{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 945,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010582010582010583,
      "grad_norm": 762.5523071289062,
      "learning_rate": 5e-05,
      "loss": 58.7334,
      "step": 1
    },
    {
      "epoch": 0.0021164021164021165,
      "grad_norm": 82.28478240966797,
      "learning_rate": 4.9947089947089946e-05,
      "loss": 49.8621,
      "step": 2
    },
    {
      "epoch": 0.0031746031746031746,
      "grad_norm": 240.16102600097656,
      "learning_rate": 4.9894179894179895e-05,
      "loss": 44.4804,
      "step": 3
    },
    {
      "epoch": 0.004232804232804233,
      "grad_norm": 224.8407745361328,
      "learning_rate": 4.9841269841269845e-05,
      "loss": 43.2508,
      "step": 4
    },
    {
      "epoch": 0.005291005291005291,
      "grad_norm": 64.33789825439453,
      "learning_rate": 4.978835978835979e-05,
      "loss": 40.2405,
      "step": 5
    },
    {
      "epoch": 0.006349206349206349,
      "grad_norm": 118.09500885009766,
      "learning_rate": 4.973544973544973e-05,
      "loss": 35.3972,
      "step": 6
    },
    {
      "epoch": 0.007407407407407408,
      "grad_norm": 111.19491577148438,
      "learning_rate": 4.968253968253969e-05,
      "loss": 33.5484,
      "step": 7
    },
    {
      "epoch": 0.008465608465608466,
      "grad_norm": 157.63052368164062,
      "learning_rate": 4.962962962962963e-05,
      "loss": 29.6581,
      "step": 8
    },
    {
      "epoch": 0.009523809523809525,
      "grad_norm": 227.2969970703125,
      "learning_rate": 4.9576719576719575e-05,
      "loss": 22.3382,
      "step": 9
    },
    {
      "epoch": 0.010582010582010581,
      "grad_norm": 169.19720458984375,
      "learning_rate": 4.9523809523809525e-05,
      "loss": 14.1501,
      "step": 10
    },
    {
      "epoch": 0.01164021164021164,
      "grad_norm": 80.01154327392578,
      "learning_rate": 4.9470899470899475e-05,
      "loss": 10.3732,
      "step": 11
    },
    {
      "epoch": 0.012698412698412698,
      "grad_norm": 74.98716735839844,
      "learning_rate": 4.941798941798942e-05,
      "loss": 10.4696,
      "step": 12
    },
    {
      "epoch": 0.013756613756613757,
      "grad_norm": 84.27120971679688,
      "learning_rate": 4.936507936507937e-05,
      "loss": 9.5193,
      "step": 13
    },
    {
      "epoch": 0.014814814814814815,
      "grad_norm": 70.76972198486328,
      "learning_rate": 4.931216931216932e-05,
      "loss": 8.4595,
      "step": 14
    },
    {
      "epoch": 0.015873015873015872,
      "grad_norm": 46.97925567626953,
      "learning_rate": 4.925925925925926e-05,
      "loss": 7.5167,
      "step": 15
    },
    {
      "epoch": 0.016931216931216932,
      "grad_norm": 22.022876739501953,
      "learning_rate": 4.9206349206349204e-05,
      "loss": 7.1356,
      "step": 16
    },
    {
      "epoch": 0.01798941798941799,
      "grad_norm": 23.84092903137207,
      "learning_rate": 4.9153439153439154e-05,
      "loss": 7.0376,
      "step": 17
    },
    {
      "epoch": 0.01904761904761905,
      "grad_norm": 21.349409103393555,
      "learning_rate": 4.9100529100529104e-05,
      "loss": 7.0128,
      "step": 18
    },
    {
      "epoch": 0.020105820105820106,
      "grad_norm": 12.040107727050781,
      "learning_rate": 4.904761904761905e-05,
      "loss": 6.4316,
      "step": 19
    },
    {
      "epoch": 0.021164021164021163,
      "grad_norm": 12.285114288330078,
      "learning_rate": 4.8994708994709e-05,
      "loss": 6.4124,
      "step": 20
    },
    {
      "epoch": 0.022222222222222223,
      "grad_norm": 17.068172454833984,
      "learning_rate": 4.894179894179895e-05,
      "loss": 6.2872,
      "step": 21
    },
    {
      "epoch": 0.02328042328042328,
      "grad_norm": 15.330672264099121,
      "learning_rate": 4.888888888888889e-05,
      "loss": 6.1388,
      "step": 22
    },
    {
      "epoch": 0.02433862433862434,
      "grad_norm": 5.708643436431885,
      "learning_rate": 4.883597883597884e-05,
      "loss": 5.9494,
      "step": 23
    },
    {
      "epoch": 0.025396825396825397,
      "grad_norm": 11.539711952209473,
      "learning_rate": 4.878306878306878e-05,
      "loss": 5.9674,
      "step": 24
    },
    {
      "epoch": 0.026455026455026454,
      "grad_norm": 10.946340560913086,
      "learning_rate": 4.873015873015873e-05,
      "loss": 5.8943,
      "step": 25
    },
    {
      "epoch": 0.027513227513227514,
      "grad_norm": 7.171319484710693,
      "learning_rate": 4.8677248677248676e-05,
      "loss": 5.7903,
      "step": 26
    },
    {
      "epoch": 0.02857142857142857,
      "grad_norm": 6.202214241027832,
      "learning_rate": 4.8624338624338626e-05,
      "loss": 5.8376,
      "step": 27
    },
    {
      "epoch": 0.02962962962962963,
      "grad_norm": 11.297914505004883,
      "learning_rate": 4.8571428571428576e-05,
      "loss": 5.492,
      "step": 28
    },
    {
      "epoch": 0.030687830687830688,
      "grad_norm": 8.590489387512207,
      "learning_rate": 4.851851851851852e-05,
      "loss": 5.9732,
      "step": 29
    },
    {
      "epoch": 0.031746031746031744,
      "grad_norm": 4.967828750610352,
      "learning_rate": 4.846560846560847e-05,
      "loss": 5.8429,
      "step": 30
    },
    {
      "epoch": 0.0328042328042328,
      "grad_norm": 7.755899429321289,
      "learning_rate": 4.841269841269841e-05,
      "loss": 5.3487,
      "step": 31
    },
    {
      "epoch": 0.033862433862433865,
      "grad_norm": 5.126328468322754,
      "learning_rate": 4.835978835978836e-05,
      "loss": 5.3793,
      "step": 32
    },
    {
      "epoch": 0.03492063492063492,
      "grad_norm": 7.324168682098389,
      "learning_rate": 4.830687830687831e-05,
      "loss": 6.0176,
      "step": 33
    },
    {
      "epoch": 0.03597883597883598,
      "grad_norm": 5.837782382965088,
      "learning_rate": 4.8253968253968255e-05,
      "loss": 5.531,
      "step": 34
    },
    {
      "epoch": 0.037037037037037035,
      "grad_norm": 7.6066460609436035,
      "learning_rate": 4.8201058201058205e-05,
      "loss": 5.617,
      "step": 35
    },
    {
      "epoch": 0.0380952380952381,
      "grad_norm": 82.80633544921875,
      "learning_rate": 4.814814814814815e-05,
      "loss": 5.664,
      "step": 36
    },
    {
      "epoch": 0.039153439153439155,
      "grad_norm": 12.928624153137207,
      "learning_rate": 4.80952380952381e-05,
      "loss": 5.2884,
      "step": 37
    },
    {
      "epoch": 0.04021164021164021,
      "grad_norm": 8.23830795288086,
      "learning_rate": 4.804232804232804e-05,
      "loss": 5.4968,
      "step": 38
    },
    {
      "epoch": 0.04126984126984127,
      "grad_norm": 8.692220687866211,
      "learning_rate": 4.798941798941799e-05,
      "loss": 5.3659,
      "step": 39
    },
    {
      "epoch": 0.042328042328042326,
      "grad_norm": 6.263346195220947,
      "learning_rate": 4.793650793650794e-05,
      "loss": 5.2251,
      "step": 40
    },
    {
      "epoch": 0.04338624338624339,
      "grad_norm": 6.244935035705566,
      "learning_rate": 4.7883597883597884e-05,
      "loss": 5.243,
      "step": 41
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 6.076992034912109,
      "learning_rate": 4.7830687830687834e-05,
      "loss": 5.4107,
      "step": 42
    },
    {
      "epoch": 0.0455026455026455,
      "grad_norm": 5.009204387664795,
      "learning_rate": 4.7777777777777784e-05,
      "loss": 5.4213,
      "step": 43
    },
    {
      "epoch": 0.04656084656084656,
      "grad_norm": 3.8069138526916504,
      "learning_rate": 4.772486772486773e-05,
      "loss": 5.2413,
      "step": 44
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 5.226925373077393,
      "learning_rate": 4.767195767195767e-05,
      "loss": 5.3811,
      "step": 45
    },
    {
      "epoch": 0.04867724867724868,
      "grad_norm": 5.156801223754883,
      "learning_rate": 4.761904761904762e-05,
      "loss": 5.5233,
      "step": 46
    },
    {
      "epoch": 0.04973544973544974,
      "grad_norm": 4.208065509796143,
      "learning_rate": 4.756613756613757e-05,
      "loss": 5.1164,
      "step": 47
    },
    {
      "epoch": 0.050793650793650794,
      "grad_norm": 4.350522994995117,
      "learning_rate": 4.7513227513227513e-05,
      "loss": 5.2106,
      "step": 48
    },
    {
      "epoch": 0.05185185185185185,
      "grad_norm": 6.086979389190674,
      "learning_rate": 4.746031746031746e-05,
      "loss": 5.2525,
      "step": 49
    },
    {
      "epoch": 0.05291005291005291,
      "grad_norm": 3.6869099140167236,
      "learning_rate": 4.740740740740741e-05,
      "loss": 5.1225,
      "step": 50
    },
    {
      "epoch": 0.05396825396825397,
      "grad_norm": 36.87986755371094,
      "learning_rate": 4.7354497354497356e-05,
      "loss": 6.1878,
      "step": 51
    },
    {
      "epoch": 0.05502645502645503,
      "grad_norm": 4.030213832855225,
      "learning_rate": 4.73015873015873e-05,
      "loss": 5.2523,
      "step": 52
    },
    {
      "epoch": 0.056084656084656084,
      "grad_norm": 6.104913711547852,
      "learning_rate": 4.7248677248677256e-05,
      "loss": 5.1113,
      "step": 53
    },
    {
      "epoch": 0.05714285714285714,
      "grad_norm": 3.9061765670776367,
      "learning_rate": 4.71957671957672e-05,
      "loss": 5.3375,
      "step": 54
    },
    {
      "epoch": 0.0582010582010582,
      "grad_norm": 4.80383825302124,
      "learning_rate": 4.714285714285714e-05,
      "loss": 4.9556,
      "step": 55
    },
    {
      "epoch": 0.05925925925925926,
      "grad_norm": 19.7348690032959,
      "learning_rate": 4.708994708994709e-05,
      "loss": 5.4731,
      "step": 56
    },
    {
      "epoch": 0.06031746031746032,
      "grad_norm": 5.619427680969238,
      "learning_rate": 4.703703703703704e-05,
      "loss": 5.2191,
      "step": 57
    },
    {
      "epoch": 0.061375661375661375,
      "grad_norm": 4.279834747314453,
      "learning_rate": 4.6984126984126986e-05,
      "loss": 5.0808,
      "step": 58
    },
    {
      "epoch": 0.06243386243386243,
      "grad_norm": 5.503246307373047,
      "learning_rate": 4.693121693121693e-05,
      "loss": 4.8678,
      "step": 59
    },
    {
      "epoch": 0.06349206349206349,
      "grad_norm": 5.174645900726318,
      "learning_rate": 4.6878306878306885e-05,
      "loss": 4.8517,
      "step": 60
    },
    {
      "epoch": 0.06455026455026455,
      "grad_norm": 4.007596492767334,
      "learning_rate": 4.682539682539683e-05,
      "loss": 5.1492,
      "step": 61
    },
    {
      "epoch": 0.0656084656084656,
      "grad_norm": 6.642776012420654,
      "learning_rate": 4.677248677248677e-05,
      "loss": 5.2119,
      "step": 62
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 6.208924293518066,
      "learning_rate": 4.671957671957672e-05,
      "loss": 5.105,
      "step": 63
    },
    {
      "epoch": 0.06772486772486773,
      "grad_norm": 5.9450531005859375,
      "learning_rate": 4.666666666666667e-05,
      "loss": 4.9118,
      "step": 64
    },
    {
      "epoch": 0.06878306878306878,
      "grad_norm": 3.375946521759033,
      "learning_rate": 4.6613756613756615e-05,
      "loss": 4.9944,
      "step": 65
    },
    {
      "epoch": 0.06984126984126984,
      "grad_norm": 4.091769218444824,
      "learning_rate": 4.656084656084656e-05,
      "loss": 5.0409,
      "step": 66
    },
    {
      "epoch": 0.07089947089947089,
      "grad_norm": 6.6762542724609375,
      "learning_rate": 4.6507936507936515e-05,
      "loss": 5.022,
      "step": 67
    },
    {
      "epoch": 0.07195767195767196,
      "grad_norm": 5.9758524894714355,
      "learning_rate": 4.645502645502646e-05,
      "loss": 4.8099,
      "step": 68
    },
    {
      "epoch": 0.07301587301587302,
      "grad_norm": 4.996827602386475,
      "learning_rate": 4.64021164021164e-05,
      "loss": 4.917,
      "step": 69
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 5.705566883087158,
      "learning_rate": 4.634920634920635e-05,
      "loss": 4.8434,
      "step": 70
    },
    {
      "epoch": 0.07513227513227513,
      "grad_norm": 11.020337104797363,
      "learning_rate": 4.62962962962963e-05,
      "loss": 5.4662,
      "step": 71
    },
    {
      "epoch": 0.0761904761904762,
      "grad_norm": 6.739878177642822,
      "learning_rate": 4.6243386243386244e-05,
      "loss": 4.5948,
      "step": 72
    },
    {
      "epoch": 0.07724867724867725,
      "grad_norm": 6.3748674392700195,
      "learning_rate": 4.6190476190476194e-05,
      "loss": 4.8899,
      "step": 73
    },
    {
      "epoch": 0.07830687830687831,
      "grad_norm": 4.912506103515625,
      "learning_rate": 4.6137566137566144e-05,
      "loss": 4.5843,
      "step": 74
    },
    {
      "epoch": 0.07936507936507936,
      "grad_norm": 5.216796875,
      "learning_rate": 4.608465608465609e-05,
      "loss": 4.926,
      "step": 75
    },
    {
      "epoch": 0.08042328042328042,
      "grad_norm": 6.009359359741211,
      "learning_rate": 4.603174603174603e-05,
      "loss": 5.1626,
      "step": 76
    },
    {
      "epoch": 0.08148148148148149,
      "grad_norm": 6.634153366088867,
      "learning_rate": 4.597883597883598e-05,
      "loss": 5.0618,
      "step": 77
    },
    {
      "epoch": 0.08253968253968254,
      "grad_norm": 5.179196834564209,
      "learning_rate": 4.592592592592593e-05,
      "loss": 5.137,
      "step": 78
    },
    {
      "epoch": 0.0835978835978836,
      "grad_norm": 4.354217529296875,
      "learning_rate": 4.587301587301587e-05,
      "loss": 4.8585,
      "step": 79
    },
    {
      "epoch": 0.08465608465608465,
      "grad_norm": 5.238408088684082,
      "learning_rate": 4.582010582010582e-05,
      "loss": 4.883,
      "step": 80
    },
    {
      "epoch": 0.08571428571428572,
      "grad_norm": 6.286460876464844,
      "learning_rate": 4.576719576719577e-05,
      "loss": 4.7807,
      "step": 81
    },
    {
      "epoch": 0.08677248677248678,
      "grad_norm": 4.681844234466553,
      "learning_rate": 4.5714285714285716e-05,
      "loss": 4.9966,
      "step": 82
    },
    {
      "epoch": 0.08783068783068783,
      "grad_norm": 4.610260963439941,
      "learning_rate": 4.5661375661375666e-05,
      "loss": 4.785,
      "step": 83
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 3.8423781394958496,
      "learning_rate": 4.560846560846561e-05,
      "loss": 5.0132,
      "step": 84
    },
    {
      "epoch": 0.08994708994708994,
      "grad_norm": 5.1862921714782715,
      "learning_rate": 4.555555555555556e-05,
      "loss": 4.8032,
      "step": 85
    },
    {
      "epoch": 0.091005291005291,
      "grad_norm": 3.509219169616699,
      "learning_rate": 4.55026455026455e-05,
      "loss": 4.8898,
      "step": 86
    },
    {
      "epoch": 0.09206349206349207,
      "grad_norm": 5.162282466888428,
      "learning_rate": 4.544973544973545e-05,
      "loss": 5.0933,
      "step": 87
    },
    {
      "epoch": 0.09312169312169312,
      "grad_norm": 8.435232162475586,
      "learning_rate": 4.5396825396825395e-05,
      "loss": 4.8731,
      "step": 88
    },
    {
      "epoch": 0.09417989417989418,
      "grad_norm": 5.817312717437744,
      "learning_rate": 4.5343915343915345e-05,
      "loss": 5.1801,
      "step": 89
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 4.302613735198975,
      "learning_rate": 4.5291005291005295e-05,
      "loss": 4.748,
      "step": 90
    },
    {
      "epoch": 0.0962962962962963,
      "grad_norm": 3.806368350982666,
      "learning_rate": 4.523809523809524e-05,
      "loss": 5.0319,
      "step": 91
    },
    {
      "epoch": 0.09735449735449736,
      "grad_norm": 5.066342830657959,
      "learning_rate": 4.518518518518519e-05,
      "loss": 4.6202,
      "step": 92
    },
    {
      "epoch": 0.09841269841269841,
      "grad_norm": 6.008955478668213,
      "learning_rate": 4.513227513227514e-05,
      "loss": 5.0549,
      "step": 93
    },
    {
      "epoch": 0.09947089947089947,
      "grad_norm": 4.634308338165283,
      "learning_rate": 4.507936507936508e-05,
      "loss": 4.825,
      "step": 94
    },
    {
      "epoch": 0.10052910052910052,
      "grad_norm": 6.513881206512451,
      "learning_rate": 4.5026455026455024e-05,
      "loss": 4.9766,
      "step": 95
    },
    {
      "epoch": 0.10158730158730159,
      "grad_norm": 4.2518815994262695,
      "learning_rate": 4.4973544973544974e-05,
      "loss": 5.0869,
      "step": 96
    },
    {
      "epoch": 0.10264550264550265,
      "grad_norm": 3.959833860397339,
      "learning_rate": 4.4920634920634924e-05,
      "loss": 4.7034,
      "step": 97
    },
    {
      "epoch": 0.1037037037037037,
      "grad_norm": 4.720004081726074,
      "learning_rate": 4.486772486772487e-05,
      "loss": 4.7978,
      "step": 98
    },
    {
      "epoch": 0.10476190476190476,
      "grad_norm": 4.846654415130615,
      "learning_rate": 4.481481481481482e-05,
      "loss": 4.8179,
      "step": 99
    },
    {
      "epoch": 0.10582010582010581,
      "grad_norm": 4.8786115646362305,
      "learning_rate": 4.476190476190477e-05,
      "loss": 4.6529,
      "step": 100
    },
    {
      "epoch": 0.10687830687830688,
      "grad_norm": 4.344315528869629,
      "learning_rate": 4.470899470899471e-05,
      "loss": 4.7327,
      "step": 101
    },
    {
      "epoch": 0.10793650793650794,
      "grad_norm": 8.0552978515625,
      "learning_rate": 4.4656084656084654e-05,
      "loss": 4.7912,
      "step": 102
    },
    {
      "epoch": 0.10899470899470899,
      "grad_norm": 6.734848976135254,
      "learning_rate": 4.460317460317461e-05,
      "loss": 4.597,
      "step": 103
    },
    {
      "epoch": 0.11005291005291006,
      "grad_norm": 3.832582950592041,
      "learning_rate": 4.4550264550264553e-05,
      "loss": 4.8456,
      "step": 104
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 3.7660605907440186,
      "learning_rate": 4.44973544973545e-05,
      "loss": 5.0973,
      "step": 105
    },
    {
      "epoch": 0.11216931216931217,
      "grad_norm": 5.155391216278076,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 4.8983,
      "step": 106
    },
    {
      "epoch": 0.11322751322751323,
      "grad_norm": 5.6872406005859375,
      "learning_rate": 4.4391534391534397e-05,
      "loss": 4.7054,
      "step": 107
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 4.285317420959473,
      "learning_rate": 4.433862433862434e-05,
      "loss": 4.8328,
      "step": 108
    },
    {
      "epoch": 0.11534391534391535,
      "grad_norm": 4.081933498382568,
      "learning_rate": 4.428571428571428e-05,
      "loss": 5.0816,
      "step": 109
    },
    {
      "epoch": 0.1164021164021164,
      "grad_norm": 4.103255748748779,
      "learning_rate": 4.423280423280424e-05,
      "loss": 4.5584,
      "step": 110
    },
    {
      "epoch": 0.11746031746031746,
      "grad_norm": 3.3774573802948,
      "learning_rate": 4.417989417989418e-05,
      "loss": 4.7316,
      "step": 111
    },
    {
      "epoch": 0.11851851851851852,
      "grad_norm": 4.408664226531982,
      "learning_rate": 4.4126984126984126e-05,
      "loss": 4.9052,
      "step": 112
    },
    {
      "epoch": 0.11957671957671957,
      "grad_norm": 4.274510383605957,
      "learning_rate": 4.4074074074074076e-05,
      "loss": 4.7359,
      "step": 113
    },
    {
      "epoch": 0.12063492063492064,
      "grad_norm": 4.7696309089660645,
      "learning_rate": 4.4021164021164026e-05,
      "loss": 4.7077,
      "step": 114
    },
    {
      "epoch": 0.12169312169312169,
      "grad_norm": 5.227766990661621,
      "learning_rate": 4.396825396825397e-05,
      "loss": 4.7518,
      "step": 115
    },
    {
      "epoch": 0.12275132275132275,
      "grad_norm": 4.35728645324707,
      "learning_rate": 4.391534391534391e-05,
      "loss": 4.7092,
      "step": 116
    },
    {
      "epoch": 0.12380952380952381,
      "grad_norm": 6.191098690032959,
      "learning_rate": 4.386243386243387e-05,
      "loss": 4.6266,
      "step": 117
    },
    {
      "epoch": 0.12486772486772486,
      "grad_norm": 9.239175796508789,
      "learning_rate": 4.380952380952381e-05,
      "loss": 5.0851,
