.0001,
      "loss": 1.5404,
      "step": 4141
    },
    {
      "epoch": 0.6650610147719974,
      "grad_norm": 0.25334295630455017,
      "learning_rate": 0.0001,
      "loss": 1.4352,
      "step": 4142
    },
    {
      "epoch": 0.6652215799614644,
      "grad_norm": 0.27983757853507996,
      "learning_rate": 0.0001,
      "loss": 1.5223,
      "step": 4143
    },
    {
      "epoch": 0.6653821451509313,
      "grad_norm": 0.2596900761127472,
      "learning_rate": 0.0001,
      "loss": 1.5112,
      "step": 4144
    },
    {
      "epoch": 0.6655427103403982,
      "grad_norm": 0.26293647289276123,
      "learning_rate": 0.0001,
      "loss": 1.5191,
      "step": 4145
    },
    {
      "epoch": 0.6657032755298651,
      "grad_norm": 0.46495234966278076,
      "learning_rate": 0.0001,
      "loss": 1.4792,
      "step": 4146
    },
    {
      "epoch": 0.665863840719332,
      "grad_norm": 0.26407575607299805,
      "learning_rate": 0.0001,
      "loss": 1.481,
      "step": 4147
    },
    {
      "epoch": 0.666024405908799,
      "grad_norm": 0.46885931491851807,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 4148
    },
    {
      "epoch": 0.6661849710982659,
      "grad_norm": 0.268364280462265,
      "learning_rate": 0.0001,
      "loss": 1.4719,
      "step": 4149
    },
    {
      "epoch": 0.6663455362877329,
      "grad_norm": 0.2675420045852661,
      "learning_rate": 0.0001,
      "loss": 1.4783,
      "step": 4150
    },
    {
      "epoch": 0.6665061014771998,
      "grad_norm": 0.26824304461479187,
      "learning_rate": 0.0001,
      "loss": 1.5892,
      "step": 4151
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.25549522042274475,
      "learning_rate": 0.0001,
      "loss": 1.5058,
      "step": 4152
    },
    {
      "epoch": 0.6668272318561336,
      "grad_norm": 0.24786025285720825,
      "learning_rate": 0.0001,
      "loss": 1.5099,
      "step": 4153
    },
    {
      "epoch": 0.6669877970456005,
      "grad_norm": 0.2666029930114746,
      "learning_rate": 0.0001,
      "loss": 1.5409,
      "step": 4154
    },
    {
      "epoch": 0.6671483622350675,
      "grad_norm": 0.28225189447402954,
      "learning_rate": 0.0001,
      "loss": 1.5035,
      "step": 4155
    },
    {
      "epoch": 0.6673089274245344,
      "grad_norm": 0.2664307653903961,
      "learning_rate": 0.0001,
      "loss": 1.5109,
      "step": 4156
    },
    {
      "epoch": 0.6674694926140012,
      "grad_norm": 0.2636362910270691,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 4157
    },
    {
      "epoch": 0.6676300578034682,
      "grad_norm": 0.27478745579719543,
      "learning_rate": 0.0001,
      "loss": 1.5109,
      "step": 4158
    },
    {
      "epoch": 0.6677906229929351,
      "grad_norm": 0.28328198194503784,
      "learning_rate": 0.0001,
      "loss": 1.4975,
      "step": 4159
    },
    {
      "epoch": 0.6679511881824021,
      "grad_norm": 0.2634706199169159,
      "learning_rate": 0.0001,
      "loss": 1.5536,
      "step": 4160
    },
    {
      "epoch": 0.668111753371869,
      "grad_norm": 0.26545238494873047,
      "learning_rate": 0.0001,
      "loss": 1.448,
      "step": 4161
    },
    {
      "epoch": 0.668272318561336,
      "grad_norm": 0.2772989869117737,
      "learning_rate": 0.0001,
      "loss": 1.5411,
      "step": 4162
    },
    {
      "epoch": 0.6684328837508028,
      "grad_norm": 0.26372072100639343,
      "learning_rate": 0.0001,
      "loss": 1.5296,
      "step": 4163
    },
    {
      "epoch": 0.6685934489402697,
      "grad_norm": 0.2692832052707672,
      "learning_rate": 0.0001,
      "loss": 1.4802,
      "step": 4164
    },
    {
      "epoch": 0.6687540141297367,
      "grad_norm": 0.2728932499885559,
      "learning_rate": 0.0001,
      "loss": 1.5573,
      "step": 4165
    },
    {
      "epoch": 0.6689145793192036,
      "grad_norm": 0.25267428159713745,
      "learning_rate": 0.0001,
      "loss": 1.5656,
      "step": 4166
    },
    {
      "epoch": 0.6690751445086706,
      "grad_norm": 0.25965332984924316,
      "learning_rate": 0.0001,
      "loss": 1.5448,
      "step": 4167
    },
    {
      "epoch": 0.6692357096981374,
      "grad_norm": 0.2715737819671631,
      "learning_rate": 0.0001,
      "loss": 1.4931,
      "step": 4168
    },
    {
      "epoch": 0.6693962748876043,
      "grad_norm": 0.263933390378952,
      "learning_rate": 0.0001,
      "loss": 1.4785,
      "step": 4169
    },
    {
      "epoch": 0.6695568400770713,
      "grad_norm": 0.26766473054885864,
      "learning_rate": 0.0001,
      "loss": 1.4657,
      "step": 4170
    },
    {
      "epoch": 0.6697174052665382,
      "grad_norm": 0.2798439860343933,
      "learning_rate": 0.0001,
      "loss": 1.5415,
      "step": 4171
    },
    {
      "epoch": 0.6698779704560052,
      "grad_norm": 0.2672432065010071,
      "learning_rate": 0.0001,
      "loss": 1.5437,
      "step": 4172
    },
    {
      "epoch": 0.6700385356454721,
      "grad_norm": 0.2547382116317749,
      "learning_rate": 0.0001,
      "loss": 1.4333,
      "step": 4173
    },
    {
      "epoch": 0.6701991008349389,
      "grad_norm": 0.26243576407432556,
      "learning_rate": 0.0001,
      "loss": 1.5311,
      "step": 4174
    },
    {
      "epoch": 0.6703596660244059,
      "grad_norm": 0.27144116163253784,
      "learning_rate": 0.0001,
      "loss": 1.5626,
      "step": 4175
    },
    {
      "epoch": 0.6705202312138728,
      "grad_norm": 0.27108368277549744,
      "learning_rate": 0.0001,
      "loss": 1.4498,
      "step": 4176
    },
    {
      "epoch": 0.6706807964033398,
      "grad_norm": 0.272928923368454,
      "learning_rate": 0.0001,
      "loss": 1.5726,
      "step": 4177
    },
    {
      "epoch": 0.6708413615928067,
      "grad_norm": 0.25330835580825806,
      "learning_rate": 0.0001,
      "loss": 1.4942,
      "step": 4178
    },
    {
      "epoch": 0.6710019267822736,
      "grad_norm": 0.25305670499801636,
      "learning_rate": 0.0001,
      "loss": 1.4918,
      "step": 4179
    },
    {
      "epoch": 0.6711624919717405,
      "grad_norm": 0.2702277600765228,
      "learning_rate": 0.0001,
      "loss": 1.5377,
      "step": 4180
    },
    {
      "epoch": 0.6713230571612074,
      "grad_norm": 0.2533726990222931,
      "learning_rate": 0.0001,
      "loss": 1.4688,
      "step": 4181
    },
    {
      "epoch": 0.6714836223506744,
      "grad_norm": 0.2574719488620758,
      "learning_rate": 0.0001,
      "loss": 1.5209,
      "step": 4182
    },
    {
      "epoch": 0.6716441875401413,
      "grad_norm": 0.2595611810684204,
      "learning_rate": 0.0001,
      "loss": 1.4995,
      "step": 4183
    },
    {
      "epoch": 0.6718047527296083,
      "grad_norm": 0.25201043486595154,
      "learning_rate": 0.0001,
      "loss": 1.471,
      "step": 4184
    },
    {
      "epoch": 0.6719653179190751,
      "grad_norm": 0.2750900089740753,
      "learning_rate": 0.0001,
      "loss": 1.4705,
      "step": 4185
    },
    {
      "epoch": 0.672125883108542,
      "grad_norm": 0.26094889640808105,
      "learning_rate": 0.0001,
      "loss": 1.5288,
      "step": 4186
    },
    {
      "epoch": 0.672286448298009,
      "grad_norm": 0.25764408707618713,
      "learning_rate": 0.0001,
      "loss": 1.4275,
      "step": 4187
    },
    {
      "epoch": 0.6724470134874759,
      "grad_norm": 0.25645217299461365,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 4188
    },
    {
      "epoch": 0.6726075786769429,
      "grad_norm": 0.2957596778869629,
      "learning_rate": 0.0001,
      "loss": 1.4583,
      "step": 4189
    },
    {
      "epoch": 0.6727681438664097,
      "grad_norm": 0.2649129629135132,
      "learning_rate": 0.0001,
      "loss": 1.5384,
      "step": 4190
    },
    {
      "epoch": 0.6729287090558767,
      "grad_norm": 0.2660101652145386,
      "learning_rate": 0.0001,
      "loss": 1.389,
      "step": 4191
    },
    {
      "epoch": 0.6730892742453436,
      "grad_norm": 0.2686706483364105,
      "learning_rate": 0.0001,
      "loss": 1.5432,
      "step": 4192
    },
    {
      "epoch": 0.6732498394348105,
      "grad_norm": 0.2836068272590637,
      "learning_rate": 0.0001,
      "loss": 1.4721,
      "step": 4193
    },
    {
      "epoch": 0.6734104046242775,
      "grad_norm": 0.25428712368011475,
      "learning_rate": 0.0001,
      "loss": 1.4336,
      "step": 4194
    },
    {
      "epoch": 0.6735709698137444,
      "grad_norm": 0.2774694859981537,
      "learning_rate": 0.0001,
      "loss": 1.4709,
      "step": 4195
    },
    {
      "epoch": 0.6737315350032113,
      "grad_norm": 0.2620084583759308,
      "learning_rate": 0.0001,
      "loss": 1.4472,
      "step": 4196
    },
    {
      "epoch": 0.6738921001926782,
      "grad_norm": 0.2742082476615906,
      "learning_rate": 0.0001,
      "loss": 1.6028,
      "step": 4197
    },
    {
      "epoch": 0.6740526653821451,
      "grad_norm": 0.2725132405757904,
      "learning_rate": 0.0001,
      "loss": 1.4722,
      "step": 4198
    },
    {
      "epoch": 0.6742132305716121,
      "grad_norm": 0.26817259192466736,
      "learning_rate": 0.0001,
      "loss": 1.5051,
      "step": 4199
    },
    {
      "epoch": 0.674373795761079,
      "grad_norm": 0.27134665846824646,
      "learning_rate": 0.0001,
      "loss": 1.4977,
      "step": 4200
    },
    {
      "epoch": 0.674534360950546,
      "grad_norm": 0.2610231935977936,
      "learning_rate": 0.0001,
      "loss": 1.5038,
      "step": 4201
    },
    {
      "epoch": 0.6746949261400128,
      "grad_norm": 0.2780536413192749,
      "learning_rate": 0.0001,
      "loss": 1.5502,
      "step": 4202
    },
    {
      "epoch": 0.6748554913294798,
      "grad_norm": 0.26236626505851746,
      "learning_rate": 0.0001,
      "loss": 1.539,
      "step": 4203
    },
    {
      "epoch": 0.6750160565189467,
      "grad_norm": 0.2582312226295471,
      "learning_rate": 0.0001,
      "loss": 1.4549,
      "step": 4204
    },
    {
      "epoch": 0.6751766217084136,
      "grad_norm": 0.26440122723579407,
      "learning_rate": 0.0001,
      "loss": 1.5539,
      "step": 4205
    },
    {
      "epoch": 0.6753371868978806,
      "grad_norm": 0.26772627234458923,
      "learning_rate": 0.0001,
      "loss": 1.4856,
      "step": 4206
    },
    {
      "epoch": 0.6754977520873474,
      "grad_norm": 0.25626301765441895,
      "learning_rate": 0.0001,
      "loss": 1.5426,
      "step": 4207
    },
    {
      "epoch": 0.6756583172768144,
      "grad_norm": 0.2645321190357208,
      "learning_rate": 0.0001,
      "loss": 1.459,
      "step": 4208
    },
    {
      "epoch": 0.6758188824662813,
      "grad_norm": 0.2824168801307678,
      "learning_rate": 0.0001,
      "loss": 1.4617,
      "step": 4209
    },
    {
      "epoch": 0.6759794476557482,
      "grad_norm": 0.25180715322494507,
      "learning_rate": 0.0001,
      "loss": 1.4945,
      "step": 4210
    },
    {
      "epoch": 0.6761400128452152,
      "grad_norm": 0.27988308668136597,
      "learning_rate": 0.0001,
      "loss": 1.4962,
      "step": 4211
    },
    {
      "epoch": 0.6763005780346821,
      "grad_norm": 0.2720843553543091,
      "learning_rate": 0.0001,
      "loss": 1.4862,
      "step": 4212
    },
    {
      "epoch": 0.676461143224149,
      "grad_norm": 0.24799558520317078,
      "learning_rate": 0.0001,
      "loss": 1.4484,
      "step": 4213
    },
    {
      "epoch": 0.6766217084136159,
      "grad_norm": 0.27907493710517883,
      "learning_rate": 0.0001,
      "loss": 1.5056,
      "step": 4214
    },
    {
      "epoch": 0.6767822736030829,
      "grad_norm": 0.26135680079460144,
      "learning_rate": 0.0001,
      "loss": 1.4847,
      "step": 4215
    },
    {
      "epoch": 0.6769428387925498,
      "grad_norm": 0.25625544786453247,
      "learning_rate": 0.0001,
      "loss": 1.5233,
      "step": 4216
    },
    {
      "epoch": 0.6771034039820167,
      "grad_norm": 0.2576616108417511,
      "learning_rate": 0.0001,
      "loss": 1.54,
      "step": 4217
    },
    {
      "epoch": 0.6772639691714836,
      "grad_norm": 0.2574950158596039,
      "learning_rate": 0.0001,
      "loss": 1.4897,
      "step": 4218
    },
    {
      "epoch": 0.6774245343609505,
      "grad_norm": 0.2586347460746765,
      "learning_rate": 0.0001,
      "loss": 1.447,
      "step": 4219
    },
    {
      "epoch": 0.6775850995504175,
      "grad_norm": 0.24843011796474457,
      "learning_rate": 0.0001,
      "loss": 1.525,
      "step": 4220
    },
    {
      "epoch": 0.6777456647398844,
      "grad_norm": 0.2628840208053589,
      "learning_rate": 0.0001,
      "loss": 1.5484,
      "step": 4221
    },
    {
      "epoch": 0.6779062299293513,
      "grad_norm": 0.2488044947385788,
      "learning_rate": 0.0001,
      "loss": 1.4058,
      "step": 4222
    },
    {
      "epoch": 0.6780667951188183,
      "grad_norm": 0.27336618304252625,
      "learning_rate": 0.0001,
      "loss": 1.554,
      "step": 4223
    },
    {
      "epoch": 0.6782273603082851,
      "grad_norm": 0.2552751898765564,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 4224
    },
    {
      "epoch": 0.6783879254977521,
      "grad_norm": 0.26962193846702576,
      "learning_rate": 0.0001,
      "loss": 1.5626,
      "step": 4225
    },
    {
      "epoch": 0.678548490687219,
      "grad_norm": 0.2638053297996521,
      "learning_rate": 0.0001,
      "loss": 1.4879,
      "step": 4226
    },
    {
      "epoch": 0.678709055876686,
      "grad_norm": 0.2734527289867401,
      "learning_rate": 0.0001,
      "loss": 1.4906,
      "step": 4227
    },
    {
      "epoch": 0.6788696210661529,
      "grad_norm": 0.2864174544811249,
      "learning_rate": 0.0001,
      "loss": 1.5056,
      "step": 4228
    },
    {
      "epoch": 0.6790301862556197,
      "grad_norm": 0.24866554141044617,
      "learning_rate": 0.0001,
      "loss": 1.5035,
      "step": 4229
    },
    {
      "epoch": 0.6791907514450867,
      "grad_norm": 0.2756553292274475,
      "learning_rate": 0.0001,
      "loss": 1.5121,
      "step": 4230
    },
    {
      "epoch": 0.6793513166345536,
      "grad_norm": 0.2636953294277191,
      "learning_rate": 0.0001,
      "loss": 1.5706,
      "step": 4231
    },
    {
      "epoch": 0.6795118818240206,
      "grad_norm": 0.2574693262577057,
      "learning_rate": 0.0001,
      "loss": 1.5137,
      "step": 4232
    },
    {
      "epoch": 0.6796724470134875,
      "grad_norm": 0.25836923718452454,
      "learning_rate": 0.0001,
      "loss": 1.5287,
      "step": 4233
    },
    {
      "epoch": 0.6798330122029544,
      "grad_norm": 0.2573105990886688,
      "learning_rate": 0.0001,
      "loss": 1.4541,
      "step": 4234
    },
    {
      "epoch": 0.6799935773924213,
      "grad_norm": 0.2674020230770111,
      "learning_rate": 0.0001,
      "loss": 1.5424,
      "step": 4235
    },
    {
      "epoch": 0.6801541425818882,
      "grad_norm": 0.269796222448349,
      "learning_rate": 0.0001,
      "loss": 1.6454,
      "step": 4236
    },
    {
      "epoch": 0.6803147077713552,
      "grad_norm": 0.2537638247013092,
      "learning_rate": 0.0001,
      "loss": 1.4866,
      "step": 4237
    },
    {
      "epoch": 0.6804752729608221,
      "grad_norm": 0.2629987299442291,
      "learning_rate": 0.0001,
      "loss": 1.4895,
      "step": 4238
    },
    {
      "epoch": 0.680635838150289,
      "grad_norm": 0.2534637749195099,
      "learning_rate": 0.0001,
      "loss": 1.4874,
      "step": 4239
    },
    {
      "epoch": 0.680796403339756,
      "grad_norm": 0.26694411039352417,
      "learning_rate": 0.0001,
      "loss": 1.5057,
      "step": 4240
    },
    {
      "epoch": 0.6809569685292228,
      "grad_norm": 0.25374630093574524,
      "learning_rate": 0.0001,
      "loss": 1.4128,
      "step": 4241
    },
    {
      "epoch": 0.6811175337186898,
      "grad_norm": 0.258924663066864,
      "learning_rate": 0.0001,
      "loss": 1.4962,
      "step": 4242
    },
    {
      "epoch": 0.6812780989081567,
      "grad_norm": 0.2512001693248749,
      "learning_rate": 0.0001,
      "loss": 1.4155,
      "step": 4243
    },
    {
      "epoch": 0.6814386640976237,
      "grad_norm": 0.2707739472389221,
      "learning_rate": 0.0001,
      "loss": 1.5236,
      "step": 4244
    },
    {
      "epoch": 0.6815992292870906,
      "grad_norm": 0.2508571445941925,
      "learning_rate": 0.0001,
      "loss": 1.5075,
      "step": 4245
    },
    {
      "epoch": 0.6817597944765574,
      "grad_norm": 0.2570149898529053,
      "learning_rate": 0.0001,
      "loss": 1.4935,
      "step": 4246
    },
    {
      "epoch": 0.6819203596660244,
      "grad_norm": 0.25217339396476746,
      "learning_rate": 0.0001,
      "loss": 1.4927,
      "step": 4247
    },
    {
      "epoch": 0.6820809248554913,
      "grad_norm": 0.26150375604629517,
      "learning_rate": 0.0001,
      "loss": 1.4548,
      "step": 4248
    },
    {
      "epoch": 0.6822414900449583,
      "grad_norm": 0.25464996695518494,
      "learning_rate": 0.0001,
      "loss": 1.4027,
      "step": 4249
    },
    {
      "epoch": 0.6824020552344252,
      "grad_norm": 0.2840500771999359,
      "learning_rate": 0.0001,
      "loss": 1.5993,
      "step": 4250
    },
    {
      "epoch": 0.6825626204238922,
      "grad_norm": 0.2636156678199768,
      "learning_rate": 0.0001,
      "loss": 1.4286,
      "step": 4251
    },
    {
      "epoch": 0.682723185613359,
      "grad_norm": 0.26296964287757874,
      "learning_rate": 0.0001,
      "loss": 1.4855,
      "step": 4252
    },
    {
      "epoch": 0.6828837508028259,
      "grad_norm": 0.2587028443813324,
      "learning_rate": 0.0001,
      "loss": 1.4446,
      "step": 4253
    },
    {
      "epoch": 0.6830443159922929,
      "grad_norm": 0.2679729759693146,
      "learning_rate": 0.0001,
      "loss": 1.5407,
      "step": 4254
    },
    {
      "epoch": 0.6832048811817598,
      "grad_norm": 0.25425592064857483,
      "learning_rate": 0.0001,
      "loss": 1.4615,
      "step": 4255
    },
    {
      "epoch": 0.6833654463712268,
      "grad_norm": 0.33618855476379395,
      "learning_rate": 0.0001,
      "loss": 1.4126,
      "step": 4256
    },
    {
      "epoch": 0.6835260115606936,
      "grad_norm": 0.5344606637954712,
      "learning_rate": 0.0001,
      "loss": 1.4545,
      "step": 4257
    },
    {
      "epoch": 0.6836865767501605,
      "grad_norm": 0.2503321170806885,
      "learning_rate": 0.0001,
      "loss": 1.4864,
      "step": 4258
    },
    {
      "epoch": 0.6838471419396275,
      "grad_norm": 0.27189552783966064,
      "learning_rate": 0.0001,
      "loss": 1.5107,
      "step": 4259
    },
    {
      "epoch": 0.6840077071290944,
      "grad_norm": 0.26606759428977966,
      "learning_rate": 0.0001,
      "loss": 1.54,
      "step": 4260
    },
    {
      "epoch": 0.6841682723185614,
      "grad_norm": 0.2644661068916321,
      "learning_rate": 0.0001,
      "loss": 1.5584,
      "step": 4261
    },
    {
      "epoch": 0.6843288375080283,
      "grad_norm": 0.26754721999168396,
      "learning_rate": 0.0001,
      "loss": 1.5246,
      "step": 4262
    },
    {
      "epoch": 0.6844894026974951,
      "grad_norm": 0.26102858781814575,
      "learning_rate": 0.0001,
      "loss": 1.4957,
      "step": 4263
    },
    {
      "epoch": 0.6846499678869621,
      "grad_norm": 0.25447434186935425,
      "learning_rate": 0.0001,
      "loss": 1.4867,
      "step": 4264
    },
    {
      "epoch": 0.684810533076429,
      "grad_norm": 0.2528344392776489,
      "learning_rate": 0.0001,
      "loss": 1.4459,
      "step": 4265
    },
    {
      "epoch": 0.684971098265896,
      "grad_norm": 0.2751440405845642,
      "learning_rate": 0.0001,
      "loss": 1.5573,
      "step": 4266
    },
    {
      "epoch": 0.6851316634553629,
      "grad_norm": 0.34374743700027466,
      "learning_rate": 0.0001,
      "loss": 1.4634,
      "step": 4267
    },
    {
      "epoch": 0.6852922286448297,
      "grad_norm": 0.4083437919616699,
      "learning_rate": 0.0001,
      "loss": 1.4515,
      "step": 4268
    },
    {
      "epoch": 0.6854527938342967,
      "grad_norm": 0.2677876651287079,
      "learning_rate": 0.0001,
      "loss": 1.573,
      "step": 4269
    },
    {
      "epoch": 0.6856133590237636,
      "grad_norm": 0.2757302224636078,
      "learning_rate": 0.0001,
      "loss": 1.6004,
      "step": 4270
    },
    {
      "epoch": 0.6857739242132306,
      "grad_norm": 0.2541894018650055,
      "learning_rate": 0.0001,
      "loss": 1.5002,
      "step": 4271
    },
    {
      "epoch": 0.6859344894026975,
      "grad_norm": 0.2496742159128189,
      "learning_rate": 0.0001,
      "loss": 1.4454,
      "step": 4272
    },
    {
      "epoch": 0.6860950545921645,
      "grad_norm": 0.27230650186538696,
      "learning_rate": 0.0001,
      "loss": 1.5389,
      "step": 4273
    },
    {
      "epoch": 0.6862556197816313,
      "grad_norm": 0.27860227227211,
      "learning_rate": 0.0001,
      "loss": 1.5486,
      "step": 4274
    },
    {
      "epoch": 0.6864161849710982,
      "grad_norm": 0.25846415758132935,
      "learning_rate": 0.0001,
      "loss": 1.427,
      "step": 4275
    },
    {
      "epoch": 0.6865767501605652,
      "grad_norm": 0.26338714361190796,
      "learning_rate": 0.0001,
      "loss": 1.5178,
      "step": 4276
    },
    {
      "epoch": 0.6867373153500321,
      "grad_norm": 0.26182571053504944,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 4277
    },
    {
      "epoch": 0.6868978805394991,
      "grad_norm": 0.26924988627433777,
      "learning_rate": 0.0001,
      "loss": 1.5526,
      "step": 4278
    },
    {
      "epoch": 0.6870584457289659,
      "grad_norm": 0.26171842217445374,
      "learning_rate": 0.0001,
      "loss": 1.4071,
      "step": 4279
    },
    {
      "epoch": 0.6872190109184328,
      "grad_norm": 0.28873947262763977,
      "learning_rate": 0.0001,
      "loss": 1.5053,
      "step": 4280
    },
    {
      "epoch": 0.6873795761078998,
      "grad_norm": 0.26880842447280884,
      "learning_rate": 0.0001,
      "loss": 1.4269,
      "step": 4281
    },
    {
      "epoch": 0.6875401412973667,
      "grad_norm": 0.274006187915802,
      "learning_rate": 0.0001,
      "loss": 1.4469,
      "step": 4282
    },
    {
      "epoch": 0.6877007064868337,
      "grad_norm": 0.27234819531440735,
      "learning_rate": 0.0001,
      "loss": 1.4983,
      "step": 4283
    },
    {
      "epoch": 0.6878612716763006,
      "grad_norm": 0.2691795229911804,
      "learning_rate": 0.0001,
      "loss": 1.5159,
      "step": 4284
    },
    {
      "epoch": 0.6880218368657675,
      "grad_norm": 0.25503280758857727,
      "learning_rate": 0.0001,
      "loss": 1.5067,
      "step": 4285
    },
    {
      "epoch": 0.6881824020552344,
      "grad_norm": 0.264271080493927,
      "learning_rate": 0.0001,
      "loss": 1.5302,
      "step": 4286
    },
    {
      "epoch": 0.6883429672447013,
      "grad_norm": 0.26026618480682373,
      "learning_rate": 0.0001,
      "loss": 1.5167,
      "step": 4287
    },
    {
      "epoch": 0.6885035324341683,
      "grad_norm": 0.2942197918891907,
      "learning_rate": 0.0001,
      "loss": 1.5174,
      "step": 4288
    },
    {
      "epoch": 0.6886640976236352,
      "grad_norm": 0.26720893383026123,
      "learning_rate": 0.0001,
      "loss": 1.4347,
      "step": 4289
    },
    {
      "epoch": 0.6888246628131022,
      "grad_norm": 0.26280948519706726,
      "learning_rate": 0.0001,
      "loss": 1.4533,
      "step": 4290
    },
    {
      "epoch": 0.688985228002569,
      "grad_norm": 0.27599456906318665,
      "learning_rate": 0.0001,
      "loss": 1.5484,
      "step": 4291
    },
    {
      "epoch": 0.689145793192036,
      "grad_norm": 0.27223873138427734,
      "learning_rate": 0.0001,
      "loss": 1.5116,
      "step": 4292
    },
    {
      "epoch": 0.6893063583815029,
      "grad_norm": 0.26796194911003113,
      "learning_rate": 0.0001,
      "loss": 1.4582,
      "step": 4293
    },
    {
      "epoch": 0.6894669235709698,
      "grad_norm": 0.2743043303489685,
      "learning_rate": 0.0001,
      "loss": 1.4849,
      "step": 4294
    },
    {
      "epoch": 0.6896274887604368,
      "grad_norm": 0.2695828974246979,
      "learning_rate": 0.0001,
      "loss": 1.5353,
      "step": 4295
    },
    {
      "epoch": 0.6897880539499036,
      "grad_norm": 0.27552083134651184,
      "learning_rate": 0.0001,
      "loss": 1.4595,
      "step": 4296
    },
    {
      "epoch": 0.6899486191393706,
      "grad_norm": 0.2738676369190216,
      "learning_rate": 0.0001,
      "loss": 1.5096,
      "step": 4297
    },
    {
      "epoch": 0.6901091843288375,
      "grad_norm": 0.42212623357772827,
      "learning_rate": 0.0001,
      "loss": 1.4909,
      "step": 4298
    },
    {
      "epoch": 0.6902697495183044,
      "grad_norm": 0.25751927495002747,
      "learning_rate": 0.0001,
      "loss": 1.5218,
      "step": 4299
    },
    {
      "epoch": 0.6904303147077714,
      "grad_norm": 0.291474848985672,
      "learning_rate": 0.0001,
      "loss": 1.5342,
      "step": 4300
    },
    {
      "epoch": 0.6905908798972383,
      "grad_norm": 0.26581260561943054,
      "learning_rate": 0.0001,
      "loss": 1.4447,
      "step": 4301
    },
    {
      "epoch": 0.6907514450867052,
      "grad_norm": 0.27196231484413147,
      "learning_rate": 0.0001,
      "loss": 1.4482,
      "step": 4302
    },
    {
      "epoch": 0.6909120102761721,
      "grad_norm": 0.26516249775886536,
      "learning_rate": 0.0001,
      "loss": 1.4661,
      "step": 4303
    },
    {
      "epoch": 0.691072575465639,
      "grad_norm": 0.2661803960800171,
      "learning_rate": 0.0001,
      "loss": 1.4621,
      "step": 4304
    },
    {
      "epoch": 0.691233140655106,
      "grad_norm": 0.26079392433166504,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 4305
    },
    {
      "epoch": 0.6913937058445729,
      "grad_norm": 0.27813488245010376,
      "learning_rate": 0.0001,
      "loss": 1.5284,
      "step": 4306
    },
    {
      "epoch": 0.6915542710340398,
      "grad_norm": 0.26817786693573,
      "learning_rate": 0.0001,
      "loss": 1.5603,
      "step": 4307
    },
    {
      "epoch": 0.6917148362235067,
      "grad_norm": 0.26582929491996765,
      "learning_rate": 0.0001,
      "loss": 1.4259,
      "step": 4308
    },
    {
      "epoch": 0.6918754014129737,
      "grad_norm": 0.26686498522758484,
      "learning_rate": 0.0001,
      "loss": 1.5148,
      "step": 4309
    },
    {
      "epoch": 0.6920359666024406,
      "grad_norm": 0.25720271468162537,
      "learning_rate": 0.0001,
      "loss": 1.4958,
      "step": 4310
    },
    {
      "epoch": 0.6921965317919075,
      "grad_norm": 0.2693347632884979,
      "learning_rate": 0.0001,
      "loss": 1.4241,
      "step": 4311
    },
    {
      "epoch": 0.6923570969813745,
      "grad_norm": 0.26887163519859314,
      "learning_rate": 0.0001,
      "loss": 1.5274,
      "step": 4312
    },
    {
      "epoch": 0.6925176621708413,
      "grad_norm": 0.2574673891067505,
      "learning_rate": 0.0001,
      "loss": 1.5347,
      "step": 4313
    },
    {
      "epoch": 0.6926782273603083,
      "grad_norm": 0.2512687146663666,
      "learning_rate": 0.0001,
      "loss": 1.3923,
      "step": 4314
    },
    {
      "epoch": 0.6928387925497752,
      "grad_norm": 0.288910835981369,
      "learning_rate": 0.0001,
      "loss": 1.4973,
      "step": 4315
    },
    {
      "epoch": 0.6929993577392421,
      "grad_norm": 0.27544349431991577,
      "learning_rate": 0.0001,
      "loss": 1.532,
      "step": 4316
    },
    {
      "epoch": 0.6931599229287091,
      "grad_norm": 0.25817129015922546,
      "learning_rate": 0.0001,
      "loss": 1.5234,
      "step": 4317
    },
    {
      "epoch": 0.6933204881181759,
      "grad_norm": 0.2737163007259369,
      "learning_rate": 0.0001,
      "loss": 1.4919,
      "step": 4318
    },
    {
      "epoch": 0.6934810533076429,
      "grad_norm": 0.26792851090431213,
      "learning_rate": 0.0001,
      "loss": 1.4809,
      "step": 4319
    },
    {
      "epoch": 0.6936416184971098,
      "grad_norm": 0.26003244519233704,
      "learning_rate": 0.0001,
      "loss": 1.4815,
      "step": 4320
    },
    {
      "epoch": 0.6938021836865768,
      "grad_norm": 0.25619208812713623,
      "learning_rate": 0.0001,
      "loss": 1.4865,
      "step": 4321
    },
    {
      "epoch": 0.6939627488760437,
      "grad_norm": 0.26232337951660156,
      "learning_rate": 0.0001,
      "loss": 1.5054,
      "step": 4322
    },
    {
      "epoch": 0.6941233140655106,
      "grad_norm": 0.2903430461883545,
      "learning_rate": 0.0001,
      "loss": 1.429,
      "step": 4323
    },
    {
      "epoch": 0.6942838792549775,
      "grad_norm": 0.2574857473373413,
      "learning_rate": 0.0001,
      "loss": 1.4722,
      "step": 4324
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.2702174782752991,
      "learning_rate": 0.0001,
      "loss": 1.4795,
      "step": 4325
    },
    {
      "epoch": 0.6946050096339114,
      "grad_norm": 0.2641509771347046,
      "learning_rate": 0.0001,
      "loss": 1.4928,
      "step": 4326
    },
    {
      "epoch": 0.6947655748233783,
      "grad_norm": 0.2596450448036194,
      "learning_rate": 0.0001,
      "loss": 1.5512,
      "step": 4327
    },
    {
      "epoch": 0.6949261400128453,
      "grad_norm": 0.2672803997993469,
      "learning_rate": 0.0001,
      "loss": 1.5115,
      "step": 4328
    },
    {
      "epoch": 0.6950867052023122,
      "grad_norm": 0.27464133501052856,
      "learning_rate": 0.0001,
      "loss": 1.5322,
      "step": 4329
    },
    {
      "epoch": 0.695247270391779,
      "grad_norm": 0.26365870237350464,
      "learning_rate": 0.0001,
      "loss": 1.5631,
      "step": 4330
    },
    {
      "epoch": 0.695407835581246,
      "grad_norm": 0.2679399847984314,
      "learning_rate": 0.0001,
      "loss": 1.5365,
      "step": 4331
    },
    {
      "epoch": 0.6955684007707129,
      "grad_norm": 0.2631140649318695,
      "learning_rate": 0.0001,
      "loss": 1.4632,
      "step": 4332
    },
    {
      "epoch": 0.6957289659601799,
      "grad_norm": 0.2667805850505829,
      "learning_rate": 0.0001,
      "loss": 1.5183,
      "step": 4333
    },
    {
      "epoch": 0.6958895311496468,
      "grad_norm": 0.25944942235946655,
      "learning_rate": 0.0001,
      "loss": 1.4965,
      "step": 4334
    },
    {
      "epoch": 0.6960500963391136,
      "grad_norm": 0.25462979078292847,
      "learning_rate": 0.0001,
      "loss": 1.5036,
      "step": 4335
    },
    {
      "epoch": 0.6962106615285806,
      "grad_norm": 0.26245835423469543,
      "learning_rate": 0.0001,
      "loss": 1.4849,
      "step": 4336
    },
    {
      "epoch": 0.6963712267180475,
      "grad_norm": 0.26964670419692993,
      "learning_rate": 0.0001,
      "loss": 1.5723,
      "step": 4337
    },
    {
      "epoch": 0.6965317919075145,
      "grad_norm": 0.2651227116584778,
      "learning_rate": 0.0001,
      "loss": 1.5487,
      "step": 4338
    },
    {
      "epoch": 0.6966923570969814,
      "grad_norm": 0.30446404218673706,
      "learning_rate": 0.0001,
      "loss": 1.5096,
      "step": 4339
    },
    {
      "epoch": 0.6968529222864484,
      "grad_norm": 0.27518704533576965,
      "learning_rate": 0.0001,
      "loss": 1.5842,
      "step": 4340
    },
    {
      "epoch": 0.6970134874759152,
      "grad_norm": 0.25260788202285767,
      "learning_rate": 0.0001,
      "loss": 1.4444,
      "step": 4341
    },
    {
      "epoch": 0.6971740526653821,
      "grad_norm": 0.25623124837875366,
      "learning_rate": 0.0001,
      "loss": 1.3767,
      "step": 4342
    },
    {
      "epoch": 0.6973346178548491,
      "grad_norm": 0.2668977975845337,
      "learning_rate": 0.0001,
      "loss": 1.5001,
      "step": 4343
    },
    {
      "epoch": 0.697495183044316,
      "grad_norm": 0.27028948068618774,
      "learning_rate": 0.0001,
      "loss": 1.4352,
      "step": 4344
    },
    {
      "epoch": 0.697655748233783,
      "grad_norm": 0.27070972323417664,
      "learning_rate": 0.0001,
      "loss": 1.6113,
      "step": 4345
    },
    {
      "epoch": 0.6978163134232498,
      "grad_norm": 0.27184128761291504,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 4346
    },
    {
      "epoch": 0.6979768786127167,
      "grad_norm": 0.2648637294769287,
      "learning_rate": 0.0001,
      "loss": 1.5043,
      "step": 4347
    },
    {
      "epoch": 0.6981374438021837,
      "grad_norm": 0.2720525562763214,
      "learning_rate": 0.0001,
      "loss": 1.5435,
      "step": 4348
    },
    {
      "epoch": 0.6982980089916506,
      "grad_norm": 0.25799763202667236,
      "learning_rate": 0.0001,
      "loss": 1.4338,
      "step": 4349
    },
    {
      "epoch": 0.6984585741811176,
      "grad_norm": 0.2532768249511719,
      "learning_rate": 0.0001,
      "loss": 1.4077,
      "step": 4350
    },
    {
      "epoch": 0.6986191393705845,
      "grad_norm": 0.2617965042591095,
      "learning_rate": 0.0001,
      "loss": 1.503,
      "step": 4351
    },
    {
      "epoch": 0.6987797045600513,
      "grad_norm": 0.27117711305618286,
      "learning_rate": 0.0001,
      "loss": 1.5566,
      "step": 4352
    },
    {
      "epoch": 0.6989402697495183,
      "grad_norm": 0.26747074723243713,
      "learning_rate": 0.0001,
      "loss": 1.4427,
      "step": 4353
    },
    {
      "epoch": 0.6991008349389852,
      "grad_norm": 0.2636476159095764,
      "learning_rate": 0.0001,
      "loss": 1.5027,
      "step": 4354
    },
    {
      "epoch": 0.6992614001284522,
      "grad_norm": 0.2568819522857666,
      "learning_rate": 0.0001,
      "loss": 1.4629,
      "step": 4355
    },
    {
      "epoch": 0.6994219653179191,
      "grad_norm": 0.2751426100730896,
      "learning_rate": 0.0001,
      "loss": 1.5391,
      "step": 4356
    },
    {
      "epoch": 0.699582530507386,
      "grad_norm": 0.2704501152038574,
      "learning_rate": 0.0001,
      "loss": 1.579,
      "step": 4357
    },
    {
      "epoch": 0.6997430956968529,
      "grad_norm": 0.2700333595275879,
      "learning_rate": 0.0001,
      "loss": 1.616,
      "step": 4358
    },
    {
      "epoch": 0.6999036608863198,
      "grad_norm": 0.268250435590744,
      "learning_rate": 0.0001,
      "loss": 1.5191,
      "step": 4359
    },
    {
      "epoch": 0.7000642260757868,
      "grad_norm": 0.26120874285697937,
      "learning_rate": 0.0001,
      "loss": 1.4953,
      "step": 4360
    },
    {
      "epoch": 0.7002247912652537,
      "grad_norm": 0.25803521275520325,
      "learning_rate": 0.0001,
      "loss": 1.4986,
      "step": 4361
    },
    {
      "epoch": 0.7003853564547207,
      "grad_norm": 0.2690806984901428,
      "learning_rate": 0.0001,
      "loss": 1.5414,
      "step": 4362
    },
    {
      "epoch": 0.7005459216441875,
      "grad_norm": 0.25062671303749084,
      "learning_rate": 0.0001,
      "loss": 1.4683,
      "step": 4363
    },
    {
      "epoch": 0.7007064868336544,
      "grad_norm": 0.265217661857605,
      "learning_rate": 0.0001,
      "loss": 1.5582,
      "step": 4364
    },
    {
      "epoch": 0.7008670520231214,
      "grad_norm": 0.2658994495868683,
      "learning_rate": 0.0001,
      "loss": 1.396,
      "step": 4365
    },
    {
      "epoch": 0.7010276172125883,
      "grad_norm": 0.2500857710838318,
      "learning_rate": 0.0001,
      "loss": 1.4334,
      "step": 4366
    },
    {
      "epoch": 0.7011881824020553,
      "grad_norm": 0.26173529028892517,
      "learning_rate": 0.0001,
      "loss": 1.48,
      "step": 4367
    },
    {
      "epoch": 0.7013487475915221,
      "grad_norm": 0.2551463544368744,
      "learning_rate": 0.0001,
      "loss": 1.4782,
      "step": 4368
    },
    {
      "epoch": 0.701509312780989,
      "grad_norm": 0.2488209754228592,
      "learning_rate": 0.0001,
      "loss": 1.3831,
      "step": 4369
    },
    {
      "epoch": 0.701669877970456,
      "grad_norm": 0.25642555952072144,
      "learning_rate": 0.0001,
      "loss": 1.5492,
      "step": 4370
    },
    {
      "epoch": 0.7018304431599229,
      "grad_norm": 0.2563942074775696,
      "learning_rate": 0.0001,
      "loss": 1.5025,
      "step": 4371
    },
    {
      "epoch": 0.7019910083493899,
      "grad_norm": 0.26586371660232544,
      "learning_rate": 0.0001,
      "loss": 1.4865,
      "step": 4372
    },
    {
      "epoch": 0.7021515735388568,
      "grad_norm": 0.25640198588371277,
      "learning_rate": 0.0001,
      "loss": 1.4257,
      "step": 4373
    },
    {
      "epoch": 0.7023121387283237,
      "grad_norm": 0.30186402797698975,
      "learning_rate": 0.0001,
      "loss": 1.515,
      "step": 4374
    },
    {
      "epoch": 0.7024727039177906,
      "grad_norm": 0.26809120178222656,
      "learning_rate": 0.0001,
      "loss": 1.4723,
      "step": 4375
    },
    {
      "epoch": 0.7026332691072575,
      "grad_norm": 0.2669029235839844,
      "learning_rate": 0.0001,
      "loss": 1.459,
      "step": 4376
    },
    {
      "epoch": 0.7027938342967245,
      "grad_norm": 0.2582818269729614,
      "learning_rate": 0.0001,
      "loss": 1.5551,
      "step": 4377
    },
    {
      "epoch": 0.7029543994861914,
      "grad_norm": 0.26471009850502014,
      "learning_rate": 0.0001,
      "loss": 1.4531,
      "step": 4378
    },
    {
      "epoch": 0.7031149646756584,
      "grad_norm": 0.25546589493751526,
      "learning_rate": 0.0001,
      "loss": 1.5049,
      "step": 4379
    },
    {
      "epoch": 0.7032755298651252,
      "grad_norm": 0.268154501914978,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 4380
    },
    {
      "epoch": 0.7034360950545921,
      "grad_norm": 0.25864431262016296,
      "learning_rate": 0.0001,
      "loss": 1.531,
      "step": 4381
    },
    {
      "epoch": 0.7035966602440591,
      "grad_norm": 0.2717253863811493,
      "learning_rate": 0.0001,
      "loss": 1.5211,
      "step": 4382
    },
    {
      "epoch": 0.703757225433526,
      "grad_norm": 0.2777341902256012,
      "learning_rate": 0.0001,
      "loss": 1.5082,
      "step": 4383
    },
    {
      "epoch": 0.703917790622993,
      "grad_norm": 0.26059722900390625,
      "learning_rate": 0.0001,
      "loss": 1.4994,
      "step": 4384
    },
    {
      "epoch": 0.7040783558124598,
      "grad_norm": 0.27167803049087524,
      "learning_rate": 0.0001,
      "loss": 1.551,
      "step": 4385
    },
    {
      "epoch": 0.7042389210019268,
      "grad_norm": 0.26073744893074036,
      "learning_rate": 0.0001,
      "loss": 1.4914,
      "step": 4386
    },
    {
      "epoch": 0.7043994861913937,
      "grad_norm": 0.2469278872013092,
      "learning_rate": 0.0001,
      "loss": 1.5015,
      "step": 4387
    },
    {
      "epoch": 0.7045600513808606,
      "grad_norm": 0.26496291160583496,
      "learning_rate": 0.0001,
      "loss": 1.5747,
      "step": 4388
    },
    {
      "epoch": 0.7047206165703276,
      "grad_norm": 0.2571849524974823,
      "learning_rate": 0.0001,
      "loss": 1.5013,
      "step": 4389
    },
    {
      "epoch": 0.7048811817597945,
      "grad_norm": 0.2778785824775696,
      "learning_rate": 0.0001,
      "loss": 1.4337,
      "step": 4390
    },
    {
      "epoch": 0.7050417469492614,
      "grad_norm": 0.2539641559123993,
      "learning_rate": 0.0001,
      "loss": 1.537,
      "step": 4391
    },
    {
      "epoch": 0.7052023121387283,
      "grad_norm": 0.27053964138031006,
      "learning_rate": 0.0001,
      "loss": 1.3811,
      "step": 4392
    },
    {
      "epoch": 0.7053628773281952,
      "grad_norm": 0.2752392292022705,
      "learning_rate": 0.0001,
      "loss": 1.4962,
      "step": 4393
    },
    {
      "epoch": 0.7055234425176622,
      "grad_norm": 0.25827983021736145,
      "learning_rate": 0.0001,
      "loss": 1.5302,
      "step": 4394
    },
    {
      "epoch": 0.7056840077071291,
      "grad_norm": 0.2535262405872345,
      "learning_rate": 0.0001,
      "loss": 1.4865,
      "step": 4395
    },
    {
      "epoch": 0.705844572896596,
      "grad_norm": 0.2754910886287689,
      "learning_rate": 0.0001,
      "loss": 1.4842,
      "step": 4396
    },
    {
      "epoch": 0.7060051380860629,
      "grad_norm": 0.26365748047828674,
      "learning_rate": 0.0001,
      "loss": 1.4433,
      "step": 4397
    },
    {
      "epoch": 0.7061657032755299,
      "grad_norm": 0.2513587176799774,
      "learning_rate": 0.0001,
      "loss": 1.4889,
      "step": 4398
    },
    {
      "epoch": 0.7063262684649968,
      "grad_norm": 0.25660595297813416,
      "learning_rate": 0.0001,
      "loss": 1.4839,
      "step": 4399
    },
    {
      "epoch": 0.7064868336544637,
      "grad_norm": 0.2599651515483856,
      "learning_rate": 0.0001,
      "loss": 1.4655,
      "step": 4400
    },
    {
      "epoch": 0.7066473988439307,
      "grad_norm": 0.27303239703178406,
      "learning_rate": 0.0001,
      "loss": 1.5043,
      "step": 4401
    },
    {
      "epoch": 0.7068079640333975,
      "grad_norm": 0.27233800292015076,
      "learning_rate": 0.0001,
      "loss": 1.4356,
      "step": 4402
    },
    {
      "epoch": 0.7069685292228645,
