{
  "best_metric": 1.3909703493118286,
  "best_model_checkpoint": "outputs/checkpoint-2955",
  "epoch": 0.2500211523817582,
  "eval_steps": 2955,
  "global_step": 2955,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 8.460952703274389e-05,
      "grad_norm": 2.34375,
      "learning_rate": 4e-05,
      "loss": 1.347,
      "step": 1
    },
    {
      "epoch": 0.00016921905406548778,
      "grad_norm": 3.140625,
      "learning_rate": 8e-05,
      "loss": 1.4858,
      "step": 2
    },
    {
      "epoch": 0.00025382858109823165,
      "grad_norm": 2.6875,
      "learning_rate": 0.00012,
      "loss": 1.1842,
      "step": 3
    },
    {
      "epoch": 0.00033843810813097557,
      "grad_norm": 2.953125,
      "learning_rate": 0.00016,
      "loss": 1.7778,
      "step": 4
    },
    {
      "epoch": 0.00042304763516371943,
      "grad_norm": 9.8125,
      "learning_rate": 0.0002,
      "loss": 1.0771,
      "step": 5
    },
    {
      "epoch": 0.0005076571621964633,
      "grad_norm": 2.15625,
      "learning_rate": 0.00024,
      "loss": 1.7402,
      "step": 6
    },
    {
      "epoch": 0.0005922666892292072,
      "grad_norm": 2.5,
      "learning_rate": 0.00028000000000000003,
      "loss": 1.6068,
      "step": 7
    },
    {
      "epoch": 0.0006768762162619511,
      "grad_norm": 2.0625,
      "learning_rate": 0.00032,
      "loss": 1.3865,
      "step": 8
    },
    {
      "epoch": 0.000761485743294695,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00035999999999999997,
      "loss": 1.5416,
      "step": 9
    },
    {
      "epoch": 0.0008460952703274389,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004,
      "loss": 1.2955,
      "step": 10
    },
    {
      "epoch": 0.0009307047973601827,
      "grad_norm": 4.28125,
      "learning_rate": 0.00044,
      "loss": 2.2166,
      "step": 11
    },
    {
      "epoch": 0.0010153143243929266,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048,
      "loss": 1.052,
      "step": 12
    },
    {
      "epoch": 0.0010999238514256705,
      "grad_norm": 2.328125,
      "learning_rate": 0.0005200000000000001,
      "loss": 1.5539,
      "step": 13
    },
    {
      "epoch": 0.0011845333784584143,
      "grad_norm": 1.59375,
      "learning_rate": 0.0005600000000000001,
      "loss": 1.4587,
      "step": 14
    },
    {
      "epoch": 0.0012691429054911584,
      "grad_norm": 1.453125,
      "learning_rate": 0.0006,
      "loss": 1.3232,
      "step": 15
    },
    {
      "epoch": 0.0013537524325239023,
      "grad_norm": 1.734375,
      "learning_rate": 0.00064,
      "loss": 1.5287,
      "step": 16
    },
    {
      "epoch": 0.0014383619595566461,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00068,
      "loss": 1.4651,
      "step": 17
    },
    {
      "epoch": 0.00152297148658939,
      "grad_norm": 1.484375,
      "learning_rate": 0.0007199999999999999,
      "loss": 1.1937,
      "step": 18
    },
    {
      "epoch": 0.0016075810136221339,
      "grad_norm": 1.671875,
      "learning_rate": 0.00076,
      "loss": 1.3257,
      "step": 19
    },
    {
      "epoch": 0.0016921905406548777,
      "grad_norm": 1.875,
      "learning_rate": 0.0008,
      "loss": 1.4073,
      "step": 20
    },
    {
      "epoch": 0.0017768000676876216,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00084,
      "loss": 1.6229,
      "step": 21
    },
    {
      "epoch": 0.0018614095947203655,
      "grad_norm": 2.0,
      "learning_rate": 0.00088,
      "loss": 1.3724,
      "step": 22
    },
    {
      "epoch": 0.0019460191217531093,
      "grad_norm": 1.734375,
      "learning_rate": 0.00092,
      "loss": 1.49,
      "step": 23
    },
    {
      "epoch": 0.002030628648785853,
      "grad_norm": 1.65625,
      "learning_rate": 0.00096,
      "loss": 1.0921,
      "step": 24
    },
    {
      "epoch": 0.0021152381758185973,
      "grad_norm": 2.6875,
      "learning_rate": 0.001,
      "loss": 1.7941,
      "step": 25
    },
    {
      "epoch": 0.002199847702851341,
      "grad_norm": 1.609375,
      "learning_rate": 0.000999999982261475,
      "loss": 1.4013,
      "step": 26
    },
    {
      "epoch": 0.002284457229884085,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009999999290459013,
      "loss": 2.0351,
      "step": 27
    },
    {
      "epoch": 0.0023690667569168287,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009999998403532825,
      "loss": 1.5376,
      "step": 28
    },
    {
      "epoch": 0.0024536762839495727,
      "grad_norm": 1.796875,
      "learning_rate": 0.000999999716183625,
      "loss": 1.2436,
      "step": 29
    },
    {
      "epoch": 0.002538285810982317,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009999995565369377,
      "loss": 1.3176,
      "step": 30
    },
    {
      "epoch": 0.0026228953380150605,
      "grad_norm": 2.109375,
      "learning_rate": 0.000999999361413232,
      "loss": 1.8695,
      "step": 31
    },
    {
      "epoch": 0.0027075048650478046,
      "grad_norm": 1.875,
      "learning_rate": 0.0009999991308125213,
      "loss": 1.3461,
      "step": 32
    },
    {
      "epoch": 0.002792114392080548,
      "grad_norm": 1.875,
      "learning_rate": 0.0009999988647348224,
      "loss": 1.1417,
      "step": 33
    },
    {
      "epoch": 0.0028767239191132923,
      "grad_norm": 2.15625,
      "learning_rate": 0.000999998563180154,
      "loss": 1.2394,
      "step": 34
    },
    {
      "epoch": 0.002961333446146036,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009999982261485374,
      "loss": 1.2879,
      "step": 35
    },
    {
      "epoch": 0.00304594297317878,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009999978536399969,
      "loss": 1.3137,
      "step": 36
    },
    {
      "epoch": 0.0031305525002115237,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009999974456545585,
      "loss": 1.4288,
      "step": 37
    },
    {
      "epoch": 0.0032151620272442677,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009999970021922515,
      "loss": 1.4801,
      "step": 38
    },
    {
      "epoch": 0.0032997715542770114,
      "grad_norm": 4.09375,
      "learning_rate": 0.000999996523253107,
      "loss": 1.4042,
      "step": 39
    },
    {
      "epoch": 0.0033843810813097555,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009999960088371593,
      "loss": 1.868,
      "step": 40
    },
    {
      "epoch": 0.0034689906083424996,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009999954589444446,
      "loss": 1.2621,
      "step": 41
    },
    {
      "epoch": 0.003553600135375243,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009999948735750022,
      "loss": 1.079,
      "step": 42
    },
    {
      "epoch": 0.0036382096624079873,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009999942527288736,
      "loss": 1.3093,
      "step": 43
    },
    {
      "epoch": 0.003722819189440731,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009999935964061027,
      "loss": 1.4501,
      "step": 44
    },
    {
      "epoch": 0.003807428716473475,
      "grad_norm": 2.65625,
      "learning_rate": 0.000999992904606736,
      "loss": 1.5224,
      "step": 45
    },
    {
      "epoch": 0.0038920382435062187,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009999921773308229,
      "loss": 1.2122,
      "step": 46
    },
    {
      "epoch": 0.003976647770538962,
      "grad_norm": 2.25,
      "learning_rate": 0.000999991414578415,
      "loss": 1.5258,
      "step": 47
    },
    {
      "epoch": 0.004061257297571706,
      "grad_norm": 2.328125,
      "learning_rate": 0.000999990616349566,
      "loss": 1.218,
      "step": 48
    },
    {
      "epoch": 0.0041458668246044505,
      "grad_norm": 2.625,
      "learning_rate": 0.0009999897826443328,
      "loss": 1.5019,
      "step": 49
    },
    {
      "epoch": 0.0042304763516371946,
      "grad_norm": 1.765625,
      "learning_rate": 0.0009999889134627747,
      "loss": 1.1043,
      "step": 50
    },
    {
      "epoch": 0.004315085878669939,
      "grad_norm": 1.7890625,
      "learning_rate": 0.000999988008804953,
      "loss": 1.6189,
      "step": 51
    },
    {
      "epoch": 0.004399695405702682,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009999870686709321,
      "loss": 1.5236,
      "step": 52
    },
    {
      "epoch": 0.004484304932735426,
      "grad_norm": 2.296875,
      "learning_rate": 0.000999986093060779,
      "loss": 1.8329,
      "step": 53
    },
    {
      "epoch": 0.00456891445976817,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009999850819745624,
      "loss": 1.5464,
      "step": 54
    },
    {
      "epoch": 0.004653523986800914,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009999840354123545,
      "loss": 1.3176,
      "step": 55
    },
    {
      "epoch": 0.004738133513833657,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009999829533742291,
      "loss": 1.2979,
      "step": 56
    },
    {
      "epoch": 0.004822743040866401,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009999818358602632,
      "loss": 1.2222,
      "step": 57
    },
    {
      "epoch": 0.0049073525678991455,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009999806828705363,
      "loss": 1.1931,
      "step": 58
    },
    {
      "epoch": 0.00499196209493189,
      "grad_norm": 1.84375,
      "learning_rate": 0.00099997949440513,
      "loss": 1.2699,
      "step": 59
    },
    {
      "epoch": 0.005076571621964634,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009999782704641284,
      "loss": 1.7302,
      "step": 60
    },
    {
      "epoch": 0.005161181148997377,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009999770110476187,
      "loss": 1.3714,
      "step": 61
    },
    {
      "epoch": 0.005245790676030121,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009999757161556903,
      "loss": 1.2651,
      "step": 62
    },
    {
      "epoch": 0.005330400203062865,
      "grad_norm": 3.921875,
      "learning_rate": 0.0009999743857884347,
      "loss": 1.1328,
      "step": 63
    },
    {
      "epoch": 0.005415009730095609,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009999730199459465,
      "loss": 1.5003,
      "step": 64
    },
    {
      "epoch": 0.005499619257128352,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009999716186283227,
      "loss": 1.4297,
      "step": 65
    },
    {
      "epoch": 0.005584228784161096,
      "grad_norm": 1.5,
      "learning_rate": 0.0009999701818356628,
      "loss": 1.3476,
      "step": 66
    },
    {
      "epoch": 0.0056688383111938405,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009999687095680685,
      "loss": 1.3231,
      "step": 67
    },
    {
      "epoch": 0.005753447838226585,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009999672018256443,
      "loss": 1.3971,
      "step": 68
    },
    {
      "epoch": 0.005838057365259328,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0009999656586084974,
      "loss": 1.0977,
      "step": 69
    },
    {
      "epoch": 0.005922666892292072,
      "grad_norm": 2.375,
      "learning_rate": 0.000999964079916737,
      "loss": 1.6166,
      "step": 70
    },
    {
      "epoch": 0.006007276419324816,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009999624657504754,
      "loss": 1.2498,
      "step": 71
    },
    {
      "epoch": 0.00609188594635756,
      "grad_norm": 3.3125,
      "learning_rate": 0.000999960816109827,
      "loss": 1.2013,
      "step": 72
    },
    {
      "epoch": 0.006176495473390304,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009999591309949086,
      "loss": 1.6118,
      "step": 73
    },
    {
      "epoch": 0.006261105000423047,
      "grad_norm": 2.078125,
      "learning_rate": 0.00099995741040584,
      "loss": 1.5442,
      "step": 74
    },
    {
      "epoch": 0.006345714527455791,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009999556543427435,
      "loss": 1.088,
      "step": 75
    },
    {
      "epoch": 0.0064303240544885355,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009999538628057433,
      "loss": 1.1979,
      "step": 76
    },
    {
      "epoch": 0.00651493358152128,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000999952035794967,
      "loss": 1.4976,
      "step": 77
    },
    {
      "epoch": 0.006599543108554023,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009999501733105438,
      "loss": 1.2319,
      "step": 78
    },
    {
      "epoch": 0.006684152635586767,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009999482753526058,
      "loss": 1.5253,
      "step": 79
    },
    {
      "epoch": 0.006768762162619511,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009999463419212882,
      "loss": 1.5161,
      "step": 80
    },
    {
      "epoch": 0.006853371689652255,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009999443730167275,
      "loss": 1.336,
      "step": 81
    },
    {
      "epoch": 0.006937981216684999,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009999423686390641,
      "loss": 1.253,
      "step": 82
    },
    {
      "epoch": 0.007022590743717742,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009999403287884396,
      "loss": 1.3102,
      "step": 83
    },
    {
      "epoch": 0.007107200270750486,
      "grad_norm": 2.25,
      "learning_rate": 0.000999938253464999,
      "loss": 1.1867,
      "step": 84
    },
    {
      "epoch": 0.0071918097977832305,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009999361426688896,
      "loss": 1.2959,
      "step": 85
    },
    {
      "epoch": 0.007276419324815975,
      "grad_norm": 2.453125,
      "learning_rate": 0.000999933996400261,
      "loss": 1.4914,
      "step": 86
    },
    {
      "epoch": 0.007361028851848718,
      "grad_norm": 1.9453125,
      "learning_rate": 0.000999931814659266,
      "loss": 1.2173,
      "step": 87
    },
    {
      "epoch": 0.007445638378881462,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009999295974460586,
      "loss": 1.5319,
      "step": 88
    },
    {
      "epoch": 0.007530247905914206,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009999273447607968,
      "loss": 1.2836,
      "step": 89
    },
    {
      "epoch": 0.00761485743294695,
      "grad_norm": 1.59375,
      "learning_rate": 0.00099992505660364,
      "loss": 1.1978,
      "step": 90
    },
    {
      "epoch": 0.007699466959979694,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009999227329747509,
      "loss": 1.1352,
      "step": 91
    },
    {
      "epoch": 0.007784076487012437,
      "grad_norm": 1.8125,
      "learning_rate": 0.000999920373874294,
      "loss": 1.6017,
      "step": 92
    },
    {
      "epoch": 0.007868686014045181,
      "grad_norm": 1.90625,
      "learning_rate": 0.000999917979302437,
      "loss": 1.2574,
      "step": 93
    },
    {
      "epoch": 0.007953295541077925,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009999155492593497,
      "loss": 1.7009,
      "step": 94
    },
    {
      "epoch": 0.00803790506811067,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009999130837452044,
      "loss": 1.1946,
      "step": 95
    },
    {
      "epoch": 0.008122514595143413,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009999105827601762,
      "loss": 1.2913,
      "step": 96
    },
    {
      "epoch": 0.008207124122176158,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009999080463044426,
      "loss": 1.2741,
      "step": 97
    },
    {
      "epoch": 0.008291733649208901,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009999054743781832,
      "loss": 1.5001,
      "step": 98
    },
    {
      "epoch": 0.008376343176241644,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009999028669815811,
      "loss": 1.1845,
      "step": 99
    },
    {
      "epoch": 0.008460952703274389,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009999002241148206,
      "loss": 1.2613,
      "step": 100
    },
    {
      "epoch": 0.008545562230307132,
      "grad_norm": 2.625,
      "learning_rate": 0.0009998975457780898,
      "loss": 1.4781,
      "step": 101
    },
    {
      "epoch": 0.008630171757339877,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009998948319715785,
      "loss": 1.3914,
      "step": 102
    },
    {
      "epoch": 0.00871478128437262,
      "grad_norm": 1.71875,
      "learning_rate": 0.0009998920826954794,
      "loss": 1.0609,
      "step": 103
    },
    {
      "epoch": 0.008799390811405364,
      "grad_norm": 3.0,
      "learning_rate": 0.0009998892979499875,
      "loss": 1.1943,
      "step": 104
    },
    {
      "epoch": 0.008884000338438109,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009998864777353,
      "loss": 1.2727,
      "step": 105
    },
    {
      "epoch": 0.008968609865470852,
      "grad_norm": 1.53125,
      "learning_rate": 0.0009998836220516177,
      "loss": 1.0181,
      "step": 106
    },
    {
      "epoch": 0.009053219392503595,
      "grad_norm": 2.21875,
      "learning_rate": 0.000999880730899143,
      "loss": 1.2839,
      "step": 107
    },
    {
      "epoch": 0.00913782891953634,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009998778042780805,
      "loss": 1.3362,
      "step": 108
    },
    {
      "epoch": 0.009222438446569083,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009998748421886383,
      "loss": 1.1448,
      "step": 109
    },
    {
      "epoch": 0.009307047973601828,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009998718446310268,
      "loss": 1.1337,
      "step": 110
    },
    {
      "epoch": 0.009391657500634571,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009998688116054583,
      "loss": 1.2494,
      "step": 111
    },
    {
      "epoch": 0.009476267027667315,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009998657431121482,
      "loss": 1.1997,
      "step": 112
    },
    {
      "epoch": 0.00956087655470006,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009998626391513141,
      "loss": 1.5433,
      "step": 113
    },
    {
      "epoch": 0.009645486081732803,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009998594997231763,
      "loss": 1.2961,
      "step": 114
    },
    {
      "epoch": 0.009730095608765548,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009998563248279574,
      "loss": 1.1966,
      "step": 115
    },
    {
      "epoch": 0.009814705135798291,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009998531144658832,
      "loss": 1.0839,
      "step": 116
    },
    {
      "epoch": 0.009899314662831034,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009998498686371808,
      "loss": 1.2408,
      "step": 117
    },
    {
      "epoch": 0.00998392418986378,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009998465873420808,
      "loss": 1.355,
      "step": 118
    },
    {
      "epoch": 0.010068533716896522,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009998432705808163,
      "loss": 1.3892,
      "step": 119
    },
    {
      "epoch": 0.010153143243929267,
      "grad_norm": 2.03125,
      "learning_rate": 0.000999839918353622,
      "loss": 1.5594,
      "step": 120
    },
    {
      "epoch": 0.01023775277096201,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009998365306607365,
      "loss": 1.1544,
      "step": 121
    },
    {
      "epoch": 0.010322362297994754,
      "grad_norm": 1.765625,
      "learning_rate": 0.0009998331075023997,
      "loss": 1.0929,
      "step": 122
    },
    {
      "epoch": 0.010406971825027499,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009998296488788545,
      "loss": 1.2034,
      "step": 123
    },
    {
      "epoch": 0.010491581352060242,
      "grad_norm": 5.625,
      "learning_rate": 0.0009998261547903464,
      "loss": 1.2612,
      "step": 124
    },
    {
      "epoch": 0.010576190879092985,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009998226252371234,
      "loss": 1.4275,
      "step": 125
    },
    {
      "epoch": 0.01066080040612573,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009998190602194357,
      "loss": 1.3639,
      "step": 126
    },
    {
      "epoch": 0.010745409933158473,
      "grad_norm": 5.09375,
      "learning_rate": 0.0009998154597375367,
      "loss": 1.5568,
      "step": 127
    },
    {
      "epoch": 0.010830019460191218,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009998118237916813,
      "loss": 1.2039,
      "step": 128
    },
    {
      "epoch": 0.010914628987223961,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009998081523821278,
      "loss": 1.2754,
      "step": 129
    },
    {
      "epoch": 0.010999238514256705,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009998044455091367,
      "loss": 1.4505,
      "step": 130
    },
    {
      "epoch": 0.01108384804128945,
      "grad_norm": 2.171875,
      "learning_rate": 0.000999800703172971,
      "loss": 1.6079,
      "step": 131
    },
    {
      "epoch": 0.011168457568322193,
      "grad_norm": 1.5,
      "learning_rate": 0.000999796925373896,
      "loss": 1.1853,
      "step": 132
    },
    {
      "epoch": 0.011253067095354938,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009997931121121801,
      "loss": 1.5865,
      "step": 133
    },
    {
      "epoch": 0.011337676622387681,
      "grad_norm": 2.015625,
      "learning_rate": 0.000999789263388094,
      "loss": 1.2847,
      "step": 134
    },
    {
      "epoch": 0.011422286149420424,
      "grad_norm": 1.875,
      "learning_rate": 0.00099978537920191,
      "loss": 1.6055,
      "step": 135
    },
    {
      "epoch": 0.01150689567645317,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009997814595539046,
      "loss": 1.0537,
      "step": 136
    },
    {
      "epoch": 0.011591505203485912,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009997775044443553,
      "loss": 1.2045,
      "step": 137
    },
    {
      "epoch": 0.011676114730518656,
      "grad_norm": 2.125,
      "learning_rate": 0.0009997735138735431,
      "loss": 1.3257,
      "step": 138
    },
    {
      "epoch": 0.0117607242575514,
      "grad_norm": 2.375,
      "learning_rate": 0.000999769487841751,
      "loss": 1.8842,
      "step": 139
    },
    {
      "epoch": 0.011845333784584144,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009997654263492646,
      "loss": 1.6685,
      "step": 140
    },
    {
      "epoch": 0.011929943311616889,
      "grad_norm": 2.09375,
      "learning_rate": 0.000999761329396372,
      "loss": 1.2609,
      "step": 141
    },
    {
      "epoch": 0.012014552838649632,
      "grad_norm": 21.625,
      "learning_rate": 0.0009997571969833643,
      "loss": 1.3058,
      "step": 142
    },
    {
      "epoch": 0.012099162365682375,
      "grad_norm": 97.5,
      "learning_rate": 0.0009997530291105344,
      "loss": 1.5144,
      "step": 143
    },
    {
      "epoch": 0.01218377189271512,
      "grad_norm": 2.765625,
      "learning_rate": 0.000999748825778178,
      "loss": 1.6006,
      "step": 144
    },
    {
      "epoch": 0.012268381419747863,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009997445869865936,
      "loss": 1.1881,
      "step": 145
    },
    {
      "epoch": 0.012352990946780608,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009997403127360817,
      "loss": 1.4521,
      "step": 146
    },
    {
      "epoch": 0.012437600473813351,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009997360030269454,
      "loss": 1.6141,
      "step": 147
    },
    {
      "epoch": 0.012522210000846095,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009997316578594909,
      "loss": 1.6004,
      "step": 148
    },
    {
      "epoch": 0.01260681952787884,
      "grad_norm": 7.0,
      "learning_rate": 0.0009997272772340265,
      "loss": 1.439,
      "step": 149
    },
    {
      "epoch": 0.012691429054911583,
      "grad_norm": 4.4375,
      "learning_rate": 0.0009997228611508628,
      "loss": 1.1298,
      "step": 150
    },
    {
      "epoch": 0.012776038581944328,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000999718409610313,
      "loss": 1.3315,
      "step": 151
    },
    {
      "epoch": 0.012860648108977071,
      "grad_norm": 172.0,
      "learning_rate": 0.0009997139226126934,
      "loss": 1.7768,
      "step": 152
    },
    {
      "epoch": 0.012945257636009814,
      "grad_norm": 1.75,
      "learning_rate": 0.0009997094001583222,
      "loss": 1.0968,
      "step": 153
    },
    {
      "epoch": 0.01302986716304256,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009997048422475201,
      "loss": 1.2392,
      "step": 154
    },
    {
      "epoch": 0.013114476690075302,
      "grad_norm": 2.25,
      "learning_rate": 0.0009997002488806108,
      "loss": 1.3551,
      "step": 155
    },
    {
      "epoch": 0.013199086217108046,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009996956200579198,
      "loss": 1.4327,
      "step": 156
    },
    {
      "epoch": 0.01328369574414079,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009996909557797762,
      "loss": 2.0459,
      "step": 157
    },
    {
      "epoch": 0.013368305271173534,
      "grad_norm": 6.03125,
      "learning_rate": 0.00099968625604651,
      "loss": 1.4953,
      "step": 158
    },
    {
      "epoch": 0.013452914798206279,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009996815208584556,
      "loss": 1.1725,
      "step": 159
    },
    {
      "epoch": 0.013537524325239022,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009996767502159485,
      "loss": 1.0878,
      "step": 160
    },
    {
      "epoch": 0.013622133852271765,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009996719441193271,
      "loss": 1.3483,
      "step": 161
    },
    {
      "epoch": 0.01370674337930451,
      "grad_norm": 2.5,
      "learning_rate": 0.0009996671025689328,
      "loss": 1.7677,
      "step": 162
    },
    {
      "epoch": 0.013791352906337253,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009996622255651088,
      "loss": 1.4143,
      "step": 163
    },
    {
      "epoch": 0.013875962433369998,
      "grad_norm": 1.75,
      "learning_rate": 0.0009996573131082014,
      "loss": 1.2013,
      "step": 164
    },
    {
      "epoch": 0.013960571960402741,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009996523651985589,
      "loss": 1.6487,
      "step": 165
    },
    {
      "epoch": 0.014045181487435485,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009996473818365327,
      "loss": 1.2965,
      "step": 166
    },
    {
      "epoch": 0.01412979101446823,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009996423630224758,
      "loss": 1.2192,
      "step": 167
    },
    {
      "epoch": 0.014214400541500973,
      "grad_norm": 2.578125,
      "learning_rate": 0.000999637308756745,
      "loss": 1.2958,
      "step": 168
    },
    {
      "epoch": 0.014299010068533716,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009996322190396986,
      "loss": 1.8261,
      "step": 169
    },
    {
      "epoch": 0.014383619595566461,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009996270938716976,
      "loss": 1.4484,
      "step": 170
    },
    {
      "epoch": 0.014468229122599204,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009996219332531059,
      "loss": 1.1201,
      "step": 171
    },
    {
      "epoch": 0.01455283864963195,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009996167371842896,
      "loss": 1.4118,
      "step": 172
    },
    {
      "epoch": 0.014637448176664692,
      "grad_norm": 1.8515625,
      "learning_rate": 0.000999611505665617,
      "loss": 1.4996,
      "step": 173
    },
    {
      "epoch": 0.014722057703697436,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009996062386974601,
      "loss": 1.471,
      "step": 174
    },
    {
      "epoch": 0.01480666723073018,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009996009362801918,
      "loss": 1.4913,
      "step": 175
    },
    {
      "epoch": 0.014891276757762924,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009995955984141891,
      "loss": 1.2942,
      "step": 176
    },
    {
      "epoch": 0.014975886284795669,
      "grad_norm": 1.875,
      "learning_rate": 0.0009995902250998301,
      "loss": 1.3884,
      "step": 177
    },
    {
      "epoch": 0.015060495811828412,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009995848163374963,
      "loss": 1.3034,
      "step": 178
    },
    {
      "epoch": 0.015145105338861155,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009995793721275714,
      "loss": 1.4782,
      "step": 179
    },
    {
      "epoch": 0.0152297148658939,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009995738924704419,
      "loss": 1.552,
      "step": 180
    },
    {
      "epoch": 0.015314324392926643,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009995683773664963,
      "loss": 1.6338,
      "step": 181
    },
    {
      "epoch": 0.015398933919959388,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009995628268161263,
      "loss": 1.6083,
      "step": 182
    },
    {
      "epoch": 0.015483543446992131,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009995572408197254,
      "loss": 1.3587,
      "step": 183
    },
    {
      "epoch": 0.015568152974024875,
      "grad_norm": 2.03125,
      "learning_rate": 0.00099955161937769,
      "loss": 1.2394,
      "step": 184
    },
    {
      "epoch": 0.01565276250105762,
      "grad_norm": 2.359375,
      "learning_rate": 0.000999545962490419,
      "loss": 1.6407,
      "step": 185
    },
    {
      "epoch": 0.015737372028090363,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009995402701583138,
      "loss": 2.0192,
      "step": 186
    },
    {
      "epoch": 0.015821981555123106,
      "grad_norm": 2.125,
      "learning_rate": 0.0009995345423817784,
      "loss": 1.3759,
      "step": 187
    },
    {
      "epoch": 0.01590659108215585,
      "grad_norm": 2.4375,
      "learning_rate": 0.000999528779161219,
      "loss": 1.5867,
      "step": 188
    },
    {
      "epoch": 0.015991200609188596,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009995229804970448,
      "loss": 1.4915,
      "step": 189
    },
    {
      "epoch": 0.01607581013622134,
      "grad_norm": 2.609375,
      "learning_rate": 0.000999517146389667,
      "loss": 1.4121,
      "step": 190
    },
    {
      "epoch": 0.016160419663254082,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0009995112768394997,
      "loss": 1.4422,
      "step": 191
    },
    {
      "epoch": 0.016245029190286826,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009995053718469592,
      "loss": 1.1671,
      "step": 192
    },
    {
      "epoch": 0.01632963871731957,
      "grad_norm": 2.125,
      "learning_rate": 0.0009994994314124646,
      "loss": 1.5485,
      "step": 193
    },
    {
      "epoch": 0.016414248244352315,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009994934555364373,
      "loss": 1.5435,
      "step": 194
    },
    {
      "epoch": 0.01649885777138506,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009994874442193014,
      "loss": 1.373,
      "step": 195
    },
    {
      "epoch": 0.016583467298417802,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009994813974614834,
      "loss": 1.4975,
      "step": 196
    },
    {