      "step": 118
    },
    {
      "epoch": 0.1259259259259259,
      "grad_norm": 3.643599510192871,
      "learning_rate": 4.3756613756613755e-05,
      "loss": 4.5317,
      "step": 119
    },
    {
      "epoch": 0.12698412698412698,
      "grad_norm": 4.724228382110596,
      "learning_rate": 4.3703703703703705e-05,
      "loss": 4.9586,
      "step": 120
    },
    {
      "epoch": 0.12804232804232804,
      "grad_norm": 8.117734909057617,
      "learning_rate": 4.3650793650793655e-05,
      "loss": 4.922,
      "step": 121
    },
    {
      "epoch": 0.1291005291005291,
      "grad_norm": 5.805531978607178,
      "learning_rate": 4.35978835978836e-05,
      "loss": 4.6248,
      "step": 122
    },
    {
      "epoch": 0.13015873015873017,
      "grad_norm": 4.43373441696167,
      "learning_rate": 4.354497354497355e-05,
      "loss": 4.7716,
      "step": 123
    },
    {
      "epoch": 0.1312169312169312,
      "grad_norm": 3.115410327911377,
      "learning_rate": 4.34920634920635e-05,
      "loss": 4.8769,
      "step": 124
    },
    {
      "epoch": 0.13227513227513227,
      "grad_norm": 3.7560129165649414,
      "learning_rate": 4.343915343915344e-05,
      "loss": 4.7546,
      "step": 125
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 3.7244677543640137,
      "learning_rate": 4.3386243386243384e-05,
      "loss": 4.8349,
      "step": 126
    },
    {
      "epoch": 0.1343915343915344,
      "grad_norm": 3.5653369426727295,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 4.6841,
      "step": 127
    },
    {
      "epoch": 0.13544973544973546,
      "grad_norm": 5.319230556488037,
      "learning_rate": 4.3280423280423284e-05,
      "loss": 4.8841,
      "step": 128
    },
    {
      "epoch": 0.1365079365079365,
      "grad_norm": 3.6799139976501465,
      "learning_rate": 4.322751322751323e-05,
      "loss": 4.6764,
      "step": 129
    },
    {
      "epoch": 0.13756613756613756,
      "grad_norm": 4.678844928741455,
      "learning_rate": 4.317460317460318e-05,
      "loss": 4.8108,
      "step": 130
    },
    {
      "epoch": 0.13862433862433862,
      "grad_norm": 4.520280838012695,
      "learning_rate": 4.312169312169313e-05,
      "loss": 4.4304,
      "step": 131
    },
    {
      "epoch": 0.13968253968253969,
      "grad_norm": 3.9787561893463135,
      "learning_rate": 4.306878306878307e-05,
      "loss": 4.7079,
      "step": 132
    },
    {
      "epoch": 0.14074074074074075,
      "grad_norm": 3.6860711574554443,
      "learning_rate": 4.301587301587302e-05,
      "loss": 4.4752,
      "step": 133
    },
    {
      "epoch": 0.14179894179894179,
      "grad_norm": 5.72208309173584,
      "learning_rate": 4.296296296296296e-05,
      "loss": 4.6981,
      "step": 134
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 3.850085735321045,
      "learning_rate": 4.291005291005291e-05,
      "loss": 4.5628,
      "step": 135
    },
    {
      "epoch": 0.1439153439153439,
      "grad_norm": 4.418358325958252,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 4.6297,
      "step": 136
    },
    {
      "epoch": 0.14497354497354498,
      "grad_norm": 5.23398494720459,
      "learning_rate": 4.2804232804232806e-05,
      "loss": 4.779,
      "step": 137
    },
    {
      "epoch": 0.14603174603174604,
      "grad_norm": 5.489072322845459,
      "learning_rate": 4.2751322751322756e-05,
      "loss": 4.6993,
      "step": 138
    },
    {
      "epoch": 0.14708994708994708,
      "grad_norm": 5.287248134613037,
      "learning_rate": 4.26984126984127e-05,
      "loss": 4.551,
      "step": 139
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 5.898606777191162,
      "learning_rate": 4.264550264550265e-05,
      "loss": 4.8247,
      "step": 140
    },
    {
      "epoch": 0.1492063492063492,
      "grad_norm": 4.300800800323486,
      "learning_rate": 4.259259259259259e-05,
      "loss": 4.4861,
      "step": 141
    },
    {
      "epoch": 0.15026455026455027,
      "grad_norm": 5.065099716186523,
      "learning_rate": 4.253968253968254e-05,
      "loss": 4.6557,
      "step": 142
    },
    {
      "epoch": 0.15132275132275133,
      "grad_norm": 4.122894287109375,
      "learning_rate": 4.248677248677249e-05,
      "loss": 4.66,
      "step": 143
    },
    {
      "epoch": 0.1523809523809524,
      "grad_norm": 5.119123458862305,
      "learning_rate": 4.2433862433862435e-05,
      "loss": 4.7191,
      "step": 144
    },
    {
      "epoch": 0.15343915343915343,
      "grad_norm": 3.865441083908081,
      "learning_rate": 4.2380952380952385e-05,
      "loss": 4.5545,
      "step": 145
    },
    {
      "epoch": 0.1544973544973545,
      "grad_norm": 5.3772406578063965,
      "learning_rate": 4.232804232804233e-05,
      "loss": 4.7646,
      "step": 146
    },
    {
      "epoch": 0.15555555555555556,
      "grad_norm": 4.440810203552246,
      "learning_rate": 4.227513227513228e-05,
      "loss": 4.6804,
      "step": 147
    },
    {
      "epoch": 0.15661375661375662,
      "grad_norm": 4.11602783203125,
      "learning_rate": 4.222222222222222e-05,
      "loss": 4.6774,
      "step": 148
    },
    {
      "epoch": 0.15767195767195769,
      "grad_norm": 3.9843058586120605,
      "learning_rate": 4.216931216931217e-05,
      "loss": 4.4568,
      "step": 149
    },
    {
      "epoch": 0.15873015873015872,
      "grad_norm": 4.767110347747803,
      "learning_rate": 4.211640211640212e-05,
      "loss": 4.7618,
      "step": 150
    },
    {
      "epoch": 0.15978835978835979,
      "grad_norm": 3.820188522338867,
      "learning_rate": 4.2063492063492065e-05,
      "loss": 4.7003,
      "step": 151
    },
    {
      "epoch": 0.16084656084656085,
      "grad_norm": 5.942455768585205,
      "learning_rate": 4.2010582010582014e-05,
      "loss": 4.3406,
      "step": 152
    },
    {
      "epoch": 0.1619047619047619,
      "grad_norm": 4.969194412231445,
      "learning_rate": 4.1957671957671964e-05,
      "loss": 4.7058,
      "step": 153
    },
    {
      "epoch": 0.16296296296296298,
      "grad_norm": 4.873613357543945,
      "learning_rate": 4.190476190476191e-05,
      "loss": 4.666,
      "step": 154
    },
    {
      "epoch": 0.164021164021164,
      "grad_norm": 3.2713990211486816,
      "learning_rate": 4.185185185185185e-05,
      "loss": 4.7406,
      "step": 155
    },
    {
      "epoch": 0.16507936507936508,
      "grad_norm": 4.688560962677002,
      "learning_rate": 4.17989417989418e-05,
      "loss": 4.5528,
      "step": 156
    },
    {
      "epoch": 0.16613756613756614,
      "grad_norm": 4.477161884307861,
      "learning_rate": 4.174603174603175e-05,
      "loss": 4.6816,
      "step": 157
    },
    {
      "epoch": 0.1671957671957672,
      "grad_norm": 3.5517561435699463,
      "learning_rate": 4.1693121693121694e-05,
      "loss": 4.615,
      "step": 158
    },
    {
      "epoch": 0.16825396825396827,
      "grad_norm": 5.814800262451172,
      "learning_rate": 4.1640211640211644e-05,
      "loss": 4.2947,
      "step": 159
    },
    {
      "epoch": 0.1693121693121693,
      "grad_norm": 4.193921089172363,
      "learning_rate": 4.1587301587301594e-05,
      "loss": 4.6911,
      "step": 160
    },
    {
      "epoch": 0.17037037037037037,
      "grad_norm": 4.776106357574463,
      "learning_rate": 4.153439153439154e-05,
      "loss": 4.5339,
      "step": 161
    },
    {
      "epoch": 0.17142857142857143,
      "grad_norm": 3.6364221572875977,
      "learning_rate": 4.148148148148148e-05,
      "loss": 4.8256,
      "step": 162
    },
    {
      "epoch": 0.1724867724867725,
      "grad_norm": 4.010839462280273,
      "learning_rate": 4.1428571428571437e-05,
      "loss": 4.6134,
      "step": 163
    },
    {
      "epoch": 0.17354497354497356,
      "grad_norm": 3.5767881870269775,
      "learning_rate": 4.137566137566138e-05,
      "loss": 4.7187,
      "step": 164
    },
    {
      "epoch": 0.1746031746031746,
      "grad_norm": 6.417559623718262,
      "learning_rate": 4.132275132275132e-05,
      "loss": 4.9141,
      "step": 165
    },
    {
      "epoch": 0.17566137566137566,
      "grad_norm": 4.13731050491333,
      "learning_rate": 4.126984126984127e-05,
      "loss": 4.8624,
      "step": 166
    },
    {
      "epoch": 0.17671957671957672,
      "grad_norm": 5.958324432373047,
      "learning_rate": 4.121693121693122e-05,
      "loss": 4.7571,
      "step": 167
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 5.988956928253174,
      "learning_rate": 4.1164021164021166e-05,
      "loss": 4.5736,
      "step": 168
    },
    {
      "epoch": 0.17883597883597885,
      "grad_norm": 7.006846904754639,
      "learning_rate": 4.111111111111111e-05,
      "loss": 4.4878,
      "step": 169
    },
    {
      "epoch": 0.17989417989417988,
      "grad_norm": 4.999007225036621,
      "learning_rate": 4.105820105820106e-05,
      "loss": 4.8561,
      "step": 170
    },
    {
      "epoch": 0.18095238095238095,
      "grad_norm": 3.71893572807312,
      "learning_rate": 4.100529100529101e-05,
      "loss": 4.5709,
      "step": 171
    },
    {
      "epoch": 0.182010582010582,
      "grad_norm": 3.3149731159210205,
      "learning_rate": 4.095238095238095e-05,
      "loss": 4.846,
      "step": 172
    },
    {
      "epoch": 0.18306878306878308,
      "grad_norm": 5.529186725616455,
      "learning_rate": 4.08994708994709e-05,
      "loss": 4.8602,
      "step": 173
    },
    {
      "epoch": 0.18412698412698414,
      "grad_norm": 4.0738983154296875,
      "learning_rate": 4.084656084656085e-05,
      "loss": 4.7176,
      "step": 174
    },
    {
      "epoch": 0.18518518518518517,
      "grad_norm": 5.210103988647461,
      "learning_rate": 4.0793650793650795e-05,
      "loss": 4.4552,
      "step": 175
    },
    {
      "epoch": 0.18624338624338624,
      "grad_norm": 7.3695068359375,
      "learning_rate": 4.074074074074074e-05,
      "loss": 4.6735,
      "step": 176
    },
    {
      "epoch": 0.1873015873015873,
      "grad_norm": 6.053244113922119,
      "learning_rate": 4.068783068783069e-05,
      "loss": 4.7158,
      "step": 177
    },
    {
      "epoch": 0.18835978835978837,
      "grad_norm": 8.235685348510742,
      "learning_rate": 4.063492063492064e-05,
      "loss": 4.676,
      "step": 178
    },
    {
      "epoch": 0.18941798941798943,
      "grad_norm": 6.053036689758301,
      "learning_rate": 4.058201058201058e-05,
      "loss": 4.8923,
      "step": 179
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 7.517977237701416,
      "learning_rate": 4.052910052910053e-05,
      "loss": 4.6929,
      "step": 180
    },
    {
      "epoch": 0.19153439153439153,
      "grad_norm": 3.8566505908966064,
      "learning_rate": 4.047619047619048e-05,
      "loss": 4.6957,
      "step": 181
    },
    {
      "epoch": 0.1925925925925926,
      "grad_norm": 5.264885425567627,
      "learning_rate": 4.0423280423280424e-05,
      "loss": 4.7976,
      "step": 182
    },
    {
      "epoch": 0.19365079365079366,
      "grad_norm": 5.861869812011719,
      "learning_rate": 4.0370370370370374e-05,
      "loss": 4.7883,
      "step": 183
    },
    {
      "epoch": 0.19470899470899472,
      "grad_norm": 3.94100022315979,
      "learning_rate": 4.031746031746032e-05,
      "loss": 4.7193,
      "step": 184
    },
    {
      "epoch": 0.19576719576719576,
      "grad_norm": 5.635229587554932,
      "learning_rate": 4.026455026455027e-05,
      "loss": 4.8176,
      "step": 185
    },
    {
      "epoch": 0.19682539682539682,
      "grad_norm": 4.604954242706299,
      "learning_rate": 4.021164021164021e-05,
      "loss": 4.7797,
      "step": 186
    },
    {
      "epoch": 0.19788359788359788,
      "grad_norm": 3.7886712551116943,
      "learning_rate": 4.015873015873016e-05,
      "loss": 4.6682,
      "step": 187
    },
    {
      "epoch": 0.19894179894179895,
      "grad_norm": 5.277453422546387,
      "learning_rate": 4.010582010582011e-05,
      "loss": 4.7632,
      "step": 188
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.224989891052246,
      "learning_rate": 4.005291005291005e-05,
      "loss": 4.8058,
      "step": 189
    },
    {
      "epoch": 0.20105820105820105,
      "grad_norm": 4.904975414276123,
      "learning_rate": 4e-05,
      "loss": 4.3213,
      "step": 190
    },
    {
      "epoch": 0.2021164021164021,
      "grad_norm": 5.446446895599365,
      "learning_rate": 3.9947089947089946e-05,
      "loss": 4.831,
      "step": 191
    },
    {
      "epoch": 0.20317460317460317,
      "grad_norm": 4.213114261627197,
      "learning_rate": 3.9894179894179896e-05,
      "loss": 4.6621,
      "step": 192
    },
    {
      "epoch": 0.20423280423280424,
      "grad_norm": 5.95498514175415,
      "learning_rate": 3.984126984126984e-05,
      "loss": 4.5721,
      "step": 193
    },
    {
      "epoch": 0.2052910052910053,
      "grad_norm": 5.409979820251465,
      "learning_rate": 3.978835978835979e-05,
      "loss": 4.8163,
      "step": 194
    },
    {
      "epoch": 0.20634920634920634,
      "grad_norm": 3.3244705200195312,
      "learning_rate": 3.973544973544974e-05,
      "loss": 4.7915,
      "step": 195
    },
    {
      "epoch": 0.2074074074074074,
      "grad_norm": 7.880211353302002,
      "learning_rate": 3.968253968253968e-05,
      "loss": 4.402,
      "step": 196
    },
    {
      "epoch": 0.20846560846560847,
      "grad_norm": 4.982080936431885,
      "learning_rate": 3.962962962962963e-05,
      "loss": 4.675,
      "step": 197
    },
    {
      "epoch": 0.20952380952380953,
      "grad_norm": 3.966547727584839,
      "learning_rate": 3.9576719576719576e-05,
      "loss": 4.4633,
      "step": 198
    },
    {
      "epoch": 0.2105820105820106,
      "grad_norm": 4.820013999938965,
      "learning_rate": 3.9523809523809526e-05,
      "loss": 4.8036,
      "step": 199
    },
    {
      "epoch": 0.21164021164021163,
      "grad_norm": 4.312441825866699,
      "learning_rate": 3.9470899470899475e-05,
      "loss": 4.3439,
      "step": 200
    },
    {
      "epoch": 0.2126984126984127,
      "grad_norm": 4.462612628936768,
      "learning_rate": 3.941798941798942e-05,
      "loss": 4.5635,
      "step": 201
    },
    {
      "epoch": 0.21375661375661376,
      "grad_norm": 4.000391960144043,
      "learning_rate": 3.936507936507937e-05,
      "loss": 4.8163,
      "step": 202
    },
    {
      "epoch": 0.21481481481481482,
      "grad_norm": 4.748673439025879,
      "learning_rate": 3.931216931216931e-05,
      "loss": 4.531,
      "step": 203
    },
    {
      "epoch": 0.21587301587301588,
      "grad_norm": 4.901307582855225,
      "learning_rate": 3.925925925925926e-05,
      "loss": 4.6288,
      "step": 204
    },
    {
      "epoch": 0.21693121693121692,
      "grad_norm": 4.037147045135498,
      "learning_rate": 3.9206349206349205e-05,
      "loss": 4.4996,
      "step": 205
    },
    {
      "epoch": 0.21798941798941798,
      "grad_norm": 3.361581802368164,
      "learning_rate": 3.9153439153439155e-05,
      "loss": 4.6278,
      "step": 206
    },
    {
      "epoch": 0.21904761904761905,
      "grad_norm": 4.140801906585693,
      "learning_rate": 3.9100529100529105e-05,
      "loss": 4.7237,
      "step": 207
    },
    {
      "epoch": 0.2201058201058201,
      "grad_norm": 4.984572410583496,
      "learning_rate": 3.904761904761905e-05,
      "loss": 4.5707,
      "step": 208
    },
    {
      "epoch": 0.22116402116402117,
      "grad_norm": 7.3860578536987305,
      "learning_rate": 3.8994708994709e-05,
      "loss": 4.716,
      "step": 209
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 6.090242385864258,
      "learning_rate": 3.894179894179895e-05,
      "loss": 4.6022,
      "step": 210
    },
    {
      "epoch": 0.22328042328042327,
      "grad_norm": 5.344094276428223,
      "learning_rate": 3.888888888888889e-05,
      "loss": 4.7154,
      "step": 211
    },
    {
      "epoch": 0.22433862433862434,
      "grad_norm": 3.6889617443084717,
      "learning_rate": 3.8835978835978834e-05,
      "loss": 4.8087,
      "step": 212
    },
    {
      "epoch": 0.2253968253968254,
      "grad_norm": 5.947376728057861,
      "learning_rate": 3.8783068783068784e-05,
      "loss": 4.7237,
      "step": 213
    },
    {
      "epoch": 0.22645502645502646,
      "grad_norm": 4.882441520690918,
      "learning_rate": 3.8730158730158734e-05,
      "loss": 4.86,
      "step": 214
    },
    {
      "epoch": 0.2275132275132275,
      "grad_norm": 4.7508625984191895,
      "learning_rate": 3.867724867724868e-05,
      "loss": 4.8796,
      "step": 215
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 4.55201530456543,
      "learning_rate": 3.862433862433863e-05,
      "loss": 4.2992,
      "step": 216
    },
    {
      "epoch": 0.22962962962962963,
      "grad_norm": 4.7580461502075195,
      "learning_rate": 3.857142857142858e-05,
      "loss": 4.7274,
      "step": 217
    },
    {
      "epoch": 0.2306878306878307,
      "grad_norm": 5.989125728607178,
      "learning_rate": 3.851851851851852e-05,
      "loss": 4.8066,
      "step": 218
    },
    {
      "epoch": 0.23174603174603176,
      "grad_norm": 6.322452545166016,
      "learning_rate": 3.846560846560846e-05,
      "loss": 4.669,
      "step": 219
    },
    {
      "epoch": 0.2328042328042328,
      "grad_norm": 4.422494888305664,
      "learning_rate": 3.841269841269842e-05,
      "loss": 4.5265,
      "step": 220
    },
    {
      "epoch": 0.23386243386243386,
      "grad_norm": 4.153601169586182,
      "learning_rate": 3.835978835978836e-05,
      "loss": 4.6917,
      "step": 221
    },
    {
      "epoch": 0.23492063492063492,
      "grad_norm": 5.383472919464111,
      "learning_rate": 3.8306878306878306e-05,
      "loss": 4.4222,
      "step": 222
    },
    {
      "epoch": 0.23597883597883598,
      "grad_norm": 6.448025703430176,
      "learning_rate": 3.8253968253968256e-05,
      "loss": 4.4118,
      "step": 223
    },
    {
      "epoch": 0.23703703703703705,
      "grad_norm": 3.9466822147369385,
      "learning_rate": 3.8201058201058206e-05,
      "loss": 4.6465,
      "step": 224
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 5.374645709991455,
      "learning_rate": 3.814814814814815e-05,
      "loss": 4.7006,
      "step": 225
    },
    {
      "epoch": 0.23915343915343915,
      "grad_norm": 4.644713878631592,
      "learning_rate": 3.809523809523809e-05,
      "loss": 4.4274,
      "step": 226
    },
    {
      "epoch": 0.2402116402116402,
      "grad_norm": 5.4732208251953125,
      "learning_rate": 3.804232804232805e-05,
      "loss": 4.5824,
      "step": 227
    },
    {
      "epoch": 0.24126984126984127,
      "grad_norm": 4.1463189125061035,
      "learning_rate": 3.798941798941799e-05,
      "loss": 4.7644,
      "step": 228
    },
    {
      "epoch": 0.24232804232804234,
      "grad_norm": 4.603626251220703,
      "learning_rate": 3.7936507936507935e-05,
      "loss": 4.4102,
      "step": 229
    },
    {
      "epoch": 0.24338624338624337,
      "grad_norm": 4.667168617248535,
      "learning_rate": 3.7883597883597885e-05,
      "loss": 4.5274,
      "step": 230
    },
    {
      "epoch": 0.24444444444444444,
      "grad_norm": 4.260847091674805,
      "learning_rate": 3.7830687830687835e-05,
      "loss": 5.0719,
      "step": 231
    },
    {
      "epoch": 0.2455026455026455,
      "grad_norm": 5.182448863983154,
      "learning_rate": 3.777777777777778e-05,
      "loss": 4.73,
      "step": 232
    },
    {
      "epoch": 0.24656084656084656,
      "grad_norm": 5.655933380126953,
      "learning_rate": 3.772486772486772e-05,
      "loss": 4.6116,
      "step": 233
    },
    {
      "epoch": 0.24761904761904763,
      "grad_norm": 5.294816493988037,
      "learning_rate": 3.767195767195768e-05,
      "loss": 4.7218,
      "step": 234
    },
    {
      "epoch": 0.24867724867724866,
      "grad_norm": 5.408359050750732,
      "learning_rate": 3.761904761904762e-05,
      "loss": 4.418,
      "step": 235
    },
    {
      "epoch": 0.24973544973544973,
      "grad_norm": 4.957231521606445,
      "learning_rate": 3.7566137566137564e-05,
      "loss": 4.8732,
      "step": 236
    },
    {
      "epoch": 0.2507936507936508,