      "grad_norm": 0.25332629680633545,
      "learning_rate": 0.0001,
      "loss": 1.4662,
      "step": 4403
    },
    {
      "epoch": 0.7071290944123314,
      "grad_norm": 0.2634056806564331,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 4404
    },
    {
      "epoch": 0.7072896596017983,
      "grad_norm": 0.26729610562324524,
      "learning_rate": 0.0001,
      "loss": 1.5289,
      "step": 4405
    },
    {
      "epoch": 0.7074502247912653,
      "grad_norm": 0.27977415919303894,
      "learning_rate": 0.0001,
      "loss": 1.58,
      "step": 4406
    },
    {
      "epoch": 0.7076107899807321,
      "grad_norm": 0.25852587819099426,
      "learning_rate": 0.0001,
      "loss": 1.4558,
      "step": 4407
    },
    {
      "epoch": 0.7077713551701991,
      "grad_norm": 0.2548057734966278,
      "learning_rate": 0.0001,
      "loss": 1.4733,
      "step": 4408
    },
    {
      "epoch": 0.707931920359666,
      "grad_norm": 0.26761743426322937,
      "learning_rate": 0.0001,
      "loss": 1.4502,
      "step": 4409
    },
    {
      "epoch": 0.708092485549133,
      "grad_norm": 0.2621440589427948,
      "learning_rate": 0.0001,
      "loss": 1.5589,
      "step": 4410
    },
    {
      "epoch": 0.7082530507385999,
      "grad_norm": 0.25833815336227417,
      "learning_rate": 0.0001,
      "loss": 1.4941,
      "step": 4411
    },
    {
      "epoch": 0.7084136159280668,
      "grad_norm": 0.25731223821640015,
      "learning_rate": 0.0001,
      "loss": 1.5585,
      "step": 4412
    },
    {
      "epoch": 0.7085741811175337,
      "grad_norm": 0.26889559626579285,
      "learning_rate": 0.0001,
      "loss": 1.5231,
      "step": 4413
    },
    {
      "epoch": 0.7087347463070006,
      "grad_norm": 0.25878679752349854,
      "learning_rate": 0.0001,
      "loss": 1.5026,
      "step": 4414
    },
    {
      "epoch": 0.7088953114964676,
      "grad_norm": 0.2717954218387604,
      "learning_rate": 0.0001,
      "loss": 1.5714,
      "step": 4415
    },
    {
      "epoch": 0.7090558766859345,
      "grad_norm": 0.26209917664527893,
      "learning_rate": 0.0001,
      "loss": 1.4316,
      "step": 4416
    },
    {
      "epoch": 0.7092164418754014,
      "grad_norm": 0.27343231439590454,
      "learning_rate": 0.0001,
      "loss": 1.5084,
      "step": 4417
    },
    {
      "epoch": 0.7093770070648683,
      "grad_norm": 0.27332833409309387,
      "learning_rate": 0.0001,
      "loss": 1.5473,
      "step": 4418
    },
    {
      "epoch": 0.7095375722543352,
      "grad_norm": 0.2737707495689392,
      "learning_rate": 0.0001,
      "loss": 1.4728,
      "step": 4419
    },
    {
      "epoch": 0.7096981374438022,
      "grad_norm": 0.2730090618133545,
      "learning_rate": 0.0001,
      "loss": 1.5087,
      "step": 4420
    },
    {
      "epoch": 0.7098587026332691,
      "grad_norm": 0.27479568123817444,
      "learning_rate": 0.0001,
      "loss": 1.5424,
      "step": 4421
    },
    {
      "epoch": 0.710019267822736,
      "grad_norm": 0.2649615406990051,
      "learning_rate": 0.0001,
      "loss": 1.449,
      "step": 4422
    },
    {
      "epoch": 0.710179833012203,
      "grad_norm": 0.28063979744911194,
      "learning_rate": 0.0001,
      "loss": 1.5263,
      "step": 4423
    },
    {
      "epoch": 0.7103403982016698,
      "grad_norm": 0.27927061915397644,
      "learning_rate": 0.0001,
      "loss": 1.4998,
      "step": 4424
    },
    {
      "epoch": 0.7105009633911368,
      "grad_norm": 0.25157079100608826,
      "learning_rate": 0.0001,
      "loss": 1.4429,
      "step": 4425
    },
    {
      "epoch": 0.7106615285806037,
      "grad_norm": 0.3160429894924164,
      "learning_rate": 0.0001,
      "loss": 1.5159,
      "step": 4426
    },
    {
      "epoch": 0.7108220937700707,
      "grad_norm": 0.28867828845977783,
      "learning_rate": 0.0001,
      "loss": 1.4464,
      "step": 4427
    },
    {
      "epoch": 0.7109826589595376,
      "grad_norm": 0.2587195038795471,
      "learning_rate": 0.0001,
      "loss": 1.4863,
      "step": 4428
    },
    {
      "epoch": 0.7111432241490045,
      "grad_norm": 0.2741629183292389,
      "learning_rate": 0.0001,
      "loss": 1.4089,
      "step": 4429
    },
    {
      "epoch": 0.7113037893384714,
      "grad_norm": 0.26618364453315735,
      "learning_rate": 0.0001,
      "loss": 1.5135,
      "step": 4430
    },
    {
      "epoch": 0.7114643545279383,
      "grad_norm": 0.2516707479953766,
      "learning_rate": 0.0001,
      "loss": 1.4103,
      "step": 4431
    },
    {
      "epoch": 0.7116249197174053,
      "grad_norm": 0.265982985496521,
      "learning_rate": 0.0001,
      "loss": 1.5198,
      "step": 4432
    },
    {
      "epoch": 0.7117854849068722,
      "grad_norm": 0.26953351497650146,
      "learning_rate": 0.0001,
      "loss": 1.4967,
      "step": 4433
    },
    {
      "epoch": 0.7119460500963392,
      "grad_norm": 0.271270751953125,
      "learning_rate": 0.0001,
      "loss": 1.4891,
      "step": 4434
    },
    {
      "epoch": 0.712106615285806,
      "grad_norm": 0.2619115710258484,
      "learning_rate": 0.0001,
      "loss": 1.5203,
      "step": 4435
    },
    {
      "epoch": 0.7122671804752729,
      "grad_norm": 0.26357564330101013,
      "learning_rate": 0.0001,
      "loss": 1.4419,
      "step": 4436
    },
    {
      "epoch": 0.7124277456647399,
      "grad_norm": 0.27811411023139954,
      "learning_rate": 0.0001,
      "loss": 1.5185,
      "step": 4437
    },
    {
      "epoch": 0.7125883108542068,
      "grad_norm": 0.27106788754463196,
      "learning_rate": 0.0001,
      "loss": 1.4602,
      "step": 4438
    },
    {
      "epoch": 0.7127488760436738,
      "grad_norm": 0.27602043747901917,
      "learning_rate": 0.0001,
      "loss": 1.4671,
      "step": 4439
    },
    {
      "epoch": 0.7129094412331407,
      "grad_norm": 0.28530940413475037,
      "learning_rate": 0.0001,
      "loss": 1.5047,
      "step": 4440
    },
    {
      "epoch": 0.7130700064226075,
      "grad_norm": 0.2758553624153137,
      "learning_rate": 0.0001,
      "loss": 1.5065,
      "step": 4441
    },
    {
      "epoch": 0.7132305716120745,
      "grad_norm": 0.2505316436290741,
      "learning_rate": 0.0001,
      "loss": 1.3736,
      "step": 4442
    },
    {
      "epoch": 0.7133911368015414,
      "grad_norm": 0.25257694721221924,
      "learning_rate": 0.0001,
      "loss": 1.5111,
      "step": 4443
    },
    {
      "epoch": 0.7135517019910084,
      "grad_norm": 0.3961699306964874,
      "learning_rate": 0.0001,
      "loss": 1.4508,
      "step": 4444
    },
    {
      "epoch": 0.7137122671804753,
      "grad_norm": 0.25373756885528564,
      "learning_rate": 0.0001,
      "loss": 1.4998,
      "step": 4445
    },
    {
      "epoch": 0.7138728323699421,
      "grad_norm": 0.2606360614299774,
      "learning_rate": 0.0001,
      "loss": 1.5664,
      "step": 4446
    },
    {
      "epoch": 0.7140333975594091,
      "grad_norm": 0.26831522583961487,
      "learning_rate": 0.0001,
      "loss": 1.511,
      "step": 4447
    },
    {
      "epoch": 0.714193962748876,
      "grad_norm": 0.2633364498615265,
      "learning_rate": 0.0001,
      "loss": 1.4681,
      "step": 4448
    },
    {
      "epoch": 0.714354527938343,
      "grad_norm": 0.27485334873199463,
      "learning_rate": 0.0001,
      "loss": 1.5337,
      "step": 4449
    },
    {
      "epoch": 0.7145150931278099,
      "grad_norm": 0.2693502604961395,
      "learning_rate": 0.0001,
      "loss": 1.4955,
      "step": 4450
    },
    {
      "epoch": 0.7146756583172769,
      "grad_norm": 0.2679779529571533,
      "learning_rate": 0.0001,
      "loss": 1.42,
      "step": 4451
    },
    {
      "epoch": 0.7148362235067437,
      "grad_norm": 0.2692188024520874,
      "learning_rate": 0.0001,
      "loss": 1.4819,
      "step": 4452
    },
    {
      "epoch": 0.7149967886962106,
      "grad_norm": 0.24806596338748932,
      "learning_rate": 0.0001,
      "loss": 1.4835,
      "step": 4453
    },
    {
      "epoch": 0.7151573538856776,
      "grad_norm": 0.2935852110385895,
      "learning_rate": 0.0001,
      "loss": 1.5861,
      "step": 4454
    },
    {
      "epoch": 0.7153179190751445,
      "grad_norm": 0.26009488105773926,
      "learning_rate": 0.0001,
      "loss": 1.5451,
      "step": 4455
    },
    {
      "epoch": 0.7154784842646115,
      "grad_norm": 0.2731396555900574,
      "learning_rate": 0.0001,
      "loss": 1.5699,
      "step": 4456
    },
    {
      "epoch": 0.7156390494540783,
      "grad_norm": 0.28009775280952454,
      "learning_rate": 0.0001,
      "loss": 1.4371,
      "step": 4457
    },
    {
      "epoch": 0.7157996146435452,
      "grad_norm": 0.26409244537353516,
      "learning_rate": 0.0001,
      "loss": 1.5191,
      "step": 4458
    },
    {
      "epoch": 0.7159601798330122,
      "grad_norm": 0.2744426429271698,
      "learning_rate": 0.0001,
      "loss": 1.5256,
      "step": 4459
    },
    {
      "epoch": 0.7161207450224791,
      "grad_norm": 0.2576165795326233,
      "learning_rate": 0.0001,
      "loss": 1.4554,
      "step": 4460
    },
    {
      "epoch": 0.7162813102119461,
      "grad_norm": 0.26500216126441956,
      "learning_rate": 0.0001,
      "loss": 1.5102,
      "step": 4461
    },
    {
      "epoch": 0.716441875401413,
      "grad_norm": 0.29779383540153503,
      "learning_rate": 0.0001,
      "loss": 1.5347,
      "step": 4462
    },
    {
      "epoch": 0.7166024405908799,
      "grad_norm": 0.2547025680541992,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 4463
    },
    {
      "epoch": 0.7167630057803468,
      "grad_norm": 0.298776239156723,
      "learning_rate": 0.0001,
      "loss": 1.5235,
      "step": 4464
    },
    {
      "epoch": 0.7169235709698137,
      "grad_norm": 0.2800496816635132,
      "learning_rate": 0.0001,
      "loss": 1.4428,
      "step": 4465
    },
    {
      "epoch": 0.7170841361592807,
      "grad_norm": 0.2626544237136841,
      "learning_rate": 0.0001,
      "loss": 1.4819,
      "step": 4466
    },
    {
      "epoch": 0.7172447013487476,
      "grad_norm": 0.25592851638793945,
      "learning_rate": 0.0001,
      "loss": 1.49,
      "step": 4467
    },
    {
      "epoch": 0.7174052665382146,
      "grad_norm": 0.2628372609615326,
      "learning_rate": 0.0001,
      "loss": 1.3694,
      "step": 4468
    },
    {
      "epoch": 0.7175658317276814,
      "grad_norm": 0.2570657730102539,
      "learning_rate": 0.0001,
      "loss": 1.4635,
      "step": 4469
    },
    {
      "epoch": 0.7177263969171483,
      "grad_norm": 0.25978749990463257,
      "learning_rate": 0.0001,
      "loss": 1.5157,
      "step": 4470
    },
    {
      "epoch": 0.7178869621066153,
      "grad_norm": 0.259258896112442,
      "learning_rate": 0.0001,
      "loss": 1.5179,
      "step": 4471
    },
    {
      "epoch": 0.7180475272960822,
      "grad_norm": 0.25281277298927307,
      "learning_rate": 0.0001,
      "loss": 1.4334,
      "step": 4472
    },
    {
      "epoch": 0.7182080924855492,
      "grad_norm": 0.2575565278530121,
      "learning_rate": 0.0001,
      "loss": 1.433,
      "step": 4473
    },
    {
      "epoch": 0.718368657675016,
      "grad_norm": 0.2693718373775482,
      "learning_rate": 0.0001,
      "loss": 1.3869,
      "step": 4474
    },
    {
      "epoch": 0.718529222864483,
      "grad_norm": 0.2633496820926666,
      "learning_rate": 0.0001,
      "loss": 1.5227,
      "step": 4475
    },
    {
      "epoch": 0.7186897880539499,
      "grad_norm": 0.2422046959400177,
      "learning_rate": 0.0001,
      "loss": 1.3841,
      "step": 4476
    },
    {
      "epoch": 0.7188503532434168,
      "grad_norm": 0.2973977029323578,
      "learning_rate": 0.0001,
      "loss": 1.5567,
      "step": 4477
    },
    {
      "epoch": 0.7190109184328838,
      "grad_norm": 0.2742076516151428,
      "learning_rate": 0.0001,
      "loss": 1.4491,
      "step": 4478
    },
    {
      "epoch": 0.7191714836223507,
      "grad_norm": 0.2550565302371979,
      "learning_rate": 0.0001,
      "loss": 1.4604,
      "step": 4479
    },
    {
      "epoch": 0.7193320488118176,
      "grad_norm": 0.2700750231742859,
      "learning_rate": 0.0001,
      "loss": 1.5673,
      "step": 4480
    },
    {
      "epoch": 0.7194926140012845,
      "grad_norm": 0.2693113386631012,
      "learning_rate": 0.0001,
      "loss": 1.5352,
      "step": 4481
    },
    {
      "epoch": 0.7196531791907514,
      "grad_norm": 0.2644595801830292,
      "learning_rate": 0.0001,
      "loss": 1.5251,
      "step": 4482
    },
    {
      "epoch": 0.7198137443802184,
      "grad_norm": 0.25771331787109375,
      "learning_rate": 0.0001,
      "loss": 1.5294,
      "step": 4483
    },
    {
      "epoch": 0.7199743095696853,
      "grad_norm": 0.2605074942111969,
      "learning_rate": 0.0001,
      "loss": 1.4934,
      "step": 4484
    },
    {
      "epoch": 0.7201348747591522,
      "grad_norm": 0.25524941086769104,
      "learning_rate": 0.0001,
      "loss": 1.5205,
      "step": 4485
    },
    {
      "epoch": 0.7202954399486191,
      "grad_norm": 0.2571198344230652,
      "learning_rate": 0.0001,
      "loss": 1.4342,
      "step": 4486
    },
    {
      "epoch": 0.720456005138086,
      "grad_norm": 0.282124787569046,
      "learning_rate": 0.0001,
      "loss": 1.5009,
      "step": 4487
    },
    {
      "epoch": 0.720616570327553,
      "grad_norm": 0.27443233132362366,
      "learning_rate": 0.0001,
      "loss": 1.4511,
      "step": 4488
    },
    {
      "epoch": 0.7207771355170199,
      "grad_norm": 0.2842823565006256,
      "learning_rate": 0.0001,
      "loss": 1.4834,
      "step": 4489
    },
    {
      "epoch": 0.7209377007064869,
      "grad_norm": 0.2627255320549011,
      "learning_rate": 0.0001,
      "loss": 1.5147,
      "step": 4490
    },
    {
      "epoch": 0.7210982658959537,
      "grad_norm": 0.25318068265914917,
      "learning_rate": 0.0001,
      "loss": 1.5308,
      "step": 4491
    },
    {
      "epoch": 0.7212588310854207,
      "grad_norm": 0.25988972187042236,
      "learning_rate": 0.0001,
      "loss": 1.4123,
      "step": 4492
    },
    {
      "epoch": 0.7214193962748876,
      "grad_norm": 0.2656155228614807,
      "learning_rate": 0.0001,
      "loss": 1.4676,
      "step": 4493
    },
    {
      "epoch": 0.7215799614643545,
      "grad_norm": 0.2692951261997223,
      "learning_rate": 0.0001,
      "loss": 1.4891,
      "step": 4494
    },
    {
      "epoch": 0.7217405266538215,
      "grad_norm": 0.267822802066803,
      "learning_rate": 0.0001,
      "loss": 1.5172,
      "step": 4495
    },
    {
      "epoch": 0.7219010918432883,
      "grad_norm": 0.26387912034988403,
      "learning_rate": 0.0001,
      "loss": 1.4045,
      "step": 4496
    },
    {
      "epoch": 0.7220616570327553,
      "grad_norm": 0.2647508978843689,
      "learning_rate": 0.0001,
      "loss": 1.4078,
      "step": 4497
    },
    {
      "epoch": 0.7222222222222222,
      "grad_norm": 0.27734246850013733,
      "learning_rate": 0.0001,
      "loss": 1.4583,
      "step": 4498
    },
    {
      "epoch": 0.7223827874116892,
      "grad_norm": 0.28359514474868774,
      "learning_rate": 0.0001,
      "loss": 1.5409,
      "step": 4499
    },
    {
      "epoch": 0.7225433526011561,
      "grad_norm": 0.26455989480018616,
      "learning_rate": 0.0001,
      "loss": 1.4627,
      "step": 4500
    },
    {
      "epoch": 0.722703917790623,
      "grad_norm": 0.28599298000335693,
      "learning_rate": 0.0001,
      "loss": 1.5766,
      "step": 4501
    },
    {
      "epoch": 0.7228644829800899,
      "grad_norm": 0.2679843604564667,
      "learning_rate": 0.0001,
      "loss": 1.5495,
      "step": 4502
    },
    {
      "epoch": 0.7230250481695568,
      "grad_norm": 0.25494423508644104,
      "learning_rate": 0.0001,
      "loss": 1.531,
      "step": 4503
    },
    {
      "epoch": 0.7231856133590238,
      "grad_norm": 0.2747451961040497,
      "learning_rate": 0.0001,
      "loss": 1.4802,
      "step": 4504
    },
    {
      "epoch": 0.7233461785484907,
      "grad_norm": 0.2605855464935303,
      "learning_rate": 0.0001,
      "loss": 1.5682,
      "step": 4505
    },
    {
      "epoch": 0.7235067437379576,
      "grad_norm": 0.27038636803627014,
      "learning_rate": 0.0001,
      "loss": 1.5644,
      "step": 4506
    },
    {
      "epoch": 0.7236673089274245,
      "grad_norm": 0.28356048464775085,
      "learning_rate": 0.0001,
      "loss": 1.5086,
      "step": 4507
    },
    {
      "epoch": 0.7238278741168914,
      "grad_norm": 0.2688811421394348,
      "learning_rate": 0.0001,
      "loss": 1.4767,
      "step": 4508
    },
    {
      "epoch": 0.7239884393063584,
      "grad_norm": 0.2893103063106537,
      "learning_rate": 0.0001,
      "loss": 1.5314,
      "step": 4509
    },
    {
      "epoch": 0.7241490044958253,
      "grad_norm": 0.27945476770401,
      "learning_rate": 0.0001,
      "loss": 1.5808,
      "step": 4510
    },
    {
      "epoch": 0.7243095696852923,
      "grad_norm": 0.2593180537223816,
      "learning_rate": 0.0001,
      "loss": 1.4916,
      "step": 4511
    },
    {
      "epoch": 0.7244701348747592,
      "grad_norm": 0.2758706212043762,
      "learning_rate": 0.0001,
      "loss": 1.4831,
      "step": 4512
    },
    {
      "epoch": 0.724630700064226,
      "grad_norm": 0.25680720806121826,
      "learning_rate": 0.0001,
      "loss": 1.4713,
      "step": 4513
    },
    {
      "epoch": 0.724791265253693,
      "grad_norm": 0.27224671840667725,
      "learning_rate": 0.0001,
      "loss": 1.4702,
      "step": 4514
    },
    {
      "epoch": 0.7249518304431599,
      "grad_norm": 0.3044149577617645,
      "learning_rate": 0.0001,
      "loss": 1.5064,
      "step": 4515
    },
    {
      "epoch": 0.7251123956326269,
      "grad_norm": 0.27164554595947266,
      "learning_rate": 0.0001,
      "loss": 1.5005,
      "step": 4516
    },
    {
      "epoch": 0.7252729608220938,
      "grad_norm": 0.276862770318985,
      "learning_rate": 0.0001,
      "loss": 1.4302,
      "step": 4517
    },
    {
      "epoch": 0.7254335260115607,
      "grad_norm": 0.267896831035614,
      "learning_rate": 0.0001,
      "loss": 1.5807,
      "step": 4518
    },
    {
      "epoch": 0.7255940912010276,
      "grad_norm": 0.28572964668273926,
      "learning_rate": 0.0001,
      "loss": 1.4622,
      "step": 4519
    },
    {
      "epoch": 0.7257546563904945,
      "grad_norm": 0.266260027885437,
      "learning_rate": 0.0001,
      "loss": 1.5402,
      "step": 4520
    },
    {
      "epoch": 0.7259152215799615,
      "grad_norm": 0.2602228820323944,
      "learning_rate": 0.0001,
      "loss": 1.4577,
      "step": 4521
    },
    {
      "epoch": 0.7260757867694284,
      "grad_norm": 0.25212356448173523,
      "learning_rate": 0.0001,
      "loss": 1.4946,
      "step": 4522
    },
    {
      "epoch": 0.7262363519588954,
      "grad_norm": 0.25058719515800476,
      "learning_rate": 0.0001,
      "loss": 1.5004,
      "step": 4523
    },
    {
      "epoch": 0.7263969171483622,
      "grad_norm": 0.25646036863327026,
      "learning_rate": 0.0001,
      "loss": 1.4732,
      "step": 4524
    },
    {
      "epoch": 0.7265574823378291,
      "grad_norm": 0.25635862350463867,
      "learning_rate": 0.0001,
      "loss": 1.4925,
      "step": 4525
    },
    {
      "epoch": 0.7267180475272961,
      "grad_norm": 0.4270424544811249,
      "learning_rate": 0.0001,
      "loss": 1.4069,
      "step": 4526
    },
    {
      "epoch": 0.726878612716763,
      "grad_norm": 0.26464807987213135,
      "learning_rate": 0.0001,
      "loss": 1.4664,
      "step": 4527
    },
    {
      "epoch": 0.72703917790623,
      "grad_norm": 0.2629585564136505,
      "learning_rate": 0.0001,
      "loss": 1.4861,
      "step": 4528
    },
    {
      "epoch": 0.7271997430956969,
      "grad_norm": 0.2564154267311096,
      "learning_rate": 0.0001,
      "loss": 1.4986,
      "step": 4529
    },
    {
      "epoch": 0.7273603082851637,
      "grad_norm": 0.28767871856689453,
      "learning_rate": 0.0001,
      "loss": 1.471,
      "step": 4530
    },
    {
      "epoch": 0.7275208734746307,
      "grad_norm": 0.2589711546897888,
      "learning_rate": 0.0001,
      "loss": 1.497,
      "step": 4531
    },
    {
      "epoch": 0.7276814386640976,
      "grad_norm": 0.2619282603263855,
      "learning_rate": 0.0001,
      "loss": 1.5319,
      "step": 4532
    },
    {
      "epoch": 0.7278420038535646,
      "grad_norm": 0.28460222482681274,
      "learning_rate": 0.0001,
      "loss": 1.534,
      "step": 4533
    },
    {
      "epoch": 0.7280025690430315,
      "grad_norm": 0.26082509756088257,
      "learning_rate": 0.0001,
      "loss": 1.5575,
      "step": 4534
    },
    {
      "epoch": 0.7281631342324983,
      "grad_norm": 0.2523138225078583,
      "learning_rate": 0.0001,
      "loss": 1.5014,
      "step": 4535
    },
    {
      "epoch": 0.7283236994219653,
      "grad_norm": 0.2843577563762665,
      "learning_rate": 0.0001,
      "loss": 1.5753,
      "step": 4536
    },
    {
      "epoch": 0.7284842646114322,
      "grad_norm": 0.26808345317840576,
      "learning_rate": 0.0001,
      "loss": 1.5482,
      "step": 4537
    },
    {
      "epoch": 0.7286448298008992,
      "grad_norm": 0.29650044441223145,
      "learning_rate": 0.0001,
      "loss": 1.5501,
      "step": 4538
    },
    {
      "epoch": 0.7288053949903661,
      "grad_norm": 0.30034783482551575,
      "learning_rate": 0.0001,
      "loss": 1.4617,
      "step": 4539
    },
    {
      "epoch": 0.7289659601798331,
      "grad_norm": 0.2697218358516693,
      "learning_rate": 0.0001,
      "loss": 1.5538,
      "step": 4540
    },
    {
      "epoch": 0.7291265253692999,
      "grad_norm": 0.26011794805526733,
      "learning_rate": 0.0001,
      "loss": 1.5113,
      "step": 4541
    },
    {
      "epoch": 0.7292870905587668,
      "grad_norm": 0.28395283222198486,
      "learning_rate": 0.0001,
      "loss": 1.5384,
      "step": 4542
    },
    {
      "epoch": 0.7294476557482338,
      "grad_norm": 0.3622431755065918,
      "learning_rate": 0.0001,
      "loss": 1.5581,
      "step": 4543
    },
    {
      "epoch": 0.7296082209377007,
      "grad_norm": 0.26604852080345154,
      "learning_rate": 0.0001,
      "loss": 1.4882,
      "step": 4544
    },
    {
      "epoch": 0.7297687861271677,
      "grad_norm": 0.2625166177749634,
      "learning_rate": 0.0001,
      "loss": 1.4636,
      "step": 4545
    },
    {
      "epoch": 0.7299293513166345,
      "grad_norm": 0.2692174017429352,
      "learning_rate": 0.0001,
      "loss": 1.4913,
      "step": 4546
    },
    {
      "epoch": 0.7300899165061014,
      "grad_norm": 0.2643943727016449,
      "learning_rate": 0.0001,
      "loss": 1.4963,
      "step": 4547
    },
    {
      "epoch": 0.7302504816955684,
      "grad_norm": 0.27789101004600525,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 4548
    },
    {
      "epoch": 0.7304110468850353,
      "grad_norm": 0.26574602723121643,
      "learning_rate": 0.0001,
      "loss": 1.4863,
      "step": 4549
    },
    {
      "epoch": 0.7305716120745023,
      "grad_norm": 0.28030118346214294,
      "learning_rate": 0.0001,
      "loss": 1.4317,
      "step": 4550
    },
    {
      "epoch": 0.7307321772639692,
      "grad_norm": 0.2659422755241394,
      "learning_rate": 0.0001,
      "loss": 1.4541,
      "step": 4551
    },
    {
      "epoch": 0.730892742453436,
      "grad_norm": 0.2769496738910675,
      "learning_rate": 0.0001,
      "loss": 1.5108,
      "step": 4552
    },
    {
      "epoch": 0.731053307642903,
      "grad_norm": 0.3044896125793457,
      "learning_rate": 0.0001,
      "loss": 1.501,
      "step": 4553
    },
    {
      "epoch": 0.7312138728323699,
      "grad_norm": 0.2771703004837036,
      "learning_rate": 0.0001,
      "loss": 1.5176,
      "step": 4554
    },
    {
      "epoch": 0.7313744380218369,
      "grad_norm": 0.27319344878196716,
      "learning_rate": 0.0001,
      "loss": 1.5388,
      "step": 4555
    },
    {
      "epoch": 0.7315350032113038,
      "grad_norm": 0.2738259732723236,
      "learning_rate": 0.0001,
      "loss": 1.48,
      "step": 4556
    },
    {
      "epoch": 0.7316955684007708,
      "grad_norm": 0.2523276209831238,
      "learning_rate": 0.0001,
      "loss": 1.3931,
      "step": 4557
    },
    {
      "epoch": 0.7318561335902376,
      "grad_norm": 0.2789495289325714,
      "learning_rate": 0.0001,
      "loss": 1.5377,
      "step": 4558
    },
    {
      "epoch": 0.7320166987797045,
      "grad_norm": 0.27878350019454956,
      "learning_rate": 0.0001,
      "loss": 1.452,
      "step": 4559
    },
    {
      "epoch": 0.7321772639691715,
      "grad_norm": 0.25896358489990234,
      "learning_rate": 0.0001,
      "loss": 1.3775,
      "step": 4560
    },
    {
      "epoch": 0.7323378291586384,
      "grad_norm": 0.27563977241516113,
      "learning_rate": 0.0001,
      "loss": 1.5585,
      "step": 4561
    },
    {
      "epoch": 0.7324983943481054,
      "grad_norm": 0.27830901741981506,
      "learning_rate": 0.0001,
      "loss": 1.5122,
      "step": 4562
    },
    {
      "epoch": 0.7326589595375722,
      "grad_norm": 0.28074386715888977,
      "learning_rate": 0.0001,
      "loss": 1.5732,
      "step": 4563
    },
    {
      "epoch": 0.7328195247270392,
      "grad_norm": 0.31096115708351135,
      "learning_rate": 0.0001,
      "loss": 1.5196,
      "step": 4564
    },
    {
      "epoch": 0.7329800899165061,
      "grad_norm": 0.2652338743209839,
      "learning_rate": 0.0001,
      "loss": 1.5123,
      "step": 4565
    },
    {
      "epoch": 0.733140655105973,
      "grad_norm": 0.255057692527771,
      "learning_rate": 0.0001,
      "loss": 1.4401,
      "step": 4566
    },
    {
      "epoch": 0.73330122029544,
      "grad_norm": 0.27116659283638,
      "learning_rate": 0.0001,
      "loss": 1.5022,
      "step": 4567
    },
    {
      "epoch": 0.7334617854849069,
      "grad_norm": 0.26373663544654846,
      "learning_rate": 0.0001,
      "loss": 1.5582,
      "step": 4568
    },
    {
      "epoch": 0.7336223506743738,
      "grad_norm": 0.2734716534614563,
      "learning_rate": 0.0001,
      "loss": 1.4851,
      "step": 4569
    },
    {
      "epoch": 0.7337829158638407,
      "grad_norm": 0.26862290501594543,
      "learning_rate": 0.0001,
      "loss": 1.4386,
      "step": 4570
    },
    {
      "epoch": 0.7339434810533076,
      "grad_norm": 0.2610526382923126,
      "learning_rate": 0.0001,
      "loss": 1.5247,
      "step": 4571
    },
    {
      "epoch": 0.7341040462427746,
      "grad_norm": 0.26719582080841064,
      "learning_rate": 0.0001,
      "loss": 1.4941,
      "step": 4572
    },
    {
      "epoch": 0.7342646114322415,
      "grad_norm": 0.25229886174201965,
      "learning_rate": 0.0001,
      "loss": 1.4897,
      "step": 4573
    },
    {
      "epoch": 0.7344251766217084,
      "grad_norm": 0.2586591839790344,
      "learning_rate": 0.0001,
      "loss": 1.5035,
      "step": 4574
    },
    {
      "epoch": 0.7345857418111753,
      "grad_norm": 0.26860129833221436,
      "learning_rate": 0.0001,
      "loss": 1.4665,
      "step": 4575
    },
    {
      "epoch": 0.7347463070006423,
      "grad_norm": 0.2788375914096832,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 4576
    },
    {
      "epoch": 0.7349068721901092,
      "grad_norm": 0.2794533967971802,
      "learning_rate": 0.0001,
      "loss": 1.5398,
      "step": 4577
    },
    {
      "epoch": 0.7350674373795761,
      "grad_norm": 0.2631220817565918,
      "learning_rate": 0.0001,
      "loss": 1.5352,
      "step": 4578
    },
    {
      "epoch": 0.7352280025690431,
      "grad_norm": 0.27125927805900574,
      "learning_rate": 0.0001,
      "loss": 1.5272,
      "step": 4579
    },
    {
      "epoch": 0.7353885677585099,
      "grad_norm": 0.26662683486938477,
      "learning_rate": 0.0001,
      "loss": 1.4715,
      "step": 4580
    },
    {
      "epoch": 0.7355491329479769,
      "grad_norm": 0.2543308138847351,
      "learning_rate": 0.0001,
      "loss": 1.4585,
      "step": 4581
    },
    {
      "epoch": 0.7357096981374438,
      "grad_norm": 0.2568158805370331,
      "learning_rate": 0.0001,
      "loss": 1.5014,
      "step": 4582
    },
    {
      "epoch": 0.7358702633269107,
      "grad_norm": 0.26792454719543457,
      "learning_rate": 0.0001,
      "loss": 1.533,
      "step": 4583
    },
    {
      "epoch": 0.7360308285163777,
      "grad_norm": 0.26723480224609375,
      "learning_rate": 0.0001,
      "loss": 1.5555,
      "step": 4584
    },
    {
      "epoch": 0.7361913937058445,
      "grad_norm": 0.24441014230251312,
      "learning_rate": 0.0001,
      "loss": 1.4151,
      "step": 4585
    },
    {
      "epoch": 0.7363519588953115,
      "grad_norm": 0.26682451367378235,
      "learning_rate": 0.0001,
      "loss": 1.579,
      "step": 4586
    },
    {
      "epoch": 0.7365125240847784,
      "grad_norm": 0.26250365376472473,
      "learning_rate": 0.0001,
      "loss": 1.5039,
      "step": 4587
    },
    {
      "epoch": 0.7366730892742454,
      "grad_norm": 0.2574078142642975,
      "learning_rate": 0.0001,
      "loss": 1.4819,
      "step": 4588
    },
    {
      "epoch": 0.7368336544637123,
      "grad_norm": 0.2646680474281311,
      "learning_rate": 0.0001,
      "loss": 1.474,
      "step": 4589
    },
    {
      "epoch": 0.7369942196531792,
      "grad_norm": 0.27785542607307434,
      "learning_rate": 0.0001,
      "loss": 1.5371,
      "step": 4590
    },
    {
      "epoch": 0.7371547848426461,
      "grad_norm": 0.2574950158596039,
      "learning_rate": 0.0001,
      "loss": 1.4823,
      "step": 4591
    },
    {
      "epoch": 0.737315350032113,
      "grad_norm": 0.2878223955631256,
      "learning_rate": 0.0001,
      "loss": 1.5013,
      "step": 4592
    },
    {
      "epoch": 0.73747591522158,
      "grad_norm": 0.269226998090744,
      "learning_rate": 0.0001,
      "loss": 1.4697,
      "step": 4593
    },
    {
      "epoch": 0.7376364804110469,
      "grad_norm": 0.2553509473800659,
      "learning_rate": 0.0001,
      "loss": 1.4206,
      "step": 4594
    },
    {
      "epoch": 0.7377970456005138,
      "grad_norm": 0.2620447278022766,
      "learning_rate": 0.0001,
      "loss": 1.4771,
      "step": 4595
    },
    {
      "epoch": 0.7379576107899807,
      "grad_norm": 0.2647787034511566,
      "learning_rate": 0.0001,
      "loss": 1.5159,
      "step": 4596
    },
    {
      "epoch": 0.7381181759794476,
      "grad_norm": 0.2749190032482147,
      "learning_rate": 0.0001,
      "loss": 1.4306,
      "step": 4597
    },
    {
      "epoch": 0.7382787411689146,
      "grad_norm": 0.26950743794441223,
      "learning_rate": 0.0001,
      "loss": 1.5582,
      "step": 4598
    },
    {
      "epoch": 0.7384393063583815,
      "grad_norm": 0.2633620500564575,
      "learning_rate": 0.0001,
      "loss": 1.492,
      "step": 4599
    },
    {
      "epoch": 0.7385998715478485,
      "grad_norm": 0.2540116310119629,
      "learning_rate": 0.0001,
      "loss": 1.4603,
      "step": 4600
    },
    {
      "epoch": 0.7387604367373154,
      "grad_norm": 0.26467224955558777,
      "learning_rate": 0.0001,
      "loss": 1.4895,
      "step": 4601
    },
    {
      "epoch": 0.7389210019267822,
      "grad_norm": 0.27076637744903564,
      "learning_rate": 0.0001,
      "loss": 1.4849,
      "step": 4602
    },
    {
      "epoch": 0.7390815671162492,
      "grad_norm": 0.26632288098335266,
      "learning_rate": 0.0001,
      "loss": 1.4634,
      "step": 4603
    },
    {
      "epoch": 0.7392421323057161,
      "grad_norm": 0.26025721430778503,
      "learning_rate": 0.0001,
      "loss": 1.5347,
      "step": 4604
    },
    {
      "epoch": 0.7394026974951831,
      "grad_norm": 0.4053424894809723,
      "learning_rate": 0.0001,
      "loss": 1.5655,
      "step": 4605
    },
    {
      "epoch": 0.73956326268465,
      "grad_norm": 0.27238404750823975,
      "learning_rate": 0.0001,
      "loss": 1.5377,
      "step": 4606
    },
    {
      "epoch": 0.739723827874117,
      "grad_norm": 0.26540979743003845,
      "learning_rate": 0.0001,
      "loss": 1.571,
      "step": 4607
    },
    {
      "epoch": 0.7398843930635838,
      "grad_norm": 0.2828616201877594,
      "learning_rate": 0.0001,
      "loss": 1.4741,
      "step": 4608
    },
    {
      "epoch": 0.7400449582530507,
      "grad_norm": 0.2742636203765869,
      "learning_rate": 0.0001,
      "loss": 1.5357,
      "step": 4609
    },
    {
      "epoch": 0.7402055234425177,
      "grad_norm": 0.2712089717388153,
      "learning_rate": 0.0001,
      "loss": 1.4794,
      "step": 4610
    },
    {
      "epoch": 0.7403660886319846,
      "grad_norm": 0.2567654848098755,
      "learning_rate": 0.0001,
      "loss": 1.432,
      "step": 4611
    },
    {
      "epoch": 0.7405266538214516,
      "grad_norm": 0.2555319368839264,
      "learning_rate": 0.0001,
      "loss": 1.5154,
      "step": 4612
    },
    {
      "epoch": 0.7406872190109184,
      "grad_norm": 0.29317590594291687,
      "learning_rate": 0.0001,
      "loss": 1.5601,
      "step": 4613
    },
    {
      "epoch": 0.7408477842003853,
      "grad_norm": 0.2835467457771301,
      "learning_rate": 0.0001,
      "loss": 1.5262,
      "step": 4614
    },
    {
      "epoch": 0.7410083493898523,
      "grad_norm": 0.26752805709838867,
      "learning_rate": 0.0001,
      "loss": 1.4735,
      "step": 4615
    },
    {
      "epoch": 0.7411689145793192,
      "grad_norm": 0.2731216847896576,
      "learning_rate": 0.0001,
      "loss": 1.4841,
      "step": 4616
    },
    {
      "epoch": 0.7413294797687862,
      "grad_norm": 0.2601814866065979,
      "learning_rate": 0.0001,
      "loss": 1.4958,
      "step": 4617
    },
    {
      "epoch": 0.7414900449582531,
      "grad_norm": 0.272246778011322,
      "learning_rate": 0.0001,
      "loss": 1.4808,
      "step": 4618
    },
    {
      "epoch": 0.7416506101477199,
      "grad_norm": 0.2681845724582672,
      "learning_rate": 0.0001,
      "loss": 1.4677,
      "step": 4619
    },
    {
      "epoch": 0.7418111753371869,
      "grad_norm": 0.2591359317302704,
      "learning_rate": 0.0001,
      "loss": 1.5269,
      "step": 4620
    },
    {
      "epoch": 0.7419717405266538,
      "grad_norm": 0.2780703902244568,
      "learning_rate": 0.0001,
      "loss": 1.4674,
      "step": 4621
    },
    {
      "epoch": 0.7421323057161208,
      "grad_norm": 0.2710312008857727,
      "learning_rate": 0.0001,
      "loss": 1.4628,
      "step": 4622
    },
    {
      "epoch": 0.7422928709055877,
      "grad_norm": 0.27332204580307007,
      "learning_rate": 0.0001,
      "loss": 1.4587,
      "step": 4623
    },
    {
      "epoch": 0.7424534360950545,
      "grad_norm": 0.2555524408817291,
      "learning_rate": 0.0001,
      "loss": 1.4517,
      "step": 4624
    },
    {
      "epoch": 0.7426140012845215,
      "grad_norm": 0.26889708638191223,
      "learning_rate": 0.0001,
      "loss": 1.4995,
      "step": 4625
    },
    {
      "epoch": 0.7427745664739884,
      "grad_norm": 0.26449641585350037,
      "learning_rate": 0.0001,
      "loss": 1.4794,
      "step": 4626
    },
    {
      "epoch": 0.7429351316634554,
      "grad_norm": 0.2725636959075928,
      "learning_rate": 0.0001,
      "loss": 1.4984,
      "step": 4627
    },
    {
      "epoch": 0.7430956968529223,
      "grad_norm": 0.2693058252334595,
      "learning_rate": 0.0001,
      "loss": 1.5372,
      "step": 4628
    },
    {
      "epoch": 0.7432562620423893,
      "grad_norm": 0.26393017172813416,
      "learning_rate": 0.0001,
      "loss": 1.4811,
      "step": 4629
    },
    {
      "epoch": 0.7434168272318561,
      "grad_norm": 0.2784920334815979,
      "learning_rate": 0.0001,
      "loss": 1.4104,
      "step": 4630
    },
    {
      "epoch": 0.743577392421323,
      "grad_norm": 0.25803062319755554,
      "learning_rate": 0.0001,
      "loss": 1.5376,
      "step": 4631
    },
    {
      "epoch": 0.74373795761079,
      "grad_norm": 0.25867918133735657,
      "learning_rate": 0.0001,
      "loss": 1.4729,
      "step": 4632
    },
    {
      "epoch": 0.7438985228002569,
      "grad_norm": 0.27000972628593445,
      "learning_rate": 0.0001,
      "loss": 1.4882,
      "step": 4633
    },
    {
      "epoch": 0.7440590879897239,
      "grad_norm": 0.28312546014785767,
      "learning_rate": 0.0001,
      "loss": 1.5758,
      "step": 4634
    },
    {
      "epoch": 0.7442196531791907,
      "grad_norm": 0.27580639719963074,
      "learning_rate": 0.0001,
      "loss": 1.4736,
      "step": 4635
    },
    {
      "epoch": 0.7443802183686576,
      "grad_norm": 0.270631343126297,
      "learning_rate": 0.0001,
      "loss": 1.4746,
      "step": 4636
    },
    {
      "epoch": 0.7445407835581246,
      "grad_norm": 0.2756424844264984,
      "learning_rate": 0.0001,
      "loss": 1.5598,
      "step": 4637
    },
    {
      "epoch": 0.7447013487475915,
      "grad_norm": 0.2685732841491699,
      "learning_rate": 0.0001,
      "loss": 1.5022,
      "step": 4638
    },
    {
      "epoch": 0.7448619139370585,
      "grad_norm": 0.2652704119682312,
      "learning_rate": 0.0001,
      "loss": 1.4959,
      "step": 4639
    },
    {
      "epoch": 0.7450224791265254,
      "grad_norm": 0.26038676500320435,
      "learning_rate": 0.0001,
      "loss": 1.4565,
      "step": 4640
    },
    {
      "epoch": 0.7451830443159922,
      "grad_norm": 0.2651382386684418,
      "learning_rate": 0.0001,
      "loss": 1.4948,
      "step": 4641
    },
    {
      "epoch": 0.7453436095054592,
      "grad_norm": 0.24904988706111908,
      "learning_rate": 0.0001,
      "loss": 1.4551,
      "step": 4642
    },
    {
      "epoch": 0.7455041746949261,
      "grad_norm": 0.26152366399765015,
      "learning_rate": 0.0001,
      "loss": 1.5302,
      "step": 4643
    },
    {
      "epoch": 0.7456647398843931,
      "grad_norm": 0.2725069224834442,
      "learning_rate": 0.0001,
      "loss": 1.5288,
      "step": 4644
    },
    {
      "epoch": 0.74582530507386,
      "grad_norm": 0.26113831996917725,
      "learning_rate": 0.0001,
      "loss": 1.5031,
      "step": 4645
    },
    {
      "epoch": 0.7459858702633269,
      "grad_norm": 0.2562262713909149,
      "learning_rate": 0.0001,
      "loss": 1.5174,
      "step": 4646
    },
    {
      "epoch": 0.7461464354527938,
      "grad_norm": 0.27277621626853943,
      "learning_rate": 0.0001,
      "loss": 1.4447,
      "step": 4647
    },
    {
      "epoch": 0.7463070006422607,
      "grad_norm": 0.27036160230636597,
      "learning_rate": 0.0001,
      "loss": 1.4749,
      "step": 4648
    },
    {
      "epoch": 0.7464675658317277,
      "grad_norm": 0.27666088938713074,
      "learning_rate": 0.0001,
      "loss": 1.5419,
      "step": 4649
    },
    {
      "epoch": 0.7466281310211946,
      "grad_norm": 0.2988976538181305,
      "learning_rate": 0.0001,
      "loss": 1.4219,
      "step": 4650
    },
    {
      "epoch": 0.7467886962106616,
      "grad_norm": 0.30231142044067383,
      "learning_rate": 0.0001,
      "loss": 1.4745,
      "step": 4651
    },
    {
      "epoch": 0.7469492614001284,
      "grad_norm": 0.279981404542923,
      "learning_rate": 0.0001,
      "loss": 1.521,
      "step": 4652
    },
    {
      "epoch": 0.7471098265895953,
      "grad_norm": 0.26500624418258667,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 4653
    },
    {
      "epoch": 0.7472703917790623,
      "grad_norm": 0.26284492015838623,
      "learning_rate": 0.0001,
      "loss": 1.4263,
      "step": 4654
    },
    {
      "epoch": 0.7474309569685292,
      "grad_norm": 0.27375173568725586,
      "learning_rate": 0.0001,
      "loss": 1.4665,
      "step": 4655
    },
    {
      "epoch": 0.7475915221579962,
      "grad_norm": 0.2888254225254059,
      "learning_rate": 0.0001,
      "loss": 1.5209,
      "step": 4656
    },
    {
      "epoch": 0.7477520873474631,
      "grad_norm": 0.31308513879776,
      "learning_rate": 0.0001,
      "loss": 1.5274,
      "step": 4657
    },
    {
      "epoch": 0.74791265253693,
      "grad_norm": 0.267828106880188,
      "learning_rate": 0.0001,
      "loss": 1.5114,
      "step": 4658
    },
    {
      "epoch": 0.7480732177263969,
      "grad_norm": 0.2816733121871948,
      "learning_rate": 0.0001,
      "loss": 1.5678,
      "step": 4659
    },
    {
      "epoch": 0.7482337829158638,
      "grad_norm": 0.26341959834098816,
      "learning_rate": 0.0001,
      "loss": 1.4596,
      "step": 4660
    },
    {
      "epoch": 0.7483943481053308,
      "grad_norm": 0.2673376798629761,
      "learning_rate": 0.0001,
      "loss": 1.4437,
      "step": 4661
    },
    {
      "epoch": 0.7485549132947977,
      "grad_norm": 0.2731376588344574,
      "learning_rate": 0.0001,
      "loss": 1.4575,
      "step": 4662
    },
    {
      "epoch": 0.7487154784842646,
      "grad_norm": 0.2772185206413269,
      "learning_rate": 0.0001,
      "loss": 1.568,
      "step": 4663
    },
    {
      "epoch": 0.7488760436737315,
      "grad_norm": 0.27636203169822693,