      "epoch": 0.016668076825450545,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009994753152634125,
      "loss": 1.294,
      "step": 197
    },
    {
      "epoch": 0.01675268635248329,
      "grad_norm": 1.578125,
      "learning_rate": 0.0009994691976255201,
      "loss": 1.1618,
      "step": 198
    },
    {
      "epoch": 0.016837295879516035,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009994630445482402,
      "loss": 1.3071,
      "step": 199
    },
    {
      "epoch": 0.016921905406548778,
      "grad_norm": 8.0625,
      "learning_rate": 0.0009994568560320095,
      "loss": 1.4145,
      "step": 200
    },
    {
      "epoch": 0.01700651493358152,
      "grad_norm": 2.125,
      "learning_rate": 0.0009994506320772669,
      "loss": 1.3224,
      "step": 201
    },
    {
      "epoch": 0.017091124460614265,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0009994443726844544,
      "loss": 1.2453,
      "step": 202
    },
    {
      "epoch": 0.017175733987647008,
      "grad_norm": 2.09375,
      "learning_rate": 0.000999438077854016,
      "loss": 2.112,
      "step": 203
    },
    {
      "epoch": 0.017260343514679755,
      "grad_norm": 1.875,
      "learning_rate": 0.000999431747586398,
      "loss": 1.336,
      "step": 204
    },
    {
      "epoch": 0.017344953041712498,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00099942538188205,
      "loss": 1.6594,
      "step": 205
    },
    {
      "epoch": 0.01742956256874524,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009994189807414236,
      "loss": 1.4261,
      "step": 206
    },
    {
      "epoch": 0.017514172095777984,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009994125441649727,
      "loss": 1.8957,
      "step": 207
    },
    {
      "epoch": 0.017598781622810727,
      "grad_norm": 2.140625,
      "learning_rate": 0.000999406072153154,
      "loss": 1.7651,
      "step": 208
    },
    {
      "epoch": 0.01768339114984347,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0009993995647064272,
      "loss": 1.1423,
      "step": 209
    },
    {
      "epoch": 0.017768000676876217,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009993930218252537,
      "loss": 1.2428,
      "step": 210
    },
    {
      "epoch": 0.01785261020390896,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009993864435100977,
      "loss": 1.5758,
      "step": 211
    },
    {
      "epoch": 0.017937219730941704,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009993798297614261,
      "loss": 1.3686,
      "step": 212
    },
    {
      "epoch": 0.018021829257974447,
      "grad_norm": 1.625,
      "learning_rate": 0.0009993731805797079,
      "loss": 1.3278,
      "step": 213
    },
    {
      "epoch": 0.01810643878500719,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009993664959654155,
      "loss": 1.4686,
      "step": 214
    },
    {
      "epoch": 0.018191048312039937,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009993597759190224,
      "loss": 1.4954,
      "step": 215
    },
    {
      "epoch": 0.01827565783907268,
      "grad_norm": 2.90625,
      "learning_rate": 0.000999353020441006,
      "loss": 2.0094,
      "step": 216
    },
    {
      "epoch": 0.018360267366105423,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009993462295318453,
      "loss": 1.3573,
      "step": 217
    },
    {
      "epoch": 0.018444876893138167,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009993394031920222,
      "loss": 1.5403,
      "step": 218
    },
    {
      "epoch": 0.01852948642017091,
      "grad_norm": 5.625,
      "learning_rate": 0.0009993325414220213,
      "loss": 1.6052,
      "step": 219
    },
    {
      "epoch": 0.018614095947203656,
      "grad_norm": 2.0,
      "learning_rate": 0.0009993256442223295,
      "loss": 1.1354,
      "step": 220
    },
    {
      "epoch": 0.0186987054742364,
      "grad_norm": 16.625,
      "learning_rate": 0.0009993187115934358,
      "loss": 1.6388,
      "step": 221
    },
    {
      "epoch": 0.018783315001269143,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009993117435358322,
      "loss": 1.5376,
      "step": 222
    },
    {
      "epoch": 0.018867924528301886,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009993047400500134,
      "loss": 1.6391,
      "step": 223
    },
    {
      "epoch": 0.01895253405533463,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009992977011364758,
      "loss": 1.3547,
      "step": 224
    },
    {
      "epoch": 0.019037143582367376,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009992906267957194,
      "loss": 1.6937,
      "step": 225
    },
    {
      "epoch": 0.01912175310940012,
      "grad_norm": 2.125,
      "learning_rate": 0.0009992835170282459,
      "loss": 1.5653,
      "step": 226
    },
    {
      "epoch": 0.019206362636432862,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009992763718345596,
      "loss": 1.2953,
      "step": 227
    },
    {
      "epoch": 0.019290972163465606,
      "grad_norm": 1.9140625,
      "learning_rate": 0.000999269191215168,
      "loss": 1.3038,
      "step": 228
    },
    {
      "epoch": 0.01937558169049835,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009992619751705801,
      "loss": 1.3373,
      "step": 229
    },
    {
      "epoch": 0.019460191217531096,
      "grad_norm": 2.609375,
      "learning_rate": 0.000999254723701308,
      "loss": 1.4272,
      "step": 230
    },
    {
      "epoch": 0.01954480074456384,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009992474368078663,
      "loss": 1.595,
      "step": 231
    },
    {
      "epoch": 0.019629410271596582,
      "grad_norm": 3.046875,
      "learning_rate": 0.000999240114490772,
      "loss": 1.5275,
      "step": 232
    },
    {
      "epoch": 0.019714019798629325,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000999232756750545,
      "loss": 1.3026,
      "step": 233
    },
    {
      "epoch": 0.01979862932566207,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009992253635877067,
      "loss": 2.0244,
      "step": 234
    },
    {
      "epoch": 0.019883238852694815,
      "grad_norm": 1.75,
      "learning_rate": 0.0009992179350027822,
      "loss": 1.7058,
      "step": 235
    },
    {
      "epoch": 0.01996784837972756,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009992104709962981,
      "loss": 1.7791,
      "step": 236
    },
    {
      "epoch": 0.0200524579067603,
      "grad_norm": 1.578125,
      "learning_rate": 0.0009992029715687847,
      "loss": 1.261,
      "step": 237
    },
    {
      "epoch": 0.020137067433793045,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009991954367207733,
      "loss": 1.322,
      "step": 238
    },
    {
      "epoch": 0.020221676960825788,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009991878664527992,
      "loss": 1.3924,
      "step": 239
    },
    {
      "epoch": 0.020306286487858535,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009991802607653992,
      "loss": 1.3023,
      "step": 240
    },
    {
      "epoch": 0.020390896014891278,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009991726196591133,
      "loss": 1.6224,
      "step": 241
    },
    {
      "epoch": 0.02047550554192402,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009991649431344831,
      "loss": 1.2995,
      "step": 242
    },
    {
      "epoch": 0.020560115068956764,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009991572311920536,
      "loss": 1.1433,
      "step": 243
    },
    {
      "epoch": 0.020644724595989507,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009991494838323723,
      "loss": 1.2607,
      "step": 244
    },
    {
      "epoch": 0.02072933412302225,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009991417010559883,
      "loss": 1.2243,
      "step": 245
    },
    {
      "epoch": 0.020813943650054997,
      "grad_norm": 1.390625,
      "learning_rate": 0.0009991338828634541,
      "loss": 1.0471,
      "step": 246
    },
    {
      "epoch": 0.02089855317708774,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009991260292553246,
      "loss": 1.7105,
      "step": 247
    },
    {
      "epoch": 0.020983162704120484,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009991181402321568,
      "loss": 1.4738,
      "step": 248
    },
    {
      "epoch": 0.021067772231153227,
      "grad_norm": 6.09375,
      "learning_rate": 0.0009991102157945106,
      "loss": 1.5079,
      "step": 249
    },
    {
      "epoch": 0.02115238175818597,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000999102255942948,
      "loss": 1.2744,
      "step": 250
    },
    {
      "epoch": 0.021236991285218717,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009990942606780343,
      "loss": 1.6377,
      "step": 251
    },
    {
      "epoch": 0.02132160081225146,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009990862300003364,
      "loss": 1.9196,
      "step": 252
    },
    {
      "epoch": 0.021406210339284203,
      "grad_norm": 2.125,
      "learning_rate": 0.0009990781639104243,
      "loss": 1.5648,
      "step": 253
    },
    {
      "epoch": 0.021490819866316947,
      "grad_norm": 2.125,
      "learning_rate": 0.00099907006240887,
      "loss": 1.3309,
      "step": 254
    },
    {
      "epoch": 0.02157542939334969,
      "grad_norm": 1.390625,
      "learning_rate": 0.0009990619254962487,
      "loss": 1.1751,
      "step": 255
    },
    {
      "epoch": 0.021660038920382436,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009990537531731375,
      "loss": 1.2317,
      "step": 256
    },
    {
      "epoch": 0.02174464844741518,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009990455454401165,
      "loss": 1.2704,
      "step": 257
    },
    {
      "epoch": 0.021829257974447923,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009990373022977681,
      "loss": 1.214,
      "step": 258
    },
    {
      "epoch": 0.021913867501480666,
      "grad_norm": 1.875,
      "learning_rate": 0.0009990290237466767,
      "loss": 1.3812,
      "step": 259
    },
    {
      "epoch": 0.02199847702851341,
      "grad_norm": 2.625,
      "learning_rate": 0.0009990207097874304,
      "loss": 1.4461,
      "step": 260
    },
    {
      "epoch": 0.022083086555546156,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009990123604206186,
      "loss": 1.7456,
      "step": 261
    },
    {
      "epoch": 0.0221676960825789,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009990039756468335,
      "loss": 1.7591,
      "step": 262
    },
    {
      "epoch": 0.022252305609611642,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009989955554666708,
      "loss": 1.3499,
      "step": 263
    },
    {
      "epoch": 0.022336915136644386,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009989870998807276,
      "loss": 1.6826,
      "step": 264
    },
    {
      "epoch": 0.02242152466367713,
      "grad_norm": 1.765625,
      "learning_rate": 0.0009989786088896036,
      "loss": 1.1738,
      "step": 265
    },
    {
      "epoch": 0.022506134190709876,
      "grad_norm": 5.46875,
      "learning_rate": 0.0009989700824939013,
      "loss": 2.0315,
      "step": 266
    },
    {
      "epoch": 0.02259074371774262,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009989615206942263,
      "loss": 1.1226,
      "step": 267
    },
    {
      "epoch": 0.022675353244775362,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009989529234911852,
      "loss": 1.2221,
      "step": 268
    },
    {
      "epoch": 0.022759962771808105,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009989442908853888,
      "loss": 1.486,
      "step": 269
    },
    {
      "epoch": 0.02284457229884085,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009989356228774491,
      "loss": 1.3799,
      "step": 270
    },
    {
      "epoch": 0.022929181825873595,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009989269194679814,
      "loss": 1.1169,
      "step": 271
    },
    {
      "epoch": 0.02301379135290634,
      "grad_norm": 3.328125,
      "learning_rate": 0.000998918180657603,
      "loss": 1.8589,
      "step": 272
    },
    {
      "epoch": 0.02309840087993908,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009989094064469341,
      "loss": 1.758,
      "step": 273
    },
    {
      "epoch": 0.023183010406971825,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009989005968365973,
      "loss": 1.4167,
      "step": 274
    },
    {
      "epoch": 0.023267619934004568,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009988917518272177,
      "loss": 1.23,
      "step": 275
    },
    {
      "epoch": 0.02335222946103731,
      "grad_norm": 2.65625,
      "learning_rate": 0.000998882871419423,
      "loss": 1.5197,
      "step": 276
    },
    {
      "epoch": 0.023436838988070058,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009988739556138428,
      "loss": 1.192,
      "step": 277
    },
    {
      "epoch": 0.0235214485151028,
      "grad_norm": 2.375,
      "learning_rate": 0.0009988650044111102,
      "loss": 1.3941,
      "step": 278
    },
    {
      "epoch": 0.023606058042135544,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009988560178118603,
      "loss": 1.9178,
      "step": 279
    },
    {
      "epoch": 0.023690667569168287,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009988469958167304,
      "loss": 1.6953,
      "step": 280
    },
    {
      "epoch": 0.02377527709620103,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009988379384263607,
      "loss": 1.6486,
      "step": 281
    },
    {
      "epoch": 0.023859886623233777,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009988288456413944,
      "loss": 1.6849,
      "step": 282
    },
    {
      "epoch": 0.02394449615026652,
      "grad_norm": 9.4375,
      "learning_rate": 0.0009988197174624761,
      "loss": 1.8455,
      "step": 283
    },
    {
      "epoch": 0.024029105677299264,
      "grad_norm": 2.75,
      "learning_rate": 0.0009988105538902537,
      "loss": 1.6421,
      "step": 284
    },
    {
      "epoch": 0.024113715204332007,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009988013549253774,
      "loss": 1.4932,
      "step": 285
    },
    {
      "epoch": 0.02419832473136475,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009987921205684997,
      "loss": 1.1856,
      "step": 286
    },
    {
      "epoch": 0.024282934258397497,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009987828508202764,
      "loss": 1.236,
      "step": 287
    },
    {
      "epoch": 0.02436754378543024,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009987735456813644,
      "loss": 1.5252,
      "step": 288
    },
    {
      "epoch": 0.024452153312462983,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009987642051524244,
      "loss": 1.1868,
      "step": 289
    },
    {
      "epoch": 0.024536762839495727,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009987548292341192,
      "loss": 1.9085,
      "step": 290
    },
    {
      "epoch": 0.02462137236652847,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009987454179271138,
      "loss": 1.4697,
      "step": 291
    },
    {
      "epoch": 0.024705981893561216,
      "grad_norm": 2.125,
      "learning_rate": 0.0009987359712320763,
      "loss": 1.3808,
      "step": 292
    },
    {
      "epoch": 0.02479059142059396,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009987264891496768,
      "loss": 1.4499,
      "step": 293
    },
    {
      "epoch": 0.024875200947626703,
      "grad_norm": 2.59375,
      "learning_rate": 0.000998716971680588,
      "loss": 1.3049,
      "step": 294
    },
    {
      "epoch": 0.024959810474659446,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009987074188254853,
      "loss": 1.3704,
      "step": 295
    },
    {
      "epoch": 0.02504442000169219,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009986978305850467,
      "loss": 1.6277,
      "step": 296
    },
    {
      "epoch": 0.025129029528724936,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009986882069599522,
      "loss": 1.4311,
      "step": 297
    },
    {
      "epoch": 0.02521363905575768,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009986785479508846,
      "loss": 1.0015,
      "step": 298
    },
    {
      "epoch": 0.025298248582790422,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009986688535585297,
      "loss": 1.2818,
      "step": 299
    },
    {
      "epoch": 0.025382858109823166,
      "grad_norm": 3.390625,
      "learning_rate": 0.000998659123783575,
      "loss": 1.459,
      "step": 300
    },
    {
      "epoch": 0.02546746763685591,
      "grad_norm": 3.453125,
      "learning_rate": 0.000998649358626711,
      "loss": 1.597,
      "step": 301
    },
    {
      "epoch": 0.025552077163888656,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009986395580886304,
      "loss": 1.6184,
      "step": 302
    },
    {
      "epoch": 0.0256366866909214,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009986297221700286,
      "loss": 1.6303,
      "step": 303
    },
    {
      "epoch": 0.025721296217954142,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009986198508716037,
      "loss": 1.2013,
      "step": 304
    },
    {
      "epoch": 0.025805905744986885,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009986099441940562,
      "loss": 1.5271,
      "step": 305
    },
    {
      "epoch": 0.02589051527201963,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009986000021380888,
      "loss": 1.5848,
      "step": 306
    },
    {
      "epoch": 0.02597512479905237,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009985900247044069,
      "loss": 1.5211,
      "step": 307
    },
    {
      "epoch": 0.02605973432608512,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009985800118937184,
      "loss": 1.5063,
      "step": 308
    },
    {
      "epoch": 0.02614434385311786,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009985699637067341,
      "loss": 1.4056,
      "step": 309
    },
    {
      "epoch": 0.026228953380150605,
      "grad_norm": 12.1875,
      "learning_rate": 0.0009985598801441665,
      "loss": 1.1841,
      "step": 310
    },
    {
      "epoch": 0.026313562907183348,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009985497612067314,
      "loss": 1.5763,
      "step": 311
    },
    {
      "epoch": 0.02639817243421609,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009985396068951467,
      "loss": 1.4368,
      "step": 312
    },
    {
      "epoch": 0.026482781961248838,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009985294172101327,
      "loss": 1.4364,
      "step": 313
    },
    {
      "epoch": 0.02656739148828158,
      "grad_norm": 2.703125,
      "learning_rate": 0.000998519192152413,
      "loss": 1.2049,
      "step": 314
    },
    {
      "epoch": 0.026652001015314324,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009985089317227122,
      "loss": 1.4638,
      "step": 315
    },
    {
      "epoch": 0.026736610542347067,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009984986359217588,
      "loss": 1.3997,
      "step": 316
    },
    {
      "epoch": 0.02682122006937981,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009984883047502835,
      "loss": 1.2781,
      "step": 317
    },
    {
      "epoch": 0.026905829596412557,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009984779382090192,
      "loss": 1.2534,
      "step": 318
    },
    {
      "epoch": 0.0269904391234453,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009984675362987013,
      "loss": 1.3566,
      "step": 319
    },
    {
      "epoch": 0.027075048650478044,
      "grad_norm": 2.59375,
      "learning_rate": 0.000998457099020068,
      "loss": 1.7275,
      "step": 320
    },
    {
      "epoch": 0.027159658177510787,
      "grad_norm": 13.25,
      "learning_rate": 0.00099844662637386,
      "loss": 1.0102,
      "step": 321
    },
    {
      "epoch": 0.02724426770454353,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009984361183608197,
      "loss": 1.6163,
      "step": 322
    },
    {
      "epoch": 0.027328877231576277,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009984255749816938,
      "loss": 1.983,
      "step": 323
    },
    {
      "epoch": 0.02741348675860902,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009984149962372294,
      "loss": 2.1572,
      "step": 324
    },
    {
      "epoch": 0.027498096285641763,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009984043821281777,
      "loss": 1.4192,
      "step": 325
    },
    {
      "epoch": 0.027582705812674507,
      "grad_norm": 3.5625,
      "learning_rate": 0.0009983937326552915,
      "loss": 1.873,
      "step": 326
    },
    {
      "epoch": 0.02766731533970725,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009983830478193264,
      "loss": 1.554,
      "step": 327
    },
    {
      "epoch": 0.027751924866739996,
      "grad_norm": 2.453125,
      "learning_rate": 0.000998372327621041,
      "loss": 1.6128,
      "step": 328
    },
    {
      "epoch": 0.02783653439377274,
      "grad_norm": 1.625,
      "learning_rate": 0.0009983615720611954,
      "loss": 1.2974,
      "step": 329
    },
    {
      "epoch": 0.027921143920805483,
      "grad_norm": 7.65625,
      "learning_rate": 0.000998350781140553,
      "loss": 1.3522,
      "step": 330
    },
    {
      "epoch": 0.028005753447838226,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009983399548598795,
      "loss": 1.6402,
      "step": 331
    },
    {
      "epoch": 0.02809036297487097,
      "grad_norm": 55.5,
      "learning_rate": 0.0009983290932199429,
      "loss": 1.5473,
      "step": 332
    },
    {
      "epoch": 0.028174972501903716,
      "grad_norm": 13.0625,
      "learning_rate": 0.000998318196221514,
      "loss": 1.3685,
      "step": 333
    },
    {
      "epoch": 0.02825958202893646,
      "grad_norm": 3.21875,
      "learning_rate": 0.000998307263865366,
      "loss": 1.3612,
      "step": 334
    },
    {
      "epoch": 0.028344191555969202,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009982962961522744,
      "loss": 1.4535,
      "step": 335
    },
    {
      "epoch": 0.028428801083001946,
      "grad_norm": 74.0,
      "learning_rate": 0.0009982852930830177,
      "loss": 1.45,
      "step": 336
    },
    {
      "epoch": 0.02851341061003469,
      "grad_norm": 4.3125,
      "learning_rate": 0.0009982742546583764,
      "loss": 1.8693,
      "step": 337
    },
    {
      "epoch": 0.028598020137067432,
      "grad_norm": 5.65625,
      "learning_rate": 0.0009982631808791338,
      "loss": 1.6081,
      "step": 338
    },
    {
      "epoch": 0.02868262966410018,
      "grad_norm": 86.0,
      "learning_rate": 0.0009982520717460757,
      "loss": 1.2881,
      "step": 339
    },
    {
      "epoch": 0.028767239191132922,
      "grad_norm": 56.25,
      "learning_rate": 0.0009982409272599902,
      "loss": 1.5507,
      "step": 340
    },
    {
      "epoch": 0.028851848718165665,
      "grad_norm": 6.1875,
      "learning_rate": 0.000998229747421668,
      "loss": 1.7009,
      "step": 341
    },
    {
      "epoch": 0.02893645824519841,
      "grad_norm": 4.96875,
      "learning_rate": 0.0009982185322319026,
      "loss": 1.7586,
      "step": 342
    },
    {
      "epoch": 0.02902106777223115,
      "grad_norm": 35.0,
      "learning_rate": 0.0009982072816914894,
      "loss": 1.3968,
      "step": 343
    },
    {
      "epoch": 0.0291056772992639,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009981959958012272,
      "loss": 1.335,
      "step": 344
    },
    {
      "epoch": 0.02919028682629664,
      "grad_norm": 4.5,
      "learning_rate": 0.0009981846745619164,
      "loss": 1.4444,
      "step": 345
    },
    {
      "epoch": 0.029274896353329385,
      "grad_norm": 28.25,
      "learning_rate": 0.0009981733179743604,
      "loss": 1.1728,
      "step": 346
    },
    {
      "epoch": 0.029359505880362128,
      "grad_norm": 256.0,
      "learning_rate": 0.0009981619260393648,
      "loss": 1.2869,
      "step": 347
    },
    {
      "epoch": 0.02944411540739487,
      "grad_norm": 3056.0,
      "learning_rate": 0.0009981504987577382,
      "loss": 1.3753,
      "step": 348
    },
    {
      "epoch": 0.029528724934427618,
      "grad_norm": 764.0,
      "learning_rate": 0.0009981390361302911,
      "loss": 1.794,
      "step": 349
    },
    {
      "epoch": 0.02961333446146036,
      "grad_norm": 1184.0,
      "learning_rate": 0.0009981275381578372,
      "loss": 2.0408,
      "step": 350
    },
    {
      "epoch": 0.029697943988493104,
      "grad_norm": 612.0,
      "learning_rate": 0.000998116004841192,
      "loss": 2.8433,
      "step": 351
    },
    {
      "epoch": 0.029782553515525848,
      "grad_norm": 544.0,
      "learning_rate": 0.000998104436181174,
      "loss": 3.2391,
      "step": 352
    },
    {
      "epoch": 0.02986716304255859,
      "grad_norm": 1808.0,
      "learning_rate": 0.000998092832178604,
      "loss": 2.8884,
      "step": 353
    },
    {
      "epoch": 0.029951772569591337,
      "grad_norm": 3456.0,
      "learning_rate": 0.0009980811928343054,
      "loss": 3.3631,
      "step": 354
    },
    {
      "epoch": 0.03003638209662408,
      "grad_norm": 478.0,
      "learning_rate": 0.000998069518149104,
      "loss": 3.3154,
      "step": 355
    },
    {
      "epoch": 0.030120991623656824,
      "grad_norm": 880.0,
      "learning_rate": 0.0009980578081238284,
      "loss": 3.3374,
      "step": 356
    },
    {
      "epoch": 0.030205601150689567,
      "grad_norm": 2608.0,
      "learning_rate": 0.0009980460627593089,
      "loss": 2.6807,
      "step": 357
    },
    {
      "epoch": 0.03029021067772231,
      "grad_norm": 83456.0,
      "learning_rate": 0.0009980342820563794,
      "loss": 2.9398,
      "step": 358
    },
    {
      "epoch": 0.030374820204755057,
      "grad_norm": 9600.0,
      "learning_rate": 0.0009980224660158757,
      "loss": 3.1462,
      "step": 359
    },
    {
      "epoch": 0.0304594297317878,
      "grad_norm": 7072.0,
      "learning_rate": 0.000998010614638636,
      "loss": 2.514,
      "step": 360
    },
    {
      "epoch": 0.030544039258820543,
      "grad_norm": 419840.0,
      "learning_rate": 0.0009979987279255013,
      "loss": 2.7013,
      "step": 361
    },
    {
      "epoch": 0.030628648785853287,
      "grad_norm": 9216.0,
      "learning_rate": 0.0009979868058773152,
      "loss": 2.6713,
      "step": 362
    },
    {
      "epoch": 0.03071325831288603,
      "grad_norm": 15232.0,
      "learning_rate": 0.0009979748484949232,
      "loss": 2.321,
      "step": 363
    },
    {
      "epoch": 0.030797867839918776,
      "grad_norm": 2768.0,
      "learning_rate": 0.0009979628557791743,
      "loss": 2.851,
      "step": 364
    },
    {
      "epoch": 0.03088247736695152,
      "grad_norm": 1440.0,
      "learning_rate": 0.0009979508277309188,
      "loss": 2.9006,
      "step": 365
    },
    {
      "epoch": 0.030967086893984263,
      "grad_norm": 1192.0,
      "learning_rate": 0.0009979387643510107,
      "loss": 2.954,
      "step": 366
    },
    {
      "epoch": 0.031051696421017006,
      "grad_norm": 1020.0,
      "learning_rate": 0.0009979266656403056,
      "loss": 2.7029,
      "step": 367
    },
    {
      "epoch": 0.03113630594804975,
      "grad_norm": 368.0,
      "learning_rate": 0.000997914531599662,
      "loss": 3.0723,
      "step": 368
    },
    {
      "epoch": 0.031220915475082493,
      "grad_norm": 118.5,
      "learning_rate": 0.000997902362229941,
      "loss": 3.4108,
      "step": 369
    },
    {
      "epoch": 0.03130552500211524,
      "grad_norm": 171.0,
      "learning_rate": 0.000997890157532006,
      "loss": 3.0421,
      "step": 370
    },
    {
      "epoch": 0.03139013452914798,
      "grad_norm": 63.75,
      "learning_rate": 0.000997877917506723,
      "loss": 3.0251,
      "step": 371
    },
    {
      "epoch": 0.031474744056180726,
      "grad_norm": 20.0,
      "learning_rate": 0.0009978656421549604,
      "loss": 2.686,
      "step": 372
    },
    {
      "epoch": 0.03155935358321347,
      "grad_norm": 49.0,
      "learning_rate": 0.0009978533314775892,
      "loss": 2.9034,
      "step": 373
    },
    {
      "epoch": 0.03164396311024621,
      "grad_norm": 14.5,
      "learning_rate": 0.000997840985475483,
      "loss": 2.2301,
      "step": 374
    },
    {
      "epoch": 0.031728572637278955,
      "grad_norm": 15.4375,
      "learning_rate": 0.0009978286041495174,
      "loss": 2.1013,
      "step": 375
    },
    {
      "epoch": 0.0318131821643117,
      "grad_norm": 211.0,
      "learning_rate": 0.0009978161875005716,
      "loss": 1.7976,
      "step": 376
    },
    {
      "epoch": 0.03189779169134445,
      "grad_norm": 34.5,
      "learning_rate": 0.000997803735529526,
      "loss": 2.2469,
      "step": 377
    },
    {
      "epoch": 0.03198240121837719,
      "grad_norm": 51.75,
      "learning_rate": 0.0009977912482372646,
      "loss": 1.7826,
      "step": 378
    },
    {
      "epoch": 0.032067010745409935,
      "grad_norm": 31.375,
      "learning_rate": 0.000997778725624673,
      "loss": 1.8353,
      "step": 379
    },
    {
      "epoch": 0.03215162027244268,
      "grad_norm": 25.25,
      "learning_rate": 0.00099776616769264,
      "loss": 1.7702,
      "step": 380
    },
    {
      "epoch": 0.03223622979947542,
      "grad_norm": 9.0625,
      "learning_rate": 0.0009977535744420563,
      "loss": 2.3,
      "step": 381
    },
    {
      "epoch": 0.032320839326508165,
      "grad_norm": 7.09375,
      "learning_rate": 0.000997740945873816,
      "loss": 1.9972,
      "step": 382
    },
    {
      "epoch": 0.03240544885354091,
      "grad_norm": 5.0625,
      "learning_rate": 0.0009977282819888145,
      "loss": 1.7951,
      "step": 383
    },
    {
      "epoch": 0.03249005838057365,
      "grad_norm": 5.15625,
      "learning_rate": 0.0009977155827879509,
      "loss": 1.5886,
      "step": 384
    },
    {
      "epoch": 0.032574667907606394,
      "grad_norm": 9.375,
      "learning_rate": 0.000997702848272126,
      "loss": 1.5342,
      "step": 385
    },
    {
      "epoch": 0.03265927743463914,
      "grad_norm": 7.34375,
      "learning_rate": 0.0009976900784422434,
      "loss": 1.9163,
      "step": 386
    },
    {
      "epoch": 0.03274388696167188,
      "grad_norm": 6.84375,
      "learning_rate": 0.0009976772732992092,
      "loss": 1.9442,
      "step": 387
    },
    {
      "epoch": 0.03282849648870463,
      "grad_norm": 5.78125,
      "learning_rate": 0.000997664432843932,
      "loss": 2.1904,
      "step": 388
    },
    {
      "epoch": 0.032913106015737374,
      "grad_norm": 9.875,
      "learning_rate": 0.0009976515570773229,
      "loss": 1.5792,
      "step": 389
    },
    {
      "epoch": 0.03299771554277012,
      "grad_norm": 6.84375,
      "learning_rate": 0.0009976386460002953,
      "loss": 2.18,
      "step": 390
    },
    {
      "epoch": 0.03308232506980286,
      "grad_norm": 6.78125,
      "learning_rate": 0.0009976256996137656,
      "loss": 1.4938,
      "step": 391
    },
    {
      "epoch": 0.033166934596835604,
      "grad_norm": 4.34375,
      "learning_rate": 0.000997612717918652,
      "loss": 1.3875,
      "step": 392
    },
    {
      "epoch": 0.03325154412386835,