      "grad_norm": 7.336751461029053,
      "learning_rate": 3.7513227513227514e-05,
      "loss": 4.1392,
      "step": 237
    },
    {
      "epoch": 0.2518518518518518,
      "grad_norm": 4.910647869110107,
      "learning_rate": 3.7460317460317464e-05,
      "loss": 4.6966,
      "step": 238
    },
    {
      "epoch": 0.2529100529100529,
      "grad_norm": 5.7639336585998535,
      "learning_rate": 3.740740740740741e-05,
      "loss": 4.6988,
      "step": 239
    },
    {
      "epoch": 0.25396825396825395,
      "grad_norm": 4.529329299926758,
      "learning_rate": 3.735449735449736e-05,
      "loss": 4.3776,
      "step": 240
    },
    {
      "epoch": 0.25502645502645505,
      "grad_norm": 5.188973903656006,
      "learning_rate": 3.730158730158731e-05,
      "loss": 4.159,
      "step": 241
    },
    {
      "epoch": 0.2560846560846561,
      "grad_norm": 3.262760877609253,
      "learning_rate": 3.724867724867725e-05,
      "loss": 4.6907,
      "step": 242
    },
    {
      "epoch": 0.2571428571428571,
      "grad_norm": 4.056471347808838,
      "learning_rate": 3.7195767195767194e-05,
      "loss": 4.6284,
      "step": 243
    },
    {
      "epoch": 0.2582010582010582,
      "grad_norm": 4.501747131347656,
      "learning_rate": 3.7142857142857143e-05,
      "loss": 4.58,
      "step": 244
    },
    {
      "epoch": 0.25925925925925924,
      "grad_norm": 4.611583709716797,
      "learning_rate": 3.7089947089947093e-05,
      "loss": 4.681,
      "step": 245
    },
    {
      "epoch": 0.26031746031746034,
      "grad_norm": 4.029916763305664,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 4.3931,
      "step": 246
    },
    {
      "epoch": 0.2613756613756614,
      "grad_norm": 3.9643843173980713,
      "learning_rate": 3.6984126984126986e-05,
      "loss": 4.6138,
      "step": 247
    },
    {
      "epoch": 0.2624338624338624,
      "grad_norm": 4.168216228485107,
      "learning_rate": 3.6931216931216936e-05,
      "loss": 4.3513,
      "step": 248
    },
    {
      "epoch": 0.2634920634920635,
      "grad_norm": 3.83542537689209,
      "learning_rate": 3.687830687830688e-05,
      "loss": 4.6283,
      "step": 249
    },
    {
      "epoch": 0.26455026455026454,
      "grad_norm": 3.7934749126434326,
      "learning_rate": 3.682539682539683e-05,
      "loss": 4.7399,
      "step": 250
    },
    {
      "epoch": 0.2656084656084656,
      "grad_norm": 4.369382381439209,
      "learning_rate": 3.677248677248677e-05,
      "loss": 4.7483,
      "step": 251
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 6.055778503417969,
      "learning_rate": 3.671957671957672e-05,
      "loss": 4.45,
      "step": 252
    },
    {
      "epoch": 0.2677248677248677,
      "grad_norm": 4.169711589813232,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 4.5306,
      "step": 253
    },
    {
      "epoch": 0.2687830687830688,
      "grad_norm": 4.536322593688965,
      "learning_rate": 3.6613756613756616e-05,
      "loss": 4.3509,
      "step": 254
    },
    {
      "epoch": 0.2698412698412698,
      "grad_norm": 6.118555068969727,
      "learning_rate": 3.6560846560846566e-05,
      "loss": 4.2821,
      "step": 255
    },
    {
      "epoch": 0.2708994708994709,
      "grad_norm": 5.678388595581055,
      "learning_rate": 3.650793650793651e-05,
      "loss": 4.5712,
      "step": 256
    },
    {
      "epoch": 0.27195767195767195,
      "grad_norm": 3.9464683532714844,
      "learning_rate": 3.645502645502646e-05,
      "loss": 4.6364,
      "step": 257
    },
    {
      "epoch": 0.273015873015873,
      "grad_norm": 4.010932922363281,
      "learning_rate": 3.64021164021164e-05,
      "loss": 4.4778,
      "step": 258
    },
    {
      "epoch": 0.2740740740740741,
      "grad_norm": 3.9546217918395996,
      "learning_rate": 3.634920634920635e-05,
      "loss": 4.7097,
      "step": 259
    },
    {
      "epoch": 0.2751322751322751,
      "grad_norm": 5.1796345710754395,
      "learning_rate": 3.62962962962963e-05,
      "loss": 4.6363,
      "step": 260
    },
    {
      "epoch": 0.2761904761904762,
      "grad_norm": 5.096473217010498,
      "learning_rate": 3.6243386243386245e-05,
      "loss": 4.5581,
      "step": 261
    },
    {
      "epoch": 0.27724867724867724,
      "grad_norm": 6.787267208099365,
      "learning_rate": 3.619047619047619e-05,
      "loss": 4.4793,
      "step": 262
    },
    {
      "epoch": 0.2783068783068783,
      "grad_norm": 3.9778239727020264,
      "learning_rate": 3.613756613756614e-05,
      "loss": 4.5243,
      "step": 263
    },
    {
      "epoch": 0.27936507936507937,
      "grad_norm": 3.2374017238616943,
      "learning_rate": 3.608465608465609e-05,
      "loss": 4.7684,
      "step": 264
    },
    {
      "epoch": 0.2804232804232804,
      "grad_norm": 4.345169544219971,
      "learning_rate": 3.603174603174603e-05,
      "loss": 4.3203,
      "step": 265
    },
    {
      "epoch": 0.2814814814814815,
      "grad_norm": 5.155961990356445,
      "learning_rate": 3.597883597883598e-05,
      "loss": 4.5482,
      "step": 266
    },
    {
      "epoch": 0.28253968253968254,
      "grad_norm": 4.2913713455200195,
      "learning_rate": 3.592592592592593e-05,
      "loss": 4.6832,
      "step": 267
    },
    {
      "epoch": 0.28359788359788357,
      "grad_norm": 4.977283477783203,
      "learning_rate": 3.5873015873015874e-05,
      "loss": 4.3257,
      "step": 268
    },
    {
      "epoch": 0.28465608465608466,
      "grad_norm": 4.866610050201416,
      "learning_rate": 3.582010582010582e-05,
      "loss": 4.5165,
      "step": 269
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 5.422711372375488,
      "learning_rate": 3.5767195767195774e-05,
      "loss": 4.6076,
      "step": 270
    },
    {
      "epoch": 0.2867724867724868,
      "grad_norm": 7.791525363922119,
      "learning_rate": 3.571428571428572e-05,
      "loss": 4.8686,
      "step": 271
    },
    {
      "epoch": 0.2878306878306878,
      "grad_norm": 6.378711700439453,
      "learning_rate": 3.566137566137566e-05,
      "loss": 4.6424,
      "step": 272
    },
    {
      "epoch": 0.28888888888888886,
      "grad_norm": 3.367389678955078,
      "learning_rate": 3.560846560846561e-05,
      "loss": 4.6617,
      "step": 273
    },
    {
      "epoch": 0.28994708994708995,
      "grad_norm": 3.890188694000244,
      "learning_rate": 3.555555555555556e-05,
      "loss": 4.2508,
      "step": 274
    },
    {
      "epoch": 0.291005291005291,
      "grad_norm": 3.5759942531585693,
      "learning_rate": 3.55026455026455e-05,
      "loss": 4.7628,
      "step": 275
    },
    {
      "epoch": 0.2920634920634921,
      "grad_norm": 3.059518337249756,
      "learning_rate": 3.5449735449735446e-05,
      "loss": 4.5359,
      "step": 276
    },
    {
      "epoch": 0.2931216931216931,
      "grad_norm": 4.013446807861328,
      "learning_rate": 3.53968253968254e-05,
      "loss": 4.5223,
      "step": 277
    },
    {
      "epoch": 0.29417989417989415,
      "grad_norm": 5.019964218139648,
      "learning_rate": 3.5343915343915346e-05,
      "loss": 4.456,
      "step": 278
    },
    {
      "epoch": 0.29523809523809524,
      "grad_norm": 3.2760136127471924,
      "learning_rate": 3.529100529100529e-05,
      "loss": 4.5636,
      "step": 279
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 5.238942623138428,
      "learning_rate": 3.523809523809524e-05,
      "loss": 4.5664,
      "step": 280
    },
    {
      "epoch": 0.29735449735449737,
      "grad_norm": 5.568114280700684,
      "learning_rate": 3.518518518518519e-05,
      "loss": 4.5683,
      "step": 281
    },
    {
      "epoch": 0.2984126984126984,
      "grad_norm": 3.24711537361145,
      "learning_rate": 3.513227513227513e-05,
      "loss": 4.5822,
      "step": 282
    },
    {
      "epoch": 0.29947089947089944,
      "grad_norm": 4.6277313232421875,
      "learning_rate": 3.5079365079365075e-05,
      "loss": 4.6366,
      "step": 283
    },
    {
      "epoch": 0.30052910052910053,
      "grad_norm": 3.8839879035949707,
      "learning_rate": 3.502645502645503e-05,
      "loss": 4.5388,
      "step": 284
    },
    {
      "epoch": 0.30158730158730157,
      "grad_norm": 7.662173271179199,
      "learning_rate": 3.4973544973544975e-05,
      "loss": 4.2736,
      "step": 285
    },
    {
      "epoch": 0.30264550264550266,
      "grad_norm": 5.089264869689941,
      "learning_rate": 3.492063492063492e-05,
      "loss": 4.6195,
      "step": 286
    },
    {
      "epoch": 0.3037037037037037,
      "grad_norm": 3.990954637527466,
      "learning_rate": 3.486772486772487e-05,
      "loss": 4.4885,
      "step": 287
    },
    {
      "epoch": 0.3047619047619048,
      "grad_norm": 4.801586151123047,
      "learning_rate": 3.481481481481482e-05,
      "loss": 4.5381,
      "step": 288
    },
    {
      "epoch": 0.3058201058201058,
      "grad_norm": 5.5773210525512695,
      "learning_rate": 3.476190476190476e-05,
      "loss": 4.5179,
      "step": 289
    },
    {
      "epoch": 0.30687830687830686,
      "grad_norm": 3.99760103225708,
      "learning_rate": 3.470899470899471e-05,
      "loss": 4.6765,
      "step": 290
    },
    {
      "epoch": 0.30793650793650795,
      "grad_norm": 4.964691638946533,
      "learning_rate": 3.465608465608466e-05,
      "loss": 4.4914,
      "step": 291
    },
    {
      "epoch": 0.308994708994709,
      "grad_norm": 3.3278160095214844,
      "learning_rate": 3.4603174603174604e-05,
      "loss": 4.4241,
      "step": 292
    },
    {
      "epoch": 0.3100529100529101,
      "grad_norm": 6.8910675048828125,
      "learning_rate": 3.455026455026455e-05,
      "loss": 4.8309,
      "step": 293
    },
    {
      "epoch": 0.3111111111111111,
      "grad_norm": 3.6537747383117676,
      "learning_rate": 3.44973544973545e-05,
      "loss": 4.4105,
      "step": 294
    },
    {
      "epoch": 0.31216931216931215,
      "grad_norm": 3.973331928253174,
      "learning_rate": 3.444444444444445e-05,
      "loss": 4.6153,
      "step": 295
    },
    {
      "epoch": 0.31322751322751324,
      "grad_norm": 6.508915424346924,
      "learning_rate": 3.439153439153439e-05,
      "loss": 4.4718,
      "step": 296
    },
    {
      "epoch": 0.3142857142857143,
      "grad_norm": 5.570568561553955,
      "learning_rate": 3.433862433862434e-05,
      "loss": 4.6205,
      "step": 297
    },
    {
      "epoch": 0.31534391534391537,
      "grad_norm": 4.15952205657959,
      "learning_rate": 3.428571428571429e-05,
      "loss": 4.7089,
      "step": 298
    },
    {
      "epoch": 0.3164021164021164,
      "grad_norm": 5.480037212371826,
      "learning_rate": 3.4232804232804234e-05,
      "loss": 4.7241,
      "step": 299
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 4.26448917388916,
      "learning_rate": 3.4179894179894184e-05,
      "loss": 4.5638,
      "step": 300
    },
    {
      "epoch": 0.31851851851851853,
      "grad_norm": 6.337195873260498,
      "learning_rate": 3.412698412698413e-05,
      "loss": 4.4273,
      "step": 301
    },
    {
      "epoch": 0.31957671957671957,
      "grad_norm": 5.988368988037109,
      "learning_rate": 3.4074074074074077e-05,
      "loss": 4.7023,
      "step": 302
    },
    {
      "epoch": 0.32063492063492066,
      "grad_norm": 5.5184245109558105,
      "learning_rate": 3.402116402116402e-05,
      "loss": 4.5858,
      "step": 303
    },
    {
      "epoch": 0.3216931216931217,
      "grad_norm": 5.176582336425781,
      "learning_rate": 3.396825396825397e-05,
      "loss": 4.4652,
      "step": 304
    },
    {
      "epoch": 0.32275132275132273,
      "grad_norm": 4.534494400024414,
      "learning_rate": 3.391534391534392e-05,
      "loss": 4.537,
      "step": 305
    },
    {
      "epoch": 0.3238095238095238,
      "grad_norm": 5.228606224060059,
      "learning_rate": 3.386243386243386e-05,
      "loss": 4.6445,
      "step": 306
    },
    {
      "epoch": 0.32486772486772486,
      "grad_norm": 3.956960678100586,
      "learning_rate": 3.380952380952381e-05,
      "loss": 4.4717,
      "step": 307
    },
    {
      "epoch": 0.32592592592592595,
      "grad_norm": 4.401339530944824,
      "learning_rate": 3.3756613756613756e-05,
      "loss": 4.3484,
      "step": 308
    },
    {
      "epoch": 0.326984126984127,
      "grad_norm": 4.009274482727051,
      "learning_rate": 3.3703703703703706e-05,
      "loss": 4.4202,
      "step": 309
    },
    {
      "epoch": 0.328042328042328,
      "grad_norm": 3.2340402603149414,
      "learning_rate": 3.3650793650793656e-05,
      "loss": 4.5176,
      "step": 310
    },
    {
      "epoch": 0.3291005291005291,
      "grad_norm": 3.553013563156128,
      "learning_rate": 3.35978835978836e-05,
      "loss": 4.598,
      "step": 311
    },
    {
      "epoch": 0.33015873015873015,
      "grad_norm": 6.81786584854126,
      "learning_rate": 3.354497354497355e-05,
      "loss": 4.2469,
      "step": 312
    },
    {
      "epoch": 0.33121693121693124,
      "grad_norm": 4.922097682952881,
      "learning_rate": 3.349206349206349e-05,
      "loss": 4.6131,
      "step": 313
    },
    {
      "epoch": 0.3322751322751323,
      "grad_norm": 4.777652263641357,
      "learning_rate": 3.343915343915344e-05,
      "loss": 4.7549,
      "step": 314
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.2790966033935547,
      "learning_rate": 3.3386243386243385e-05,
      "loss": 4.4934,
      "step": 315
    },
    {
      "epoch": 0.3343915343915344,
      "grad_norm": 4.401710033416748,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 4.5042,
      "step": 316
    },
    {
      "epoch": 0.33544973544973544,
      "grad_norm": 5.545312404632568,
      "learning_rate": 3.3280423280423285e-05,
      "loss": 4.3179,
      "step": 317
    },
    {
      "epoch": 0.33650793650793653,
      "grad_norm": 3.8146395683288574,
      "learning_rate": 3.322751322751323e-05,
      "loss": 4.5358,
      "step": 318
    },
    {
      "epoch": 0.33756613756613757,
      "grad_norm": 4.654644966125488,
      "learning_rate": 3.317460317460318e-05,
      "loss": 4.5204,
      "step": 319
    },
    {
      "epoch": 0.3386243386243386,
      "grad_norm": 4.144741535186768,
      "learning_rate": 3.312169312169313e-05,
      "loss": 4.93,
      "step": 320
    },
    {
      "epoch": 0.3396825396825397,
      "grad_norm": 4.340035915374756,
      "learning_rate": 3.306878306878307e-05,
      "loss": 4.2467,
      "step": 321
    },
    {
      "epoch": 0.34074074074074073,
      "grad_norm": 4.59688663482666,
      "learning_rate": 3.3015873015873014e-05,
      "loss": 4.3413,
      "step": 322
    },
    {
      "epoch": 0.3417989417989418,
      "grad_norm": 3.6761655807495117,
      "learning_rate": 3.2962962962962964e-05,
      "loss": 4.4834,
      "step": 323
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 5.661898612976074,
      "learning_rate": 3.2910052910052914e-05,
      "loss": 4.516,
      "step": 324
    },
    {
      "epoch": 0.3439153439153439,
      "grad_norm": 2.8504817485809326,
      "learning_rate": 3.285714285714286e-05,
      "loss": 4.55,
      "step": 325
    },
    {
      "epoch": 0.344973544973545,
      "grad_norm": 3.2584402561187744,
      "learning_rate": 3.280423280423281e-05,
      "loss": 4.4524,
      "step": 326
    },
    {
      "epoch": 0.346031746031746,
      "grad_norm": 3.2479212284088135,
      "learning_rate": 3.275132275132276e-05,
      "loss": 4.5226,
      "step": 327
    },
    {
      "epoch": 0.3470899470899471,
      "grad_norm": 5.987076282501221,
      "learning_rate": 3.26984126984127e-05,
      "loss": 4.4857,
      "step": 328
    },
    {
      "epoch": 0.34814814814814815,
      "grad_norm": 5.704360485076904,
      "learning_rate": 3.264550264550264e-05,
      "loss": 4.5853,
      "step": 329
    },
    {
      "epoch": 0.3492063492063492,
      "grad_norm": 3.879491090774536,
      "learning_rate": 3.25925925925926e-05,
      "loss": 4.6353,
      "step": 330
    },
    {
      "epoch": 0.3502645502645503,
      "grad_norm": 4.888803958892822,
      "learning_rate": 3.253968253968254e-05,
      "loss": 4.777,
      "step": 331
    },
    {
      "epoch": 0.3513227513227513,
      "grad_norm": 3.1391172409057617,
      "learning_rate": 3.2486772486772486e-05,
      "loss": 4.6797,
      "step": 332
    },
    {
      "epoch": 0.3523809523809524,
      "grad_norm": 3.772940158843994,
      "learning_rate": 3.2433862433862436e-05,
      "loss": 4.4289,
      "step": 333
    },
    {
      "epoch": 0.35343915343915344,
      "grad_norm": 6.232966423034668,
      "learning_rate": 3.2380952380952386e-05,
      "loss": 4.7493,
      "step": 334
    },
    {
      "epoch": 0.3544973544973545,
      "grad_norm": 3.59843373298645,
      "learning_rate": 3.232804232804233e-05,
      "loss": 4.6664,
      "step": 335
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 4.1545538902282715,
      "learning_rate": 3.227513227513227e-05,
      "loss": 4.3846,
      "step": 336
    },
    {
      "epoch": 0.3566137566137566,
      "grad_norm": 3.4982540607452393,
      "learning_rate": 3.222222222222223e-05,
      "loss": 4.5178,
      "step": 337
    },
    {
      "epoch": 0.3576719576719577,
      "grad_norm": 3.8578543663024902,
      "learning_rate": 3.216931216931217e-05,
      "loss": 4.4057,
      "step": 338
    },
    {
      "epoch": 0.35873015873015873,
      "grad_norm": 4.9274678230285645,
      "learning_rate": 3.2116402116402115e-05,
      "loss": 4.6315,
      "step": 339
    },
    {
      "epoch": 0.35978835978835977,
      "grad_norm": 3.3767309188842773,
      "learning_rate": 3.2063492063492065e-05,
      "loss": 4.4462,
      "step": 340
    },
    {
      "epoch": 0.36084656084656086,
      "grad_norm": 5.290660858154297,
      "learning_rate": 3.2010582010582015e-05,
      "loss": 4.4087,
      "step": 341
    },
    {
      "epoch": 0.3619047619047619,
      "grad_norm": 3.6890292167663574,
      "learning_rate": 3.195767195767196e-05,
      "loss": 4.2977,
      "step": 342
    },
    {
      "epoch": 0.362962962962963,
      "grad_norm": 4.656151294708252,
      "learning_rate": 3.19047619047619e-05,
      "loss": 4.2439,
      "step": 343
    },
    {
      "epoch": 0.364021164021164,
      "grad_norm": 3.621307611465454,
      "learning_rate": 3.185185185185185e-05,
      "loss": 4.4836,
      "step": 344
    },
    {
      "epoch": 0.36507936507936506,
      "grad_norm": 3.977259874343872,
      "learning_rate": 3.17989417989418e-05,
      "loss": 4.4694,
      "step": 345
    },
    {
      "epoch": 0.36613756613756615,
      "grad_norm": 3.787940263748169,
      "learning_rate": 3.1746031746031745e-05,
      "loss": 4.3896,
      "step": 346
    },
    {
      "epoch": 0.3671957671957672,
      "grad_norm": 6.7594451904296875,
      "learning_rate": 3.1693121693121695e-05,
      "loss": 4.4299,
      "step": 347
    },
    {
      "epoch": 0.3682539682539683,
      "grad_norm": 4.46245813369751,
      "learning_rate": 3.1640211640211645e-05,
      "loss": 4.9394,
      "step": 348
    },
    {
      "epoch": 0.3693121693121693,
      "grad_norm": 4.181535720825195,
      "learning_rate": 3.158730158730159e-05,
      "loss": 4.6598,
      "step": 349
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 2.9622275829315186,
      "learning_rate": 3.153439153439154e-05,
      "loss": 4.6615,
      "step": 350
    },
    {
      "epoch": 0.37142857142857144,
      "grad_norm": 6.954951763153076,
      "learning_rate": 3.148148148148148e-05,
      "loss": 4.5004,
      "step": 351
    },
    {
      "epoch": 0.3724867724867725,
      "grad_norm": 5.163333415985107,
      "learning_rate": 3.142857142857143e-05,
      "loss": 4.4917,
      "step": 352
    },
    {
      "epoch": 0.37354497354497357,
      "grad_norm": 4.577300548553467,
      "learning_rate": 3.1375661375661374e-05,
      "loss": 4.5335,
      "step": 353
    },
    {
      "epoch": 0.3746031746031746,
      "grad_norm": 3.999112367630005,
      "learning_rate": 3.1322751322751324e-05,
      "loss": 4.4476,
      "step": 354
    },
    {
      "epoch": 0.37566137566137564,
      "grad_norm": 4.00216007232666,