      "learning_rate": 0.0001,
      "loss": 1.4845,
      "step": 4664
    },
    {
      "epoch": 0.7490366088631984,
      "grad_norm": 0.28590911626815796,
      "learning_rate": 0.0001,
      "loss": 1.5398,
      "step": 4665
    },
    {
      "epoch": 0.7491971740526654,
      "grad_norm": 0.26970577239990234,
      "learning_rate": 0.0001,
      "loss": 1.4664,
      "step": 4666
    },
    {
      "epoch": 0.7493577392421323,
      "grad_norm": 0.2564913332462311,
      "learning_rate": 0.0001,
      "loss": 1.41,
      "step": 4667
    },
    {
      "epoch": 0.7495183044315993,
      "grad_norm": 0.2824428081512451,
      "learning_rate": 0.0001,
      "loss": 1.5264,
      "step": 4668
    },
    {
      "epoch": 0.7496788696210661,
      "grad_norm": 0.2598324120044708,
      "learning_rate": 0.0001,
      "loss": 1.4318,
      "step": 4669
    },
    {
      "epoch": 0.7498394348105331,
      "grad_norm": 0.2638610899448395,
      "learning_rate": 0.0001,
      "loss": 1.5312,
      "step": 4670
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.26793932914733887,
      "learning_rate": 0.0001,
      "loss": 1.5179,
      "step": 4671
    },
    {
      "epoch": 0.7501605651894669,
      "grad_norm": 0.2577224671840668,
      "learning_rate": 0.0001,
      "loss": 1.4373,
      "step": 4672
    },
    {
      "epoch": 0.7503211303789339,
      "grad_norm": 0.27305158972740173,
      "learning_rate": 0.0001,
      "loss": 1.4649,
      "step": 4673
    },
    {
      "epoch": 0.7504816955684007,
      "grad_norm": 0.24940225481987,
      "learning_rate": 0.0001,
      "loss": 1.4178,
      "step": 4674
    },
    {
      "epoch": 0.7506422607578677,
      "grad_norm": 0.2555713355541229,
      "learning_rate": 0.0001,
      "loss": 1.5347,
      "step": 4675
    },
    {
      "epoch": 0.7508028259473346,
      "grad_norm": 0.2697537839412689,
      "learning_rate": 0.0001,
      "loss": 1.499,
      "step": 4676
    },
    {
      "epoch": 0.7509633911368016,
      "grad_norm": 0.2856290638446808,
      "learning_rate": 0.0001,
      "loss": 1.4792,
      "step": 4677
    },
    {
      "epoch": 0.7511239563262685,
      "grad_norm": 0.2914230227470398,
      "learning_rate": 0.0001,
      "loss": 1.5277,
      "step": 4678
    },
    {
      "epoch": 0.7512845215157354,
      "grad_norm": 0.2725703716278076,
      "learning_rate": 0.0001,
      "loss": 1.5056,
      "step": 4679
    },
    {
      "epoch": 0.7514450867052023,
      "grad_norm": 0.26528775691986084,
      "learning_rate": 0.0001,
      "loss": 1.5179,
      "step": 4680
    },
    {
      "epoch": 0.7516056518946692,
      "grad_norm": 0.27045440673828125,
      "learning_rate": 0.0001,
      "loss": 1.5296,
      "step": 4681
    },
    {
      "epoch": 0.7517662170841362,
      "grad_norm": 0.2855928838253021,
      "learning_rate": 0.0001,
      "loss": 1.481,
      "step": 4682
    },
    {
      "epoch": 0.7519267822736031,
      "grad_norm": 0.2561173141002655,
      "learning_rate": 0.0001,
      "loss": 1.3798,
      "step": 4683
    },
    {
      "epoch": 0.75208734746307,
      "grad_norm": 0.2532760202884674,
      "learning_rate": 0.0001,
      "loss": 1.4825,
      "step": 4684
    },
    {
      "epoch": 0.7522479126525369,
      "grad_norm": 0.27161532640457153,
      "learning_rate": 0.0001,
      "loss": 1.4858,
      "step": 4685
    },
    {
      "epoch": 0.7524084778420038,
      "grad_norm": 0.2532717287540436,
      "learning_rate": 0.0001,
      "loss": 1.4916,
      "step": 4686
    },
    {
      "epoch": 0.7525690430314708,
      "grad_norm": 0.257660835981369,
      "learning_rate": 0.0001,
      "loss": 1.4006,
      "step": 4687
    },
    {
      "epoch": 0.7527296082209377,
      "grad_norm": 0.26825979351997375,
      "learning_rate": 0.0001,
      "loss": 1.5558,
      "step": 4688
    },
    {
      "epoch": 0.7528901734104047,
      "grad_norm": 0.2663351595401764,
      "learning_rate": 0.0001,
      "loss": 1.5599,
      "step": 4689
    },
    {
      "epoch": 0.7530507385998716,
      "grad_norm": 0.269704133272171,
      "learning_rate": 0.0001,
      "loss": 1.4263,
      "step": 4690
    },
    {
      "epoch": 0.7532113037893384,
      "grad_norm": 0.27241092920303345,
      "learning_rate": 0.0001,
      "loss": 1.5113,
      "step": 4691
    },
    {
      "epoch": 0.7533718689788054,
      "grad_norm": 0.2649691700935364,
      "learning_rate": 0.0001,
      "loss": 1.5441,
      "step": 4692
    },
    {
      "epoch": 0.7535324341682723,
      "grad_norm": 0.27136510610580444,
      "learning_rate": 0.0001,
      "loss": 1.515,
      "step": 4693
    },
    {
      "epoch": 0.7536929993577393,
      "grad_norm": 0.2539619207382202,
      "learning_rate": 0.0001,
      "loss": 1.4025,
      "step": 4694
    },
    {
      "epoch": 0.7538535645472062,
      "grad_norm": 0.27253806591033936,
      "learning_rate": 0.0001,
      "loss": 1.5135,
      "step": 4695
    },
    {
      "epoch": 0.7540141297366731,
      "grad_norm": 0.2730512320995331,
      "learning_rate": 0.0001,
      "loss": 1.5218,
      "step": 4696
    },
    {
      "epoch": 0.75417469492614,
      "grad_norm": 0.26921844482421875,
      "learning_rate": 0.0001,
      "loss": 1.3778,
      "step": 4697
    },
    {
      "epoch": 0.7543352601156069,
      "grad_norm": 0.264752596616745,
      "learning_rate": 0.0001,
      "loss": 1.4958,
      "step": 4698
    },
    {
      "epoch": 0.7544958253050739,
      "grad_norm": 0.27301615476608276,
      "learning_rate": 0.0001,
      "loss": 1.5358,
      "step": 4699
    },
    {
      "epoch": 0.7546563904945408,
      "grad_norm": 0.2675076425075531,
      "learning_rate": 0.0001,
      "loss": 1.4918,
      "step": 4700
    },
    {
      "epoch": 0.7548169556840078,
      "grad_norm": 0.2580350637435913,
      "learning_rate": 0.0001,
      "loss": 1.4876,
      "step": 4701
    },
    {
      "epoch": 0.7549775208734746,
      "grad_norm": 0.2589607536792755,
      "learning_rate": 0.0001,
      "loss": 1.4948,
      "step": 4702
    },
    {
      "epoch": 0.7551380860629415,
      "grad_norm": 0.26058530807495117,
      "learning_rate": 0.0001,
      "loss": 1.3502,
      "step": 4703
    },
    {
      "epoch": 0.7552986512524085,
      "grad_norm": 0.2683979570865631,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 4704
    },
    {
      "epoch": 0.7554592164418754,
      "grad_norm": 0.26223766803741455,
      "learning_rate": 0.0001,
      "loss": 1.5359,
      "step": 4705
    },
    {
      "epoch": 0.7556197816313424,
      "grad_norm": 0.3043115735054016,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 4706
    },
    {
      "epoch": 0.7557803468208093,
      "grad_norm": 0.2684251666069031,
      "learning_rate": 0.0001,
      "loss": 1.5332,
      "step": 4707
    },
    {
      "epoch": 0.7559409120102761,
      "grad_norm": 0.25287577509880066,
      "learning_rate": 0.0001,
      "loss": 1.3995,
      "step": 4708
    },
    {
      "epoch": 0.7561014771997431,
      "grad_norm": 0.25544285774230957,
      "learning_rate": 0.0001,
      "loss": 1.5004,
      "step": 4709
    },
    {
      "epoch": 0.75626204238921,
      "grad_norm": 0.25161290168762207,
      "learning_rate": 0.0001,
      "loss": 1.4749,
      "step": 4710
    },
    {
      "epoch": 0.756422607578677,
      "grad_norm": 0.280152291059494,
      "learning_rate": 0.0001,
      "loss": 1.4558,
      "step": 4711
    },
    {
      "epoch": 0.7565831727681439,
      "grad_norm": 0.2701803743839264,
      "learning_rate": 0.0001,
      "loss": 1.4781,
      "step": 4712
    },
    {
      "epoch": 0.7567437379576107,
      "grad_norm": 0.254133015871048,
      "learning_rate": 0.0001,
      "loss": 1.4657,
      "step": 4713
    },
    {
      "epoch": 0.7569043031470777,
      "grad_norm": 0.2725158929824829,
      "learning_rate": 0.0001,
      "loss": 1.5022,
      "step": 4714
    },
    {
      "epoch": 0.7570648683365446,
      "grad_norm": 0.2598293721675873,
      "learning_rate": 0.0001,
      "loss": 1.446,
      "step": 4715
    },
    {
      "epoch": 0.7572254335260116,
      "grad_norm": 0.26696163415908813,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 4716
    },
    {
      "epoch": 0.7573859987154785,
      "grad_norm": 0.26862427592277527,
      "learning_rate": 0.0001,
      "loss": 1.4589,
      "step": 4717
    },
    {
      "epoch": 0.7575465639049455,
      "grad_norm": 0.2665650248527527,
      "learning_rate": 0.0001,
      "loss": 1.5172,
      "step": 4718
    },
    {
      "epoch": 0.7577071290944123,
      "grad_norm": 0.2625967860221863,
      "learning_rate": 0.0001,
      "loss": 1.4693,
      "step": 4719
    },
    {
      "epoch": 0.7578676942838792,
      "grad_norm": 0.25631463527679443,
      "learning_rate": 0.0001,
      "loss": 1.5017,
      "step": 4720
    },
    {
      "epoch": 0.7580282594733462,
      "grad_norm": 0.2516089081764221,
      "learning_rate": 0.0001,
      "loss": 1.4522,
      "step": 4721
    },
    {
      "epoch": 0.7581888246628131,
      "grad_norm": 0.25942090153694153,
      "learning_rate": 0.0001,
      "loss": 1.5062,
      "step": 4722
    },
    {
      "epoch": 0.7583493898522801,
      "grad_norm": 0.2715856432914734,
      "learning_rate": 0.0001,
      "loss": 1.455,
      "step": 4723
    },
    {
      "epoch": 0.7585099550417469,
      "grad_norm": 0.24603420495986938,
      "learning_rate": 0.0001,
      "loss": 1.4145,
      "step": 4724
    },
    {
      "epoch": 0.7586705202312138,
      "grad_norm": 0.2652493715286255,
      "learning_rate": 0.0001,
      "loss": 1.5064,
      "step": 4725
    },
    {
      "epoch": 0.7588310854206808,
      "grad_norm": 0.2675178647041321,
      "learning_rate": 0.0001,
      "loss": 1.4618,
      "step": 4726
    },
    {
      "epoch": 0.7589916506101477,
      "grad_norm": 0.25946810841560364,
      "learning_rate": 0.0001,
      "loss": 1.4265,
      "step": 4727
    },
    {
      "epoch": 0.7591522157996147,
      "grad_norm": 0.3071325421333313,
      "learning_rate": 0.0001,
      "loss": 1.6179,
      "step": 4728
    },
    {
      "epoch": 0.7593127809890816,
      "grad_norm": 0.2836286425590515,
      "learning_rate": 0.0001,
      "loss": 1.5251,
      "step": 4729
    },
    {
      "epoch": 0.7594733461785484,
      "grad_norm": 0.2590748369693756,
      "learning_rate": 0.0001,
      "loss": 1.4962,
      "step": 4730
    },
    {
      "epoch": 0.7596339113680154,
      "grad_norm": 0.2758234143257141,
      "learning_rate": 0.0001,
      "loss": 1.505,
      "step": 4731
    },
    {
      "epoch": 0.7597944765574823,
      "grad_norm": 0.2708079218864441,
      "learning_rate": 0.0001,
      "loss": 1.5904,
      "step": 4732
    },
    {
      "epoch": 0.7599550417469493,
      "grad_norm": 0.27538663148880005,
      "learning_rate": 0.0001,
      "loss": 1.5742,
      "step": 4733
    },
    {
      "epoch": 0.7601156069364162,
      "grad_norm": 0.2765510380268097,
      "learning_rate": 0.0001,
      "loss": 1.4783,
      "step": 4734
    },
    {
      "epoch": 0.760276172125883,
      "grad_norm": 0.25619640946388245,
      "learning_rate": 0.0001,
      "loss": 1.4826,
      "step": 4735
    },
    {
      "epoch": 0.76043673731535,
      "grad_norm": 0.2517492473125458,
      "learning_rate": 0.0001,
      "loss": 1.4504,
      "step": 4736
    },
    {
      "epoch": 0.7605973025048169,
      "grad_norm": 0.25144222378730774,
      "learning_rate": 0.0001,
      "loss": 1.4623,
      "step": 4737
    },
    {
      "epoch": 0.7607578676942839,
      "grad_norm": 0.26716700196266174,
      "learning_rate": 0.0001,
      "loss": 1.5116,
      "step": 4738
    },
    {
      "epoch": 0.7609184328837508,
      "grad_norm": 0.2455921471118927,
      "learning_rate": 0.0001,
      "loss": 1.4347,
      "step": 4739
    },
    {
      "epoch": 0.7610789980732178,
      "grad_norm": 0.2558843791484833,
      "learning_rate": 0.0001,
      "loss": 1.4003,
      "step": 4740
    },
    {
      "epoch": 0.7612395632626846,
      "grad_norm": 0.2663254141807556,
      "learning_rate": 0.0001,
      "loss": 1.4465,
      "step": 4741
    },
    {
      "epoch": 0.7614001284521515,
      "grad_norm": 0.2517986595630646,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 4742
    },
    {
      "epoch": 0.7615606936416185,
      "grad_norm": 0.2688898742198944,
      "learning_rate": 0.0001,
      "loss": 1.5246,
      "step": 4743
    },
    {
      "epoch": 0.7617212588310854,
      "grad_norm": 0.2682409882545471,
      "learning_rate": 0.0001,
      "loss": 1.5623,
      "step": 4744
    },
    {
      "epoch": 0.7618818240205524,
      "grad_norm": 0.2704203128814697,
      "learning_rate": 0.0001,
      "loss": 1.4919,
      "step": 4745
    },
    {
      "epoch": 0.7620423892100193,
      "grad_norm": 0.27167192101478577,
      "learning_rate": 0.0001,
      "loss": 1.5474,
      "step": 4746
    },
    {
      "epoch": 0.7622029543994862,
      "grad_norm": 0.27285847067832947,
      "learning_rate": 0.0001,
      "loss": 1.5328,
      "step": 4747
    },
    {
      "epoch": 0.7623635195889531,
      "grad_norm": 0.25148075819015503,
      "learning_rate": 0.0001,
      "loss": 1.4122,
      "step": 4748
    },
    {
      "epoch": 0.76252408477842,
      "grad_norm": 0.31446388363838196,
      "learning_rate": 0.0001,
      "loss": 1.4981,
      "step": 4749
    },
    {
      "epoch": 0.762684649967887,
      "grad_norm": 0.2945290803909302,
      "learning_rate": 0.0001,
      "loss": 1.5408,
      "step": 4750
    },
    {
      "epoch": 0.7628452151573539,
      "grad_norm": 0.25229719281196594,
      "learning_rate": 0.0001,
      "loss": 1.4355,
      "step": 4751
    },
    {
      "epoch": 0.7630057803468208,
      "grad_norm": 0.2849136292934418,
      "learning_rate": 0.0001,
      "loss": 1.4678,
      "step": 4752
    },
    {
      "epoch": 0.7631663455362877,
      "grad_norm": 0.27459242939949036,
      "learning_rate": 0.0001,
      "loss": 1.4002,
      "step": 4753
    },
    {
      "epoch": 0.7633269107257546,
      "grad_norm": 0.2552124857902527,
      "learning_rate": 0.0001,
      "loss": 1.545,
      "step": 4754
    },
    {
      "epoch": 0.7634874759152216,
      "grad_norm": 0.2512931525707245,
      "learning_rate": 0.0001,
      "loss": 1.4921,
      "step": 4755
    },
    {
      "epoch": 0.7636480411046885,
      "grad_norm": 0.2796391248703003,
      "learning_rate": 0.0001,
      "loss": 1.5396,
      "step": 4756
    },
    {
      "epoch": 0.7638086062941555,
      "grad_norm": 0.2868058383464813,
      "learning_rate": 0.0001,
      "loss": 1.5936,
      "step": 4757
    },
    {
      "epoch": 0.7639691714836223,
      "grad_norm": 0.25818485021591187,
      "learning_rate": 0.0001,
      "loss": 1.4779,
      "step": 4758
    },
    {
      "epoch": 0.7641297366730893,
      "grad_norm": 0.2648962736129761,
      "learning_rate": 0.0001,
      "loss": 1.5517,
      "step": 4759
    },
    {
      "epoch": 0.7642903018625562,
      "grad_norm": 0.305493026971817,
      "learning_rate": 0.0001,
      "loss": 1.5399,
      "step": 4760
    },
    {
      "epoch": 0.7644508670520231,
      "grad_norm": 0.2696012854576111,
      "learning_rate": 0.0001,
      "loss": 1.4914,
      "step": 4761
    },
    {
      "epoch": 0.7646114322414901,
      "grad_norm": 0.26010653376579285,
      "learning_rate": 0.0001,
      "loss": 1.3876,
      "step": 4762
    },
    {
      "epoch": 0.7647719974309569,
      "grad_norm": 0.27666372060775757,
      "learning_rate": 0.0001,
      "loss": 1.494,
      "step": 4763
    },
    {
      "epoch": 0.7649325626204239,
      "grad_norm": 0.27314937114715576,
      "learning_rate": 0.0001,
      "loss": 1.4817,
      "step": 4764
    },
    {
      "epoch": 0.7650931278098908,
      "grad_norm": 0.2704566419124603,
      "learning_rate": 0.0001,
      "loss": 1.4727,
      "step": 4765
    },
    {
      "epoch": 0.7652536929993577,
      "grad_norm": 0.2436332404613495,
      "learning_rate": 0.0001,
      "loss": 1.4332,
      "step": 4766
    },
    {
      "epoch": 0.7654142581888247,
      "grad_norm": 0.28788483142852783,
      "learning_rate": 0.0001,
      "loss": 1.6135,
      "step": 4767
    },
    {
      "epoch": 0.7655748233782916,
      "grad_norm": 0.26671111583709717,
      "learning_rate": 0.0001,
      "loss": 1.4954,
      "step": 4768
    },
    {
      "epoch": 0.7657353885677585,
      "grad_norm": 0.2638619840145111,
      "learning_rate": 0.0001,
      "loss": 1.4693,
      "step": 4769
    },
    {
      "epoch": 0.7658959537572254,
      "grad_norm": 0.26721131801605225,
      "learning_rate": 0.0001,
      "loss": 1.4429,
      "step": 4770
    },
    {
      "epoch": 0.7660565189466924,
      "grad_norm": 0.2585608661174774,
      "learning_rate": 0.0001,
      "loss": 1.4793,
      "step": 4771
    },
    {
      "epoch": 0.7662170841361593,
      "grad_norm": 0.2614308297634125,
      "learning_rate": 0.0001,
      "loss": 1.5581,
      "step": 4772
    },
    {
      "epoch": 0.7663776493256262,
      "grad_norm": 0.3514977693557739,
      "learning_rate": 0.0001,
      "loss": 1.4327,
      "step": 4773
    },
    {
      "epoch": 0.7665382145150931,
      "grad_norm": 0.26077356934547424,
      "learning_rate": 0.0001,
      "loss": 1.4794,
      "step": 4774
    },
    {
      "epoch": 0.76669877970456,
      "grad_norm": 0.2752145826816559,
      "learning_rate": 0.0001,
      "loss": 1.4869,
      "step": 4775
    },
    {
      "epoch": 0.766859344894027,
      "grad_norm": 0.2619808316230774,
      "learning_rate": 0.0001,
      "loss": 1.49,
      "step": 4776
    },
    {
      "epoch": 0.7670199100834939,
      "grad_norm": 0.2664008140563965,
      "learning_rate": 0.0001,
      "loss": 1.524,
      "step": 4777
    },
    {
      "epoch": 0.7671804752729608,
      "grad_norm": 0.2534385323524475,
      "learning_rate": 0.0001,
      "loss": 1.4437,
      "step": 4778
    },
    {
      "epoch": 0.7673410404624278,
      "grad_norm": 0.26758888363838196,
      "learning_rate": 0.0001,
      "loss": 1.4656,
      "step": 4779
    },
    {
      "epoch": 0.7675016056518946,
      "grad_norm": 0.26431041955947876,
      "learning_rate": 0.0001,
      "loss": 1.4692,
      "step": 4780
    },
    {
      "epoch": 0.7676621708413616,
      "grad_norm": 0.27445584535598755,
      "learning_rate": 0.0001,
      "loss": 1.5469,
      "step": 4781
    },
    {
      "epoch": 0.7678227360308285,
      "grad_norm": 0.24833647906780243,
      "learning_rate": 0.0001,
      "loss": 1.4113,
      "step": 4782
    },
    {
      "epoch": 0.7679833012202955,
      "grad_norm": 0.2731981873512268,
      "learning_rate": 0.0001,
      "loss": 1.5062,
      "step": 4783
    },
    {
      "epoch": 0.7681438664097624,
      "grad_norm": 0.2720390856266022,
      "learning_rate": 0.0001,
      "loss": 1.5158,
      "step": 4784
    },
    {
      "epoch": 0.7683044315992292,
      "grad_norm": 0.2624323070049286,
      "learning_rate": 0.0001,
      "loss": 1.5222,
      "step": 4785
    },
    {
      "epoch": 0.7684649967886962,
      "grad_norm": 0.26187559962272644,
      "learning_rate": 0.0001,
      "loss": 1.4671,
      "step": 4786
    },
    {
      "epoch": 0.7686255619781631,
      "grad_norm": 0.2950212061405182,
      "learning_rate": 0.0001,
      "loss": 1.5074,
      "step": 4787
    },
    {
      "epoch": 0.7687861271676301,
      "grad_norm": 0.26819872856140137,
      "learning_rate": 0.0001,
      "loss": 1.5208,
      "step": 4788
    },
    {
      "epoch": 0.768946692357097,
      "grad_norm": 0.2585736811161041,
      "learning_rate": 0.0001,
      "loss": 1.3985,
      "step": 4789
    },
    {
      "epoch": 0.769107257546564,
      "grad_norm": 0.2602596580982208,
      "learning_rate": 0.0001,
      "loss": 1.5587,
      "step": 4790
    },
    {
      "epoch": 0.7692678227360308,
      "grad_norm": 0.2900967597961426,
      "learning_rate": 0.0001,
      "loss": 1.4779,
      "step": 4791
    },
    {
      "epoch": 0.7694283879254977,
      "grad_norm": 0.2687471807003021,
      "learning_rate": 0.0001,
      "loss": 1.4822,
      "step": 4792
    },
    {
      "epoch": 0.7695889531149647,
      "grad_norm": 0.286593496799469,
      "learning_rate": 0.0001,
      "loss": 1.4339,
      "step": 4793
    },
    {
      "epoch": 0.7697495183044316,
      "grad_norm": 0.2533605694770813,
      "learning_rate": 0.0001,
      "loss": 1.4495,
      "step": 4794
    },
    {
      "epoch": 0.7699100834938986,
      "grad_norm": 0.25842908024787903,
      "learning_rate": 0.0001,
      "loss": 1.5387,
      "step": 4795
    },
    {
      "epoch": 0.7700706486833655,
      "grad_norm": 0.2921748161315918,
      "learning_rate": 0.0001,
      "loss": 1.5241,
      "step": 4796
    },
    {
      "epoch": 0.7702312138728323,
      "grad_norm": 0.2845020592212677,
      "learning_rate": 0.0001,
      "loss": 1.5323,
      "step": 4797
    },
    {
      "epoch": 0.7703917790622993,
      "grad_norm": 0.2742638885974884,
      "learning_rate": 0.0001,
      "loss": 1.4803,
      "step": 4798
    },
    {
      "epoch": 0.7705523442517662,
      "grad_norm": 0.2830826938152313,
      "learning_rate": 0.0001,
      "loss": 1.4818,
      "step": 4799
    },
    {
      "epoch": 0.7707129094412332,
      "grad_norm": 0.27695101499557495,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 4800
    },
    {
      "epoch": 0.7708734746307001,
      "grad_norm": 0.2770475447177887,
      "learning_rate": 0.0001,
      "loss": 1.4154,
      "step": 4801
    },
    {
      "epoch": 0.7710340398201669,
      "grad_norm": 0.26153096556663513,
      "learning_rate": 0.0001,
      "loss": 1.4582,
      "step": 4802
    },
    {
      "epoch": 0.7711946050096339,
      "grad_norm": 0.26818645000457764,
      "learning_rate": 0.0001,
      "loss": 1.4709,
      "step": 4803
    },
    {
      "epoch": 0.7713551701991008,
      "grad_norm": 0.27246400713920593,
      "learning_rate": 0.0001,
      "loss": 1.441,
      "step": 4804
    },
    {
      "epoch": 0.7715157353885678,
      "grad_norm": 0.26476332545280457,
      "learning_rate": 0.0001,
      "loss": 1.5333,
      "step": 4805
    },
    {
      "epoch": 0.7716763005780347,
      "grad_norm": 0.26698577404022217,
      "learning_rate": 0.0001,
      "loss": 1.5364,
      "step": 4806
    },
    {
      "epoch": 0.7718368657675017,
      "grad_norm": 0.2695167660713196,
      "learning_rate": 0.0001,
      "loss": 1.4789,
      "step": 4807
    },
    {
      "epoch": 0.7719974309569685,
      "grad_norm": 0.2772916853427887,
      "learning_rate": 0.0001,
      "loss": 1.502,
      "step": 4808
    },
    {
      "epoch": 0.7721579961464354,
      "grad_norm": 0.2715173363685608,
      "learning_rate": 0.0001,
      "loss": 1.4884,
      "step": 4809
    },
    {
      "epoch": 0.7723185613359024,
      "grad_norm": 0.27370476722717285,
      "learning_rate": 0.0001,
      "loss": 1.4717,
      "step": 4810
    },
    {
      "epoch": 0.7724791265253693,
      "grad_norm": 0.26012903451919556,
      "learning_rate": 0.0001,
      "loss": 1.5177,
      "step": 4811
    },
    {
      "epoch": 0.7726396917148363,
      "grad_norm": 0.2607390582561493,
      "learning_rate": 0.0001,
      "loss": 1.4756,
      "step": 4812
    },
    {
      "epoch": 0.7728002569043031,
      "grad_norm": 0.28200775384902954,
      "learning_rate": 0.0001,
      "loss": 1.524,
      "step": 4813
    },
    {
      "epoch": 0.77296082209377,
      "grad_norm": 0.2612951695919037,
      "learning_rate": 0.0001,
      "loss": 1.4588,
      "step": 4814
    },
    {
      "epoch": 0.773121387283237,
      "grad_norm": 0.26932767033576965,
      "learning_rate": 0.0001,
      "loss": 1.4984,
      "step": 4815
    },
    {
      "epoch": 0.7732819524727039,
      "grad_norm": 0.26361799240112305,
      "learning_rate": 0.0001,
      "loss": 1.4578,
      "step": 4816
    },
    {
      "epoch": 0.7734425176621709,
      "grad_norm": 0.2659071385860443,
      "learning_rate": 0.0001,
      "loss": 1.5742,
      "step": 4817
    },
    {
      "epoch": 0.7736030828516378,
      "grad_norm": 0.26427409052848816,
      "learning_rate": 0.0001,
      "loss": 1.5407,
      "step": 4818
    },
    {
      "epoch": 0.7737636480411046,
      "grad_norm": 0.2671090066432953,
      "learning_rate": 0.0001,
      "loss": 1.5054,
      "step": 4819
    },
    {
      "epoch": 0.7739242132305716,
      "grad_norm": 0.26824119687080383,
      "learning_rate": 0.0001,
      "loss": 1.527,
      "step": 4820
    },
    {
      "epoch": 0.7740847784200385,
      "grad_norm": 0.25563180446624756,
      "learning_rate": 0.0001,
      "loss": 1.4901,
      "step": 4821
    },
    {
      "epoch": 0.7742453436095055,
      "grad_norm": 0.25880300998687744,
      "learning_rate": 0.0001,
      "loss": 1.4817,
      "step": 4822
    },
    {
      "epoch": 0.7744059087989724,
      "grad_norm": 0.2709629237651825,
      "learning_rate": 0.0001,
      "loss": 1.6155,
      "step": 4823
    },
    {
      "epoch": 0.7745664739884393,
      "grad_norm": 0.26997342705726624,
      "learning_rate": 0.0001,
      "loss": 1.5066,
      "step": 4824
    },
    {
      "epoch": 0.7747270391779062,
      "grad_norm": 0.26577097177505493,
      "learning_rate": 0.0001,
      "loss": 1.5907,
      "step": 4825
    },
    {
      "epoch": 0.7748876043673731,
      "grad_norm": 0.26206445693969727,
      "learning_rate": 0.0001,
      "loss": 1.5525,
      "step": 4826
    },
    {
      "epoch": 0.7750481695568401,
      "grad_norm": 0.26891064643859863,
      "learning_rate": 0.0001,
      "loss": 1.585,
      "step": 4827
    },
    {
      "epoch": 0.775208734746307,
      "grad_norm": 0.2602813243865967,
      "learning_rate": 0.0001,
      "loss": 1.5347,
      "step": 4828
    },
    {
      "epoch": 0.775369299935774,
      "grad_norm": 0.2580989897251129,
      "learning_rate": 0.0001,
      "loss": 1.4591,
      "step": 4829
    },
    {
      "epoch": 0.7755298651252408,
      "grad_norm": 0.2754908800125122,
      "learning_rate": 0.0001,
      "loss": 1.5025,
      "step": 4830
    },
    {
      "epoch": 0.7756904303147077,
      "grad_norm": 0.4796976149082184,
      "learning_rate": 0.0001,
      "loss": 1.5579,
      "step": 4831
    },
    {
      "epoch": 0.7758509955041747,
      "grad_norm": 0.2524743378162384,
      "learning_rate": 0.0001,
      "loss": 1.3753,
      "step": 4832
    },
    {
      "epoch": 0.7760115606936416,
      "grad_norm": 0.25471341609954834,
      "learning_rate": 0.0001,
      "loss": 1.4611,
      "step": 4833
    },
    {
      "epoch": 0.7761721258831086,
      "grad_norm": 0.27099505066871643,
      "learning_rate": 0.0001,
      "loss": 1.5078,
      "step": 4834
    },
    {
      "epoch": 0.7763326910725755,
      "grad_norm": 0.28913718461990356,
      "learning_rate": 0.0001,
      "loss": 1.5626,
      "step": 4835
    },
    {
      "epoch": 0.7764932562620424,
      "grad_norm": 0.29221460223197937,
      "learning_rate": 0.0001,
      "loss": 1.5956,
      "step": 4836
    },
    {
      "epoch": 0.7766538214515093,
      "grad_norm": 0.26697152853012085,
      "learning_rate": 0.0001,
      "loss": 1.5113,
      "step": 4837
    },
    {
      "epoch": 0.7768143866409762,
      "grad_norm": 0.26228007674217224,
      "learning_rate": 0.0001,
      "loss": 1.4794,
      "step": 4838
    },
    {
      "epoch": 0.7769749518304432,
      "grad_norm": 0.2753364145755768,
      "learning_rate": 0.0001,
      "loss": 1.5021,
      "step": 4839
    },
    {
      "epoch": 0.7771355170199101,
      "grad_norm": 0.27867865562438965,
      "learning_rate": 0.0001,
      "loss": 1.4644,
      "step": 4840
    },
    {
      "epoch": 0.777296082209377,
      "grad_norm": 0.276676207780838,
      "learning_rate": 0.0001,
      "loss": 1.4687,
      "step": 4841
    },
    {
      "epoch": 0.7774566473988439,
      "grad_norm": 0.26723018288612366,
      "learning_rate": 0.0001,
      "loss": 1.4774,
      "step": 4842
    },
    {
      "epoch": 0.7776172125883108,
      "grad_norm": 0.2919076383113861,
      "learning_rate": 0.0001,
      "loss": 1.4366,
      "step": 4843
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.257845014333725,
      "learning_rate": 0.0001,
      "loss": 1.4358,
      "step": 4844
    },
    {
      "epoch": 0.7779383429672447,
      "grad_norm": 0.2613871395587921,
      "learning_rate": 0.0001,
      "loss": 1.4362,
      "step": 4845
    },
    {
      "epoch": 0.7780989081567117,
      "grad_norm": 0.2684071362018585,
      "learning_rate": 0.0001,
      "loss": 1.4661,
      "step": 4846
    },
    {
      "epoch": 0.7782594733461785,
      "grad_norm": 0.2790563702583313,
      "learning_rate": 0.0001,
      "loss": 1.5563,
      "step": 4847
    },
    {
      "epoch": 0.7784200385356455,
      "grad_norm": 0.27308058738708496,
      "learning_rate": 0.0001,
      "loss": 1.4427,
      "step": 4848
    },
    {
      "epoch": 0.7785806037251124,
      "grad_norm": 0.2813565135002136,
      "learning_rate": 0.0001,
      "loss": 1.544,
      "step": 4849
    },
    {
      "epoch": 0.7787411689145793,
      "grad_norm": 0.27245306968688965,
      "learning_rate": 0.0001,
      "loss": 1.5118,
      "step": 4850
    },
    {
      "epoch": 0.7789017341040463,
      "grad_norm": 0.2604306936264038,
      "learning_rate": 0.0001,
      "loss": 1.4451,
      "step": 4851
    },
    {
      "epoch": 0.7790622992935131,
      "grad_norm": 0.24787943065166473,
      "learning_rate": 0.0001,
      "loss": 1.4135,
      "step": 4852
    },
    {
      "epoch": 0.7792228644829801,
      "grad_norm": 0.26028093695640564,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 4853
    },
    {
      "epoch": 0.779383429672447,
      "grad_norm": 0.2668739855289459,
      "learning_rate": 0.0001,
      "loss": 1.4557,
      "step": 4854
    },
    {
      "epoch": 0.779543994861914,
      "grad_norm": 0.2671746015548706,
      "learning_rate": 0.0001,
      "loss": 1.5064,
      "step": 4855
    },
    {
      "epoch": 0.7797045600513809,
      "grad_norm": 0.25071027874946594,
      "learning_rate": 0.0001,
      "loss": 1.4033,
      "step": 4856
    },
    {
      "epoch": 0.7798651252408478,
      "grad_norm": 0.27966752648353577,
      "learning_rate": 0.0001,
      "loss": 1.5315,
      "step": 4857
    },
    {
      "epoch": 0.7800256904303147,
      "grad_norm": 0.264328271150589,
      "learning_rate": 0.0001,
      "loss": 1.4493,
      "step": 4858
    },
    {
      "epoch": 0.7801862556197816,
      "grad_norm": 0.2697535753250122,
      "learning_rate": 0.0001,
      "loss": 1.4789,
      "step": 4859
    },
    {
      "epoch": 0.7803468208092486,
      "grad_norm": 0.2685675024986267,
      "learning_rate": 0.0001,
      "loss": 1.4747,
      "step": 4860
    },
    {
      "epoch": 0.7805073859987155,
      "grad_norm": 0.25597450137138367,
      "learning_rate": 0.0001,
      "loss": 1.5201,
      "step": 4861
    },
    {
      "epoch": 0.7806679511881824,
      "grad_norm": 0.2750207483768463,
      "learning_rate": 0.0001,
      "loss": 1.4902,
      "step": 4862
    },
    {
      "epoch": 0.7808285163776493,
      "grad_norm": 0.2637827694416046,
      "learning_rate": 0.0001,
      "loss": 1.534,
      "step": 4863
    },
    {
      "epoch": 0.7809890815671162,
      "grad_norm": 0.2612398564815521,
      "learning_rate": 0.0001,
      "loss": 1.4627,
      "step": 4864
    },
    {
      "epoch": 0.7811496467565832,
      "grad_norm": 0.25394272804260254,
      "learning_rate": 0.0001,
      "loss": 1.452,
      "step": 4865
    },
    {
      "epoch": 0.7813102119460501,
      "grad_norm": 0.2608264684677124,
      "learning_rate": 0.0001,
      "loss": 1.4986,
      "step": 4866
    },
    {
      "epoch": 0.781470777135517,
      "grad_norm": 0.2770000398159027,
      "learning_rate": 0.0001,
      "loss": 1.5841,
      "step": 4867
    },
    {
      "epoch": 0.781631342324984,
      "grad_norm": 0.2652257978916168,
      "learning_rate": 0.0001,
      "loss": 1.5138,
      "step": 4868
    },
    {
      "epoch": 0.7817919075144508,
      "grad_norm": 0.2749718129634857,
      "learning_rate": 0.0001,
      "loss": 1.4816,
      "step": 4869
    },
    {
      "epoch": 0.7819524727039178,
      "grad_norm": 0.25634610652923584,
      "learning_rate": 0.0001,
      "loss": 1.5285,
      "step": 4870
    },
    {
      "epoch": 0.7821130378933847,
      "grad_norm": 0.26548346877098083,
      "learning_rate": 0.0001,
      "loss": 1.4286,
      "step": 4871
    },
    {
      "epoch": 0.7822736030828517,
      "grad_norm": 0.2843562066555023,
      "learning_rate": 0.0001,
      "loss": 1.5951,
      "step": 4872
    },
    {
      "epoch": 0.7824341682723186,
      "grad_norm": 0.2538413107395172,
      "learning_rate": 0.0001,
      "loss": 1.4865,
      "step": 4873
    },
    {
      "epoch": 0.7825947334617854,
      "grad_norm": 0.2846985161304474,
      "learning_rate": 0.0001,
      "loss": 1.5546,
      "step": 4874
    },
    {
      "epoch": 0.7827552986512524,
      "grad_norm": 0.27610915899276733,
      "learning_rate": 0.0001,
      "loss": 1.5497,
      "step": 4875
    },
    {
      "epoch": 0.7829158638407193,
      "grad_norm": 0.27627769112586975,
      "learning_rate": 0.0001,
      "loss": 1.5147,
      "step": 4876
    },
    {
      "epoch": 0.7830764290301863,
      "grad_norm": 0.2938385307788849,
      "learning_rate": 0.0001,
      "loss": 1.4364,
      "step": 4877
    },
    {
      "epoch": 0.7832369942196532,
      "grad_norm": 0.2528034746646881,
      "learning_rate": 0.0001,
      "loss": 1.4407,
      "step": 4878
    },
    {
      "epoch": 0.7833975594091201,
      "grad_norm": 0.2692616581916809,
      "learning_rate": 0.0001,
      "loss": 1.4386,
      "step": 4879
    },
    {
      "epoch": 0.783558124598587,
      "grad_norm": 0.26693812012672424,
      "learning_rate": 0.0001,
      "loss": 1.5488,
      "step": 4880
    },
    {
      "epoch": 0.7837186897880539,
      "grad_norm": 0.3015109598636627,
      "learning_rate": 0.0001,
      "loss": 1.4796,
      "step": 4881
    },
    {
      "epoch": 0.7838792549775209,
      "grad_norm": 0.29318591952323914,
      "learning_rate": 0.0001,
      "loss": 1.5622,
      "step": 4882
    },
    {
      "epoch": 0.7840398201669878,
      "grad_norm": 0.2597436308860779,
      "learning_rate": 0.0001,
      "loss": 1.4494,
      "step": 4883
    },
    {
      "epoch": 0.7842003853564548,
      "grad_norm": 0.25885215401649475,
      "learning_rate": 0.0001,
      "loss": 1.5154,
      "step": 4884
    },
    {
      "epoch": 0.7843609505459217,
      "grad_norm": 0.2527919411659241,
      "learning_rate": 0.0001,
      "loss": 1.4266,
      "step": 4885
    },
    {
      "epoch": 0.7845215157353885,
      "grad_norm": 0.288007527589798,
      "learning_rate": 0.0001,
      "loss": 1.5183,
      "step": 4886
    },
    {
      "epoch": 0.7846820809248555,
      "grad_norm": 0.2749107778072357,
      "learning_rate": 0.0001,
      "loss": 1.4981,
      "step": 4887
    },
    {
      "epoch": 0.7848426461143224,
      "grad_norm": 0.2630927264690399,
      "learning_rate": 0.0001,
      "loss": 1.4514,
      "step": 4888
    },
    {
      "epoch": 0.7850032113037894,
      "grad_norm": 0.27400052547454834,
      "learning_rate": 0.0001,
      "loss": 1.5732,
      "step": 4889
    },
    {
      "epoch": 0.7851637764932563,
      "grad_norm": 0.2656361758708954,
      "learning_rate": 0.0001,
      "loss": 1.4284,
      "step": 4890
    },
    {
      "epoch": 0.7853243416827231,
      "grad_norm": 0.26750198006629944,
      "learning_rate": 0.0001,
      "loss": 1.4855,
      "step": 4891
    },
    {
      "epoch": 0.7854849068721901,
      "grad_norm": 0.2690584361553192,
      "learning_rate": 0.0001,
      "loss": 1.5387,
      "step": 4892
    },
    {
      "epoch": 0.785645472061657,
      "grad_norm": 0.2951248586177826,
      "learning_rate": 0.0001,
      "loss": 1.4459,
      "step": 4893
    },
    {
      "epoch": 0.785806037251124,
      "grad_norm": 0.2853166162967682,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 4894
    },
    {
      "epoch": 0.7859666024405909,
      "grad_norm": 0.25854942202568054,
      "learning_rate": 0.0001,
      "loss": 1.3991,
      "step": 4895
    },
    {
      "epoch": 0.7861271676300579,
      "grad_norm": 0.27628210186958313,
      "learning_rate": 0.0001,
      "loss": 1.5019,
      "step": 4896
    },
    {
      "epoch": 0.7862877328195247,
      "grad_norm": 0.2692117989063263,
      "learning_rate": 0.0001,
      "loss": 1.453,
      "step": 4897
    },
    {
      "epoch": 0.7864482980089916,
      "grad_norm": 0.2619384527206421,
      "learning_rate": 0.0001,
      "loss": 1.474,
      "step": 4898
    },
    {
      "epoch": 0.7866088631984586,
      "grad_norm": 0.26570191979408264,
      "learning_rate": 0.0001,
      "loss": 1.4745,
      "step": 4899
    },
    {
      "epoch": 0.7867694283879255,
      "grad_norm": 0.2800072729587555,
      "learning_rate": 0.0001,
      "loss": 1.5106,
      "step": 4900
    },
    {
      "epoch": 0.7869299935773925,
      "grad_norm": 0.2808997631072998,
      "learning_rate": 0.0001,
      "loss": 1.526,
      "step": 4901
    },
    {
      "epoch": 0.7870905587668593,
      "grad_norm": 0.2560919523239136,
      "learning_rate": 0.0001,
      "loss": 1.4728,
      "step": 4902
    },
    {
      "epoch": 0.7872511239563262,
      "grad_norm": 0.27413082122802734,
      "learning_rate": 0.0001,
      "loss": 1.5347,
      "step": 4903
    },
    {
      "epoch": 0.7874116891457932,
      "grad_norm": 0.26692163944244385,
      "learning_rate": 0.0001,
      "loss": 1.4403,
      "step": 4904
    },
    {
      "epoch": 0.7875722543352601,
      "grad_norm": 0.28223931789398193,
      "learning_rate": 0.0001,
      "loss": 1.4638,
      "step": 4905
    },
    {
      "epoch": 0.7877328195247271,
      "grad_norm": 0.2553137242794037,
      "learning_rate": 0.0001,
      "loss": 1.4612,
      "step": 4906
    },
    {
      "epoch": 0.787893384714194,
      "grad_norm": 0.30425265431404114,
      "learning_rate": 0.0001,
      "loss": 1.5459,
      "step": 4907
    },
    {
      "epoch": 0.7880539499036608,
      "grad_norm": 0.25780966877937317,
      "learning_rate": 0.0001,
      "loss": 1.3951,
      "step": 4908
    },
    {
      "epoch": 0.7882145150931278,
      "grad_norm": 0.27276623249053955,
      "learning_rate": 0.0001,
      "loss": 1.5178,
      "step": 4909
    },
    {
      "epoch": 0.7883750802825947,
      "grad_norm": 0.27824667096138,
      "learning_rate": 0.0001,
      "loss": 1.4709,
      "step": 4910
    },
    {
      "epoch": 0.7885356454720617,
      "grad_norm": 0.27756956219673157,
      "learning_rate": 0.0001,
      "loss": 1.53,
      "step": 4911
    },
    {
      "epoch": 0.7886962106615286,
      "grad_norm": 0.2589087188243866,
      "learning_rate": 0.0001,
      "loss": 1.42,
      "step": 4912
    },
    {
      "epoch": 0.7888567758509955,
      "grad_norm": 0.26565444469451904,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 4913
    },
    {
      "epoch": 0.7890173410404624,
      "grad_norm": 0.272074431180954,
      "learning_rate": 0.0001,
      "loss": 1.4637,
      "step": 4914
    },
    {
      "epoch": 0.7891779062299293,
      "grad_norm": 0.257182240486145,
      "learning_rate": 0.0001,
      "loss": 1.5132,
      "step": 4915
    },
    {
      "epoch": 0.7893384714193963,
      "grad_norm": 0.2661198079586029,
      "learning_rate": 0.0001,
      "loss": 1.4332,
      "step": 4916
    },
    {
      "epoch": 0.7894990366088632,
      "grad_norm": 0.3132728040218353,
      "learning_rate": 0.0001,
      "loss": 1.4547,
      "step": 4917
    },
    {
      "epoch": 0.7896596017983302,
      "grad_norm": 0.27426013350486755,
      "learning_rate": 0.0001,
      "loss": 1.4991,
      "step": 4918
    },
    {
      "epoch": 0.789820166987797,
      "grad_norm": 0.27901703119277954,
      "learning_rate": 0.0001,
      "loss": 1.5383,
      "step": 4919
    },
    {
      "epoch": 0.789980732177264,
      "grad_norm": 0.25175678730010986,
      "learning_rate": 0.0001,
      "loss": 1.5175,
      "step": 4920
    },
    {
      "epoch": 0.7901412973667309,
      "grad_norm": 0.2794884145259857,
      "learning_rate": 0.0001,
      "loss": 1.5227,
      "step": 4921
    },
    {
      "epoch": 0.7903018625561978,
      "grad_norm": 0.25591138005256653,
      "learning_rate": 0.0001,
      "loss": 1.4893,
      "step": 4922
    },
    {
      "epoch": 0.7904624277456648,
      "grad_norm": 0.26282110810279846,
      "learning_rate": 0.0001,
      "loss": 1.4878,
      "step": 4923
    },
    {
      "epoch": 0.7906229929351317,
      "grad_norm": 0.3139071464538574,
      "learning_rate": 0.0001,
      "loss": 1.4877,
      "step": 4924
    },
    {
      "epoch": 0.7907835581245986,
      "grad_norm": 0.2633921504020691,
      "learning_rate": 0.0001,
      "loss": 1.3999,
      "step": 4925
    },
    {