      "grad_norm": 4.28125,
      "learning_rate": 0.000997599700915876,
      "loss": 1.724,
      "step": 393
    },
    {
      "epoch": 0.03333615365090109,
      "grad_norm": 3.296875,
      "learning_rate": 0.000997586648606361,
      "loss": 1.7405,
      "step": 394
    },
    {
      "epoch": 0.033420763177933833,
      "grad_norm": 3.25,
      "learning_rate": 0.0009975735609910332,
      "loss": 1.8569,
      "step": 395
    },
    {
      "epoch": 0.03350537270496658,
      "grad_norm": 5.03125,
      "learning_rate": 0.0009975604380708211,
      "loss": 2.2328,
      "step": 396
    },
    {
      "epoch": 0.03358998223199932,
      "grad_norm": 35.25,
      "learning_rate": 0.0009975472798466562,
      "loss": 1.3729,
      "step": 397
    },
    {
      "epoch": 0.03367459175903207,
      "grad_norm": 55.25,
      "learning_rate": 0.0009975340863194713,
      "loss": 1.679,
      "step": 398
    },
    {
      "epoch": 0.03375920128606481,
      "grad_norm": 21.375,
      "learning_rate": 0.0009975208574902034,
      "loss": 1.6629,
      "step": 399
    },
    {
      "epoch": 0.033843810813097557,
      "grad_norm": 5.09375,
      "learning_rate": 0.000997507593359791,
      "loss": 1.5835,
      "step": 400
    },
    {
      "epoch": 0.0339284203401303,
      "grad_norm": 4.53125,
      "learning_rate": 0.0009974942939291746,
      "loss": 2.1673,
      "step": 401
    },
    {
      "epoch": 0.03401302986716304,
      "grad_norm": 11.875,
      "learning_rate": 0.0009974809591992986,
      "loss": 1.9143,
      "step": 402
    },
    {
      "epoch": 0.034097639394195786,
      "grad_norm": 6.375,
      "learning_rate": 0.0009974675891711087,
      "loss": 1.2408,
      "step": 403
    },
    {
      "epoch": 0.03418224892122853,
      "grad_norm": 6.4375,
      "learning_rate": 0.0009974541838455538,
      "loss": 1.779,
      "step": 404
    },
    {
      "epoch": 0.03426685844826127,
      "grad_norm": 9.6875,
      "learning_rate": 0.000997440743223585,
      "loss": 1.5609,
      "step": 405
    },
    {
      "epoch": 0.034351467975294016,
      "grad_norm": 10.8125,
      "learning_rate": 0.0009974272673061557,
      "loss": 1.7199,
      "step": 406
    },
    {
      "epoch": 0.03443607750232676,
      "grad_norm": 11.1875,
      "learning_rate": 0.0009974137560942228,
      "loss": 1.6702,
      "step": 407
    },
    {
      "epoch": 0.03452068702935951,
      "grad_norm": 6.5625,
      "learning_rate": 0.000997400209588744,
      "loss": 1.49,
      "step": 408
    },
    {
      "epoch": 0.03460529655639225,
      "grad_norm": 21.875,
      "learning_rate": 0.000997386627790681,
      "loss": 2.2663,
      "step": 409
    },
    {
      "epoch": 0.034689906083424996,
      "grad_norm": 7.125,
      "learning_rate": 0.0009973730107009978,
      "loss": 1.6606,
      "step": 410
    },
    {
      "epoch": 0.03477451561045774,
      "grad_norm": 15.125,
      "learning_rate": 0.0009973593583206602,
      "loss": 1.6024,
      "step": 411
    },
    {
      "epoch": 0.03485912513749048,
      "grad_norm": 9.5625,
      "learning_rate": 0.0009973456706506368,
      "loss": 1.6296,
      "step": 412
    },
    {
      "epoch": 0.034943734664523225,
      "grad_norm": 7.8125,
      "learning_rate": 0.0009973319476918989,
      "loss": 1.5316,
      "step": 413
    },
    {
      "epoch": 0.03502834419155597,
      "grad_norm": 6.0,
      "learning_rate": 0.0009973181894454201,
      "loss": 1.6886,
      "step": 414
    },
    {
      "epoch": 0.03511295371858871,
      "grad_norm": 7.34375,
      "learning_rate": 0.000997304395912177,
      "loss": 1.5605,
      "step": 415
    },
    {
      "epoch": 0.035197563245621455,
      "grad_norm": 5.96875,
      "learning_rate": 0.0009972905670931481,
      "loss": 1.6617,
      "step": 416
    },
    {
      "epoch": 0.0352821727726542,
      "grad_norm": 115.0,
      "learning_rate": 0.0009972767029893144,
      "loss": 1.4393,
      "step": 417
    },
    {
      "epoch": 0.03536678229968694,
      "grad_norm": 7.3125,
      "learning_rate": 0.0009972628036016597,
      "loss": 1.658,
      "step": 418
    },
    {
      "epoch": 0.03545139182671969,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009972488689311703,
      "loss": 1.4791,
      "step": 419
    },
    {
      "epoch": 0.035536001353752435,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009972348989788352,
      "loss": 1.9138,
      "step": 420
    },
    {
      "epoch": 0.03562061088078518,
      "grad_norm": 3.9375,
      "learning_rate": 0.000997220893745645,
      "loss": 1.6945,
      "step": 421
    },
    {
      "epoch": 0.03570522040781792,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009972068532325938,
      "loss": 1.531,
      "step": 422
    },
    {
      "epoch": 0.035789829934850664,
      "grad_norm": 6.375,
      "learning_rate": 0.0009971927774406779,
      "loss": 1.2508,
      "step": 423
    },
    {
      "epoch": 0.03587443946188341,
      "grad_norm": 4.15625,
      "learning_rate": 0.000997178666370896,
      "loss": 1.8392,
      "step": 424
    },
    {
      "epoch": 0.03595904898891615,
      "grad_norm": 6.21875,
      "learning_rate": 0.0009971645200242492,
      "loss": 2.0438,
      "step": 425
    },
    {
      "epoch": 0.036043658515948894,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009971503384017413,
      "loss": 1.1425,
      "step": 426
    },
    {
      "epoch": 0.03612826804298164,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009971361215043784,
      "loss": 1.6439,
      "step": 427
    },
    {
      "epoch": 0.03621287757001438,
      "grad_norm": 3.484375,
      "learning_rate": 0.0009971218693331696,
      "loss": 1.4733,
      "step": 428
    },
    {
      "epoch": 0.03629748709704713,
      "grad_norm": 2.75,
      "learning_rate": 0.000997107581889126,
      "loss": 1.4184,
      "step": 429
    },
    {
      "epoch": 0.036382096624079874,
      "grad_norm": 3.78125,
      "learning_rate": 0.0009970932591732613,
      "loss": 1.8347,
      "step": 430
    },
    {
      "epoch": 0.03646670615111262,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009970789011865915,
      "loss": 1.5557,
      "step": 431
    },
    {
      "epoch": 0.03655131567814536,
      "grad_norm": 2.875,
      "learning_rate": 0.0009970645079301359,
      "loss": 1.624,
      "step": 432
    },
    {
      "epoch": 0.0366359252051781,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009970500794049153,
      "loss": 1.3857,
      "step": 433
    },
    {
      "epoch": 0.03672053473221085,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009970356156119538,
      "loss": 1.4808,
      "step": 434
    },
    {
      "epoch": 0.03680514425924359,
      "grad_norm": 5.34375,
      "learning_rate": 0.0009970211165522774,
      "loss": 1.2682,
      "step": 435
    },
    {
      "epoch": 0.03688975378627633,
      "grad_norm": 7.84375,
      "learning_rate": 0.000997006582226915,
      "loss": 1.4656,
      "step": 436
    },
    {
      "epoch": 0.036974363313309076,
      "grad_norm": 4.21875,
      "learning_rate": 0.000996992012636898,
      "loss": 1.4898,
      "step": 437
    },
    {
      "epoch": 0.03705897284034182,
      "grad_norm": 6.46875,
      "learning_rate": 0.0009969774077832597,
      "loss": 1.5521,
      "step": 438
    },
    {
      "epoch": 0.03714358236737457,
      "grad_norm": 2.921875,
      "learning_rate": 0.000996962767667037,
      "loss": 1.1943,
      "step": 439
    },
    {
      "epoch": 0.03722819189440731,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009969480922892681,
      "loss": 1.3804,
      "step": 440
    },
    {
      "epoch": 0.037312801421440056,
      "grad_norm": 3.8125,
      "learning_rate": 0.0009969333816509946,
      "loss": 1.6806,
      "step": 441
    },
    {
      "epoch": 0.0373974109484728,
      "grad_norm": 10.1875,
      "learning_rate": 0.0009969186357532604,
      "loss": 2.4735,
      "step": 442
    },
    {
      "epoch": 0.03748202047550554,
      "grad_norm": 4.5625,
      "learning_rate": 0.0009969038545971115,
      "loss": 1.4318,
      "step": 443
    },
    {
      "epoch": 0.037566630002538286,
      "grad_norm": 6.40625,
      "learning_rate": 0.0009968890381835968,
      "loss": 1.7157,
      "step": 444
    },
    {
      "epoch": 0.03765123952957103,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009968741865137676,
      "loss": 2.0397,
      "step": 445
    },
    {
      "epoch": 0.03773584905660377,
      "grad_norm": 3.984375,
      "learning_rate": 0.0009968592995886778,
      "loss": 1.4634,
      "step": 446
    },
    {
      "epoch": 0.037820458583636515,
      "grad_norm": 4.28125,
      "learning_rate": 0.0009968443774093835,
      "loss": 1.4054,
      "step": 447
    },
    {
      "epoch": 0.03790506811066926,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009968294199769433,
      "loss": 1.3067,
      "step": 448
    },
    {
      "epoch": 0.037989677637702,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009968144272924188,
      "loss": 1.1902,
      "step": 449
    },
    {
      "epoch": 0.03807428716473475,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009967993993568738,
      "loss": 1.284,
      "step": 450
    },
    {
      "epoch": 0.038158896691767495,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009967843361713747,
      "loss": 1.7728,
      "step": 451
    },
    {
      "epoch": 0.03824350621880024,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009967692377369898,
      "loss": 2.0907,
      "step": 452
    },
    {
      "epoch": 0.03832811574583298,
      "grad_norm": 2.1875,
      "learning_rate": 0.000996754104054791,
      "loss": 1.5884,
      "step": 453
    },
    {
      "epoch": 0.038412725272865725,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009967389351258516,
      "loss": 1.4421,
      "step": 454
    },
    {
      "epoch": 0.03849733479989847,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009967237309512482,
      "loss": 1.4248,
      "step": 455
    },
    {
      "epoch": 0.03858194432693121,
      "grad_norm": 7.09375,
      "learning_rate": 0.0009967084915320594,
      "loss": 1.588,
      "step": 456
    },
    {
      "epoch": 0.038666553853963954,
      "grad_norm": 1112.0,
      "learning_rate": 0.0009966932168693668,
      "loss": 1.2521,
      "step": 457
    },
    {
      "epoch": 0.0387511633809967,
      "grad_norm": 266.0,
      "learning_rate": 0.0009966779069642538,
      "loss": 1.5216,
      "step": 458
    },
    {
      "epoch": 0.03883577290802944,
      "grad_norm": 5.5625,
      "learning_rate": 0.0009966625618178068,
      "loss": 2.8209,
      "step": 459
    },
    {
      "epoch": 0.03892038243506219,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009966471814311152,
      "loss": 1.7797,
      "step": 460
    },
    {
      "epoch": 0.039004991962094934,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009966317658052695,
      "loss": 2.0008,
      "step": 461
    },
    {
      "epoch": 0.03908960148912768,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009966163149413637,
      "loss": 1.947,
      "step": 462
    },
    {
      "epoch": 0.03917421101616042,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009966008288404943,
      "loss": 1.8034,
      "step": 463
    },
    {
      "epoch": 0.039258820543193164,
      "grad_norm": 2.71875,
      "learning_rate": 0.00099658530750376,
      "loss": 1.4772,
      "step": 464
    },
    {
      "epoch": 0.03934343007022591,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009965697509322622,
      "loss": 1.603,
      "step": 465
    },
    {
      "epoch": 0.03942803959725865,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009965541591271045,
      "loss": 1.4341,
      "step": 466
    },
    {
      "epoch": 0.039512649124291394,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009965385320893933,
      "loss": 1.594,
      "step": 467
    },
    {
      "epoch": 0.03959725865132414,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009965228698202374,
      "loss": 1.7873,
      "step": 468
    },
    {
      "epoch": 0.03968186817835688,
      "grad_norm": 2.0,
      "learning_rate": 0.0009965071723207484,
      "loss": 1.5389,
      "step": 469
    },
    {
      "epoch": 0.03976647770538963,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0009964914395920396,
      "loss": 1.3452,
      "step": 470
    },
    {
      "epoch": 0.03985108723242237,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009964756716352276,
      "loss": 1.5352,
      "step": 471
    },
    {
      "epoch": 0.03993569675945512,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009964598684514313,
      "loss": 1.0926,
      "step": 472
    },
    {
      "epoch": 0.04002030628648786,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009964440300417716,
      "loss": 1.4208,
      "step": 473
    },
    {
      "epoch": 0.0401049158135206,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009964281564073725,
      "loss": 1.2632,
      "step": 474
    },
    {
      "epoch": 0.040189525340553346,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009964122475493605,
      "loss": 1.5355,
      "step": 475
    },
    {
      "epoch": 0.04027413486758609,
      "grad_norm": 2.546875,
      "learning_rate": 0.000996396303468864,
      "loss": 1.8134,
      "step": 476
    },
    {
      "epoch": 0.04035874439461883,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009963803241670147,
      "loss": 1.6089,
      "step": 477
    },
    {
      "epoch": 0.040443353921651576,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009963643096449462,
      "loss": 1.5777,
      "step": 478
    },
    {
      "epoch": 0.04052796344868432,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009963482599037946,
      "loss": 1.6556,
      "step": 479
    },
    {
      "epoch": 0.04061257297571707,
      "grad_norm": 18.625,
      "learning_rate": 0.0009963321749446993,
      "loss": 1.472,
      "step": 480
    },
    {
      "epoch": 0.04069718250274981,
      "grad_norm": 2.203125,
      "learning_rate": 0.000996316054768801,
      "loss": 1.401,
      "step": 481
    },
    {
      "epoch": 0.040781792029782556,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009962998993772439,
      "loss": 1.4282,
      "step": 482
    },
    {
      "epoch": 0.0408664015568153,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009962837087711738,
      "loss": 1.6441,
      "step": 483
    },
    {
      "epoch": 0.04095101108384804,
      "grad_norm": 3.6875,
      "learning_rate": 0.00099626748295174,
      "loss": 2.364,
      "step": 484
    },
    {
      "epoch": 0.041035620610880785,
      "grad_norm": 4.21875,
      "learning_rate": 0.0009962512219200935,
      "loss": 2.0495,
      "step": 485
    },
    {
      "epoch": 0.04112023013791353,
      "grad_norm": 3.34375,
      "learning_rate": 0.000996234925677388,
      "loss": 1.4075,
      "step": 486
    },
    {
      "epoch": 0.04120483966494627,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009962185942247802,
      "loss": 1.7289,
      "step": 487
    },
    {
      "epoch": 0.041289449191979015,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009962022275634285,
      "loss": 1.3854,
      "step": 488
    },
    {
      "epoch": 0.04137405871901176,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009961858256944942,
      "loss": 1.6097,
      "step": 489
    },
    {
      "epoch": 0.0414586682460445,
      "grad_norm": 6.03125,
      "learning_rate": 0.0009961693886191413,
      "loss": 1.5453,
      "step": 490
    },
    {
      "epoch": 0.04154327777307725,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009961529163385358,
      "loss": 1.6348,
      "step": 491
    },
    {
      "epoch": 0.041627887300109995,
      "grad_norm": 2.015625,
      "learning_rate": 0.000996136408853847,
      "loss": 1.3966,
      "step": 492
    },
    {
      "epoch": 0.04171249682714274,
      "grad_norm": 2.75,
      "learning_rate": 0.0009961198661662455,
      "loss": 1.8525,
      "step": 493
    },
    {
      "epoch": 0.04179710635417548,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009961032882769055,
      "loss": 1.5937,
      "step": 494
    },
    {
      "epoch": 0.041881715881208224,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009960866751870033,
      "loss": 1.9505,
      "step": 495
    },
    {
      "epoch": 0.04196632540824097,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009960700268977173,
      "loss": 1.7248,
      "step": 496
    },
    {
      "epoch": 0.04205093493527371,
      "grad_norm": 2.875,
      "learning_rate": 0.000996053343410229,
      "loss": 1.8714,
      "step": 497
    },
    {
      "epoch": 0.042135544462306454,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009960366247257222,
      "loss": 1.2586,
      "step": 498
    },
    {
      "epoch": 0.0422201539893392,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009960198708453831,
      "loss": 1.796,
      "step": 499
    },
    {
      "epoch": 0.04230476351637194,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009960030817704004,
      "loss": 1.4065,
      "step": 500
    },
    {
      "epoch": 0.04238937304340469,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009959862575019656,
      "loss": 1.2108,
      "step": 501
    },
    {
      "epoch": 0.042473982570437434,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009959693980412722,
      "loss": 1.6824,
      "step": 502
    },
    {
      "epoch": 0.04255859209747018,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009959525033895165,
      "loss": 1.4973,
      "step": 503
    },
    {
      "epoch": 0.04264320162450292,
      "grad_norm": 9.9375,
      "learning_rate": 0.0009959355735478972,
      "loss": 1.399,
      "step": 504
    },
    {
      "epoch": 0.04272781115153566,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009959186085176158,
      "loss": 2.0055,
      "step": 505
    },
    {
      "epoch": 0.04281242067856841,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009959016082998755,
      "loss": 1.3033,
      "step": 506
    },
    {
      "epoch": 0.04289703020560115,
      "grad_norm": 2.625,
      "learning_rate": 0.0009958845728958831,
      "loss": 2.1366,
      "step": 507
    },
    {
      "epoch": 0.04298163973263389,
      "grad_norm": 2.15625,
      "learning_rate": 0.000995867502306847,
      "loss": 1.4802,
      "step": 508
    },
    {
      "epoch": 0.043066249259666636,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009958503965339788,
      "loss": 2.423,
      "step": 509
    },
    {
      "epoch": 0.04315085878669938,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009958332555784918,
      "loss": 1.7215,
      "step": 510
    },
    {
      "epoch": 0.04323546831373213,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009958160794416022,
      "loss": 1.6019,
      "step": 511
    },
    {
      "epoch": 0.04332007784076487,
      "grad_norm": 2.109375,
      "learning_rate": 0.000995798868124529,
      "loss": 1.3312,
      "step": 512
    },
    {
      "epoch": 0.043404687367797616,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009957816216284935,
      "loss": 1.6573,
      "step": 513
    },
    {
      "epoch": 0.04348929689483036,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009957643399547192,
      "loss": 1.3591,
      "step": 514
    },
    {
      "epoch": 0.0435739064218631,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009957470231044322,
      "loss": 1.4667,
      "step": 515
    },
    {
      "epoch": 0.043658515948895846,
      "grad_norm": 2.125,
      "learning_rate": 0.0009957296710788612,
      "loss": 1.4413,
      "step": 516
    },
    {
      "epoch": 0.04374312547592859,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009957122838792378,
      "loss": 1.1131,
      "step": 517
    },
    {
      "epoch": 0.04382773500296133,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009956948615067952,
      "loss": 1.2172,
      "step": 518
    },
    {
      "epoch": 0.043912344529994075,
      "grad_norm": 2.75,
      "learning_rate": 0.00099567740396277,
      "loss": 1.8067,
      "step": 519
    },
    {
      "epoch": 0.04399695405702682,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009956599112484006,
      "loss": 1.5125,
      "step": 520
    },
    {
      "epoch": 0.04408156358405956,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009956423833649283,
      "loss": 1.5085,
      "step": 521
    },
    {
      "epoch": 0.04416617311109231,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009956248203135966,
      "loss": 1.5852,
      "step": 522
    },
    {
      "epoch": 0.044250782638125055,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009956072220956521,
      "loss": 1.1601,
      "step": 523
    },
    {
      "epoch": 0.0443353921651578,
      "grad_norm": 2.953125,
      "learning_rate": 0.000995589588712343,
      "loss": 1.6977,
      "step": 524
    },
    {
      "epoch": 0.04442000169219054,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009955719201649206,
      "loss": 1.9036,
      "step": 525
    },
    {
      "epoch": 0.044504611219223285,
      "grad_norm": 2.5,
      "learning_rate": 0.0009955542164546388,
      "loss": 1.4687,
      "step": 526
    },
    {
      "epoch": 0.04458922074625603,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009955364775827535,
      "loss": 1.9439,
      "step": 527
    },
    {
      "epoch": 0.04467383027328877,
      "grad_norm": 466.0,
      "learning_rate": 0.0009955187035505233,
      "loss": 1.6088,
      "step": 528
    },
    {
      "epoch": 0.044758439800321514,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009955008943592094,
      "loss": 1.9631,
      "step": 529
    },
    {
      "epoch": 0.04484304932735426,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009954830500100756,
      "loss": 1.3634,
      "step": 530
    },
    {
      "epoch": 0.044927658854387,
      "grad_norm": 3.140625,
      "learning_rate": 0.0009954651705043878,
      "loss": 1.8,
      "step": 531
    },
    {
      "epoch": 0.04501226838141975,
      "grad_norm": 6.59375,
      "learning_rate": 0.0009954472558434148,
      "loss": 1.5134,
      "step": 532
    },
    {
      "epoch": 0.045096877908452494,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009954293060284276,
      "loss": 1.6015,
      "step": 533
    },
    {
      "epoch": 0.04518148743548524,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009954113210606998,
      "loss": 1.4167,
      "step": 534
    },
    {
      "epoch": 0.04526609696251798,
      "grad_norm": 4.65625,
      "learning_rate": 0.0009953933009415076,
      "loss": 1.9478,
      "step": 535
    },
    {
      "epoch": 0.045350706489550724,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009953752456721297,
      "loss": 1.6296,
      "step": 536
    },
    {
      "epoch": 0.04543531601658347,
      "grad_norm": 2.640625,
      "learning_rate": 0.000995357155253847,
      "loss": 1.7564,
      "step": 537
    },
    {
      "epoch": 0.04551992554361621,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009953390296879431,
      "loss": 1.3776,
      "step": 538
    },
    {
      "epoch": 0.045604535070648954,
      "grad_norm": 3.984375,
      "learning_rate": 0.0009953208689757044,
      "loss": 1.5307,
      "step": 539
    },
    {
      "epoch": 0.0456891445976817,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009953026731184189,
      "loss": 1.4973,
      "step": 540
    },
    {
      "epoch": 0.04577375412471444,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009952844421173782,
      "loss": 1.9153,
      "step": 541
    },
    {
      "epoch": 0.04585836365174719,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009952661759738753,
      "loss": 1.9208,
      "step": 542
    },
    {
      "epoch": 0.04594297317877993,
      "grad_norm": 4.5,
      "learning_rate": 0.0009952478746892068,
      "loss": 1.2596,
      "step": 543
    },
    {
      "epoch": 0.04602758270581268,
      "grad_norm": 46.75,
      "learning_rate": 0.0009952295382646712,
      "loss": 1.3377,
      "step": 544
    },
    {
      "epoch": 0.04611219223284542,
      "grad_norm": 10.125,
      "learning_rate": 0.0009952111667015694,
      "loss": 2.066,
      "step": 545
    },
    {
      "epoch": 0.04619680175987816,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009951927600012048,
      "loss": 1.248,
      "step": 546
    },
    {
      "epoch": 0.046281411286910906,
      "grad_norm": 4.78125,
      "learning_rate": 0.0009951743181648835,
      "loss": 1.3788,
      "step": 547
    },
    {
      "epoch": 0.04636602081394365,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009951558411939142,
      "loss": 1.508,
      "step": 548
    },
    {
      "epoch": 0.04645063034097639,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009951373290896077,
      "loss": 1.1485,
      "step": 549
    },
    {
      "epoch": 0.046535239868009136,
      "grad_norm": 5.25,
      "learning_rate": 0.0009951187818532778,
      "loss": 1.14,
      "step": 550
    },
    {
      "epoch": 0.04661984939504188,
      "grad_norm": 2.796875,
      "learning_rate": 0.00099510019948624,
      "loss": 1.6612,
      "step": 551
    },
    {
      "epoch": 0.04670445892207462,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009950815819898136,
      "loss": 1.694,
      "step": 552
    },
    {
      "epoch": 0.04678906844910737,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009950629293653188,
      "loss": 1.5725,
      "step": 553
    },
    {
      "epoch": 0.046873677976140116,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009950442416140795,
      "loss": 1.4235,
      "step": 554
    },
    {
      "epoch": 0.04695828750317286,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009950255187374216,
      "loss": 1.1938,
      "step": 555
    },
    {
      "epoch": 0.0470428970302056,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009950067607366732,
      "loss": 1.648,
      "step": 556
    },
    {
      "epoch": 0.047127506557238345,
      "grad_norm": 2.09375,
      "learning_rate": 0.000994987967613166,
      "loss": 1.5901,
      "step": 557
    },
    {
      "epoch": 0.04721211608427109,
      "grad_norm": 3.125,
      "learning_rate": 0.0009949691393682327,
      "loss": 2.1817,
      "step": 558
    },
    {
      "epoch": 0.04729672561130383,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009949502760032097,
      "loss": 1.7478,
      "step": 559
    },
    {
      "epoch": 0.047381335138336575,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009949313775194353,
      "loss": 1.5831,
      "step": 560
    },
    {
      "epoch": 0.04746594466536932,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009949124439182506,
      "loss": 1.1959,
      "step": 561
    },
    {
      "epoch": 0.04755055419240206,
      "grad_norm": 5.40625,
      "learning_rate": 0.0009948934752009985,
      "loss": 1.5904,
      "step": 562
    },
    {
      "epoch": 0.04763516371943481,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009948744713690256,
      "loss": 1.669,
      "step": 563
    },
    {
      "epoch": 0.047719773246467555,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009948554324236798,
      "loss": 1.5525,
      "step": 564
    },
    {
      "epoch": 0.0478043827735003,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009948363583663121,
      "loss": 1.3652,
      "step": 565
    },
    {
      "epoch": 0.04788899230053304,
      "grad_norm": 23.75,
      "learning_rate": 0.000994817249198276,
      "loss": 1.809,
      "step": 566
    },
    {
      "epoch": 0.047973601827565784,
      "grad_norm": 6.9375,
      "learning_rate": 0.000994798104920927,
      "loss": 2.4332,
      "step": 567
    },
    {
      "epoch": 0.04805821135459853,
      "grad_norm": 14.375,
      "learning_rate": 0.0009947789255356243,
      "loss": 1.3239,
      "step": 568
    },
    {
      "epoch": 0.04814282088163127,
      "grad_norm": 5.0,
      "learning_rate": 0.0009947597110437278,
      "loss": 1.5916,
      "step": 569
    },
    {
      "epoch": 0.048227430408664014,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009947404614466017,
      "loss": 1.8177,
      "step": 570
    },
    {
      "epoch": 0.04831203993569676,
      "grad_norm": 10.9375,
      "learning_rate": 0.000994721176745611,
      "loss": 1.3777,
      "step": 571
    },
    {
      "epoch": 0.0483966494627295,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009947018569421246,
      "loss": 2.0263,
      "step": 572
    },
    {
      "epoch": 0.04848125898976225,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009946825020375131,
      "loss": 1.9254,
      "step": 573
    },
    {
      "epoch": 0.048565868516794994,
      "grad_norm": 3.34375,
      "learning_rate": 0.00099466311203315,
      "loss": 1.3581,
      "step": 574
    },
    {
      "epoch": 0.04865047804382774,
      "grad_norm": 3.75,
      "learning_rate": 0.0009946436869304107,
      "loss": 2.6882,
      "step": 575
    },
    {
      "epoch": 0.04873508757086048,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009946242267306739,
      "loss": 1.1264,
      "step": 576
    },
    {
      "epoch": 0.048819697097893223,
      "grad_norm": 6.34375,
      "learning_rate": 0.0009946047314353204,
      "loss": 1.405,
      "step": 577
    },
    {
      "epoch": 0.04890430662492597,
      "grad_norm": 2.484375,
      "learning_rate": 0.000994585201045733,
      "loss": 1.4567,
      "step": 578
    },
    {
      "epoch": 0.04898891615195871,
      "grad_norm": 2.65625,
      "learning_rate": 0.000994565635563298,
      "loss": 1.2529,
      "step": 579
    },
    {
      "epoch": 0.04907352567899145,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009945460349894033,
      "loss": 1.725,
      "step": 580
    },
    {
      "epoch": 0.049158135206024196,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009945263993254398,
      "loss": 1.7213,
      "step": 581
    },
    {
      "epoch": 0.04924274473305694,
      "grad_norm": 3.8125,
      "learning_rate": 0.0009945067285728007,
      "loss": 2.0613,
      "step": 582
    },
    {
      "epoch": 0.04932735426008968,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009944870227328816,
      "loss": 1.7861,
      "step": 583
    },
    {
      "epoch": 0.04941196378712243,
      "grad_norm": 3.09375,
      "learning_rate": 0.000994467281807081,
      "loss": 1.6237,
      "step": 584
    },
    {
      "epoch": 0.049496573314155176,
      "grad_norm": 4.125,
      "learning_rate": 0.0009944475057967995,
      "loss": 1.434,
      "step": 585
    },
    {
      "epoch": 0.04958118284118792,
      "grad_norm": 2.5,
      "learning_rate": 0.00099442769470344,
      "loss": 1.276,
      "step": 586
    },
    {
      "epoch": 0.04966579236822066,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009944078485284084,
      "loss": 1.6379,
      "step": 587
    },
    {
      "epoch": 0.049750401895253406,
      "grad_norm": 2.90625,
      "learning_rate": 0.000994387967273113,
      "loss": 2.2163,
      "step": 588
    },
    {
      "epoch": 0.04983501142228615,
      "grad_norm": 2.53125,
      "learning_rate": 0.000994368050938964,