      "learning_rate": 3.1269841269841274e-05,
      "loss": 4.4549,
      "step": 355
    },
    {
      "epoch": 0.37671957671957673,
      "grad_norm": 5.307511329650879,
      "learning_rate": 3.121693121693122e-05,
      "loss": 4.6019,
      "step": 356
    },
    {
      "epoch": 0.37777777777777777,
      "grad_norm": 2.6185190677642822,
      "learning_rate": 3.116402116402117e-05,
      "loss": 4.5182,
      "step": 357
    },
    {
      "epoch": 0.37883597883597886,
      "grad_norm": 4.905261039733887,
      "learning_rate": 3.111111111111111e-05,
      "loss": 4.4365,
      "step": 358
    },
    {
      "epoch": 0.3798941798941799,
      "grad_norm": 3.4827020168304443,
      "learning_rate": 3.105820105820106e-05,
      "loss": 4.5379,
      "step": 359
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 5.549962997436523,
      "learning_rate": 3.100529100529101e-05,
      "loss": 4.4842,
      "step": 360
    },
    {
      "epoch": 0.382010582010582,
      "grad_norm": 3.405832052230835,
      "learning_rate": 3.095238095238095e-05,
      "loss": 4.4383,
      "step": 361
    },
    {
      "epoch": 0.38306878306878306,
      "grad_norm": 4.3347673416137695,
      "learning_rate": 3.08994708994709e-05,
      "loss": 4.5793,
      "step": 362
    },
    {
      "epoch": 0.38412698412698415,
      "grad_norm": 4.637452602386475,
      "learning_rate": 3.0846560846560846e-05,
      "loss": 4.4116,
      "step": 363
    },
    {
      "epoch": 0.3851851851851852,
      "grad_norm": 3.653627395629883,
      "learning_rate": 3.0793650793650796e-05,
      "loss": 4.608,
      "step": 364
    },
    {
      "epoch": 0.3862433862433862,
      "grad_norm": 3.076195240020752,
      "learning_rate": 3.074074074074074e-05,
      "loss": 4.4001,
      "step": 365
    },
    {
      "epoch": 0.3873015873015873,
      "grad_norm": 5.166315078735352,
      "learning_rate": 3.068783068783069e-05,
      "loss": 4.1011,
      "step": 366
    },
    {
      "epoch": 0.38835978835978835,
      "grad_norm": 6.1348795890808105,
      "learning_rate": 3.063492063492064e-05,
      "loss": 4.6797,
      "step": 367
    },
    {
      "epoch": 0.38941798941798944,
      "grad_norm": 2.963944673538208,
      "learning_rate": 3.058201058201058e-05,
      "loss": 4.3261,
      "step": 368
    },
    {
      "epoch": 0.3904761904761905,
      "grad_norm": 4.195929527282715,
      "learning_rate": 3.052910052910053e-05,
      "loss": 4.5108,
      "step": 369
    },
    {
      "epoch": 0.3915343915343915,
      "grad_norm": 4.026662349700928,
      "learning_rate": 3.0476190476190482e-05,
      "loss": 4.2837,
      "step": 370
    },
    {
      "epoch": 0.3925925925925926,
      "grad_norm": 4.093654155731201,
      "learning_rate": 3.0423280423280425e-05,
      "loss": 4.6069,
      "step": 371
    },
    {
      "epoch": 0.39365079365079364,
      "grad_norm": 5.716070652008057,
      "learning_rate": 3.037037037037037e-05,
      "loss": 4.4523,
      "step": 372
    },
    {
      "epoch": 0.39470899470899473,
      "grad_norm": 3.701359748840332,
      "learning_rate": 3.0317460317460318e-05,
      "loss": 4.4073,
      "step": 373
    },
    {
      "epoch": 0.39576719576719577,
      "grad_norm": 7.2860260009765625,
      "learning_rate": 3.0264550264550268e-05,
      "loss": 4.3031,
      "step": 374
    },
    {
      "epoch": 0.3968253968253968,
      "grad_norm": 6.570174217224121,
      "learning_rate": 3.0211640211640215e-05,
      "loss": 4.4551,
      "step": 375
    },
    {
      "epoch": 0.3978835978835979,
      "grad_norm": 4.895565509796143,
      "learning_rate": 3.0158730158730158e-05,
      "loss": 4.4219,
      "step": 376
    },
    {
      "epoch": 0.39894179894179893,
      "grad_norm": 7.9695844650268555,
      "learning_rate": 3.010582010582011e-05,
      "loss": 4.4131,
      "step": 377
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.441410064697266,
      "learning_rate": 3.0052910052910054e-05,
      "loss": 4.4745,
      "step": 378
    },
    {
      "epoch": 0.40105820105820106,
      "grad_norm": 3.7005107402801514,
      "learning_rate": 3e-05,
      "loss": 4.3727,
      "step": 379
    },
    {
      "epoch": 0.4021164021164021,
      "grad_norm": 5.960515975952148,
      "learning_rate": 2.9947089947089947e-05,
      "loss": 4.5026,
      "step": 380
    },
    {
      "epoch": 0.4031746031746032,
      "grad_norm": 3.2702596187591553,
      "learning_rate": 2.9894179894179897e-05,
      "loss": 4.5372,
      "step": 381
    },
    {
      "epoch": 0.4042328042328042,
      "grad_norm": 4.725318908691406,
      "learning_rate": 2.9841269841269844e-05,
      "loss": 4.4181,
      "step": 382
    },
    {
      "epoch": 0.4052910052910053,
      "grad_norm": 6.9476847648620605,
      "learning_rate": 2.9788359788359787e-05,
      "loss": 4.2138,
      "step": 383
    },
    {
      "epoch": 0.40634920634920635,
      "grad_norm": 3.7959463596343994,
      "learning_rate": 2.973544973544974e-05,
      "loss": 4.4974,
      "step": 384
    },
    {
      "epoch": 0.4074074074074074,
      "grad_norm": 2.9990975856781006,
      "learning_rate": 2.9682539682539683e-05,
      "loss": 4.4096,
      "step": 385
    },
    {
      "epoch": 0.4084656084656085,
      "grad_norm": 6.560277462005615,
      "learning_rate": 2.962962962962963e-05,
      "loss": 4.9725,
      "step": 386
    },
    {
      "epoch": 0.4095238095238095,
      "grad_norm": 3.6386265754699707,
      "learning_rate": 2.957671957671958e-05,
      "loss": 4.5124,
      "step": 387
    },
    {
      "epoch": 0.4105820105820106,
      "grad_norm": 3.7434399127960205,
      "learning_rate": 2.9523809523809526e-05,
      "loss": 4.308,
      "step": 388
    },
    {
      "epoch": 0.41164021164021164,
      "grad_norm": 4.926546096801758,
      "learning_rate": 2.9470899470899473e-05,
      "loss": 4.44,
      "step": 389
    },
    {
      "epoch": 0.4126984126984127,
      "grad_norm": 4.1928510665893555,
      "learning_rate": 2.9417989417989416e-05,
      "loss": 4.588,
      "step": 390
    },
    {
      "epoch": 0.41375661375661377,
      "grad_norm": 3.4188411235809326,
      "learning_rate": 2.9365079365079366e-05,
      "loss": 4.3135,
      "step": 391
    },
    {
      "epoch": 0.4148148148148148,
      "grad_norm": 4.299923896789551,
      "learning_rate": 2.9312169312169313e-05,
      "loss": 4.4664,
      "step": 392
    },
    {
      "epoch": 0.4158730158730159,
      "grad_norm": 5.515903472900391,
      "learning_rate": 2.925925925925926e-05,
      "loss": 4.5932,
      "step": 393
    },
    {
      "epoch": 0.41693121693121693,
      "grad_norm": 4.561779022216797,
      "learning_rate": 2.920634920634921e-05,
      "loss": 4.4384,
      "step": 394
    },
    {
      "epoch": 0.41798941798941797,
      "grad_norm": 4.012057781219482,
      "learning_rate": 2.9153439153439156e-05,
      "loss": 4.3051,
      "step": 395
    },
    {
      "epoch": 0.41904761904761906,
      "grad_norm": 4.529195785522461,
      "learning_rate": 2.91005291005291e-05,
      "loss": 4.0737,
      "step": 396
    },
    {
      "epoch": 0.4201058201058201,
      "grad_norm": 4.308627128601074,
      "learning_rate": 2.9047619047619052e-05,
      "loss": 4.2985,
      "step": 397
    },
    {
      "epoch": 0.4211640211640212,
      "grad_norm": 3.731900930404663,
      "learning_rate": 2.8994708994708995e-05,
      "loss": 4.5682,
      "step": 398
    },
    {
      "epoch": 0.4222222222222222,
      "grad_norm": 3.0067977905273438,
      "learning_rate": 2.8941798941798942e-05,
      "loss": 4.4598,
      "step": 399
    },
    {
      "epoch": 0.42328042328042326,
      "grad_norm": 5.660052299499512,
      "learning_rate": 2.8888888888888888e-05,
      "loss": 4.2262,
      "step": 400
    },
    {
      "epoch": 0.42433862433862435,
      "grad_norm": 3.870854139328003,
      "learning_rate": 2.8835978835978838e-05,
      "loss": 4.4143,
      "step": 401
    },
    {
      "epoch": 0.4253968253968254,
      "grad_norm": 4.174057483673096,
      "learning_rate": 2.8783068783068785e-05,
      "loss": 4.3944,
      "step": 402
    },
    {
      "epoch": 0.4264550264550265,
      "grad_norm": 3.0973563194274902,
      "learning_rate": 2.8730158730158728e-05,
      "loss": 4.6079,
      "step": 403
    },
    {
      "epoch": 0.4275132275132275,
      "grad_norm": 3.542325735092163,
      "learning_rate": 2.867724867724868e-05,
      "loss": 4.5103,
      "step": 404
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 5.6434712409973145,
      "learning_rate": 2.8624338624338624e-05,
      "loss": 4.6433,
      "step": 405
    },
    {
      "epoch": 0.42962962962962964,
      "grad_norm": 4.372058868408203,
      "learning_rate": 2.857142857142857e-05,
      "loss": 4.4411,
      "step": 406
    },
    {
      "epoch": 0.4306878306878307,
      "grad_norm": 4.075821876525879,
      "learning_rate": 2.851851851851852e-05,
      "loss": 4.5535,
      "step": 407
    },
    {
      "epoch": 0.43174603174603177,
      "grad_norm": 3.91575288772583,
      "learning_rate": 2.8465608465608467e-05,
      "loss": 4.697,
      "step": 408
    },
    {
      "epoch": 0.4328042328042328,
      "grad_norm": 5.666733741760254,
      "learning_rate": 2.8412698412698414e-05,
      "loss": 4.4831,
      "step": 409
    },
    {
      "epoch": 0.43386243386243384,
      "grad_norm": 6.003392696380615,
      "learning_rate": 2.8359788359788357e-05,
      "loss": 4.1948,
      "step": 410
    },
    {
      "epoch": 0.43492063492063493,
      "grad_norm": 3.3927369117736816,
      "learning_rate": 2.830687830687831e-05,
      "loss": 4.361,
      "step": 411
    },
    {
      "epoch": 0.43597883597883597,
      "grad_norm": 4.443639278411865,
      "learning_rate": 2.8253968253968253e-05,
      "loss": 4.1115,
      "step": 412
    },
    {
      "epoch": 0.43703703703703706,
      "grad_norm": 3.2199268341064453,
      "learning_rate": 2.82010582010582e-05,
      "loss": 4.5649,
      "step": 413
    },
    {
      "epoch": 0.4380952380952381,
      "grad_norm": 5.219230651855469,
      "learning_rate": 2.814814814814815e-05,
      "loss": 4.3149,
      "step": 414
    },
    {
      "epoch": 0.43915343915343913,
      "grad_norm": 4.251458644866943,
      "learning_rate": 2.8095238095238096e-05,
      "loss": 4.5838,
      "step": 415
    },
    {
      "epoch": 0.4402116402116402,
      "grad_norm": 3.8369529247283936,
      "learning_rate": 2.8042328042328043e-05,
      "loss": 4.6192,
      "step": 416
    },
    {
      "epoch": 0.44126984126984126,
      "grad_norm": 5.307433128356934,
      "learning_rate": 2.7989417989417993e-05,
      "loss": 4.434,
      "step": 417
    },
    {
      "epoch": 0.44232804232804235,
      "grad_norm": 4.317311763763428,
      "learning_rate": 2.793650793650794e-05,
      "loss": 4.4022,
      "step": 418
    },
    {
      "epoch": 0.4433862433862434,
      "grad_norm": 4.3074259757995605,
      "learning_rate": 2.7883597883597883e-05,
      "loss": 4.5237,
      "step": 419
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 5.363450050354004,
      "learning_rate": 2.783068783068783e-05,
      "loss": 4.6406,
      "step": 420
    },
    {
      "epoch": 0.4455026455026455,
      "grad_norm": 4.091902256011963,
      "learning_rate": 2.777777777777778e-05,
      "loss": 4.3964,
      "step": 421
    },
    {
      "epoch": 0.44656084656084655,
      "grad_norm": 4.199675559997559,
      "learning_rate": 2.7724867724867726e-05,
      "loss": 4.5985,
      "step": 422
    },
    {
      "epoch": 0.44761904761904764,
      "grad_norm": 3.6524128913879395,
      "learning_rate": 2.7671957671957672e-05,
      "loss": 4.7435,
      "step": 423
    },
    {
      "epoch": 0.4486772486772487,
      "grad_norm": 3.6887259483337402,
      "learning_rate": 2.7619047619047622e-05,
      "loss": 4.5237,
      "step": 424
    },
    {
      "epoch": 0.4497354497354497,
      "grad_norm": 5.04093599319458,
      "learning_rate": 2.756613756613757e-05,
      "loss": 4.3478,
      "step": 425
    },
    {
      "epoch": 0.4507936507936508,
      "grad_norm": 4.271005153656006,
      "learning_rate": 2.7513227513227512e-05,
      "loss": 4.2349,
      "step": 426
    },
    {
      "epoch": 0.45185185185185184,
      "grad_norm": 4.623360633850098,
      "learning_rate": 2.7460317460317465e-05,
      "loss": 4.0299,
      "step": 427
    },
    {
      "epoch": 0.45291005291005293,
      "grad_norm": 3.6036221981048584,
      "learning_rate": 2.7407407407407408e-05,
      "loss": 4.5202,
      "step": 428
    },
    {
      "epoch": 0.45396825396825397,
      "grad_norm": 5.633344650268555,
      "learning_rate": 2.7354497354497355e-05,
      "loss": 4.4543,
      "step": 429
    },
    {
      "epoch": 0.455026455026455,
      "grad_norm": 5.114171504974365,
      "learning_rate": 2.73015873015873e-05,
      "loss": 4.4791,
      "step": 430
    },
    {
      "epoch": 0.4560846560846561,
      "grad_norm": 4.855916500091553,
      "learning_rate": 2.724867724867725e-05,
      "loss": 4.4991,
      "step": 431
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 5.824271202087402,
      "learning_rate": 2.7195767195767198e-05,
      "loss": 4.248,
      "step": 432
    },
    {
      "epoch": 0.4582010582010582,
      "grad_norm": 4.154948711395264,
      "learning_rate": 2.714285714285714e-05,
      "loss": 4.2861,
      "step": 433
    },
    {
      "epoch": 0.45925925925925926,
      "grad_norm": 3.779582977294922,
      "learning_rate": 2.7089947089947094e-05,
      "loss": 4.2374,
      "step": 434
    },
    {
      "epoch": 0.4603174603174603,
      "grad_norm": 2.884605884552002,
      "learning_rate": 2.7037037037037037e-05,
      "loss": 4.2192,
      "step": 435
    },
    {
      "epoch": 0.4613756613756614,
      "grad_norm": 4.426102638244629,
      "learning_rate": 2.6984126984126984e-05,
      "loss": 4.23,
      "step": 436
    },
    {
      "epoch": 0.4624338624338624,
      "grad_norm": 4.26328706741333,
      "learning_rate": 2.6931216931216934e-05,
      "loss": 4.4972,
      "step": 437
    },
    {
      "epoch": 0.4634920634920635,
      "grad_norm": 4.900135517120361,
      "learning_rate": 2.687830687830688e-05,
      "loss": 4.3481,
      "step": 438
    },
    {
      "epoch": 0.46455026455026455,
      "grad_norm": 3.687103033065796,
      "learning_rate": 2.6825396825396827e-05,
      "loss": 4.5606,
      "step": 439
    },
    {
      "epoch": 0.4656084656084656,
      "grad_norm": 5.069578170776367,
      "learning_rate": 2.677248677248677e-05,
      "loss": 4.3747,
      "step": 440
    },
    {
      "epoch": 0.4666666666666667,
      "grad_norm": 3.736144542694092,
      "learning_rate": 2.6719576719576723e-05,
      "loss": 4.4117,
      "step": 441
    },
    {
      "epoch": 0.4677248677248677,
      "grad_norm": 3.2292275428771973,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 4.6203,
      "step": 442
    },
    {
      "epoch": 0.4687830687830688,
      "grad_norm": 4.4302778244018555,
      "learning_rate": 2.6613756613756613e-05,
      "loss": 4.3818,
      "step": 443
    },
    {
      "epoch": 0.46984126984126984,
      "grad_norm": 5.112982273101807,
      "learning_rate": 2.6560846560846563e-05,
      "loss": 4.2857,
      "step": 444
    },
    {
      "epoch": 0.4708994708994709,
      "grad_norm": 4.428035736083984,
      "learning_rate": 2.650793650793651e-05,
      "loss": 4.2864,
      "step": 445
    },
    {
      "epoch": 0.47195767195767196,
      "grad_norm": 2.9887120723724365,
      "learning_rate": 2.6455026455026456e-05,
      "loss": 4.4141,
      "step": 446
    },
    {
      "epoch": 0.473015873015873,
      "grad_norm": 4.006242275238037,
      "learning_rate": 2.6402116402116406e-05,
      "loss": 4.4684,
      "step": 447
    },
    {
      "epoch": 0.4740740740740741,
      "grad_norm": 6.761603355407715,
      "learning_rate": 2.6349206349206353e-05,
      "loss": 3.7916,
      "step": 448
    },
    {
      "epoch": 0.47513227513227513,
      "grad_norm": 4.260864734649658,
      "learning_rate": 2.6296296296296296e-05,
      "loss": 4.5701,
      "step": 449
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 4.10745906829834,
      "learning_rate": 2.6243386243386242e-05,
      "loss": 4.5473,
      "step": 450
    },
    {
      "epoch": 0.47724867724867726,
      "grad_norm": 5.057222843170166,
      "learning_rate": 2.6190476190476192e-05,
      "loss": 4.0021,
      "step": 451
    },
    {
      "epoch": 0.4783068783068783,
      "grad_norm": 3.276907205581665,
      "learning_rate": 2.613756613756614e-05,
      "loss": 4.3013,
      "step": 452
    },
    {
      "epoch": 0.4793650793650794,
      "grad_norm": 4.663476467132568,
      "learning_rate": 2.6084656084656085e-05,
      "loss": 4.0696,
      "step": 453
    },
    {
      "epoch": 0.4804232804232804,
      "grad_norm": 3.4896063804626465,
      "learning_rate": 2.6031746031746035e-05,
      "loss": 4.5279,
      "step": 454
    },
    {
      "epoch": 0.48148148148148145,
      "grad_norm": 4.717653274536133,
      "learning_rate": 2.5978835978835982e-05,
      "loss": 4.256,
      "step": 455
    },
    {
      "epoch": 0.48253968253968255,
      "grad_norm": 2.9412264823913574,
      "learning_rate": 2.5925925925925925e-05,
      "loss": 4.5379,
      "step": 456
    },
    {
      "epoch": 0.4835978835978836,
      "grad_norm": 5.3735671043396,
      "learning_rate": 2.5873015873015878e-05,
      "loss": 4.5428,
      "step": 457
    },
    {
      "epoch": 0.4846560846560847,
      "grad_norm": 3.9320805072784424,
      "learning_rate": 2.582010582010582e-05,
      "loss": 4.2526,
      "step": 458
    },
    {
      "epoch": 0.4857142857142857,
      "grad_norm": 5.301612854003906,
      "learning_rate": 2.5767195767195768e-05,
      "loss": 4.3464,
      "step": 459
    },
    {
      "epoch": 0.48677248677248675,
      "grad_norm": 6.028964996337891,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 4.2735,
      "step": 460
    },
    {
      "epoch": 0.48783068783068784,
      "grad_norm": 3.6883628368377686,
      "learning_rate": 2.5661375661375664e-05,
      "loss": 4.2878,
      "step": 461
    },
    {
      "epoch": 0.4888888888888889,
      "grad_norm": 3.1534693241119385,
      "learning_rate": 2.560846560846561e-05,
      "loss": 4.6231,
      "step": 462
    },
    {
      "epoch": 0.48994708994708996,
      "grad_norm": 3.1988255977630615,
      "learning_rate": 2.5555555555555554e-05,
      "loss": 4.4659,
      "step": 463
    },
    {
      "epoch": 0.491005291005291,
      "grad_norm": 6.167055606842041,
      "learning_rate": 2.5502645502645507e-05,
      "loss": 4.0693,
      "step": 464
    },
    {
      "epoch": 0.49206349206349204,
      "grad_norm": 4.961990833282471,
      "learning_rate": 2.544973544973545e-05,
      "loss": 4.6526,
      "step": 465
    },
    {
      "epoch": 0.4931216931216931,
      "grad_norm": 4.159661293029785,
      "learning_rate": 2.5396825396825397e-05,
      "loss": 4.3308,
      "step": 466
    },
    {
      "epoch": 0.49417989417989416,
      "grad_norm": 3.0661191940307617,
      "learning_rate": 2.5343915343915347e-05,
      "loss": 4.3929,
      "step": 467
    },
    {
      "epoch": 0.49523809523809526,
      "grad_norm": 3.240173816680908,
      "learning_rate": 2.5291005291005294e-05,
      "loss": 4.3422,
      "step": 468
    },
    {
      "epoch": 0.4962962962962963,
      "grad_norm": 3.9534902572631836,
      "learning_rate": 2.523809523809524e-05,
      "loss": 4.4378,
      "step": 469
    },
    {
      "epoch": 0.4973544973544973,
      "grad_norm": 4.109508037567139,
      "learning_rate": 2.5185185185185183e-05,
      "loss": 4.313,
      "step": 470
    },
    {
      "epoch": 0.4984126984126984,
      "grad_norm": 3.0627992153167725,
      "learning_rate": 2.5132275132275137e-05,
      "loss": 4.4831,
      "step": 471
    },
    {
      "epoch": 0.49947089947089945,
      "grad_norm": 3.873530149459839,
      "learning_rate": 2.507936507936508e-05,
      "loss": 4.4164,
      "step": 472
    },
    {
      "epoch": 0.5005291005291005,
      "grad_norm": 3.79267954826355,