      "epoch": 0.7909441233140655,
      "grad_norm": 0.2790989577770233,
      "learning_rate": 0.0001,
      "loss": 1.4996,
      "step": 4926
    },
    {
      "epoch": 0.7911046885035324,
      "grad_norm": 0.26268425583839417,
      "learning_rate": 0.0001,
      "loss": 1.4982,
      "step": 4927
    },
    {
      "epoch": 0.7912652536929994,
      "grad_norm": 0.27156296372413635,
      "learning_rate": 0.0001,
      "loss": 1.445,
      "step": 4928
    },
    {
      "epoch": 0.7914258188824663,
      "grad_norm": 0.250273734331131,
      "learning_rate": 0.0001,
      "loss": 1.4499,
      "step": 4929
    },
    {
      "epoch": 0.7915863840719332,
      "grad_norm": 0.2585732042789459,
      "learning_rate": 0.0001,
      "loss": 1.419,
      "step": 4930
    },
    {
      "epoch": 0.7917469492614001,
      "grad_norm": 0.2718723714351654,
      "learning_rate": 0.0001,
      "loss": 1.4687,
      "step": 4931
    },
    {
      "epoch": 0.791907514450867,
      "grad_norm": 0.26102644205093384,
      "learning_rate": 0.0001,
      "loss": 1.4524,
      "step": 4932
    },
    {
      "epoch": 0.792068079640334,
      "grad_norm": 0.2890603840351105,
      "learning_rate": 0.0001,
      "loss": 1.4379,
      "step": 4933
    },
    {
      "epoch": 0.7922286448298009,
      "grad_norm": 0.28953224420547485,
      "learning_rate": 0.0001,
      "loss": 1.5449,
      "step": 4934
    },
    {
      "epoch": 0.7923892100192679,
      "grad_norm": 0.26467910408973694,
      "learning_rate": 0.0001,
      "loss": 1.4644,
      "step": 4935
    },
    {
      "epoch": 0.7925497752087347,
      "grad_norm": 0.610118567943573,
      "learning_rate": 0.0001,
      "loss": 1.4814,
      "step": 4936
    },
    {
      "epoch": 0.7927103403982017,
      "grad_norm": 0.2836211323738098,
      "learning_rate": 0.0001,
      "loss": 1.5604,
      "step": 4937
    },
    {
      "epoch": 0.7928709055876686,
      "grad_norm": 0.26038748025894165,
      "learning_rate": 0.0001,
      "loss": 1.4935,
      "step": 4938
    },
    {
      "epoch": 0.7930314707771355,
      "grad_norm": 0.2619526982307434,
      "learning_rate": 0.0001,
      "loss": 1.5362,
      "step": 4939
    },
    {
      "epoch": 0.7931920359666025,
      "grad_norm": 0.2593645751476288,
      "learning_rate": 0.0001,
      "loss": 1.3833,
      "step": 4940
    },
    {
      "epoch": 0.7933526011560693,
      "grad_norm": 0.27356091141700745,
      "learning_rate": 0.0001,
      "loss": 1.4723,
      "step": 4941
    },
    {
      "epoch": 0.7935131663455363,
      "grad_norm": 0.3041076362133026,
      "learning_rate": 0.0001,
      "loss": 1.4685,
      "step": 4942
    },
    {
      "epoch": 0.7936737315350032,
      "grad_norm": 0.2665919065475464,
      "learning_rate": 0.0001,
      "loss": 1.4729,
      "step": 4943
    },
    {
      "epoch": 0.7938342967244701,
      "grad_norm": 0.27540159225463867,
      "learning_rate": 0.0001,
      "loss": 1.4914,
      "step": 4944
    },
    {
      "epoch": 0.7939948619139371,
      "grad_norm": 0.2577289938926697,
      "learning_rate": 0.0001,
      "loss": 1.4944,
      "step": 4945
    },
    {
      "epoch": 0.794155427103404,
      "grad_norm": 0.261925607919693,
      "learning_rate": 0.0001,
      "loss": 1.451,
      "step": 4946
    },
    {
      "epoch": 0.7943159922928709,
      "grad_norm": 0.2669730484485626,
      "learning_rate": 0.0001,
      "loss": 1.5532,
      "step": 4947
    },
    {
      "epoch": 0.7944765574823378,
      "grad_norm": 0.2583950459957123,
      "learning_rate": 0.0001,
      "loss": 1.4655,
      "step": 4948
    },
    {
      "epoch": 0.7946371226718048,
      "grad_norm": 0.2625668942928314,
      "learning_rate": 0.0001,
      "loss": 1.4777,
      "step": 4949
    },
    {
      "epoch": 0.7947976878612717,
      "grad_norm": 0.2661846876144409,
      "learning_rate": 0.0001,
      "loss": 1.5237,
      "step": 4950
    },
    {
      "epoch": 0.7949582530507386,
      "grad_norm": 0.2708057463169098,
      "learning_rate": 0.0001,
      "loss": 1.5202,
      "step": 4951
    },
    {
      "epoch": 0.7951188182402055,
      "grad_norm": 0.26769402623176575,
      "learning_rate": 0.0001,
      "loss": 1.5222,
      "step": 4952
    },
    {
      "epoch": 0.7952793834296724,
      "grad_norm": 0.267253577709198,
      "learning_rate": 0.0001,
      "loss": 1.4605,
      "step": 4953
    },
    {
      "epoch": 0.7954399486191394,
      "grad_norm": 0.2620091438293457,
      "learning_rate": 0.0001,
      "loss": 1.4502,
      "step": 4954
    },
    {
      "epoch": 0.7956005138086063,
      "grad_norm": 0.27369511127471924,
      "learning_rate": 0.0001,
      "loss": 1.4489,
      "step": 4955
    },
    {
      "epoch": 0.7957610789980732,
      "grad_norm": 0.2532361149787903,
      "learning_rate": 0.0001,
      "loss": 1.4876,
      "step": 4956
    },
    {
      "epoch": 0.7959216441875402,
      "grad_norm": 0.258556991815567,
      "learning_rate": 0.0001,
      "loss": 1.4314,
      "step": 4957
    },
    {
      "epoch": 0.796082209377007,
      "grad_norm": 0.2675907611846924,
      "learning_rate": 0.0001,
      "loss": 1.4406,
      "step": 4958
    },
    {
      "epoch": 0.796242774566474,
      "grad_norm": 0.2844993770122528,
      "learning_rate": 0.0001,
      "loss": 1.4963,
      "step": 4959
    },
    {
      "epoch": 0.7964033397559409,
      "grad_norm": 0.2730778455734253,
      "learning_rate": 0.0001,
      "loss": 1.5229,
      "step": 4960
    },
    {
      "epoch": 0.7965639049454079,
      "grad_norm": 0.29573535919189453,
      "learning_rate": 0.0001,
      "loss": 1.5178,
      "step": 4961
    },
    {
      "epoch": 0.7967244701348748,
      "grad_norm": 0.27346470952033997,
      "learning_rate": 0.0001,
      "loss": 1.4969,
      "step": 4962
    },
    {
      "epoch": 0.7968850353243416,
      "grad_norm": 0.29766732454299927,
      "learning_rate": 0.0001,
      "loss": 1.4557,
      "step": 4963
    },
    {
      "epoch": 0.7970456005138086,
      "grad_norm": 0.262014776468277,
      "learning_rate": 0.0001,
      "loss": 1.4524,
      "step": 4964
    },
    {
      "epoch": 0.7972061657032755,
      "grad_norm": 0.2678148150444031,
      "learning_rate": 0.0001,
      "loss": 1.4671,
      "step": 4965
    },
    {
      "epoch": 0.7973667308927425,
      "grad_norm": 0.2627437114715576,
      "learning_rate": 0.0001,
      "loss": 1.4819,
      "step": 4966
    },
    {
      "epoch": 0.7975272960822094,
      "grad_norm": 0.2854844927787781,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 4967
    },
    {
      "epoch": 0.7976878612716763,
      "grad_norm": 0.26114368438720703,
      "learning_rate": 0.0001,
      "loss": 1.4042,
      "step": 4968
    },
    {
      "epoch": 0.7978484264611432,
      "grad_norm": 0.2642273008823395,
      "learning_rate": 0.0001,
      "loss": 1.5096,
      "step": 4969
    },
    {
      "epoch": 0.7980089916506101,
      "grad_norm": 0.25962576270103455,
      "learning_rate": 0.0001,
      "loss": 1.3862,
      "step": 4970
    },
    {
      "epoch": 0.7981695568400771,
      "grad_norm": 0.27187272906303406,
      "learning_rate": 0.0001,
      "loss": 1.4273,
      "step": 4971
    },
    {
      "epoch": 0.798330122029544,
      "grad_norm": 0.26994588971138,
      "learning_rate": 0.0001,
      "loss": 1.5165,
      "step": 4972
    },
    {
      "epoch": 0.798490687219011,
      "grad_norm": 0.2657373249530792,
      "learning_rate": 0.0001,
      "loss": 1.5219,
      "step": 4973
    },
    {
      "epoch": 0.7986512524084779,
      "grad_norm": 0.2619072496891022,
      "learning_rate": 0.0001,
      "loss": 1.4555,
      "step": 4974
    },
    {
      "epoch": 0.7988118175979447,
      "grad_norm": 0.283348947763443,
      "learning_rate": 0.0001,
      "loss": 1.5939,
      "step": 4975
    },
    {
      "epoch": 0.7989723827874117,
      "grad_norm": 0.2702791392803192,
      "learning_rate": 0.0001,
      "loss": 1.484,
      "step": 4976
    },
    {
      "epoch": 0.7991329479768786,
      "grad_norm": 0.26904091238975525,
      "learning_rate": 0.0001,
      "loss": 1.4473,
      "step": 4977
    },
    {
      "epoch": 0.7992935131663456,
      "grad_norm": 0.2811717987060547,
      "learning_rate": 0.0001,
      "loss": 1.445,
      "step": 4978
    },
    {
      "epoch": 0.7994540783558125,
      "grad_norm": 0.28999173641204834,
      "learning_rate": 0.0001,
      "loss": 1.5023,
      "step": 4979
    },
    {
      "epoch": 0.7996146435452793,
      "grad_norm": 0.2871098518371582,
      "learning_rate": 0.0001,
      "loss": 1.4955,
      "step": 4980
    },
    {
      "epoch": 0.7997752087347463,
      "grad_norm": 0.2720760107040405,
      "learning_rate": 0.0001,
      "loss": 1.4889,
      "step": 4981
    },
    {
      "epoch": 0.7999357739242132,
      "grad_norm": 0.25334689021110535,
      "learning_rate": 0.0001,
      "loss": 1.3643,
      "step": 4982
    },
    {
      "epoch": 0.8000963391136802,
      "grad_norm": 0.2747771441936493,
      "learning_rate": 0.0001,
      "loss": 1.4334,
      "step": 4983
    },
    {
      "epoch": 0.8002569043031471,
      "grad_norm": 0.2666880190372467,
      "learning_rate": 0.0001,
      "loss": 1.578,
      "step": 4984
    },
    {
      "epoch": 0.800417469492614,
      "grad_norm": 0.279110848903656,
      "learning_rate": 0.0001,
      "loss": 1.4749,
      "step": 4985
    },
    {
      "epoch": 0.8005780346820809,
      "grad_norm": 0.27905863523483276,
      "learning_rate": 0.0001,
      "loss": 1.4997,
      "step": 4986
    },
    {
      "epoch": 0.8007385998715478,
      "grad_norm": 0.2762279212474823,
      "learning_rate": 0.0001,
      "loss": 1.5345,
      "step": 4987
    },
    {
      "epoch": 0.8008991650610148,
      "grad_norm": 0.2663026750087738,
      "learning_rate": 0.0001,
      "loss": 1.4433,
      "step": 4988
    },
    {
      "epoch": 0.8010597302504817,
      "grad_norm": 0.275191068649292,
      "learning_rate": 0.0001,
      "loss": 1.5455,
      "step": 4989
    },
    {
      "epoch": 0.8012202954399487,
      "grad_norm": 0.2972000241279602,
      "learning_rate": 0.0001,
      "loss": 1.5177,
      "step": 4990
    },
    {
      "epoch": 0.8013808606294155,
      "grad_norm": 0.2799358367919922,
      "learning_rate": 0.0001,
      "loss": 1.582,
      "step": 4991
    },
    {
      "epoch": 0.8015414258188824,
      "grad_norm": 0.28580302000045776,
      "learning_rate": 0.0001,
      "loss": 1.5184,
      "step": 4992
    },
    {
      "epoch": 0.8017019910083494,
      "grad_norm": 0.26291462779045105,
      "learning_rate": 0.0001,
      "loss": 1.4824,
      "step": 4993
    },
    {
      "epoch": 0.8018625561978163,
      "grad_norm": 0.29605501890182495,
      "learning_rate": 0.0001,
      "loss": 1.5611,
      "step": 4994
    },
    {
      "epoch": 0.8020231213872833,
      "grad_norm": 0.2709025740623474,
      "learning_rate": 0.0001,
      "loss": 1.5195,
      "step": 4995
    },
    {
      "epoch": 0.8021836865767502,
      "grad_norm": 0.26770129799842834,
      "learning_rate": 0.0001,
      "loss": 1.4864,
      "step": 4996
    },
    {
      "epoch": 0.802344251766217,
      "grad_norm": 0.26983755826950073,
      "learning_rate": 0.0001,
      "loss": 1.511,
      "step": 4997
    },
    {
      "epoch": 0.802504816955684,
      "grad_norm": 0.2597654163837433,
      "learning_rate": 0.0001,
      "loss": 1.5056,
      "step": 4998
    },
    {
      "epoch": 0.8026653821451509,
      "grad_norm": 0.274734765291214,
      "learning_rate": 0.0001,
      "loss": 1.5336,
      "step": 4999
    },
    {
      "epoch": 0.8028259473346179,
      "grad_norm": 0.2811043858528137,
      "learning_rate": 0.0001,
      "loss": 1.4991,
      "step": 5000
    },
    {
      "epoch": 0.8029865125240848,
      "grad_norm": 0.26507386565208435,
      "learning_rate": 0.0001,
      "loss": 1.3735,
      "step": 5001
    },
    {
      "epoch": 0.8031470777135516,
      "grad_norm": 0.27205589413642883,
      "learning_rate": 0.0001,
      "loss": 1.4867,
      "step": 5002
    },
    {
      "epoch": 0.8033076429030186,
      "grad_norm": 0.2646946609020233,
      "learning_rate": 0.0001,
      "loss": 1.505,
      "step": 5003
    },
    {
      "epoch": 0.8034682080924855,
      "grad_norm": 0.27561819553375244,
      "learning_rate": 0.0001,
      "loss": 1.522,
      "step": 5004
    },
    {
      "epoch": 0.8036287732819525,
      "grad_norm": 0.25112950801849365,
      "learning_rate": 0.0001,
      "loss": 1.4336,
      "step": 5005
    },
    {
      "epoch": 0.8037893384714194,
      "grad_norm": 0.2693318724632263,
      "learning_rate": 0.0001,
      "loss": 1.4009,
      "step": 5006
    },
    {
      "epoch": 0.8039499036608864,
      "grad_norm": 0.2844517230987549,
      "learning_rate": 0.0001,
      "loss": 1.5084,
      "step": 5007
    },
    {
      "epoch": 0.8041104688503532,
      "grad_norm": 2.04303240776062,
      "learning_rate": 0.0001,
      "loss": 1.4859,
      "step": 5008
    },
    {
      "epoch": 0.8042710340398201,
      "grad_norm": 0.2620633542537689,
      "learning_rate": 0.0001,
      "loss": 1.5667,
      "step": 5009
    },
    {
      "epoch": 0.8044315992292871,
      "grad_norm": 0.28380200266838074,
      "learning_rate": 0.0001,
      "loss": 1.4166,
      "step": 5010
    },
    {
      "epoch": 0.804592164418754,
      "grad_norm": 0.31242579221725464,
      "learning_rate": 0.0001,
      "loss": 1.5084,
      "step": 5011
    },
    {
      "epoch": 0.804752729608221,
      "grad_norm": 0.28257498145103455,
      "learning_rate": 0.0001,
      "loss": 1.4796,
      "step": 5012
    },
    {
      "epoch": 0.8049132947976878,
      "grad_norm": 0.3021131455898285,
      "learning_rate": 0.0001,
      "loss": 1.4714,
      "step": 5013
    },
    {
      "epoch": 0.8050738599871547,
      "grad_norm": 0.2672417461872101,
      "learning_rate": 0.0001,
      "loss": 1.4839,
      "step": 5014
    },
    {
      "epoch": 0.8052344251766217,
      "grad_norm": 0.27186620235443115,
      "learning_rate": 0.0001,
      "loss": 1.4964,
      "step": 5015
    },
    {
      "epoch": 0.8053949903660886,
      "grad_norm": 0.26172634959220886,
      "learning_rate": 0.0001,
      "loss": 1.4633,
      "step": 5016
    },
    {
      "epoch": 0.8055555555555556,
      "grad_norm": 0.2712869346141815,
      "learning_rate": 0.0001,
      "loss": 1.4988,
      "step": 5017
    },
    {
      "epoch": 0.8057161207450225,
      "grad_norm": 0.2918786108493805,
      "learning_rate": 0.0001,
      "loss": 1.5346,
      "step": 5018
    },
    {
      "epoch": 0.8058766859344894,
      "grad_norm": 0.28895458579063416,
      "learning_rate": 0.0001,
      "loss": 1.4888,
      "step": 5019
    },
    {
      "epoch": 0.8060372511239563,
      "grad_norm": 0.27583470940589905,
      "learning_rate": 0.0001,
      "loss": 1.4827,
      "step": 5020
    },
    {
      "epoch": 0.8061978163134232,
      "grad_norm": 0.26342907547950745,
      "learning_rate": 0.0001,
      "loss": 1.4998,
      "step": 5021
    },
    {
      "epoch": 0.8063583815028902,
      "grad_norm": 0.2677418887615204,
      "learning_rate": 0.0001,
      "loss": 1.4933,
      "step": 5022
    },
    {
      "epoch": 0.8065189466923571,
      "grad_norm": 0.289798378944397,
      "learning_rate": 0.0001,
      "loss": 1.5267,
      "step": 5023
    },
    {
      "epoch": 0.8066795118818241,
      "grad_norm": 0.28941789269447327,
      "learning_rate": 0.0001,
      "loss": 1.5593,
      "step": 5024
    },
    {
      "epoch": 0.8068400770712909,
      "grad_norm": 0.270546019077301,
      "learning_rate": 0.0001,
      "loss": 1.5388,
      "step": 5025
    },
    {
      "epoch": 0.8070006422607579,
      "grad_norm": 0.2650315463542938,
      "learning_rate": 0.0001,
      "loss": 1.4661,
      "step": 5026
    },
    {
      "epoch": 0.8071612074502248,
      "grad_norm": 0.289795458316803,
      "learning_rate": 0.0001,
      "loss": 1.5281,
      "step": 5027
    },
    {
      "epoch": 0.8073217726396917,
      "grad_norm": 0.28381696343421936,
      "learning_rate": 0.0001,
      "loss": 1.5533,
      "step": 5028
    },
    {
      "epoch": 0.8074823378291587,
      "grad_norm": 0.2664477527141571,
      "learning_rate": 0.0001,
      "loss": 1.4749,
      "step": 5029
    },
    {
      "epoch": 0.8076429030186255,
      "grad_norm": 0.26828449964523315,
      "learning_rate": 0.0001,
      "loss": 1.5559,
      "step": 5030
    },
    {
      "epoch": 0.8078034682080925,
      "grad_norm": 0.28125303983688354,
      "learning_rate": 0.0001,
      "loss": 1.4843,
      "step": 5031
    },
    {
      "epoch": 0.8079640333975594,
      "grad_norm": 0.32039839029312134,
      "learning_rate": 0.0001,
      "loss": 1.4856,
      "step": 5032
    },
    {
      "epoch": 0.8081245985870263,
      "grad_norm": 0.26907145977020264,
      "learning_rate": 0.0001,
      "loss": 1.4669,
      "step": 5033
    },
    {
      "epoch": 0.8082851637764933,
      "grad_norm": 0.258063405752182,
      "learning_rate": 0.0001,
      "loss": 1.5126,
      "step": 5034
    },
    {
      "epoch": 0.8084457289659602,
      "grad_norm": 0.2588548958301544,
      "learning_rate": 0.0001,
      "loss": 1.465,
      "step": 5035
    },
    {
      "epoch": 0.8086062941554271,
      "grad_norm": 0.255706787109375,
      "learning_rate": 0.0001,
      "loss": 1.433,
      "step": 5036
    },
    {
      "epoch": 0.808766859344894,
      "grad_norm": 0.27706262469291687,
      "learning_rate": 0.0001,
      "loss": 1.5994,
      "step": 5037
    },
    {
      "epoch": 0.808927424534361,
      "grad_norm": 0.31538891792297363,
      "learning_rate": 0.0001,
      "loss": 1.4693,
      "step": 5038
    },
    {
      "epoch": 0.8090879897238279,
      "grad_norm": 0.27659207582473755,
      "learning_rate": 0.0001,
      "loss": 1.4528,
      "step": 5039
    },
    {
      "epoch": 0.8092485549132948,
      "grad_norm": 0.2625981569290161,
      "learning_rate": 0.0001,
      "loss": 1.5124,
      "step": 5040
    },
    {
      "epoch": 0.8094091201027617,
      "grad_norm": 0.2831464111804962,
      "learning_rate": 0.0001,
      "loss": 1.6326,
      "step": 5041
    },
    {
      "epoch": 0.8095696852922286,
      "grad_norm": 0.29719650745391846,
      "learning_rate": 0.0001,
      "loss": 1.4677,
      "step": 5042
    },
    {
      "epoch": 0.8097302504816956,
      "grad_norm": 0.33954355120658875,
      "learning_rate": 0.0001,
      "loss": 1.5095,
      "step": 5043
    },
    {
      "epoch": 0.8098908156711625,
      "grad_norm": 0.2682896852493286,
      "learning_rate": 0.0001,
      "loss": 1.4545,
      "step": 5044
    },
    {
      "epoch": 0.8100513808606294,
      "grad_norm": 0.2693544030189514,
      "learning_rate": 0.0001,
      "loss": 1.4792,
      "step": 5045
    },
    {
      "epoch": 0.8102119460500964,
      "grad_norm": 0.31356045603752136,
      "learning_rate": 0.0001,
      "loss": 1.4621,
      "step": 5046
    },
    {
      "epoch": 0.8103725112395632,
      "grad_norm": 0.2878251075744629,
      "learning_rate": 0.0001,
      "loss": 1.5351,
      "step": 5047
    },
    {
      "epoch": 0.8105330764290302,
      "grad_norm": 0.27377381920814514,
      "learning_rate": 0.0001,
      "loss": 1.4388,
      "step": 5048
    },
    {
      "epoch": 0.8106936416184971,
      "grad_norm": 0.27631136775016785,
      "learning_rate": 0.0001,
      "loss": 1.5438,
      "step": 5049
    },
    {
      "epoch": 0.810854206807964,
      "grad_norm": 0.2838387191295624,
      "learning_rate": 0.0001,
      "loss": 1.4786,
      "step": 5050
    },
    {
      "epoch": 0.811014771997431,
      "grad_norm": 0.2612224817276001,
      "learning_rate": 0.0001,
      "loss": 1.5201,
      "step": 5051
    },
    {
      "epoch": 0.8111753371868978,
      "grad_norm": 0.26157259941101074,
      "learning_rate": 0.0001,
      "loss": 1.5067,
      "step": 5052
    },
    {
      "epoch": 0.8113359023763648,
      "grad_norm": 0.278354287147522,
      "learning_rate": 0.0001,
      "loss": 1.5035,
      "step": 5053
    },
    {
      "epoch": 0.8114964675658317,
      "grad_norm": 0.2769748270511627,
      "learning_rate": 0.0001,
      "loss": 1.4643,
      "step": 5054
    },
    {
      "epoch": 0.8116570327552987,
      "grad_norm": 0.2752477824687958,
      "learning_rate": 0.0001,
      "loss": 1.5277,
      "step": 5055
    },
    {
      "epoch": 0.8118175979447656,
      "grad_norm": 0.2681583762168884,
      "learning_rate": 0.0001,
      "loss": 1.5103,
      "step": 5056
    },
    {
      "epoch": 0.8119781631342325,
      "grad_norm": 0.26612240076065063,
      "learning_rate": 0.0001,
      "loss": 1.5354,
      "step": 5057
    },
    {
      "epoch": 0.8121387283236994,
      "grad_norm": 0.2747218608856201,
      "learning_rate": 0.0001,
      "loss": 1.4617,
      "step": 5058
    },
    {
      "epoch": 0.8122992935131663,
      "grad_norm": 0.24961914122104645,
      "learning_rate": 0.0001,
      "loss": 1.2948,
      "step": 5059
    },
    {
      "epoch": 0.8124598587026333,
      "grad_norm": 0.2838800847530365,
      "learning_rate": 0.0001,
      "loss": 1.5084,
      "step": 5060
    },
    {
      "epoch": 0.8126204238921002,
      "grad_norm": 0.2648646831512451,
      "learning_rate": 0.0001,
      "loss": 1.506,
      "step": 5061
    },
    {
      "epoch": 0.8127809890815672,
      "grad_norm": 0.2676868438720703,
      "learning_rate": 0.0001,
      "loss": 1.4899,
      "step": 5062
    },
    {
      "epoch": 0.8129415542710341,
      "grad_norm": 0.24974873661994934,
      "learning_rate": 0.0001,
      "loss": 1.4207,
      "step": 5063
    },
    {
      "epoch": 0.8131021194605009,
      "grad_norm": 0.2762000262737274,
      "learning_rate": 0.0001,
      "loss": 1.4827,
      "step": 5064
    },
    {
      "epoch": 0.8132626846499679,
      "grad_norm": 0.2552899420261383,
      "learning_rate": 0.0001,
      "loss": 1.5017,
      "step": 5065
    },
    {
      "epoch": 0.8134232498394348,
      "grad_norm": 0.2649138867855072,
      "learning_rate": 0.0001,
      "loss": 1.375,
      "step": 5066
    },
    {
      "epoch": 0.8135838150289018,
      "grad_norm": 0.26335644721984863,
      "learning_rate": 0.0001,
      "loss": 1.4779,
      "step": 5067
    },
    {
      "epoch": 0.8137443802183687,
      "grad_norm": 0.26281577348709106,
      "learning_rate": 0.0001,
      "loss": 1.4712,
      "step": 5068
    },
    {
      "epoch": 0.8139049454078355,
      "grad_norm": 0.26597362756729126,
      "learning_rate": 0.0001,
      "loss": 1.4837,
      "step": 5069
    },
    {
      "epoch": 0.8140655105973025,
      "grad_norm": 0.27353256940841675,
      "learning_rate": 0.0001,
      "loss": 1.5319,
      "step": 5070
    },
    {
      "epoch": 0.8142260757867694,
      "grad_norm": 0.2876351773738861,
      "learning_rate": 0.0001,
      "loss": 1.4816,
      "step": 5071
    },
    {
      "epoch": 0.8143866409762364,
      "grad_norm": 0.25910821557044983,
      "learning_rate": 0.0001,
      "loss": 1.4715,
      "step": 5072
    },
    {
      "epoch": 0.8145472061657033,
      "grad_norm": 0.2710203528404236,
      "learning_rate": 0.0001,
      "loss": 1.4432,
      "step": 5073
    },
    {
      "epoch": 0.8147077713551703,
      "grad_norm": 0.2655072808265686,
      "learning_rate": 0.0001,
      "loss": 1.5376,
      "step": 5074
    },
    {
      "epoch": 0.8148683365446371,
      "grad_norm": 0.2728174030780792,
      "learning_rate": 0.0001,
      "loss": 1.5714,
      "step": 5075
    },
    {
      "epoch": 0.815028901734104,
      "grad_norm": 0.26975101232528687,
      "learning_rate": 0.0001,
      "loss": 1.4301,
      "step": 5076
    },
    {
      "epoch": 0.815189466923571,
      "grad_norm": 0.25748392939567566,
      "learning_rate": 0.0001,
      "loss": 1.5084,
      "step": 5077
    },
    {
      "epoch": 0.8153500321130379,
      "grad_norm": 0.25959932804107666,
      "learning_rate": 0.0001,
      "loss": 1.5004,
      "step": 5078
    },
    {
      "epoch": 0.8155105973025049,
      "grad_norm": 0.25693467259407043,
      "learning_rate": 0.0001,
      "loss": 1.4671,
      "step": 5079
    },
    {
      "epoch": 0.8156711624919717,
      "grad_norm": 0.26165667176246643,
      "learning_rate": 0.0001,
      "loss": 1.3918,
      "step": 5080
    },
    {
      "epoch": 0.8158317276814386,
      "grad_norm": 0.27073153853416443,
      "learning_rate": 0.0001,
      "loss": 1.5717,
      "step": 5081
    },
    {
      "epoch": 0.8159922928709056,
      "grad_norm": 0.27127423882484436,
      "learning_rate": 0.0001,
      "loss": 1.5349,
      "step": 5082
    },
    {
      "epoch": 0.8161528580603725,
      "grad_norm": 0.26971322298049927,
      "learning_rate": 0.0001,
      "loss": 1.454,
      "step": 5083
    },
    {
      "epoch": 0.8163134232498395,
      "grad_norm": 0.26631149649620056,
      "learning_rate": 0.0001,
      "loss": 1.5591,
      "step": 5084
    },
    {
      "epoch": 0.8164739884393064,
      "grad_norm": 0.25626811385154724,
      "learning_rate": 0.0001,
      "loss": 1.4841,
      "step": 5085
    },
    {
      "epoch": 0.8166345536287732,
      "grad_norm": 0.274842768907547,
      "learning_rate": 0.0001,
      "loss": 1.4183,
      "step": 5086
    },
    {
      "epoch": 0.8167951188182402,
      "grad_norm": 0.2677263021469116,
      "learning_rate": 0.0001,
      "loss": 1.4909,
      "step": 5087
    },
    {
      "epoch": 0.8169556840077071,
      "grad_norm": 0.26319411396980286,
      "learning_rate": 0.0001,
      "loss": 1.4075,
      "step": 5088
    },
    {
      "epoch": 0.8171162491971741,
      "grad_norm": 0.2668224573135376,
      "learning_rate": 0.0001,
      "loss": 1.4753,
      "step": 5089
    },
    {
      "epoch": 0.817276814386641,
      "grad_norm": 0.27852606773376465,
      "learning_rate": 0.0001,
      "loss": 1.5219,
      "step": 5090
    },
    {
      "epoch": 0.8174373795761078,
      "grad_norm": 0.26831531524658203,
      "learning_rate": 0.0001,
      "loss": 1.5035,
      "step": 5091
    },
    {
      "epoch": 0.8175979447655748,
      "grad_norm": 0.29036858677864075,
      "learning_rate": 0.0001,
      "loss": 1.4808,
      "step": 5092
    },
    {
      "epoch": 0.8177585099550417,
      "grad_norm": 0.2693397104740143,
      "learning_rate": 0.0001,
      "loss": 1.4563,
      "step": 5093
    },
    {
      "epoch": 0.8179190751445087,
      "grad_norm": 0.26656654477119446,
      "learning_rate": 0.0001,
      "loss": 1.459,
      "step": 5094
    },
    {
      "epoch": 0.8180796403339756,
      "grad_norm": 0.4419085681438446,
      "learning_rate": 0.0001,
      "loss": 1.6211,
      "step": 5095
    },
    {
      "epoch": 0.8182402055234426,
      "grad_norm": 0.246019184589386,
      "learning_rate": 0.0001,
      "loss": 1.3657,
      "step": 5096
    },
    {
      "epoch": 0.8184007707129094,
      "grad_norm": 0.2562299966812134,
      "learning_rate": 0.0001,
      "loss": 1.4701,
      "step": 5097
    },
    {
      "epoch": 0.8185613359023763,
      "grad_norm": 0.2575107514858246,
      "learning_rate": 0.0001,
      "loss": 1.5029,
      "step": 5098
    },
    {
      "epoch": 0.8187219010918433,
      "grad_norm": 0.2756263017654419,
      "learning_rate": 0.0001,
      "loss": 1.5079,
      "step": 5099
    },
    {
      "epoch": 0.8188824662813102,
      "grad_norm": 0.2649906873703003,
      "learning_rate": 0.0001,
      "loss": 1.5112,
      "step": 5100
    },
    {
      "epoch": 0.8190430314707772,
      "grad_norm": 0.2766910195350647,
      "learning_rate": 0.0001,
      "loss": 1.4806,
      "step": 5101
    },
    {
      "epoch": 0.819203596660244,
      "grad_norm": 0.2611122131347656,
      "learning_rate": 0.0001,
      "loss": 1.4494,
      "step": 5102
    },
    {
      "epoch": 0.819364161849711,
      "grad_norm": 0.2742558717727661,
      "learning_rate": 0.0001,
      "loss": 1.5182,
      "step": 5103
    },
    {
      "epoch": 0.8195247270391779,
      "grad_norm": 0.272408127784729,
      "learning_rate": 0.0001,
      "loss": 1.5638,
      "step": 5104
    },
    {
      "epoch": 0.8196852922286448,
      "grad_norm": 0.25343772768974304,
      "learning_rate": 0.0001,
      "loss": 1.4802,
      "step": 5105
    },
    {
      "epoch": 0.8198458574181118,
      "grad_norm": 0.29670315980911255,
      "learning_rate": 0.0001,
      "loss": 1.3951,
      "step": 5106
    },
    {
      "epoch": 0.8200064226075787,
      "grad_norm": 0.2548818290233612,
      "learning_rate": 0.0001,
      "loss": 1.4713,
      "step": 5107
    },
    {
      "epoch": 0.8201669877970456,
      "grad_norm": 0.26189446449279785,
      "learning_rate": 0.0001,
      "loss": 1.4966,
      "step": 5108
    },
    {
      "epoch": 0.8203275529865125,
      "grad_norm": 0.8882119059562683,
      "learning_rate": 0.0001,
      "loss": 1.4813,
      "step": 5109
    },
    {
      "epoch": 0.8204881181759794,
      "grad_norm": 0.29879793524742126,
      "learning_rate": 0.0001,
      "loss": 1.5289,
      "step": 5110
    },
    {
      "epoch": 0.8206486833654464,
      "grad_norm": 0.27714061737060547,
      "learning_rate": 0.0001,
      "loss": 1.5195,
      "step": 5111
    },
    {
      "epoch": 0.8208092485549133,
      "grad_norm": 0.2755475342273712,
      "learning_rate": 0.0001,
      "loss": 1.5512,
      "step": 5112
    },
    {
      "epoch": 0.8209698137443803,
      "grad_norm": 0.2726532518863678,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 5113
    },
    {
      "epoch": 0.8211303789338471,
      "grad_norm": 0.2651592791080475,
      "learning_rate": 0.0001,
      "loss": 1.4032,
      "step": 5114
    },
    {
      "epoch": 0.821290944123314,
      "grad_norm": 0.2701057195663452,
      "learning_rate": 0.0001,
      "loss": 1.4587,
      "step": 5115
    },
    {
      "epoch": 0.821451509312781,
      "grad_norm": 0.2695329487323761,
      "learning_rate": 0.0001,
      "loss": 1.4754,
      "step": 5116
    },
    {
      "epoch": 0.8216120745022479,
      "grad_norm": 0.2873031795024872,
      "learning_rate": 0.0001,
      "loss": 1.4787,
      "step": 5117
    },
    {
      "epoch": 0.8217726396917149,
      "grad_norm": 0.29290664196014404,
      "learning_rate": 0.0001,
      "loss": 1.4805,
      "step": 5118
    },
    {
      "epoch": 0.8219332048811817,
      "grad_norm": 0.270831823348999,
      "learning_rate": 0.0001,
      "loss": 1.3993,
      "step": 5119
    },
    {
      "epoch": 0.8220937700706487,
      "grad_norm": 0.2933433949947357,
      "learning_rate": 0.0001,
      "loss": 1.5612,
      "step": 5120
    },
    {
      "epoch": 0.8222543352601156,
      "grad_norm": 0.2748069763183594,
      "learning_rate": 0.0001,
      "loss": 1.5219,
      "step": 5121
    },
    {
      "epoch": 0.8224149004495825,
      "grad_norm": 0.2588675022125244,
      "learning_rate": 0.0001,
      "loss": 1.4801,
      "step": 5122
    },
    {
      "epoch": 0.8225754656390495,
      "grad_norm": 0.2799936830997467,
      "learning_rate": 0.0001,
      "loss": 1.438,
      "step": 5123
    },
    {
      "epoch": 0.8227360308285164,
      "grad_norm": 0.27927446365356445,
      "learning_rate": 0.0001,
      "loss": 1.4703,
      "step": 5124
    },
    {
      "epoch": 0.8228965960179833,
      "grad_norm": 0.27305909991264343,
      "learning_rate": 0.0001,
      "loss": 1.5344,
      "step": 5125
    },
    {
      "epoch": 0.8230571612074502,
      "grad_norm": 0.2809090316295624,
      "learning_rate": 0.0001,
      "loss": 1.4558,
      "step": 5126
    },
    {
      "epoch": 0.8232177263969171,
      "grad_norm": 0.29485854506492615,
      "learning_rate": 0.0001,
      "loss": 1.5036,
      "step": 5127
    },
    {
      "epoch": 0.8233782915863841,
      "grad_norm": 0.2724483013153076,
      "learning_rate": 0.0001,
      "loss": 1.462,
      "step": 5128
    },
    {
      "epoch": 0.823538856775851,
      "grad_norm": 0.26400211453437805,
      "learning_rate": 0.0001,
      "loss": 1.5174,
      "step": 5129
    },
    {
      "epoch": 0.8236994219653179,
      "grad_norm": 0.26416003704071045,
      "learning_rate": 0.0001,
      "loss": 1.4405,
      "step": 5130
    },
    {
      "epoch": 0.8238599871547848,
      "grad_norm": 0.2855091094970703,
      "learning_rate": 0.0001,
      "loss": 1.4165,
      "step": 5131
    },
    {
      "epoch": 0.8240205523442518,
      "grad_norm": 0.27835193276405334,
      "learning_rate": 0.0001,
      "loss": 1.4652,
      "step": 5132
    },
    {
      "epoch": 0.8241811175337187,
      "grad_norm": 0.258139044046402,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 5133
    },
    {
      "epoch": 0.8243416827231856,
      "grad_norm": 0.2563215494155884,
      "learning_rate": 0.0001,
      "loss": 1.4405,
      "step": 5134
    },
    {
      "epoch": 0.8245022479126526,
      "grad_norm": 0.2658300995826721,
      "learning_rate": 0.0001,
      "loss": 1.4993,
      "step": 5135
    },
    {
      "epoch": 0.8246628131021194,
      "grad_norm": 0.2655284106731415,
      "learning_rate": 0.0001,
      "loss": 1.4788,
      "step": 5136
    },
    {
      "epoch": 0.8248233782915864,
      "grad_norm": 0.27811136841773987,
      "learning_rate": 0.0001,
      "loss": 1.5086,
      "step": 5137
    },
    {
      "epoch": 0.8249839434810533,
      "grad_norm": 0.28676706552505493,
      "learning_rate": 0.0001,
      "loss": 1.4531,
      "step": 5138
    },
    {
      "epoch": 0.8251445086705202,
      "grad_norm": 0.2639085650444031,
      "learning_rate": 0.0001,
      "loss": 1.4882,
      "step": 5139
    },
    {
      "epoch": 0.8253050738599872,
      "grad_norm": 0.28511327505111694,
      "learning_rate": 0.0001,
      "loss": 1.4999,
      "step": 5140
    },
    {
      "epoch": 0.825465639049454,
      "grad_norm": 0.26077866554260254,
      "learning_rate": 0.0001,
      "loss": 1.4877,
      "step": 5141
    },
    {
      "epoch": 0.825626204238921,
      "grad_norm": 0.28479114174842834,
      "learning_rate": 0.0001,
      "loss": 1.596,
      "step": 5142
    },
    {
      "epoch": 0.8257867694283879,
      "grad_norm": 0.2742884159088135,
      "learning_rate": 0.0001,
      "loss": 1.4773,
      "step": 5143
    },
    {
      "epoch": 0.8259473346178549,
      "grad_norm": 0.26783105731010437,
      "learning_rate": 0.0001,
      "loss": 1.4657,
      "step": 5144
    },
    {
      "epoch": 0.8261078998073218,
      "grad_norm": 0.2745811939239502,
      "learning_rate": 0.0001,
      "loss": 1.5226,
      "step": 5145
    },
    {
      "epoch": 0.8262684649967887,
      "grad_norm": 0.2664661705493927,
      "learning_rate": 0.0001,
      "loss": 1.4311,
      "step": 5146
    },
    {
      "epoch": 0.8264290301862556,
      "grad_norm": 0.2732374370098114,
      "learning_rate": 0.0001,
      "loss": 1.3745,
      "step": 5147
    },
    {
      "epoch": 0.8265895953757225,
      "grad_norm": 0.2661466896533966,
      "learning_rate": 0.0001,
      "loss": 1.3327,
      "step": 5148
    },
    {
      "epoch": 0.8267501605651895,
      "grad_norm": 0.26395609974861145,
      "learning_rate": 0.0001,
      "loss": 1.474,
      "step": 5149
    },
    {
      "epoch": 0.8269107257546564,
      "grad_norm": 0.2627910375595093,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 5150
    },
    {
      "epoch": 0.8270712909441233,
      "grad_norm": 0.2805642783641815,
      "learning_rate": 0.0001,
      "loss": 1.5115,
      "step": 5151
    },
    {
      "epoch": 0.8272318561335903,
      "grad_norm": 0.26596057415008545,
      "learning_rate": 0.0001,
      "loss": 1.4544,
      "step": 5152
    },
    {
      "epoch": 0.8273924213230571,
      "grad_norm": 0.2646350860595703,
      "learning_rate": 0.0001,
      "loss": 1.519,
      "step": 5153
    },
    {
      "epoch": 0.8275529865125241,
      "grad_norm": 0.2681911289691925,
      "learning_rate": 0.0001,
      "loss": 1.5043,
      "step": 5154
    },
    {
      "epoch": 0.827713551701991,
      "grad_norm": 0.2620398998260498,
      "learning_rate": 0.0001,
      "loss": 1.5099,
      "step": 5155
    },
    {
      "epoch": 0.827874116891458,
      "grad_norm": 0.26036378741264343,
      "learning_rate": 0.0001,
      "loss": 1.4766,
      "step": 5156
    },
    {
      "epoch": 0.8280346820809249,
      "grad_norm": 0.2933289408683777,
      "learning_rate": 0.0001,
      "loss": 1.5055,
      "step": 5157
    },
    {
      "epoch": 0.8281952472703917,
      "grad_norm": 0.26431766152381897,
      "learning_rate": 0.0001,
      "loss": 1.4785,
      "step": 5158
    },
    {
      "epoch": 0.8283558124598587,
      "grad_norm": 0.27542543411254883,
      "learning_rate": 0.0001,
      "loss": 1.4591,
      "step": 5159
    },
    {
      "epoch": 0.8285163776493256,
      "grad_norm": 0.26881033182144165,
      "learning_rate": 0.0001,
      "loss": 1.4321,
      "step": 5160
    },
    {
      "epoch": 0.8286769428387926,
      "grad_norm": 0.27814653515815735,
      "learning_rate": 0.0001,
      "loss": 1.4129,
      "step": 5161
    },
    {
      "epoch": 0.8288375080282595,
      "grad_norm": 0.259012371301651,
      "learning_rate": 0.0001,
      "loss": 1.4539,
      "step": 5162
    },
    {
      "epoch": 0.8289980732177264,
      "grad_norm": 0.2711510956287384,
      "learning_rate": 0.0001,
      "loss": 1.4343,
      "step": 5163
    },
    {
      "epoch": 0.8291586384071933,
      "grad_norm": 0.27185365557670593,
      "learning_rate": 0.0001,
      "loss": 1.4643,
      "step": 5164
    },
    {
      "epoch": 0.8293192035966602,
      "grad_norm": 0.25620895624160767,
      "learning_rate": 0.0001,
      "loss": 1.4095,
      "step": 5165
    },
    {
      "epoch": 0.8294797687861272,
      "grad_norm": 0.2644282281398773,
      "learning_rate": 0.0001,
      "loss": 1.5839,
      "step": 5166
    },
    {
      "epoch": 0.8296403339755941,
      "grad_norm": 0.2591506838798523,
      "learning_rate": 0.0001,
      "loss": 1.4084,
      "step": 5167
    },
    {
      "epoch": 0.8298008991650611,
      "grad_norm": 0.27087241411209106,
      "learning_rate": 0.0001,
      "loss": 1.5111,
      "step": 5168
    },
    {
      "epoch": 0.8299614643545279,
      "grad_norm": 0.2518240213394165,
      "learning_rate": 0.0001,
      "loss": 1.4937,
      "step": 5169
    },
    {
      "epoch": 0.8301220295439948,
      "grad_norm": 0.260097473859787,
      "learning_rate": 0.0001,
      "loss": 1.452,
      "step": 5170
    },
    {
      "epoch": 0.8302825947334618,
      "grad_norm": 0.292864590883255,
      "learning_rate": 0.0001,
      "loss": 1.4754,
      "step": 5171
    },
    {
      "epoch": 0.8304431599229287,
      "grad_norm": 0.2785932123661041,
      "learning_rate": 0.0001,
      "loss": 1.4306,
      "step": 5172
    },
    {
      "epoch": 0.8306037251123957,
      "grad_norm": 0.2712307572364807,
      "learning_rate": 0.0001,
      "loss": 1.4387,
      "step": 5173
    },
    {
      "epoch": 0.8307642903018626,
      "grad_norm": 0.26541319489479065,
      "learning_rate": 0.0001,
      "loss": 1.47,
      "step": 5174
    },
    {
      "epoch": 0.8309248554913294,
      "grad_norm": 0.2870188355445862,
      "learning_rate": 0.0001,
      "loss": 1.4839,
      "step": 5175
    },
    {
      "epoch": 0.8310854206807964,
      "grad_norm": 0.2913323938846588,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 5176
    },
    {
      "epoch": 0.8312459858702633,
      "grad_norm": 0.2722484767436981,
      "learning_rate": 0.0001,
      "loss": 1.445,
      "step": 5177
    },
    {
      "epoch": 0.8314065510597303,
      "grad_norm": 0.2657018005847931,
      "learning_rate": 0.0001,
      "loss": 1.4855,
      "step": 5178
    },
    {
      "epoch": 0.8315671162491972,
      "grad_norm": 0.2667604684829712,
      "learning_rate": 0.0001,
      "loss": 1.5824,
      "step": 5179
    },
    {
      "epoch": 0.831727681438664,
      "grad_norm": 0.2654268741607666,
      "learning_rate": 0.0001,
      "loss": 1.5505,
      "step": 5180
    },
    {
      "epoch": 0.831888246628131,
      "grad_norm": 0.2547430992126465,
      "learning_rate": 0.0001,
      "loss": 1.4161,
      "step": 5181
    },
    {
      "epoch": 0.8320488118175979,
      "grad_norm": 0.2565046548843384,
      "learning_rate": 0.0001,
      "loss": 1.4418,
      "step": 5182
    },
    {
      "epoch": 0.8322093770070649,
      "grad_norm": 0.26058104634284973,
      "learning_rate": 0.0001,
      "loss": 1.545,
      "step": 5183
    },
    {
      "epoch": 0.8323699421965318,
      "grad_norm": 0.2802268862724304,
      "learning_rate": 0.0001,
      "loss": 1.399,
      "step": 5184
    },
    {
      "epoch": 0.8325305073859988,
      "grad_norm": 0.25977227091789246,
      "learning_rate": 0.0001,
      "loss": 1.5027,
      "step": 5185
    },
    {
      "epoch": 0.8326910725754656,
      "grad_norm": 0.2635708451271057,
      "learning_rate": 0.0001,
      "loss": 1.4115,
      "step": 5186
    },
    {
      "epoch": 0.8328516377649325,
      "grad_norm": 0.2661842405796051,