      "loss": 1.1714,
      "step": 589
    },
    {
      "epoch": 0.04991962094931889,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009943480995273752,
      "loss": 1.6011,
      "step": 590
    },
    {
      "epoch": 0.050004230476351635,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009943281130397618,
      "loss": 1.8517,
      "step": 591
    },
    {
      "epoch": 0.05008884000338438,
      "grad_norm": 2.328125,
      "learning_rate": 0.000994308091477542,
      "loss": 1.7369,
      "step": 592
    },
    {
      "epoch": 0.05017344953041712,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009942880348421364,
      "loss": 1.5299,
      "step": 593
    },
    {
      "epoch": 0.05025805905744987,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009942679431349682,
      "loss": 1.8462,
      "step": 594
    },
    {
      "epoch": 0.050342668584482615,
      "grad_norm": 1.9921875,
      "learning_rate": 0.000994247816357463,
      "loss": 1.1763,
      "step": 595
    },
    {
      "epoch": 0.05042727811151536,
      "grad_norm": 2.125,
      "learning_rate": 0.0009942276545110485,
      "loss": 1.6205,
      "step": 596
    },
    {
      "epoch": 0.0505118876385481,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009942074575971557,
      "loss": 1.674,
      "step": 597
    },
    {
      "epoch": 0.050596497165580845,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009941872256172175,
      "loss": 1.2339,
      "step": 598
    },
    {
      "epoch": 0.05068110669261359,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009941669585726697,
      "loss": 1.473,
      "step": 599
    },
    {
      "epoch": 0.05076571621964633,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009941466564649497,
      "loss": 1.6188,
      "step": 600
    },
    {
      "epoch": 0.050850325746679075,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009941263192954986,
      "loss": 1.3364,
      "step": 601
    },
    {
      "epoch": 0.05093493527371182,
      "grad_norm": 2.625,
      "learning_rate": 0.0009941059470657593,
      "loss": 1.3214,
      "step": 602
    },
    {
      "epoch": 0.05101954480074456,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009940855397771772,
      "loss": 1.5279,
      "step": 603
    },
    {
      "epoch": 0.05110415432777731,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009940650974312,
      "loss": 1.4447,
      "step": 604
    },
    {
      "epoch": 0.051188763854810054,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009940446200292787,
      "loss": 1.9854,
      "step": 605
    },
    {
      "epoch": 0.0512733733818428,
      "grad_norm": 2.890625,
      "learning_rate": 0.000994024107572866,
      "loss": 2.004,
      "step": 606
    },
    {
      "epoch": 0.05135798290887554,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009940035600634171,
      "loss": 1.4318,
      "step": 607
    },
    {
      "epoch": 0.051442592435908284,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009939829775023905,
      "loss": 2.39,
      "step": 608
    },
    {
      "epoch": 0.05152720196294103,
      "grad_norm": 2.640625,
      "learning_rate": 0.000993962359891246,
      "loss": 1.607,
      "step": 609
    },
    {
      "epoch": 0.05161181148997377,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009939417072314472,
      "loss": 1.7664,
      "step": 610
    },
    {
      "epoch": 0.051696421017006514,
      "grad_norm": 2.03125,
      "learning_rate": 0.000993921019524459,
      "loss": 1.537,
      "step": 611
    },
    {
      "epoch": 0.05178103054403926,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009939002967717491,
      "loss": 1.7743,
      "step": 612
    },
    {
      "epoch": 0.051865640071072,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009938795389747884,
      "loss": 1.266,
      "step": 613
    },
    {
      "epoch": 0.05195024959810474,
      "grad_norm": 6.75,
      "learning_rate": 0.0009938587461350492,
      "loss": 2.0028,
      "step": 614
    },
    {
      "epoch": 0.05203485912513749,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009938379182540074,
      "loss": 1.4361,
      "step": 615
    },
    {
      "epoch": 0.05211946865217024,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009938170553331406,
      "loss": 1.9379,
      "step": 616
    },
    {
      "epoch": 0.05220407817920298,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009937961573739287,
      "loss": 2.0253,
      "step": 617
    },
    {
      "epoch": 0.05228868770623572,
      "grad_norm": 2.515625,
      "learning_rate": 0.000993775224377855,
      "loss": 1.8613,
      "step": 618
    },
    {
      "epoch": 0.052373297233268466,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009937542563464047,
      "loss": 1.1693,
      "step": 619
    },
    {
      "epoch": 0.05245790676030121,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009937332532810655,
      "loss": 1.6421,
      "step": 620
    },
    {
      "epoch": 0.05254251628733395,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009937122151833279,
      "loss": 1.3297,
      "step": 621
    },
    {
      "epoch": 0.052627125814366696,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009936911420546839,
      "loss": 2.0026,
      "step": 622
    },
    {
      "epoch": 0.05271173534139944,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009936700338966295,
      "loss": 1.5152,
      "step": 623
    },
    {
      "epoch": 0.05279634486843218,
      "grad_norm": 2.25,
      "learning_rate": 0.000993648890710662,
      "loss": 1.6053,
      "step": 624
    },
    {
      "epoch": 0.05288095439546493,
      "grad_norm": 3.40625,
      "learning_rate": 0.000993627712498282,
      "loss": 1.2965,
      "step": 625
    },
    {
      "epoch": 0.052965563922497676,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009936064992609917,
      "loss": 1.3589,
      "step": 626
    },
    {
      "epoch": 0.05305017344953042,
      "grad_norm": 5.09375,
      "learning_rate": 0.0009935852510002965,
      "loss": 2.1085,
      "step": 627
    },
    {
      "epoch": 0.05313478297656316,
      "grad_norm": 3.21875,
      "learning_rate": 0.000993563967717704,
      "loss": 1.7981,
      "step": 628
    },
    {
      "epoch": 0.053219392503595905,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009935426494147243,
      "loss": 1.3866,
      "step": 629
    },
    {
      "epoch": 0.05330400203062865,
      "grad_norm": 3.078125,
      "learning_rate": 0.00099352129609287,
      "loss": 1.7455,
      "step": 630
    },
    {
      "epoch": 0.05338861155766139,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009934999077536566,
      "loss": 1.417,
      "step": 631
    },
    {
      "epoch": 0.053473221084694135,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009934784843986014,
      "loss": 1.609,
      "step": 632
    },
    {
      "epoch": 0.05355783061172688,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009934570260292242,
      "loss": 1.5699,
      "step": 633
    },
    {
      "epoch": 0.05364244013875962,
      "grad_norm": 4.125,
      "learning_rate": 0.000993435532647048,
      "loss": 1.3726,
      "step": 634
    },
    {
      "epoch": 0.05372704966579237,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009934140042535974,
      "loss": 1.7899,
      "step": 635
    },
    {
      "epoch": 0.053811659192825115,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009933924408504003,
      "loss": 1.9819,
      "step": 636
    },
    {
      "epoch": 0.05389626871985786,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009933708424389868,
      "loss": 2.1737,
      "step": 637
    },
    {
      "epoch": 0.0539808782468906,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009933492090208888,
      "loss": 1.491,
      "step": 638
    },
    {
      "epoch": 0.054065487773923344,
      "grad_norm": 2.234375,
      "learning_rate": 0.000993327540597642,
      "loss": 1.7265,
      "step": 639
    },
    {
      "epoch": 0.05415009730095609,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009933058371707833,
      "loss": 1.9788,
      "step": 640
    },
    {
      "epoch": 0.05423470682798883,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009932840987418531,
      "loss": 1.5674,
      "step": 641
    },
    {
      "epoch": 0.054319316355021574,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009932623253123935,
      "loss": 1.7566,
      "step": 642
    },
    {
      "epoch": 0.05440392588205432,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009932405168839495,
      "loss": 1.6038,
      "step": 643
    },
    {
      "epoch": 0.05448853540908706,
      "grad_norm": 2.625,
      "learning_rate": 0.0009932186734580685,
      "loss": 1.7238,
      "step": 644
    },
    {
      "epoch": 0.054573144936119804,
      "grad_norm": 2.125,
      "learning_rate": 0.0009931967950363005,
      "loss": 1.1529,
      "step": 645
    },
    {
      "epoch": 0.054657754463152554,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009931748816201977,
      "loss": 1.6076,
      "step": 646
    },
    {
      "epoch": 0.0547423639901853,
      "grad_norm": 2.375,
      "learning_rate": 0.0009931529332113153,
      "loss": 1.53,
      "step": 647
    },
    {
      "epoch": 0.05482697351721804,
      "grad_norm": 2.109375,
      "learning_rate": 0.00099313094981121,
      "loss": 1.2956,
      "step": 648
    },
    {
      "epoch": 0.054911583044250784,
      "grad_norm": 2.15625,
      "learning_rate": 0.000993108931421442,
      "loss": 1.419,
      "step": 649
    },
    {
      "epoch": 0.05499619257128353,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009930868780435736,
      "loss": 1.7091,
      "step": 650
    },
    {
      "epoch": 0.05508080209831627,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009930647896791696,
      "loss": 1.5703,
      "step": 651
    },
    {
      "epoch": 0.05516541162534901,
      "grad_norm": 1.9921875,
      "learning_rate": 0.000993042666329797,
      "loss": 1.4627,
      "step": 652
    },
    {
      "epoch": 0.055250021152381756,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000993020507997026,
      "loss": 1.1559,
      "step": 653
    },
    {
      "epoch": 0.0553346306794145,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009929983146824285,
      "loss": 1.4138,
      "step": 654
    },
    {
      "epoch": 0.05541924020644724,
      "grad_norm": 2.484375,
      "learning_rate": 0.000992976086387579,
      "loss": 1.4449,
      "step": 655
    },
    {
      "epoch": 0.05550384973347999,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009929538231140552,
      "loss": 1.5909,
      "step": 656
    },
    {
      "epoch": 0.055588459260512736,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009929315248634362,
      "loss": 1.1685,
      "step": 657
    },
    {
      "epoch": 0.05567306878754548,
      "grad_norm": 1.7890625,
      "learning_rate": 0.000992909191637305,
      "loss": 1.3446,
      "step": 658
    },
    {
      "epoch": 0.05575767831457822,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009928868234372452,
      "loss": 1.6815,
      "step": 659
    },
    {
      "epoch": 0.055842287841610966,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009928644202648446,
      "loss": 1.393,
      "step": 660
    },
    {
      "epoch": 0.05592689736864371,
      "grad_norm": 5.28125,
      "learning_rate": 0.000992841982121693,
      "loss": 2.9083,
      "step": 661
    },
    {
      "epoch": 0.05601150689567645,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009928195090093818,
      "loss": 1.6078,
      "step": 662
    },
    {
      "epoch": 0.056096116422709195,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009927970009295057,
      "loss": 1.7307,
      "step": 663
    },
    {
      "epoch": 0.05618072594974194,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009927744578836618,
      "loss": 1.7337,
      "step": 664
    },
    {
      "epoch": 0.05626533547677468,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00099275187987345,
      "loss": 1.6907,
      "step": 665
    },
    {
      "epoch": 0.05634994500380743,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009927292669004721,
      "loss": 1.4351,
      "step": 666
    },
    {
      "epoch": 0.056434554530840175,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009927066189663323,
      "loss": 1.7726,
      "step": 667
    },
    {
      "epoch": 0.05651916405787292,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009926839360726379,
      "loss": 1.2951,
      "step": 668
    },
    {
      "epoch": 0.05660377358490566,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009926612182209983,
      "loss": 1.2836,
      "step": 669
    },
    {
      "epoch": 0.056688383111938405,
      "grad_norm": 2.65625,
      "learning_rate": 0.000992638465413025,
      "loss": 1.9349,
      "step": 670
    },
    {
      "epoch": 0.05677299263897115,
      "grad_norm": 2.3125,
      "learning_rate": 0.000992615677650333,
      "loss": 1.4464,
      "step": 671
    },
    {
      "epoch": 0.05685760216600389,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009925928549345388,
      "loss": 1.9914,
      "step": 672
    },
    {
      "epoch": 0.056942211693036635,
      "grad_norm": 1.8984375,
      "learning_rate": 0.000992569997267262,
      "loss": 1.186,
      "step": 673
    },
    {
      "epoch": 0.05702682122006938,
      "grad_norm": 21.875,
      "learning_rate": 0.0009925471046501244,
      "loss": 1.4508,
      "step": 674
    },
    {
      "epoch": 0.05711143074710212,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009925241770847501,
      "loss": 1.4313,
      "step": 675
    },
    {
      "epoch": 0.057196040274134864,
      "grad_norm": 2.078125,
      "learning_rate": 0.000992501214572766,
      "loss": 1.356,
      "step": 676
    },
    {
      "epoch": 0.057280649801167614,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009924782171158019,
      "loss": 1.6079,
      "step": 677
    },
    {
      "epoch": 0.05736525932820036,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009924551847154885,
      "loss": 1.2967,
      "step": 678
    },
    {
      "epoch": 0.0574498688552331,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009924321173734611,
      "loss": 1.246,
      "step": 679
    },
    {
      "epoch": 0.057534478382265844,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009924090150913558,
      "loss": 1.5219,
      "step": 680
    },
    {
      "epoch": 0.05761908790929859,
      "grad_norm": 2.875,
      "learning_rate": 0.0009923858778708122,
      "loss": 1.6032,
      "step": 681
    },
    {
      "epoch": 0.05770369743633133,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009923627057134715,
      "loss": 2.0224,
      "step": 682
    },
    {
      "epoch": 0.057788306963364074,
      "grad_norm": 1.796875,
      "learning_rate": 0.0009923394986209781,
      "loss": 1.0825,
      "step": 683
    },
    {
      "epoch": 0.05787291649039682,
      "grad_norm": 2.828125,
      "learning_rate": 0.000992316256594979,
      "loss": 1.7923,
      "step": 684
    },
    {
      "epoch": 0.05795752601742956,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009922929796371228,
      "loss": 1.2554,
      "step": 685
    },
    {
      "epoch": 0.0580421355444623,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009922696677490612,
      "loss": 1.477,
      "step": 686
    },
    {
      "epoch": 0.05812674507149505,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009922463209324484,
      "loss": 1.9843,
      "step": 687
    },
    {
      "epoch": 0.0582113545985278,
      "grad_norm": 1.71875,
      "learning_rate": 0.0009922229391889409,
      "loss": 1.2976,
      "step": 688
    },
    {
      "epoch": 0.05829596412556054,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009921995225201977,
      "loss": 1.3527,
      "step": 689
    },
    {
      "epoch": 0.05838057365259328,
      "grad_norm": 2.125,
      "learning_rate": 0.0009921760709278803,
      "loss": 1.4138,
      "step": 690
    },
    {
      "epoch": 0.058465183179626026,
      "grad_norm": 2.03125,
      "learning_rate": 0.000992152584413653,
      "loss": 1.2841,
      "step": 691
    },
    {
      "epoch": 0.05854979270665877,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009921290629791818,
      "loss": 1.4886,
      "step": 692
    },
    {
      "epoch": 0.05863440223369151,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009921055066261356,
      "loss": 1.489,
      "step": 693
    },
    {
      "epoch": 0.058719011760724256,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009920819153561864,
      "loss": 1.6962,
      "step": 694
    },
    {
      "epoch": 0.058803621287757,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009920582891710075,
      "loss": 1.5303,
      "step": 695
    },
    {
      "epoch": 0.05888823081478974,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009920346280722758,
      "loss": 1.5211,
      "step": 696
    },
    {
      "epoch": 0.05897284034182249,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009920109320616698,
      "loss": 1.1595,
      "step": 697
    },
    {
      "epoch": 0.059057449868855236,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009919872011408708,
      "loss": 1.8798,
      "step": 698
    },
    {
      "epoch": 0.05914205939588798,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009919634353115628,
      "loss": 2.0343,
      "step": 699
    },
    {
      "epoch": 0.05922666892292072,
      "grad_norm": 2.734375,
      "learning_rate": 0.000991939634575432,
      "loss": 1.7153,
      "step": 700
    },
    {
      "epoch": 0.059311278449953465,
      "grad_norm": 2.59375,
      "learning_rate": 0.000991915798934167,
      "loss": 1.6021,
      "step": 701
    },
    {
      "epoch": 0.05939588797698621,
      "grad_norm": 2.75,
      "learning_rate": 0.0009918919283894594,
      "loss": 2.0145,
      "step": 702
    },
    {
      "epoch": 0.05948049750401895,
      "grad_norm": 10.8125,
      "learning_rate": 0.0009918680229430027,
      "loss": 1.7758,
      "step": 703
    },
    {
      "epoch": 0.059565107031051695,
      "grad_norm": 87.0,
      "learning_rate": 0.000991844082596493,
      "loss": 1.5483,
      "step": 704
    },
    {
      "epoch": 0.05964971655808444,
      "grad_norm": 17.0,
      "learning_rate": 0.000991820107351629,
      "loss": 1.4835,
      "step": 705
    },
    {
      "epoch": 0.05973432608511718,
      "grad_norm": 3.0,
      "learning_rate": 0.0009917960972101118,
      "loss": 1.5788,
      "step": 706
    },
    {
      "epoch": 0.059818935612149925,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009917720521736453,
      "loss": 1.7573,
      "step": 707
    },
    {
      "epoch": 0.059903545139182675,
      "grad_norm": 4.625,
      "learning_rate": 0.0009917479722439356,
      "loss": 1.4461,
      "step": 708
    },
    {
      "epoch": 0.05998815466621542,
      "grad_norm": 3.140625,
      "learning_rate": 0.0009917238574226907,
      "loss": 2.0573,
      "step": 709
    },
    {
      "epoch": 0.06007276419324816,
      "grad_norm": 9.0625,
      "learning_rate": 0.0009916997077116223,
      "loss": 1.5221,
      "step": 710
    },
    {
      "epoch": 0.060157373720280904,
      "grad_norm": 20.625,
      "learning_rate": 0.0009916755231124438,
      "loss": 1.5474,
      "step": 711
    },
    {
      "epoch": 0.06024198324731365,
      "grad_norm": 37.5,
      "learning_rate": 0.0009916513036268707,
      "loss": 1.5596,
      "step": 712
    },
    {
      "epoch": 0.06032659277434639,
      "grad_norm": 9.0625,
      "learning_rate": 0.000991627049256622,
      "loss": 1.5698,
      "step": 713
    },
    {
      "epoch": 0.060411202301379134,
      "grad_norm": 5.5625,
      "learning_rate": 0.0009916027600034183,
      "loss": 1.4073,
      "step": 714
    },
    {
      "epoch": 0.06049581182841188,
      "grad_norm": 11.1875,
      "learning_rate": 0.0009915784358689834,
      "loss": 2.1335,
      "step": 715
    },
    {
      "epoch": 0.06058042135544462,
      "grad_norm": 6.34375,
      "learning_rate": 0.000991554076855043,
      "loss": 1.6378,
      "step": 716
    },
    {
      "epoch": 0.060665030882477364,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009915296829633253,
      "loss": 1.3378,
      "step": 717
    },
    {
      "epoch": 0.060749640409510114,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009915052541955612,
      "loss": 1.7154,
      "step": 718
    },
    {
      "epoch": 0.06083424993654286,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009914807905534843,
      "loss": 1.574,
      "step": 719
    },
    {
      "epoch": 0.0609188594635756,
      "grad_norm": 5.125,
      "learning_rate": 0.0009914562920388303,
      "loss": 1.0806,
      "step": 720
    },
    {
      "epoch": 0.061003468990608344,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009914317586533373,
      "loss": 1.4744,
      "step": 721
    },
    {
      "epoch": 0.06108807851764109,
      "grad_norm": 2.390625,
      "learning_rate": 0.000991407190398746,
      "loss": 1.752,
      "step": 722
    },
    {
      "epoch": 0.06117268804467383,
      "grad_norm": 4.25,
      "learning_rate": 0.0009913825872768,
      "loss": 1.405,
      "step": 723
    },
    {
      "epoch": 0.06125729757170657,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009913579492892447,
      "loss": 1.5219,
      "step": 724
    },
    {
      "epoch": 0.061341907098739316,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009913332764378282,
      "loss": 1.3847,
      "step": 725
    },
    {
      "epoch": 0.06142651662577206,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009913085687243014,
      "loss": 2.0721,
      "step": 726
    },
    {
      "epoch": 0.0615111261528048,
      "grad_norm": 4.75,
      "learning_rate": 0.000991283826150417,
      "loss": 1.373,
      "step": 727
    },
    {
      "epoch": 0.06159573567983755,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009912590487179311,
      "loss": 1.8838,
      "step": 728
    },
    {
      "epoch": 0.061680345206870296,
      "grad_norm": 2.125,
      "learning_rate": 0.0009912342364286014,
      "loss": 1.6944,
      "step": 729
    },
    {
      "epoch": 0.06176495473390304,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009912093892841885,
      "loss": 1.4671,
      "step": 730
    },
    {
      "epoch": 0.06184956426093578,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009911845072864557,
      "loss": 1.8892,
      "step": 731
    },
    {
      "epoch": 0.061934173787968526,
      "grad_norm": 2.453125,
      "learning_rate": 0.000991159590437168,
      "loss": 1.6079,
      "step": 732
    },
    {
      "epoch": 0.06201878331500127,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009911346387380935,
      "loss": 1.2397,
      "step": 733
    },
    {
      "epoch": 0.06210339284203401,
      "grad_norm": 1.9765625,
      "learning_rate": 0.000991109652191003,
      "loss": 1.5609,
      "step": 734
    },
    {
      "epoch": 0.062188002369066755,
      "grad_norm": 1.640625,
      "learning_rate": 0.000991084630797669,
      "loss": 1.1347,
      "step": 735
    },
    {
      "epoch": 0.0622726118960995,
      "grad_norm": 2.703125,
      "learning_rate": 0.000991059574559867,
      "loss": 1.2393,
      "step": 736
    },
    {
      "epoch": 0.06235722142313224,
      "grad_norm": 2.0,
      "learning_rate": 0.0009910344834793748,
      "loss": 1.5733,
      "step": 737
    },
    {
      "epoch": 0.062441830950164985,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009910093575579725,
      "loss": 1.396,
      "step": 738
    },
    {
      "epoch": 0.06252644047719773,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009909841967974435,
      "loss": 1.7831,
      "step": 739
    },
    {
      "epoch": 0.06261105000423048,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009909590011995726,
      "loss": 1.3194,
      "step": 740
    },
    {
      "epoch": 0.06269565953126321,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009909337707661474,
      "loss": 1.2303,
      "step": 741
    },
    {
      "epoch": 0.06278026905829596,
      "grad_norm": 3.96875,
      "learning_rate": 0.0009909085054989584,
      "loss": 2.2183,
      "step": 742
    },
    {
      "epoch": 0.0628648785853287,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009908832053997984,
      "loss": 1.1788,
      "step": 743
    },
    {
      "epoch": 0.06294948811236145,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009908578704704621,
      "loss": 1.4907,
      "step": 744
    },
    {
      "epoch": 0.0630340976393942,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009908325007127474,
      "loss": 1.3706,
      "step": 745
    },
    {
      "epoch": 0.06311870716642694,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009908070961284546,
      "loss": 1.5826,
      "step": 746
    },
    {
      "epoch": 0.06320331669345969,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009907816567193857,
      "loss": 1.4454,
      "step": 747
    },
    {
      "epoch": 0.06328792622049242,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009907561824873462,
      "loss": 1.2485,
      "step": 748
    },
    {
      "epoch": 0.06337253574752517,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009907306734341433,
      "loss": 1.3329,
      "step": 749
    },
    {
      "epoch": 0.06345714527455791,
      "grad_norm": 41.0,
      "learning_rate": 0.0009907051295615873,
      "loss": 1.9753,
      "step": 750
    },
    {
      "epoch": 0.06354175480159066,
      "grad_norm": 1.875,
      "learning_rate": 0.00099067955087149,
      "loss": 1.1702,
      "step": 751
    },
    {
      "epoch": 0.0636263643286234,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009906539373656673,
      "loss": 1.4072,
      "step": 752
    },
    {
      "epoch": 0.06371097385565615,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009906282890459357,
      "loss": 1.3549,
      "step": 753
    },
    {
      "epoch": 0.0637955833826889,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009906026059141155,
      "loss": 1.5511,
      "step": 754
    },
    {
      "epoch": 0.06388019290972163,
      "grad_norm": 2.828125,
      "learning_rate": 0.000990576887972029,
      "loss": 1.7139,
      "step": 755
    },
    {
      "epoch": 0.06396480243675438,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009905511352215008,
      "loss": 1.4449,
      "step": 756
    },
    {
      "epoch": 0.06404941196378712,
      "grad_norm": 2.0,
      "learning_rate": 0.0009905253476643586,
      "loss": 1.7991,
      "step": 757
    },
    {
      "epoch": 0.06413402149081987,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009904995253024313,
      "loss": 2.0944,
      "step": 758
    },
    {
      "epoch": 0.0642186310178526,
      "grad_norm": 2.65625,
      "learning_rate": 0.000990473668137552,
      "loss": 1.7127,
      "step": 759
    },
    {
      "epoch": 0.06430324054488536,
      "grad_norm": 1.9453125,
      "learning_rate": 0.000990447776171555,
      "loss": 1.3375,
      "step": 760
    },
    {
      "epoch": 0.06438785007191809,
      "grad_norm": 1.953125,
      "learning_rate": 0.000990421849406277,
      "loss": 1.7721,
      "step": 761
    },
    {
      "epoch": 0.06447245959895084,
      "grad_norm": 2.375,
      "learning_rate": 0.0009903958878435585,
      "loss": 1.3878,
      "step": 762
    },
    {
      "epoch": 0.06455706912598358,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009903698914852407,
      "loss": 1.4458,
      "step": 763
    },
    {
      "epoch": 0.06464167865301633,
      "grad_norm": 1.9765625,
      "learning_rate": 0.000990343860333169,
      "loss": 1.2819,
      "step": 764
    },
    {
      "epoch": 0.06472628818004908,
      "grad_norm": 2.75,
      "learning_rate": 0.0009903177943891897,
      "loss": 1.6695,
      "step": 765
    },
    {
      "epoch": 0.06481089770708182,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009902916936551528,
      "loss": 1.7797,
      "step": 766
    },
    {
      "epoch": 0.06489550723411457,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009902655581329098,
      "loss": 1.3855,
      "step": 767
    },
    {
      "epoch": 0.0649801167611473,
      "grad_norm": 2.0,
      "learning_rate": 0.0009902393878243154,
      "loss": 1.3078,
      "step": 768
    },
    {
      "epoch": 0.06506472628818005,
      "grad_norm": 3.25,
      "learning_rate": 0.0009902131827312264,
      "loss": 1.7763,
      "step": 769
    },
    {
      "epoch": 0.06514933581521279,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009901869428555023,
      "loss": 1.8768,
      "step": 770
    },
    {
      "epoch": 0.06523394534224554,
      "grad_norm": 1.71875,
      "learning_rate": 0.0009901606681990047,
      "loss": 1.3035,
      "step": 771
    },
    {
      "epoch": 0.06531855486927828,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009901343587635982,
      "loss": 1.7163,
      "step": 772
    },
    {
      "epoch": 0.06540316439631103,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000990108014551149,
      "loss": 1.1899,
      "step": 773
    },
    {
      "epoch": 0.06548777392334376,
      "grad_norm": 1.609375,
      "learning_rate": 0.0009900816355635272,
      "loss": 1.1712,
      "step": 774
    },
    {
      "epoch": 0.06557238345037651,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009900552218026037,
      "loss": 1.4713,
      "step": 775
    },
    {
      "epoch": 0.06565699297740926,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009900287732702532,
      "loss": 1.2781,
      "step": 776
    },
    {
      "epoch": 0.065741602504442,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009900022899683519,
      "loss": 1.8668,
      "step": 777
    },
    {
      "epoch": 0.06582621203147475,
      "grad_norm": 2.359375,
      "learning_rate": 0.000989975771898779,
      "loss": 1.7424,
      "step": 778
    },
    {
      "epoch": 0.06591082155850748,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009899492190634164,
      "loss": 2.0886,
      "step": 779
    },
    {
      "epoch": 0.06599543108554023,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009899226314641477,
      "loss": 1.5442,
      "step": 780
    },
    {
      "epoch": 0.06608004061257297,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009898960091028596,
      "loss": 1.5617,
      "step": 781
    },
    {
      "epoch": 0.06616465013960572,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009898693519814414,
      "loss": 1.5265,
      "step": 782
    },
    {
      "epoch": 0.06624925966663846,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009898426601017839,
      "loss": 1.3403,
      "step": 783
    },
    {
      "epoch": 0.06633386919367121,
      "grad_norm": 1.5625,
      "learning_rate": 0.0009898159334657815,
      "loss": 1.2974,
      "step": 784
    },
    {
      "epoch": 0.06641847872070396,
      "grad_norm": 2.140625,
      "learning_rate": 0.00098978917207533,
      "loss": 1.6667,
      "step": 785
    },
    {