      "learning_rate": 2.5026455026455026e-05,
      "loss": 4.4306,
      "step": 473
    },
    {
      "epoch": 0.5015873015873016,
      "grad_norm": 4.215601921081543,
      "learning_rate": 2.4973544973544973e-05,
      "loss": 4.4788,
      "step": 474
    },
    {
      "epoch": 0.5026455026455027,
      "grad_norm": 3.260300874710083,
      "learning_rate": 2.4920634920634923e-05,
      "loss": 4.3694,
      "step": 475
    },
    {
      "epoch": 0.5037037037037037,
      "grad_norm": 4.4810404777526855,
      "learning_rate": 2.4867724867724866e-05,
      "loss": 4.4525,
      "step": 476
    },
    {
      "epoch": 0.5047619047619047,
      "grad_norm": 4.280769348144531,
      "learning_rate": 2.4814814814814816e-05,
      "loss": 4.2847,
      "step": 477
    },
    {
      "epoch": 0.5058201058201058,
      "grad_norm": 5.8798298835754395,
      "learning_rate": 2.4761904761904762e-05,
      "loss": 4.3589,
      "step": 478
    },
    {
      "epoch": 0.5068783068783069,
      "grad_norm": 4.679714679718018,
      "learning_rate": 2.470899470899471e-05,
      "loss": 4.3346,
      "step": 479
    },
    {
      "epoch": 0.5079365079365079,
      "grad_norm": 7.5125322341918945,
      "learning_rate": 2.465608465608466e-05,
      "loss": 4.2475,
      "step": 480
    },
    {
      "epoch": 0.508994708994709,
      "grad_norm": 5.064920425415039,
      "learning_rate": 2.4603174603174602e-05,
      "loss": 4.2715,
      "step": 481
    },
    {
      "epoch": 0.5100529100529101,
      "grad_norm": 3.553298234939575,
      "learning_rate": 2.4550264550264552e-05,
      "loss": 4.3644,
      "step": 482
    },
    {
      "epoch": 0.5111111111111111,
      "grad_norm": 5.128535747528076,
      "learning_rate": 2.44973544973545e-05,
      "loss": 4.5012,
      "step": 483
    },
    {
      "epoch": 0.5121693121693122,
      "grad_norm": 6.287095546722412,
      "learning_rate": 2.4444444444444445e-05,
      "loss": 4.3774,
      "step": 484
    },
    {
      "epoch": 0.5132275132275133,
      "grad_norm": 6.201144218444824,
      "learning_rate": 2.439153439153439e-05,
      "loss": 4.542,
      "step": 485
    },
    {
      "epoch": 0.5142857142857142,
      "grad_norm": 5.993530750274658,
      "learning_rate": 2.4338624338624338e-05,
      "loss": 4.6664,
      "step": 486
    },
    {
      "epoch": 0.5153439153439153,
      "grad_norm": 5.866735458374023,
      "learning_rate": 2.4285714285714288e-05,
      "loss": 4.4555,
      "step": 487
    },
    {
      "epoch": 0.5164021164021164,
      "grad_norm": 4.489416599273682,
      "learning_rate": 2.4232804232804234e-05,
      "loss": 4.5308,
      "step": 488
    },
    {
      "epoch": 0.5174603174603175,
      "grad_norm": 6.804817199707031,
      "learning_rate": 2.417989417989418e-05,
      "loss": 4.2786,
      "step": 489
    },
    {
      "epoch": 0.5185185185185185,
      "grad_norm": 4.931401252746582,
      "learning_rate": 2.4126984126984128e-05,
      "loss": 4.3546,
      "step": 490
    },
    {
      "epoch": 0.5195767195767196,
      "grad_norm": 4.534653186798096,
      "learning_rate": 2.4074074074074074e-05,
      "loss": 4.4706,
      "step": 491
    },
    {
      "epoch": 0.5206349206349207,
      "grad_norm": 3.931093692779541,
      "learning_rate": 2.402116402116402e-05,
      "loss": 4.2993,
      "step": 492
    },
    {
      "epoch": 0.5216931216931217,
      "grad_norm": 4.178028106689453,
      "learning_rate": 2.396825396825397e-05,
      "loss": 4.3037,
      "step": 493
    },
    {
      "epoch": 0.5227513227513227,
      "grad_norm": 3.1427457332611084,
      "learning_rate": 2.3915343915343917e-05,
      "loss": 4.377,
      "step": 494
    },
    {
      "epoch": 0.5238095238095238,
      "grad_norm": 4.532064437866211,
      "learning_rate": 2.3862433862433864e-05,
      "loss": 4.5601,
      "step": 495
    },
    {
      "epoch": 0.5248677248677248,
      "grad_norm": 5.367987155914307,
      "learning_rate": 2.380952380952381e-05,
      "loss": 4.3662,
      "step": 496
    },
    {
      "epoch": 0.5259259259259259,
      "grad_norm": 4.349072456359863,
      "learning_rate": 2.3756613756613757e-05,
      "loss": 4.2585,
      "step": 497
    },
    {
      "epoch": 0.526984126984127,
      "grad_norm": 4.508454322814941,
      "learning_rate": 2.3703703703703707e-05,
      "loss": 4.3033,
      "step": 498
    },
    {
      "epoch": 0.5280423280423281,
      "grad_norm": 3.51326060295105,
      "learning_rate": 2.365079365079365e-05,
      "loss": 4.3074,
      "step": 499
    },
    {
      "epoch": 0.5291005291005291,
      "grad_norm": 4.635214328765869,
      "learning_rate": 2.35978835978836e-05,
      "loss": 4.5061,
      "step": 500
    },
    {
      "epoch": 0.5301587301587302,
      "grad_norm": 7.175034523010254,
      "learning_rate": 2.3544973544973546e-05,
      "loss": 4.5246,
      "step": 501
    },
    {
      "epoch": 0.5312169312169313,
      "grad_norm": 7.556386947631836,
      "learning_rate": 2.3492063492063493e-05,
      "loss": 4.5628,
      "step": 502
    },
    {
      "epoch": 0.5322751322751322,
      "grad_norm": 5.281272888183594,
      "learning_rate": 2.3439153439153443e-05,
      "loss": 4.5511,
      "step": 503
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 4.228816986083984,
      "learning_rate": 2.3386243386243386e-05,
      "loss": 4.7682,
      "step": 504
    },
    {
      "epoch": 0.5343915343915344,
      "grad_norm": 4.464179992675781,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 4.2103,
      "step": 505
    },
    {
      "epoch": 0.5354497354497354,
      "grad_norm": 2.9484353065490723,
      "learning_rate": 2.328042328042328e-05,
      "loss": 4.3896,
      "step": 506
    },
    {
      "epoch": 0.5365079365079365,
      "grad_norm": 3.5408449172973633,
      "learning_rate": 2.322751322751323e-05,
      "loss": 4.5458,
      "step": 507
    },
    {
      "epoch": 0.5375661375661376,
      "grad_norm": 4.470260143280029,
      "learning_rate": 2.3174603174603175e-05,
      "loss": 4.3186,
      "step": 508
    },
    {
      "epoch": 0.5386243386243387,
      "grad_norm": 4.546544551849365,
      "learning_rate": 2.3121693121693122e-05,
      "loss": 4.6774,
      "step": 509
    },
    {
      "epoch": 0.5396825396825397,
      "grad_norm": 3.3700125217437744,
      "learning_rate": 2.3068783068783072e-05,
      "loss": 4.3627,
      "step": 510
    },
    {
      "epoch": 0.5407407407407407,
      "grad_norm": 4.171989440917969,
      "learning_rate": 2.3015873015873015e-05,
      "loss": 4.5353,
      "step": 511
    },
    {
      "epoch": 0.5417989417989418,
      "grad_norm": 3.360245943069458,
      "learning_rate": 2.2962962962962965e-05,
      "loss": 4.2469,
      "step": 512
    },
    {
      "epoch": 0.5428571428571428,
      "grad_norm": 4.016007900238037,
      "learning_rate": 2.291005291005291e-05,
      "loss": 4.7331,
      "step": 513
    },
    {
      "epoch": 0.5439153439153439,
      "grad_norm": 4.515334606170654,
      "learning_rate": 2.2857142857142858e-05,
      "loss": 4.1892,
      "step": 514
    },
    {
      "epoch": 0.544973544973545,
      "grad_norm": 5.877944469451904,
      "learning_rate": 2.2804232804232805e-05,
      "loss": 3.928,
      "step": 515
    },
    {
      "epoch": 0.546031746031746,
      "grad_norm": 4.498457908630371,
      "learning_rate": 2.275132275132275e-05,
      "loss": 4.667,
      "step": 516
    },
    {
      "epoch": 0.5470899470899471,
      "grad_norm": 4.965441703796387,
      "learning_rate": 2.2698412698412698e-05,
      "loss": 4.2019,
      "step": 517
    },
    {
      "epoch": 0.5481481481481482,
      "grad_norm": 3.545275926589966,
      "learning_rate": 2.2645502645502648e-05,
      "loss": 4.3709,
      "step": 518
    },
    {
      "epoch": 0.5492063492063493,
      "grad_norm": 4.3779191970825195,
      "learning_rate": 2.2592592592592594e-05,
      "loss": 4.4913,
      "step": 519
    },
    {
      "epoch": 0.5502645502645502,
      "grad_norm": 4.215112209320068,
      "learning_rate": 2.253968253968254e-05,
      "loss": 4.5574,
      "step": 520
    },
    {
      "epoch": 0.5513227513227513,
      "grad_norm": 5.649901866912842,
      "learning_rate": 2.2486772486772487e-05,
      "loss": 4.2828,
      "step": 521
    },
    {
      "epoch": 0.5523809523809524,
      "grad_norm": 3.5006558895111084,
      "learning_rate": 2.2433862433862434e-05,
      "loss": 4.4051,
      "step": 522
    },
    {
      "epoch": 0.5534391534391534,
      "grad_norm": 3.895487070083618,
      "learning_rate": 2.2380952380952384e-05,
      "loss": 4.4244,
      "step": 523
    },
    {
      "epoch": 0.5544973544973545,
      "grad_norm": 3.6373565196990967,
      "learning_rate": 2.2328042328042327e-05,
      "loss": 4.363,
      "step": 524
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 3.6034340858459473,
      "learning_rate": 2.2275132275132277e-05,
      "loss": 4.5101,
      "step": 525
    },
    {
      "epoch": 0.5566137566137566,
      "grad_norm": 6.335474967956543,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 4.2665,
      "step": 526
    },
    {
      "epoch": 0.5576719576719577,
      "grad_norm": 3.84147047996521,
      "learning_rate": 2.216931216931217e-05,
      "loss": 4.0104,
      "step": 527
    },
    {
      "epoch": 0.5587301587301587,
      "grad_norm": 5.233031272888184,
      "learning_rate": 2.211640211640212e-05,
      "loss": 4.6061,
      "step": 528
    },
    {
      "epoch": 0.5597883597883598,
      "grad_norm": 5.9673638343811035,
      "learning_rate": 2.2063492063492063e-05,
      "loss": 4.3624,
      "step": 529
    },
    {
      "epoch": 0.5608465608465608,
      "grad_norm": 6.370858669281006,
      "learning_rate": 2.2010582010582013e-05,
      "loss": 4.1914,
      "step": 530
    },
    {
      "epoch": 0.5619047619047619,
      "grad_norm": 4.488510608673096,
      "learning_rate": 2.1957671957671956e-05,
      "loss": 4.4786,
      "step": 531
    },
    {
      "epoch": 0.562962962962963,
      "grad_norm": 5.369960308074951,
      "learning_rate": 2.1904761904761906e-05,
      "loss": 4.4175,
      "step": 532
    },
    {
      "epoch": 0.564021164021164,
      "grad_norm": 4.92180061340332,
      "learning_rate": 2.1851851851851852e-05,
      "loss": 4.4534,
      "step": 533
    },
    {
      "epoch": 0.5650793650793651,
      "grad_norm": 4.838220596313477,
      "learning_rate": 2.17989417989418e-05,
      "loss": 4.5532,
      "step": 534
    },
    {
      "epoch": 0.5661375661375662,
      "grad_norm": 5.853415489196777,
      "learning_rate": 2.174603174603175e-05,
      "loss": 4.1377,
      "step": 535
    },
    {
      "epoch": 0.5671957671957671,
      "grad_norm": 4.8340349197387695,
      "learning_rate": 2.1693121693121692e-05,
      "loss": 4.1588,
      "step": 536
    },
    {
      "epoch": 0.5682539682539682,
      "grad_norm": 3.3916378021240234,
      "learning_rate": 2.1640211640211642e-05,
      "loss": 4.3666,
      "step": 537
    },
    {
      "epoch": 0.5693121693121693,
      "grad_norm": 3.867189407348633,
      "learning_rate": 2.158730158730159e-05,
      "loss": 4.2795,
      "step": 538
    },
    {
      "epoch": 0.5703703703703704,
      "grad_norm": 4.104162693023682,
      "learning_rate": 2.1534391534391535e-05,
      "loss": 4.3084,
      "step": 539
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 3.797863721847534,
      "learning_rate": 2.148148148148148e-05,
      "loss": 4.3259,
      "step": 540
    },
    {
      "epoch": 0.5724867724867725,
      "grad_norm": 6.428219795227051,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 4.502,
      "step": 541
    },
    {
      "epoch": 0.5735449735449736,
      "grad_norm": 3.7687265872955322,
      "learning_rate": 2.1375661375661378e-05,
      "loss": 4.2587,
      "step": 542
    },
    {
      "epoch": 0.5746031746031746,
      "grad_norm": 4.52725887298584,
      "learning_rate": 2.1322751322751325e-05,
      "loss": 4.6748,
      "step": 543
    },
    {
      "epoch": 0.5756613756613757,
      "grad_norm": 7.533745288848877,
      "learning_rate": 2.126984126984127e-05,
      "loss": 4.4511,
      "step": 544
    },
    {
      "epoch": 0.5767195767195767,
      "grad_norm": 6.036813735961914,
      "learning_rate": 2.1216931216931218e-05,
      "loss": 4.6686,
      "step": 545
    },
    {
      "epoch": 0.5777777777777777,
      "grad_norm": 5.2509331703186035,
      "learning_rate": 2.1164021164021164e-05,
      "loss": 4.1922,
      "step": 546
    },
    {
      "epoch": 0.5788359788359788,
      "grad_norm": 4.836385726928711,
      "learning_rate": 2.111111111111111e-05,
      "loss": 4.3825,
      "step": 547
    },
    {
      "epoch": 0.5798941798941799,
      "grad_norm": 3.305849075317383,
      "learning_rate": 2.105820105820106e-05,
      "loss": 4.4511,
      "step": 548
    },
    {
      "epoch": 0.580952380952381,
      "grad_norm": 6.00985050201416,
      "learning_rate": 2.1005291005291007e-05,
      "loss": 4.6058,
      "step": 549
    },
    {
      "epoch": 0.582010582010582,
      "grad_norm": 3.1658153533935547,
      "learning_rate": 2.0952380952380954e-05,
      "loss": 4.4339,
      "step": 550
    },
    {
      "epoch": 0.5830687830687831,
      "grad_norm": 3.461515426635742,
      "learning_rate": 2.08994708994709e-05,
      "loss": 4.1928,
      "step": 551
    },
    {
      "epoch": 0.5841269841269842,
      "grad_norm": 3.686345100402832,
      "learning_rate": 2.0846560846560847e-05,
      "loss": 4.2196,
      "step": 552
    },
    {
      "epoch": 0.5851851851851851,
      "grad_norm": 4.964058876037598,
      "learning_rate": 2.0793650793650797e-05,
      "loss": 4.4196,
      "step": 553
    },
    {
      "epoch": 0.5862433862433862,
      "grad_norm": 3.2014639377593994,
      "learning_rate": 2.074074074074074e-05,
      "loss": 4.3415,
      "step": 554
    },
    {
      "epoch": 0.5873015873015873,
      "grad_norm": 5.361993789672852,
      "learning_rate": 2.068783068783069e-05,
      "loss": 4.3946,
      "step": 555
    },
    {
      "epoch": 0.5883597883597883,
      "grad_norm": 6.569308757781982,
      "learning_rate": 2.0634920634920636e-05,
      "loss": 4.1728,
      "step": 556
    },
    {
      "epoch": 0.5894179894179894,
      "grad_norm": 4.793816089630127,
      "learning_rate": 2.0582010582010583e-05,
      "loss": 4.5882,
      "step": 557
    },
    {
      "epoch": 0.5904761904761905,
      "grad_norm": 3.3498852252960205,
      "learning_rate": 2.052910052910053e-05,
      "loss": 4.3371,
      "step": 558
    },
    {
      "epoch": 0.5915343915343916,
      "grad_norm": 3.4059863090515137,
      "learning_rate": 2.0476190476190476e-05,
      "loss": 4.384,
      "step": 559
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 5.030102729797363,
      "learning_rate": 2.0423280423280426e-05,
      "loss": 4.3385,
      "step": 560
    },
    {
      "epoch": 0.5936507936507937,
      "grad_norm": 3.9506351947784424,
      "learning_rate": 2.037037037037037e-05,
      "loss": 4.3949,
      "step": 561
    },
    {
      "epoch": 0.5947089947089947,
      "grad_norm": 4.350306034088135,
      "learning_rate": 2.031746031746032e-05,
      "loss": 4.2672,
      "step": 562
    },
    {
      "epoch": 0.5957671957671957,
      "grad_norm": 3.6148154735565186,
      "learning_rate": 2.0264550264550266e-05,
      "loss": 4.3542,
      "step": 563
    },
    {
      "epoch": 0.5968253968253968,
      "grad_norm": 4.015986442565918,
      "learning_rate": 2.0211640211640212e-05,
      "loss": 4.3468,
      "step": 564
    },
    {
      "epoch": 0.5978835978835979,
      "grad_norm": 4.623772621154785,
      "learning_rate": 2.015873015873016e-05,
      "loss": 4.3602,
      "step": 565
    },
    {
      "epoch": 0.5989417989417989,
      "grad_norm": 4.602967739105225,
      "learning_rate": 2.0105820105820105e-05,
      "loss": 4.2878,
      "step": 566
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.3940911293029785,
      "learning_rate": 2.0052910052910055e-05,
      "loss": 4.4939,
      "step": 567
    },
    {
      "epoch": 0.6010582010582011,
      "grad_norm": 6.775349140167236,
      "learning_rate": 2e-05,
      "loss": 4.2697,
      "step": 568
    },
    {
      "epoch": 0.6021164021164022,
      "grad_norm": 3.592540979385376,
      "learning_rate": 1.9947089947089948e-05,
      "loss": 4.3831,
      "step": 569
    },
    {
      "epoch": 0.6031746031746031,
      "grad_norm": 3.6361138820648193,
      "learning_rate": 1.9894179894179895e-05,
      "loss": 4.459,
      "step": 570
    },
    {
      "epoch": 0.6042328042328042,
      "grad_norm": 5.545779228210449,
      "learning_rate": 1.984126984126984e-05,
      "loss": 4.5091,
      "step": 571
    },
    {
      "epoch": 0.6052910052910053,
      "grad_norm": 4.679067134857178,
      "learning_rate": 1.9788359788359788e-05,
      "loss": 4.1179,
      "step": 572
    },
    {
      "epoch": 0.6063492063492063,
      "grad_norm": 4.026285171508789,
      "learning_rate": 1.9735449735449738e-05,
      "loss": 4.4972,
      "step": 573
    },
    {
      "epoch": 0.6074074074074074,
      "grad_norm": 4.826168060302734,
      "learning_rate": 1.9682539682539684e-05,
      "loss": 4.326,
      "step": 574
    },
    {
      "epoch": 0.6084656084656085,
      "grad_norm": 4.132863521575928,
      "learning_rate": 1.962962962962963e-05,
      "loss": 4.4658,
      "step": 575
    },
    {
      "epoch": 0.6095238095238096,
      "grad_norm": 3.1971607208251953,
      "learning_rate": 1.9576719576719577e-05,
      "loss": 4.3489,
      "step": 576
    },
    {
      "epoch": 0.6105820105820106,
      "grad_norm": 4.324203014373779,
      "learning_rate": 1.9523809523809524e-05,
      "loss": 4.5387,
      "step": 577
    },
    {
      "epoch": 0.6116402116402117,
      "grad_norm": 4.452277660369873,
      "learning_rate": 1.9470899470899474e-05,
      "loss": 4.371,
      "step": 578
    },
    {
      "epoch": 0.6126984126984127,
      "grad_norm": 5.406804084777832,
      "learning_rate": 1.9417989417989417e-05,
      "loss": 4.3936,
      "step": 579
    },
    {
      "epoch": 0.6137566137566137,
      "grad_norm": 3.683032751083374,
      "learning_rate": 1.9365079365079367e-05,
      "loss": 4.5795,
      "step": 580
    },
    {
      "epoch": 0.6148148148148148,
      "grad_norm": 4.9315667152404785,
      "learning_rate": 1.9312169312169313e-05,
      "loss": 4.3404,
      "step": 581
    },
    {
      "epoch": 0.6158730158730159,
      "grad_norm": 3.669973611831665,
      "learning_rate": 1.925925925925926e-05,
      "loss": 4.479,
      "step": 582
    },
    {
      "epoch": 0.6169312169312169,
      "grad_norm": 6.914597034454346,
      "learning_rate": 1.920634920634921e-05,
      "loss": 5.2586,
      "step": 583
    },
    {
      "epoch": 0.617989417989418,
      "grad_norm": 6.880323886871338,
      "learning_rate": 1.9153439153439153e-05,
      "loss": 4.2503,
      "step": 584
    },
    {
      "epoch": 0.6190476190476191,
      "grad_norm": 3.348428726196289,
      "learning_rate": 1.9100529100529103e-05,
      "loss": 4.2295,
      "step": 585
    },
    {
      "epoch": 0.6201058201058202,
      "grad_norm": 3.1387596130371094,
      "learning_rate": 1.9047619047619046e-05,
      "loss": 4.437,
      "step": 586
    },
    {
      "epoch": 0.6211640211640211,
      "grad_norm": 3.1645758152008057,
      "learning_rate": 1.8994708994708996e-05,
      "loss": 4.3887,
      "step": 587
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 3.8095288276672363,
      "learning_rate": 1.8941798941798943e-05,
      "loss": 4.3497,
      "step": 588
    },
    {
      "epoch": 0.6232804232804233,
      "grad_norm": 3.393749237060547,
      "learning_rate": 1.888888888888889e-05,
      "loss": 4.2098,
      "step": 589
    },
    {
      "epoch": 0.6243386243386243,
      "grad_norm": 5.955323696136475,
      "learning_rate": 1.883597883597884e-05,
      "loss": 4.3409,
      "step": 590
    },
    {
      "epoch": 0.6253968253968254,
      "grad_norm": 3.0141565799713135,
      "learning_rate": 1.8783068783068782e-05,