      "learning_rate": 0.0001,
      "loss": 1.4521,
      "step": 5187
    },
    {
      "epoch": 0.8330122029543995,
      "grad_norm": 0.266492635011673,
      "learning_rate": 0.0001,
      "loss": 1.4381,
      "step": 5188
    },
    {
      "epoch": 0.8331727681438664,
      "grad_norm": 0.2667500376701355,
      "learning_rate": 0.0001,
      "loss": 1.4577,
      "step": 5189
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.27786850929260254,
      "learning_rate": 0.0001,
      "loss": 1.5013,
      "step": 5190
    },
    {
      "epoch": 0.8334938985228002,
      "grad_norm": 0.28187161684036255,
      "learning_rate": 0.0001,
      "loss": 1.4082,
      "step": 5191
    },
    {
      "epoch": 0.8336544637122671,
      "grad_norm": 0.2588532865047455,
      "learning_rate": 0.0001,
      "loss": 1.4118,
      "step": 5192
    },
    {
      "epoch": 0.8338150289017341,
      "grad_norm": 0.2951449751853943,
      "learning_rate": 0.0001,
      "loss": 1.5414,
      "step": 5193
    },
    {
      "epoch": 0.833975594091201,
      "grad_norm": 0.2797166705131531,
      "learning_rate": 0.0001,
      "loss": 1.4807,
      "step": 5194
    },
    {
      "epoch": 0.834136159280668,
      "grad_norm": 0.2833901643753052,
      "learning_rate": 0.0001,
      "loss": 1.4815,
      "step": 5195
    },
    {
      "epoch": 0.8342967244701349,
      "grad_norm": 0.28295475244522095,
      "learning_rate": 0.0001,
      "loss": 1.4543,
      "step": 5196
    },
    {
      "epoch": 0.8344572896596018,
      "grad_norm": 0.268713116645813,
      "learning_rate": 0.0001,
      "loss": 1.4507,
      "step": 5197
    },
    {
      "epoch": 0.8346178548490687,
      "grad_norm": 0.2791979908943176,
      "learning_rate": 0.0001,
      "loss": 1.4525,
      "step": 5198
    },
    {
      "epoch": 0.8347784200385356,
      "grad_norm": 0.2836192846298218,
      "learning_rate": 0.0001,
      "loss": 1.5645,
      "step": 5199
    },
    {
      "epoch": 0.8349389852280026,
      "grad_norm": 0.2725848853588104,
      "learning_rate": 0.0001,
      "loss": 1.4823,
      "step": 5200
    },
    {
      "epoch": 0.8350995504174695,
      "grad_norm": 0.2682066559791565,
      "learning_rate": 0.0001,
      "loss": 1.5213,
      "step": 5201
    },
    {
      "epoch": 0.8352601156069365,
      "grad_norm": 0.2783905863761902,
      "learning_rate": 0.0001,
      "loss": 1.555,
      "step": 5202
    },
    {
      "epoch": 0.8354206807964033,
      "grad_norm": 0.26786595582962036,
      "learning_rate": 0.0001,
      "loss": 1.4646,
      "step": 5203
    },
    {
      "epoch": 0.8355812459858702,
      "grad_norm": 0.2640950083732605,
      "learning_rate": 0.0001,
      "loss": 1.4267,
      "step": 5204
    },
    {
      "epoch": 0.8357418111753372,
      "grad_norm": 0.27672556042671204,
      "learning_rate": 0.0001,
      "loss": 1.4804,
      "step": 5205
    },
    {
      "epoch": 0.8359023763648041,
      "grad_norm": 0.27699586749076843,
      "learning_rate": 0.0001,
      "loss": 1.4971,
      "step": 5206
    },
    {
      "epoch": 0.8360629415542711,
      "grad_norm": 0.26214927434921265,
      "learning_rate": 0.0001,
      "loss": 1.4213,
      "step": 5207
    },
    {
      "epoch": 0.8362235067437379,
      "grad_norm": 0.28681671619415283,
      "learning_rate": 0.0001,
      "loss": 1.4586,
      "step": 5208
    },
    {
      "epoch": 0.8363840719332049,
      "grad_norm": 0.2791830599308014,
      "learning_rate": 0.0001,
      "loss": 1.477,
      "step": 5209
    },
    {
      "epoch": 0.8365446371226718,
      "grad_norm": 0.2742231786251068,
      "learning_rate": 0.0001,
      "loss": 1.5096,
      "step": 5210
    },
    {
      "epoch": 0.8367052023121387,
      "grad_norm": 0.2677064836025238,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 5211
    },
    {
      "epoch": 0.8368657675016057,
      "grad_norm": 0.28335240483283997,
      "learning_rate": 0.0001,
      "loss": 1.5283,
      "step": 5212
    },
    {
      "epoch": 0.8370263326910726,
      "grad_norm": 0.2656855583190918,
      "learning_rate": 0.0001,
      "loss": 1.4812,
      "step": 5213
    },
    {
      "epoch": 0.8371868978805395,
      "grad_norm": 0.281790167093277,
      "learning_rate": 0.0001,
      "loss": 1.5169,
      "step": 5214
    },
    {
      "epoch": 0.8373474630700064,
      "grad_norm": 0.277771919965744,
      "learning_rate": 0.0001,
      "loss": 1.471,
      "step": 5215
    },
    {
      "epoch": 0.8375080282594733,
      "grad_norm": 0.29074445366859436,
      "learning_rate": 0.0001,
      "loss": 1.5112,
      "step": 5216
    },
    {
      "epoch": 0.8376685934489403,
      "grad_norm": 0.2728709876537323,
      "learning_rate": 0.0001,
      "loss": 1.4522,
      "step": 5217
    },
    {
      "epoch": 0.8378291586384072,
      "grad_norm": 0.2697131037712097,
      "learning_rate": 0.0001,
      "loss": 1.5388,
      "step": 5218
    },
    {
      "epoch": 0.8379897238278741,
      "grad_norm": 0.26410043239593506,
      "learning_rate": 0.0001,
      "loss": 1.4333,
      "step": 5219
    },
    {
      "epoch": 0.838150289017341,
      "grad_norm": 0.27733278274536133,
      "learning_rate": 0.0001,
      "loss": 1.5169,
      "step": 5220
    },
    {
      "epoch": 0.838310854206808,
      "grad_norm": 0.26385006308555603,
      "learning_rate": 0.0001,
      "loss": 1.4538,
      "step": 5221
    },
    {
      "epoch": 0.8384714193962749,
      "grad_norm": 0.2629423439502716,
      "learning_rate": 0.0001,
      "loss": 1.5075,
      "step": 5222
    },
    {
      "epoch": 0.8386319845857418,
      "grad_norm": 0.27457037568092346,
      "learning_rate": 0.0001,
      "loss": 1.4637,
      "step": 5223
    },
    {
      "epoch": 0.8387925497752088,
      "grad_norm": 0.2441031038761139,
      "learning_rate": 0.0001,
      "loss": 1.4459,
      "step": 5224
    },
    {
      "epoch": 0.8389531149646756,
      "grad_norm": 0.2537875473499298,
      "learning_rate": 0.0001,
      "loss": 1.4015,
      "step": 5225
    },
    {
      "epoch": 0.8391136801541426,
      "grad_norm": 0.26629117131233215,
      "learning_rate": 0.0001,
      "loss": 1.4603,
      "step": 5226
    },
    {
      "epoch": 0.8392742453436095,
      "grad_norm": 0.26998355984687805,
      "learning_rate": 0.0001,
      "loss": 1.4563,
      "step": 5227
    },
    {
      "epoch": 0.8394348105330764,
      "grad_norm": 0.2816150188446045,
      "learning_rate": 0.0001,
      "loss": 1.6008,
      "step": 5228
    },
    {
      "epoch": 0.8395953757225434,
      "grad_norm": 0.2685852348804474,
      "learning_rate": 0.0001,
      "loss": 1.5131,
      "step": 5229
    },
    {
      "epoch": 0.8397559409120102,
      "grad_norm": 0.27722010016441345,
      "learning_rate": 0.0001,
      "loss": 1.5362,
      "step": 5230
    },
    {
      "epoch": 0.8399165061014772,
      "grad_norm": 0.2766026258468628,
      "learning_rate": 0.0001,
      "loss": 1.5552,
      "step": 5231
    },
    {
      "epoch": 0.8400770712909441,
      "grad_norm": 0.2711658477783203,
      "learning_rate": 0.0001,
      "loss": 1.5026,
      "step": 5232
    },
    {
      "epoch": 0.840237636480411,
      "grad_norm": 0.27401483058929443,
      "learning_rate": 0.0001,
      "loss": 1.4661,
      "step": 5233
    },
    {
      "epoch": 0.840398201669878,
      "grad_norm": 0.2823639512062073,
      "learning_rate": 0.0001,
      "loss": 1.475,
      "step": 5234
    },
    {
      "epoch": 0.8405587668593449,
      "grad_norm": 0.256363183259964,
      "learning_rate": 0.0001,
      "loss": 1.485,
      "step": 5235
    },
    {
      "epoch": 0.8407193320488118,
      "grad_norm": 0.2676762640476227,
      "learning_rate": 0.0001,
      "loss": 1.4965,
      "step": 5236
    },
    {
      "epoch": 0.8408798972382787,
      "grad_norm": 0.25752776861190796,
      "learning_rate": 0.0001,
      "loss": 1.4888,
      "step": 5237
    },
    {
      "epoch": 0.8410404624277457,
      "grad_norm": 0.2785666882991791,
      "learning_rate": 0.0001,
      "loss": 1.4931,
      "step": 5238
    },
    {
      "epoch": 0.8412010276172126,
      "grad_norm": 0.303597092628479,
      "learning_rate": 0.0001,
      "loss": 1.5779,
      "step": 5239
    },
    {
      "epoch": 0.8413615928066795,
      "grad_norm": 0.2613723874092102,
      "learning_rate": 0.0001,
      "loss": 1.4444,
      "step": 5240
    },
    {
      "epoch": 0.8415221579961464,
      "grad_norm": 0.25317010283470154,
      "learning_rate": 0.0001,
      "loss": 1.4211,
      "step": 5241
    },
    {
      "epoch": 0.8416827231856133,
      "grad_norm": 0.2762047052383423,
      "learning_rate": 0.0001,
      "loss": 1.5074,
      "step": 5242
    },
    {
      "epoch": 0.8418432883750803,
      "grad_norm": 0.38027575612068176,
      "learning_rate": 0.0001,
      "loss": 1.4986,
      "step": 5243
    },
    {
      "epoch": 0.8420038535645472,
      "grad_norm": 0.28426018357276917,
      "learning_rate": 0.0001,
      "loss": 1.5045,
      "step": 5244
    },
    {
      "epoch": 0.8421644187540142,
      "grad_norm": 0.26807907223701477,
      "learning_rate": 0.0001,
      "loss": 1.508,
      "step": 5245
    },
    {
      "epoch": 0.8423249839434811,
      "grad_norm": 0.26778993010520935,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 5246
    },
    {
      "epoch": 0.8424855491329479,
      "grad_norm": 0.26153409481048584,
      "learning_rate": 0.0001,
      "loss": 1.51,
      "step": 5247
    },
    {
      "epoch": 0.8426461143224149,
      "grad_norm": 0.26222196221351624,
      "learning_rate": 0.0001,
      "loss": 1.5181,
      "step": 5248
    },
    {
      "epoch": 0.8428066795118818,
      "grad_norm": 0.26973864436149597,
      "learning_rate": 0.0001,
      "loss": 1.4557,
      "step": 5249
    },
    {
      "epoch": 0.8429672447013488,
      "grad_norm": 0.27253398299217224,
      "learning_rate": 0.0001,
      "loss": 1.4572,
      "step": 5250
    },
    {
      "epoch": 0.8431278098908157,
      "grad_norm": 0.26086390018463135,
      "learning_rate": 0.0001,
      "loss": 1.4947,
      "step": 5251
    },
    {
      "epoch": 0.8432883750802826,
      "grad_norm": 0.25629332661628723,
      "learning_rate": 0.0001,
      "loss": 1.4801,
      "step": 5252
    },
    {
      "epoch": 0.8434489402697495,
      "grad_norm": 0.27789345383644104,
      "learning_rate": 0.0001,
      "loss": 1.5344,
      "step": 5253
    },
    {
      "epoch": 0.8436095054592164,
      "grad_norm": 0.25398340821266174,
      "learning_rate": 0.0001,
      "loss": 1.5206,
      "step": 5254
    },
    {
      "epoch": 0.8437700706486834,
      "grad_norm": 0.2797142565250397,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 5255
    },
    {
      "epoch": 0.8439306358381503,
      "grad_norm": 0.2748796343803406,
      "learning_rate": 0.0001,
      "loss": 1.5764,
      "step": 5256
    },
    {
      "epoch": 0.8440912010276173,
      "grad_norm": 0.255555123090744,
      "learning_rate": 0.0001,
      "loss": 1.4444,
      "step": 5257
    },
    {
      "epoch": 0.8442517662170841,
      "grad_norm": 0.27899670600891113,
      "learning_rate": 0.0001,
      "loss": 1.4988,
      "step": 5258
    },
    {
      "epoch": 0.844412331406551,
      "grad_norm": 0.2732127904891968,
      "learning_rate": 0.0001,
      "loss": 1.4811,
      "step": 5259
    },
    {
      "epoch": 0.844572896596018,
      "grad_norm": 0.2523353099822998,
      "learning_rate": 0.0001,
      "loss": 1.4327,
      "step": 5260
    },
    {
      "epoch": 0.8447334617854849,
      "grad_norm": 0.26799678802490234,
      "learning_rate": 0.0001,
      "loss": 1.419,
      "step": 5261
    },
    {
      "epoch": 0.8448940269749519,
      "grad_norm": 0.26532357931137085,
      "learning_rate": 0.0001,
      "loss": 1.4337,
      "step": 5262
    },
    {
      "epoch": 0.8450545921644188,
      "grad_norm": 0.2633708119392395,
      "learning_rate": 0.0001,
      "loss": 1.559,
      "step": 5263
    },
    {
      "epoch": 0.8452151573538856,
      "grad_norm": 0.27879685163497925,
      "learning_rate": 0.0001,
      "loss": 1.4835,
      "step": 5264
    },
    {
      "epoch": 0.8453757225433526,
      "grad_norm": 0.27665314078330994,
      "learning_rate": 0.0001,
      "loss": 1.4895,
      "step": 5265
    },
    {
      "epoch": 0.8455362877328195,
      "grad_norm": 0.2681369483470917,
      "learning_rate": 0.0001,
      "loss": 1.5189,
      "step": 5266
    },
    {
      "epoch": 0.8456968529222865,
      "grad_norm": 0.27890172600746155,
      "learning_rate": 0.0001,
      "loss": 1.4846,
      "step": 5267
    },
    {
      "epoch": 0.8458574181117534,
      "grad_norm": 0.2770402729511261,
      "learning_rate": 0.0001,
      "loss": 1.4708,
      "step": 5268
    },
    {
      "epoch": 0.8460179833012202,
      "grad_norm": 0.28067901730537415,
      "learning_rate": 0.0001,
      "loss": 1.493,
      "step": 5269
    },
    {
      "epoch": 0.8461785484906872,
      "grad_norm": 0.2720443308353424,
      "learning_rate": 0.0001,
      "loss": 1.465,
      "step": 5270
    },
    {
      "epoch": 0.8463391136801541,
      "grad_norm": 0.2602487802505493,
      "learning_rate": 0.0001,
      "loss": 1.4784,
      "step": 5271
    },
    {
      "epoch": 0.8464996788696211,
      "grad_norm": 0.2525050938129425,
      "learning_rate": 0.0001,
      "loss": 1.383,
      "step": 5272
    },
    {
      "epoch": 0.846660244059088,
      "grad_norm": 0.28399643301963806,
      "learning_rate": 0.0001,
      "loss": 1.5532,
      "step": 5273
    },
    {
      "epoch": 0.846820809248555,
      "grad_norm": 0.27307626605033875,
      "learning_rate": 0.0001,
      "loss": 1.5325,
      "step": 5274
    },
    {
      "epoch": 0.8469813744380218,
      "grad_norm": 0.26894059777259827,
      "learning_rate": 0.0001,
      "loss": 1.4124,
      "step": 5275
    },
    {
      "epoch": 0.8471419396274887,
      "grad_norm": 0.2619190514087677,
      "learning_rate": 0.0001,
      "loss": 1.5041,
      "step": 5276
    },
    {
      "epoch": 0.8473025048169557,
      "grad_norm": 0.25521716475486755,
      "learning_rate": 0.0001,
      "loss": 1.492,
      "step": 5277
    },
    {
      "epoch": 0.8474630700064226,
      "grad_norm": 0.25996091961860657,
      "learning_rate": 0.0001,
      "loss": 1.4228,
      "step": 5278
    },
    {
      "epoch": 0.8476236351958896,
      "grad_norm": 0.2658037841320038,
      "learning_rate": 0.0001,
      "loss": 1.5182,
      "step": 5279
    },
    {
      "epoch": 0.8477842003853564,
      "grad_norm": 0.2619093060493469,
      "learning_rate": 0.0001,
      "loss": 1.4209,
      "step": 5280
    },
    {
      "epoch": 0.8479447655748233,
      "grad_norm": 0.2617737054824829,
      "learning_rate": 0.0001,
      "loss": 1.4443,
      "step": 5281
    },
    {
      "epoch": 0.8481053307642903,
      "grad_norm": 0.2692483365535736,
      "learning_rate": 0.0001,
      "loss": 1.5543,
      "step": 5282
    },
    {
      "epoch": 0.8482658959537572,
      "grad_norm": 0.2643255293369293,
      "learning_rate": 0.0001,
      "loss": 1.4965,
      "step": 5283
    },
    {
      "epoch": 0.8484264611432242,
      "grad_norm": 0.2760925889015198,
      "learning_rate": 0.0001,
      "loss": 1.4917,
      "step": 5284
    },
    {
      "epoch": 0.8485870263326911,
      "grad_norm": 0.26432475447654724,
      "learning_rate": 0.0001,
      "loss": 1.5258,
      "step": 5285
    },
    {
      "epoch": 0.848747591522158,
      "grad_norm": 0.2962948679924011,
      "learning_rate": 0.0001,
      "loss": 1.5083,
      "step": 5286
    },
    {
      "epoch": 0.8489081567116249,
      "grad_norm": 0.2693207859992981,
      "learning_rate": 0.0001,
      "loss": 1.4983,
      "step": 5287
    },
    {
      "epoch": 0.8490687219010918,
      "grad_norm": 0.27385634183883667,
      "learning_rate": 0.0001,
      "loss": 1.4977,
      "step": 5288
    },
    {
      "epoch": 0.8492292870905588,
      "grad_norm": 0.2681661546230316,
      "learning_rate": 0.0001,
      "loss": 1.4905,
      "step": 5289
    },
    {
      "epoch": 0.8493898522800257,
      "grad_norm": 0.26129087805747986,
      "learning_rate": 0.0001,
      "loss": 1.4372,
      "step": 5290
    },
    {
      "epoch": 0.8495504174694927,
      "grad_norm": 0.2743181586265564,
      "learning_rate": 0.0001,
      "loss": 1.4226,
      "step": 5291
    },
    {
      "epoch": 0.8497109826589595,
      "grad_norm": 0.2775701880455017,
      "learning_rate": 0.0001,
      "loss": 1.5435,
      "step": 5292
    },
    {
      "epoch": 0.8498715478484264,
      "grad_norm": 0.2661895751953125,
      "learning_rate": 0.0001,
      "loss": 1.522,
      "step": 5293
    },
    {
      "epoch": 0.8500321130378934,
      "grad_norm": 0.27059099078178406,
      "learning_rate": 0.0001,
      "loss": 1.5029,
      "step": 5294
    },
    {
      "epoch": 0.8501926782273603,
      "grad_norm": 0.2673824727535248,
      "learning_rate": 0.0001,
      "loss": 1.4741,
      "step": 5295
    },
    {
      "epoch": 0.8503532434168273,
      "grad_norm": 0.25712454319000244,
      "learning_rate": 0.0001,
      "loss": 1.4721,
      "step": 5296
    },
    {
      "epoch": 0.8505138086062941,
      "grad_norm": 0.28328362107276917,
      "learning_rate": 0.0001,
      "loss": 1.4103,
      "step": 5297
    },
    {
      "epoch": 0.850674373795761,
      "grad_norm": 0.26156899333000183,
      "learning_rate": 0.0001,
      "loss": 1.5011,
      "step": 5298
    },
    {
      "epoch": 0.850834938985228,
      "grad_norm": 0.27011993527412415,
      "learning_rate": 0.0001,
      "loss": 1.4274,
      "step": 5299
    },
    {
      "epoch": 0.8509955041746949,
      "grad_norm": 0.26966166496276855,
      "learning_rate": 0.0001,
      "loss": 1.5654,
      "step": 5300
    },
    {
      "epoch": 0.8511560693641619,
      "grad_norm": 0.25330469012260437,
      "learning_rate": 0.0001,
      "loss": 1.4416,
      "step": 5301
    },
    {
      "epoch": 0.8513166345536288,
      "grad_norm": 0.2578624188899994,
      "learning_rate": 0.0001,
      "loss": 1.4339,
      "step": 5302
    },
    {
      "epoch": 0.8514771997430957,
      "grad_norm": 0.268004834651947,
      "learning_rate": 0.0001,
      "loss": 1.5447,
      "step": 5303
    },
    {
      "epoch": 0.8516377649325626,
      "grad_norm": 0.2628016173839569,
      "learning_rate": 0.0001,
      "loss": 1.4418,
      "step": 5304
    },
    {
      "epoch": 0.8517983301220295,
      "grad_norm": 0.26189881563186646,
      "learning_rate": 0.0001,
      "loss": 1.5414,
      "step": 5305
    },
    {
      "epoch": 0.8519588953114965,
      "grad_norm": 0.26972198486328125,
      "learning_rate": 0.0001,
      "loss": 1.5235,
      "step": 5306
    },
    {
      "epoch": 0.8521194605009634,
      "grad_norm": 0.27024611830711365,
      "learning_rate": 0.0001,
      "loss": 1.4779,
      "step": 5307
    },
    {
      "epoch": 0.8522800256904303,
      "grad_norm": 0.2719064950942993,
      "learning_rate": 0.0001,
      "loss": 1.5013,
      "step": 5308
    },
    {
      "epoch": 0.8524405908798972,
      "grad_norm": 0.27928489446640015,
      "learning_rate": 0.0001,
      "loss": 1.4162,
      "step": 5309
    },
    {
      "epoch": 0.8526011560693642,
      "grad_norm": 0.2659405767917633,
      "learning_rate": 0.0001,
      "loss": 1.4961,
      "step": 5310
    },
    {
      "epoch": 0.8527617212588311,
      "grad_norm": 0.28146472573280334,
      "learning_rate": 0.0001,
      "loss": 1.4631,
      "step": 5311
    },
    {
      "epoch": 0.852922286448298,
      "grad_norm": 0.2521761655807495,
      "learning_rate": 0.0001,
      "loss": 1.4895,
      "step": 5312
    },
    {
      "epoch": 0.853082851637765,
      "grad_norm": 0.26773086190223694,
      "learning_rate": 0.0001,
      "loss": 1.4861,
      "step": 5313
    },
    {
      "epoch": 0.8532434168272318,
      "grad_norm": 0.28108856081962585,
      "learning_rate": 0.0001,
      "loss": 1.4696,
      "step": 5314
    },
    {
      "epoch": 0.8534039820166988,
      "grad_norm": 0.277148574590683,
      "learning_rate": 0.0001,
      "loss": 1.4788,
      "step": 5315
    },
    {
      "epoch": 0.8535645472061657,
      "grad_norm": 0.2859897017478943,
      "learning_rate": 0.0001,
      "loss": 1.5683,
      "step": 5316
    },
    {
      "epoch": 0.8537251123956326,
      "grad_norm": 0.27081987261772156,
      "learning_rate": 0.0001,
      "loss": 1.4866,
      "step": 5317
    },
    {
      "epoch": 0.8538856775850996,
      "grad_norm": 0.2703607678413391,
      "learning_rate": 0.0001,
      "loss": 1.5545,
      "step": 5318
    },
    {
      "epoch": 0.8540462427745664,
      "grad_norm": 0.29761916399002075,
      "learning_rate": 0.0001,
      "loss": 1.4591,
      "step": 5319
    },
    {
      "epoch": 0.8542068079640334,
      "grad_norm": 0.27993783354759216,
      "learning_rate": 0.0001,
      "loss": 1.5114,
      "step": 5320
    },
    {
      "epoch": 0.8543673731535003,
      "grad_norm": 0.29130029678344727,
      "learning_rate": 0.0001,
      "loss": 1.4765,
      "step": 5321
    },
    {
      "epoch": 0.8545279383429673,
      "grad_norm": 0.2789912223815918,
      "learning_rate": 0.0001,
      "loss": 1.4779,
      "step": 5322
    },
    {
      "epoch": 0.8546885035324342,
      "grad_norm": 0.26253095269203186,
      "learning_rate": 0.0001,
      "loss": 1.4685,
      "step": 5323
    },
    {
      "epoch": 0.8548490687219011,
      "grad_norm": 0.26710742712020874,
      "learning_rate": 0.0001,
      "loss": 1.4584,
      "step": 5324
    },
    {
      "epoch": 0.855009633911368,
      "grad_norm": 0.25740182399749756,
      "learning_rate": 0.0001,
      "loss": 1.4335,
      "step": 5325
    },
    {
      "epoch": 0.8551701991008349,
      "grad_norm": 0.2696593403816223,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 5326
    },
    {
      "epoch": 0.8553307642903019,
      "grad_norm": 0.2718164920806885,
      "learning_rate": 0.0001,
      "loss": 1.5228,
      "step": 5327
    },
    {
      "epoch": 0.8554913294797688,
      "grad_norm": 0.2829359173774719,
      "learning_rate": 0.0001,
      "loss": 1.5784,
      "step": 5328
    },
    {
      "epoch": 0.8556518946692357,
      "grad_norm": 0.2718700170516968,
      "learning_rate": 0.0001,
      "loss": 1.5205,
      "step": 5329
    },
    {
      "epoch": 0.8558124598587026,
      "grad_norm": 0.2533983886241913,
      "learning_rate": 0.0001,
      "loss": 1.468,
      "step": 5330
    },
    {
      "epoch": 0.8559730250481695,
      "grad_norm": 0.25820839405059814,
      "learning_rate": 0.0001,
      "loss": 1.4253,
      "step": 5331
    },
    {
      "epoch": 0.8561335902376365,
      "grad_norm": 0.2800899147987366,
      "learning_rate": 0.0001,
      "loss": 1.5077,
      "step": 5332
    },
    {
      "epoch": 0.8562941554271034,
      "grad_norm": 0.26723790168762207,
      "learning_rate": 0.0001,
      "loss": 1.4986,
      "step": 5333
    },
    {
      "epoch": 0.8564547206165704,
      "grad_norm": 0.34934258460998535,
      "learning_rate": 0.0001,
      "loss": 1.4808,
      "step": 5334
    },
    {
      "epoch": 0.8566152858060373,
      "grad_norm": 0.26037484407424927,
      "learning_rate": 0.0001,
      "loss": 1.4963,
      "step": 5335
    },
    {
      "epoch": 0.8567758509955041,
      "grad_norm": 0.26629379391670227,
      "learning_rate": 0.0001,
      "loss": 1.4993,
      "step": 5336
    },
    {
      "epoch": 0.8569364161849711,
      "grad_norm": 0.28581124544143677,
      "learning_rate": 0.0001,
      "loss": 1.4839,
      "step": 5337
    },
    {
      "epoch": 0.857096981374438,
      "grad_norm": 0.26524630188941956,
      "learning_rate": 0.0001,
      "loss": 1.4362,
      "step": 5338
    },
    {
      "epoch": 0.857257546563905,
      "grad_norm": 0.2664186954498291,
      "learning_rate": 0.0001,
      "loss": 1.5323,
      "step": 5339
    },
    {
      "epoch": 0.8574181117533719,
      "grad_norm": 0.26289859414100647,
      "learning_rate": 0.0001,
      "loss": 1.529,
      "step": 5340
    },
    {
      "epoch": 0.8575786769428388,
      "grad_norm": 0.2632129490375519,
      "learning_rate": 0.0001,
      "loss": 1.5224,
      "step": 5341
    },
    {
      "epoch": 0.8577392421323057,
      "grad_norm": 0.26697301864624023,
      "learning_rate": 0.0001,
      "loss": 1.4813,
      "step": 5342
    },
    {
      "epoch": 0.8578998073217726,
      "grad_norm": 0.2633684575557709,
      "learning_rate": 0.0001,
      "loss": 1.5293,
      "step": 5343
    },
    {
      "epoch": 0.8580603725112396,
      "grad_norm": 0.26969778537750244,
      "learning_rate": 0.0001,
      "loss": 1.5256,
      "step": 5344
    },
    {
      "epoch": 0.8582209377007065,
      "grad_norm": 0.28679952025413513,
      "learning_rate": 0.0001,
      "loss": 1.4575,
      "step": 5345
    },
    {
      "epoch": 0.8583815028901735,
      "grad_norm": 0.2709237039089203,
      "learning_rate": 0.0001,
      "loss": 1.4327,
      "step": 5346
    },
    {
      "epoch": 0.8585420680796403,
      "grad_norm": 0.2711273431777954,
      "learning_rate": 0.0001,
      "loss": 1.4518,
      "step": 5347
    },
    {
      "epoch": 0.8587026332691072,
      "grad_norm": 0.2717254161834717,
      "learning_rate": 0.0001,
      "loss": 1.5657,
      "step": 5348
    },
    {
      "epoch": 0.8588631984585742,
      "grad_norm": 0.2722921073436737,
      "learning_rate": 0.0001,
      "loss": 1.4986,
      "step": 5349
    },
    {
      "epoch": 0.8590237636480411,
      "grad_norm": 0.27039068937301636,
      "learning_rate": 0.0001,
      "loss": 1.5133,
      "step": 5350
    },
    {
      "epoch": 0.8591843288375081,
      "grad_norm": 0.2701614201068878,
      "learning_rate": 0.0001,
      "loss": 1.5577,
      "step": 5351
    },
    {
      "epoch": 0.859344894026975,
      "grad_norm": 0.3063983619213104,
      "learning_rate": 0.0001,
      "loss": 1.4986,
      "step": 5352
    },
    {
      "epoch": 0.8595054592164418,
      "grad_norm": 0.2711946666240692,
      "learning_rate": 0.0001,
      "loss": 1.4582,
      "step": 5353
    },
    {
      "epoch": 0.8596660244059088,
      "grad_norm": 0.2693926990032196,
      "learning_rate": 0.0001,
      "loss": 1.477,
      "step": 5354
    },
    {
      "epoch": 0.8598265895953757,
      "grad_norm": 0.2748611867427826,
      "learning_rate": 0.0001,
      "loss": 1.5143,
      "step": 5355
    },
    {
      "epoch": 0.8599871547848427,
      "grad_norm": 0.2937244474887848,
      "learning_rate": 0.0001,
      "loss": 1.4983,
      "step": 5356
    },
    {
      "epoch": 0.8601477199743096,
      "grad_norm": 0.26753324270248413,
      "learning_rate": 0.0001,
      "loss": 1.5532,
      "step": 5357
    },
    {
      "epoch": 0.8603082851637764,
      "grad_norm": 0.2588178813457489,
      "learning_rate": 0.0001,
      "loss": 1.4297,
      "step": 5358
    },
    {
      "epoch": 0.8604688503532434,
      "grad_norm": 0.27366507053375244,
      "learning_rate": 0.0001,
      "loss": 1.438,
      "step": 5359
    },
    {
      "epoch": 0.8606294155427103,
      "grad_norm": 0.26366257667541504,
      "learning_rate": 0.0001,
      "loss": 1.437,
      "step": 5360
    },
    {
      "epoch": 0.8607899807321773,
      "grad_norm": 0.2775055468082428,
      "learning_rate": 0.0001,
      "loss": 1.4391,
      "step": 5361
    },
    {
      "epoch": 0.8609505459216442,
      "grad_norm": 0.27710622549057007,
      "learning_rate": 0.0001,
      "loss": 1.5381,
      "step": 5362
    },
    {
      "epoch": 0.8611111111111112,
      "grad_norm": 0.258648544549942,
      "learning_rate": 0.0001,
      "loss": 1.4236,
      "step": 5363
    },
    {
      "epoch": 0.861271676300578,
      "grad_norm": 0.2793954312801361,
      "learning_rate": 0.0001,
      "loss": 1.5147,
      "step": 5364
    },
    {
      "epoch": 0.8614322414900449,
      "grad_norm": 0.2777060568332672,
      "learning_rate": 0.0001,
      "loss": 1.4437,
      "step": 5365
    },
    {
      "epoch": 0.8615928066795119,
      "grad_norm": 0.28187084197998047,
      "learning_rate": 0.0001,
      "loss": 1.5089,
      "step": 5366
    },
    {
      "epoch": 0.8617533718689788,
      "grad_norm": 0.26416015625,
      "learning_rate": 0.0001,
      "loss": 1.4458,
      "step": 5367
    },
    {
      "epoch": 0.8619139370584458,
      "grad_norm": 0.2784457802772522,
      "learning_rate": 0.0001,
      "loss": 1.5126,
      "step": 5368
    },
    {
      "epoch": 0.8620745022479126,
      "grad_norm": 0.2705570161342621,
      "learning_rate": 0.0001,
      "loss": 1.4383,
      "step": 5369
    },
    {
      "epoch": 0.8622350674373795,
      "grad_norm": 0.2800073027610779,
      "learning_rate": 0.0001,
      "loss": 1.5053,
      "step": 5370
    },
    {
      "epoch": 0.8623956326268465,
      "grad_norm": 0.30694282054901123,
      "learning_rate": 0.0001,
      "loss": 1.5594,
      "step": 5371
    },
    {
      "epoch": 0.8625561978163134,
      "grad_norm": 0.25870153307914734,
      "learning_rate": 0.0001,
      "loss": 1.424,
      "step": 5372
    },
    {
      "epoch": 0.8627167630057804,
      "grad_norm": 0.29331105947494507,
      "learning_rate": 0.0001,
      "loss": 1.5723,
      "step": 5373
    },
    {
      "epoch": 0.8628773281952473,
      "grad_norm": 0.3343738615512848,
      "learning_rate": 0.0001,
      "loss": 1.5026,
      "step": 5374
    },
    {
      "epoch": 0.8630378933847141,
      "grad_norm": 0.2702012360095978,
      "learning_rate": 0.0001,
      "loss": 1.5063,
      "step": 5375
    },
    {
      "epoch": 0.8631984585741811,
      "grad_norm": 0.28902357816696167,
      "learning_rate": 0.0001,
      "loss": 1.5046,
      "step": 5376
    },
    {
      "epoch": 0.863359023763648,
      "grad_norm": 0.2942296266555786,
      "learning_rate": 0.0001,
      "loss": 1.5046,
      "step": 5377
    },
    {
      "epoch": 0.863519588953115,
      "grad_norm": 0.28400853276252747,
      "learning_rate": 0.0001,
      "loss": 1.51,
      "step": 5378
    },
    {
      "epoch": 0.8636801541425819,
      "grad_norm": 0.2927720546722412,
      "learning_rate": 0.0001,
      "loss": 1.4818,
      "step": 5379
    },
    {
      "epoch": 0.8638407193320488,
      "grad_norm": 0.26653483510017395,
      "learning_rate": 0.0001,
      "loss": 1.4673,
      "step": 5380
    },
    {
      "epoch": 0.8640012845215157,
      "grad_norm": 0.2643788456916809,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 5381
    },
    {
      "epoch": 0.8641618497109826,
      "grad_norm": 0.26809826493263245,
      "learning_rate": 0.0001,
      "loss": 1.4062,
      "step": 5382
    },
    {
      "epoch": 0.8643224149004496,
      "grad_norm": 0.26230669021606445,
      "learning_rate": 0.0001,
      "loss": 1.5291,
      "step": 5383
    },
    {
      "epoch": 0.8644829800899165,
      "grad_norm": 0.2660991847515106,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 5384
    },
    {
      "epoch": 0.8646435452793835,
      "grad_norm": 0.27956071496009827,
      "learning_rate": 0.0001,
      "loss": 1.4282,
      "step": 5385
    },
    {
      "epoch": 0.8648041104688503,
      "grad_norm": 0.2823096513748169,
      "learning_rate": 0.0001,
      "loss": 1.543,
      "step": 5386
    },
    {
      "epoch": 0.8649646756583173,
      "grad_norm": 0.2795092463493347,
      "learning_rate": 0.0001,
      "loss": 1.4906,
      "step": 5387
    },
    {
      "epoch": 0.8651252408477842,
      "grad_norm": 0.26812344789505005,
      "learning_rate": 0.0001,
      "loss": 1.5251,
      "step": 5388
    },
    {
      "epoch": 0.8652858060372511,
      "grad_norm": 0.2653723359107971,
      "learning_rate": 0.0001,
      "loss": 1.4801,
      "step": 5389
    },
    {
      "epoch": 0.8654463712267181,
      "grad_norm": 0.272517591714859,
      "learning_rate": 0.0001,
      "loss": 1.4916,
      "step": 5390
    },
    {
      "epoch": 0.865606936416185,
      "grad_norm": 0.28307634592056274,
      "learning_rate": 0.0001,
      "loss": 1.4491,
      "step": 5391
    },
    {
      "epoch": 0.8657675016056519,
      "grad_norm": 0.26914727687835693,
      "learning_rate": 0.0001,
      "loss": 1.4699,
      "step": 5392
    },
    {
      "epoch": 0.8659280667951188,
      "grad_norm": 0.29564759135246277,
      "learning_rate": 0.0001,
      "loss": 1.4507,
      "step": 5393
    },
    {
      "epoch": 0.8660886319845857,
      "grad_norm": 0.2670200765132904,
      "learning_rate": 0.0001,
      "loss": 1.4429,
      "step": 5394
    },
    {
      "epoch": 0.8662491971740527,
      "grad_norm": 0.2828080654144287,
      "learning_rate": 0.0001,
      "loss": 1.4994,
      "step": 5395
    },
    {
      "epoch": 0.8664097623635196,
      "grad_norm": 0.27270907163619995,
      "learning_rate": 0.0001,
      "loss": 1.5513,
      "step": 5396
    },
    {
      "epoch": 0.8665703275529865,
      "grad_norm": 0.2735515534877777,
      "learning_rate": 0.0001,
      "loss": 1.4291,
      "step": 5397
    },
    {
      "epoch": 0.8667308927424534,
      "grad_norm": 0.2732189893722534,
      "learning_rate": 0.0001,
      "loss": 1.4891,
      "step": 5398
    },
    {
      "epoch": 0.8668914579319204,
      "grad_norm": 0.26425090432167053,
      "learning_rate": 0.0001,
      "loss": 1.5195,
      "step": 5399
    },
    {
      "epoch": 0.8670520231213873,
      "grad_norm": 0.2832198441028595,
      "learning_rate": 0.0001,
      "loss": 1.4958,
      "step": 5400
    },
    {
      "epoch": 0.8672125883108542,
      "grad_norm": 0.2647020220756531,
      "learning_rate": 0.0001,
      "loss": 1.5135,
      "step": 5401
    },
    {
      "epoch": 0.8673731535003212,
      "grad_norm": 0.26475292444229126,
      "learning_rate": 0.0001,
      "loss": 1.4592,
      "step": 5402
    },
    {
      "epoch": 0.867533718689788,
      "grad_norm": 0.2587742209434509,
      "learning_rate": 0.0001,
      "loss": 1.5061,
      "step": 5403
    },
    {
      "epoch": 0.867694283879255,
      "grad_norm": 0.2691299617290497,
      "learning_rate": 0.0001,
      "loss": 1.4102,
      "step": 5404
    },
    {
      "epoch": 0.8678548490687219,
      "grad_norm": 0.26208651065826416,
      "learning_rate": 0.0001,
      "loss": 1.4109,
      "step": 5405
    },
    {
      "epoch": 0.8680154142581888,
      "grad_norm": 0.2596539556980133,
      "learning_rate": 0.0001,
      "loss": 1.4853,
      "step": 5406
    },
    {
      "epoch": 0.8681759794476558,
      "grad_norm": 0.27161863446235657,
      "learning_rate": 0.0001,
      "loss": 1.4881,
      "step": 5407
    },
    {
      "epoch": 0.8683365446371226,
      "grad_norm": 0.2674539089202881,
      "learning_rate": 0.0001,
      "loss": 1.4426,
      "step": 5408
    },
    {
      "epoch": 0.8684971098265896,
      "grad_norm": 0.26671645045280457,
      "learning_rate": 0.0001,
      "loss": 1.4391,
      "step": 5409
    },
    {
      "epoch": 0.8686576750160565,
      "grad_norm": 0.3180280923843384,
      "learning_rate": 0.0001,
      "loss": 1.5427,
      "step": 5410
    },
    {
      "epoch": 0.8688182402055235,
      "grad_norm": 0.2686658799648285,
      "learning_rate": 0.0001,
      "loss": 1.4433,
      "step": 5411
    },
    {
      "epoch": 0.8689788053949904,
      "grad_norm": 0.2700963020324707,
      "learning_rate": 0.0001,
      "loss": 1.4939,
      "step": 5412
    },
    {
      "epoch": 0.8691393705844573,
      "grad_norm": 0.2653445601463318,
      "learning_rate": 0.0001,
      "loss": 1.524,
      "step": 5413
    },
    {
      "epoch": 0.8692999357739242,
      "grad_norm": 0.25825613737106323,
      "learning_rate": 0.0001,
      "loss": 1.4503,
      "step": 5414
    },
    {
      "epoch": 0.8694605009633911,
      "grad_norm": 0.2721810042858124,
      "learning_rate": 0.0001,
      "loss": 1.3788,
      "step": 5415
    },
    {
      "epoch": 0.8696210661528581,
      "grad_norm": 0.26655474305152893,
      "learning_rate": 0.0001,
      "loss": 1.5133,
      "step": 5416
    },
    {
      "epoch": 0.869781631342325,
      "grad_norm": 0.2792733907699585,
      "learning_rate": 0.0001,
      "loss": 1.4712,
      "step": 5417
    },
    {
      "epoch": 0.869942196531792,
      "grad_norm": 0.2694145143032074,
      "learning_rate": 0.0001,
      "loss": 1.4402,
      "step": 5418
    },
    {
      "epoch": 0.8701027617212588,
      "grad_norm": 0.28874000906944275,
      "learning_rate": 0.0001,
      "loss": 1.4661,
      "step": 5419
    },
    {
      "epoch": 0.8702633269107257,
      "grad_norm": 0.2771264314651489,
      "learning_rate": 0.0001,
      "loss": 1.5002,
      "step": 5420
    },
    {
      "epoch": 0.8704238921001927,
      "grad_norm": 0.2902165353298187,
      "learning_rate": 0.0001,
      "loss": 1.4846,
      "step": 5421
    },
    {
      "epoch": 0.8705844572896596,
      "grad_norm": 0.28441616892814636,
      "learning_rate": 0.0001,
      "loss": 1.4281,
      "step": 5422
    },
    {
      "epoch": 0.8707450224791266,
      "grad_norm": 0.4823160767555237,
      "learning_rate": 0.0001,
      "loss": 1.4394,
      "step": 5423
    },
    {
      "epoch": 0.8709055876685935,
      "grad_norm": 0.269285649061203,
      "learning_rate": 0.0001,
      "loss": 1.474,
      "step": 5424
    },
    {
      "epoch": 0.8710661528580603,
      "grad_norm": 0.26998424530029297,
      "learning_rate": 0.0001,
      "loss": 1.3978,
      "step": 5425
    },
    {
      "epoch": 0.8712267180475273,
      "grad_norm": 0.28708550333976746,
      "learning_rate": 0.0001,
      "loss": 1.4277,
      "step": 5426
    },
    {
      "epoch": 0.8713872832369942,
      "grad_norm": 0.2990915775299072,
      "learning_rate": 0.0001,
      "loss": 1.4551,
      "step": 5427
    },
    {
      "epoch": 0.8715478484264612,
      "grad_norm": 0.32246556878089905,
      "learning_rate": 0.0001,
      "loss": 1.4494,
      "step": 5428
    },
    {
      "epoch": 0.8717084136159281,
      "grad_norm": 0.26776349544525146,
      "learning_rate": 0.0001,
      "loss": 1.4335,
      "step": 5429
    },
    {
      "epoch": 0.871868978805395,
      "grad_norm": 0.2630974054336548,
      "learning_rate": 0.0001,
      "loss": 1.3887,
      "step": 5430
    },
    {
      "epoch": 0.8720295439948619,
      "grad_norm": 0.29397687315940857,
      "learning_rate": 0.0001,
      "loss": 1.5027,
      "step": 5431
    },
    {
      "epoch": 0.8721901091843288,
      "grad_norm": 0.264883428812027,
      "learning_rate": 0.0001,
      "loss": 1.5002,
      "step": 5432
    },
    {
      "epoch": 0.8723506743737958,
      "grad_norm": 0.2784595489501953,
      "learning_rate": 0.0001,
      "loss": 1.4356,
      "step": 5433
    },
    {
      "epoch": 0.8725112395632627,
      "grad_norm": 0.2815721929073334,
      "learning_rate": 0.0001,
      "loss": 1.4899,
      "step": 5434
    },
    {
      "epoch": 0.8726718047527297,
      "grad_norm": 0.27264702320098877,
      "learning_rate": 0.0001,
      "loss": 1.495,
      "step": 5435
    },
    {
      "epoch": 0.8728323699421965,
      "grad_norm": 0.38980332016944885,
      "learning_rate": 0.0001,
      "loss": 1.3819,
      "step": 5436
    },
    {
      "epoch": 0.8729929351316634,
      "grad_norm": 0.2716798186302185,
      "learning_rate": 0.0001,
      "loss": 1.3688,
      "step": 5437
    },
    {
      "epoch": 0.8731535003211304,
      "grad_norm": 0.2638702094554901,
      "learning_rate": 0.0001,
      "loss": 1.4922,
      "step": 5438
    },
    {
      "epoch": 0.8733140655105973,
      "grad_norm": 0.27662479877471924,
      "learning_rate": 0.0001,
      "loss": 1.4759,
      "step": 5439
    },
    {
      "epoch": 0.8734746307000643,
      "grad_norm": 0.27404460310935974,
      "learning_rate": 0.0001,
      "loss": 1.4806,
      "step": 5440
    },
    {
      "epoch": 0.8736351958895312,
      "grad_norm": 0.29231128096580505,
      "learning_rate": 0.0001,
      "loss": 1.5648,
      "step": 5441
    },
    {
      "epoch": 0.873795761078998,
      "grad_norm": 0.26786231994628906,
      "learning_rate": 0.0001,
      "loss": 1.4983,
      "step": 5442
    },
    {
      "epoch": 0.873956326268465,
      "grad_norm": 0.2640931010246277,
      "learning_rate": 0.0001,
      "loss": 1.4557,
      "step": 5443
    },
    {
      "epoch": 0.8741168914579319,
      "grad_norm": 0.27399834990501404,
      "learning_rate": 0.0001,
      "loss": 1.4503,
      "step": 5444
    },
    {
      "epoch": 0.8742774566473989,
      "grad_norm": 0.28965431451797485,
      "learning_rate": 0.0001,
      "loss": 1.5404,
      "step": 5445
    },
    {
      "epoch": 0.8744380218368658,
      "grad_norm": 0.34253400564193726,
      "learning_rate": 0.0001,
      "loss": 1.445,
      "step": 5446
    },
    {
      "epoch": 0.8745985870263326,
      "grad_norm": 0.2716148793697357,
      "learning_rate": 0.0001,
      "loss": 1.4432,
      "step": 5447
    },
    {
      "epoch": 0.8747591522157996,
      "grad_norm": 0.2725996673107147,
      "learning_rate": 0.0001,
      "loss": 1.4203,