      "epoch": 0.0665030882477367,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009897623759323288,
      "loss": 1.6101,
      "step": 786
    },
    {
      "epoch": 0.06658769777476944,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009897355450386792,
      "loss": 1.5363,
      "step": 787
    },
    {
      "epoch": 0.06667230730180218,
      "grad_norm": 2.125,
      "learning_rate": 0.0009897086793962844,
      "loss": 1.3417,
      "step": 788
    },
    {
      "epoch": 0.06675691682883493,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009896817790070512,
      "loss": 1.5175,
      "step": 789
    },
    {
      "epoch": 0.06684152635586767,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009896548438728878,
      "loss": 2.0061,
      "step": 790
    },
    {
      "epoch": 0.06692613588290042,
      "grad_norm": 2.0625,
      "learning_rate": 0.000989627873995706,
      "loss": 1.3307,
      "step": 791
    },
    {
      "epoch": 0.06701074540993315,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009896008693774188,
      "loss": 1.243,
      "step": 792
    },
    {
      "epoch": 0.0670953549369659,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009895738300199424,
      "loss": 1.1859,
      "step": 793
    },
    {
      "epoch": 0.06717996446399864,
      "grad_norm": 2.234375,
      "learning_rate": 0.000989546755925196,
      "loss": 1.9148,
      "step": 794
    },
    {
      "epoch": 0.06726457399103139,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009895196470950996,
      "loss": 1.9708,
      "step": 795
    },
    {
      "epoch": 0.06734918351806414,
      "grad_norm": 2.125,
      "learning_rate": 0.0009894925035315773,
      "loss": 1.4743,
      "step": 796
    },
    {
      "epoch": 0.06743379304509688,
      "grad_norm": 2.09375,
      "learning_rate": 0.000989465325236555,
      "loss": 1.3024,
      "step": 797
    },
    {
      "epoch": 0.06751840257212963,
      "grad_norm": 1.875,
      "learning_rate": 0.0009894381122119611,
      "loss": 1.3952,
      "step": 798
    },
    {
      "epoch": 0.06760301209916236,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009894108644597265,
      "loss": 1.6025,
      "step": 799
    },
    {
      "epoch": 0.06768762162619511,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009893835819817843,
      "loss": 1.4616,
      "step": 800
    },
    {
      "epoch": 0.06777223115322785,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009893562647800705,
      "loss": 1.4377,
      "step": 801
    },
    {
      "epoch": 0.0678568406802606,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009893289128565233,
      "loss": 1.1917,
      "step": 802
    },
    {
      "epoch": 0.06794145020729334,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009893015262130835,
      "loss": 1.6836,
      "step": 803
    },
    {
      "epoch": 0.06802605973432609,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009892741048516942,
      "loss": 1.8366,
      "step": 804
    },
    {
      "epoch": 0.06811066926135882,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009892466487743013,
      "loss": 1.2407,
      "step": 805
    },
    {
      "epoch": 0.06819527878839157,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009892191579828527,
      "loss": 1.3242,
      "step": 806
    },
    {
      "epoch": 0.06827988831542432,
      "grad_norm": 2.328125,
      "learning_rate": 0.000989191632479299,
      "loss": 1.3561,
      "step": 807
    },
    {
      "epoch": 0.06836449784245706,
      "grad_norm": 2.765625,
      "learning_rate": 0.000989164072265593,
      "loss": 1.7873,
      "step": 808
    },
    {
      "epoch": 0.06844910736948981,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009891364773436909,
      "loss": 1.866,
      "step": 809
    },
    {
      "epoch": 0.06853371689652255,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009891088477155501,
      "loss": 1.301,
      "step": 810
    },
    {
      "epoch": 0.0686183264235553,
      "grad_norm": 2.203125,
      "learning_rate": 0.000989081183383131,
      "loss": 1.9391,
      "step": 811
    },
    {
      "epoch": 0.06870293595058803,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009890534843483968,
      "loss": 1.538,
      "step": 812
    },
    {
      "epoch": 0.06878754547762078,
      "grad_norm": 2.703125,
      "learning_rate": 0.000989025750613313,
      "loss": 1.7845,
      "step": 813
    },
    {
      "epoch": 0.06887215500465352,
      "grad_norm": 2.375,
      "learning_rate": 0.000988997982179847,
      "loss": 1.6156,
      "step": 814
    },
    {
      "epoch": 0.06895676453168627,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009889701790499691,
      "loss": 1.8843,
      "step": 815
    },
    {
      "epoch": 0.06904137405871902,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009889423412256524,
      "loss": 2.4142,
      "step": 816
    },
    {
      "epoch": 0.06912598358575175,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009889144687088719,
      "loss": 1.3258,
      "step": 817
    },
    {
      "epoch": 0.0692105931127845,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009888865615016052,
      "loss": 1.7502,
      "step": 818
    },
    {
      "epoch": 0.06929520263981724,
      "grad_norm": 256.0,
      "learning_rate": 0.0009888586196058324,
      "loss": 2.3116,
      "step": 819
    },
    {
      "epoch": 0.06937981216684999,
      "grad_norm": 3200.0,
      "learning_rate": 0.0009888306430235363,
      "loss": 1.4503,
      "step": 820
    },
    {
      "epoch": 0.06946442169388273,
      "grad_norm": 1336.0,
      "learning_rate": 0.000988802631756702,
      "loss": 1.4526,
      "step": 821
    },
    {
      "epoch": 0.06954903122091548,
      "grad_norm": 239.0,
      "learning_rate": 0.0009887745858073167,
      "loss": 1.7473,
      "step": 822
    },
    {
      "epoch": 0.06963364074794821,
      "grad_norm": 8896.0,
      "learning_rate": 0.0009887465051773708,
      "loss": 1.654,
      "step": 823
    },
    {
      "epoch": 0.06971825027498096,
      "grad_norm": 260.0,
      "learning_rate": 0.0009887183898688561,
      "loss": 1.7425,
      "step": 824
    },
    {
      "epoch": 0.0698028598020137,
      "grad_norm": 82.5,
      "learning_rate": 0.000988690239883768,
      "loss": 1.2647,
      "step": 825
    },
    {
      "epoch": 0.06988746932904645,
      "grad_norm": 7.21875,
      "learning_rate": 0.000988662055224104,
      "loss": 1.8347,
      "step": 826
    },
    {
      "epoch": 0.0699720788560792,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009886338358918634,
      "loss": 1.7102,
      "step": 827
    },
    {
      "epoch": 0.07005668838311194,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009886055818890487,
      "loss": 1.2199,
      "step": 828
    },
    {
      "epoch": 0.07014129791014469,
      "grad_norm": 2.125,
      "learning_rate": 0.0009885772932176646,
      "loss": 1.6728,
      "step": 829
    },
    {
      "epoch": 0.07022590743717742,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009885489698797185,
      "loss": 1.6751,
      "step": 830
    },
    {
      "epoch": 0.07031051696421017,
      "grad_norm": 2.96875,
      "learning_rate": 0.00098852061187722,
      "loss": 2.0866,
      "step": 831
    },
    {
      "epoch": 0.07039512649124291,
      "grad_norm": 2.140625,
      "learning_rate": 0.000988492219212181,
      "loss": 1.8536,
      "step": 832
    },
    {
      "epoch": 0.07047973601827566,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009884637918866164,
      "loss": 1.3587,
      "step": 833
    },
    {
      "epoch": 0.0705643455453084,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009884353299025427,
      "loss": 1.8319,
      "step": 834
    },
    {
      "epoch": 0.07064895507234115,
      "grad_norm": 2.53125,
      "learning_rate": 0.00098840683326198,
      "loss": 1.502,
      "step": 835
    },
    {
      "epoch": 0.07073356459937388,
      "grad_norm": 2.078125,
      "learning_rate": 0.00098837830196695,
      "loss": 1.3094,
      "step": 836
    },
    {
      "epoch": 0.07081817412640663,
      "grad_norm": 2.296875,
      "learning_rate": 0.000988349736019477,
      "loss": 1.7868,
      "step": 837
    },
    {
      "epoch": 0.07090278365343938,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009883211354215881,
      "loss": 1.7548,
      "step": 838
    },
    {
      "epoch": 0.07098739318047212,
      "grad_norm": 19.5,
      "learning_rate": 0.0009882925001753124,
      "loss": 1.4973,
      "step": 839
    },
    {
      "epoch": 0.07107200270750487,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009882638302826819,
      "loss": 1.4404,
      "step": 840
    },
    {
      "epoch": 0.0711566122345376,
      "grad_norm": 2.25,
      "learning_rate": 0.0009882351257457305,
      "loss": 1.2324,
      "step": 841
    },
    {
      "epoch": 0.07124122176157036,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009882063865664953,
      "loss": 1.2775,
      "step": 842
    },
    {
      "epoch": 0.07132583128860309,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009881776127470155,
      "loss": 1.2671,
      "step": 843
    },
    {
      "epoch": 0.07141044081563584,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009881488042893323,
      "loss": 1.5563,
      "step": 844
    },
    {
      "epoch": 0.07149505034266858,
      "grad_norm": 3.703125,
      "learning_rate": 0.0009881199611954901,
      "loss": 1.2645,
      "step": 845
    },
    {
      "epoch": 0.07157965986970133,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009880910834675352,
      "loss": 1.1689,
      "step": 846
    },
    {
      "epoch": 0.07166426939673408,
      "grad_norm": 7.90625,
      "learning_rate": 0.0009880621711075169,
      "loss": 2.4346,
      "step": 847
    },
    {
      "epoch": 0.07174887892376682,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009880332241174864,
      "loss": 1.6908,
      "step": 848
    },
    {
      "epoch": 0.07183348845079957,
      "grad_norm": 15.9375,
      "learning_rate": 0.0009880042424994977,
      "loss": 1.3413,
      "step": 849
    },
    {
      "epoch": 0.0719180979778323,
      "grad_norm": 2.375,
      "learning_rate": 0.0009879752262556072,
      "loss": 1.2217,
      "step": 850
    },
    {
      "epoch": 0.07200270750486505,
      "grad_norm": 16.375,
      "learning_rate": 0.000987946175387874,
      "loss": 1.9302,
      "step": 851
    },
    {
      "epoch": 0.07208731703189779,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009879170898983586,
      "loss": 1.5108,
      "step": 852
    },
    {
      "epoch": 0.07217192655893054,
      "grad_norm": 2.625,
      "learning_rate": 0.000987887969789125,
      "loss": 1.8845,
      "step": 853
    },
    {
      "epoch": 0.07225653608596327,
      "grad_norm": 2.0,
      "learning_rate": 0.00098785881506224,
      "loss": 1.4357,
      "step": 854
    },
    {
      "epoch": 0.07234114561299602,
      "grad_norm": 2.703125,
      "learning_rate": 0.000987829625719772,
      "loss": 1.9615,
      "step": 855
    },
    {
      "epoch": 0.07242575514002876,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009878004017637916,
      "loss": 1.3379,
      "step": 856
    },
    {
      "epoch": 0.07251036466706151,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009877711431963726,
      "loss": 1.2542,
      "step": 857
    },
    {
      "epoch": 0.07259497419409426,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009877418500195915,
      "loss": 1.6995,
      "step": 858
    },
    {
      "epoch": 0.072679583721127,
      "grad_norm": 2.34375,
      "learning_rate": 0.000987712522235526,
      "loss": 1.3208,
      "step": 859
    },
    {
      "epoch": 0.07276419324815975,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009876831598462576,
      "loss": 1.4122,
      "step": 860
    },
    {
      "epoch": 0.07284880277519248,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009876537628538695,
      "loss": 1.3085,
      "step": 861
    },
    {
      "epoch": 0.07293341230222523,
      "grad_norm": 3.5,
      "learning_rate": 0.0009876243312604476,
      "loss": 1.6158,
      "step": 862
    },
    {
      "epoch": 0.07301802182925797,
      "grad_norm": 2.71875,
      "learning_rate": 0.00098759486506808,
      "loss": 2.0759,
      "step": 863
    },
    {
      "epoch": 0.07310263135629072,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009875653642788574,
      "loss": 1.1766,
      "step": 864
    },
    {
      "epoch": 0.07318724088332346,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009875358288948734,
      "loss": 1.644,
      "step": 865
    },
    {
      "epoch": 0.0732718504103562,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009875062589182234,
      "loss": 1.6867,
      "step": 866
    },
    {
      "epoch": 0.07335645993738894,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009874766543510056,
      "loss": 1.5969,
      "step": 867
    },
    {
      "epoch": 0.0734410694644217,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009874470151953203,
      "loss": 1.4714,
      "step": 868
    },
    {
      "epoch": 0.07352567899145444,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009874173414532708,
      "loss": 1.4911,
      "step": 869
    },
    {
      "epoch": 0.07361028851848718,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009873876331269627,
      "loss": 1.256,
      "step": 870
    },
    {
      "epoch": 0.07369489804551993,
      "grad_norm": 2.0,
      "learning_rate": 0.0009873578902185034,
      "loss": 1.3236,
      "step": 871
    },
    {
      "epoch": 0.07377950757255267,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009873281127300037,
      "loss": 1.5322,
      "step": 872
    },
    {
      "epoch": 0.07386411709958542,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009872983006635765,
      "loss": 1.3789,
      "step": 873
    },
    {
      "epoch": 0.07394872662661815,
      "grad_norm": 2.0,
      "learning_rate": 0.0009872684540213367,
      "loss": 1.5366,
      "step": 874
    },
    {
      "epoch": 0.0740333361536509,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009872385728054026,
      "loss": 1.3476,
      "step": 875
    },
    {
      "epoch": 0.07411794568068364,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009872086570178937,
      "loss": 1.3212,
      "step": 876
    },
    {
      "epoch": 0.07420255520771639,
      "grad_norm": 2.0,
      "learning_rate": 0.0009871787066609333,
      "loss": 1.5854,
      "step": 877
    },
    {
      "epoch": 0.07428716473474914,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0009871487217366461,
      "loss": 1.9873,
      "step": 878
    },
    {
      "epoch": 0.07437177426178188,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009871187022471597,
      "loss": 1.6567,
      "step": 879
    },
    {
      "epoch": 0.07445638378881463,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009870886481946042,
      "loss": 2.2102,
      "step": 880
    },
    {
      "epoch": 0.07454099331584736,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009870585595811122,
      "loss": 1.6831,
      "step": 881
    },
    {
      "epoch": 0.07462560284288011,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009870284364088182,
      "loss": 1.3647,
      "step": 882
    },
    {
      "epoch": 0.07471021236991285,
      "grad_norm": 1.921875,
      "learning_rate": 0.00098699827867986,
      "loss": 1.6757,
      "step": 883
    },
    {
      "epoch": 0.0747948218969456,
      "grad_norm": 2.25,
      "learning_rate": 0.000986968086396377,
      "loss": 1.7809,
      "step": 884
    },
    {
      "epoch": 0.07487943142397833,
      "grad_norm": 88.0,
      "learning_rate": 0.0009869378595605118,
      "loss": 1.5137,
      "step": 885
    },
    {
      "epoch": 0.07496404095101109,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009869075981744089,
      "loss": 1.2163,
      "step": 886
    },
    {
      "epoch": 0.07504865047804382,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009868773022402159,
      "loss": 2.0105,
      "step": 887
    },
    {
      "epoch": 0.07513326000507657,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009868469717600817,
      "loss": 1.2363,
      "step": 888
    },
    {
      "epoch": 0.07521786953210932,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009868166067361589,
      "loss": 1.373,
      "step": 889
    },
    {
      "epoch": 0.07530247905914206,
      "grad_norm": 2.765625,
      "learning_rate": 0.000986786207170602,
      "loss": 1.7164,
      "step": 890
    },
    {
      "epoch": 0.07538708858617481,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009867557730655677,
      "loss": 1.5228,
      "step": 891
    },
    {
      "epoch": 0.07547169811320754,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009867253044232155,
      "loss": 1.3065,
      "step": 892
    },
    {
      "epoch": 0.0755563076402403,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009866948012457078,
      "loss": 1.2471,
      "step": 893
    },
    {
      "epoch": 0.07564091716727303,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009866642635352081,
      "loss": 1.4287,
      "step": 894
    },
    {
      "epoch": 0.07572552669430578,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009866336912938837,
      "loss": 1.2643,
      "step": 895
    },
    {
      "epoch": 0.07581013622133852,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009866030845239037,
      "loss": 1.5419,
      "step": 896
    },
    {
      "epoch": 0.07589474574837127,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009865724432274396,
      "loss": 2.086,
      "step": 897
    },
    {
      "epoch": 0.075979355275404,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009865417674066658,
      "loss": 1.5489,
      "step": 898
    },
    {
      "epoch": 0.07606396480243675,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009865110570637587,
      "loss": 1.5994,
      "step": 899
    },
    {
      "epoch": 0.0761485743294695,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009864803122008976,
      "loss": 1.5424,
      "step": 900
    },
    {
      "epoch": 0.07623318385650224,
      "grad_norm": 5.0,
      "learning_rate": 0.0009864495328202635,
      "loss": 1.8551,
      "step": 901
    },
    {
      "epoch": 0.07631779338353499,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009864187189240407,
      "loss": 1.3119,
      "step": 902
    },
    {
      "epoch": 0.07640240291056773,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009863878705144156,
      "loss": 1.3557,
      "step": 903
    },
    {
      "epoch": 0.07648701243760048,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009863569875935767,
      "loss": 1.1767,
      "step": 904
    },
    {
      "epoch": 0.07657162196463321,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009863260701637154,
      "loss": 1.6619,
      "step": 905
    },
    {
      "epoch": 0.07665623149166596,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009862951182270257,
      "loss": 1.4369,
      "step": 906
    },
    {
      "epoch": 0.0767408410186987,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009862641317857034,
      "loss": 1.2868,
      "step": 907
    },
    {
      "epoch": 0.07682545054573145,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009862331108419472,
      "loss": 1.4289,
      "step": 908
    },
    {
      "epoch": 0.0769100600727642,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009862020553979582,
      "loss": 1.4375,
      "step": 909
    },
    {
      "epoch": 0.07699466959979694,
      "grad_norm": 3.484375,
      "learning_rate": 0.00098617096545594,
      "loss": 1.2853,
      "step": 910
    },
    {
      "epoch": 0.07707927912682969,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009861398410180983,
      "loss": 1.5263,
      "step": 911
    },
    {
      "epoch": 0.07716388865386242,
      "grad_norm": 4.4375,
      "learning_rate": 0.000986108682086642,
      "loss": 1.3333,
      "step": 912
    },
    {
      "epoch": 0.07724849818089517,
      "grad_norm": 2.75,
      "learning_rate": 0.0009860774886637814,
      "loss": 1.8517,
      "step": 913
    },
    {
      "epoch": 0.07733310770792791,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009860462607517303,
      "loss": 1.933,
      "step": 914
    },
    {
      "epoch": 0.07741771723496066,
      "grad_norm": 3.15625,
      "learning_rate": 0.000986014998352704,
      "loss": 1.8137,
      "step": 915
    },
    {
      "epoch": 0.0775023267619934,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009859837014689209,
      "loss": 1.5091,
      "step": 916
    },
    {
      "epoch": 0.07758693628902615,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009859523701026016,
      "loss": 1.5885,
      "step": 917
    },
    {
      "epoch": 0.07767154581605888,
      "grad_norm": 3.484375,
      "learning_rate": 0.0009859210042559693,
      "loss": 1.7174,
      "step": 918
    },
    {
      "epoch": 0.07775615534309163,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009858896039312494,
      "loss": 1.4618,
      "step": 919
    },
    {
      "epoch": 0.07784076487012438,
      "grad_norm": 2.125,
      "learning_rate": 0.00098585816913067,
      "loss": 1.5152,
      "step": 920
    },
    {
      "epoch": 0.07792537439715712,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009858266998564614,
      "loss": 1.9132,
      "step": 921
    },
    {
      "epoch": 0.07800998392418987,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009857951961108565,
      "loss": 1.9144,
      "step": 922
    },
    {
      "epoch": 0.0780945934512226,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009857636578960909,
      "loss": 1.5054,
      "step": 923
    },
    {
      "epoch": 0.07817920297825535,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009857320852144017,
      "loss": 1.6493,
      "step": 924
    },
    {
      "epoch": 0.07826381250528809,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009857004780680298,
      "loss": 1.305,
      "step": 925
    },
    {
      "epoch": 0.07834842203232084,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009856688364592177,
      "loss": 1.5856,
      "step": 926
    },
    {
      "epoch": 0.07843303155935358,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009856371603902104,
      "loss": 1.8333,
      "step": 927
    },
    {
      "epoch": 0.07851764108638633,
      "grad_norm": 2.0,
      "learning_rate": 0.0009856054498632552,
      "loss": 1.3659,
      "step": 928
    },
    {
      "epoch": 0.07860225061341906,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009855737048806025,
      "loss": 1.6661,
      "step": 929
    },
    {
      "epoch": 0.07868686014045181,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009855419254445044,
      "loss": 1.576,
      "step": 930
    },
    {
      "epoch": 0.07877146966748456,
      "grad_norm": 1.75,
      "learning_rate": 0.0009855101115572161,
      "loss": 1.5323,
      "step": 931
    },
    {
      "epoch": 0.0788560791945173,
      "grad_norm": 1.53125,
      "learning_rate": 0.0009854782632209946,
      "loss": 1.1929,
      "step": 932
    },
    {
      "epoch": 0.07894068872155005,
      "grad_norm": 2.5,
      "learning_rate": 0.0009854463804381001,
      "loss": 1.8038,
      "step": 933
    },
    {
      "epoch": 0.07902529824858279,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009854144632107944,
      "loss": 1.2447,
      "step": 934
    },
    {
      "epoch": 0.07910990777561554,
      "grad_norm": 6.90625,
      "learning_rate": 0.0009853825115413424,
      "loss": 1.8586,
      "step": 935
    },
    {
      "epoch": 0.07919451730264827,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009853505254320111,
      "loss": 1.1978,
      "step": 936
    },
    {
      "epoch": 0.07927912682968102,
      "grad_norm": 2.21875,
      "learning_rate": 0.00098531850488507,
      "loss": 1.8378,
      "step": 937
    },
    {
      "epoch": 0.07936373635671376,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009852864499027912,
      "loss": 1.6816,
      "step": 938
    },
    {
      "epoch": 0.07944834588374651,
      "grad_norm": 2.234375,
      "learning_rate": 0.000985254360487449,
      "loss": 2.4547,
      "step": 939
    },
    {
      "epoch": 0.07953295541077926,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0009852222366413207,
      "loss": 1.337,
      "step": 940
    },
    {
      "epoch": 0.079617564937812,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000985190078366685,
      "loss": 1.3058,
      "step": 941
    },
    {
      "epoch": 0.07970217446484475,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009851578856658238,
      "loss": 1.9367,
      "step": 942
    },
    {
      "epoch": 0.07978678399187748,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009851256585410217,
      "loss": 1.9327,
      "step": 943
    },
    {
      "epoch": 0.07987139351891023,
      "grad_norm": 1.78125,
      "learning_rate": 0.000985093396994565,
      "loss": 1.3476,
      "step": 944
    },
    {
      "epoch": 0.07995600304594297,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009850611010287428,
      "loss": 1.6144,
      "step": 945
    },
    {
      "epoch": 0.08004061257297572,
      "grad_norm": 2.5,
      "learning_rate": 0.000985028770645847,
      "loss": 1.2468,
      "step": 946
    },
    {
      "epoch": 0.08012522210000846,
      "grad_norm": 1.875,
      "learning_rate": 0.0009849964058481712,
      "loss": 1.4379,
      "step": 947
    },
    {
      "epoch": 0.0802098316270412,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0009849640066380116,
      "loss": 1.1792,
      "step": 948
    },
    {
      "epoch": 0.08029444115407394,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009849315730176675,
      "loss": 1.2601,
      "step": 949
    },
    {
      "epoch": 0.08037905068110669,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009848991049894403,
      "loss": 1.5897,
      "step": 950
    },
    {
      "epoch": 0.08046366020813944,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009848666025556333,
      "loss": 1.3158,
      "step": 951
    },
    {
      "epoch": 0.08054826973517218,
      "grad_norm": 1.75,
      "learning_rate": 0.0009848340657185527,
      "loss": 1.5459,
      "step": 952
    },
    {
      "epoch": 0.08063287926220493,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009848014944805077,
      "loss": 1.8678,
      "step": 953
    },
    {
      "epoch": 0.08071748878923767,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009847688888438087,
      "loss": 1.9256,
      "step": 954
    },
    {
      "epoch": 0.08080209831627042,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009847362488107697,
      "loss": 1.6365,
      "step": 955
    },
    {
      "epoch": 0.08088670784330315,
      "grad_norm": 2.5,
      "learning_rate": 0.0009847035743837061,
      "loss": 2.1988,
      "step": 956
    },
    {
      "epoch": 0.0809713173703359,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009846708655649368,
      "loss": 1.7661,
      "step": 957
    },
    {
      "epoch": 0.08105592689736864,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009846381223567824,
      "loss": 1.6287,
      "step": 958
    },
    {
      "epoch": 0.08114053642440139,
      "grad_norm": 1.7734375,
      "learning_rate": 0.000984605344761566,
      "loss": 1.3312,
      "step": 959
    },
    {
      "epoch": 0.08122514595143414,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009845725327816137,
      "loss": 1.4061,
      "step": 960
    },
    {
      "epoch": 0.08130975547846687,
      "grad_norm": 3.625,
      "learning_rate": 0.0009845396864192532,
      "loss": 1.923,
      "step": 961
    },
    {
      "epoch": 0.08139436500549962,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009845068056768156,
      "loss": 2.2527,
      "step": 962
    },
    {
      "epoch": 0.08147897453253236,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009844738905566335,
      "loss": 2.0868,
      "step": 963
    },
    {
      "epoch": 0.08156358405956511,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009844409410610427,
      "loss": 1.3897,
      "step": 964
    },
    {
      "epoch": 0.08164819358659785,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009844079571923808,
      "loss": 1.3911,
      "step": 965
    },
    {
      "epoch": 0.0817328031136306,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009843749389529882,
      "loss": 1.8,
      "step": 966
    },
    {
      "epoch": 0.08181741264066333,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009843418863452076,
      "loss": 1.6755,
      "step": 967
    },
    {
      "epoch": 0.08190202216769608,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009843087993713843,
      "loss": 1.3793,
      "step": 968
    },
    {
      "epoch": 0.08198663169472882,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009842756780338662,
      "loss": 2.2142,
      "step": 969
    },
    {
      "epoch": 0.08207124122176157,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009842425223350031,
      "loss": 1.5566,
      "step": 970
    },
    {
      "epoch": 0.08215585074879432,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0009842093322771478,
      "loss": 1.2594,
      "step": 971
    },
    {
      "epoch": 0.08224046027582706,
      "grad_norm": 1.8828125,
      "learning_rate": 0.000984176107862655,
      "loss": 1.4557,
      "step": 972
    },
    {
      "epoch": 0.08232506980285981,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009841428490938821,
      "loss": 1.5143,
      "step": 973
    },
    {
      "epoch": 0.08240967932989254,
      "grad_norm": 1.96875,
      "learning_rate": 0.000984109555973189,
      "loss": 1.453,
      "step": 974
    },
    {
      "epoch": 0.0824942888569253,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009840762285029385,
      "loss": 1.7673,
      "step": 975
    },
    {
      "epoch": 0.08257889838395803,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009840428666854945,
      "loss": 1.4655,
      "step": 976
    },
    {
      "epoch": 0.08266350791099078,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009840094705232244,
      "loss": 1.4066,
      "step": 977
    },
    {
      "epoch": 0.08274811743802352,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009839760400184982,
      "loss": 1.7262,
      "step": 978
    },
    {
      "epoch": 0.08283272696505627,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009839425751736876,
      "loss": 1.7313,
      "step": 979
    },
    {
      "epoch": 0.082917336492089,
      "grad_norm": 1.640625,
      "learning_rate": 0.000983909075991167,
      "loss": 1.2565,
      "step": 980
    },
    {
      "epoch": 0.08300194601912175,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009838755424733134,
      "loss": 1.6774,
      "step": 981
    },
    {
      "epoch": 0.0830865555461545,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009838419746225062,
      "loss": 2.0531,
      "step": 982
    },