      "loss": 4.35,
      "step": 591
    },
    {
      "epoch": 0.6264550264550265,
      "grad_norm": 5.6626176834106445,
      "learning_rate": 1.8730158730158732e-05,
      "loss": 4.3674,
      "step": 592
    },
    {
      "epoch": 0.6275132275132275,
      "grad_norm": 3.7289211750030518,
      "learning_rate": 1.867724867724868e-05,
      "loss": 4.2095,
      "step": 593
    },
    {
      "epoch": 0.6285714285714286,
      "grad_norm": 4.1014862060546875,
      "learning_rate": 1.8624338624338625e-05,
      "loss": 4.2905,
      "step": 594
    },
    {
      "epoch": 0.6296296296296297,
      "grad_norm": 5.644645690917969,
      "learning_rate": 1.8571428571428572e-05,
      "loss": 4.2984,
      "step": 595
    },
    {
      "epoch": 0.6306878306878307,
      "grad_norm": 5.378518104553223,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 4.1897,
      "step": 596
    },
    {
      "epoch": 0.6317460317460317,
      "grad_norm": 4.600176811218262,
      "learning_rate": 1.8465608465608468e-05,
      "loss": 4.2564,
      "step": 597
    },
    {
      "epoch": 0.6328042328042328,
      "grad_norm": 6.3858137130737305,
      "learning_rate": 1.8412698412698415e-05,
      "loss": 4.1599,
      "step": 598
    },
    {
      "epoch": 0.6338624338624339,
      "grad_norm": 4.172644138336182,
      "learning_rate": 1.835978835978836e-05,
      "loss": 4.2698,
      "step": 599
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 4.711319923400879,
      "learning_rate": 1.8306878306878308e-05,
      "loss": 4.4446,
      "step": 600
    },
    {
      "epoch": 0.635978835978836,
      "grad_norm": 4.155757427215576,
      "learning_rate": 1.8253968253968254e-05,
      "loss": 4.4898,
      "step": 601
    },
    {
      "epoch": 0.6370370370370371,
      "grad_norm": 4.684317588806152,
      "learning_rate": 1.82010582010582e-05,
      "loss": 4.2174,
      "step": 602
    },
    {
      "epoch": 0.638095238095238,
      "grad_norm": 4.73888635635376,
      "learning_rate": 1.814814814814815e-05,
      "loss": 4.397,
      "step": 603
    },
    {
      "epoch": 0.6391534391534391,
      "grad_norm": 4.162045478820801,
      "learning_rate": 1.8095238095238094e-05,
      "loss": 4.4202,
      "step": 604
    },
    {
      "epoch": 0.6402116402116402,
      "grad_norm": 6.125421524047852,
      "learning_rate": 1.8042328042328044e-05,
      "loss": 4.2264,
      "step": 605
    },
    {
      "epoch": 0.6412698412698413,
      "grad_norm": 5.661704063415527,
      "learning_rate": 1.798941798941799e-05,
      "loss": 4.0215,
      "step": 606
    },
    {
      "epoch": 0.6423280423280423,
      "grad_norm": 6.217195510864258,
      "learning_rate": 1.7936507936507937e-05,
      "loss": 4.2445,
      "step": 607
    },
    {
      "epoch": 0.6433862433862434,
      "grad_norm": 4.448593616485596,
      "learning_rate": 1.7883597883597887e-05,
      "loss": 3.9882,
      "step": 608
    },
    {
      "epoch": 0.6444444444444445,
      "grad_norm": 6.415726184844971,
      "learning_rate": 1.783068783068783e-05,
      "loss": 4.4289,
      "step": 609
    },
    {
      "epoch": 0.6455026455026455,
      "grad_norm": 4.124817848205566,
      "learning_rate": 1.777777777777778e-05,
      "loss": 4.3764,
      "step": 610
    },
    {
      "epoch": 0.6465608465608466,
      "grad_norm": 4.285531520843506,
      "learning_rate": 1.7724867724867723e-05,
      "loss": 4.4348,
      "step": 611
    },
    {
      "epoch": 0.6476190476190476,
      "grad_norm": 4.148486614227295,
      "learning_rate": 1.7671957671957673e-05,
      "loss": 4.448,
      "step": 612
    },
    {
      "epoch": 0.6486772486772486,
      "grad_norm": 6.1382060050964355,
      "learning_rate": 1.761904761904762e-05,
      "loss": 4.249,
      "step": 613
    },
    {
      "epoch": 0.6497354497354497,
      "grad_norm": 4.567010879516602,
      "learning_rate": 1.7566137566137566e-05,
      "loss": 4.2096,
      "step": 614
    },
    {
      "epoch": 0.6507936507936508,
      "grad_norm": 7.315486431121826,
      "learning_rate": 1.7513227513227516e-05,
      "loss": 4.1564,
      "step": 615
    },
    {
      "epoch": 0.6518518518518519,
      "grad_norm": 5.819315433502197,
      "learning_rate": 1.746031746031746e-05,
      "loss": 4.6175,
      "step": 616
    },
    {
      "epoch": 0.6529100529100529,
      "grad_norm": 3.4257044792175293,
      "learning_rate": 1.740740740740741e-05,
      "loss": 4.4864,
      "step": 617
    },
    {
      "epoch": 0.653968253968254,
      "grad_norm": 4.5550947189331055,
      "learning_rate": 1.7354497354497356e-05,
      "loss": 4.3435,
      "step": 618
    },
    {
      "epoch": 0.6550264550264551,
      "grad_norm": 5.184391498565674,
      "learning_rate": 1.7301587301587302e-05,
      "loss": 4.294,
      "step": 619
    },
    {
      "epoch": 0.656084656084656,
      "grad_norm": 5.654160022735596,
      "learning_rate": 1.724867724867725e-05,
      "loss": 4.3935,
      "step": 620
    },
    {
      "epoch": 0.6571428571428571,
      "grad_norm": 5.634194850921631,
      "learning_rate": 1.7195767195767195e-05,
      "loss": 4.0203,
      "step": 621
    },
    {
      "epoch": 0.6582010582010582,
      "grad_norm": 5.462583065032959,
      "learning_rate": 1.7142857142857145e-05,
      "loss": 4.4987,
      "step": 622
    },
    {
      "epoch": 0.6592592592592592,
      "grad_norm": 4.333361625671387,
      "learning_rate": 1.7089947089947092e-05,
      "loss": 4.4802,
      "step": 623
    },
    {
      "epoch": 0.6603174603174603,
      "grad_norm": 3.9108376502990723,
      "learning_rate": 1.7037037037037038e-05,
      "loss": 4.2963,
      "step": 624
    },
    {
      "epoch": 0.6613756613756614,
      "grad_norm": 4.688434600830078,
      "learning_rate": 1.6984126984126985e-05,
      "loss": 4.2159,
      "step": 625
    },
    {
      "epoch": 0.6624338624338625,
      "grad_norm": 3.797638177871704,
      "learning_rate": 1.693121693121693e-05,
      "loss": 4.3626,
      "step": 626
    },
    {
      "epoch": 0.6634920634920635,
      "grad_norm": 3.1089894771575928,
      "learning_rate": 1.6878306878306878e-05,
      "loss": 4.4314,
      "step": 627
    },
    {
      "epoch": 0.6645502645502646,
      "grad_norm": 3.5496623516082764,
      "learning_rate": 1.6825396825396828e-05,
      "loss": 4.4034,
      "step": 628
    },
    {
      "epoch": 0.6656084656084656,
      "grad_norm": 3.350839138031006,
      "learning_rate": 1.6772486772486774e-05,
      "loss": 4.4295,
      "step": 629
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 5.997464179992676,
      "learning_rate": 1.671957671957672e-05,
      "loss": 4.4228,
      "step": 630
    },
    {
      "epoch": 0.6677248677248677,
      "grad_norm": 4.045958042144775,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 4.507,
      "step": 631
    },
    {
      "epoch": 0.6687830687830688,
      "grad_norm": 6.075403690338135,
      "learning_rate": 1.6613756613756614e-05,
      "loss": 4.4149,
      "step": 632
    },
    {
      "epoch": 0.6698412698412698,
      "grad_norm": 4.632997512817383,
      "learning_rate": 1.6560846560846564e-05,
      "loss": 4.2041,
      "step": 633
    },
    {
      "epoch": 0.6708994708994709,
      "grad_norm": 6.3005499839782715,
      "learning_rate": 1.6507936507936507e-05,
      "loss": 4.2768,
      "step": 634
    },
    {
      "epoch": 0.671957671957672,
      "grad_norm": 3.847238063812256,
      "learning_rate": 1.6455026455026457e-05,
      "loss": 4.4107,
      "step": 635
    },
    {
      "epoch": 0.6730158730158731,
      "grad_norm": 3.6848814487457275,
      "learning_rate": 1.6402116402116404e-05,
      "loss": 4.2626,
      "step": 636
    },
    {
      "epoch": 0.674074074074074,
      "grad_norm": 6.838409900665283,
      "learning_rate": 1.634920634920635e-05,
      "loss": 4.3674,
      "step": 637
    },
    {
      "epoch": 0.6751322751322751,
      "grad_norm": 6.633561134338379,
      "learning_rate": 1.62962962962963e-05,
      "loss": 4.626,
      "step": 638
    },
    {
      "epoch": 0.6761904761904762,
      "grad_norm": 3.2217533588409424,
      "learning_rate": 1.6243386243386243e-05,
      "loss": 4.4099,
      "step": 639
    },
    {
      "epoch": 0.6772486772486772,
      "grad_norm": 4.3829145431518555,
      "learning_rate": 1.6190476190476193e-05,
      "loss": 4.2519,
      "step": 640
    },
    {
      "epoch": 0.6783068783068783,
      "grad_norm": 4.161040306091309,
      "learning_rate": 1.6137566137566136e-05,
      "loss": 4.4131,
      "step": 641
    },
    {
      "epoch": 0.6793650793650794,
      "grad_norm": 5.863997459411621,
      "learning_rate": 1.6084656084656086e-05,
      "loss": 3.9776,
      "step": 642
    },
    {
      "epoch": 0.6804232804232804,
      "grad_norm": 4.011388301849365,
      "learning_rate": 1.6031746031746033e-05,
      "loss": 4.2827,
      "step": 643
    },
    {
      "epoch": 0.6814814814814815,
      "grad_norm": 5.71457052230835,
      "learning_rate": 1.597883597883598e-05,
      "loss": 4.1867,
      "step": 644
    },
    {
      "epoch": 0.6825396825396826,
      "grad_norm": 2.9896559715270996,
      "learning_rate": 1.5925925925925926e-05,
      "loss": 4.2617,
      "step": 645
    },
    {
      "epoch": 0.6835978835978836,
      "grad_norm": 3.395284652709961,
      "learning_rate": 1.5873015873015872e-05,
      "loss": 4.4314,
      "step": 646
    },
    {
      "epoch": 0.6846560846560846,
      "grad_norm": 5.6781086921691895,
      "learning_rate": 1.5820105820105822e-05,
      "loss": 4.1137,
      "step": 647
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 5.346649646759033,
      "learning_rate": 1.576719576719577e-05,
      "loss": 4.186,
      "step": 648
    },
    {
      "epoch": 0.6867724867724868,
      "grad_norm": 4.502099990844727,
      "learning_rate": 1.5714285714285715e-05,
      "loss": 4.2691,
      "step": 649
    },
    {
      "epoch": 0.6878306878306878,
      "grad_norm": 4.361939907073975,
      "learning_rate": 1.5661375661375662e-05,
      "loss": 4.3178,
      "step": 650
    },
    {
      "epoch": 0.6888888888888889,
      "grad_norm": 4.989232540130615,
      "learning_rate": 1.560846560846561e-05,
      "loss": 4.5618,
      "step": 651
    },
    {
      "epoch": 0.68994708994709,
      "grad_norm": 4.986234188079834,
      "learning_rate": 1.5555555555555555e-05,
      "loss": 4.141,
      "step": 652
    },
    {
      "epoch": 0.691005291005291,
      "grad_norm": 4.698188781738281,
      "learning_rate": 1.5502645502645505e-05,
      "loss": 4.2413,
      "step": 653
    },
    {
      "epoch": 0.692063492063492,
      "grad_norm": 3.489686965942383,
      "learning_rate": 1.544973544973545e-05,
      "loss": 4.1997,
      "step": 654
    },
    {
      "epoch": 0.6931216931216931,
      "grad_norm": 5.456641674041748,
      "learning_rate": 1.5396825396825398e-05,
      "loss": 3.8636,
      "step": 655
    },
    {
      "epoch": 0.6941798941798942,
      "grad_norm": 5.964907169342041,
      "learning_rate": 1.5343915343915344e-05,
      "loss": 3.9141,
      "step": 656
    },
    {
      "epoch": 0.6952380952380952,
      "grad_norm": 4.022069931030273,
      "learning_rate": 1.529100529100529e-05,
      "loss": 4.222,
      "step": 657
    },
    {
      "epoch": 0.6962962962962963,
      "grad_norm": 4.964052677154541,
      "learning_rate": 1.5238095238095241e-05,
      "loss": 4.4431,
      "step": 658
    },
    {
      "epoch": 0.6973544973544974,
      "grad_norm": 4.9138712882995605,
      "learning_rate": 1.5185185185185186e-05,
      "loss": 4.3141,
      "step": 659
    },
    {
      "epoch": 0.6984126984126984,
      "grad_norm": 5.288991928100586,
      "learning_rate": 1.5132275132275134e-05,
      "loss": 4.29,
      "step": 660
    },
    {
      "epoch": 0.6994708994708995,
      "grad_norm": 2.88527512550354,
      "learning_rate": 1.5079365079365079e-05,
      "loss": 4.4974,
      "step": 661
    },
    {
      "epoch": 0.7005291005291006,
      "grad_norm": 3.6969175338745117,
      "learning_rate": 1.5026455026455027e-05,
      "loss": 4.3049,
      "step": 662
    },
    {
      "epoch": 0.7015873015873015,
      "grad_norm": 3.2847836017608643,
      "learning_rate": 1.4973544973544974e-05,
      "loss": 4.1501,
      "step": 663
    },
    {
      "epoch": 0.7026455026455026,
      "grad_norm": 4.742918014526367,
      "learning_rate": 1.4920634920634922e-05,
      "loss": 4.1301,
      "step": 664
    },
    {
      "epoch": 0.7037037037037037,
      "grad_norm": 6.213130474090576,
      "learning_rate": 1.486772486772487e-05,
      "loss": 3.9858,
      "step": 665
    },
    {
      "epoch": 0.7047619047619048,
      "grad_norm": 8.191361427307129,
      "learning_rate": 1.4814814814814815e-05,
      "loss": 5.2602,
      "step": 666
    },
    {
      "epoch": 0.7058201058201058,
      "grad_norm": 5.620320796966553,
      "learning_rate": 1.4761904761904763e-05,
      "loss": 4.3565,
      "step": 667
    },
    {
      "epoch": 0.7068783068783069,
      "grad_norm": 3.7786309719085693,
      "learning_rate": 1.4708994708994708e-05,
      "loss": 4.3774,
      "step": 668
    },
    {
      "epoch": 0.707936507936508,
      "grad_norm": 6.146083831787109,
      "learning_rate": 1.4656084656084656e-05,
      "loss": 4.4039,
      "step": 669
    },
    {
      "epoch": 0.708994708994709,
      "grad_norm": 4.173206329345703,
      "learning_rate": 1.4603174603174605e-05,
      "loss": 4.495,
      "step": 670
    },
    {
      "epoch": 0.71005291005291,
      "grad_norm": 4.497081279754639,
      "learning_rate": 1.455026455026455e-05,
      "loss": 4.2545,
      "step": 671
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 6.875660419464111,
      "learning_rate": 1.4497354497354498e-05,
      "loss": 4.1121,
      "step": 672
    },
    {
      "epoch": 0.7121693121693121,
      "grad_norm": 5.249729633331299,
      "learning_rate": 1.4444444444444444e-05,
      "loss": 4.4577,
      "step": 673
    },
    {
      "epoch": 0.7132275132275132,
      "grad_norm": 6.506221294403076,
      "learning_rate": 1.4391534391534392e-05,
      "loss": 4.5862,
      "step": 674
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 3.437570571899414,
      "learning_rate": 1.433862433862434e-05,
      "loss": 4.2684,
      "step": 675
    },
    {
      "epoch": 0.7153439153439154,
      "grad_norm": 3.463667631149292,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 4.5315,
      "step": 676
    },
    {
      "epoch": 0.7164021164021164,
      "grad_norm": 5.436329364776611,
      "learning_rate": 1.4232804232804234e-05,
      "loss": 4.5016,
      "step": 677
    },
    {
      "epoch": 0.7174603174603175,
      "grad_norm": 3.9234187602996826,
      "learning_rate": 1.4179894179894179e-05,
      "loss": 4.5041,
      "step": 678
    },
    {
      "epoch": 0.7185185185185186,
      "grad_norm": 3.112135410308838,
      "learning_rate": 1.4126984126984127e-05,
      "loss": 4.3273,
      "step": 679
    },
    {
      "epoch": 0.7195767195767195,
      "grad_norm": 3.713372230529785,
      "learning_rate": 1.4074074074074075e-05,
      "loss": 4.3923,
      "step": 680
    },
    {
      "epoch": 0.7206349206349206,
      "grad_norm": 5.0203375816345215,
      "learning_rate": 1.4021164021164022e-05,
      "loss": 4.4324,
      "step": 681
    },
    {
      "epoch": 0.7216931216931217,
      "grad_norm": 5.588266849517822,
      "learning_rate": 1.396825396825397e-05,
      "loss": 4.6234,
      "step": 682
    },
    {
      "epoch": 0.7227513227513227,
      "grad_norm": 6.488243103027344,
      "learning_rate": 1.3915343915343915e-05,
      "loss": 4.9581,
      "step": 683
    },
    {
      "epoch": 0.7238095238095238,
      "grad_norm": 5.537155628204346,
      "learning_rate": 1.3862433862433863e-05,
      "loss": 4.4419,
      "step": 684
    },
    {
      "epoch": 0.7248677248677249,
      "grad_norm": 4.607074737548828,
      "learning_rate": 1.3809523809523811e-05,
      "loss": 4.3356,
      "step": 685
    },
    {
      "epoch": 0.725925925925926,
      "grad_norm": 4.534053325653076,
      "learning_rate": 1.3756613756613756e-05,
      "loss": 4.4978,
      "step": 686
    },
    {
      "epoch": 0.726984126984127,
      "grad_norm": 5.670014381408691,
      "learning_rate": 1.3703703703703704e-05,
      "loss": 4.4208,
      "step": 687
    },
    {
      "epoch": 0.728042328042328,
      "grad_norm": 4.900901794433594,
      "learning_rate": 1.365079365079365e-05,
      "loss": 4.2623,
      "step": 688
    },
    {
      "epoch": 0.7291005291005291,
      "grad_norm": 5.048198699951172,
      "learning_rate": 1.3597883597883599e-05,
      "loss": 4.3792,
      "step": 689
    },
    {
      "epoch": 0.7301587301587301,
      "grad_norm": 3.673882007598877,
      "learning_rate": 1.3544973544973547e-05,
      "loss": 4.5557,
      "step": 690
    },
    {
      "epoch": 0.7312169312169312,
      "grad_norm": 4.310447692871094,
      "learning_rate": 1.3492063492063492e-05,
      "loss": 4.1925,
      "step": 691
    },
    {
      "epoch": 0.7322751322751323,
      "grad_norm": 3.8118319511413574,
      "learning_rate": 1.343915343915344e-05,
      "loss": 4.25,
      "step": 692
    },
    {
      "epoch": 0.7333333333333333,
      "grad_norm": 3.351060628890991,
      "learning_rate": 1.3386243386243385e-05,
      "loss": 4.4192,
      "step": 693
    },
    {
      "epoch": 0.7343915343915344,
      "grad_norm": 4.297890663146973,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 4.208,
      "step": 694
    },
    {
      "epoch": 0.7354497354497355,
      "grad_norm": 5.1479315757751465,
      "learning_rate": 1.3280423280423282e-05,
      "loss": 4.38,
      "step": 695
    },
    {
      "epoch": 0.7365079365079366,
      "grad_norm": 5.45789909362793,
      "learning_rate": 1.3227513227513228e-05,
      "loss": 4.4045,
      "step": 696
    },
    {
      "epoch": 0.7375661375661375,
      "grad_norm": 2.974043369293213,
      "learning_rate": 1.3174603174603176e-05,
      "loss": 4.5253,
      "step": 697
    },
    {
      "epoch": 0.7386243386243386,
      "grad_norm": 5.557780742645264,
      "learning_rate": 1.3121693121693121e-05,
      "loss": 4.569,
      "step": 698
    },
    {
      "epoch": 0.7396825396825397,
      "grad_norm": 3.755084991455078,
      "learning_rate": 1.306878306878307e-05,
      "loss": 4.3673,
      "step": 699
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 4.555693626403809,
      "learning_rate": 1.3015873015873018e-05,
      "loss": 4.2973,
      "step": 700
    },
    {
      "epoch": 0.7417989417989418,
      "grad_norm": 5.5160651206970215,
      "learning_rate": 1.2962962962962962e-05,
      "loss": 4.399,
      "step": 701
    },
    {
      "epoch": 0.7428571428571429,
      "grad_norm": 4.394107818603516,
      "learning_rate": 1.291005291005291e-05,
      "loss": 4.3341,
      "step": 702
    },
    {
      "epoch": 0.7439153439153439,
      "grad_norm": 5.06443452835083,
      "learning_rate": 1.2857142857142857e-05,
      "loss": 4.1375,
      "step": 703
    },
    {
      "epoch": 0.744973544973545,
      "grad_norm": 3.5541141033172607,
      "learning_rate": 1.2804232804232805e-05,
      "loss": 4.203,
      "step": 704
    },
    {
      "epoch": 0.746031746031746,
      "grad_norm": 4.325985431671143,
      "learning_rate": 1.2751322751322754e-05,
      "loss": 4.5174,
      "step": 705
    },
    {
      "epoch": 0.7470899470899471,
      "grad_norm": 2.9800527095794678,
      "learning_rate": 1.2698412698412699e-05,
      "loss": 4.4557,
      "step": 706
    },
    {
      "epoch": 0.7481481481481481,
      "grad_norm": 3.3406500816345215,
      "learning_rate": 1.2645502645502647e-05,
      "loss": 4.453,
      "step": 707
    },
    {
      "epoch": 0.7492063492063492,
      "grad_norm": 5.763829231262207,
      "learning_rate": 1.2592592592592592e-05,
      "loss": 4.3594,
      "step": 708
    },
    {
      "epoch": 0.7502645502645503,
      "grad_norm": 4.623786449432373,
      "learning_rate": 1.253968253968254e-05,
      "loss": 4.291,
      "step": 709
    },
    {