      "step": 5448
    },
    {
      "epoch": 0.8749197174052665,
      "grad_norm": 0.26596102118492126,
      "learning_rate": 0.0001,
      "loss": 1.4237,
      "step": 5449
    },
    {
      "epoch": 0.8750802825947335,
      "grad_norm": 0.279045432806015,
      "learning_rate": 0.0001,
      "loss": 1.4495,
      "step": 5450
    },
    {
      "epoch": 0.8752408477842004,
      "grad_norm": 0.25917115807533264,
      "learning_rate": 0.0001,
      "loss": 1.4089,
      "step": 5451
    },
    {
      "epoch": 0.8754014129736674,
      "grad_norm": 0.25699543952941895,
      "learning_rate": 0.0001,
      "loss": 1.438,
      "step": 5452
    },
    {
      "epoch": 0.8755619781631342,
      "grad_norm": 0.28414681553840637,
      "learning_rate": 0.0001,
      "loss": 1.4592,
      "step": 5453
    },
    {
      "epoch": 0.8757225433526011,
      "grad_norm": 0.2951044738292694,
      "learning_rate": 0.0001,
      "loss": 1.4188,
      "step": 5454
    },
    {
      "epoch": 0.8758831085420681,
      "grad_norm": 0.3004177510738373,
      "learning_rate": 0.0001,
      "loss": 1.4602,
      "step": 5455
    },
    {
      "epoch": 0.876043673731535,
      "grad_norm": 0.274126261472702,
      "learning_rate": 0.0001,
      "loss": 1.5185,
      "step": 5456
    },
    {
      "epoch": 0.876204238921002,
      "grad_norm": 0.2812749743461609,
      "learning_rate": 0.0001,
      "loss": 1.4698,
      "step": 5457
    },
    {
      "epoch": 0.8763648041104688,
      "grad_norm": 0.26154085993766785,
      "learning_rate": 0.0001,
      "loss": 1.3933,
      "step": 5458
    },
    {
      "epoch": 0.8765253692999357,
      "grad_norm": 0.2749265432357788,
      "learning_rate": 0.0001,
      "loss": 1.4817,
      "step": 5459
    },
    {
      "epoch": 0.8766859344894027,
      "grad_norm": 0.2600361406803131,
      "learning_rate": 0.0001,
      "loss": 1.405,
      "step": 5460
    },
    {
      "epoch": 0.8768464996788696,
      "grad_norm": 0.29384899139404297,
      "learning_rate": 0.0001,
      "loss": 1.5376,
      "step": 5461
    },
    {
      "epoch": 0.8770070648683366,
      "grad_norm": 0.2780305743217468,
      "learning_rate": 0.0001,
      "loss": 1.4352,
      "step": 5462
    },
    {
      "epoch": 0.8771676300578035,
      "grad_norm": 0.2738387882709503,
      "learning_rate": 0.0001,
      "loss": 1.5633,
      "step": 5463
    },
    {
      "epoch": 0.8773281952472703,
      "grad_norm": 0.27933788299560547,
      "learning_rate": 0.0001,
      "loss": 1.54,
      "step": 5464
    },
    {
      "epoch": 0.8774887604367373,
      "grad_norm": 0.26793208718299866,
      "learning_rate": 0.0001,
      "loss": 1.5069,
      "step": 5465
    },
    {
      "epoch": 0.8776493256262042,
      "grad_norm": 0.2673221528530121,
      "learning_rate": 0.0001,
      "loss": 1.4992,
      "step": 5466
    },
    {
      "epoch": 0.8778098908156712,
      "grad_norm": 0.25794437527656555,
      "learning_rate": 0.0001,
      "loss": 1.3249,
      "step": 5467
    },
    {
      "epoch": 0.8779704560051381,
      "grad_norm": 0.2692946493625641,
      "learning_rate": 0.0001,
      "loss": 1.4745,
      "step": 5468
    },
    {
      "epoch": 0.878131021194605,
      "grad_norm": 0.27269116044044495,
      "learning_rate": 0.0001,
      "loss": 1.4185,
      "step": 5469
    },
    {
      "epoch": 0.8782915863840719,
      "grad_norm": 0.27077004313468933,
      "learning_rate": 0.0001,
      "loss": 1.4149,
      "step": 5470
    },
    {
      "epoch": 0.8784521515735388,
      "grad_norm": 0.27809661626815796,
      "learning_rate": 0.0001,
      "loss": 1.5599,
      "step": 5471
    },
    {
      "epoch": 0.8786127167630058,
      "grad_norm": 0.2573157250881195,
      "learning_rate": 0.0001,
      "loss": 1.4652,
      "step": 5472
    },
    {
      "epoch": 0.8787732819524727,
      "grad_norm": 0.27147430181503296,
      "learning_rate": 0.0001,
      "loss": 1.468,
      "step": 5473
    },
    {
      "epoch": 0.8789338471419397,
      "grad_norm": 0.27140012383461,
      "learning_rate": 0.0001,
      "loss": 1.4872,
      "step": 5474
    },
    {
      "epoch": 0.8790944123314065,
      "grad_norm": 0.2830095589160919,
      "learning_rate": 0.0001,
      "loss": 1.5283,
      "step": 5475
    },
    {
      "epoch": 0.8792549775208734,
      "grad_norm": 0.2714841663837433,
      "learning_rate": 0.0001,
      "loss": 1.4509,
      "step": 5476
    },
    {
      "epoch": 0.8794155427103404,
      "grad_norm": 0.27208662033081055,
      "learning_rate": 0.0001,
      "loss": 1.4991,
      "step": 5477
    },
    {
      "epoch": 0.8795761078998073,
      "grad_norm": 0.268839031457901,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 5478
    },
    {
      "epoch": 0.8797366730892743,
      "grad_norm": 0.2627919018268585,
      "learning_rate": 0.0001,
      "loss": 1.4628,
      "step": 5479
    },
    {
      "epoch": 0.8798972382787412,
      "grad_norm": 0.2893525958061218,
      "learning_rate": 0.0001,
      "loss": 1.5424,
      "step": 5480
    },
    {
      "epoch": 0.880057803468208,
      "grad_norm": 0.27234259247779846,
      "learning_rate": 0.0001,
      "loss": 1.3436,
      "step": 5481
    },
    {
      "epoch": 0.880218368657675,
      "grad_norm": 0.265053927898407,
      "learning_rate": 0.0001,
      "loss": 1.4102,
      "step": 5482
    },
    {
      "epoch": 0.8803789338471419,
      "grad_norm": 0.27132099866867065,
      "learning_rate": 0.0001,
      "loss": 1.4566,
      "step": 5483
    },
    {
      "epoch": 0.8805394990366089,
      "grad_norm": 0.26990145444869995,
      "learning_rate": 0.0001,
      "loss": 1.4016,
      "step": 5484
    },
    {
      "epoch": 0.8807000642260758,
      "grad_norm": 0.26688551902770996,
      "learning_rate": 0.0001,
      "loss": 1.4271,
      "step": 5485
    },
    {
      "epoch": 0.8808606294155427,
      "grad_norm": 0.2790434956550598,
      "learning_rate": 0.0001,
      "loss": 1.492,
      "step": 5486
    },
    {
      "epoch": 0.8810211946050096,
      "grad_norm": 0.2871028482913971,
      "learning_rate": 0.0001,
      "loss": 1.4772,
      "step": 5487
    },
    {
      "epoch": 0.8811817597944765,
      "grad_norm": 0.2595863342285156,
      "learning_rate": 0.0001,
      "loss": 1.4739,
      "step": 5488
    },
    {
      "epoch": 0.8813423249839435,
      "grad_norm": 0.28704044222831726,
      "learning_rate": 0.0001,
      "loss": 1.5201,
      "step": 5489
    },
    {
      "epoch": 0.8815028901734104,
      "grad_norm": 0.2778429090976715,
      "learning_rate": 0.0001,
      "loss": 1.4335,
      "step": 5490
    },
    {
      "epoch": 0.8816634553628774,
      "grad_norm": 0.270887553691864,
      "learning_rate": 0.0001,
      "loss": 1.4402,
      "step": 5491
    },
    {
      "epoch": 0.8818240205523442,
      "grad_norm": 0.26834598183631897,
      "learning_rate": 0.0001,
      "loss": 1.4492,
      "step": 5492
    },
    {
      "epoch": 0.8819845857418112,
      "grad_norm": 0.26983681321144104,
      "learning_rate": 0.0001,
      "loss": 1.5221,
      "step": 5493
    },
    {
      "epoch": 0.8821451509312781,
      "grad_norm": 0.26623982191085815,
      "learning_rate": 0.0001,
      "loss": 1.5069,
      "step": 5494
    },
    {
      "epoch": 0.882305716120745,
      "grad_norm": 0.2848629951477051,
      "learning_rate": 0.0001,
      "loss": 1.4249,
      "step": 5495
    },
    {
      "epoch": 0.882466281310212,
      "grad_norm": 0.26681122183799744,
      "learning_rate": 0.0001,
      "loss": 1.4899,
      "step": 5496
    },
    {
      "epoch": 0.8826268464996788,
      "grad_norm": 0.26040562987327576,
      "learning_rate": 0.0001,
      "loss": 1.5291,
      "step": 5497
    },
    {
      "epoch": 0.8827874116891458,
      "grad_norm": 1.1018840074539185,
      "learning_rate": 0.0001,
      "loss": 1.4965,
      "step": 5498
    },
    {
      "epoch": 0.8829479768786127,
      "grad_norm": 0.26510193943977356,
      "learning_rate": 0.0001,
      "loss": 1.5338,
      "step": 5499
    },
    {
      "epoch": 0.8831085420680796,
      "grad_norm": 0.2770121693611145,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 5500
    },
    {
      "epoch": 0.8832691072575466,
      "grad_norm": 0.2851237952709198,
      "learning_rate": 0.0001,
      "loss": 1.397,
      "step": 5501
    },
    {
      "epoch": 0.8834296724470135,
      "grad_norm": 0.2876242399215698,
      "learning_rate": 0.0001,
      "loss": 1.5258,
      "step": 5502
    },
    {
      "epoch": 0.8835902376364804,
      "grad_norm": 0.28714463114738464,
      "learning_rate": 0.0001,
      "loss": 1.4906,
      "step": 5503
    },
    {
      "epoch": 0.8837508028259473,
      "grad_norm": 0.2737811207771301,
      "learning_rate": 0.0001,
      "loss": 1.4187,
      "step": 5504
    },
    {
      "epoch": 0.8839113680154143,
      "grad_norm": 0.2581880986690521,
      "learning_rate": 0.0001,
      "loss": 1.4158,
      "step": 5505
    },
    {
      "epoch": 0.8840719332048812,
      "grad_norm": 0.27452948689460754,
      "learning_rate": 0.0001,
      "loss": 1.5494,
      "step": 5506
    },
    {
      "epoch": 0.8842324983943481,
      "grad_norm": 0.25586310029029846,
      "learning_rate": 0.0001,
      "loss": 1.5023,
      "step": 5507
    },
    {
      "epoch": 0.884393063583815,
      "grad_norm": 0.2672421634197235,
      "learning_rate": 0.0001,
      "loss": 1.4421,
      "step": 5508
    },
    {
      "epoch": 0.8845536287732819,
      "grad_norm": 0.2715543508529663,
      "learning_rate": 0.0001,
      "loss": 1.5156,
      "step": 5509
    },
    {
      "epoch": 0.8847141939627489,
      "grad_norm": 0.27912983298301697,
      "learning_rate": 0.0001,
      "loss": 1.52,
      "step": 5510
    },
    {
      "epoch": 0.8848747591522158,
      "grad_norm": 0.27395251393318176,
      "learning_rate": 0.0001,
      "loss": 1.5105,
      "step": 5511
    },
    {
      "epoch": 0.8850353243416827,
      "grad_norm": 0.28986644744873047,
      "learning_rate": 0.0001,
      "loss": 1.4493,
      "step": 5512
    },
    {
      "epoch": 0.8851958895311497,
      "grad_norm": 0.277619332075119,
      "learning_rate": 0.0001,
      "loss": 1.4742,
      "step": 5513
    },
    {
      "epoch": 0.8853564547206165,
      "grad_norm": 0.2849483788013458,
      "learning_rate": 0.0001,
      "loss": 1.4366,
      "step": 5514
    },
    {
      "epoch": 0.8855170199100835,
      "grad_norm": 0.26593446731567383,
      "learning_rate": 0.0001,
      "loss": 1.4509,
      "step": 5515
    },
    {
      "epoch": 0.8856775850995504,
      "grad_norm": 0.2862519323825836,
      "learning_rate": 0.0001,
      "loss": 1.5016,
      "step": 5516
    },
    {
      "epoch": 0.8858381502890174,
      "grad_norm": 0.2797463834285736,
      "learning_rate": 0.0001,
      "loss": 1.5768,
      "step": 5517
    },
    {
      "epoch": 0.8859987154784843,
      "grad_norm": 0.2575477659702301,
      "learning_rate": 0.0001,
      "loss": 1.437,
      "step": 5518
    },
    {
      "epoch": 0.8861592806679512,
      "grad_norm": 0.3024967610836029,
      "learning_rate": 0.0001,
      "loss": 1.4959,
      "step": 5519
    },
    {
      "epoch": 0.8863198458574181,
      "grad_norm": 0.28635838627815247,
      "learning_rate": 0.0001,
      "loss": 1.5708,
      "step": 5520
    },
    {
      "epoch": 0.886480411046885,
      "grad_norm": 0.2505148947238922,
      "learning_rate": 0.0001,
      "loss": 1.4379,
      "step": 5521
    },
    {
      "epoch": 0.886640976236352,
      "grad_norm": 0.2753680348396301,
      "learning_rate": 0.0001,
      "loss": 1.5203,
      "step": 5522
    },
    {
      "epoch": 0.8868015414258189,
      "grad_norm": 0.29059112071990967,
      "learning_rate": 0.0001,
      "loss": 1.4797,
      "step": 5523
    },
    {
      "epoch": 0.8869621066152859,
      "grad_norm": 0.2863348722457886,
      "learning_rate": 0.0001,
      "loss": 1.5352,
      "step": 5524
    },
    {
      "epoch": 0.8871226718047527,
      "grad_norm": 0.2873809039592743,
      "learning_rate": 0.0001,
      "loss": 1.5321,
      "step": 5525
    },
    {
      "epoch": 0.8872832369942196,
      "grad_norm": 0.2848975658416748,
      "learning_rate": 0.0001,
      "loss": 1.4525,
      "step": 5526
    },
    {
      "epoch": 0.8874438021836866,
      "grad_norm": 0.2567618489265442,
      "learning_rate": 0.0001,
      "loss": 1.4443,
      "step": 5527
    },
    {
      "epoch": 0.8876043673731535,
      "grad_norm": 0.26232290267944336,
      "learning_rate": 0.0001,
      "loss": 1.4997,
      "step": 5528
    },
    {
      "epoch": 0.8877649325626205,
      "grad_norm": 0.5042279958724976,
      "learning_rate": 0.0001,
      "loss": 1.4704,
      "step": 5529
    },
    {
      "epoch": 0.8879254977520874,
      "grad_norm": 0.27147457003593445,
      "learning_rate": 0.0001,
      "loss": 1.4107,
      "step": 5530
    },
    {
      "epoch": 0.8880860629415542,
      "grad_norm": 0.25734785199165344,
      "learning_rate": 0.0001,
      "loss": 1.4891,
      "step": 5531
    },
    {
      "epoch": 0.8882466281310212,
      "grad_norm": 0.2572897672653198,
      "learning_rate": 0.0001,
      "loss": 1.3871,
      "step": 5532
    },
    {
      "epoch": 0.8884071933204881,
      "grad_norm": 0.27331671118736267,
      "learning_rate": 0.0001,
      "loss": 1.5044,
      "step": 5533
    },
    {
      "epoch": 0.8885677585099551,
      "grad_norm": 0.2718226909637451,
      "learning_rate": 0.0001,
      "loss": 1.4272,
      "step": 5534
    },
    {
      "epoch": 0.888728323699422,
      "grad_norm": 0.29373684525489807,
      "learning_rate": 0.0001,
      "loss": 1.4484,
      "step": 5535
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.29399439692497253,
      "learning_rate": 0.0001,
      "loss": 1.4547,
      "step": 5536
    },
    {
      "epoch": 0.8890494540783558,
      "grad_norm": 0.27119654417037964,
      "learning_rate": 0.0001,
      "loss": 1.4411,
      "step": 5537
    },
    {
      "epoch": 0.8892100192678227,
      "grad_norm": 0.2952917218208313,
      "learning_rate": 0.0001,
      "loss": 1.4182,
      "step": 5538
    },
    {
      "epoch": 0.8893705844572897,
      "grad_norm": 0.28254881501197815,
      "learning_rate": 0.0001,
      "loss": 1.4993,
      "step": 5539
    },
    {
      "epoch": 0.8895311496467566,
      "grad_norm": 0.27757877111434937,
      "learning_rate": 0.0001,
      "loss": 1.5142,
      "step": 5540
    },
    {
      "epoch": 0.8896917148362236,
      "grad_norm": 0.27630168199539185,
      "learning_rate": 0.0001,
      "loss": 1.4694,
      "step": 5541
    },
    {
      "epoch": 0.8898522800256904,
      "grad_norm": 0.2867511808872223,
      "learning_rate": 0.0001,
      "loss": 1.3884,
      "step": 5542
    },
    {
      "epoch": 0.8900128452151573,
      "grad_norm": 0.27782851457595825,
      "learning_rate": 0.0001,
      "loss": 1.4658,
      "step": 5543
    },
    {
      "epoch": 0.8901734104046243,
      "grad_norm": 0.3377024233341217,
      "learning_rate": 0.0001,
      "loss": 1.4425,
      "step": 5544
    },
    {
      "epoch": 0.8903339755940912,
      "grad_norm": 0.27609983086586,
      "learning_rate": 0.0001,
      "loss": 1.3812,
      "step": 5545
    },
    {
      "epoch": 0.8904945407835582,
      "grad_norm": 0.29063501954078674,
      "learning_rate": 0.0001,
      "loss": 1.5037,
      "step": 5546
    },
    {
      "epoch": 0.890655105973025,
      "grad_norm": 0.2954513132572174,
      "learning_rate": 0.0001,
      "loss": 1.5111,
      "step": 5547
    },
    {
      "epoch": 0.8908156711624919,
      "grad_norm": 0.2803086042404175,
      "learning_rate": 0.0001,
      "loss": 1.4915,
      "step": 5548
    },
    {
      "epoch": 0.8909762363519589,
      "grad_norm": 0.2909174859523773,
      "learning_rate": 0.0001,
      "loss": 1.4672,
      "step": 5549
    },
    {
      "epoch": 0.8911368015414258,
      "grad_norm": 0.28933873772621155,
      "learning_rate": 0.0001,
      "loss": 1.518,
      "step": 5550
    },
    {
      "epoch": 0.8912973667308928,
      "grad_norm": 0.27435481548309326,
      "learning_rate": 0.0001,
      "loss": 1.4816,
      "step": 5551
    },
    {
      "epoch": 0.8914579319203597,
      "grad_norm": 0.2842663824558258,
      "learning_rate": 0.0001,
      "loss": 1.4744,
      "step": 5552
    },
    {
      "epoch": 0.8916184971098265,
      "grad_norm": 0.27131035923957825,
      "learning_rate": 0.0001,
      "loss": 1.4683,
      "step": 5553
    },
    {
      "epoch": 0.8917790622992935,
      "grad_norm": 0.2606917917728424,
      "learning_rate": 0.0001,
      "loss": 1.3991,
      "step": 5554
    },
    {
      "epoch": 0.8919396274887604,
      "grad_norm": 0.2685220241546631,
      "learning_rate": 0.0001,
      "loss": 1.4473,
      "step": 5555
    },
    {
      "epoch": 0.8921001926782274,
      "grad_norm": 0.2765880227088928,
      "learning_rate": 0.0001,
      "loss": 1.4593,
      "step": 5556
    },
    {
      "epoch": 0.8922607578676943,
      "grad_norm": 0.26068785786628723,
      "learning_rate": 0.0001,
      "loss": 1.4555,
      "step": 5557
    },
    {
      "epoch": 0.8924213230571612,
      "grad_norm": 0.2859265208244324,
      "learning_rate": 0.0001,
      "loss": 1.5097,
      "step": 5558
    },
    {
      "epoch": 0.8925818882466281,
      "grad_norm": 0.2766442894935608,
      "learning_rate": 0.0001,
      "loss": 1.465,
      "step": 5559
    },
    {
      "epoch": 0.892742453436095,
      "grad_norm": 0.27757006883621216,
      "learning_rate": 0.0001,
      "loss": 1.502,
      "step": 5560
    },
    {
      "epoch": 0.892903018625562,
      "grad_norm": 0.2708435654640198,
      "learning_rate": 0.0001,
      "loss": 1.5407,
      "step": 5561
    },
    {
      "epoch": 0.8930635838150289,
      "grad_norm": 0.27640071511268616,
      "learning_rate": 0.0001,
      "loss": 1.4641,
      "step": 5562
    },
    {
      "epoch": 0.8932241490044959,
      "grad_norm": 0.26688772439956665,
      "learning_rate": 0.0001,
      "loss": 1.4338,
      "step": 5563
    },
    {
      "epoch": 0.8933847141939627,
      "grad_norm": 0.2749989628791809,
      "learning_rate": 0.0001,
      "loss": 1.4659,
      "step": 5564
    },
    {
      "epoch": 0.8935452793834296,
      "grad_norm": 0.29130905866622925,
      "learning_rate": 0.0001,
      "loss": 1.4989,
      "step": 5565
    },
    {
      "epoch": 0.8937058445728966,
      "grad_norm": 0.28483328223228455,
      "learning_rate": 0.0001,
      "loss": 1.4407,
      "step": 5566
    },
    {
      "epoch": 0.8938664097623635,
      "grad_norm": 0.42114609479904175,
      "learning_rate": 0.0001,
      "loss": 1.5139,
      "step": 5567
    },
    {
      "epoch": 0.8940269749518305,
      "grad_norm": 0.2821149230003357,
      "learning_rate": 0.0001,
      "loss": 1.4913,
      "step": 5568
    },
    {
      "epoch": 0.8941875401412974,
      "grad_norm": 0.2953449487686157,
      "learning_rate": 0.0001,
      "loss": 1.4786,
      "step": 5569
    },
    {
      "epoch": 0.8943481053307643,
      "grad_norm": 0.26782578229904175,
      "learning_rate": 0.0001,
      "loss": 1.4649,
      "step": 5570
    },
    {
      "epoch": 0.8945086705202312,
      "grad_norm": 0.27817678451538086,
      "learning_rate": 0.0001,
      "loss": 1.52,
      "step": 5571
    },
    {
      "epoch": 0.8946692357096981,
      "grad_norm": 0.27889594435691833,
      "learning_rate": 0.0001,
      "loss": 1.4755,
      "step": 5572
    },
    {
      "epoch": 0.8948298008991651,
      "grad_norm": 0.25449198484420776,
      "learning_rate": 0.0001,
      "loss": 1.3736,
      "step": 5573
    },
    {
      "epoch": 0.894990366088632,
      "grad_norm": 0.27022528648376465,
      "learning_rate": 0.0001,
      "loss": 1.4576,
      "step": 5574
    },
    {
      "epoch": 0.8951509312780989,
      "grad_norm": 0.2697735130786896,
      "learning_rate": 0.0001,
      "loss": 1.4779,
      "step": 5575
    },
    {
      "epoch": 0.8953114964675658,
      "grad_norm": 0.28412875533103943,
      "learning_rate": 0.0001,
      "loss": 1.5155,
      "step": 5576
    },
    {
      "epoch": 0.8954720616570327,
      "grad_norm": 0.26869145035743713,
      "learning_rate": 0.0001,
      "loss": 1.4827,
      "step": 5577
    },
    {
      "epoch": 0.8956326268464997,
      "grad_norm": 0.26857876777648926,
      "learning_rate": 0.0001,
      "loss": 1.4641,
      "step": 5578
    },
    {
      "epoch": 0.8957931920359666,
      "grad_norm": 0.2718310058116913,
      "learning_rate": 0.0001,
      "loss": 1.5202,
      "step": 5579
    },
    {
      "epoch": 0.8959537572254336,
      "grad_norm": 0.27814191579818726,
      "learning_rate": 0.0001,
      "loss": 1.5414,
      "step": 5580
    },
    {
      "epoch": 0.8961143224149004,
      "grad_norm": 0.2594946324825287,
      "learning_rate": 0.0001,
      "loss": 1.3946,
      "step": 5581
    },
    {
      "epoch": 0.8962748876043674,
      "grad_norm": 0.2645629942417145,
      "learning_rate": 0.0001,
      "loss": 1.4806,
      "step": 5582
    },
    {
      "epoch": 0.8964354527938343,
      "grad_norm": 0.2640096843242645,
      "learning_rate": 0.0001,
      "loss": 1.5146,
      "step": 5583
    },
    {
      "epoch": 0.8965960179833012,
      "grad_norm": 0.2750597894191742,
      "learning_rate": 0.0001,
      "loss": 1.523,
      "step": 5584
    },
    {
      "epoch": 0.8967565831727682,
      "grad_norm": 0.26160359382629395,
      "learning_rate": 0.0001,
      "loss": 1.3128,
      "step": 5585
    },
    {
      "epoch": 0.896917148362235,
      "grad_norm": 0.2674008905887604,
      "learning_rate": 0.0001,
      "loss": 1.4988,
      "step": 5586
    },
    {
      "epoch": 0.897077713551702,
      "grad_norm": 0.3044244945049286,
      "learning_rate": 0.0001,
      "loss": 1.4849,
      "step": 5587
    },
    {
      "epoch": 0.8972382787411689,
      "grad_norm": 0.27884620428085327,
      "learning_rate": 0.0001,
      "loss": 1.3733,
      "step": 5588
    },
    {
      "epoch": 0.8973988439306358,
      "grad_norm": 0.27493929862976074,
      "learning_rate": 0.0001,
      "loss": 1.5163,
      "step": 5589
    },
    {
      "epoch": 0.8975594091201028,
      "grad_norm": 0.2740201950073242,
      "learning_rate": 0.0001,
      "loss": 1.5709,
      "step": 5590
    },
    {
      "epoch": 0.8977199743095697,
      "grad_norm": 0.2665272355079651,
      "learning_rate": 0.0001,
      "loss": 1.4594,
      "step": 5591
    },
    {
      "epoch": 0.8978805394990366,
      "grad_norm": 0.28412479162216187,
      "learning_rate": 0.0001,
      "loss": 1.3765,
      "step": 5592
    },
    {
      "epoch": 0.8980411046885035,
      "grad_norm": 0.25726547837257385,
      "learning_rate": 0.0001,
      "loss": 1.4562,
      "step": 5593
    },
    {
      "epoch": 0.8982016698779705,
      "grad_norm": 0.26011887192726135,
      "learning_rate": 0.0001,
      "loss": 1.4944,
      "step": 5594
    },
    {
      "epoch": 0.8983622350674374,
      "grad_norm": 0.27630525827407837,
      "learning_rate": 0.0001,
      "loss": 1.546,
      "step": 5595
    },
    {
      "epoch": 0.8985228002569043,
      "grad_norm": 0.27536508440971375,
      "learning_rate": 0.0001,
      "loss": 1.4899,
      "step": 5596
    },
    {
      "epoch": 0.8986833654463712,
      "grad_norm": 0.2938755452632904,
      "learning_rate": 0.0001,
      "loss": 1.2796,
      "step": 5597
    },
    {
      "epoch": 0.8988439306358381,
      "grad_norm": 0.2800268232822418,
      "learning_rate": 0.0001,
      "loss": 1.421,
      "step": 5598
    },
    {
      "epoch": 0.8990044958253051,
      "grad_norm": 0.2824115753173828,
      "learning_rate": 0.0001,
      "loss": 1.4583,
      "step": 5599
    },
    {
      "epoch": 0.899165061014772,
      "grad_norm": 0.28322410583496094,
      "learning_rate": 0.0001,
      "loss": 1.4766,
      "step": 5600
    },
    {
      "epoch": 0.899325626204239,
      "grad_norm": 0.2641003727912903,
      "learning_rate": 0.0001,
      "loss": 1.4357,
      "step": 5601
    },
    {
      "epoch": 0.8994861913937059,
      "grad_norm": 0.2806249260902405,
      "learning_rate": 0.0001,
      "loss": 1.4854,
      "step": 5602
    },
    {
      "epoch": 0.8996467565831727,
      "grad_norm": 0.27235904335975647,
      "learning_rate": 0.0001,
      "loss": 1.5044,
      "step": 5603
    },
    {
      "epoch": 0.8998073217726397,
      "grad_norm": 0.2679192125797272,
      "learning_rate": 0.0001,
      "loss": 1.508,
      "step": 5604
    },
    {
      "epoch": 0.8999678869621066,
      "grad_norm": 0.2759861350059509,
      "learning_rate": 0.0001,
      "loss": 1.4847,
      "step": 5605
    },
    {
      "epoch": 0.9001284521515736,
      "grad_norm": 0.26617512106895447,
      "learning_rate": 0.0001,
      "loss": 1.4159,
      "step": 5606
    },
    {
      "epoch": 0.9002890173410405,
      "grad_norm": 0.27368101477622986,
      "learning_rate": 0.0001,
      "loss": 1.4694,
      "step": 5607
    },
    {
      "epoch": 0.9004495825305073,
      "grad_norm": 0.2679339349269867,
      "learning_rate": 0.0001,
      "loss": 1.4271,
      "step": 5608
    },
    {
      "epoch": 0.9006101477199743,
      "grad_norm": 0.2649863660335541,
      "learning_rate": 0.0001,
      "loss": 1.43,
      "step": 5609
    },
    {
      "epoch": 0.9007707129094412,
      "grad_norm": 0.26748520135879517,
      "learning_rate": 0.0001,
      "loss": 1.5332,
      "step": 5610
    },
    {
      "epoch": 0.9009312780989082,
      "grad_norm": 0.26997894048690796,
      "learning_rate": 0.0001,
      "loss": 1.3761,
      "step": 5611
    },
    {
      "epoch": 0.9010918432883751,
      "grad_norm": 0.2782628536224365,
      "learning_rate": 0.0001,
      "loss": 1.4201,
      "step": 5612
    },
    {
      "epoch": 0.901252408477842,
      "grad_norm": 0.2644610106945038,
      "learning_rate": 0.0001,
      "loss": 1.4314,
      "step": 5613
    },
    {
      "epoch": 0.9014129736673089,
      "grad_norm": 0.26767414808273315,
      "learning_rate": 0.0001,
      "loss": 1.4027,
      "step": 5614
    },
    {
      "epoch": 0.9015735388567758,
      "grad_norm": 0.2820192277431488,
      "learning_rate": 0.0001,
      "loss": 1.4734,
      "step": 5615
    },
    {
      "epoch": 0.9017341040462428,
      "grad_norm": 0.27115315198898315,
      "learning_rate": 0.0001,
      "loss": 1.4578,
      "step": 5616
    },
    {
      "epoch": 0.9018946692357097,
      "grad_norm": 0.26880043745040894,
      "learning_rate": 0.0001,
      "loss": 1.4978,
      "step": 5617
    },
    {
      "epoch": 0.9020552344251767,
      "grad_norm": 0.49644196033477783,
      "learning_rate": 0.0001,
      "loss": 1.5041,
      "step": 5618
    },
    {
      "epoch": 0.9022157996146436,
      "grad_norm": 0.2766619920730591,
      "learning_rate": 0.0001,
      "loss": 1.4895,
      "step": 5619
    },
    {
      "epoch": 0.9023763648041104,
      "grad_norm": 0.25244495272636414,
      "learning_rate": 0.0001,
      "loss": 1.4198,
      "step": 5620
    },
    {
      "epoch": 0.9025369299935774,
      "grad_norm": 0.2686665952205658,
      "learning_rate": 0.0001,
      "loss": 1.4735,
      "step": 5621
    },
    {
      "epoch": 0.9026974951830443,
      "grad_norm": 0.2957390546798706,
      "learning_rate": 0.0001,
      "loss": 1.5258,
      "step": 5622
    },
    {
      "epoch": 0.9028580603725113,
      "grad_norm": 0.30311015248298645,
      "learning_rate": 0.0001,
      "loss": 1.4177,
      "step": 5623
    },
    {
      "epoch": 0.9030186255619782,
      "grad_norm": 0.2766197919845581,
      "learning_rate": 0.0001,
      "loss": 1.4683,
      "step": 5624
    },
    {
      "epoch": 0.903179190751445,
      "grad_norm": 0.2730553150177002,
      "learning_rate": 0.0001,
      "loss": 1.4722,
      "step": 5625
    },
    {
      "epoch": 0.903339755940912,
      "grad_norm": 0.28147438168525696,
      "learning_rate": 0.0001,
      "loss": 1.5865,
      "step": 5626
    },
    {
      "epoch": 0.9035003211303789,
      "grad_norm": 0.2749589681625366,
      "learning_rate": 0.0001,
      "loss": 1.4322,
      "step": 5627
    },
    {
      "epoch": 0.9036608863198459,
      "grad_norm": 0.2956511080265045,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 5628
    },
    {
      "epoch": 0.9038214515093128,
      "grad_norm": 0.2774571180343628,
      "learning_rate": 0.0001,
      "loss": 1.4553,
      "step": 5629
    },
    {
      "epoch": 0.9039820166987798,
      "grad_norm": 0.2679741084575653,
      "learning_rate": 0.0001,
      "loss": 1.4936,
      "step": 5630
    },
    {
      "epoch": 0.9041425818882466,
      "grad_norm": 0.3090623915195465,
      "learning_rate": 0.0001,
      "loss": 1.5493,
      "step": 5631
    },
    {
      "epoch": 0.9043031470777135,
      "grad_norm": 0.26591628789901733,
      "learning_rate": 0.0001,
      "loss": 1.5005,
      "step": 5632
    },
    {
      "epoch": 0.9044637122671805,
      "grad_norm": 0.2733018100261688,
      "learning_rate": 0.0001,
      "loss": 1.4519,
      "step": 5633
    },
    {
      "epoch": 0.9046242774566474,
      "grad_norm": 0.2661973834037781,
      "learning_rate": 0.0001,
      "loss": 1.5416,
      "step": 5634
    },
    {
      "epoch": 0.9047848426461144,
      "grad_norm": 0.27166926860809326,
      "learning_rate": 0.0001,
      "loss": 1.4797,
      "step": 5635
    },
    {
      "epoch": 0.9049454078355812,
      "grad_norm": 0.2655140459537506,
      "learning_rate": 0.0001,
      "loss": 1.4686,
      "step": 5636
    },
    {
      "epoch": 0.9051059730250481,
      "grad_norm": 0.28489869832992554,
      "learning_rate": 0.0001,
      "loss": 1.4468,
      "step": 5637
    },
    {
      "epoch": 0.9052665382145151,
      "grad_norm": 0.2994062304496765,
      "learning_rate": 0.0001,
      "loss": 1.4438,
      "step": 5638
    },
    {
      "epoch": 0.905427103403982,
      "grad_norm": 0.2887355089187622,
      "learning_rate": 0.0001,
      "loss": 1.4588,
      "step": 5639
    },
    {
      "epoch": 0.905587668593449,
      "grad_norm": 0.29034167528152466,
      "learning_rate": 0.0001,
      "loss": 1.5389,
      "step": 5640
    },
    {
      "epoch": 0.9057482337829159,
      "grad_norm": 0.28640443086624146,
      "learning_rate": 0.0001,
      "loss": 1.4877,
      "step": 5641
    },
    {
      "epoch": 0.9059087989723827,
      "grad_norm": 0.28215524554252625,
      "learning_rate": 0.0001,
      "loss": 1.4479,
      "step": 5642
    },
    {
      "epoch": 0.9060693641618497,
      "grad_norm": 0.32695475220680237,
      "learning_rate": 0.0001,
      "loss": 1.4924,
      "step": 5643
    },
    {
      "epoch": 0.9062299293513166,
      "grad_norm": 0.29609331488609314,
      "learning_rate": 0.0001,
      "loss": 1.5179,
      "step": 5644
    },
    {
      "epoch": 0.9063904945407836,
      "grad_norm": 0.2883678078651428,
      "learning_rate": 0.0001,
      "loss": 1.4947,
      "step": 5645
    },
    {
      "epoch": 0.9065510597302505,
      "grad_norm": 0.2817707657814026,
      "learning_rate": 0.0001,
      "loss": 1.4946,
      "step": 5646
    },
    {
      "epoch": 0.9067116249197174,
      "grad_norm": 0.26052823662757874,
      "learning_rate": 0.0001,
      "loss": 1.4611,
      "step": 5647
    },
    {
      "epoch": 0.9068721901091843,
      "grad_norm": 0.26856547594070435,
      "learning_rate": 0.0001,
      "loss": 1.4527,
      "step": 5648
    },
    {
      "epoch": 0.9070327552986512,
      "grad_norm": 0.2800810933113098,
      "learning_rate": 0.0001,
      "loss": 1.4586,
      "step": 5649
    },
    {
      "epoch": 0.9071933204881182,
      "grad_norm": 0.2645875811576843,
      "learning_rate": 0.0001,
      "loss": 1.4309,
      "step": 5650
    },
    {
      "epoch": 0.9073538856775851,
      "grad_norm": 0.28114935755729675,
      "learning_rate": 0.0001,
      "loss": 1.5123,
      "step": 5651
    },
    {
      "epoch": 0.9075144508670521,
      "grad_norm": 0.2695562243461609,
      "learning_rate": 0.0001,
      "loss": 1.4983,
      "step": 5652
    },
    {
      "epoch": 0.9076750160565189,
      "grad_norm": 0.27862122654914856,
      "learning_rate": 0.0001,
      "loss": 1.4686,
      "step": 5653
    },
    {
      "epoch": 0.9078355812459858,
      "grad_norm": 0.9690446257591248,
      "learning_rate": 0.0001,
      "loss": 1.5502,
      "step": 5654
    },
    {
      "epoch": 0.9079961464354528,
      "grad_norm": 0.27264535427093506,
      "learning_rate": 0.0001,
      "loss": 1.4717,
      "step": 5655
    },
    {
      "epoch": 0.9081567116249197,
      "grad_norm": 0.27225714921951294,
      "learning_rate": 0.0001,
      "loss": 1.4961,
      "step": 5656
    },
    {
      "epoch": 0.9083172768143867,
      "grad_norm": 0.28575438261032104,
      "learning_rate": 0.0001,
      "loss": 1.4547,
      "step": 5657
    },
    {
      "epoch": 0.9084778420038536,
      "grad_norm": 0.29622870683670044,
      "learning_rate": 0.0001,
      "loss": 1.3789,
      "step": 5658
    },
    {
      "epoch": 0.9086384071933205,
      "grad_norm": 0.28283485770225525,
      "learning_rate": 0.0001,
      "loss": 1.4775,
      "step": 5659
    },
    {
      "epoch": 0.9087989723827874,
      "grad_norm": 0.27316781878471375,
      "learning_rate": 0.0001,
      "loss": 1.4421,
      "step": 5660
    },
    {
      "epoch": 0.9089595375722543,
      "grad_norm": 0.2823334038257599,
      "learning_rate": 0.0001,
      "loss": 1.4682,
      "step": 5661
    },
    {
      "epoch": 0.9091201027617213,
      "grad_norm": 0.27554428577423096,
      "learning_rate": 0.0001,
      "loss": 1.4361,
      "step": 5662
    },
    {
      "epoch": 0.9092806679511882,
      "grad_norm": 0.27766990661621094,
      "learning_rate": 0.0001,
      "loss": 1.527,
      "step": 5663
    },
    {
      "epoch": 0.9094412331406551,
      "grad_norm": 0.29925546050071716,
      "learning_rate": 0.0001,
      "loss": 1.5193,
      "step": 5664
    },
    {
      "epoch": 0.909601798330122,
      "grad_norm": 0.27168726921081543,
      "learning_rate": 0.0001,
      "loss": 1.4366,
      "step": 5665
    },
    {
      "epoch": 0.909762363519589,
      "grad_norm": 0.2684997022151947,
      "learning_rate": 0.0001,
      "loss": 1.4119,
      "step": 5666
    },
    {
      "epoch": 0.9099229287090559,
      "grad_norm": 0.2845095694065094,
      "learning_rate": 0.0001,
      "loss": 1.5543,
      "step": 5667
    },
    {
      "epoch": 0.9100834938985228,
      "grad_norm": 0.2682120203971863,
      "learning_rate": 0.0001,
      "loss": 1.4631,
      "step": 5668
    },
    {
      "epoch": 0.9102440590879898,
      "grad_norm": 0.26111850142478943,
      "learning_rate": 0.0001,
      "loss": 1.4614,
      "step": 5669
    },
    {
      "epoch": 0.9104046242774566,
      "grad_norm": 0.26749932765960693,
      "learning_rate": 0.0001,
      "loss": 1.5122,
      "step": 5670
    },
    {
      "epoch": 0.9105651894669236,
      "grad_norm": 0.27375659346580505,
      "learning_rate": 0.0001,
      "loss": 1.4932,
      "step": 5671
    },
    {
      "epoch": 0.9107257546563905,
      "grad_norm": 0.2742297947406769,
      "learning_rate": 0.0001,
      "loss": 1.4931,
      "step": 5672
    },
    {
      "epoch": 0.9108863198458574,
      "grad_norm": 0.2770094573497772,
      "learning_rate": 0.0001,
      "loss": 1.3913,
      "step": 5673
    },
    {
      "epoch": 0.9110468850353244,
      "grad_norm": 0.284294068813324,
      "learning_rate": 0.0001,
      "loss": 1.5404,
      "step": 5674
    },
    {
      "epoch": 0.9112074502247912,
      "grad_norm": 0.27823662757873535,
      "learning_rate": 0.0001,
      "loss": 1.5444,
      "step": 5675
    },
    {
      "epoch": 0.9113680154142582,
      "grad_norm": 0.27007046341896057,
      "learning_rate": 0.0001,
      "loss": 1.4577,
      "step": 5676
    },
    {
      "epoch": 0.9115285806037251,
      "grad_norm": 0.27728602290153503,
      "learning_rate": 0.0001,
      "loss": 1.531,
      "step": 5677
    },
    {
      "epoch": 0.911689145793192,
      "grad_norm": 0.3271082043647766,
      "learning_rate": 0.0001,
      "loss": 1.4856,
      "step": 5678
    },
    {
      "epoch": 0.911849710982659,
      "grad_norm": 0.27263563871383667,
      "learning_rate": 0.0001,
      "loss": 1.4884,
      "step": 5679
    },
    {
      "epoch": 0.9120102761721259,
      "grad_norm": 0.2838371992111206,
      "learning_rate": 0.0001,
      "loss": 1.5213,
      "step": 5680
    },
    {
      "epoch": 0.9121708413615928,
      "grad_norm": 0.26938414573669434,
      "learning_rate": 0.0001,
      "loss": 1.4646,
      "step": 5681
    },
    {
      "epoch": 0.9123314065510597,
      "grad_norm": 0.283206045627594,
      "learning_rate": 0.0001,
      "loss": 1.482,
      "step": 5682
    },
    {
      "epoch": 0.9124919717405267,
      "grad_norm": 0.2698117196559906,
      "learning_rate": 0.0001,
      "loss": 1.4331,
      "step": 5683
    },
    {
      "epoch": 0.9126525369299936,
      "grad_norm": 0.2849421203136444,
      "learning_rate": 0.0001,
      "loss": 1.5028,
      "step": 5684
    },
    {
      "epoch": 0.9128131021194605,
      "grad_norm": 0.27039268612861633,
      "learning_rate": 0.0001,
      "loss": 1.4567,
      "step": 5685
    },
    {
      "epoch": 0.9129736673089274,
      "grad_norm": 0.26833176612854004,
      "learning_rate": 0.0001,
      "loss": 1.4978,
      "step": 5686
    },
    {
      "epoch": 0.9131342324983943,
      "grad_norm": 0.2842022776603699,
      "learning_rate": 0.0001,
      "loss": 1.4712,
      "step": 5687
    },
    {
      "epoch": 0.9132947976878613,
      "grad_norm": 0.28925463557243347,
      "learning_rate": 0.0001,
      "loss": 1.5191,
      "step": 5688
    },
    {
      "epoch": 0.9134553628773282,
      "grad_norm": 0.2731495201587677,
      "learning_rate": 0.0001,
      "loss": 1.456,
      "step": 5689
    },
    {
      "epoch": 0.9136159280667951,
      "grad_norm": 0.2853472828865051,
      "learning_rate": 0.0001,
      "loss": 1.5377,
      "step": 5690
    },
    {
      "epoch": 0.9137764932562621,
      "grad_norm": 0.2628810703754425,
      "learning_rate": 0.0001,
      "loss": 1.4697,
      "step": 5691
    },
    {
      "epoch": 0.9139370584457289,
      "grad_norm": 0.2630572021007538,
      "learning_rate": 0.0001,
      "loss": 1.4778,
      "step": 5692
    },
    {
      "epoch": 0.9140976236351959,
      "grad_norm": 0.27523350715637207,
      "learning_rate": 0.0001,
      "loss": 1.4572,
      "step": 5693
    },
    {
      "epoch": 0.9142581888246628,
      "grad_norm": 0.267199844121933,
      "learning_rate": 0.0001,
      "loss": 1.5051,
      "step": 5694
    },
    {
      "epoch": 0.9144187540141298,
      "grad_norm": 0.2666296362876892,
      "learning_rate": 0.0001,
      "loss": 1.4595,
      "step": 5695
    },
    {
      "epoch": 0.9145793192035967,
      "grad_norm": 0.2623789608478546,
      "learning_rate": 0.0001,
      "loss": 1.4275,
      "step": 5696
    },
    {
      "epoch": 0.9147398843930635,
      "grad_norm": 0.2783721387386322,
      "learning_rate": 0.0001,
      "loss": 1.4741,
      "step": 5697
    },
    {
      "epoch": 0.9149004495825305,
      "grad_norm": 0.2690523862838745,
      "learning_rate": 0.0001,
      "loss": 1.4386,
      "step": 5698
    },
    {
      "epoch": 0.9150610147719974,
      "grad_norm": 0.27985507249832153,
      "learning_rate": 0.0001,
      "loss": 1.4814,
      "step": 5699
    },
    {
      "epoch": 0.9152215799614644,
      "grad_norm": 0.2744964063167572,
      "learning_rate": 0.0001,
      "loss": 1.5168,
      "step": 5700
    },
    {
      "epoch": 0.9153821451509313,
      "grad_norm": 0.29585015773773193,
      "learning_rate": 0.0001,
      "loss": 1.4577,
      "step": 5701
    },
    {
      "epoch": 0.9155427103403982,
      "grad_norm": 0.6999393105506897,
      "learning_rate": 0.0001,
      "loss": 1.4938,
      "step": 5702
    },
    {
      "epoch": 0.9157032755298651,
      "grad_norm": 0.26724332571029663,
      "learning_rate": 0.0001,
      "loss": 1.4642,
      "step": 5703
    },
    {
      "epoch": 0.915863840719332,
      "grad_norm": 0.26325684785842896,
      "learning_rate": 0.0001,
      "loss": 1.5335,
      "step": 5704
    },
    {
      "epoch": 0.916024405908799,
      "grad_norm": 0.2777913808822632,
      "learning_rate": 0.0001,
      "loss": 1.4725,
      "step": 5705
    },
    {
      "epoch": 0.9161849710982659,
      "grad_norm": 0.277014821767807,
      "learning_rate": 0.0001,
      "loss": 1.4831,
      "step": 5706
    },
    {
      "epoch": 0.9163455362877329,
      "grad_norm": 0.2649841010570526,
      "learning_rate": 0.0001,
      "loss": 1.4683,
      "step": 5707
    },
    {
      "epoch": 0.9165061014771998,
      "grad_norm": 0.25996530055999756,
      "learning_rate": 0.0001,
      "loss": 1.4099,
      "step": 5708
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.27567023038864136,
      "learning_rate": 0.0001,
      "loss": 1.4417,
      "step": 5709