    {
      "epoch": 0.08317116507318724,
      "grad_norm": 1.421875,
      "learning_rate": 0.000983808372441127,
      "loss": 1.2864,
      "step": 983
    },
    {
      "epoch": 0.08325577460021999,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0009837747359315602,
      "loss": 1.2051,
      "step": 984
    },
    {
      "epoch": 0.08334038412725273,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009837410650961928,
      "loss": 1.4152,
      "step": 985
    },
    {
      "epoch": 0.08342499365428548,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009837073599374129,
      "loss": 1.3432,
      "step": 986
    },
    {
      "epoch": 0.08350960318131821,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009836736204576128,
      "loss": 1.134,
      "step": 987
    },
    {
      "epoch": 0.08359421270835096,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0009836398466591862,
      "loss": 1.2807,
      "step": 988
    },
    {
      "epoch": 0.0836788222353837,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009836060385445297,
      "loss": 1.7219,
      "step": 989
    },
    {
      "epoch": 0.08376343176241645,
      "grad_norm": 2.078125,
      "learning_rate": 0.000983572196116042,
      "loss": 1.4981,
      "step": 990
    },
    {
      "epoch": 0.0838480412894492,
      "grad_norm": 1.7890625,
      "learning_rate": 0.000983538319376124,
      "loss": 1.4966,
      "step": 991
    },
    {
      "epoch": 0.08393265081648194,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00098350440832718,
      "loss": 1.1321,
      "step": 992
    },
    {
      "epoch": 0.08401726034351469,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009834704629716159,
      "loss": 1.5848,
      "step": 993
    },
    {
      "epoch": 0.08410186987054742,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009834364833118398,
      "loss": 1.4115,
      "step": 994
    },
    {
      "epoch": 0.08418647939758017,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009834024693502635,
      "loss": 1.1149,
      "step": 995
    },
    {
      "epoch": 0.08427108892461291,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009833684210893,
      "loss": 1.2468,
      "step": 996
    },
    {
      "epoch": 0.08435569845164566,
      "grad_norm": 2.65625,
      "learning_rate": 0.000983334338531365,
      "loss": 1.7233,
      "step": 997
    },
    {
      "epoch": 0.0844403079786784,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009833002216788772,
      "loss": 1.279,
      "step": 998
    },
    {
      "epoch": 0.08452491750571114,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009832660705342568,
      "loss": 1.492,
      "step": 999
    },
    {
      "epoch": 0.08460952703274388,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009832318850999276,
      "loss": 1.5952,
      "step": 1000
    },
    {
      "epoch": 0.08469413655977663,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009831976653783148,
      "loss": 2.1617,
      "step": 1001
    },
    {
      "epoch": 0.08477874608680938,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009831634113718465,
      "loss": 1.8623,
      "step": 1002
    },
    {
      "epoch": 0.08486335561384212,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009831291230829532,
      "loss": 1.2179,
      "step": 1003
    },
    {
      "epoch": 0.08494796514087487,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009830948005140677,
      "loss": 2.7212,
      "step": 1004
    },
    {
      "epoch": 0.0850325746679076,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009830604436676254,
      "loss": 1.3704,
      "step": 1005
    },
    {
      "epoch": 0.08511718419494035,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009830260525460642,
      "loss": 1.4233,
      "step": 1006
    },
    {
      "epoch": 0.08520179372197309,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000982991627151824,
      "loss": 1.2145,
      "step": 1007
    },
    {
      "epoch": 0.08528640324900584,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009829571674873476,
      "loss": 1.7418,
      "step": 1008
    },
    {
      "epoch": 0.08537101277603858,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00098292267355508,
      "loss": 1.3812,
      "step": 1009
    },
    {
      "epoch": 0.08545562230307133,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009828881453574688,
      "loss": 1.4297,
      "step": 1010
    },
    {
      "epoch": 0.08554023183010406,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009828535828969639,
      "loss": 1.407,
      "step": 1011
    },
    {
      "epoch": 0.08562484135713681,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009828189861760175,
      "loss": 2.0553,
      "step": 1012
    },
    {
      "epoch": 0.08570945088416956,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009827843551970844,
      "loss": 1.5334,
      "step": 1013
    },
    {
      "epoch": 0.0857940604112023,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009827496899626218,
      "loss": 1.6022,
      "step": 1014
    },
    {
      "epoch": 0.08587866993823505,
      "grad_norm": 37.25,
      "learning_rate": 0.0009827149904750895,
      "loss": 1.5598,
      "step": 1015
    },
    {
      "epoch": 0.08596327946526779,
      "grad_norm": 12416.0,
      "learning_rate": 0.0009826802567369495,
      "loss": 2.3789,
      "step": 1016
    },
    {
      "epoch": 0.08604788899230054,
      "grad_norm": 24832.0,
      "learning_rate": 0.0009826454887506661,
      "loss": 1.453,
      "step": 1017
    },
    {
      "epoch": 0.08613249851933327,
      "grad_norm": 1520.0,
      "learning_rate": 0.0009826106865187068,
      "loss": 2.0715,
      "step": 1018
    },
    {
      "epoch": 0.08621710804636602,
      "grad_norm": 268288.0,
      "learning_rate": 0.0009825758500435402,
      "loss": 1.5979,
      "step": 1019
    },
    {
      "epoch": 0.08630171757339876,
      "grad_norm": 724.0,
      "learning_rate": 0.0009825409793276386,
      "loss": 2.3891,
      "step": 1020
    },
    {
      "epoch": 0.08638632710043151,
      "grad_norm": 1032.0,
      "learning_rate": 0.0009825060743734759,
      "loss": 1.74,
      "step": 1021
    },
    {
      "epoch": 0.08647093662746426,
      "grad_norm": 532.0,
      "learning_rate": 0.0009824711351835293,
      "loss": 1.8621,
      "step": 1022
    },
    {
      "epoch": 0.086555546154497,
      "grad_norm": 41472.0,
      "learning_rate": 0.0009824361617602772,
      "loss": 2.0616,
      "step": 1023
    },
    {
      "epoch": 0.08664015568152975,
      "grad_norm": 5632.0,
      "learning_rate": 0.0009824011541062015,
      "loss": 2.4814,
      "step": 1024
    },
    {
      "epoch": 0.08672476520856248,
      "grad_norm": 1728.0,
      "learning_rate": 0.0009823661122237862,
      "loss": 2.1939,
      "step": 1025
    },
    {
      "epoch": 0.08680937473559523,
      "grad_norm": 230.0,
      "learning_rate": 0.0009823310361155174,
      "loss": 2.3367,
      "step": 1026
    },
    {
      "epoch": 0.08689398426262797,
      "grad_norm": 4672.0,
      "learning_rate": 0.000982295925783884,
      "loss": 1.7879,
      "step": 1027
    },
    {
      "epoch": 0.08697859378966072,
      "grad_norm": 258.0,
      "learning_rate": 0.000982260781231377,
      "loss": 2.3785,
      "step": 1028
    },
    {
      "epoch": 0.08706320331669345,
      "grad_norm": 6080.0,
      "learning_rate": 0.0009822256024604908,
      "loss": 1.9995,
      "step": 1029
    },
    {
      "epoch": 0.0871478128437262,
      "grad_norm": 209.0,
      "learning_rate": 0.0009821903894737206,
      "loss": 2.6536,
      "step": 1030
    },
    {
      "epoch": 0.08723242237075894,
      "grad_norm": 416.0,
      "learning_rate": 0.0009821551422735653,
      "loss": 2.2265,
      "step": 1031
    },
    {
      "epoch": 0.08731703189779169,
      "grad_norm": 354.0,
      "learning_rate": 0.0009821198608625258,
      "loss": 2.0657,
      "step": 1032
    },
    {
      "epoch": 0.08740164142482444,
      "grad_norm": 6.875,
      "learning_rate": 0.0009820845452431057,
      "loss": 2.2475,
      "step": 1033
    },
    {
      "epoch": 0.08748625095185718,
      "grad_norm": 27.875,
      "learning_rate": 0.0009820491954178104,
      "loss": 2.2301,
      "step": 1034
    },
    {
      "epoch": 0.08757086047888993,
      "grad_norm": 28.0,
      "learning_rate": 0.000982013811389148,
      "loss": 1.9133,
      "step": 1035
    },
    {
      "epoch": 0.08765547000592266,
      "grad_norm": 3.8125,
      "learning_rate": 0.00098197839315963,
      "loss": 1.8696,
      "step": 1036
    },
    {
      "epoch": 0.08774007953295541,
      "grad_norm": 3.703125,
      "learning_rate": 0.0009819429407317683,
      "loss": 2.1083,
      "step": 1037
    },
    {
      "epoch": 0.08782468905998815,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009819074541080792,
      "loss": 1.8446,
      "step": 1038
    },
    {
      "epoch": 0.0879092985870209,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009818719332910805,
      "loss": 1.4616,
      "step": 1039
    },
    {
      "epoch": 0.08799390811405364,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009818363782832923,
      "loss": 1.2599,
      "step": 1040
    },
    {
      "epoch": 0.08807851764108639,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009818007890872375,
      "loss": 1.5326,
      "step": 1041
    },
    {
      "epoch": 0.08816312716811912,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009817651657054415,
      "loss": 2.1217,
      "step": 1042
    },
    {
      "epoch": 0.08824773669515187,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009817295081404316,
      "loss": 1.3865,
      "step": 1043
    },
    {
      "epoch": 0.08833234622218462,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000981693816394738,
      "loss": 1.0872,
      "step": 1044
    },
    {
      "epoch": 0.08841695574921736,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009816580904708932,
      "loss": 1.2501,
      "step": 1045
    },
    {
      "epoch": 0.08850156527625011,
      "grad_norm": 2.25,
      "learning_rate": 0.000981622330371432,
      "loss": 1.747,
      "step": 1046
    },
    {
      "epoch": 0.08858617480328285,
      "grad_norm": 2.375,
      "learning_rate": 0.0009815865360988918,
      "loss": 1.8884,
      "step": 1047
    },
    {
      "epoch": 0.0886707843303156,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009815507076558123,
      "loss": 1.4927,
      "step": 1048
    },
    {
      "epoch": 0.08875539385734833,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009815148450447357,
      "loss": 1.63,
      "step": 1049
    },
    {
      "epoch": 0.08884000338438108,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009814789482682068,
      "loss": 1.4192,
      "step": 1050
    },
    {
      "epoch": 0.08892461291141382,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009814430173287723,
      "loss": 1.3324,
      "step": 1051
    },
    {
      "epoch": 0.08900922243844657,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009814070522289819,
      "loss": 1.3373,
      "step": 1052
    },
    {
      "epoch": 0.08909383196547932,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009813710529713872,
      "loss": 1.2407,
      "step": 1053
    },
    {
      "epoch": 0.08917844149251206,
      "grad_norm": 5.3125,
      "learning_rate": 0.0009813350195585428,
      "loss": 1.6743,
      "step": 1054
    },
    {
      "epoch": 0.0892630510195448,
      "grad_norm": 7.9375,
      "learning_rate": 0.000981298951993005,
      "loss": 1.708,
      "step": 1055
    },
    {
      "epoch": 0.08934766054657754,
      "grad_norm": 3.5625,
      "learning_rate": 0.0009812628502773335,
      "loss": 1.9571,
      "step": 1056
    },
    {
      "epoch": 0.08943227007361029,
      "grad_norm": 4.71875,
      "learning_rate": 0.0009812267144140897,
      "loss": 1.6252,
      "step": 1057
    },
    {
      "epoch": 0.08951687960064303,
      "grad_norm": 3.71875,
      "learning_rate": 0.000981190544405837,
      "loss": 1.5166,
      "step": 1058
    },
    {
      "epoch": 0.08960148912767578,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009811543402551426,
      "loss": 1.2253,
      "step": 1059
    },
    {
      "epoch": 0.08968609865470852,
      "grad_norm": 8.9375,
      "learning_rate": 0.000981118101964575,
      "loss": 1.6067,
      "step": 1060
    },
    {
      "epoch": 0.08977070818174127,
      "grad_norm": 4.78125,
      "learning_rate": 0.0009810818295367054,
      "loss": 2.1231,
      "step": 1061
    },
    {
      "epoch": 0.089855317708774,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009810455229741076,
      "loss": 2.0207,
      "step": 1062
    },
    {
      "epoch": 0.08993992723580675,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009810091822793575,
      "loss": 1.3291,
      "step": 1063
    },
    {
      "epoch": 0.0900245367628395,
      "grad_norm": 2.140625,
      "learning_rate": 0.000980972807455034,
      "loss": 1.372,
      "step": 1064
    },
    {
      "epoch": 0.09010914628987224,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009809363985037176,
      "loss": 1.6328,
      "step": 1065
    },
    {
      "epoch": 0.09019375581690499,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0009808999554279919,
      "loss": 1.2573,
      "step": 1066
    },
    {
      "epoch": 0.09027836534393772,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009808634782304427,
      "loss": 1.8395,
      "step": 1067
    },
    {
      "epoch": 0.09036297487097047,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009808269669136583,
      "loss": 1.7144,
      "step": 1068
    },
    {
      "epoch": 0.09044758439800321,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009807904214802288,
      "loss": 1.6188,
      "step": 1069
    },
    {
      "epoch": 0.09053219392503596,
      "grad_norm": 2.125,
      "learning_rate": 0.000980753841932748,
      "loss": 1.5533,
      "step": 1070
    },
    {
      "epoch": 0.0906168034520687,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009807172282738108,
      "loss": 1.3053,
      "step": 1071
    },
    {
      "epoch": 0.09070141297910145,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0009806805805060155,
      "loss": 1.2844,
      "step": 1072
    },
    {
      "epoch": 0.09078602250613418,
      "grad_norm": 2.25,
      "learning_rate": 0.000980643898631962,
      "loss": 1.3895,
      "step": 1073
    },
    {
      "epoch": 0.09087063203316693,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009806071826542534,
      "loss": 2.4032,
      "step": 1074
    },
    {
      "epoch": 0.09095524156019968,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009805704325754946,
      "loss": 1.4699,
      "step": 1075
    },
    {
      "epoch": 0.09103985108723242,
      "grad_norm": 2.0,
      "learning_rate": 0.0009805336483982932,
      "loss": 1.5715,
      "step": 1076
    },
    {
      "epoch": 0.09112446061426517,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009804968301252593,
      "loss": 1.7278,
      "step": 1077
    },
    {
      "epoch": 0.09120907014129791,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009804599777590052,
      "loss": 1.8499,
      "step": 1078
    },
    {
      "epoch": 0.09129367966833066,
      "grad_norm": 1.9296875,
      "learning_rate": 0.000980423091302146,
      "loss": 1.4527,
      "step": 1079
    },
    {
      "epoch": 0.0913782891953634,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009803861707572985,
      "loss": 1.2488,
      "step": 1080
    },
    {
      "epoch": 0.09146289872239614,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009803492161270825,
      "loss": 1.1751,
      "step": 1081
    },
    {
      "epoch": 0.09154750824942888,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009803122274141202,
      "loss": 1.6812,
      "step": 1082
    },
    {
      "epoch": 0.09163211777646163,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009802752046210363,
      "loss": 1.3578,
      "step": 1083
    },
    {
      "epoch": 0.09171672730349438,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009802381477504573,
      "loss": 1.748,
      "step": 1084
    },
    {
      "epoch": 0.09180133683052712,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009802010568050127,
      "loss": 1.4377,
      "step": 1085
    },
    {
      "epoch": 0.09188594635755987,
      "grad_norm": 1.375,
      "learning_rate": 0.0009801639317873346,
      "loss": 1.1001,
      "step": 1086
    },
    {
      "epoch": 0.0919705558845926,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009801267727000565,
      "loss": 1.5289,
      "step": 1087
    },
    {
      "epoch": 0.09205516541162535,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009800895795458154,
      "loss": 1.2743,
      "step": 1088
    },
    {
      "epoch": 0.09213977493865809,
      "grad_norm": 2.0,
      "learning_rate": 0.0009800523523272504,
      "loss": 1.3311,
      "step": 1089
    },
    {
      "epoch": 0.09222438446569084,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009800150910470025,
      "loss": 1.9524,
      "step": 1090
    },
    {
      "epoch": 0.09230899399272358,
      "grad_norm": 2.59375,
      "learning_rate": 0.000979977795707716,
      "loss": 1.273,
      "step": 1091
    },
    {
      "epoch": 0.09239360351975633,
      "grad_norm": 2.78125,
      "learning_rate": 0.000979940466312037,
      "loss": 1.3509,
      "step": 1092
    },
    {
      "epoch": 0.09247821304678906,
      "grad_norm": 3.53125,
      "learning_rate": 0.000979903102862614,
      "loss": 2.1755,
      "step": 1093
    },
    {
      "epoch": 0.09256282257382181,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009798657053620983,
      "loss": 1.535,
      "step": 1094
    },
    {
      "epoch": 0.09264743210085456,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009798282738131435,
      "loss": 1.8625,
      "step": 1095
    },
    {
      "epoch": 0.0927320416278873,
      "grad_norm": 2.203125,
      "learning_rate": 0.000979790808218405,
      "loss": 1.3442,
      "step": 1096
    },
    {
      "epoch": 0.09281665115492005,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009797533085805419,
      "loss": 1.3641,
      "step": 1097
    },
    {
      "epoch": 0.09290126068195279,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009797157749022144,
      "loss": 1.4349,
      "step": 1098
    },
    {
      "epoch": 0.09298587020898554,
      "grad_norm": 2.5,
      "learning_rate": 0.0009796782071860857,
      "loss": 1.5218,
      "step": 1099
    },
    {
      "epoch": 0.09307047973601827,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009796406054348216,
      "loss": 1.777,
      "step": 1100
    },
    {
      "epoch": 0.09315508926305102,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009796029696510899,
      "loss": 1.616,
      "step": 1101
    },
    {
      "epoch": 0.09323969879008376,
      "grad_norm": 5.5,
      "learning_rate": 0.0009795652998375612,
      "loss": 1.91,
      "step": 1102
    },
    {
      "epoch": 0.09332430831711651,
      "grad_norm": 62.75,
      "learning_rate": 0.0009795275959969083,
      "loss": 1.4356,
      "step": 1103
    },
    {
      "epoch": 0.09340891784414924,
      "grad_norm": 4.125,
      "learning_rate": 0.0009794898581318063,
      "loss": 1.2968,
      "step": 1104
    },
    {
      "epoch": 0.093493527371182,
      "grad_norm": 2.34375,
      "learning_rate": 0.000979452086244933,
      "loss": 1.3711,
      "step": 1105
    },
    {
      "epoch": 0.09357813689821474,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009794142803389683,
      "loss": 1.4057,
      "step": 1106
    },
    {
      "epoch": 0.09366274642524748,
      "grad_norm": 33.75,
      "learning_rate": 0.0009793764404165948,
      "loss": 1.554,
      "step": 1107
    },
    {
      "epoch": 0.09374735595228023,
      "grad_norm": 42.5,
      "learning_rate": 0.0009793385664804974,
      "loss": 1.7888,
      "step": 1108
    },
    {
      "epoch": 0.09383196547931297,
      "grad_norm": 16.25,
      "learning_rate": 0.0009793006585333635,
      "loss": 2.6923,
      "step": 1109
    },
    {
      "epoch": 0.09391657500634572,
      "grad_norm": 65.0,
      "learning_rate": 0.0009792627165778826,
      "loss": 1.4619,
      "step": 1110
    },
    {
      "epoch": 0.09400118453337845,
      "grad_norm": 2.78125,
      "learning_rate": 0.000979224740616747,
      "loss": 1.7485,
      "step": 1111
    },
    {
      "epoch": 0.0940857940604112,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009791867306526514,
      "loss": 1.4483,
      "step": 1112
    },
    {
      "epoch": 0.09417040358744394,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009791486866882926,
      "loss": 1.5252,
      "step": 1113
    },
    {
      "epoch": 0.09425501311447669,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009791106087263697,
      "loss": 1.3683,
      "step": 1114
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009790724967695848,
      "loss": 1.6058,
      "step": 1115
    },
    {
      "epoch": 0.09442423216854218,
      "grad_norm": 2.625,
      "learning_rate": 0.000979034350820642,
      "loss": 1.7291,
      "step": 1116
    },
    {
      "epoch": 0.09450884169557493,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009789961708822482,
      "loss": 1.5204,
      "step": 1117
    },
    {
      "epoch": 0.09459345122260766,
      "grad_norm": 2.21875,
      "learning_rate": 0.000978957956957112,
      "loss": 1.8387,
      "step": 1118
    },
    {
      "epoch": 0.09467806074964041,
      "grad_norm": 2.109375,
      "learning_rate": 0.000978919709047945,
      "loss": 1.9336,
      "step": 1119
    },
    {
      "epoch": 0.09476267027667315,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009788814271574611,
      "loss": 1.4482,
      "step": 1120
    },
    {
      "epoch": 0.0948472798037059,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009788431112883764,
      "loss": 1.6606,
      "step": 1121
    },
    {
      "epoch": 0.09493188933073864,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009788047614434098,
      "loss": 1.5378,
      "step": 1122
    },
    {
      "epoch": 0.09501649885777139,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009787663776252822,
      "loss": 1.5033,
      "step": 1123
    },
    {
      "epoch": 0.09510110838480412,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000978727959836717,
      "loss": 1.2545,
      "step": 1124
    },
    {
      "epoch": 0.09518571791183687,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009786895080804406,
      "loss": 2.2442,
      "step": 1125
    },
    {
      "epoch": 0.09527032743886962,
      "grad_norm": 1.765625,
      "learning_rate": 0.0009786510223591808,
      "loss": 1.2678,
      "step": 1126
    },
    {
      "epoch": 0.09535493696590236,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009786125026756685,
      "loss": 1.4403,
      "step": 1127
    },
    {
      "epoch": 0.09543954649293511,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009785739490326367,
      "loss": 1.2975,
      "step": 1128
    },
    {
      "epoch": 0.09552415601996785,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009785353614328212,
      "loss": 2.0455,
      "step": 1129
    },
    {
      "epoch": 0.0956087655470006,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009784967398789597,
      "loss": 1.8203,
      "step": 1130
    },
    {
      "epoch": 0.09569337507403333,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009784580843737924,
      "loss": 1.4559,
      "step": 1131
    },
    {
      "epoch": 0.09577798460106608,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009784193949200627,
      "loss": 1.647,
      "step": 1132
    },
    {
      "epoch": 0.09586259412809882,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009783806715205152,
      "loss": 1.4784,
      "step": 1133
    },
    {
      "epoch": 0.09594720365513157,
      "grad_norm": 2.375,
      "learning_rate": 0.0009783419141778979,
      "loss": 1.7829,
      "step": 1134
    },
    {
      "epoch": 0.0960318131821643,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009783031228949603,
      "loss": 1.3547,
      "step": 1135
    },
    {
      "epoch": 0.09611642270919706,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009782642976744551,
      "loss": 1.7349,
      "step": 1136
    },
    {
      "epoch": 0.0962010322362298,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009782254385191373,
      "loss": 1.7968,
      "step": 1137
    },
    {
      "epoch": 0.09628564176326254,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009781865454317638,
      "loss": 1.6307,
      "step": 1138
    },
    {
      "epoch": 0.09637025129029529,
      "grad_norm": 2.140625,
      "learning_rate": 0.000978147618415094,
      "loss": 1.742,
      "step": 1139
    },
    {
      "epoch": 0.09645486081732803,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009781086574718906,
      "loss": 1.7787,
      "step": 1140
    },
    {
      "epoch": 0.09653947034436078,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009780696626049178,
      "loss": 1.4011,
      "step": 1141
    },
    {
      "epoch": 0.09662407987139351,
      "grad_norm": 1.8671875,
      "learning_rate": 0.000978030633816942,
      "loss": 1.3218,
      "step": 1142
    },
    {
      "epoch": 0.09670868939842626,
      "grad_norm": 2.015625,
      "learning_rate": 0.000977991571110733,
      "loss": 1.4626,
      "step": 1143
    },
    {
      "epoch": 0.096793298925459,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009779524744890622,
      "loss": 1.7715,
      "step": 1144
    },
    {
      "epoch": 0.09687790845249175,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009779133439547037,
      "loss": 1.9679,
      "step": 1145
    },
    {
      "epoch": 0.0969625179795245,
      "grad_norm": 1.703125,
      "learning_rate": 0.000977874179510434,
      "loss": 1.3441,
      "step": 1146
    },
    {
      "epoch": 0.09704712750655724,
      "grad_norm": 2.25,
      "learning_rate": 0.000977834981159032,
      "loss": 1.8169,
      "step": 1147
    },
    {
      "epoch": 0.09713173703358999,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009777957489032789,
      "loss": 2.0271,
      "step": 1148
    },
    {
      "epoch": 0.09721634656062272,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009777564827459584,
      "loss": 1.3061,
      "step": 1149
    },
    {
      "epoch": 0.09730095608765547,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009777171826898566,
      "loss": 1.3582,
      "step": 1150
    },
    {
      "epoch": 0.09738556561468821,
      "grad_norm": 2.625,
      "learning_rate": 0.000977677848737762,
      "loss": 1.9084,
      "step": 1151
    },
    {
      "epoch": 0.09747017514172096,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009776384808924657,
      "loss": 1.3422,
      "step": 1152
    },
    {
      "epoch": 0.0975547846687537,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009775990791567608,
      "loss": 1.7844,
      "step": 1153
    },
    {
      "epoch": 0.09763939419578645,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009775596435334431,
      "loss": 1.7936,
      "step": 1154
    },
    {
      "epoch": 0.09772400372281918,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009775201740253106,
      "loss": 1.4572,
      "step": 1155
    },
    {
      "epoch": 0.09780861324985193,
      "grad_norm": 2.109375,
      "learning_rate": 0.000977480670635164,
      "loss": 1.4436,
      "step": 1156
    },
    {
      "epoch": 0.09789322277688468,
      "grad_norm": 13.25,
      "learning_rate": 0.000977441133365806,
      "loss": 1.9157,
      "step": 1157
    },
    {
      "epoch": 0.09797783230391742,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009774015622200421,
      "loss": 1.58,
      "step": 1158
    },
    {
      "epoch": 0.09806244183095017,
      "grad_norm": 3.484375,
      "learning_rate": 0.00097736195720068,
      "loss": 1.1432,
      "step": 1159
    },
    {
      "epoch": 0.0981470513579829,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009773223183105297,
      "loss": 1.3685,
      "step": 1160
    },
    {
      "epoch": 0.09823166088501566,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009772826455524042,
      "loss": 2.6012,
      "step": 1161
    },
    {
      "epoch": 0.09831627041204839,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000977242938929118,
      "loss": 1.5334,
      "step": 1162
    },
    {
      "epoch": 0.09840087993908114,
      "grad_norm": 1.234375,
      "learning_rate": 0.0009772031984434885,
      "loss": 1.0355,
      "step": 1163
    },
    {
      "epoch": 0.09848548946611388,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009771634240983353,
      "loss": 1.378,
      "step": 1164
    },
    {
      "epoch": 0.09857009899314663,
      "grad_norm": 1.21875,
      "learning_rate": 0.000977123615896481,
      "loss": 1.0549,
      "step": 1165
    },
    {
      "epoch": 0.09865470852017937,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00097708377384075,
      "loss": 1.7946,
      "step": 1166
    },
    {
      "epoch": 0.09873931804721212,
      "grad_norm": 20.625,
      "learning_rate": 0.0009770438979339693,
      "loss": 1.2534,
      "step": 1167
    },
    {
      "epoch": 0.09882392757424487,
      "grad_norm": 27.25,
      "learning_rate": 0.0009770039881789678,
      "loss": 1.6131,
      "step": 1168
    },
    {
      "epoch": 0.0989085371012776,
      "grad_norm": 161.0,
      "learning_rate": 0.000976964044578578,
      "loss": 1.7596,
      "step": 1169
    },
    {
      "epoch": 0.09899314662831035,
      "grad_norm": 222.0,
      "learning_rate": 0.000976924067135633,
      "loss": 1.6324,
      "step": 1170
    },
    {
      "epoch": 0.09907775615534309,
      "grad_norm": 212.0,
      "learning_rate": 0.0009768840558529708,
      "loss": 1.6193,
      "step": 1171
    },
    {
      "epoch": 0.09916236568237584,
      "grad_norm": 36.25,
      "learning_rate": 0.000976844010733429,
      "loss": 1.4593,
      "step": 1172
    },
    {
      "epoch": 0.09924697520940857,
      "grad_norm": 4.21875,
      "learning_rate": 0.0009768039317798502,
      "loss": 1.7706,
      "step": 1173
    },
    {
      "epoch": 0.09933158473644133,
      "grad_norm": 7.03125,
      "learning_rate": 0.000976763818995077,
      "loss": 1.3412,
      "step": 1174
    },
    {
      "epoch": 0.09941619426347406,
      "grad_norm": 4.5,
      "learning_rate": 0.0009767236723819562,
      "loss": 2.4674,
      "step": 1175
    },
    {
      "epoch": 0.09950080379050681,
      "grad_norm": 2.75,
      "learning_rate": 0.0009766834919433362,
      "loss": 2.0204,
      "step": 1176
    },
    {
      "epoch": 0.09958541331753956,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009766432776820682,
      "loss": 1.6212,
      "step": 1177
    },
    {
      "epoch": 0.0996700228445723,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009766030296010054,
      "loss": 1.5863,
      "step": 1178
    },
    {