      "epoch": 0.7513227513227513,
      "grad_norm": 4.377042293548584,
      "learning_rate": 1.2486772486772486e-05,
      "loss": 4.0562,
      "step": 710
    },
    {
      "epoch": 0.7523809523809524,
      "grad_norm": 3.437060594558716,
      "learning_rate": 1.2433862433862433e-05,
      "loss": 4.4313,
      "step": 711
    },
    {
      "epoch": 0.7534391534391535,
      "grad_norm": 6.42732572555542,
      "learning_rate": 1.2380952380952381e-05,
      "loss": 4.3612,
      "step": 712
    },
    {
      "epoch": 0.7544973544973544,
      "grad_norm": 5.4876627922058105,
      "learning_rate": 1.232804232804233e-05,
      "loss": 4.3101,
      "step": 713
    },
    {
      "epoch": 0.7555555555555555,
      "grad_norm": 4.041126728057861,
      "learning_rate": 1.2275132275132276e-05,
      "loss": 4.3653,
      "step": 714
    },
    {
      "epoch": 0.7566137566137566,
      "grad_norm": 4.583195209503174,
      "learning_rate": 1.2222222222222222e-05,
      "loss": 4.164,
      "step": 715
    },
    {
      "epoch": 0.7576719576719577,
      "grad_norm": 4.028945446014404,
      "learning_rate": 1.2169312169312169e-05,
      "loss": 4.2116,
      "step": 716
    },
    {
      "epoch": 0.7587301587301587,
      "grad_norm": 6.026222229003906,
      "learning_rate": 1.2116402116402117e-05,
      "loss": 4.2678,
      "step": 717
    },
    {
      "epoch": 0.7597883597883598,
      "grad_norm": 3.280238151550293,
      "learning_rate": 1.2063492063492064e-05,
      "loss": 4.1293,
      "step": 718
    },
    {
      "epoch": 0.7608465608465609,
      "grad_norm": 3.384033203125,
      "learning_rate": 1.201058201058201e-05,
      "loss": 4.3763,
      "step": 719
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 3.618666648864746,
      "learning_rate": 1.1957671957671959e-05,
      "loss": 4.5046,
      "step": 720
    },
    {
      "epoch": 0.762962962962963,
      "grad_norm": 3.716320514678955,
      "learning_rate": 1.1904761904761905e-05,
      "loss": 4.4088,
      "step": 721
    },
    {
      "epoch": 0.764021164021164,
      "grad_norm": 4.554460048675537,
      "learning_rate": 1.1851851851851853e-05,
      "loss": 4.1846,
      "step": 722
    },
    {
      "epoch": 0.765079365079365,
      "grad_norm": 4.681301116943359,
      "learning_rate": 1.17989417989418e-05,
      "loss": 4.2343,
      "step": 723
    },
    {
      "epoch": 0.7661375661375661,
      "grad_norm": 5.567061901092529,
      "learning_rate": 1.1746031746031746e-05,
      "loss": 3.9467,
      "step": 724
    },
    {
      "epoch": 0.7671957671957672,
      "grad_norm": 3.0023434162139893,
      "learning_rate": 1.1693121693121693e-05,
      "loss": 4.1829,
      "step": 725
    },
    {
      "epoch": 0.7682539682539683,
      "grad_norm": 6.105156898498535,
      "learning_rate": 1.164021164021164e-05,
      "loss": 4.0128,
      "step": 726
    },
    {
      "epoch": 0.7693121693121693,
      "grad_norm": 3.2592265605926514,
      "learning_rate": 1.1587301587301588e-05,
      "loss": 4.4044,
      "step": 727
    },
    {
      "epoch": 0.7703703703703704,
      "grad_norm": 4.033621311187744,
      "learning_rate": 1.1534391534391536e-05,
      "loss": 4.5054,
      "step": 728
    },
    {
      "epoch": 0.7714285714285715,
      "grad_norm": 6.191745281219482,
      "learning_rate": 1.1481481481481482e-05,
      "loss": 4.7064,
      "step": 729
    },
    {
      "epoch": 0.7724867724867724,
      "grad_norm": 6.648366928100586,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 4.815,
      "step": 730
    },
    {
      "epoch": 0.7735449735449735,
      "grad_norm": 3.693485736846924,
      "learning_rate": 1.1375661375661376e-05,
      "loss": 4.1928,
      "step": 731
    },
    {
      "epoch": 0.7746031746031746,
      "grad_norm": 3.7810678482055664,
      "learning_rate": 1.1322751322751324e-05,
      "loss": 4.3781,
      "step": 732
    },
    {
      "epoch": 0.7756613756613756,
      "grad_norm": 4.211301803588867,
      "learning_rate": 1.126984126984127e-05,
      "loss": 4.1813,
      "step": 733
    },
    {
      "epoch": 0.7767195767195767,
      "grad_norm": 3.337557792663574,
      "learning_rate": 1.1216931216931217e-05,
      "loss": 4.3928,
      "step": 734
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 3.5120623111724854,
      "learning_rate": 1.1164021164021163e-05,
      "loss": 4.272,
      "step": 735
    },
    {
      "epoch": 0.7788359788359789,
      "grad_norm": 5.32808780670166,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 4.3989,
      "step": 736
    },
    {
      "epoch": 0.7798941798941799,
      "grad_norm": 5.039298057556152,
      "learning_rate": 1.105820105820106e-05,
      "loss": 4.1569,
      "step": 737
    },
    {
      "epoch": 0.780952380952381,
      "grad_norm": 4.3160223960876465,
      "learning_rate": 1.1005291005291006e-05,
      "loss": 4.0202,
      "step": 738
    },
    {
      "epoch": 0.782010582010582,
      "grad_norm": 6.317615985870361,
      "learning_rate": 1.0952380952380953e-05,
      "loss": 4.3689,
      "step": 739
    },
    {
      "epoch": 0.783068783068783,
      "grad_norm": 4.654201030731201,
      "learning_rate": 1.08994708994709e-05,
      "loss": 4.2718,
      "step": 740
    },
    {
      "epoch": 0.7841269841269841,
      "grad_norm": 4.299462795257568,
      "learning_rate": 1.0846560846560846e-05,
      "loss": 4.2899,
      "step": 741
    },
    {
      "epoch": 0.7851851851851852,
      "grad_norm": 3.5088319778442383,
      "learning_rate": 1.0793650793650794e-05,
      "loss": 4.453,
      "step": 742
    },
    {
      "epoch": 0.7862433862433862,
      "grad_norm": 4.868035316467285,
      "learning_rate": 1.074074074074074e-05,
      "loss": 4.3791,
      "step": 743
    },
    {
      "epoch": 0.7873015873015873,
      "grad_norm": 3.0440139770507812,
      "learning_rate": 1.0687830687830689e-05,
      "loss": 4.2274,
      "step": 744
    },
    {
      "epoch": 0.7883597883597884,
      "grad_norm": 3.8341310024261475,
      "learning_rate": 1.0634920634920636e-05,
      "loss": 4.5376,
      "step": 745
    },
    {
      "epoch": 0.7894179894179895,
      "grad_norm": 4.445629596710205,
      "learning_rate": 1.0582010582010582e-05,
      "loss": 4.4937,
      "step": 746
    },
    {
      "epoch": 0.7904761904761904,
      "grad_norm": 3.684857130050659,
      "learning_rate": 1.052910052910053e-05,
      "loss": 4.8373,
      "step": 747
    },
    {
      "epoch": 0.7915343915343915,
      "grad_norm": 3.8749489784240723,
      "learning_rate": 1.0476190476190477e-05,
      "loss": 4.5327,
      "step": 748
    },
    {
      "epoch": 0.7925925925925926,
      "grad_norm": 3.256950855255127,
      "learning_rate": 1.0423280423280423e-05,
      "loss": 4.6,
      "step": 749
    },
    {
      "epoch": 0.7936507936507936,
      "grad_norm": 5.716484069824219,
      "learning_rate": 1.037037037037037e-05,
      "loss": 4.7403,
      "step": 750
    },
    {
      "epoch": 0.7947089947089947,
      "grad_norm": 6.667811870574951,
      "learning_rate": 1.0317460317460318e-05,
      "loss": 4.1508,
      "step": 751
    },
    {
      "epoch": 0.7957671957671958,
      "grad_norm": 6.336728572845459,
      "learning_rate": 1.0264550264550265e-05,
      "loss": 3.9915,
      "step": 752
    },
    {
      "epoch": 0.7968253968253968,
      "grad_norm": 3.4443869590759277,
      "learning_rate": 1.0211640211640213e-05,
      "loss": 4.2412,
      "step": 753
    },
    {
      "epoch": 0.7978835978835979,
      "grad_norm": 3.8380494117736816,
      "learning_rate": 1.015873015873016e-05,
      "loss": 4.1926,
      "step": 754
    },
    {
      "epoch": 0.798941798941799,
      "grad_norm": 4.2469401359558105,
      "learning_rate": 1.0105820105820106e-05,
      "loss": 4.2798,
      "step": 755
    },
    {
      "epoch": 0.8,
      "grad_norm": 3.973093032836914,
      "learning_rate": 1.0052910052910053e-05,
      "loss": 4.2463,
      "step": 756
    },
    {
      "epoch": 0.801058201058201,
      "grad_norm": 3.1232118606567383,
      "learning_rate": 1e-05,
      "loss": 4.1261,
      "step": 757
    },
    {
      "epoch": 0.8021164021164021,
      "grad_norm": 6.105686187744141,
      "learning_rate": 9.947089947089947e-06,
      "loss": 4.2619,
      "step": 758
    },
    {
      "epoch": 0.8031746031746032,
      "grad_norm": 6.121016502380371,
      "learning_rate": 9.894179894179894e-06,
      "loss": 4.431,
      "step": 759
    },
    {
      "epoch": 0.8042328042328042,
      "grad_norm": 6.174698352813721,
      "learning_rate": 9.841269841269842e-06,
      "loss": 4.3786,
      "step": 760
    },
    {
      "epoch": 0.8052910052910053,
      "grad_norm": 5.5360941886901855,
      "learning_rate": 9.788359788359789e-06,
      "loss": 4.2559,
      "step": 761
    },
    {
      "epoch": 0.8063492063492064,
      "grad_norm": 3.6265597343444824,
      "learning_rate": 9.735449735449737e-06,
      "loss": 4.4958,
      "step": 762
    },
    {
      "epoch": 0.8074074074074075,
      "grad_norm": 6.606620788574219,
      "learning_rate": 9.682539682539683e-06,
      "loss": 4.5506,
      "step": 763
    },
    {
      "epoch": 0.8084656084656084,
      "grad_norm": 5.32057523727417,
      "learning_rate": 9.62962962962963e-06,
      "loss": 4.3155,
      "step": 764
    },
    {
      "epoch": 0.8095238095238095,
      "grad_norm": 3.1966543197631836,
      "learning_rate": 9.576719576719577e-06,
      "loss": 4.4766,
      "step": 765
    },
    {
      "epoch": 0.8105820105820106,
      "grad_norm": 4.467657566070557,
      "learning_rate": 9.523809523809523e-06,
      "loss": 4.4167,
      "step": 766
    },
    {
      "epoch": 0.8116402116402116,
      "grad_norm": 3.145953416824341,
      "learning_rate": 9.470899470899471e-06,
      "loss": 4.2835,
      "step": 767
    },
    {
      "epoch": 0.8126984126984127,
      "grad_norm": 4.974372386932373,
      "learning_rate": 9.41798941798942e-06,
      "loss": 4.5889,
      "step": 768
    },
    {
      "epoch": 0.8137566137566138,
      "grad_norm": 3.7245638370513916,
      "learning_rate": 9.365079365079366e-06,
      "loss": 4.1537,
      "step": 769
    },
    {
      "epoch": 0.8148148148148148,
      "grad_norm": 3.2511162757873535,
      "learning_rate": 9.312169312169313e-06,
      "loss": 4.4419,
      "step": 770
    },
    {
      "epoch": 0.8158730158730159,
      "grad_norm": 6.026091575622559,
      "learning_rate": 9.259259259259259e-06,
      "loss": 4.209,
      "step": 771
    },
    {
      "epoch": 0.816931216931217,
      "grad_norm": 3.5133731365203857,
      "learning_rate": 9.206349206349207e-06,
      "loss": 4.3719,
      "step": 772
    },
    {
      "epoch": 0.817989417989418,
      "grad_norm": 4.241281509399414,
      "learning_rate": 9.153439153439154e-06,
      "loss": 4.5862,
      "step": 773
    },
    {
      "epoch": 0.819047619047619,
      "grad_norm": 4.5300517082214355,
      "learning_rate": 9.1005291005291e-06,
      "loss": 4.4843,
      "step": 774
    },
    {
      "epoch": 0.8201058201058201,
      "grad_norm": 5.496360778808594,
      "learning_rate": 9.047619047619047e-06,
      "loss": 4.2999,
      "step": 775
    },
    {
      "epoch": 0.8211640211640212,
      "grad_norm": 3.988664388656616,
      "learning_rate": 8.994708994708995e-06,
      "loss": 4.2362,
      "step": 776
    },
    {
      "epoch": 0.8222222222222222,
      "grad_norm": 3.7009782791137695,
      "learning_rate": 8.941798941798943e-06,
      "loss": 4.2936,
      "step": 777
    },
    {
      "epoch": 0.8232804232804233,
      "grad_norm": 5.037206649780273,
      "learning_rate": 8.88888888888889e-06,
      "loss": 4.3659,
      "step": 778
    },
    {
      "epoch": 0.8243386243386244,
      "grad_norm": 3.1383752822875977,
      "learning_rate": 8.835978835978837e-06,
      "loss": 4.4558,
      "step": 779
    },
    {
      "epoch": 0.8253968253968254,
      "grad_norm": 3.314852714538574,
      "learning_rate": 8.783068783068783e-06,
      "loss": 4.556,
      "step": 780
    },
    {
      "epoch": 0.8264550264550264,
      "grad_norm": 3.957002878189087,
      "learning_rate": 8.73015873015873e-06,
      "loss": 4.4447,
      "step": 781
    },
    {
      "epoch": 0.8275132275132275,
      "grad_norm": 3.4280648231506348,
      "learning_rate": 8.677248677248678e-06,
      "loss": 4.245,
      "step": 782
    },
    {
      "epoch": 0.8285714285714286,
      "grad_norm": 3.471327304840088,
      "learning_rate": 8.624338624338624e-06,
      "loss": 4.4841,
      "step": 783
    },
    {
      "epoch": 0.8296296296296296,
      "grad_norm": 4.864382743835449,
      "learning_rate": 8.571428571428573e-06,
      "loss": 4.2108,
      "step": 784
    },
    {
      "epoch": 0.8306878306878307,
      "grad_norm": 5.851284027099609,
      "learning_rate": 8.518518518518519e-06,
      "loss": 4.0281,
      "step": 785
    },
    {
      "epoch": 0.8317460317460318,
      "grad_norm": 4.928969860076904,
      "learning_rate": 8.465608465608466e-06,
      "loss": 4.2632,
      "step": 786
    },
    {
      "epoch": 0.8328042328042328,
      "grad_norm": 3.591388702392578,
      "learning_rate": 8.412698412698414e-06,
      "loss": 4.3222,
      "step": 787
    },
    {
      "epoch": 0.8338624338624339,
      "grad_norm": 3.6822257041931152,
      "learning_rate": 8.35978835978836e-06,
      "loss": 4.687,
      "step": 788
    },
    {
      "epoch": 0.834920634920635,
      "grad_norm": 3.3557658195495605,
      "learning_rate": 8.306878306878307e-06,
      "loss": 4.3424,
      "step": 789
    },
    {
      "epoch": 0.8359788359788359,
      "grad_norm": 3.654463529586792,
      "learning_rate": 8.253968253968254e-06,
      "loss": 4.0186,
      "step": 790
    },
    {
      "epoch": 0.837037037037037,
      "grad_norm": 4.440916538238525,
      "learning_rate": 8.201058201058202e-06,
      "loss": 4.1733,
      "step": 791
    },
    {
      "epoch": 0.8380952380952381,
      "grad_norm": 5.019598007202148,
      "learning_rate": 8.14814814814815e-06,
      "loss": 4.2378,
      "step": 792
    },
    {
      "epoch": 0.8391534391534392,
      "grad_norm": 5.27589225769043,
      "learning_rate": 8.095238095238097e-06,
      "loss": 4.2217,
      "step": 793
    },
    {
      "epoch": 0.8402116402116402,
      "grad_norm": 3.2513492107391357,
      "learning_rate": 8.042328042328043e-06,
      "loss": 4.2982,
      "step": 794
    },
    {
      "epoch": 0.8412698412698413,
      "grad_norm": 3.5115115642547607,
      "learning_rate": 7.98941798941799e-06,
      "loss": 4.2832,
      "step": 795
    },
    {
      "epoch": 0.8423280423280424,
      "grad_norm": 4.6819000244140625,
      "learning_rate": 7.936507936507936e-06,
      "loss": 4.3657,
      "step": 796
    },
    {
      "epoch": 0.8433862433862434,
      "grad_norm": 4.580070972442627,
      "learning_rate": 7.883597883597884e-06,
      "loss": 4.3363,
      "step": 797
    },
    {
      "epoch": 0.8444444444444444,
      "grad_norm": 4.679523468017578,
      "learning_rate": 7.830687830687831e-06,
      "loss": 4.2533,
      "step": 798
    },
    {
      "epoch": 0.8455026455026455,
      "grad_norm": 3.7114503383636475,
      "learning_rate": 7.777777777777777e-06,
      "loss": 4.2543,
      "step": 799
    },
    {
      "epoch": 0.8465608465608465,
      "grad_norm": 3.6227071285247803,
      "learning_rate": 7.724867724867726e-06,
      "loss": 4.3794,
      "step": 800
    },
    {
      "epoch": 0.8476190476190476,
      "grad_norm": 3.9101266860961914,
      "learning_rate": 7.671957671957672e-06,
      "loss": 4.4385,
      "step": 801
    },
    {
      "epoch": 0.8486772486772487,
      "grad_norm": 4.538809299468994,
      "learning_rate": 7.6190476190476205e-06,
      "loss": 4.1798,
      "step": 802
    },
    {
      "epoch": 0.8497354497354498,
      "grad_norm": 3.392962694168091,
      "learning_rate": 7.566137566137567e-06,
      "loss": 4.3668,
      "step": 803
    },
    {
      "epoch": 0.8507936507936508,
      "grad_norm": 3.7657711505889893,
      "learning_rate": 7.5132275132275136e-06,
      "loss": 4.2039,
      "step": 804
    },
    {
      "epoch": 0.8518518518518519,
      "grad_norm": 4.12888240814209,
      "learning_rate": 7.460317460317461e-06,
      "loss": 4.4272,
      "step": 805
    },
    {
      "epoch": 0.852910052910053,
      "grad_norm": 4.325788974761963,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 4.3938,
      "step": 806
    },
    {
      "epoch": 0.8539682539682539,
      "grad_norm": 6.836640357971191,
      "learning_rate": 7.354497354497354e-06,
      "loss": 4.481,
      "step": 807
    },
    {
      "epoch": 0.855026455026455,
      "grad_norm": 3.2702300548553467,
      "learning_rate": 7.301587301587302e-06,
      "loss": 4.2527,
      "step": 808
    },
    {
      "epoch": 0.8560846560846561,
      "grad_norm": 3.6480207443237305,
      "learning_rate": 7.248677248677249e-06,
      "loss": 4.3472,
      "step": 809
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 3.0763864517211914,
      "learning_rate": 7.195767195767196e-06,
      "loss": 4.1348,
      "step": 810
    },
    {
      "epoch": 0.8582010582010582,
      "grad_norm": 4.69089937210083,
      "learning_rate": 7.142857142857143e-06,
      "loss": 4.4034,
      "step": 811
    },
    {
      "epoch": 0.8592592592592593,
      "grad_norm": 6.656426429748535,
      "learning_rate": 7.089947089947089e-06,
      "loss": 4.4632,
      "step": 812
    },
    {
      "epoch": 0.8603174603174604,
      "grad_norm": 5.9208760261535645,
      "learning_rate": 7.0370370370370375e-06,
      "loss": 4.4209,
      "step": 813
    },
    {
      "epoch": 0.8613756613756614,
      "grad_norm": 5.307591438293457,
      "learning_rate": 6.984126984126985e-06,
      "loss": 4.3608,
      "step": 814
    },
    {
      "epoch": 0.8624338624338624,
      "grad_norm": 3.104219675064087,
      "learning_rate": 6.931216931216931e-06,
      "loss": 4.2559,
      "step": 815
    },
    {
      "epoch": 0.8634920634920635,
      "grad_norm": 5.242783069610596,
      "learning_rate": 6.878306878306878e-06,
      "loss": 4.426,
      "step": 816
    },
    {
      "epoch": 0.8645502645502645,
      "grad_norm": 3.6152803897857666,
      "learning_rate": 6.825396825396825e-06,
      "loss": 4.2805,
      "step": 817
    },
    {
      "epoch": 0.8656084656084656,
      "grad_norm": 5.046814918518066,
      "learning_rate": 6.7724867724867736e-06,
      "loss": 4.2196,
      "step": 818
    },
    {
      "epoch": 0.8666666666666667,
      "grad_norm": 4.008995532989502,
      "learning_rate": 6.71957671957672e-06,
      "loss": 4.3369,
      "step": 819
    },
    {
      "epoch": 0.8677248677248677,
      "grad_norm": 3.5504777431488037,
      "learning_rate": 6.666666666666667e-06,
      "loss": 4.2037,
      "step": 820
    },
    {
      "epoch": 0.8687830687830688,
      "grad_norm": 5.669442653656006,
      "learning_rate": 6.613756613756614e-06,
      "loss": 4.2309,
      "step": 821
    },
    {
      "epoch": 0.8698412698412699,
      "grad_norm": 5.117327690124512,
      "learning_rate": 6.5608465608465606e-06,
      "loss": 4.3519,
      "step": 822
    },
    {
      "epoch": 0.870899470899471,
      "grad_norm": 4.981729507446289,
      "learning_rate": 6.507936507936509e-06,
      "loss": 4.01,
      "step": 823
    },
    {
      "epoch": 0.8719576719576719,
      "grad_norm": 3.664243459701538,
      "learning_rate": 6.455026455026455e-06,
      "loss": 4.2071,
      "step": 824
    },
    {
      "epoch": 0.873015873015873,
      "grad_norm": 3.5531766414642334,
      "learning_rate": 6.402116402116403e-06,
      "loss": 4.4418,
      "step": 825
    },
    {
      "epoch": 0.8740740740740741,
      "grad_norm": 3.976097583770752,
      "learning_rate": 6.349206349206349e-06,
      "loss": 4.3688,
      "step": 826
    },
    {
      "epoch": 0.8751322751322751,
      "grad_norm": 3.7528669834136963,
      "learning_rate": 6.296296296296296e-06,
      "loss": 4.3587,
      "step": 827
    },
    {
      "epoch": 0.8761904761904762,
      "grad_norm": 5.616343021392822,