    },
    {
      "epoch": 0.9168272318561336,
      "grad_norm": 0.26955941319465637,
      "learning_rate": 0.0001,
      "loss": 1.4445,
      "step": 5710
    },
    {
      "epoch": 0.9169877970456005,
      "grad_norm": 0.27277833223342896,
      "learning_rate": 0.0001,
      "loss": 1.5454,
      "step": 5711
    },
    {
      "epoch": 0.9171483622350675,
      "grad_norm": 0.26426592469215393,
      "learning_rate": 0.0001,
      "loss": 1.4566,
      "step": 5712
    },
    {
      "epoch": 0.9173089274245344,
      "grad_norm": 0.26648545265197754,
      "learning_rate": 0.0001,
      "loss": 1.49,
      "step": 5713
    },
    {
      "epoch": 0.9174694926140012,
      "grad_norm": 0.266801118850708,
      "learning_rate": 0.0001,
      "loss": 1.3964,
      "step": 5714
    },
    {
      "epoch": 0.9176300578034682,
      "grad_norm": 0.2627018690109253,
      "learning_rate": 0.0001,
      "loss": 1.486,
      "step": 5715
    },
    {
      "epoch": 0.9177906229929351,
      "grad_norm": 0.2842526137828827,
      "learning_rate": 0.0001,
      "loss": 1.4861,
      "step": 5716
    },
    {
      "epoch": 0.9179511881824021,
      "grad_norm": 0.26749587059020996,
      "learning_rate": 0.0001,
      "loss": 1.424,
      "step": 5717
    },
    {
      "epoch": 0.918111753371869,
      "grad_norm": 0.28474441170692444,
      "learning_rate": 0.0001,
      "loss": 1.4852,
      "step": 5718
    },
    {
      "epoch": 0.918272318561336,
      "grad_norm": 0.26730814576148987,
      "learning_rate": 0.0001,
      "loss": 1.3754,
      "step": 5719
    },
    {
      "epoch": 0.9184328837508028,
      "grad_norm": 0.27947551012039185,
      "learning_rate": 0.0001,
      "loss": 1.4497,
      "step": 5720
    },
    {
      "epoch": 0.9185934489402697,
      "grad_norm": 0.28426793217658997,
      "learning_rate": 0.0001,
      "loss": 1.4121,
      "step": 5721
    },
    {
      "epoch": 0.9187540141297367,
      "grad_norm": 0.28511571884155273,
      "learning_rate": 0.0001,
      "loss": 1.4188,
      "step": 5722
    },
    {
      "epoch": 0.9189145793192036,
      "grad_norm": 0.29709392786026,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 5723
    },
    {
      "epoch": 0.9190751445086706,
      "grad_norm": 0.2649746537208557,
      "learning_rate": 0.0001,
      "loss": 1.4985,
      "step": 5724
    },
    {
      "epoch": 0.9192357096981374,
      "grad_norm": 0.2661975026130676,
      "learning_rate": 0.0001,
      "loss": 1.5049,
      "step": 5725
    },
    {
      "epoch": 0.9193962748876043,
      "grad_norm": 0.29220449924468994,
      "learning_rate": 0.0001,
      "loss": 1.4678,
      "step": 5726
    },
    {
      "epoch": 0.9195568400770713,
      "grad_norm": 0.26839110255241394,
      "learning_rate": 0.0001,
      "loss": 1.3938,
      "step": 5727
    },
    {
      "epoch": 0.9197174052665382,
      "grad_norm": 0.2792374789714813,
      "learning_rate": 0.0001,
      "loss": 1.4432,
      "step": 5728
    },
    {
      "epoch": 0.9198779704560052,
      "grad_norm": 0.2627927362918854,
      "learning_rate": 0.0001,
      "loss": 1.4709,
      "step": 5729
    },
    {
      "epoch": 0.9200385356454721,
      "grad_norm": 0.2641960084438324,
      "learning_rate": 0.0001,
      "loss": 1.5006,
      "step": 5730
    },
    {
      "epoch": 0.9201991008349389,
      "grad_norm": 0.27540579438209534,
      "learning_rate": 0.0001,
      "loss": 1.4866,
      "step": 5731
    },
    {
      "epoch": 0.9203596660244059,
      "grad_norm": 0.2716390788555145,
      "learning_rate": 0.0001,
      "loss": 1.4407,
      "step": 5732
    },
    {
      "epoch": 0.9205202312138728,
      "grad_norm": 0.2709677219390869,
      "learning_rate": 0.0001,
      "loss": 1.4148,
      "step": 5733
    },
    {
      "epoch": 0.9206807964033398,
      "grad_norm": 0.284290075302124,
      "learning_rate": 0.0001,
      "loss": 1.5233,
      "step": 5734
    },
    {
      "epoch": 0.9208413615928067,
      "grad_norm": 0.2621372938156128,
      "learning_rate": 0.0001,
      "loss": 1.4556,
      "step": 5735
    },
    {
      "epoch": 0.9210019267822736,
      "grad_norm": 0.27332156896591187,
      "learning_rate": 0.0001,
      "loss": 1.376,
      "step": 5736
    },
    {
      "epoch": 0.9211624919717405,
      "grad_norm": 0.27216067910194397,
      "learning_rate": 0.0001,
      "loss": 1.5591,
      "step": 5737
    },
    {
      "epoch": 0.9213230571612074,
      "grad_norm": 0.27463793754577637,
      "learning_rate": 0.0001,
      "loss": 1.5493,
      "step": 5738
    },
    {
      "epoch": 0.9214836223506744,
      "grad_norm": 0.28009170293807983,
      "learning_rate": 0.0001,
      "loss": 1.5244,
      "step": 5739
    },
    {
      "epoch": 0.9216441875401413,
      "grad_norm": 0.2847888767719269,
      "learning_rate": 0.0001,
      "loss": 1.49,
      "step": 5740
    },
    {
      "epoch": 0.9218047527296083,
      "grad_norm": 0.26402273774147034,
      "learning_rate": 0.0001,
      "loss": 1.5125,
      "step": 5741
    },
    {
      "epoch": 0.9219653179190751,
      "grad_norm": 0.2641732096672058,
      "learning_rate": 0.0001,
      "loss": 1.5069,
      "step": 5742
    },
    {
      "epoch": 0.922125883108542,
      "grad_norm": 0.29763108491897583,
      "learning_rate": 0.0001,
      "loss": 1.4892,
      "step": 5743
    },
    {
      "epoch": 0.922286448298009,
      "grad_norm": 0.2714250385761261,
      "learning_rate": 0.0001,
      "loss": 1.516,
      "step": 5744
    },
    {
      "epoch": 0.9224470134874759,
      "grad_norm": 0.2819075584411621,
      "learning_rate": 0.0001,
      "loss": 1.5841,
      "step": 5745
    },
    {
      "epoch": 0.9226075786769429,
      "grad_norm": 0.2795034945011139,
      "learning_rate": 0.0001,
      "loss": 1.4864,
      "step": 5746
    },
    {
      "epoch": 0.9227681438664097,
      "grad_norm": 0.2678384482860565,
      "learning_rate": 0.0001,
      "loss": 1.4533,
      "step": 5747
    },
    {
      "epoch": 0.9229287090558767,
      "grad_norm": 0.2672341763973236,
      "learning_rate": 0.0001,
      "loss": 1.5334,
      "step": 5748
    },
    {
      "epoch": 0.9230892742453436,
      "grad_norm": 0.27180591225624084,
      "learning_rate": 0.0001,
      "loss": 1.4305,
      "step": 5749
    },
    {
      "epoch": 0.9232498394348105,
      "grad_norm": 0.2737089693546295,
      "learning_rate": 0.0001,
      "loss": 1.5157,
      "step": 5750
    },
    {
      "epoch": 0.9234104046242775,
      "grad_norm": 0.2823292911052704,
      "learning_rate": 0.0001,
      "loss": 1.4531,
      "step": 5751
    },
    {
      "epoch": 0.9235709698137444,
      "grad_norm": 0.2742321789264679,
      "learning_rate": 0.0001,
      "loss": 1.4078,
      "step": 5752
    },
    {
      "epoch": 0.9237315350032113,
      "grad_norm": 0.2717142105102539,
      "learning_rate": 0.0001,
      "loss": 1.4479,
      "step": 5753
    },
    {
      "epoch": 0.9238921001926782,
      "grad_norm": 0.2947547435760498,
      "learning_rate": 0.0001,
      "loss": 1.4127,
      "step": 5754
    },
    {
      "epoch": 0.9240526653821451,
      "grad_norm": 0.2838168144226074,
      "learning_rate": 0.0001,
      "loss": 1.476,
      "step": 5755
    },
    {
      "epoch": 0.9242132305716121,
      "grad_norm": 0.27356472611427307,
      "learning_rate": 0.0001,
      "loss": 1.4374,
      "step": 5756
    },
    {
      "epoch": 0.924373795761079,
      "grad_norm": 0.303329735994339,
      "learning_rate": 0.0001,
      "loss": 1.5093,
      "step": 5757
    },
    {
      "epoch": 0.924534360950546,
      "grad_norm": 0.32209450006484985,
      "learning_rate": 0.0001,
      "loss": 1.5002,
      "step": 5758
    },
    {
      "epoch": 0.9246949261400128,
      "grad_norm": 0.2569775879383087,
      "learning_rate": 0.0001,
      "loss": 1.5086,
      "step": 5759
    },
    {
      "epoch": 0.9248554913294798,
      "grad_norm": 0.27903053164482117,
      "learning_rate": 0.0001,
      "loss": 1.4443,
      "step": 5760
    },
    {
      "epoch": 0.9250160565189467,
      "grad_norm": 0.25670740008354187,
      "learning_rate": 0.0001,
      "loss": 1.5015,
      "step": 5761
    },
    {
      "epoch": 0.9251766217084136,
      "grad_norm": 0.28923937678337097,
      "learning_rate": 0.0001,
      "loss": 1.4439,
      "step": 5762
    },
    {
      "epoch": 0.9253371868978806,
      "grad_norm": 0.2757799029350281,
      "learning_rate": 0.0001,
      "loss": 1.4689,
      "step": 5763
    },
    {
      "epoch": 0.9254977520873474,
      "grad_norm": 0.2755475342273712,
      "learning_rate": 0.0001,
      "loss": 1.473,
      "step": 5764
    },
    {
      "epoch": 0.9256583172768144,
      "grad_norm": 0.2650447189807892,
      "learning_rate": 0.0001,
      "loss": 1.3937,
      "step": 5765
    },
    {
      "epoch": 0.9258188824662813,
      "grad_norm": 0.2696771025657654,
      "learning_rate": 0.0001,
      "loss": 1.4672,
      "step": 5766
    },
    {
      "epoch": 0.9259794476557482,
      "grad_norm": 0.29964062571525574,
      "learning_rate": 0.0001,
      "loss": 1.4711,
      "step": 5767
    },
    {
      "epoch": 0.9261400128452152,
      "grad_norm": 0.32170623540878296,
      "learning_rate": 0.0001,
      "loss": 1.4489,
      "step": 5768
    },
    {
      "epoch": 0.9263005780346821,
      "grad_norm": 0.2805902361869812,
      "learning_rate": 0.0001,
      "loss": 1.4482,
      "step": 5769
    },
    {
      "epoch": 0.926461143224149,
      "grad_norm": 0.2696913778781891,
      "learning_rate": 0.0001,
      "loss": 1.4797,
      "step": 5770
    },
    {
      "epoch": 0.9266217084136159,
      "grad_norm": 0.27366599440574646,
      "learning_rate": 0.0001,
      "loss": 1.4612,
      "step": 5771
    },
    {
      "epoch": 0.9267822736030829,
      "grad_norm": 0.30340951681137085,
      "learning_rate": 0.0001,
      "loss": 1.4558,
      "step": 5772
    },
    {
      "epoch": 0.9269428387925498,
      "grad_norm": 0.2692902386188507,
      "learning_rate": 0.0001,
      "loss": 1.445,
      "step": 5773
    },
    {
      "epoch": 0.9271034039820167,
      "grad_norm": 0.29561668634414673,
      "learning_rate": 0.0001,
      "loss": 1.3912,
      "step": 5774
    },
    {
      "epoch": 0.9272639691714836,
      "grad_norm": 0.2912541925907135,
      "learning_rate": 0.0001,
      "loss": 1.4678,
      "step": 5775
    },
    {
      "epoch": 0.9274245343609505,
      "grad_norm": 0.27322182059288025,
      "learning_rate": 0.0001,
      "loss": 1.4082,
      "step": 5776
    },
    {
      "epoch": 0.9275850995504175,
      "grad_norm": 0.29552537202835083,
      "learning_rate": 0.0001,
      "loss": 1.4656,
      "step": 5777
    },
    {
      "epoch": 0.9277456647398844,
      "grad_norm": 0.2890673577785492,
      "learning_rate": 0.0001,
      "loss": 1.4762,
      "step": 5778
    },
    {
      "epoch": 0.9279062299293513,
      "grad_norm": 0.27739542722702026,
      "learning_rate": 0.0001,
      "loss": 1.556,
      "step": 5779
    },
    {
      "epoch": 0.9280667951188183,
      "grad_norm": 0.2785024046897888,
      "learning_rate": 0.0001,
      "loss": 1.5496,
      "step": 5780
    },
    {
      "epoch": 0.9282273603082851,
      "grad_norm": 0.25842979550361633,
      "learning_rate": 0.0001,
      "loss": 1.4732,
      "step": 5781
    },
    {
      "epoch": 0.9283879254977521,
      "grad_norm": 0.29370245337486267,
      "learning_rate": 0.0001,
      "loss": 1.4841,
      "step": 5782
    },
    {
      "epoch": 0.928548490687219,
      "grad_norm": 0.27456536889076233,
      "learning_rate": 0.0001,
      "loss": 1.401,
      "step": 5783
    },
    {
      "epoch": 0.928709055876686,
      "grad_norm": 0.2704974412918091,
      "learning_rate": 0.0001,
      "loss": 1.4826,
      "step": 5784
    },
    {
      "epoch": 0.9288696210661529,
      "grad_norm": 0.30428287386894226,
      "learning_rate": 0.0001,
      "loss": 1.4592,
      "step": 5785
    },
    {
      "epoch": 0.9290301862556197,
      "grad_norm": 0.29401740431785583,
      "learning_rate": 0.0001,
      "loss": 1.4012,
      "step": 5786
    },
    {
      "epoch": 0.9291907514450867,
      "grad_norm": 0.2812541723251343,
      "learning_rate": 0.0001,
      "loss": 1.3853,
      "step": 5787
    },
    {
      "epoch": 0.9293513166345536,
      "grad_norm": 0.28248366713523865,
      "learning_rate": 0.0001,
      "loss": 1.4619,
      "step": 5788
    },
    {
      "epoch": 0.9295118818240206,
      "grad_norm": 0.2675034999847412,
      "learning_rate": 0.0001,
      "loss": 1.4516,
      "step": 5789
    },
    {
      "epoch": 0.9296724470134875,
      "grad_norm": 0.26978638768196106,
      "learning_rate": 0.0001,
      "loss": 1.4693,
      "step": 5790
    },
    {
      "epoch": 0.9298330122029544,
      "grad_norm": 0.2856466770172119,
      "learning_rate": 0.0001,
      "loss": 1.5821,
      "step": 5791
    },
    {
      "epoch": 0.9299935773924213,
      "grad_norm": 0.271677166223526,
      "learning_rate": 0.0001,
      "loss": 1.3973,
      "step": 5792
    },
    {
      "epoch": 0.9301541425818882,
      "grad_norm": 0.28180110454559326,
      "learning_rate": 0.0001,
      "loss": 1.4597,
      "step": 5793
    },
    {
      "epoch": 0.9303147077713552,
      "grad_norm": 0.27508625388145447,
      "learning_rate": 0.0001,
      "loss": 1.5676,
      "step": 5794
    },
    {
      "epoch": 0.9304752729608221,
      "grad_norm": 0.2649001479148865,
      "learning_rate": 0.0001,
      "loss": 1.4636,
      "step": 5795
    },
    {
      "epoch": 0.930635838150289,
      "grad_norm": 0.28027021884918213,
      "learning_rate": 0.0001,
      "loss": 1.4499,
      "step": 5796
    },
    {
      "epoch": 0.930796403339756,
      "grad_norm": 0.28741174936294556,
      "learning_rate": 0.0001,
      "loss": 1.4998,
      "step": 5797
    },
    {
      "epoch": 0.9309569685292228,
      "grad_norm": 0.2758386433124542,
      "learning_rate": 0.0001,
      "loss": 1.5158,
      "step": 5798
    },
    {
      "epoch": 0.9311175337186898,
      "grad_norm": 0.2677361071109772,
      "learning_rate": 0.0001,
      "loss": 1.4557,
      "step": 5799
    },
    {
      "epoch": 0.9312780989081567,
      "grad_norm": 0.2974143326282501,
      "learning_rate": 0.0001,
      "loss": 1.425,
      "step": 5800
    },
    {
      "epoch": 0.9314386640976237,
      "grad_norm": 0.3010811507701874,
      "learning_rate": 0.0001,
      "loss": 1.3815,
      "step": 5801
    },
    {
      "epoch": 0.9315992292870906,
      "grad_norm": 0.2760634124279022,
      "learning_rate": 0.0001,
      "loss": 1.48,
      "step": 5802
    },
    {
      "epoch": 0.9317597944765574,
      "grad_norm": 0.2719084322452545,
      "learning_rate": 0.0001,
      "loss": 1.4971,
      "step": 5803
    },
    {
      "epoch": 0.9319203596660244,
      "grad_norm": 0.2982003092765808,
      "learning_rate": 0.0001,
      "loss": 1.5477,
      "step": 5804
    },
    {
      "epoch": 0.9320809248554913,
      "grad_norm": 0.30464425683021545,
      "learning_rate": 0.0001,
      "loss": 1.498,
      "step": 5805
    },
    {
      "epoch": 0.9322414900449583,
      "grad_norm": 0.2761540114879608,
      "learning_rate": 0.0001,
      "loss": 1.4721,
      "step": 5806
    },
    {
      "epoch": 0.9324020552344252,
      "grad_norm": 0.2752193510532379,
      "learning_rate": 0.0001,
      "loss": 1.4132,
      "step": 5807
    },
    {
      "epoch": 0.9325626204238922,
      "grad_norm": 0.26726093888282776,
      "learning_rate": 0.0001,
      "loss": 1.4526,
      "step": 5808
    },
    {
      "epoch": 0.932723185613359,
      "grad_norm": 0.26164042949676514,
      "learning_rate": 0.0001,
      "loss": 1.4471,
      "step": 5809
    },
    {
      "epoch": 0.9328837508028259,
      "grad_norm": 0.268328994512558,
      "learning_rate": 0.0001,
      "loss": 1.4582,
      "step": 5810
    },
    {
      "epoch": 0.9330443159922929,
      "grad_norm": 0.2759345471858978,
      "learning_rate": 0.0001,
      "loss": 1.5367,
      "step": 5811
    },
    {
      "epoch": 0.9332048811817598,
      "grad_norm": 0.2661013901233673,
      "learning_rate": 0.0001,
      "loss": 1.558,
      "step": 5812
    },
    {
      "epoch": 0.9333654463712268,
      "grad_norm": 0.2785830795764923,
      "learning_rate": 0.0001,
      "loss": 1.456,
      "step": 5813
    },
    {
      "epoch": 0.9335260115606936,
      "grad_norm": 0.26263684034347534,
      "learning_rate": 0.0001,
      "loss": 1.4605,
      "step": 5814
    },
    {
      "epoch": 0.9336865767501605,
      "grad_norm": 0.27695077657699585,
      "learning_rate": 0.0001,
      "loss": 1.4791,
      "step": 5815
    },
    {
      "epoch": 0.9338471419396275,
      "grad_norm": 0.28673410415649414,
      "learning_rate": 0.0001,
      "loss": 1.4811,
      "step": 5816
    },
    {
      "epoch": 0.9340077071290944,
      "grad_norm": 0.28872042894363403,
      "learning_rate": 0.0001,
      "loss": 1.5044,
      "step": 5817
    },
    {
      "epoch": 0.9341682723185614,
      "grad_norm": 0.27367937564849854,
      "learning_rate": 0.0001,
      "loss": 1.5029,
      "step": 5818
    },
    {
      "epoch": 0.9343288375080283,
      "grad_norm": 0.2637515366077423,
      "learning_rate": 0.0001,
      "loss": 1.4633,
      "step": 5819
    },
    {
      "epoch": 0.9344894026974951,
      "grad_norm": 0.2860463559627533,
      "learning_rate": 0.0001,
      "loss": 1.4937,
      "step": 5820
    },
    {
      "epoch": 0.9346499678869621,
      "grad_norm": 0.28555914759635925,
      "learning_rate": 0.0001,
      "loss": 1.445,
      "step": 5821
    },
    {
      "epoch": 0.934810533076429,
      "grad_norm": 0.27750441431999207,
      "learning_rate": 0.0001,
      "loss": 1.4198,
      "step": 5822
    },
    {
      "epoch": 0.934971098265896,
      "grad_norm": 0.2655699551105499,
      "learning_rate": 0.0001,
      "loss": 1.4535,
      "step": 5823
    },
    {
      "epoch": 0.9351316634553629,
      "grad_norm": 0.27347785234451294,
      "learning_rate": 0.0001,
      "loss": 1.5319,
      "step": 5824
    },
    {
      "epoch": 0.9352922286448297,
      "grad_norm": 0.28993481397628784,
      "learning_rate": 0.0001,
      "loss": 1.4661,
      "step": 5825
    },
    {
      "epoch": 0.9354527938342967,
      "grad_norm": 0.273605614900589,
      "learning_rate": 0.0001,
      "loss": 1.52,
      "step": 5826
    },
    {
      "epoch": 0.9356133590237636,
      "grad_norm": 0.25676229596138,
      "learning_rate": 0.0001,
      "loss": 1.4204,
      "step": 5827
    },
    {
      "epoch": 0.9357739242132306,
      "grad_norm": 0.27479469776153564,
      "learning_rate": 0.0001,
      "loss": 1.441,
      "step": 5828
    },
    {
      "epoch": 0.9359344894026975,
      "grad_norm": 0.2789990305900574,
      "learning_rate": 0.0001,
      "loss": 1.472,
      "step": 5829
    },
    {
      "epoch": 0.9360950545921645,
      "grad_norm": 0.27907949686050415,
      "learning_rate": 0.0001,
      "loss": 1.4628,
      "step": 5830
    },
    {
      "epoch": 0.9362556197816313,
      "grad_norm": 0.27137044072151184,
      "learning_rate": 0.0001,
      "loss": 1.4569,
      "step": 5831
    },
    {
      "epoch": 0.9364161849710982,
      "grad_norm": 0.2736453115940094,
      "learning_rate": 0.0001,
      "loss": 1.4437,
      "step": 5832
    },
    {
      "epoch": 0.9365767501605652,
      "grad_norm": 0.28897377848625183,
      "learning_rate": 0.0001,
      "loss": 1.4673,
      "step": 5833
    },
    {
      "epoch": 0.9367373153500321,
      "grad_norm": 0.26451051235198975,
      "learning_rate": 0.0001,
      "loss": 1.4595,
      "step": 5834
    },
    {
      "epoch": 0.9368978805394991,
      "grad_norm": 0.27823954820632935,
      "learning_rate": 0.0001,
      "loss": 1.4499,
      "step": 5835
    },
    {
      "epoch": 0.9370584457289659,
      "grad_norm": 0.27579158544540405,
      "learning_rate": 0.0001,
      "loss": 1.4965,
      "step": 5836
    },
    {
      "epoch": 0.9372190109184328,
      "grad_norm": 0.2658326029777527,
      "learning_rate": 0.0001,
      "loss": 1.4783,
      "step": 5837
    },
    {
      "epoch": 0.9373795761078998,
      "grad_norm": 0.28364139795303345,
      "learning_rate": 0.0001,
      "loss": 1.4208,
      "step": 5838
    },
    {
      "epoch": 0.9375401412973667,
      "grad_norm": 0.26561906933784485,
      "learning_rate": 0.0001,
      "loss": 1.4145,
      "step": 5839
    },
    {
      "epoch": 0.9377007064868337,
      "grad_norm": 0.2735253870487213,
      "learning_rate": 0.0001,
      "loss": 1.4599,
      "step": 5840
    },
    {
      "epoch": 0.9378612716763006,
      "grad_norm": 0.27882084250450134,
      "learning_rate": 0.0001,
      "loss": 1.5012,
      "step": 5841
    },
    {
      "epoch": 0.9380218368657675,
      "grad_norm": 0.2897747755050659,
      "learning_rate": 0.0001,
      "loss": 1.5606,
      "step": 5842
    },
    {
      "epoch": 0.9381824020552344,
      "grad_norm": 0.28041335940361023,
      "learning_rate": 0.0001,
      "loss": 1.4903,
      "step": 5843
    },
    {
      "epoch": 0.9383429672447013,
      "grad_norm": 0.2811106741428375,
      "learning_rate": 0.0001,
      "loss": 1.4833,
      "step": 5844
    },
    {
      "epoch": 0.9385035324341683,
      "grad_norm": 0.41006171703338623,
      "learning_rate": 0.0001,
      "loss": 1.4621,
      "step": 5845
    },
    {
      "epoch": 0.9386640976236352,
      "grad_norm": 0.2849881052970886,
      "learning_rate": 0.0001,
      "loss": 1.4206,
      "step": 5846
    },
    {
      "epoch": 0.9388246628131022,
      "grad_norm": 0.27344533801078796,
      "learning_rate": 0.0001,
      "loss": 1.45,
      "step": 5847
    },
    {
      "epoch": 0.938985228002569,
      "grad_norm": 0.2861812710762024,
      "learning_rate": 0.0001,
      "loss": 1.396,
      "step": 5848
    },
    {
      "epoch": 0.939145793192036,
      "grad_norm": 0.2781996428966522,
      "learning_rate": 0.0001,
      "loss": 1.5337,
      "step": 5849
    },
    {
      "epoch": 0.9393063583815029,
      "grad_norm": 0.27811169624328613,
      "learning_rate": 0.0001,
      "loss": 1.4811,
      "step": 5850
    },
    {
      "epoch": 0.9394669235709698,
      "grad_norm": 0.26856598258018494,
      "learning_rate": 0.0001,
      "loss": 1.3898,
      "step": 5851
    },
    {
      "epoch": 0.9396274887604368,
      "grad_norm": 0.2990454137325287,
      "learning_rate": 0.0001,
      "loss": 1.4769,
      "step": 5852
    },
    {
      "epoch": 0.9397880539499036,
      "grad_norm": 0.2712068259716034,
      "learning_rate": 0.0001,
      "loss": 1.4573,
      "step": 5853
    },
    {
      "epoch": 0.9399486191393706,
      "grad_norm": 0.2844711244106293,
      "learning_rate": 0.0001,
      "loss": 1.5406,
      "step": 5854
    },
    {
      "epoch": 0.9401091843288375,
      "grad_norm": 0.2788091003894806,
      "learning_rate": 0.0001,
      "loss": 1.5116,
      "step": 5855
    },
    {
      "epoch": 0.9402697495183044,
      "grad_norm": 0.2758863866329193,
      "learning_rate": 0.0001,
      "loss": 1.4672,
      "step": 5856
    },
    {
      "epoch": 0.9404303147077714,
      "grad_norm": 0.28007230162620544,
      "learning_rate": 0.0001,
      "loss": 1.4855,
      "step": 5857
    },
    {
      "epoch": 0.9405908798972383,
      "grad_norm": 0.32679662108421326,
      "learning_rate": 0.0001,
      "loss": 1.4784,
      "step": 5858
    },
    {
      "epoch": 0.9407514450867052,
      "grad_norm": 0.2912706732749939,
      "learning_rate": 0.0001,
      "loss": 1.3704,
      "step": 5859
    },
    {
      "epoch": 0.9409120102761721,
      "grad_norm": 0.27617281675338745,
      "learning_rate": 0.0001,
      "loss": 1.489,
      "step": 5860
    },
    {
      "epoch": 0.941072575465639,
      "grad_norm": 0.3089311718940735,
      "learning_rate": 0.0001,
      "loss": 1.5022,
      "step": 5861
    },
    {
      "epoch": 0.941233140655106,
      "grad_norm": 0.3159063160419464,
      "learning_rate": 0.0001,
      "loss": 1.44,
      "step": 5862
    },
    {
      "epoch": 0.9413937058445729,
      "grad_norm": 0.2852429151535034,
      "learning_rate": 0.0001,
      "loss": 1.4906,
      "step": 5863
    },
    {
      "epoch": 0.9415542710340398,
      "grad_norm": 0.2734900712966919,
      "learning_rate": 0.0001,
      "loss": 1.4781,
      "step": 5864
    },
    {
      "epoch": 0.9417148362235067,
      "grad_norm": 0.28580528497695923,
      "learning_rate": 0.0001,
      "loss": 1.5271,
      "step": 5865
    },
    {
      "epoch": 0.9418754014129737,
      "grad_norm": 0.28720587491989136,
      "learning_rate": 0.0001,
      "loss": 1.4516,
      "step": 5866
    },
    {
      "epoch": 0.9420359666024406,
      "grad_norm": 0.2807033061981201,
      "learning_rate": 0.0001,
      "loss": 1.4719,
      "step": 5867
    },
    {
      "epoch": 0.9421965317919075,
      "grad_norm": 0.29748672246932983,
      "learning_rate": 0.0001,
      "loss": 1.4244,
      "step": 5868
    },
    {
      "epoch": 0.9423570969813745,
      "grad_norm": 0.27877122163772583,
      "learning_rate": 0.0001,
      "loss": 1.5303,
      "step": 5869
    },
    {
      "epoch": 0.9425176621708413,
      "grad_norm": 0.26318901777267456,
      "learning_rate": 0.0001,
      "loss": 1.4143,
      "step": 5870
    },
    {
      "epoch": 0.9426782273603083,
      "grad_norm": 0.2971193492412567,
      "learning_rate": 0.0001,
      "loss": 1.4202,
      "step": 5871
    },
    {
      "epoch": 0.9428387925497752,
      "grad_norm": 0.2822686433792114,
      "learning_rate": 0.0001,
      "loss": 1.4454,
      "step": 5872
    },
    {
      "epoch": 0.9429993577392421,
      "grad_norm": 0.27671289443969727,
      "learning_rate": 0.0001,
      "loss": 1.455,
      "step": 5873
    },
    {
      "epoch": 0.9431599229287091,
      "grad_norm": 0.2803163230419159,
      "learning_rate": 0.0001,
      "loss": 1.4861,
      "step": 5874
    },
    {
      "epoch": 0.9433204881181759,
      "grad_norm": 0.3060593008995056,
      "learning_rate": 0.0001,
      "loss": 1.4971,
      "step": 5875
    },
    {
      "epoch": 0.9434810533076429,
      "grad_norm": 0.27979764342308044,
      "learning_rate": 0.0001,
      "loss": 1.5331,
      "step": 5876
    },
    {
      "epoch": 0.9436416184971098,
      "grad_norm": 0.27894946932792664,
      "learning_rate": 0.0001,
      "loss": 1.522,
      "step": 5877
    },
    {
      "epoch": 0.9438021836865768,
      "grad_norm": 0.28618311882019043,
      "learning_rate": 0.0001,
      "loss": 1.4215,
      "step": 5878
    },
    {
      "epoch": 0.9439627488760437,
      "grad_norm": 0.2913786470890045,
      "learning_rate": 0.0001,
      "loss": 1.5194,
      "step": 5879
    },
    {
      "epoch": 0.9441233140655106,
      "grad_norm": 0.2834182679653168,
      "learning_rate": 0.0001,
      "loss": 1.4395,
      "step": 5880
    },
    {
      "epoch": 0.9442838792549775,
      "grad_norm": 0.29673516750335693,
      "learning_rate": 0.0001,
      "loss": 1.4773,
      "step": 5881
    },
    {
      "epoch": 0.9444444444444444,
      "grad_norm": 0.2880716919898987,
      "learning_rate": 0.0001,
      "loss": 1.4757,
      "step": 5882
    },
    {
      "epoch": 0.9446050096339114,
      "grad_norm": 0.27656540274620056,
      "learning_rate": 0.0001,
      "loss": 1.4813,
      "step": 5883
    },
    {
      "epoch": 0.9447655748233783,
      "grad_norm": 0.259142130613327,
      "learning_rate": 0.0001,
      "loss": 1.4003,
      "step": 5884
    },
    {
      "epoch": 0.9449261400128453,
      "grad_norm": 0.2958175539970398,
      "learning_rate": 0.0001,
      "loss": 1.5299,
      "step": 5885
    },
    {
      "epoch": 0.9450867052023122,
      "grad_norm": 0.2727348208427429,
      "learning_rate": 0.0001,
      "loss": 1.3805,
      "step": 5886
    },
    {
      "epoch": 0.945247270391779,
      "grad_norm": 0.28077074885368347,
      "learning_rate": 0.0001,
      "loss": 1.5182,
      "step": 5887
    },
    {
      "epoch": 0.945407835581246,
      "grad_norm": 0.28943517804145813,
      "learning_rate": 0.0001,
      "loss": 1.4916,
      "step": 5888
    },
    {
      "epoch": 0.9455684007707129,
      "grad_norm": 0.28285327553749084,
      "learning_rate": 0.0001,
      "loss": 1.5546,
      "step": 5889
    },
    {
      "epoch": 0.9457289659601799,
      "grad_norm": 0.2768253982067108,
      "learning_rate": 0.0001,
      "loss": 1.4987,
      "step": 5890
    },
    {
      "epoch": 0.9458895311496468,
      "grad_norm": 0.2773067057132721,
      "learning_rate": 0.0001,
      "loss": 1.4761,
      "step": 5891
    },
    {
      "epoch": 0.9460500963391136,
      "grad_norm": 0.27290916442871094,
      "learning_rate": 0.0001,
      "loss": 1.4429,
      "step": 5892
    },
    {
      "epoch": 0.9462106615285806,
      "grad_norm": 0.2910511791706085,
      "learning_rate": 0.0001,
      "loss": 1.494,
      "step": 5893
    },
    {
      "epoch": 0.9463712267180475,
      "grad_norm": 0.27037549018859863,
      "learning_rate": 0.0001,
      "loss": 1.49,
      "step": 5894
    },
    {
      "epoch": 0.9465317919075145,
      "grad_norm": 0.2779509425163269,
      "learning_rate": 0.0001,
      "loss": 1.4289,
      "step": 5895
    },
    {
      "epoch": 0.9466923570969814,
      "grad_norm": 0.27673134207725525,
      "learning_rate": 0.0001,
      "loss": 1.4547,
      "step": 5896
    },
    {
      "epoch": 0.9468529222864484,
      "grad_norm": 0.2648793160915375,
      "learning_rate": 0.0001,
      "loss": 1.4743,
      "step": 5897
    },
    {
      "epoch": 0.9470134874759152,
      "grad_norm": 0.27191540598869324,
      "learning_rate": 0.0001,
      "loss": 1.5512,
      "step": 5898
    },
    {
      "epoch": 0.9471740526653821,
      "grad_norm": 0.26977187395095825,
      "learning_rate": 0.0001,
      "loss": 1.4684,
      "step": 5899
    },
    {
      "epoch": 0.9473346178548491,
      "grad_norm": 0.273308664560318,
      "learning_rate": 0.0001,
      "loss": 1.4252,
      "step": 5900
    },
    {
      "epoch": 0.947495183044316,
      "grad_norm": 0.27446800470352173,
      "learning_rate": 0.0001,
      "loss": 1.4646,
      "step": 5901
    },
    {
      "epoch": 0.947655748233783,
      "grad_norm": 0.29054462909698486,
      "learning_rate": 0.0001,
      "loss": 1.5228,
      "step": 5902
    },
    {
      "epoch": 0.9478163134232498,
      "grad_norm": 0.27276232838630676,
      "learning_rate": 0.0001,
      "loss": 1.4498,
      "step": 5903
    },
    {
      "epoch": 0.9479768786127167,
      "grad_norm": 0.2760717272758484,
      "learning_rate": 0.0001,
      "loss": 1.4693,
      "step": 5904
    },
    {
      "epoch": 0.9481374438021837,
      "grad_norm": 0.2777491509914398,
      "learning_rate": 0.0001,
      "loss": 1.4673,
      "step": 5905
    },
    {
      "epoch": 0.9482980089916506,
      "grad_norm": 0.2675125300884247,
      "learning_rate": 0.0001,
      "loss": 1.5352,
      "step": 5906
    },
    {
      "epoch": 0.9484585741811176,
      "grad_norm": 0.2754668891429901,
      "learning_rate": 0.0001,
      "loss": 1.4192,
      "step": 5907
    },
    {
      "epoch": 0.9486191393705845,
      "grad_norm": 0.28691086173057556,
      "learning_rate": 0.0001,
      "loss": 1.4273,
      "step": 5908
    },
    {
      "epoch": 0.9487797045600513,
      "grad_norm": 0.2644820809364319,
      "learning_rate": 0.0001,
      "loss": 1.4089,
      "step": 5909
    },
    {
      "epoch": 0.9489402697495183,
      "grad_norm": 0.27131006121635437,
      "learning_rate": 0.0001,
      "loss": 1.4656,
      "step": 5910
    },
    {
      "epoch": 0.9491008349389852,
      "grad_norm": 0.2807711660861969,
      "learning_rate": 0.0001,
      "loss": 1.4128,
      "step": 5911
    },
    {
      "epoch": 0.9492614001284522,
      "grad_norm": 0.26891958713531494,
      "learning_rate": 0.0001,
      "loss": 1.4235,
      "step": 5912
    },
    {
      "epoch": 0.9494219653179191,
      "grad_norm": 0.6112186312675476,
      "learning_rate": 0.0001,
      "loss": 1.4226,
      "step": 5913
    },
    {
      "epoch": 0.949582530507386,
      "grad_norm": 0.2824949026107788,
      "learning_rate": 0.0001,
      "loss": 1.5121,
      "step": 5914
    },
    {
      "epoch": 0.9497430956968529,
      "grad_norm": 0.3157612979412079,
      "learning_rate": 0.0001,
      "loss": 1.4592,
      "step": 5915
    },
    {
      "epoch": 0.9499036608863198,
      "grad_norm": 0.3086928129196167,
      "learning_rate": 0.0001,
      "loss": 1.4821,
      "step": 5916
    },
    {
      "epoch": 0.9500642260757868,
      "grad_norm": 0.26061418652534485,
      "learning_rate": 0.0001,
      "loss": 1.5238,
      "step": 5917
    },
    {
      "epoch": 0.9502247912652537,
      "grad_norm": 0.27031055092811584,
      "learning_rate": 0.0001,
      "loss": 1.386,
      "step": 5918
    },
    {
      "epoch": 0.9503853564547207,
      "grad_norm": 0.28135839104652405,
      "learning_rate": 0.0001,
      "loss": 1.5086,
      "step": 5919
    },
    {
      "epoch": 0.9505459216441875,
      "grad_norm": 0.31137263774871826,
      "learning_rate": 0.0001,
      "loss": 1.4556,
      "step": 5920
    },
    {
      "epoch": 0.9507064868336544,
      "grad_norm": 0.2834000289440155,
      "learning_rate": 0.0001,
      "loss": 1.4445,
      "step": 5921
    },
    {
      "epoch": 0.9508670520231214,
      "grad_norm": 0.27586954832077026,
      "learning_rate": 0.0001,
      "loss": 1.4798,
      "step": 5922
    },
    {
      "epoch": 0.9510276172125883,
      "grad_norm": 0.25340452790260315,
      "learning_rate": 0.0001,
      "loss": 1.3766,
      "step": 5923
    },
    {
      "epoch": 0.9511881824020553,
      "grad_norm": 0.29076021909713745,
      "learning_rate": 0.0001,
      "loss": 1.4756,
      "step": 5924
    },
    {
      "epoch": 0.9513487475915221,
      "grad_norm": 0.2764754295349121,
      "learning_rate": 0.0001,
      "loss": 1.534,
      "step": 5925
    },
    {
      "epoch": 0.951509312780989,
      "grad_norm": 0.28199848532676697,
      "learning_rate": 0.0001,
      "loss": 1.4777,
      "step": 5926
    },
    {
      "epoch": 0.951669877970456,
      "grad_norm": 0.2712484896183014,
      "learning_rate": 0.0001,
      "loss": 1.5353,
      "step": 5927
    },
    {
      "epoch": 0.9518304431599229,
      "grad_norm": 0.29423338174819946,
      "learning_rate": 0.0001,
      "loss": 1.5036,
      "step": 5928
    },
    {
      "epoch": 0.9519910083493899,
      "grad_norm": 0.2705962061882019,
      "learning_rate": 0.0001,
      "loss": 1.3911,
      "step": 5929
    },
    {
      "epoch": 0.9521515735388568,
      "grad_norm": 0.26917755603790283,
      "learning_rate": 0.0001,
      "loss": 1.4662,
      "step": 5930
    },
    {
      "epoch": 0.9523121387283237,
      "grad_norm": 0.30825498700141907,
      "learning_rate": 0.0001,
      "loss": 1.4396,
      "step": 5931
    },
    {
      "epoch": 0.9524727039177906,
      "grad_norm": 0.2915974259376526,
      "learning_rate": 0.0001,
      "loss": 1.4902,
      "step": 5932
    },
    {
      "epoch": 0.9526332691072575,
      "grad_norm": 0.2787433862686157,
      "learning_rate": 0.0001,
      "loss": 1.4666,
      "step": 5933
    },
    {
      "epoch": 0.9527938342967245,
      "grad_norm": 0.2745096981525421,
      "learning_rate": 0.0001,
      "loss": 1.4713,
      "step": 5934
    },
    {
      "epoch": 0.9529543994861914,
      "grad_norm": 0.26666146516799927,
      "learning_rate": 0.0001,
      "loss": 1.4314,
      "step": 5935
    },
    {
      "epoch": 0.9531149646756584,
      "grad_norm": 0.2656130790710449,
      "learning_rate": 0.0001,
      "loss": 1.4568,
      "step": 5936
    },
    {
      "epoch": 0.9532755298651252,
      "grad_norm": 0.2698099613189697,
      "learning_rate": 0.0001,
      "loss": 1.4545,
      "step": 5937
    },
    {
      "epoch": 0.9534360950545921,
      "grad_norm": 0.26967161893844604,
      "learning_rate": 0.0001,
      "loss": 1.4674,
      "step": 5938
    },
    {
      "epoch": 0.9535966602440591,
      "grad_norm": 0.27115342020988464,
      "learning_rate": 0.0001,
      "loss": 1.4547,
      "step": 5939
    },
    {
      "epoch": 0.953757225433526,
      "grad_norm": 0.2653388977050781,
      "learning_rate": 0.0001,
      "loss": 1.4779,
      "step": 5940
    },
    {
      "epoch": 0.953917790622993,
      "grad_norm": 0.2982099950313568,
      "learning_rate": 0.0001,
      "loss": 1.5727,
      "step": 5941
    },
    {
      "epoch": 0.9540783558124598,
      "grad_norm": 0.28064000606536865,
      "learning_rate": 0.0001,
      "loss": 1.4975,
      "step": 5942
    },
    {
      "epoch": 0.9542389210019268,
      "grad_norm": 0.2620096504688263,
      "learning_rate": 0.0001,
      "loss": 1.3736,
      "step": 5943
    },
    {
      "epoch": 0.9543994861913937,
      "grad_norm": 0.3034951686859131,
      "learning_rate": 0.0001,
      "loss": 1.4932,
      "step": 5944
    },
    {
      "epoch": 0.9545600513808606,
      "grad_norm": 0.2841450572013855,
      "learning_rate": 0.0001,
      "loss": 1.42,
      "step": 5945
    },
    {
      "epoch": 0.9547206165703276,
      "grad_norm": 0.9656533598899841,
      "learning_rate": 0.0001,
      "loss": 1.4952,
      "step": 5946
    },
    {
      "epoch": 0.9548811817597945,
      "grad_norm": 0.2824438214302063,
      "learning_rate": 0.0001,
      "loss": 1.4824,
      "step": 5947
    },
    {
      "epoch": 0.9550417469492614,
      "grad_norm": 0.29613274335861206,
      "learning_rate": 0.0001,
      "loss": 1.4816,
      "step": 5948
    },
    {
      "epoch": 0.9552023121387283,
      "grad_norm": 0.32699763774871826,
      "learning_rate": 0.0001,
      "loss": 1.545,
      "step": 5949
    },
    {
      "epoch": 0.9553628773281952,
      "grad_norm": 0.34357208013534546,
      "learning_rate": 0.0001,
      "loss": 1.4593,
      "step": 5950
    },
    {
      "epoch": 0.9555234425176622,
      "grad_norm": 0.2892925441265106,
      "learning_rate": 0.0001,
      "loss": 1.4865,
      "step": 5951
    },
    {
      "epoch": 0.9556840077071291,
      "grad_norm": 0.2864363491535187,
      "learning_rate": 0.0001,
      "loss": 1.5113,
      "step": 5952
    },
    {
      "epoch": 0.955844572896596,
      "grad_norm": 0.2872791886329651,
      "learning_rate": 0.0001,
      "loss": 1.4583,
      "step": 5953
    },
    {
      "epoch": 0.9560051380860629,
      "grad_norm": 0.2911718189716339,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 5954
    },
    {
      "epoch": 0.9561657032755299,
      "grad_norm": 0.28609058260917664,
      "learning_rate": 0.0001,
      "loss": 1.4682,
      "step": 5955
    },
    {
      "epoch": 0.9563262684649968,
      "grad_norm": 0.2892938554286957,
      "learning_rate": 0.0001,
      "loss": 1.5025,
      "step": 5956
    },
    {
      "epoch": 0.9564868336544637,
      "grad_norm": 0.2699921429157257,
      "learning_rate": 0.0001,
      "loss": 1.4165,
      "step": 5957
    },
    {
      "epoch": 0.9566473988439307,
      "grad_norm": 0.27306798100471497,
      "learning_rate": 0.0001,
      "loss": 1.4736,
      "step": 5958
    },
    {
      "epoch": 0.9568079640333975,
      "grad_norm": 0.26198604702949524,
      "learning_rate": 0.0001,
      "loss": 1.348,
      "step": 5959
    },
    {
      "epoch": 0.9569685292228645,
      "grad_norm": 0.2712896466255188,
      "learning_rate": 0.0001,
      "loss": 1.5062,
      "step": 5960
    },
    {
      "epoch": 0.9571290944123314,
      "grad_norm": 0.2693850100040436,
      "learning_rate": 0.0001,
      "loss": 1.4626,
      "step": 5961
    },
    {
      "epoch": 0.9572896596017983,
      "grad_norm": 0.28435471653938293,
      "learning_rate": 0.0001,
      "loss": 1.4935,
      "step": 5962
    },
    {
      "epoch": 0.9574502247912653,
      "grad_norm": 0.280477374792099,
      "learning_rate": 0.0001,
      "loss": 1.4133,
      "step": 5963
    },
    {
      "epoch": 0.9576107899807321,
      "grad_norm": 0.28667399287223816,
      "learning_rate": 0.0001,
      "loss": 1.4673,
      "step": 5964
    },
    {
      "epoch": 0.9577713551701991,
      "grad_norm": 0.2773427367210388,
      "learning_rate": 0.0001,
      "loss": 1.4987,
      "step": 5965
    },
    {
      "epoch": 0.957931920359666,
      "grad_norm": 0.25841793417930603,
      "learning_rate": 0.0001,
      "loss": 1.4617,
      "step": 5966
    },
    {
      "epoch": 0.958092485549133,
      "grad_norm": 0.2560836970806122,
      "learning_rate": 0.0001,
      "loss": 1.4402,
      "step": 5967
    },
    {
      "epoch": 0.9582530507385999,
      "grad_norm": 0.2719283998012543,
      "learning_rate": 0.0001,
      "loss": 1.4475,
      "step": 5968
    },
    {
      "epoch": 0.9584136159280668,
      "grad_norm": 0.26865220069885254,
      "learning_rate": 0.0001,
      "loss": 1.4518,
      "step": 5969
    },
    {
      "epoch": 0.9585741811175337,
      "grad_norm": 0.2628419101238251,
      "learning_rate": 0.0001,
      "loss": 1.514,