      "epoch": 0.09975463237160505,
      "grad_norm": 5.03125,
      "learning_rate": 0.0009765627477030035,
      "loss": 1.8125,
      "step": 1179
    },
    {
      "epoch": 0.09983924189863778,
      "grad_norm": 5.21875,
      "learning_rate": 0.0009765224319909207,
      "loss": 1.5478,
      "step": 1180
    },
    {
      "epoch": 0.09992385142567053,
      "grad_norm": 3.140625,
      "learning_rate": 0.0009764820824676176,
      "loss": 1.7456,
      "step": 1181
    },
    {
      "epoch": 0.10000846095270327,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009764416991359571,
      "loss": 1.5472,
      "step": 1182
    },
    {
      "epoch": 0.10009307047973602,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009764012819988047,
      "loss": 1.5653,
      "step": 1183
    },
    {
      "epoch": 0.10017768000676876,
      "grad_norm": 1.953125,
      "learning_rate": 0.000976360831059028,
      "loss": 1.1331,
      "step": 1184
    },
    {
      "epoch": 0.10026228953380151,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009763203463194974,
      "loss": 1.5689,
      "step": 1185
    },
    {
      "epoch": 0.10034689906083424,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009762798277830851,
      "loss": 1.5637,
      "step": 1186
    },
    {
      "epoch": 0.100431508587867,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009762392754526665,
      "loss": 1.3388,
      "step": 1187
    },
    {
      "epoch": 0.10051611811489974,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009761986893311185,
      "loss": 1.6358,
      "step": 1188
    },
    {
      "epoch": 0.10060072764193248,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009761580694213211,
      "loss": 1.9164,
      "step": 1189
    },
    {
      "epoch": 0.10068533716896523,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009761174157261564,
      "loss": 1.2581,
      "step": 1190
    },
    {
      "epoch": 0.10076994669599797,
      "grad_norm": 1.9921875,
      "learning_rate": 0.000976076728248509,
      "loss": 1.7589,
      "step": 1191
    },
    {
      "epoch": 0.10085455622303072,
      "grad_norm": 8.875,
      "learning_rate": 0.0009760360069912659,
      "loss": 1.4762,
      "step": 1192
    },
    {
      "epoch": 0.10093916575006345,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009759952519573161,
      "loss": 1.2083,
      "step": 1193
    },
    {
      "epoch": 0.1010237752770962,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009759544631495518,
      "loss": 1.7694,
      "step": 1194
    },
    {
      "epoch": 0.10110838480412894,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009759136405708669,
      "loss": 1.7721,
      "step": 1195
    },
    {
      "epoch": 0.10119299433116169,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009758727842241578,
      "loss": 1.4142,
      "step": 1196
    },
    {
      "epoch": 0.10127760385819443,
      "grad_norm": 3.90625,
      "learning_rate": 0.0009758318941123235,
      "loss": 1.8119,
      "step": 1197
    },
    {
      "epoch": 0.10136221338522718,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009757909702382654,
      "loss": 1.4722,
      "step": 1198
    },
    {
      "epoch": 0.10144682291225993,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009757500126048875,
      "loss": 1.9893,
      "step": 1199
    },
    {
      "epoch": 0.10153143243929266,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009757090212150954,
      "loss": 1.6535,
      "step": 1200
    },
    {
      "epoch": 0.10161604196632541,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009756679960717978,
      "loss": 1.5505,
      "step": 1201
    },
    {
      "epoch": 0.10170065149335815,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009756269371779056,
      "loss": 1.2525,
      "step": 1202
    },
    {
      "epoch": 0.1017852610203909,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009755858445363321,
      "loss": 1.2738,
      "step": 1203
    },
    {
      "epoch": 0.10186987054742364,
      "grad_norm": 7.9375,
      "learning_rate": 0.0009755447181499932,
      "loss": 2.0184,
      "step": 1204
    },
    {
      "epoch": 0.10195448007445639,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009755035580218067,
      "loss": 1.5274,
      "step": 1205
    },
    {
      "epoch": 0.10203908960148912,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009754623641546931,
      "loss": 1.43,
      "step": 1206
    },
    {
      "epoch": 0.10212369912852187,
      "grad_norm": 4.625,
      "learning_rate": 0.0009754211365515755,
      "loss": 1.9757,
      "step": 1207
    },
    {
      "epoch": 0.10220830865555462,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009753798752153789,
      "loss": 1.4827,
      "step": 1208
    },
    {
      "epoch": 0.10229291818258736,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009753385801490311,
      "loss": 1.4773,
      "step": 1209
    },
    {
      "epoch": 0.10237752770962011,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009752972513554621,
      "loss": 1.0962,
      "step": 1210
    },
    {
      "epoch": 0.10246213723665284,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009752558888376044,
      "loss": 1.6157,
      "step": 1211
    },
    {
      "epoch": 0.1025467467636856,
      "grad_norm": 3.03125,
      "learning_rate": 0.000975214492598393,
      "loss": 2.0855,
      "step": 1212
    },
    {
      "epoch": 0.10263135629071833,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009751730626407646,
      "loss": 1.7336,
      "step": 1213
    },
    {
      "epoch": 0.10271596581775108,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009751315989676595,
      "loss": 1.2921,
      "step": 1214
    },
    {
      "epoch": 0.10280057534478382,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009750901015820192,
      "loss": 1.7493,
      "step": 1215
    },
    {
      "epoch": 0.10288518487181657,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009750485704867884,
      "loss": 1.5447,
      "step": 1216
    },
    {
      "epoch": 0.1029697943988493,
      "grad_norm": 1.625,
      "learning_rate": 0.0009750070056849139,
      "loss": 1.2892,
      "step": 1217
    },
    {
      "epoch": 0.10305440392588205,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009749654071793445,
      "loss": 1.7133,
      "step": 1218
    },
    {
      "epoch": 0.1031390134529148,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009749237749730324,
      "loss": 1.5131,
      "step": 1219
    },
    {
      "epoch": 0.10322362297994754,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009748821090689311,
      "loss": 1.7161,
      "step": 1220
    },
    {
      "epoch": 0.10330823250698029,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009748404094699972,
      "loss": 1.1276,
      "step": 1221
    },
    {
      "epoch": 0.10339284203401303,
      "grad_norm": 1.875,
      "learning_rate": 0.0009747986761791894,
      "loss": 1.2483,
      "step": 1222
    },
    {
      "epoch": 0.10347745156104578,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009747569091994688,
      "loss": 1.5523,
      "step": 1223
    },
    {
      "epoch": 0.10356206108807851,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009747151085337988,
      "loss": 1.4974,
      "step": 1224
    },
    {
      "epoch": 0.10364667061511126,
      "grad_norm": 3.78125,
      "learning_rate": 0.0009746732741851458,
      "loss": 1.578,
      "step": 1225
    },
    {
      "epoch": 0.103731280142144,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009746314061564776,
      "loss": 1.4485,
      "step": 1226
    },
    {
      "epoch": 0.10381588966917675,
      "grad_norm": 3.0,
      "learning_rate": 0.0009745895044507653,
      "loss": 1.6686,
      "step": 1227
    },
    {
      "epoch": 0.10390049919620949,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009745475690709817,
      "loss": 2.2154,
      "step": 1228
    },
    {
      "epoch": 0.10398510872324224,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009745056000201025,
      "loss": 1.263,
      "step": 1229
    },
    {
      "epoch": 0.10406971825027499,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009744635973011053,
      "loss": 1.6155,
      "step": 1230
    },
    {
      "epoch": 0.10415432777730772,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009744215609169708,
      "loss": 1.6001,
      "step": 1231
    },
    {
      "epoch": 0.10423893730434047,
      "grad_norm": 3.75,
      "learning_rate": 0.0009743794908706813,
      "loss": 1.5411,
      "step": 1232
    },
    {
      "epoch": 0.10432354683137321,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009743373871652219,
      "loss": 1.705,
      "step": 1233
    },
    {
      "epoch": 0.10440815635840596,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009742952498035802,
      "loss": 1.4604,
      "step": 1234
    },
    {
      "epoch": 0.1044927658854387,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009742530787887458,
      "loss": 1.7508,
      "step": 1235
    },
    {
      "epoch": 0.10457737541247145,
      "grad_norm": 2.4375,
      "learning_rate": 0.000974210874123711,
      "loss": 1.531,
      "step": 1236
    },
    {
      "epoch": 0.10466198493950418,
      "grad_norm": 12.4375,
      "learning_rate": 0.0009741686358114704,
      "loss": 1.711,
      "step": 1237
    },
    {
      "epoch": 0.10474659446653693,
      "grad_norm": 2.0,
      "learning_rate": 0.0009741263638550211,
      "loss": 1.2051,
      "step": 1238
    },
    {
      "epoch": 0.10483120399356968,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009740840582573622,
      "loss": 1.6329,
      "step": 1239
    },
    {
      "epoch": 0.10491581352060242,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009740417190214956,
      "loss": 1.3478,
      "step": 1240
    },
    {
      "epoch": 0.10500042304763517,
      "grad_norm": 3.5625,
      "learning_rate": 0.0009739993461504255,
      "loss": 2.0825,
      "step": 1241
    },
    {
      "epoch": 0.1050850325746679,
      "grad_norm": 3.484375,
      "learning_rate": 0.0009739569396471584,
      "loss": 1.5619,
      "step": 1242
    },
    {
      "epoch": 0.10516964210170066,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009739144995147032,
      "loss": 1.6857,
      "step": 1243
    },
    {
      "epoch": 0.10525425162873339,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009738720257560712,
      "loss": 2.055,
      "step": 1244
    },
    {
      "epoch": 0.10533886115576614,
      "grad_norm": 3.75,
      "learning_rate": 0.0009738295183742761,
      "loss": 1.5064,
      "step": 1245
    },
    {
      "epoch": 0.10542347068279888,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009737869773723338,
      "loss": 1.6104,
      "step": 1246
    },
    {
      "epoch": 0.10550808020983163,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009737444027532632,
      "loss": 1.384,
      "step": 1247
    },
    {
      "epoch": 0.10559268973686436,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009737017945200846,
      "loss": 1.2434,
      "step": 1248
    },
    {
      "epoch": 0.10567729926389711,
      "grad_norm": 10.5,
      "learning_rate": 0.0009736591526758215,
      "loss": 1.3119,
      "step": 1249
    },
    {
      "epoch": 0.10576190879092986,
      "grad_norm": 4.46875,
      "learning_rate": 0.0009736164772234996,
      "loss": 1.5795,
      "step": 1250
    },
    {
      "epoch": 0.1058465183179626,
      "grad_norm": 2.125,
      "learning_rate": 0.0009735737681661467,
      "loss": 1.509,
      "step": 1251
    },
    {
      "epoch": 0.10593112784499535,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009735310255067933,
      "loss": 1.8188,
      "step": 1252
    },
    {
      "epoch": 0.10601573737202809,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009734882492484722,
      "loss": 1.4994,
      "step": 1253
    },
    {
      "epoch": 0.10610034689906084,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009734454393942185,
      "loss": 1.4028,
      "step": 1254
    },
    {
      "epoch": 0.10618495642609357,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009734025959470697,
      "loss": 1.4703,
      "step": 1255
    },
    {
      "epoch": 0.10626956595312632,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009733597189100658,
      "loss": 1.5421,
      "step": 1256
    },
    {
      "epoch": 0.10635417548015906,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009733168082862489,
      "loss": 1.2942,
      "step": 1257
    },
    {
      "epoch": 0.10643878500719181,
      "grad_norm": 60.25,
      "learning_rate": 0.0009732738640786641,
      "loss": 1.1292,
      "step": 1258
    },
    {
      "epoch": 0.10652339453422455,
      "grad_norm": 7.40625,
      "learning_rate": 0.000973230886290358,
      "loss": 1.8368,
      "step": 1259
    },
    {
      "epoch": 0.1066080040612573,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009731878749243804,
      "loss": 1.3996,
      "step": 1260
    },
    {
      "epoch": 0.10669261358829005,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009731448299837829,
      "loss": 1.2125,
      "step": 1261
    },
    {
      "epoch": 0.10677722311532278,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009731017514716196,
      "loss": 2.307,
      "step": 1262
    },
    {
      "epoch": 0.10686183264235553,
      "grad_norm": 3.0,
      "learning_rate": 0.0009730586393909475,
      "loss": 1.9541,
      "step": 1263
    },
    {
      "epoch": 0.10694644216938827,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009730154937448254,
      "loss": 1.3867,
      "step": 1264
    },
    {
      "epoch": 0.10703105169642102,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009729723145363146,
      "loss": 1.5045,
      "step": 1265
    },
    {
      "epoch": 0.10711566122345376,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009729291017684789,
      "loss": 1.5067,
      "step": 1266
    },
    {
      "epoch": 0.1072002707504865,
      "grad_norm": 3.75,
      "learning_rate": 0.0009728858554443844,
      "loss": 1.6486,
      "step": 1267
    },
    {
      "epoch": 0.10728488027751924,
      "grad_norm": 1.875,
      "learning_rate": 0.0009728425755670995,
      "loss": 1.3745,
      "step": 1268
    },
    {
      "epoch": 0.10736948980455199,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009727992621396954,
      "loss": 1.9907,
      "step": 1269
    },
    {
      "epoch": 0.10745409933158474,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009727559151652451,
      "loss": 1.2928,
      "step": 1270
    },
    {
      "epoch": 0.10753870885861748,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009727125346468241,
      "loss": 1.7407,
      "step": 1271
    },
    {
      "epoch": 0.10762331838565023,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009726691205875109,
      "loss": 1.6605,
      "step": 1272
    },
    {
      "epoch": 0.10770792791268297,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009726256729903855,
      "loss": 2.0841,
      "step": 1273
    },
    {
      "epoch": 0.10779253743971572,
      "grad_norm": 2.875,
      "learning_rate": 0.000972582191858531,
      "loss": 2.1977,
      "step": 1274
    },
    {
      "epoch": 0.10787714696674845,
      "grad_norm": 1.921875,
      "learning_rate": 0.000972538677195032,
      "loss": 1.8519,
      "step": 1275
    },
    {
      "epoch": 0.1079617564937812,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009724951290029767,
      "loss": 2.0959,
      "step": 1276
    },
    {
      "epoch": 0.10804636602081394,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009724515472854547,
      "loss": 2.4057,
      "step": 1277
    },
    {
      "epoch": 0.10813097554784669,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009724079320455584,
      "loss": 1.0793,
      "step": 1278
    },
    {
      "epoch": 0.10821558507487943,
      "grad_norm": 2.5,
      "learning_rate": 0.0009723642832863824,
      "loss": 1.5975,
      "step": 1279
    },
    {
      "epoch": 0.10830019460191218,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009723206010110238,
      "loss": 1.6228,
      "step": 1280
    },
    {
      "epoch": 0.10838480412894493,
      "grad_norm": 1.75,
      "learning_rate": 0.0009722768852225819,
      "loss": 1.3604,
      "step": 1281
    },
    {
      "epoch": 0.10846941365597766,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009722331359241587,
      "loss": 1.8574,
      "step": 1282
    },
    {
      "epoch": 0.10855402318301041,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009721893531188584,
      "loss": 1.2707,
      "step": 1283
    },
    {
      "epoch": 0.10863863271004315,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009721455368097876,
      "loss": 1.3574,
      "step": 1284
    },
    {
      "epoch": 0.1087232422370759,
      "grad_norm": 9.5625,
      "learning_rate": 0.000972101687000055,
      "loss": 1.623,
      "step": 1285
    },
    {
      "epoch": 0.10880785176410863,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009720578036927721,
      "loss": 1.2452,
      "step": 1286
    },
    {
      "epoch": 0.10889246129114138,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009720138868910524,
      "loss": 1.2837,
      "step": 1287
    },
    {
      "epoch": 0.10897707081817412,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009719699365980124,
      "loss": 1.3889,
      "step": 1288
    },
    {
      "epoch": 0.10906168034520687,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009719259528167704,
      "loss": 2.0841,
      "step": 1289
    },
    {
      "epoch": 0.10914628987223961,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009718819355504468,
      "loss": 1.3764,
      "step": 1290
    },
    {
      "epoch": 0.10923089939927236,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009718378848021654,
      "loss": 1.4692,
      "step": 1291
    },
    {
      "epoch": 0.10931550892630511,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009717938005750513,
      "loss": 1.4256,
      "step": 1292
    },
    {
      "epoch": 0.10940011845333784,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009717496828722329,
      "loss": 1.5266,
      "step": 1293
    },
    {
      "epoch": 0.1094847279803706,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009717055316968402,
      "loss": 1.4898,
      "step": 1294
    },
    {
      "epoch": 0.10956933750740333,
      "grad_norm": 1.71875,
      "learning_rate": 0.0009716613470520059,
      "loss": 1.4224,
      "step": 1295
    },
    {
      "epoch": 0.10965394703443608,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009716171289408654,
      "loss": 1.3272,
      "step": 1296
    },
    {
      "epoch": 0.10973855656146882,
      "grad_norm": 2.15625,
      "learning_rate": 0.000971572877366556,
      "loss": 2.1372,
      "step": 1297
    },
    {
      "epoch": 0.10982316608850157,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009715285923322173,
      "loss": 1.3308,
      "step": 1298
    },
    {
      "epoch": 0.1099077756155343,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009714842738409917,
      "loss": 1.1388,
      "step": 1299
    },
    {
      "epoch": 0.10999238514256705,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000971439921896024,
      "loss": 1.3074,
      "step": 1300
    },
    {
      "epoch": 0.1100769946695998,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009713955365004605,
      "loss": 1.4713,
      "step": 1301
    },
    {
      "epoch": 0.11016160419663254,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009713511176574513,
      "loss": 1.4709,
      "step": 1302
    },
    {
      "epoch": 0.11024621372366529,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009713066653701477,
      "loss": 2.0952,
      "step": 1303
    },
    {
      "epoch": 0.11033082325069803,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009712621796417037,
      "loss": 1.7456,
      "step": 1304
    },
    {
      "epoch": 0.11041543277773078,
      "grad_norm": 2.078125,
      "learning_rate": 0.000971217660475276,
      "loss": 1.4681,
      "step": 1305
    },
    {
      "epoch": 0.11050004230476351,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009711731078740232,
      "loss": 1.3407,
      "step": 1306
    },
    {
      "epoch": 0.11058465183179626,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009711285218411066,
      "loss": 1.6956,
      "step": 1307
    },
    {
      "epoch": 0.110669261358829,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009710839023796896,
      "loss": 1.3339,
      "step": 1308
    },
    {
      "epoch": 0.11075387088586175,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009710392494929383,
      "loss": 1.8294,
      "step": 1309
    },
    {
      "epoch": 0.11083848041289449,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009709945631840211,
      "loss": 1.6254,
      "step": 1310
    },
    {
      "epoch": 0.11092308993992724,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009709498434561085,
      "loss": 1.7818,
      "step": 1311
    },
    {
      "epoch": 0.11100769946695999,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009709050903123735,
      "loss": 1.2549,
      "step": 1312
    },
    {
      "epoch": 0.11109230899399272,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009708603037559918,
      "loss": 1.4361,
      "step": 1313
    },
    {
      "epoch": 0.11117691852102547,
      "grad_norm": 2.25,
      "learning_rate": 0.0009708154837901409,
      "loss": 1.6412,
      "step": 1314
    },
    {
      "epoch": 0.11126152804805821,
      "grad_norm": 1.9453125,
      "learning_rate": 0.000970770630418001,
      "loss": 1.2866,
      "step": 1315
    },
    {
      "epoch": 0.11134613757509096,
      "grad_norm": 1.421875,
      "learning_rate": 0.0009707257436427548,
      "loss": 1.0412,
      "step": 1316
    },
    {
      "epoch": 0.1114307471021237,
      "grad_norm": 4.3125,
      "learning_rate": 0.000970680823467587,
      "loss": 1.2454,
      "step": 1317
    },
    {
      "epoch": 0.11151535662915645,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009706358698956851,
      "loss": 2.0625,
      "step": 1318
    },
    {
      "epoch": 0.11159996615618918,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009705908829302386,
      "loss": 1.6725,
      "step": 1319
    },
    {
      "epoch": 0.11168457568322193,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009705458625744395,
      "loss": 1.2903,
      "step": 1320
    },
    {
      "epoch": 0.11176918521025467,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009705008088314819,
      "loss": 1.3689,
      "step": 1321
    },
    {
      "epoch": 0.11185379473728742,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009704557217045632,
      "loss": 1.4749,
      "step": 1322
    },
    {
      "epoch": 0.11193840426432017,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009704106011968821,
      "loss": 1.4468,
      "step": 1323
    },
    {
      "epoch": 0.1120230137913529,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009703654473116401,
      "loss": 1.4558,
      "step": 1324
    },
    {
      "epoch": 0.11210762331838565,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009703202600520411,
      "loss": 1.3395,
      "step": 1325
    },
    {
      "epoch": 0.11219223284541839,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009702750394212913,
      "loss": 1.2188,
      "step": 1326
    },
    {
      "epoch": 0.11227684237245114,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009702297854225992,
      "loss": 1.7909,
      "step": 1327
    },
    {
      "epoch": 0.11236145189948388,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009701844980591761,
      "loss": 1.3553,
      "step": 1328
    },
    {
      "epoch": 0.11244606142651663,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009701391773342349,
      "loss": 1.2229,
      "step": 1329
    },
    {
      "epoch": 0.11253067095354936,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009700938232509916,
      "loss": 1.7963,
      "step": 1330
    },
    {
      "epoch": 0.11261528048058211,
      "grad_norm": 1.59375,
      "learning_rate": 0.000970048435812664,
      "loss": 1.1655,
      "step": 1331
    },
    {
      "epoch": 0.11269989000761486,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009700030150224728,
      "loss": 1.1911,
      "step": 1332
    },
    {
      "epoch": 0.1127844995346476,
      "grad_norm": 7.03125,
      "learning_rate": 0.0009699575608836407,
      "loss": 1.716,
      "step": 1333
    },
    {
      "epoch": 0.11286910906168035,
      "grad_norm": 5.5625,
      "learning_rate": 0.0009699120733993926,
      "loss": 1.4594,
      "step": 1334
    },
    {
      "epoch": 0.11295371858871309,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009698665525729565,
      "loss": 1.6004,
      "step": 1335
    },
    {
      "epoch": 0.11303832811574584,
      "grad_norm": 3.90625,
      "learning_rate": 0.0009698209984075618,
      "loss": 1.4957,
      "step": 1336
    },
    {
      "epoch": 0.11312293764277857,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009697754109064411,
      "loss": 1.6579,
      "step": 1337
    },
    {
      "epoch": 0.11320754716981132,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009697297900728288,
      "loss": 1.1856,
      "step": 1338
    },
    {
      "epoch": 0.11329215669684406,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009696841359099622,
      "loss": 1.5672,
      "step": 1339
    },
    {
      "epoch": 0.11337676622387681,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009696384484210803,
      "loss": 1.6304,
      "step": 1340
    },
    {
      "epoch": 0.11346137575090955,
      "grad_norm": 1.5,
      "learning_rate": 0.0009695927276094249,
      "loss": 1.3056,
      "step": 1341
    },
    {
      "epoch": 0.1135459852779423,
      "grad_norm": 2.453125,
      "learning_rate": 0.00096954697347824,
      "loss": 1.8999,
      "step": 1342
    },
    {
      "epoch": 0.11363059480497505,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009695011860307725,
      "loss": 1.573,
      "step": 1343
    },
    {
      "epoch": 0.11371520433200778,
      "grad_norm": 1.875,
      "learning_rate": 0.0009694553652702707,
      "loss": 1.1649,
      "step": 1344
    },
    {
      "epoch": 0.11379981385904053,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0009694095111999858,
      "loss": 1.1248,
      "step": 1345
    },
    {
      "epoch": 0.11388442338607327,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009693636238231718,
      "loss": 1.5645,
      "step": 1346
    },
    {
      "epoch": 0.11396903291310602,
      "grad_norm": 5.1875,
      "learning_rate": 0.0009693177031430842,
      "loss": 1.5147,
      "step": 1347
    },
    {
      "epoch": 0.11405364244013876,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009692717491629812,
      "loss": 1.2638,
      "step": 1348
    },
    {
      "epoch": 0.1141382519671715,
      "grad_norm": 8.125,
      "learning_rate": 0.0009692257618861236,
      "loss": 1.2806,
      "step": 1349
    },
    {
      "epoch": 0.11422286149420424,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009691797413157744,
      "loss": 1.6189,
      "step": 1350
    },
    {
      "epoch": 0.11430747102123699,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009691336874551991,
      "loss": 1.6729,
      "step": 1351
    },
    {
      "epoch": 0.11439208054826973,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009690876003076649,
      "loss": 1.3271,
      "step": 1352
    },
    {
      "epoch": 0.11447669007530248,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009690414798764422,
      "loss": 1.3906,
      "step": 1353
    },
    {
      "epoch": 0.11456129960233523,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009689953261648036,
      "loss": 1.3052,
      "step": 1354
    },
    {
      "epoch": 0.11464590912936796,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009689491391760236,
      "loss": 1.2964,
      "step": 1355
    },
    {
      "epoch": 0.11473051865640072,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009689029189133794,
      "loss": 2.0084,
      "step": 1356
    },
    {
      "epoch": 0.11481512818343345,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009688566653801508,
      "loss": 1.3417,
      "step": 1357
    },
    {
      "epoch": 0.1148997377104662,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009688103785796193,
      "loss": 1.3176,
      "step": 1358
    },
    {
      "epoch": 0.11498434723749894,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009687640585150693,
      "loss": 1.642,
      "step": 1359
    },
    {
      "epoch": 0.11506895676453169,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009687177051897875,
      "loss": 1.4681,
      "step": 1360
    },
    {
      "epoch": 0.11515356629156442,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009686713186070625,
      "loss": 1.4532,
      "step": 1361
    },
    {
      "epoch": 0.11523817581859717,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009686248987701863,
      "loss": 2.2981,
      "step": 1362
    },
    {
      "epoch": 0.11532278534562992,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009685784456824518,
      "loss": 1.9279,
      "step": 1363
    },
    {
      "epoch": 0.11540739487266266,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009685319593471555,
      "loss": 2.327,
      "step": 1364
    },
    {
      "epoch": 0.11549200439969541,
      "grad_norm": 1.46875,
      "learning_rate": 0.0009684854397675958,
      "loss": 1.3184,
      "step": 1365
    },
    {
      "epoch": 0.11557661392672815,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009684388869470731,
      "loss": 1.2853,
      "step": 1366
    },
    {
      "epoch": 0.1156612234537609,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0009683923008888909,
      "loss": 1.4037,
      "step": 1367
    },
    {
      "epoch": 0.11574583298079363,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009683456815963545,
      "loss": 1.4846,
      "step": 1368
    },
    {
      "epoch": 0.11583044250782638,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009682990290727718,
      "loss": 1.6489,
      "step": 1369
    },
    {
      "epoch": 0.11591505203485912,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009682523433214529,
      "loss": 1.8534,
      "step": 1370
    },
    {
      "epoch": 0.11599966156189187,
      "grad_norm": 1.421875,
      "learning_rate": 0.0009682056243457104,
      "loss": 1.2607,
      "step": 1371
    },
    {
      "epoch": 0.1160842710889246,
      "grad_norm": 1.875,
      "learning_rate": 0.0009681588721488592,
      "loss": 1.6378,
      "step": 1372
    },
    {
      "epoch": 0.11616888061595736,
      "grad_norm": 2.5,
      "learning_rate": 0.0009681120867342166,
      "loss": 1.2047,
      "step": 1373
    },
    {
      "epoch": 0.1162534901429901,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009680652681051021,