      "learning_rate": 6.243386243386243e-06,
      "loss": 4.4252,
      "step": 828
    },
    {
      "epoch": 0.8772486772486773,
      "grad_norm": 5.017111301422119,
      "learning_rate": 6.190476190476191e-06,
      "loss": 4.2342,
      "step": 829
    },
    {
      "epoch": 0.8783068783068783,
      "grad_norm": 4.545386791229248,
      "learning_rate": 6.137566137566138e-06,
      "loss": 4.1429,
      "step": 830
    },
    {
      "epoch": 0.8793650793650793,
      "grad_norm": 5.04618501663208,
      "learning_rate": 6.0846560846560845e-06,
      "loss": 4.3157,
      "step": 831
    },
    {
      "epoch": 0.8804232804232804,
      "grad_norm": 5.4261884689331055,
      "learning_rate": 6.031746031746032e-06,
      "loss": 4.1527,
      "step": 832
    },
    {
      "epoch": 0.8814814814814815,
      "grad_norm": 4.209038734436035,
      "learning_rate": 5.978835978835979e-06,
      "loss": 4.1613,
      "step": 833
    },
    {
      "epoch": 0.8825396825396825,
      "grad_norm": 3.330681562423706,
      "learning_rate": 5.925925925925927e-06,
      "loss": 4.3347,
      "step": 834
    },
    {
      "epoch": 0.8835978835978836,
      "grad_norm": 4.095284938812256,
      "learning_rate": 5.873015873015873e-06,
      "loss": 4.4005,
      "step": 835
    },
    {
      "epoch": 0.8846560846560847,
      "grad_norm": 5.182604789733887,
      "learning_rate": 5.82010582010582e-06,
      "loss": 4.0961,
      "step": 836
    },
    {
      "epoch": 0.8857142857142857,
      "grad_norm": 4.743280410766602,
      "learning_rate": 5.767195767195768e-06,
      "loss": 4.4677,
      "step": 837
    },
    {
      "epoch": 0.8867724867724868,
      "grad_norm": 5.909420967102051,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 4.0023,
      "step": 838
    },
    {
      "epoch": 0.8878306878306879,
      "grad_norm": 4.317636966705322,
      "learning_rate": 5.661375661375662e-06,
      "loss": 4.2847,
      "step": 839
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 4.787182331085205,
      "learning_rate": 5.6084656084656084e-06,
      "loss": 4.3972,
      "step": 840
    },
    {
      "epoch": 0.8899470899470899,
      "grad_norm": 4.967015266418457,
      "learning_rate": 5.555555555555556e-06,
      "loss": 4.23,
      "step": 841
    },
    {
      "epoch": 0.891005291005291,
      "grad_norm": 3.515166997909546,
      "learning_rate": 5.502645502645503e-06,
      "loss": 4.5061,
      "step": 842
    },
    {
      "epoch": 0.8920634920634921,
      "grad_norm": 5.033231258392334,
      "learning_rate": 5.44973544973545e-06,
      "loss": 4.416,
      "step": 843
    },
    {
      "epoch": 0.8931216931216931,
      "grad_norm": 3.4822046756744385,
      "learning_rate": 5.396825396825397e-06,
      "loss": 4.345,
      "step": 844
    },
    {
      "epoch": 0.8941798941798942,
      "grad_norm": 4.365127086639404,
      "learning_rate": 5.3439153439153445e-06,
      "loss": 4.2098,
      "step": 845
    },
    {
      "epoch": 0.8952380952380953,
      "grad_norm": 2.938222646713257,
      "learning_rate": 5.291005291005291e-06,
      "loss": 4.1394,
      "step": 846
    },
    {
      "epoch": 0.8962962962962963,
      "grad_norm": 3.7429540157318115,
      "learning_rate": 5.2380952380952384e-06,
      "loss": 4.3739,
      "step": 847
    },
    {
      "epoch": 0.8973544973544973,
      "grad_norm": 5.029417991638184,
      "learning_rate": 5.185185185185185e-06,
      "loss": 4.4277,
      "step": 848
    },
    {
      "epoch": 0.8984126984126984,
      "grad_norm": 3.043536424636841,
      "learning_rate": 5.132275132275132e-06,
      "loss": 4.1825,
      "step": 849
    },
    {
      "epoch": 0.8994708994708994,
      "grad_norm": 2.5307250022888184,
      "learning_rate": 5.07936507936508e-06,
      "loss": 4.4159,
      "step": 850
    },
    {
      "epoch": 0.9005291005291005,
      "grad_norm": 4.139734268188477,
      "learning_rate": 5.026455026455026e-06,
      "loss": 4.4281,
      "step": 851
    },
    {
      "epoch": 0.9015873015873016,
      "grad_norm": 4.402191638946533,
      "learning_rate": 4.973544973544974e-06,
      "loss": 4.2871,
      "step": 852
    },
    {
      "epoch": 0.9026455026455027,
      "grad_norm": 3.3426613807678223,
      "learning_rate": 4.920634920634921e-06,
      "loss": 4.2474,
      "step": 853
    },
    {
      "epoch": 0.9037037037037037,
      "grad_norm": 3.7858426570892334,
      "learning_rate": 4.8677248677248685e-06,
      "loss": 4.2868,
      "step": 854
    },
    {
      "epoch": 0.9047619047619048,
      "grad_norm": 4.034852981567383,
      "learning_rate": 4.814814814814815e-06,
      "loss": 3.9792,
      "step": 855
    },
    {
      "epoch": 0.9058201058201059,
      "grad_norm": 3.4261882305145264,
      "learning_rate": 4.7619047619047615e-06,
      "loss": 4.1932,
      "step": 856
    },
    {
      "epoch": 0.9068783068783068,
      "grad_norm": 4.608288288116455,
      "learning_rate": 4.70899470899471e-06,
      "loss": 4.165,
      "step": 857
    },
    {
      "epoch": 0.9079365079365079,
      "grad_norm": 3.5496902465820312,
      "learning_rate": 4.656084656084656e-06,
      "loss": 4.5227,
      "step": 858
    },
    {
      "epoch": 0.908994708994709,
      "grad_norm": 5.273512840270996,
      "learning_rate": 4.603174603174604e-06,
      "loss": 4.2818,
      "step": 859
    },
    {
      "epoch": 0.91005291005291,
      "grad_norm": 5.324268817901611,
      "learning_rate": 4.55026455026455e-06,
      "loss": 4.3419,
      "step": 860
    },
    {
      "epoch": 0.9111111111111111,
      "grad_norm": 3.7706358432769775,
      "learning_rate": 4.497354497354498e-06,
      "loss": 4.1445,
      "step": 861
    },
    {
      "epoch": 0.9121693121693122,
      "grad_norm": 3.275144100189209,
      "learning_rate": 4.444444444444445e-06,
      "loss": 4.4849,
      "step": 862
    },
    {
      "epoch": 0.9132275132275133,
      "grad_norm": 3.1599481105804443,
      "learning_rate": 4.3915343915343915e-06,
      "loss": 4.2842,
      "step": 863
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 4.762506484985352,
      "learning_rate": 4.338624338624339e-06,
      "loss": 3.9712,
      "step": 864
    },
    {
      "epoch": 0.9153439153439153,
      "grad_norm": 3.7932703495025635,
      "learning_rate": 4.285714285714286e-06,
      "loss": 4.3115,
      "step": 865
    },
    {
      "epoch": 0.9164021164021164,
      "grad_norm": 4.2084736824035645,
      "learning_rate": 4.232804232804233e-06,
      "loss": 4.294,
      "step": 866
    },
    {
      "epoch": 0.9174603174603174,
      "grad_norm": 3.4369008541107178,
      "learning_rate": 4.17989417989418e-06,
      "loss": 4.3855,
      "step": 867
    },
    {
      "epoch": 0.9185185185185185,
      "grad_norm": 5.6549177169799805,
      "learning_rate": 4.126984126984127e-06,
      "loss": 4.3273,
      "step": 868
    },
    {
      "epoch": 0.9195767195767196,
      "grad_norm": 3.5706095695495605,
      "learning_rate": 4.074074074074075e-06,
      "loss": 3.9828,
      "step": 869
    },
    {
      "epoch": 0.9206349206349206,
      "grad_norm": 5.644643306732178,
      "learning_rate": 4.0211640211640215e-06,
      "loss": 4.1306,
      "step": 870
    },
    {
      "epoch": 0.9216931216931217,
      "grad_norm": 5.228826522827148,
      "learning_rate": 3.968253968253968e-06,
      "loss": 4.1117,
      "step": 871
    },
    {
      "epoch": 0.9227513227513228,
      "grad_norm": 3.1740591526031494,
      "learning_rate": 3.9153439153439155e-06,
      "loss": 4.4053,
      "step": 872
    },
    {
      "epoch": 0.9238095238095239,
      "grad_norm": 3.6859006881713867,
      "learning_rate": 3.862433862433863e-06,
      "loss": 4.3222,
      "step": 873
    },
    {
      "epoch": 0.9248677248677248,
      "grad_norm": 7.347630023956299,
      "learning_rate": 3.8095238095238102e-06,
      "loss": 4.3781,
      "step": 874
    },
    {
      "epoch": 0.9259259259259259,
      "grad_norm": 4.983848571777344,
      "learning_rate": 3.7566137566137568e-06,
      "loss": 4.2057,
      "step": 875
    },
    {
      "epoch": 0.926984126984127,
      "grad_norm": 6.287315368652344,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 4.1598,
      "step": 876
    },
    {
      "epoch": 0.928042328042328,
      "grad_norm": 5.142515182495117,
      "learning_rate": 3.650793650793651e-06,
      "loss": 4.5319,
      "step": 877
    },
    {
      "epoch": 0.9291005291005291,
      "grad_norm": 3.5645596981048584,
      "learning_rate": 3.597883597883598e-06,
      "loss": 4.3179,
      "step": 878
    },
    {
      "epoch": 0.9301587301587302,
      "grad_norm": 4.008954048156738,
      "learning_rate": 3.5449735449735446e-06,
      "loss": 4.0794,
      "step": 879
    },
    {
      "epoch": 0.9312169312169312,
      "grad_norm": 5.768702030181885,
      "learning_rate": 3.4920634920634924e-06,
      "loss": 4.5562,
      "step": 880
    },
    {
      "epoch": 0.9322751322751323,
      "grad_norm": 4.129014015197754,
      "learning_rate": 3.439153439153439e-06,
      "loss": 4.3659,
      "step": 881
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 6.950862407684326,
      "learning_rate": 3.3862433862433868e-06,
      "loss": 4.3704,
      "step": 882
    },
    {
      "epoch": 0.9343915343915344,
      "grad_norm": 6.936004638671875,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 4.6306,
      "step": 883
    },
    {
      "epoch": 0.9354497354497354,
      "grad_norm": 6.682819366455078,
      "learning_rate": 3.2804232804232803e-06,
      "loss": 4.3953,
      "step": 884
    },
    {
      "epoch": 0.9365079365079365,
      "grad_norm": 4.455723762512207,
      "learning_rate": 3.2275132275132277e-06,
      "loss": 4.3125,
      "step": 885
    },
    {
      "epoch": 0.9375661375661376,
      "grad_norm": 5.623180389404297,
      "learning_rate": 3.1746031746031746e-06,
      "loss": 4.2719,
      "step": 886
    },
    {
      "epoch": 0.9386243386243386,
      "grad_norm": 3.4735963344573975,
      "learning_rate": 3.1216931216931216e-06,
      "loss": 4.4579,
      "step": 887
    },
    {
      "epoch": 0.9396825396825397,
      "grad_norm": 3.9385921955108643,
      "learning_rate": 3.068783068783069e-06,
      "loss": 4.3531,
      "step": 888
    },
    {
      "epoch": 0.9407407407407408,
      "grad_norm": 4.1221818923950195,
      "learning_rate": 3.015873015873016e-06,
      "loss": 4.2012,
      "step": 889
    },
    {
      "epoch": 0.9417989417989417,
      "grad_norm": 4.046891212463379,
      "learning_rate": 2.9629629629629633e-06,
      "loss": 4.4402,
      "step": 890
    },
    {
      "epoch": 0.9428571428571428,
      "grad_norm": 4.89032506942749,
      "learning_rate": 2.91005291005291e-06,
      "loss": 4.2284,
      "step": 891
    },
    {
      "epoch": 0.9439153439153439,
      "grad_norm": 3.5154287815093994,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 4.0714,
      "step": 892
    },
    {
      "epoch": 0.944973544973545,
      "grad_norm": 4.873518466949463,
      "learning_rate": 2.8042328042328042e-06,
      "loss": 4.2028,
      "step": 893
    },
    {
      "epoch": 0.946031746031746,
      "grad_norm": 2.748612403869629,
      "learning_rate": 2.7513227513227516e-06,
      "loss": 4.3582,
      "step": 894
    },
    {
      "epoch": 0.9470899470899471,
      "grad_norm": 3.607381820678711,
      "learning_rate": 2.6984126984126986e-06,
      "loss": 4.7712,
      "step": 895
    },
    {
      "epoch": 0.9481481481481482,
      "grad_norm": 3.9943275451660156,
      "learning_rate": 2.6455026455026455e-06,
      "loss": 4.3027,
      "step": 896
    },
    {
      "epoch": 0.9492063492063492,
      "grad_norm": 4.708794116973877,
      "learning_rate": 2.5925925925925925e-06,
      "loss": 4.3027,
      "step": 897
    },
    {
      "epoch": 0.9502645502645503,
      "grad_norm": 3.1916885375976562,
      "learning_rate": 2.53968253968254e-06,
      "loss": 4.346,
      "step": 898
    },
    {
      "epoch": 0.9513227513227513,
      "grad_norm": 4.0273566246032715,
      "learning_rate": 2.486772486772487e-06,
      "loss": 4.2844,
      "step": 899
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 4.110088348388672,
      "learning_rate": 2.4338624338624342e-06,
      "loss": 4.1634,
      "step": 900
    },
    {
      "epoch": 0.9534391534391534,
      "grad_norm": 3.984637975692749,
      "learning_rate": 2.3809523809523808e-06,
      "loss": 4.0886,
      "step": 901
    },
    {
      "epoch": 0.9544973544973545,
      "grad_norm": 5.0471391677856445,
      "learning_rate": 2.328042328042328e-06,
      "loss": 4.3436,
      "step": 902
    },
    {
      "epoch": 0.9555555555555556,
      "grad_norm": 2.9655942916870117,
      "learning_rate": 2.275132275132275e-06,
      "loss": 4.5204,
      "step": 903
    },
    {
      "epoch": 0.9566137566137566,
      "grad_norm": 6.42230224609375,
      "learning_rate": 2.2222222222222225e-06,
      "loss": 4.61,
      "step": 904
    },
    {
      "epoch": 0.9576719576719577,
      "grad_norm": 5.822199821472168,
      "learning_rate": 2.1693121693121695e-06,
      "loss": 4.1584,
      "step": 905
    },
    {
      "epoch": 0.9587301587301588,
      "grad_norm": 4.80122709274292,
      "learning_rate": 2.1164021164021164e-06,
      "loss": 4.2502,
      "step": 906
    },
    {
      "epoch": 0.9597883597883597,
      "grad_norm": 6.633864402770996,
      "learning_rate": 2.0634920634920634e-06,
      "loss": 4.3063,
      "step": 907
    },
    {
      "epoch": 0.9608465608465608,
      "grad_norm": 4.278034210205078,
      "learning_rate": 2.0105820105820108e-06,
      "loss": 4.3158,
      "step": 908
    },
    {
      "epoch": 0.9619047619047619,
      "grad_norm": 5.122796058654785,
      "learning_rate": 1.9576719576719577e-06,
      "loss": 4.4717,
      "step": 909
    },
    {
      "epoch": 0.9629629629629629,
      "grad_norm": 4.548975467681885,
      "learning_rate": 1.9047619047619051e-06,
      "loss": 4.356,
      "step": 910
    },
    {
      "epoch": 0.964021164021164,
      "grad_norm": 5.9967827796936035,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 4.062,
      "step": 911
    },
    {
      "epoch": 0.9650793650793651,
      "grad_norm": 2.504676580429077,
      "learning_rate": 1.798941798941799e-06,
      "loss": 4.2251,
      "step": 912
    },
    {
      "epoch": 0.9661375661375662,
      "grad_norm": 3.5260939598083496,
      "learning_rate": 1.7460317460317462e-06,
      "loss": 4.3478,
      "step": 913
    },
    {
      "epoch": 0.9671957671957672,
      "grad_norm": 4.228456974029541,
      "learning_rate": 1.6931216931216934e-06,
      "loss": 4.2843,
      "step": 914
    },
    {
      "epoch": 0.9682539682539683,
      "grad_norm": 3.96049165725708,
      "learning_rate": 1.6402116402116401e-06,
      "loss": 4.4371,
      "step": 915
    },
    {
      "epoch": 0.9693121693121693,
      "grad_norm": 4.0121846199035645,
      "learning_rate": 1.5873015873015873e-06,
      "loss": 4.1606,
      "step": 916
    },
    {
      "epoch": 0.9703703703703703,
      "grad_norm": 4.30086612701416,
      "learning_rate": 1.5343915343915345e-06,
      "loss": 4.3627,
      "step": 917
    },
    {
      "epoch": 0.9714285714285714,
      "grad_norm": 3.6604509353637695,
      "learning_rate": 1.4814814814814817e-06,
      "loss": 4.4242,
      "step": 918
    },
    {
      "epoch": 0.9724867724867725,
      "grad_norm": 4.073081970214844,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 4.1937,
      "step": 919
    },
    {
      "epoch": 0.9735449735449735,
      "grad_norm": 5.492738246917725,
      "learning_rate": 1.3756613756613758e-06,
      "loss": 4.0173,
      "step": 920
    },
    {
      "epoch": 0.9746031746031746,
      "grad_norm": 3.94403076171875,
      "learning_rate": 1.3227513227513228e-06,
      "loss": 4.321,
      "step": 921
    },
    {
      "epoch": 0.9756613756613757,
      "grad_norm": 4.628042697906494,
      "learning_rate": 1.26984126984127e-06,
      "loss": 4.2686,
      "step": 922
    },
    {
      "epoch": 0.9767195767195768,
      "grad_norm": 7.038707733154297,
      "learning_rate": 1.2169312169312171e-06,
      "loss": 4.0722,
      "step": 923
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 3.864443778991699,
      "learning_rate": 1.164021164021164e-06,
      "loss": 4.2553,
      "step": 924
    },
    {
      "epoch": 0.9788359788359788,
      "grad_norm": 3.2768311500549316,
      "learning_rate": 1.1111111111111112e-06,
      "loss": 4.1619,
      "step": 925
    },
    {
      "epoch": 0.9798941798941799,
      "grad_norm": 6.034976482391357,
      "learning_rate": 1.0582010582010582e-06,
      "loss": 4.2766,
      "step": 926
    },
    {
      "epoch": 0.9809523809523809,
      "grad_norm": 3.5577170848846436,
      "learning_rate": 1.0052910052910054e-06,
      "loss": 4.2038,
      "step": 927
    },
    {
      "epoch": 0.982010582010582,
      "grad_norm": 3.4226174354553223,
      "learning_rate": 9.523809523809526e-07,
      "loss": 4.2256,
      "step": 928
    },
    {
      "epoch": 0.9830687830687831,
      "grad_norm": 3.801210880279541,
      "learning_rate": 8.994708994708995e-07,
      "loss": 4.6662,
      "step": 929
    },
    {
      "epoch": 0.9841269841269841,
      "grad_norm": 3.7857511043548584,
      "learning_rate": 8.465608465608467e-07,
      "loss": 4.349,
      "step": 930
    },
    {
      "epoch": 0.9851851851851852,
      "grad_norm": 4.348080635070801,
      "learning_rate": 7.936507936507937e-07,
      "loss": 4.2466,
      "step": 931
    },
    {
      "epoch": 0.9862433862433863,
      "grad_norm": 3.3915085792541504,
      "learning_rate": 7.407407407407408e-07,
      "loss": 4.4284,
      "step": 932
    },
    {
      "epoch": 0.9873015873015873,
      "grad_norm": 5.1464972496032715,
      "learning_rate": 6.878306878306879e-07,
      "loss": 4.2641,
      "step": 933
    },
    {
      "epoch": 0.9883597883597883,
      "grad_norm": 4.361679553985596,
      "learning_rate": 6.34920634920635e-07,
      "loss": 4.1923,
      "step": 934
    },
    {
      "epoch": 0.9894179894179894,
      "grad_norm": 4.495355606079102,
      "learning_rate": 5.82010582010582e-07,
      "loss": 4.1683,
      "step": 935
    },
    {
      "epoch": 0.9904761904761905,
      "grad_norm": 4.874444484710693,
      "learning_rate": 5.291005291005291e-07,
      "loss": 4.2167,
      "step": 936
    },
    {
      "epoch": 0.9915343915343915,
      "grad_norm": 3.7013728618621826,
      "learning_rate": 4.761904761904763e-07,
      "loss": 4.0992,
      "step": 937
    },
    {
      "epoch": 0.9925925925925926,
      "grad_norm": 6.150048732757568,
      "learning_rate": 4.2328042328042335e-07,
      "loss": 4.2488,
      "step": 938
    },
    {
      "epoch": 0.9936507936507937,
      "grad_norm": 5.740020751953125,
      "learning_rate": 3.703703703703704e-07,
      "loss": 4.4361,
      "step": 939
    },
    {
      "epoch": 0.9947089947089947,
      "grad_norm": 5.7900800704956055,
      "learning_rate": 3.174603174603175e-07,
      "loss": 4.5688,
      "step": 940
    },
    {
      "epoch": 0.9957671957671957,
      "grad_norm": 3.77296781539917,
      "learning_rate": 2.6455026455026455e-07,
      "loss": 4.3662,
      "step": 941
    },
    {
      "epoch": 0.9968253968253968,
      "grad_norm": 4.104562759399414,
      "learning_rate": 2.1164021164021167e-07,
      "loss": 4.3561,
      "step": 942
    },
    {
      "epoch": 0.9978835978835979,
      "grad_norm": 5.027563095092773,
      "learning_rate": 1.5873015873015874e-07,
      "loss": 4.3309,
      "step": 943
    },
    {
      "epoch": 0.9989417989417989,
      "grad_norm": 4.270631313323975,
      "learning_rate": 1.0582010582010584e-07,
      "loss": 4.3465,
      "step": 944
    },
    {
      "epoch": 1.0,
      "grad_norm": 4.4612956047058105,
      "learning_rate": 5.291005291005292e-08,
      "loss": 4.3776,
      "step": 945
    }
  ],
  "logging_steps": 1,
  "max_steps": 945,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1382850220916736.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}