      "step": 5970
    },
    {
      "epoch": 0.9587347463070006,
      "grad_norm": 0.2671055495738983,
      "learning_rate": 0.0001,
      "loss": 1.5483,
      "step": 5971
    },
    {
      "epoch": 0.9588953114964676,
      "grad_norm": 0.28136101365089417,
      "learning_rate": 0.0001,
      "loss": 1.5116,
      "step": 5972
    },
    {
      "epoch": 0.9590558766859345,
      "grad_norm": 0.27849438786506653,
      "learning_rate": 0.0001,
      "loss": 1.5061,
      "step": 5973
    },
    {
      "epoch": 0.9592164418754014,
      "grad_norm": 0.2710857689380646,
      "learning_rate": 0.0001,
      "loss": 1.3976,
      "step": 5974
    },
    {
      "epoch": 0.9593770070648683,
      "grad_norm": 0.269164115190506,
      "learning_rate": 0.0001,
      "loss": 1.4451,
      "step": 5975
    },
    {
      "epoch": 0.9595375722543352,
      "grad_norm": 0.2771841287612915,
      "learning_rate": 0.0001,
      "loss": 1.4537,
      "step": 5976
    },
    {
      "epoch": 0.9596981374438022,
      "grad_norm": 0.2635327875614166,
      "learning_rate": 0.0001,
      "loss": 1.4565,
      "step": 5977
    },
    {
      "epoch": 0.9598587026332691,
      "grad_norm": 0.2794688045978546,
      "learning_rate": 0.0001,
      "loss": 1.4913,
      "step": 5978
    },
    {
      "epoch": 0.960019267822736,
      "grad_norm": 0.286764919757843,
      "learning_rate": 0.0001,
      "loss": 1.4734,
      "step": 5979
    },
    {
      "epoch": 0.960179833012203,
      "grad_norm": 0.2917470335960388,
      "learning_rate": 0.0001,
      "loss": 1.4634,
      "step": 5980
    },
    {
      "epoch": 0.9603403982016698,
      "grad_norm": 0.27942582964897156,
      "learning_rate": 0.0001,
      "loss": 1.5373,
      "step": 5981
    },
    {
      "epoch": 0.9605009633911368,
      "grad_norm": 0.27506059408187866,
      "learning_rate": 0.0001,
      "loss": 1.4676,
      "step": 5982
    },
    {
      "epoch": 0.9606615285806037,
      "grad_norm": 0.2907000184059143,
      "learning_rate": 0.0001,
      "loss": 1.4463,
      "step": 5983
    },
    {
      "epoch": 0.9608220937700707,
      "grad_norm": 0.27734360098838806,
      "learning_rate": 0.0001,
      "loss": 1.4884,
      "step": 5984
    },
    {
      "epoch": 0.9609826589595376,
      "grad_norm": 0.29087257385253906,
      "learning_rate": 0.0001,
      "loss": 1.4645,
      "step": 5985
    },
    {
      "epoch": 0.9611432241490045,
      "grad_norm": 0.26189374923706055,
      "learning_rate": 0.0001,
      "loss": 1.3998,
      "step": 5986
    },
    {
      "epoch": 0.9613037893384714,
      "grad_norm": 0.2801060080528259,
      "learning_rate": 0.0001,
      "loss": 1.4756,
      "step": 5987
    },
    {
      "epoch": 0.9614643545279383,
      "grad_norm": 0.2744179964065552,
      "learning_rate": 0.0001,
      "loss": 1.465,
      "step": 5988
    },
    {
      "epoch": 0.9616249197174053,
      "grad_norm": 0.28119373321533203,
      "learning_rate": 0.0001,
      "loss": 1.472,
      "step": 5989
    },
    {
      "epoch": 0.9617854849068722,
      "grad_norm": 0.2790205776691437,
      "learning_rate": 0.0001,
      "loss": 1.4537,
      "step": 5990
    },
    {
      "epoch": 0.9619460500963392,
      "grad_norm": 0.27779579162597656,
      "learning_rate": 0.0001,
      "loss": 1.4187,
      "step": 5991
    },
    {
      "epoch": 0.962106615285806,
      "grad_norm": 0.29540571570396423,
      "learning_rate": 0.0001,
      "loss": 1.4923,
      "step": 5992
    },
    {
      "epoch": 0.9622671804752729,
      "grad_norm": 0.2819844186306,
      "learning_rate": 0.0001,
      "loss": 1.4482,
      "step": 5993
    },
    {
      "epoch": 0.9624277456647399,
      "grad_norm": 0.27781808376312256,
      "learning_rate": 0.0001,
      "loss": 1.3551,
      "step": 5994
    },
    {
      "epoch": 0.9625883108542068,
      "grad_norm": 0.26764678955078125,
      "learning_rate": 0.0001,
      "loss": 1.3977,
      "step": 5995
    },
    {
      "epoch": 0.9627488760436738,
      "grad_norm": 0.2768016755580902,
      "learning_rate": 0.0001,
      "loss": 1.5034,
      "step": 5996
    },
    {
      "epoch": 0.9629094412331407,
      "grad_norm": 0.28940582275390625,
      "learning_rate": 0.0001,
      "loss": 1.4729,
      "step": 5997
    },
    {
      "epoch": 0.9630700064226075,
      "grad_norm": 0.27764105796813965,
      "learning_rate": 0.0001,
      "loss": 1.484,
      "step": 5998
    },
    {
      "epoch": 0.9632305716120745,
      "grad_norm": 0.2838978171348572,
      "learning_rate": 0.0001,
      "loss": 1.3841,
      "step": 5999
    },
    {
      "epoch": 0.9633911368015414,
      "grad_norm": 0.29951396584510803,
      "learning_rate": 0.0001,
      "loss": 1.4491,
      "step": 6000
    },
    {
      "epoch": 0.9635517019910084,
      "grad_norm": 0.2669565677642822,
      "learning_rate": 0.0001,
      "loss": 1.4029,
      "step": 6001
    },
    {
      "epoch": 0.9637122671804753,
      "grad_norm": 0.3049006164073944,
      "learning_rate": 0.0001,
      "loss": 1.4756,
      "step": 6002
    },
    {
      "epoch": 0.9638728323699421,
      "grad_norm": 0.27438488602638245,
      "learning_rate": 0.0001,
      "loss": 1.5432,
      "step": 6003
    },
    {
      "epoch": 0.9640333975594091,
      "grad_norm": 0.26807013154029846,
      "learning_rate": 0.0001,
      "loss": 1.4198,
      "step": 6004
    },
    {
      "epoch": 0.964193962748876,
      "grad_norm": 0.27279675006866455,
      "learning_rate": 0.0001,
      "loss": 1.5018,
      "step": 6005
    },
    {
      "epoch": 0.964354527938343,
      "grad_norm": 0.2722545564174652,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 6006
    },
    {
      "epoch": 0.9645150931278099,
      "grad_norm": 0.276894748210907,
      "learning_rate": 0.0001,
      "loss": 1.5087,
      "step": 6007
    },
    {
      "epoch": 0.9646756583172769,
      "grad_norm": 0.272074818611145,
      "learning_rate": 0.0001,
      "loss": 1.4721,
      "step": 6008
    },
    {
      "epoch": 0.9648362235067437,
      "grad_norm": 0.2679750621318817,
      "learning_rate": 0.0001,
      "loss": 1.4397,
      "step": 6009
    },
    {
      "epoch": 0.9649967886962106,
      "grad_norm": 0.26656609773635864,
      "learning_rate": 0.0001,
      "loss": 1.4337,
      "step": 6010
    },
    {
      "epoch": 0.9651573538856776,
      "grad_norm": 0.2664980888366699,
      "learning_rate": 0.0001,
      "loss": 1.4173,
      "step": 6011
    },
    {
      "epoch": 0.9653179190751445,
      "grad_norm": 0.2975117862224579,
      "learning_rate": 0.0001,
      "loss": 1.4243,
      "step": 6012
    },
    {
      "epoch": 0.9654784842646115,
      "grad_norm": 0.2981017529964447,
      "learning_rate": 0.0001,
      "loss": 1.463,
      "step": 6013
    },
    {
      "epoch": 0.9656390494540783,
      "grad_norm": 0.27719423174858093,
      "learning_rate": 0.0001,
      "loss": 1.4027,
      "step": 6014
    },
    {
      "epoch": 0.9657996146435452,
      "grad_norm": 0.2824347913265228,
      "learning_rate": 0.0001,
      "loss": 1.3975,
      "step": 6015
    },
    {
      "epoch": 0.9659601798330122,
      "grad_norm": 0.28527072072029114,
      "learning_rate": 0.0001,
      "loss": 1.44,
      "step": 6016
    },
    {
      "epoch": 0.9661207450224791,
      "grad_norm": 0.26541316509246826,
      "learning_rate": 0.0001,
      "loss": 1.4369,
      "step": 6017
    },
    {
      "epoch": 0.9662813102119461,
      "grad_norm": 0.3047367334365845,
      "learning_rate": 0.0001,
      "loss": 1.5166,
      "step": 6018
    },
    {
      "epoch": 0.966441875401413,
      "grad_norm": 0.2990919351577759,
      "learning_rate": 0.0001,
      "loss": 1.4694,
      "step": 6019
    },
    {
      "epoch": 0.9666024405908799,
      "grad_norm": 0.27594760060310364,
      "learning_rate": 0.0001,
      "loss": 1.4195,
      "step": 6020
    },
    {
      "epoch": 0.9667630057803468,
      "grad_norm": 0.28485724329948425,
      "learning_rate": 0.0001,
      "loss": 1.5256,
      "step": 6021
    },
    {
      "epoch": 0.9669235709698137,
      "grad_norm": 0.296840637922287,
      "learning_rate": 0.0001,
      "loss": 1.4589,
      "step": 6022
    },
    {
      "epoch": 0.9670841361592807,
      "grad_norm": 0.27788132429122925,
      "learning_rate": 0.0001,
      "loss": 1.4796,
      "step": 6023
    },
    {
      "epoch": 0.9672447013487476,
      "grad_norm": 0.27840691804885864,
      "learning_rate": 0.0001,
      "loss": 1.4823,
      "step": 6024
    },
    {
      "epoch": 0.9674052665382146,
      "grad_norm": 0.29735177755355835,
      "learning_rate": 0.0001,
      "loss": 1.3861,
      "step": 6025
    },
    {
      "epoch": 0.9675658317276814,
      "grad_norm": 0.27756813168525696,
      "learning_rate": 0.0001,
      "loss": 1.4538,
      "step": 6026
    },
    {
      "epoch": 0.9677263969171483,
      "grad_norm": 0.27351388335227966,
      "learning_rate": 0.0001,
      "loss": 1.4276,
      "step": 6027
    },
    {
      "epoch": 0.9678869621066153,
      "grad_norm": 0.27169686555862427,
      "learning_rate": 0.0001,
      "loss": 1.4239,
      "step": 6028
    },
    {
      "epoch": 0.9680475272960822,
      "grad_norm": 0.3168017268180847,
      "learning_rate": 0.0001,
      "loss": 1.4175,
      "step": 6029
    },
    {
      "epoch": 0.9682080924855492,
      "grad_norm": 0.264495313167572,
      "learning_rate": 0.0001,
      "loss": 1.4514,
      "step": 6030
    },
    {
      "epoch": 0.968368657675016,
      "grad_norm": 0.26328355073928833,
      "learning_rate": 0.0001,
      "loss": 1.3986,
      "step": 6031
    },
    {
      "epoch": 0.968529222864483,
      "grad_norm": 0.2844201326370239,
      "learning_rate": 0.0001,
      "loss": 1.4448,
      "step": 6032
    },
    {
      "epoch": 0.9686897880539499,
      "grad_norm": 0.27860555052757263,
      "learning_rate": 0.0001,
      "loss": 1.4985,
      "step": 6033
    },
    {
      "epoch": 0.9688503532434168,
      "grad_norm": 0.27706608176231384,
      "learning_rate": 0.0001,
      "loss": 1.4341,
      "step": 6034
    },
    {
      "epoch": 0.9690109184328838,
      "grad_norm": 0.3051508665084839,
      "learning_rate": 0.0001,
      "loss": 1.457,
      "step": 6035
    },
    {
      "epoch": 0.9691714836223507,
      "grad_norm": 0.28054526448249817,
      "learning_rate": 0.0001,
      "loss": 1.449,
      "step": 6036
    },
    {
      "epoch": 0.9693320488118176,
      "grad_norm": 0.26974231004714966,
      "learning_rate": 0.0001,
      "loss": 1.4576,
      "step": 6037
    },
    {
      "epoch": 0.9694926140012845,
      "grad_norm": 0.2820809781551361,
      "learning_rate": 0.0001,
      "loss": 1.4367,
      "step": 6038
    },
    {
      "epoch": 0.9696531791907514,
      "grad_norm": 0.2799365222454071,
      "learning_rate": 0.0001,
      "loss": 1.5023,
      "step": 6039
    },
    {
      "epoch": 0.9698137443802184,
      "grad_norm": 0.2809252142906189,
      "learning_rate": 0.0001,
      "loss": 1.4693,
      "step": 6040
    },
    {
      "epoch": 0.9699743095696853,
      "grad_norm": 0.27401256561279297,
      "learning_rate": 0.0001,
      "loss": 1.5606,
      "step": 6041
    },
    {
      "epoch": 0.9701348747591522,
      "grad_norm": 0.27618643641471863,
      "learning_rate": 0.0001,
      "loss": 1.4711,
      "step": 6042
    },
    {
      "epoch": 0.9702954399486191,
      "grad_norm": 0.2657073140144348,
      "learning_rate": 0.0001,
      "loss": 1.4804,
      "step": 6043
    },
    {
      "epoch": 0.970456005138086,
      "grad_norm": 0.26313135027885437,
      "learning_rate": 0.0001,
      "loss": 1.4401,
      "step": 6044
    },
    {
      "epoch": 0.970616570327553,
      "grad_norm": 0.790553629398346,
      "learning_rate": 0.0001,
      "loss": 1.465,
      "step": 6045
    },
    {
      "epoch": 0.9707771355170199,
      "grad_norm": 0.28462934494018555,
      "learning_rate": 0.0001,
      "loss": 1.4823,
      "step": 6046
    },
    {
      "epoch": 0.9709377007064869,
      "grad_norm": 0.29648351669311523,
      "learning_rate": 0.0001,
      "loss": 1.3966,
      "step": 6047
    },
    {
      "epoch": 0.9710982658959537,
      "grad_norm": 0.2969985604286194,
      "learning_rate": 0.0001,
      "loss": 1.5062,
      "step": 6048
    },
    {
      "epoch": 0.9712588310854207,
      "grad_norm": 0.2779256999492645,
      "learning_rate": 0.0001,
      "loss": 1.5596,
      "step": 6049
    },
    {
      "epoch": 0.9714193962748876,
      "grad_norm": 0.2802236080169678,
      "learning_rate": 0.0001,
      "loss": 1.5852,
      "step": 6050
    },
    {
      "epoch": 0.9715799614643545,
      "grad_norm": 0.27366623282432556,
      "learning_rate": 0.0001,
      "loss": 1.5129,
      "step": 6051
    },
    {
      "epoch": 0.9717405266538215,
      "grad_norm": 0.2757594585418701,
      "learning_rate": 0.0001,
      "loss": 1.5257,
      "step": 6052
    },
    {
      "epoch": 0.9719010918432883,
      "grad_norm": 0.28436192870140076,
      "learning_rate": 0.0001,
      "loss": 1.4486,
      "step": 6053
    },
    {
      "epoch": 0.9720616570327553,
      "grad_norm": 0.2683652341365814,
      "learning_rate": 0.0001,
      "loss": 1.4946,
      "step": 6054
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 0.2610686123371124,
      "learning_rate": 0.0001,
      "loss": 1.4334,
      "step": 6055
    },
    {
      "epoch": 0.9723827874116892,
      "grad_norm": 0.28330203890800476,
      "learning_rate": 0.0001,
      "loss": 1.4521,
      "step": 6056
    },
    {
      "epoch": 0.9725433526011561,
      "grad_norm": 0.3239646852016449,
      "learning_rate": 0.0001,
      "loss": 1.4915,
      "step": 6057
    },
    {
      "epoch": 0.972703917790623,
      "grad_norm": 0.27439191937446594,
      "learning_rate": 0.0001,
      "loss": 1.4892,
      "step": 6058
    },
    {
      "epoch": 0.9728644829800899,
      "grad_norm": 0.266892671585083,
      "learning_rate": 0.0001,
      "loss": 1.4453,
      "step": 6059
    },
    {
      "epoch": 0.9730250481695568,
      "grad_norm": 0.2938152849674225,
      "learning_rate": 0.0001,
      "loss": 1.4933,
      "step": 6060
    },
    {
      "epoch": 0.9731856133590238,
      "grad_norm": 0.2957319915294647,
      "learning_rate": 0.0001,
      "loss": 1.5158,
      "step": 6061
    },
    {
      "epoch": 0.9733461785484907,
      "grad_norm": 0.27349087595939636,
      "learning_rate": 0.0001,
      "loss": 1.5462,
      "step": 6062
    },
    {
      "epoch": 0.9735067437379576,
      "grad_norm": 0.28939199447631836,
      "learning_rate": 0.0001,
      "loss": 1.4771,
      "step": 6063
    },
    {
      "epoch": 0.9736673089274245,
      "grad_norm": 0.2705721855163574,
      "learning_rate": 0.0001,
      "loss": 1.416,
      "step": 6064
    },
    {
      "epoch": 0.9738278741168914,
      "grad_norm": 0.2997393012046814,
      "learning_rate": 0.0001,
      "loss": 1.5235,
      "step": 6065
    },
    {
      "epoch": 0.9739884393063584,
      "grad_norm": 0.2901417315006256,
      "learning_rate": 0.0001,
      "loss": 1.382,
      "step": 6066
    },
    {
      "epoch": 0.9741490044958253,
      "grad_norm": 0.284340500831604,
      "learning_rate": 0.0001,
      "loss": 1.4807,
      "step": 6067
    },
    {
      "epoch": 0.9743095696852923,
      "grad_norm": 0.2745599150657654,
      "learning_rate": 0.0001,
      "loss": 1.3844,
      "step": 6068
    },
    {
      "epoch": 0.9744701348747592,
      "grad_norm": 0.29601961374282837,
      "learning_rate": 0.0001,
      "loss": 1.5104,
      "step": 6069
    },
    {
      "epoch": 0.974630700064226,
      "grad_norm": 0.29848021268844604,
      "learning_rate": 0.0001,
      "loss": 1.5022,
      "step": 6070
    },
    {
      "epoch": 0.974791265253693,
      "grad_norm": 0.26561540365219116,
      "learning_rate": 0.0001,
      "loss": 1.4765,
      "step": 6071
    },
    {
      "epoch": 0.9749518304431599,
      "grad_norm": 0.2823335826396942,
      "learning_rate": 0.0001,
      "loss": 1.4665,
      "step": 6072
    },
    {
      "epoch": 0.9751123956326269,
      "grad_norm": 0.28383684158325195,
      "learning_rate": 0.0001,
      "loss": 1.5105,
      "step": 6073
    },
    {
      "epoch": 0.9752729608220938,
      "grad_norm": 0.2845028042793274,
      "learning_rate": 0.0001,
      "loss": 1.4474,
      "step": 6074
    },
    {
      "epoch": 0.9754335260115607,
      "grad_norm": 0.2673282027244568,
      "learning_rate": 0.0001,
      "loss": 1.4024,
      "step": 6075
    },
    {
      "epoch": 0.9755940912010276,
      "grad_norm": 0.2782292664051056,
      "learning_rate": 0.0001,
      "loss": 1.4336,
      "step": 6076
    },
    {
      "epoch": 0.9757546563904945,
      "grad_norm": 0.2785802185535431,
      "learning_rate": 0.0001,
      "loss": 1.4614,
      "step": 6077
    },
    {
      "epoch": 0.9759152215799615,
      "grad_norm": 0.26905113458633423,
      "learning_rate": 0.0001,
      "loss": 1.3642,
      "step": 6078
    },
    {
      "epoch": 0.9760757867694284,
      "grad_norm": 0.2865033447742462,
      "learning_rate": 0.0001,
      "loss": 1.4281,
      "step": 6079
    },
    {
      "epoch": 0.9762363519588954,
      "grad_norm": 0.2800520956516266,
      "learning_rate": 0.0001,
      "loss": 1.4609,
      "step": 6080
    },
    {
      "epoch": 0.9763969171483622,
      "grad_norm": 0.3049313426017761,
      "learning_rate": 0.0001,
      "loss": 1.5001,
      "step": 6081
    },
    {
      "epoch": 0.9765574823378291,
      "grad_norm": 0.28505152463912964,
      "learning_rate": 0.0001,
      "loss": 1.4622,
      "step": 6082
    },
    {
      "epoch": 0.9767180475272961,
      "grad_norm": 0.2728041410446167,
      "learning_rate": 0.0001,
      "loss": 1.4396,
      "step": 6083
    },
    {
      "epoch": 0.976878612716763,
      "grad_norm": 0.27396929264068604,
      "learning_rate": 0.0001,
      "loss": 1.4796,
      "step": 6084
    },
    {
      "epoch": 0.97703917790623,
      "grad_norm": 0.5953158140182495,
      "learning_rate": 0.0001,
      "loss": 1.3769,
      "step": 6085
    },
    {
      "epoch": 0.9771997430956969,
      "grad_norm": 0.2573602497577667,
      "learning_rate": 0.0001,
      "loss": 1.4832,
      "step": 6086
    },
    {
      "epoch": 0.9773603082851637,
      "grad_norm": 0.2817201614379883,
      "learning_rate": 0.0001,
      "loss": 1.5091,
      "step": 6087
    },
    {
      "epoch": 0.9775208734746307,
      "grad_norm": 0.2709033489227295,
      "learning_rate": 0.0001,
      "loss": 1.4692,
      "step": 6088
    },
    {
      "epoch": 0.9776814386640976,
      "grad_norm": 0.30877017974853516,
      "learning_rate": 0.0001,
      "loss": 1.4484,
      "step": 6089
    },
    {
      "epoch": 0.9778420038535646,
      "grad_norm": 0.26173847913742065,
      "learning_rate": 0.0001,
      "loss": 1.4751,
      "step": 6090
    },
    {
      "epoch": 0.9780025690430315,
      "grad_norm": 0.2807842791080475,
      "learning_rate": 0.0001,
      "loss": 1.4638,
      "step": 6091
    },
    {
      "epoch": 0.9781631342324983,
      "grad_norm": 0.2715712785720825,
      "learning_rate": 0.0001,
      "loss": 1.4712,
      "step": 6092
    },
    {
      "epoch": 0.9783236994219653,
      "grad_norm": 0.2671475410461426,
      "learning_rate": 0.0001,
      "loss": 1.4161,
      "step": 6093
    },
    {
      "epoch": 0.9784842646114322,
      "grad_norm": 0.2942464053630829,
      "learning_rate": 0.0001,
      "loss": 1.4988,
      "step": 6094
    },
    {
      "epoch": 0.9786448298008992,
      "grad_norm": 0.2681715190410614,
      "learning_rate": 0.0001,
      "loss": 1.4191,
      "step": 6095
    },
    {
      "epoch": 0.9788053949903661,
      "grad_norm": 0.2870206832885742,
      "learning_rate": 0.0001,
      "loss": 1.3867,
      "step": 6096
    },
    {
      "epoch": 0.9789659601798331,
      "grad_norm": 0.2776786983013153,
      "learning_rate": 0.0001,
      "loss": 1.4712,
      "step": 6097
    },
    {
      "epoch": 0.9791265253692999,
      "grad_norm": 0.27417871356010437,
      "learning_rate": 0.0001,
      "loss": 1.4053,
      "step": 6098
    },
    {
      "epoch": 0.9792870905587668,
      "grad_norm": 0.2741701006889343,
      "learning_rate": 0.0001,
      "loss": 1.5258,
      "step": 6099
    },
    {
      "epoch": 0.9794476557482338,
      "grad_norm": 0.2945915758609772,
      "learning_rate": 0.0001,
      "loss": 1.549,
      "step": 6100
    },
    {
      "epoch": 0.9796082209377007,
      "grad_norm": 0.2797362208366394,
      "learning_rate": 0.0001,
      "loss": 1.5345,
      "step": 6101
    },
    {
      "epoch": 0.9797687861271677,
      "grad_norm": 0.2899368107318878,
      "learning_rate": 0.0001,
      "loss": 1.4929,
      "step": 6102
    },
    {
      "epoch": 0.9799293513166345,
      "grad_norm": 0.28004252910614014,
      "learning_rate": 0.0001,
      "loss": 1.4797,
      "step": 6103
    },
    {
      "epoch": 0.9800899165061014,
      "grad_norm": 0.3024308681488037,
      "learning_rate": 0.0001,
      "loss": 1.4515,
      "step": 6104
    },
    {
      "epoch": 0.9802504816955684,
      "grad_norm": 0.30811381340026855,
      "learning_rate": 0.0001,
      "loss": 1.4288,
      "step": 6105
    },
    {
      "epoch": 0.9804110468850353,
      "grad_norm": 0.29313984513282776,
      "learning_rate": 0.0001,
      "loss": 1.5526,
      "step": 6106
    },
    {
      "epoch": 0.9805716120745023,
      "grad_norm": 0.25765109062194824,
      "learning_rate": 0.0001,
      "loss": 1.4112,
      "step": 6107
    },
    {
      "epoch": 0.9807321772639692,
      "grad_norm": 0.2743788957595825,
      "learning_rate": 0.0001,
      "loss": 1.4229,
      "step": 6108
    },
    {
      "epoch": 0.980892742453436,
      "grad_norm": 0.28418049216270447,
      "learning_rate": 0.0001,
      "loss": 1.5385,
      "step": 6109
    },
    {
      "epoch": 0.981053307642903,
      "grad_norm": 0.2828102707862854,
      "learning_rate": 0.0001,
      "loss": 1.5066,
      "step": 6110
    },
    {
      "epoch": 0.9812138728323699,
      "grad_norm": 0.27831506729125977,
      "learning_rate": 0.0001,
      "loss": 1.4488,
      "step": 6111
    },
    {
      "epoch": 0.9813744380218369,
      "grad_norm": 0.2679727375507355,
      "learning_rate": 0.0001,
      "loss": 1.5022,
      "step": 6112
    },
    {
      "epoch": 0.9815350032113038,
      "grad_norm": 0.2713260054588318,
      "learning_rate": 0.0001,
      "loss": 1.4424,
      "step": 6113
    },
    {
      "epoch": 0.9816955684007708,
      "grad_norm": 0.2776862382888794,
      "learning_rate": 0.0001,
      "loss": 1.5187,
      "step": 6114
    },
    {
      "epoch": 0.9818561335902376,
      "grad_norm": 0.2952801585197449,
      "learning_rate": 0.0001,
      "loss": 1.5083,
      "step": 6115
    },
    {
      "epoch": 0.9820166987797045,
      "grad_norm": 0.3064367175102234,
      "learning_rate": 0.0001,
      "loss": 1.4903,
      "step": 6116
    },
    {
      "epoch": 0.9821772639691715,
      "grad_norm": 0.2772613763809204,
      "learning_rate": 0.0001,
      "loss": 1.4656,
      "step": 6117
    },
    {
      "epoch": 0.9823378291586384,
      "grad_norm": 0.283323734998703,
      "learning_rate": 0.0001,
      "loss": 1.463,
      "step": 6118
    },
    {
      "epoch": 0.9824983943481054,
      "grad_norm": 0.3010428547859192,
      "learning_rate": 0.0001,
      "loss": 1.5127,
      "step": 6119
    },
    {
      "epoch": 0.9826589595375722,
      "grad_norm": 0.27567583322525024,
      "learning_rate": 0.0001,
      "loss": 1.4729,
      "step": 6120
    },
    {
      "epoch": 0.9828195247270392,
      "grad_norm": 0.26461541652679443,
      "learning_rate": 0.0001,
      "loss": 1.4573,
      "step": 6121
    },
    {
      "epoch": 0.9829800899165061,
      "grad_norm": 0.2829727232456207,
      "learning_rate": 0.0001,
      "loss": 1.484,
      "step": 6122
    },
    {
      "epoch": 0.983140655105973,
      "grad_norm": 0.2693321108818054,
      "learning_rate": 0.0001,
      "loss": 1.4139,
      "step": 6123
    },
    {
      "epoch": 0.98330122029544,
      "grad_norm": 0.29115068912506104,
      "learning_rate": 0.0001,
      "loss": 1.5082,
      "step": 6124
    },
    {
      "epoch": 0.9834617854849069,
      "grad_norm": 0.28654834628105164,
      "learning_rate": 0.0001,
      "loss": 1.4419,
      "step": 6125
    },
    {
      "epoch": 0.9836223506743738,
      "grad_norm": 0.28508302569389343,
      "learning_rate": 0.0001,
      "loss": 1.4822,
      "step": 6126
    },
    {
      "epoch": 0.9837829158638407,
      "grad_norm": 0.2751412093639374,
      "learning_rate": 0.0001,
      "loss": 1.4292,
      "step": 6127
    },
    {
      "epoch": 0.9839434810533076,
      "grad_norm": 0.2865232229232788,
      "learning_rate": 0.0001,
      "loss": 1.544,
      "step": 6128
    },
    {
      "epoch": 0.9841040462427746,
      "grad_norm": 0.2914300858974457,
      "learning_rate": 0.0001,
      "loss": 1.5164,
      "step": 6129
    },
    {
      "epoch": 0.9842646114322415,
      "grad_norm": 0.28615498542785645,
      "learning_rate": 0.0001,
      "loss": 1.4364,
      "step": 6130
    },
    {
      "epoch": 0.9844251766217084,
      "grad_norm": 0.28262922167778015,
      "learning_rate": 0.0001,
      "loss": 1.4928,
      "step": 6131
    },
    {
      "epoch": 0.9845857418111753,
      "grad_norm": 0.27573537826538086,
      "learning_rate": 0.0001,
      "loss": 1.4675,
      "step": 6132
    },
    {
      "epoch": 0.9847463070006423,
      "grad_norm": 0.2692889869213104,
      "learning_rate": 0.0001,
      "loss": 1.4212,
      "step": 6133
    },
    {
      "epoch": 0.9849068721901092,
      "grad_norm": 0.2852250039577484,
      "learning_rate": 0.0001,
      "loss": 1.5296,
      "step": 6134
    },
    {
      "epoch": 0.9850674373795761,
      "grad_norm": 0.2724355161190033,
      "learning_rate": 0.0001,
      "loss": 1.4882,
      "step": 6135
    },
    {
      "epoch": 0.9852280025690431,
      "grad_norm": 0.29830068349838257,
      "learning_rate": 0.0001,
      "loss": 1.4987,
      "step": 6136
    },
    {
      "epoch": 0.9853885677585099,
      "grad_norm": 0.29177263379096985,
      "learning_rate": 0.0001,
      "loss": 1.6092,
      "step": 6137
    },
    {
      "epoch": 0.9855491329479769,
      "grad_norm": 0.2806275486946106,
      "learning_rate": 0.0001,
      "loss": 1.4853,
      "step": 6138
    },
    {
      "epoch": 0.9857096981374438,
      "grad_norm": 0.29522383213043213,
      "learning_rate": 0.0001,
      "loss": 1.4888,
      "step": 6139
    },
    {
      "epoch": 0.9858702633269107,
      "grad_norm": 0.2776612639427185,
      "learning_rate": 0.0001,
      "loss": 1.412,
      "step": 6140
    },
    {
      "epoch": 0.9860308285163777,
      "grad_norm": 0.2890492081642151,
      "learning_rate": 0.0001,
      "loss": 1.4377,
      "step": 6141
    },
    {
      "epoch": 0.9861913937058445,
      "grad_norm": 0.2801278829574585,
      "learning_rate": 0.0001,
      "loss": 1.4832,
      "step": 6142
    },
    {
      "epoch": 0.9863519588953115,
      "grad_norm": 0.2889750599861145,
      "learning_rate": 0.0001,
      "loss": 1.501,
      "step": 6143
    },
    {
      "epoch": 0.9865125240847784,
      "grad_norm": 0.2753765285015106,
      "learning_rate": 0.0001,
      "loss": 1.3762,
      "step": 6144
    },
    {
      "epoch": 0.9866730892742454,
      "grad_norm": 0.2816472053527832,
      "learning_rate": 0.0001,
      "loss": 1.517,
      "step": 6145
    },
    {
      "epoch": 0.9868336544637123,
      "grad_norm": 0.2731228172779083,
      "learning_rate": 0.0001,
      "loss": 1.4484,
      "step": 6146
    },
    {
      "epoch": 0.9869942196531792,
      "grad_norm": 0.3136705160140991,
      "learning_rate": 0.0001,
      "loss": 1.458,
      "step": 6147
    },
    {
      "epoch": 0.9871547848426461,
      "grad_norm": 0.26373156905174255,
      "learning_rate": 0.0001,
      "loss": 1.4254,
      "step": 6148
    },
    {
      "epoch": 0.987315350032113,
      "grad_norm": 0.28131967782974243,
      "learning_rate": 0.0001,
      "loss": 1.4722,
      "step": 6149
    },
    {
      "epoch": 0.98747591522158,
      "grad_norm": 0.2643895447254181,
      "learning_rate": 0.0001,
      "loss": 1.2916,
      "step": 6150
    },
    {
      "epoch": 0.9876364804110469,
      "grad_norm": 0.26116758584976196,
      "learning_rate": 0.0001,
      "loss": 1.4171,
      "step": 6151
    },
    {
      "epoch": 0.9877970456005138,
      "grad_norm": 0.2842550277709961,
      "learning_rate": 0.0001,
      "loss": 1.5123,
      "step": 6152
    },
    {
      "epoch": 0.9879576107899807,
      "grad_norm": 0.28289249539375305,
      "learning_rate": 0.0001,
      "loss": 1.4759,
      "step": 6153
    },
    {
      "epoch": 0.9881181759794476,
      "grad_norm": 0.2961236536502838,
      "learning_rate": 0.0001,
      "loss": 1.5121,
      "step": 6154
    },
    {
      "epoch": 0.9882787411689146,
      "grad_norm": 0.27023664116859436,
      "learning_rate": 0.0001,
      "loss": 1.4453,
      "step": 6155
    },
    {
      "epoch": 0.9884393063583815,
      "grad_norm": 0.28188565373420715,
      "learning_rate": 0.0001,
      "loss": 1.5199,
      "step": 6156
    },
    {
      "epoch": 0.9885998715478485,
      "grad_norm": 0.2707480490207672,
      "learning_rate": 0.0001,
      "loss": 1.4374,
      "step": 6157
    },
    {
      "epoch": 0.9887604367373154,
      "grad_norm": 0.2875424921512604,
      "learning_rate": 0.0001,
      "loss": 1.4498,
      "step": 6158
    },
    {
      "epoch": 0.9889210019267822,
      "grad_norm": 0.26978111267089844,
      "learning_rate": 0.0001,
      "loss": 1.4972,
      "step": 6159
    },
    {
      "epoch": 0.9890815671162492,
      "grad_norm": 0.26786309480667114,
      "learning_rate": 0.0001,
      "loss": 1.4748,
      "step": 6160
    },
    {
      "epoch": 0.9892421323057161,
      "grad_norm": 0.2710617780685425,
      "learning_rate": 0.0001,
      "loss": 1.4253,
      "step": 6161
    },
    {
      "epoch": 0.9894026974951831,
      "grad_norm": 0.26757073402404785,
      "learning_rate": 0.0001,
      "loss": 1.5044,
      "step": 6162
    },
    {
      "epoch": 0.98956326268465,
      "grad_norm": 0.2969173491001129,
      "learning_rate": 0.0001,
      "loss": 1.521,
      "step": 6163
    },
    {
      "epoch": 0.989723827874117,
      "grad_norm": 0.263449490070343,
      "learning_rate": 0.0001,
      "loss": 1.4488,
      "step": 6164
    },
    {
      "epoch": 0.9898843930635838,
      "grad_norm": 0.26113778352737427,
      "learning_rate": 0.0001,
      "loss": 1.4237,
      "step": 6165
    },
    {
      "epoch": 0.9900449582530507,
      "grad_norm": 0.281708300113678,
      "learning_rate": 0.0001,
      "loss": 1.4929,
      "step": 6166
    },
    {
      "epoch": 0.9902055234425177,
      "grad_norm": 0.274546355009079,
      "learning_rate": 0.0001,
      "loss": 1.5144,
      "step": 6167
    },
    {
      "epoch": 0.9903660886319846,
      "grad_norm": 0.2744201421737671,
      "learning_rate": 0.0001,
      "loss": 1.4446,
      "step": 6168
    },
    {
      "epoch": 0.9905266538214516,
      "grad_norm": 0.26994210481643677,
      "learning_rate": 0.0001,
      "loss": 1.4654,
      "step": 6169
    },
    {
      "epoch": 0.9906872190109184,
      "grad_norm": 0.2769588232040405,
      "learning_rate": 0.0001,
      "loss": 1.4318,
      "step": 6170
    },
    {
      "epoch": 0.9908477842003853,
      "grad_norm": 0.29553619027137756,
      "learning_rate": 0.0001,
      "loss": 1.5025,
      "step": 6171
    },
    {
      "epoch": 0.9910083493898523,
      "grad_norm": 0.2770235538482666,
      "learning_rate": 0.0001,
      "loss": 1.5148,
      "step": 6172
    },
    {
      "epoch": 0.9911689145793192,
      "grad_norm": 0.2792353630065918,
      "learning_rate": 0.0001,
      "loss": 1.5359,
      "step": 6173
    },
    {
      "epoch": 0.9913294797687862,
      "grad_norm": 0.276390016078949,
      "learning_rate": 0.0001,
      "loss": 1.4846,
      "step": 6174
    },
    {
      "epoch": 0.9914900449582531,
      "grad_norm": 0.27889484167099,
      "learning_rate": 0.0001,
      "loss": 1.4428,
      "step": 6175
    },
    {
      "epoch": 0.9916506101477199,
      "grad_norm": 0.30174416303634644,
      "learning_rate": 0.0001,
      "loss": 1.4598,
      "step": 6176
    },
    {
      "epoch": 0.9918111753371869,
      "grad_norm": 0.27532634139060974,
      "learning_rate": 0.0001,
      "loss": 1.4345,
      "step": 6177
    },
    {
      "epoch": 0.9919717405266538,
      "grad_norm": 0.27762967348098755,
      "learning_rate": 0.0001,
      "loss": 1.4002,
      "step": 6178
    },
    {
      "epoch": 0.9921323057161208,
      "grad_norm": 0.29852789640426636,
      "learning_rate": 0.0001,
      "loss": 1.4166,
      "step": 6179
    },
    {
      "epoch": 0.9922928709055877,
      "grad_norm": 0.27629998326301575,
      "learning_rate": 0.0001,
      "loss": 1.4837,
      "step": 6180
    },
    {
      "epoch": 0.9924534360950545,
      "grad_norm": 0.3078816533088684,
      "learning_rate": 0.0001,
      "loss": 1.5066,
      "step": 6181
    },
    {
      "epoch": 0.9926140012845215,
      "grad_norm": 0.2954596281051636,
      "learning_rate": 0.0001,
      "loss": 1.4908,
      "step": 6182
    },
    {
      "epoch": 0.9927745664739884,
      "grad_norm": 0.2811075747013092,
      "learning_rate": 0.0001,
      "loss": 1.5397,
      "step": 6183
    },
    {
      "epoch": 0.9929351316634554,
      "grad_norm": 0.2870076894760132,
      "learning_rate": 0.0001,
      "loss": 1.544,
      "step": 6184
    },
    {
      "epoch": 0.9930956968529223,
      "grad_norm": 0.32010337710380554,
      "learning_rate": 0.0001,
      "loss": 1.5037,
      "step": 6185
    },
    {
      "epoch": 0.9932562620423893,
      "grad_norm": 0.29296836256980896,
      "learning_rate": 0.0001,
      "loss": 1.5605,
      "step": 6186
    },
    {
      "epoch": 0.9934168272318561,
      "grad_norm": 0.2698921263217926,
      "learning_rate": 0.0001,
      "loss": 1.5104,
      "step": 6187
    },
    {
      "epoch": 0.993577392421323,
      "grad_norm": 0.2633209228515625,
      "learning_rate": 0.0001,
      "loss": 1.4218,
      "step": 6188
    },
    {
      "epoch": 0.99373795761079,
      "grad_norm": 0.2665596306324005,
      "learning_rate": 0.0001,
      "loss": 1.5071,
      "step": 6189
    },
    {
      "epoch": 0.9938985228002569,
      "grad_norm": 0.2854216396808624,
      "learning_rate": 0.0001,
      "loss": 1.537,
      "step": 6190
    },
    {
      "epoch": 0.9940590879897239,
      "grad_norm": 0.28087618947029114,
      "learning_rate": 0.0001,
      "loss": 1.4403,
      "step": 6191
    },
    {
      "epoch": 0.9942196531791907,
      "grad_norm": 0.2625194489955902,
      "learning_rate": 0.0001,
      "loss": 1.4798,
      "step": 6192
    },
    {
      "epoch": 0.9943802183686576,
      "grad_norm": 0.269114226102829,
      "learning_rate": 0.0001,
      "loss": 1.3957,
      "step": 6193
    },
    {
      "epoch": 0.9945407835581246,
      "grad_norm": 0.26467272639274597,
      "learning_rate": 0.0001,
      "loss": 1.4213,
      "step": 6194
    },
    {
      "epoch": 0.9947013487475915,
      "grad_norm": 0.2702041566371918,
      "learning_rate": 0.0001,
      "loss": 1.4794,
      "step": 6195
    },
    {
      "epoch": 0.9948619139370585,
      "grad_norm": 0.30384930968284607,
      "learning_rate": 0.0001,
      "loss": 1.4239,
      "step": 6196
    },
    {
      "epoch": 0.9950224791265254,
      "grad_norm": 0.27441248297691345,
      "learning_rate": 0.0001,
      "loss": 1.4826,
      "step": 6197
    },
    {
      "epoch": 0.9951830443159922,
      "grad_norm": 0.30752405524253845,
      "learning_rate": 0.0001,
      "loss": 1.459,
      "step": 6198
    },
    {
      "epoch": 0.9953436095054592,
      "grad_norm": 0.3062322437763214,
      "learning_rate": 0.0001,
      "loss": 1.4969,
      "step": 6199
    },
    {
      "epoch": 0.9955041746949261,
      "grad_norm": 0.3011613190174103,
      "learning_rate": 0.0001,
      "loss": 1.4248,
      "step": 6200
    },
    {
      "epoch": 0.9956647398843931,
      "grad_norm": 0.2832570970058441,
      "learning_rate": 0.0001,
      "loss": 1.5697,
      "step": 6201
    },
    {
      "epoch": 0.99582530507386,
      "grad_norm": 0.262753963470459,
      "learning_rate": 0.0001,
      "loss": 1.4788,
      "step": 6202
    },
    {
      "epoch": 0.9959858702633269,
      "grad_norm": 0.2819031774997711,
      "learning_rate": 0.0001,
      "loss": 1.4638,
      "step": 6203
    },
    {
      "epoch": 0.9961464354527938,
      "grad_norm": 0.2841491401195526,
      "learning_rate": 0.0001,
      "loss": 1.4599,
      "step": 6204
    },
    {
      "epoch": 0.9963070006422607,
      "grad_norm": 0.2673529386520386,
      "learning_rate": 0.0001,
      "loss": 1.491,
      "step": 6205
    },
    {
      "epoch": 0.9964675658317277,
      "grad_norm": 0.2715817987918854,
      "learning_rate": 0.0001,
      "loss": 1.5009,
      "step": 6206
    },
    {
      "epoch": 0.9966281310211946,
      "grad_norm": 0.30681803822517395,
      "learning_rate": 0.0001,
      "loss": 1.4673,
      "step": 6207
    },
    {
      "epoch": 0.9967886962106616,
      "grad_norm": 0.27776339650154114,
      "learning_rate": 0.0001,
      "loss": 1.3702,
      "step": 6208
    },
    {
      "epoch": 0.9969492614001284,
      "grad_norm": 0.26564541459083557,
      "learning_rate": 0.0001,
      "loss": 1.404,
      "step": 6209
    },
    {
      "epoch": 0.9971098265895953,
      "grad_norm": 0.2827150225639343,
      "learning_rate": 0.0001,
      "loss": 1.4291,
      "step": 6210
    },
    {
      "epoch": 0.9972703917790623,
      "grad_norm": 0.28041255474090576,
      "learning_rate": 0.0001,
      "loss": 1.4986,
      "step": 6211
    },
    {
      "epoch": 0.9974309569685292,
      "grad_norm": 0.2790243923664093,
      "learning_rate": 0.0001,
      "loss": 1.4789,
      "step": 6212
    },
    {
      "epoch": 0.9975915221579962,
      "grad_norm": 0.2771053910255432,
      "learning_rate": 0.0001,
      "loss": 1.4705,
      "step": 6213
    },
    {
      "epoch": 0.9977520873474631,
      "grad_norm": 0.27453845739364624,
      "learning_rate": 0.0001,
      "loss": 1.5308,
      "step": 6214
    },
    {
      "epoch": 0.99791265253693,
      "grad_norm": 0.28533750772476196,
      "learning_rate": 0.0001,
      "loss": 1.3929,
      "step": 6215
    },
    {
      "epoch": 0.9980732177263969,
      "grad_norm": 0.28414100408554077,
      "learning_rate": 0.0001,
      "loss": 1.471,
      "step": 6216
    },
    {
      "epoch": 0.9982337829158638,
      "grad_norm": 0.2757773995399475,
      "learning_rate": 0.0001,
      "loss": 1.4661,
      "step": 6217
    },
    {
      "epoch": 0.9983943481053308,
      "grad_norm": 0.27444469928741455,
      "learning_rate": 0.0001,
      "loss": 1.4961,
      "step": 6218
    },
    {
      "epoch": 0.9985549132947977,
      "grad_norm": 0.27805936336517334,
      "learning_rate": 0.0001,
      "loss": 1.6059,
      "step": 6219
    },
    {
      "epoch": 0.9987154784842646,
      "grad_norm": 0.2738643288612366,
      "learning_rate": 0.0001,
      "loss": 1.4357,
      "step": 6220
    },
    {
      "epoch": 0.9988760436737315,
      "grad_norm": 0.2820572555065155,
      "learning_rate": 0.0001,
      "loss": 1.4146,
      "step": 6221
    },
    {
      "epoch": 0.9990366088631984,
      "grad_norm": 0.2885551154613495,
      "learning_rate": 0.0001,
      "loss": 1.4504,
      "step": 6222
    },
    {
      "epoch": 0.9991971740526654,
      "grad_norm": 0.2716134488582611,
      "learning_rate": 0.0001,
      "loss": 1.4,
      "step": 6223
    },
    {
      "epoch": 0.9993577392421323,
      "grad_norm": 0.270205020904541,
      "learning_rate": 0.0001,
      "loss": 1.4337,
      "step": 6224
    },
    {
      "epoch": 0.9995183044315993,
      "grad_norm": 0.2815423607826233,
      "learning_rate": 0.0001,
      "loss": 1.5145,
      "step": 6225
    },
    {
      "epoch": 0.9996788696210661,
      "grad_norm": 0.3001120686531067,
      "learning_rate": 0.0001,
      "loss": 1.5119,
      "step": 6226
    },
    {
      "epoch": 0.9998394348105331,
      "grad_norm": 0.27873408794403076,
      "learning_rate": 0.0001,
      "loss": 1.5616,
      "step": 6227
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3003462553024292,
      "learning_rate": 0.0001,
      "loss": 1.4898,
      "step": 6228
    },
    {
      "epoch": 1.0,
      "step": 6228,
      "total_flos": 3.5626252265708323e+19,
      "train_loss": 1.5359360199558023,
      "train_runtime": 115641.057,
      "train_samples_per_second": 0.431,
      "train_steps_per_second": 0.054
    }
  ],
  "logging_steps": 1,
  "max_steps": 6228,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 750,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.5626252265708323e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}