      "loss": 1.4987,
      "step": 1374
    },
    {
      "epoch": 0.11633809967002284,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009680184162648378,
      "loss": 2.2783,
      "step": 1375
    },
    {
      "epoch": 0.1164227091970556,
      "grad_norm": 2.15625,
      "learning_rate": 0.000967971531216748,
      "loss": 1.6945,
      "step": 1376
    },
    {
      "epoch": 0.11650731872408833,
      "grad_norm": 1.765625,
      "learning_rate": 0.0009679246129641592,
      "loss": 1.4369,
      "step": 1377
    },
    {
      "epoch": 0.11659192825112108,
      "grad_norm": 1.875,
      "learning_rate": 0.0009678776615104008,
      "loss": 1.2187,
      "step": 1378
    },
    {
      "epoch": 0.11667653777815382,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009678306768588039,
      "loss": 1.3711,
      "step": 1379
    },
    {
      "epoch": 0.11676114730518657,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009677836590127024,
      "loss": 1.5878,
      "step": 1380
    },
    {
      "epoch": 0.1168457568322193,
      "grad_norm": 2.25,
      "learning_rate": 0.0009677366079754323,
      "loss": 2.3004,
      "step": 1381
    },
    {
      "epoch": 0.11693036635925205,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009676895237503322,
      "loss": 1.9087,
      "step": 1382
    },
    {
      "epoch": 0.11701497588628479,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009676424063407427,
      "loss": 1.231,
      "step": 1383
    },
    {
      "epoch": 0.11709958541331754,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009675952557500071,
      "loss": 1.8616,
      "step": 1384
    },
    {
      "epoch": 0.11718419494035029,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009675480719814712,
      "loss": 1.3295,
      "step": 1385
    },
    {
      "epoch": 0.11726880446738303,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009675008550384826,
      "loss": 1.6486,
      "step": 1386
    },
    {
      "epoch": 0.11735341399441578,
      "grad_norm": 1.875,
      "learning_rate": 0.0009674536049243913,
      "loss": 1.4837,
      "step": 1387
    },
    {
      "epoch": 0.11743802352144851,
      "grad_norm": 1.40625,
      "learning_rate": 0.0009674063216425503,
      "loss": 1.0784,
      "step": 1388
    },
    {
      "epoch": 0.11752263304848126,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009673590051963145,
      "loss": 1.2668,
      "step": 1389
    },
    {
      "epoch": 0.117607242575514,
      "grad_norm": 2.15625,
      "learning_rate": 0.000967311655589041,
      "loss": 1.7921,
      "step": 1390
    },
    {
      "epoch": 0.11769185210254675,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009672642728240896,
      "loss": 1.5237,
      "step": 1391
    },
    {
      "epoch": 0.11777646162957948,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009672168569048221,
      "loss": 1.3162,
      "step": 1392
    },
    {
      "epoch": 0.11786107115661223,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009671694078346031,
      "loss": 2.2811,
      "step": 1393
    },
    {
      "epoch": 0.11794568068364499,
      "grad_norm": 1.796875,
      "learning_rate": 0.0009671219256167993,
      "loss": 1.4129,
      "step": 1394
    },
    {
      "epoch": 0.11803029021067772,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009670744102547795,
      "loss": 1.5846,
      "step": 1395
    },
    {
      "epoch": 0.11811489973771047,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009670268617519153,
      "loss": 1.3402,
      "step": 1396
    },
    {
      "epoch": 0.11819950926474321,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009669792801115804,
      "loss": 1.4835,
      "step": 1397
    },
    {
      "epoch": 0.11828411879177596,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000966931665337151,
      "loss": 1.1357,
      "step": 1398
    },
    {
      "epoch": 0.1183687283188087,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009668840174320055,
      "loss": 1.1846,
      "step": 1399
    },
    {
      "epoch": 0.11845333784584144,
      "grad_norm": 2.25,
      "learning_rate": 0.0009668363363995248,
      "loss": 1.4786,
      "step": 1400
    },
    {
      "epoch": 0.11853794737287418,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009667886222430919,
      "loss": 2.3769,
      "step": 1401
    },
    {
      "epoch": 0.11862255689990693,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009667408749660924,
      "loss": 1.7232,
      "step": 1402
    },
    {
      "epoch": 0.11870716642693967,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009666930945719141,
      "loss": 2.4226,
      "step": 1403
    },
    {
      "epoch": 0.11879177595397242,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009666452810639474,
      "loss": 2.1532,
      "step": 1404
    },
    {
      "epoch": 0.11887638548100517,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009665974344455848,
      "loss": 1.4863,
      "step": 1405
    },
    {
      "epoch": 0.1189609950080379,
      "grad_norm": 1.7734375,
      "learning_rate": 0.000966549554720221,
      "loss": 1.7287,
      "step": 1406
    },
    {
      "epoch": 0.11904560453507065,
      "grad_norm": 1.4375,
      "learning_rate": 0.0009665016418912535,
      "loss": 1.4212,
      "step": 1407
    },
    {
      "epoch": 0.11913021406210339,
      "grad_norm": 2.0,
      "learning_rate": 0.0009664536959620817,
      "loss": 1.7071,
      "step": 1408
    },
    {
      "epoch": 0.11921482358913614,
      "grad_norm": 2.140625,
      "learning_rate": 0.000966405716936108,
      "loss": 2.0373,
      "step": 1409
    },
    {
      "epoch": 0.11929943311616888,
      "grad_norm": 2.125,
      "learning_rate": 0.0009663577048167363,
      "loss": 1.4342,
      "step": 1410
    },
    {
      "epoch": 0.11938404264320163,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009663096596073731,
      "loss": 1.349,
      "step": 1411
    },
    {
      "epoch": 0.11946865217023436,
      "grad_norm": 2.078125,
      "learning_rate": 0.000966261581311428,
      "loss": 1.5168,
      "step": 1412
    },
    {
      "epoch": 0.11955326169726711,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009662134699323118,
      "loss": 1.5704,
      "step": 1413
    },
    {
      "epoch": 0.11963787122429985,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009661653254734385,
      "loss": 1.589,
      "step": 1414
    },
    {
      "epoch": 0.1197224807513326,
      "grad_norm": 1.703125,
      "learning_rate": 0.000966117147938224,
      "loss": 1.2434,
      "step": 1415
    },
    {
      "epoch": 0.11980709027836535,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009660689373300869,
      "loss": 1.7883,
      "step": 1416
    },
    {
      "epoch": 0.11989169980539809,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009660206936524476,
      "loss": 1.288,
      "step": 1417
    },
    {
      "epoch": 0.11997630933243084,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009659724169087295,
      "loss": 1.8011,
      "step": 1418
    },
    {
      "epoch": 0.12006091885946357,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009659241071023579,
      "loss": 1.3013,
      "step": 1419
    },
    {
      "epoch": 0.12014552838649632,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009658757642367606,
      "loss": 1.1419,
      "step": 1420
    },
    {
      "epoch": 0.12023013791352906,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009658273883153678,
      "loss": 1.8795,
      "step": 1421
    },
    {
      "epoch": 0.12031474744056181,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009657789793416117,
      "loss": 1.4308,
      "step": 1422
    },
    {
      "epoch": 0.12039935696759455,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009657305373189275,
      "loss": 1.6976,
      "step": 1423
    },
    {
      "epoch": 0.1204839664946273,
      "grad_norm": 2.296875,
      "learning_rate": 0.000965682062250752,
      "loss": 1.9072,
      "step": 1424
    },
    {
      "epoch": 0.12056857602166005,
      "grad_norm": 1.8515625,
      "learning_rate": 0.000965633554140525,
      "loss": 1.4905,
      "step": 1425
    },
    {
      "epoch": 0.12065318554869278,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000965585012991688,
      "loss": 1.483,
      "step": 1426
    },
    {
      "epoch": 0.12073779507572553,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009655364388076857,
      "loss": 1.7728,
      "step": 1427
    },
    {
      "epoch": 0.12082240460275827,
      "grad_norm": 2.25,
      "learning_rate": 0.0009654878315919642,
      "loss": 1.2446,
      "step": 1428
    },
    {
      "epoch": 0.12090701412979102,
      "grad_norm": 2.0,
      "learning_rate": 0.0009654391913479726,
      "loss": 1.2488,
      "step": 1429
    },
    {
      "epoch": 0.12099162365682375,
      "grad_norm": 1.875,
      "learning_rate": 0.000965390518079162,
      "loss": 1.4446,
      "step": 1430
    },
    {
      "epoch": 0.1210762331838565,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0009653418117889861,
      "loss": 1.3892,
      "step": 1431
    },
    {
      "epoch": 0.12116084271088924,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009652930724809007,
      "loss": 1.4618,
      "step": 1432
    },
    {
      "epoch": 0.12124545223792199,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009652443001583641,
      "loss": 1.324,
      "step": 1433
    },
    {
      "epoch": 0.12133006176495473,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009651954948248368,
      "loss": 1.3312,
      "step": 1434
    },
    {
      "epoch": 0.12141467129198748,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009651466564837819,
      "loss": 1.7812,
      "step": 1435
    },
    {
      "epoch": 0.12149928081902023,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009650977851386646,
      "loss": 1.6304,
      "step": 1436
    },
    {
      "epoch": 0.12158389034605296,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009650488807929525,
      "loss": 1.4517,
      "step": 1437
    },
    {
      "epoch": 0.12166849987308571,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009649999434501156,
      "loss": 1.6517,
      "step": 1438
    },
    {
      "epoch": 0.12175310940011845,
      "grad_norm": 1.421875,
      "learning_rate": 0.0009649509731136263,
      "loss": 1.1628,
      "step": 1439
    },
    {
      "epoch": 0.1218377189271512,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000964901969786959,
      "loss": 1.2173,
      "step": 1440
    },
    {
      "epoch": 0.12192232845418394,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009648529334735908,
      "loss": 1.3575,
      "step": 1441
    },
    {
      "epoch": 0.12200693798121669,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009648038641770012,
      "loss": 1.3882,
      "step": 1442
    },
    {
      "epoch": 0.12209154750824942,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009647547619006715,
      "loss": 1.3644,
      "step": 1443
    },
    {
      "epoch": 0.12217615703528217,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009647056266480862,
      "loss": 1.7855,
      "step": 1444
    },
    {
      "epoch": 0.12226076656231491,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009646564584227313,
      "loss": 1.7909,
      "step": 1445
    },
    {
      "epoch": 0.12234537608934766,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009646072572280955,
      "loss": 1.7245,
      "step": 1446
    },
    {
      "epoch": 0.12242998561638041,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00096455802306767,
      "loss": 1.3732,
      "step": 1447
    },
    {
      "epoch": 0.12251459514341315,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009645087559449479,
      "loss": 1.6771,
      "step": 1448
    },
    {
      "epoch": 0.1225992046704459,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009644594558634252,
      "loss": 1.5598,
      "step": 1449
    },
    {
      "epoch": 0.12268381419747863,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009644101228265997,
      "loss": 1.7255,
      "step": 1450
    },
    {
      "epoch": 0.12276842372451138,
      "grad_norm": 2.078125,
      "learning_rate": 0.000964360756837972,
      "loss": 1.4548,
      "step": 1451
    },
    {
      "epoch": 0.12285303325154412,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009643113579010446,
      "loss": 1.209,
      "step": 1452
    },
    {
      "epoch": 0.12293764277857687,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009642619260193226,
      "loss": 1.5805,
      "step": 1453
    },
    {
      "epoch": 0.1230222523056096,
      "grad_norm": 2.5,
      "learning_rate": 0.0009642124611963137,
      "loss": 1.4141,
      "step": 1454
    },
    {
      "epoch": 0.12310686183264236,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009641629634355271,
      "loss": 1.4156,
      "step": 1455
    },
    {
      "epoch": 0.1231914713596751,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009641134327404755,
      "loss": 1.3704,
      "step": 1456
    },
    {
      "epoch": 0.12327608088670784,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009640638691146727,
      "loss": 1.4825,
      "step": 1457
    },
    {
      "epoch": 0.12336069041374059,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009640142725616357,
      "loss": 1.9116,
      "step": 1458
    },
    {
      "epoch": 0.12344529994077333,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009639646430848838,
      "loss": 1.2592,
      "step": 1459
    },
    {
      "epoch": 0.12352990946780608,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000963914980687938,
      "loss": 1.325,
      "step": 1460
    },
    {
      "epoch": 0.12361451899483882,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009638652853743224,
      "loss": 1.3751,
      "step": 1461
    },
    {
      "epoch": 0.12369912852187157,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009638155571475628,
      "loss": 1.9815,
      "step": 1462
    },
    {
      "epoch": 0.1237837380489043,
      "grad_norm": 2.375,
      "learning_rate": 0.0009637657960111878,
      "loss": 1.9995,
      "step": 1463
    },
    {
      "epoch": 0.12386834757593705,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009637160019687282,
      "loss": 1.1554,
      "step": 1464
    },
    {
      "epoch": 0.12395295710296979,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000963666175023717,
      "loss": 1.5514,
      "step": 1465
    },
    {
      "epoch": 0.12403756663000254,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009636163151796895,
      "loss": 1.7037,
      "step": 1466
    },
    {
      "epoch": 0.12412217615703529,
      "grad_norm": 1.625,
      "learning_rate": 0.0009635664224401838,
      "loss": 1.0985,
      "step": 1467
    },
    {
      "epoch": 0.12420678568406802,
      "grad_norm": 1.3125,
      "learning_rate": 0.0009635164968087396,
      "loss": 1.164,
      "step": 1468
    },
    {
      "epoch": 0.12429139521110077,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009634665382888996,
      "loss": 1.614,
      "step": 1469
    },
    {
      "epoch": 0.12437600473813351,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009634165468842084,
      "loss": 2.0477,
      "step": 1470
    },
    {
      "epoch": 0.12446061426516626,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009633665225982133,
      "loss": 1.7466,
      "step": 1471
    },
    {
      "epoch": 0.124545223792199,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009633164654344636,
      "loss": 2.6692,
      "step": 1472
    },
    {
      "epoch": 0.12462983331923175,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009632663753965109,
      "loss": 1.8574,
      "step": 1473
    },
    {
      "epoch": 0.12471444284626448,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009632162524879096,
      "loss": 1.8344,
      "step": 1474
    },
    {
      "epoch": 0.12479905237329723,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000963166096712216,
      "loss": 1.117,
      "step": 1475
    },
    {
      "epoch": 0.12488366190032997,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009631159080729887,
      "loss": 1.7059,
      "step": 1476
    },
    {
      "epoch": 0.12496827142736272,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009630656865737892,
      "loss": 1.3479,
      "step": 1477
    },
    {
      "epoch": 0.12505288095439546,
      "grad_norm": 1.375,
      "learning_rate": 0.0009630154322181803,
      "loss": 1.1603,
      "step": 1478
    },
    {
      "epoch": 0.12513749048142822,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009629651450097283,
      "loss": 2.1208,
      "step": 1479
    },
    {
      "epoch": 0.12522210000846096,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009629148249520012,
      "loss": 1.7463,
      "step": 1480
    },
    {
      "epoch": 0.1253067095354937,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009628644720485691,
      "loss": 1.5732,
      "step": 1481
    },
    {
      "epoch": 0.12539131906252643,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0009628140863030051,
      "loss": 1.307,
      "step": 1482
    },
    {
      "epoch": 0.1254759285895592,
      "grad_norm": 5.15625,
      "learning_rate": 0.000962763667718884,
      "loss": 1.1641,
      "step": 1483
    },
    {
      "epoch": 0.12556053811659193,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009627132162997833,
      "loss": 2.9223,
      "step": 1484
    },
    {
      "epoch": 0.12564514764362467,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009626627320492829,
      "loss": 1.8814,
      "step": 1485
    },
    {
      "epoch": 0.1257297571706574,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009626122149709646,
      "loss": 1.1738,
      "step": 1486
    },
    {
      "epoch": 0.12581436669769017,
      "grad_norm": 2.09375,
      "learning_rate": 0.000962561665068413,
      "loss": 1.9127,
      "step": 1487
    },
    {
      "epoch": 0.1258989762247229,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009625110823452148,
      "loss": 1.6481,
      "step": 1488
    },
    {
      "epoch": 0.12598358575175564,
      "grad_norm": 1.71875,
      "learning_rate": 0.0009624604668049588,
      "loss": 1.5491,
      "step": 1489
    },
    {
      "epoch": 0.1260681952787884,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009624098184512368,
      "loss": 1.5094,
      "step": 1490
    },
    {
      "epoch": 0.12615280480582114,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0009623591372876422,
      "loss": 1.1754,
      "step": 1491
    },
    {
      "epoch": 0.12623741433285388,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009623084233177711,
      "loss": 1.7325,
      "step": 1492
    },
    {
      "epoch": 0.1263220238598866,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009622576765452218,
      "loss": 2.2729,
      "step": 1493
    },
    {
      "epoch": 0.12640663338691938,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009622068969735951,
      "loss": 1.4665,
      "step": 1494
    },
    {
      "epoch": 0.1264912429139521,
      "grad_norm": 1.578125,
      "learning_rate": 0.000962156084606494,
      "loss": 1.0787,
      "step": 1495
    },
    {
      "epoch": 0.12657585244098485,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009621052394475241,
      "loss": 1.4844,
      "step": 1496
    },
    {
      "epoch": 0.1266604619680176,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009620543615002925,
      "loss": 1.7645,
      "step": 1497
    },
    {
      "epoch": 0.12674507149505035,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009620034507684097,
      "loss": 1.3078,
      "step": 1498
    },
    {
      "epoch": 0.12682968102208309,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009619525072554878,
      "loss": 1.6513,
      "step": 1499
    },
    {
      "epoch": 0.12691429054911582,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0009619015309651414,
      "loss": 1.4951,
      "step": 1500
    },
    {
      "epoch": 0.12699890007614859,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009618505219009876,
      "loss": 1.335,
      "step": 1501
    },
    {
      "epoch": 0.12708350960318132,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009617994800666457,
      "loss": 1.5282,
      "step": 1502
    },
    {
      "epoch": 0.12716811913021406,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009617484054657373,
      "loss": 1.6293,
      "step": 1503
    },
    {
      "epoch": 0.1272527286572468,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009616972981018862,
      "loss": 1.3002,
      "step": 1504
    },
    {
      "epoch": 0.12733733818427956,
      "grad_norm": 1.546875,
      "learning_rate": 0.000961646157978719,
      "loss": 1.2245,
      "step": 1505
    },
    {
      "epoch": 0.1274219477113123,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000961594985099864,
      "loss": 1.5624,
      "step": 1506
    },
    {
      "epoch": 0.12750655723834503,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009615437794689523,
      "loss": 1.5946,
      "step": 1507
    },
    {
      "epoch": 0.1275911667653778,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009614925410896171,
      "loss": 1.5274,
      "step": 1508
    },
    {
      "epoch": 0.12767577629241053,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0009614412699654941,
      "loss": 1.3838,
      "step": 1509
    },
    {
      "epoch": 0.12776038581944327,
      "grad_norm": 3.203125,
      "learning_rate": 0.000961389966100221,
      "loss": 1.8053,
      "step": 1510
    },
    {
      "epoch": 0.127844995346476,
      "grad_norm": 1.671875,
      "learning_rate": 0.000961338629497438,
      "loss": 1.2179,
      "step": 1511
    },
    {
      "epoch": 0.12792960487350877,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009612872601607878,
      "loss": 2.118,
      "step": 1512
    },
    {
      "epoch": 0.1280142144005415,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009612358580939151,
      "loss": 2.1665,
      "step": 1513
    },
    {
      "epoch": 0.12809882392757424,
      "grad_norm": 1.578125,
      "learning_rate": 0.0009611844233004674,
      "loss": 1.1916,
      "step": 1514
    },
    {
      "epoch": 0.12818343345460698,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009611329557840938,
      "loss": 1.7597,
      "step": 1515
    },
    {
      "epoch": 0.12826804298163974,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009610814555484464,
      "loss": 2.5467,
      "step": 1516
    },
    {
      "epoch": 0.12835265250867248,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009610299225971793,
      "loss": 1.7181,
      "step": 1517
    },
    {
      "epoch": 0.1284372620357052,
      "grad_norm": 1.734375,
      "learning_rate": 0.000960978356933949,
      "loss": 1.5758,
      "step": 1518
    },
    {
      "epoch": 0.12852187156273798,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009609267585624141,
      "loss": 1.3388,
      "step": 1519
    },
    {
      "epoch": 0.1286064810897707,
      "grad_norm": 1.65625,
      "learning_rate": 0.000960875127486236,
      "loss": 1.2266,
      "step": 1520
    },
    {
      "epoch": 0.12869109061680345,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009608234637090778,
      "loss": 1.3682,
      "step": 1521
    },
    {
      "epoch": 0.12877570014383619,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009607717672346057,
      "loss": 1.4821,
      "step": 1522
    },
    {
      "epoch": 0.12886030967086895,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009607200380664873,
      "loss": 1.7901,
      "step": 1523
    },
    {
      "epoch": 0.1289449191979017,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009606682762083933,
      "loss": 1.5262,
      "step": 1524
    },
    {
      "epoch": 0.12902952872493442,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009606164816639966,
      "loss": 1.2778,
      "step": 1525
    },
    {
      "epoch": 0.12911413825196716,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009605646544369718,
      "loss": 1.2163,
      "step": 1526
    },
    {
      "epoch": 0.12919874777899992,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009605127945309964,
      "loss": 1.3738,
      "step": 1527
    },
    {
      "epoch": 0.12928335730603266,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009604609019497501,
      "loss": 1.2378,
      "step": 1528
    },
    {
      "epoch": 0.1293679668330654,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009604089766969148,
      "loss": 1.4164,
      "step": 1529
    },
    {
      "epoch": 0.12945257636009816,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009603570187761752,
      "loss": 1.8496,
      "step": 1530
    },
    {
      "epoch": 0.1295371858871309,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009603050281912174,
      "loss": 1.7854,
      "step": 1531
    },
    {
      "epoch": 0.12962179541416363,
      "grad_norm": 2.5,
      "learning_rate": 0.0009602530049457308,
      "loss": 1.4132,
      "step": 1532
    },
    {
      "epoch": 0.12970640494119637,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009602009490434063,
      "loss": 1.1467,
      "step": 1533
    },
    {
      "epoch": 0.12979101446822913,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009601488604879376,
      "loss": 1.423,
      "step": 1534
    },
    {
      "epoch": 0.12987562399526187,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009600967392830207,
      "loss": 1.4438,
      "step": 1535
    },
    {
      "epoch": 0.1299602335222946,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009600445854323536,
      "loss": 2.2134,
      "step": 1536
    },
    {
      "epoch": 0.13004484304932734,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009599923989396371,
      "loss": 1.6875,
      "step": 1537
    },
    {
      "epoch": 0.1301294525763601,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0009599401798085738,
      "loss": 1.1595,
      "step": 1538
    },
    {
      "epoch": 0.13021406210339284,
      "grad_norm": 1.796875,
      "learning_rate": 0.0009598879280428691,
      "loss": 1.8692,
      "step": 1539
    },
    {
      "epoch": 0.13029867163042558,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009598356436462303,
      "loss": 1.2502,
      "step": 1540
    },
    {
      "epoch": 0.13038328115745834,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0009597833266223673,
      "loss": 1.1216,
      "step": 1541
    },
    {
      "epoch": 0.13046789068449108,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000959730976974992,
      "loss": 1.565,
      "step": 1542
    },
    {
      "epoch": 0.13055250021152381,
      "grad_norm": 1.75,
      "learning_rate": 0.0009596785947078191,
      "loss": 1.701,
      "step": 1543
    },
    {
      "epoch": 0.13063710973855655,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009596261798245651,
      "loss": 1.611,
      "step": 1544
    },
    {
      "epoch": 0.13072171926558931,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009595737323289491,
      "loss": 1.7228,
      "step": 1545
    },
    {
      "epoch": 0.13080632879262205,
      "grad_norm": 1.75,
      "learning_rate": 0.0009595212522246927,
      "loss": 1.1685,
      "step": 1546
    },
    {
      "epoch": 0.1308909383196548,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009594687395155192,
      "loss": 1.9043,
      "step": 1547
    },
    {
      "epoch": 0.13097554784668752,
      "grad_norm": 2.296875,
      "learning_rate": 0.000959416194205155,
      "loss": 1.9119,
      "step": 1548
    },
    {
      "epoch": 0.1310601573737203,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009593636162973281,
      "loss": 1.1613,
      "step": 1549
    },
    {
      "epoch": 0.13114476690075302,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009593110057957693,
      "loss": 1.3034,
      "step": 1550
    },
    {
      "epoch": 0.13122937642778576,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009592583627042114,
      "loss": 1.7489,
      "step": 1551
    },
    {
      "epoch": 0.13131398595481852,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009592056870263896,
      "loss": 1.3276,
      "step": 1552
    },
    {
      "epoch": 0.13139859548185126,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0009591529787660416,
      "loss": 1.3964,
      "step": 1553
    },
    {
      "epoch": 0.131483205008884,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009591002379269074,
      "loss": 1.9128,
      "step": 1554
    },
    {
      "epoch": 0.13156781453591673,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009590474645127287,
      "loss": 1.1463,
      "step": 1555
    },
    {
      "epoch": 0.1316524240629495,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009589946585272504,
      "loss": 1.4311,
      "step": 1556
    },
    {
      "epoch": 0.13173703358998223,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009589418199742192,
      "loss": 1.566,
      "step": 1557
    },
    {
      "epoch": 0.13182164311701497,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009588889488573842,
      "loss": 2.4008,
      "step": 1558
    },
    {
      "epoch": 0.13190625264404773,
      "grad_norm": 2.015625,
      "learning_rate": 0.000958836045180497,
      "loss": 1.7791,
      "step": 1559
    },
    {
      "epoch": 0.13199086217108047,
      "grad_norm": 1.796875,
      "learning_rate": 0.0009587831089473108,
      "loss": 1.6562,
      "step": 1560
    },
    {
      "epoch": 0.1320754716981132,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009587301401615822,
      "loss": 2.122,
      "step": 1561
    },
    {
      "epoch": 0.13216008122514594,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009586771388270692,
      "loss": 1.5242,
      "step": 1562
    },
    {
      "epoch": 0.1322446907521787,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009586241049475328,
      "loss": 2.136,
      "step": 1563
    },
    {
      "epoch": 0.13232930027921144,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009585710385267357,
      "loss": 1.8985,
      "step": 1564
    },
    {
      "epoch": 0.13241390980624418,
      "grad_norm": 5.625,
      "learning_rate": 0.0009585179395684432,
      "loss": 2.2641,
      "step": 1565
    },
    {
      "epoch": 0.13249851933327692,
      "grad_norm": 1.703125,
      "learning_rate": 0.000958464808076423,
      "loss": 1.0959,
      "step": 1566
    },
    {
      "epoch": 0.13258312886030968,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009584116440544449,
      "loss": 1.2457,
      "step": 1567
    },
    {
      "epoch": 0.13266773838734242,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009583584475062812,
      "loss": 1.6552,
      "step": 1568
    },
    {
      "epoch": 0.13275234791437515,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009583052184357064,
      "loss": 1.7083,
      "step": 1569
    },
    {
      "epoch": 0.13283695744140792,
      "grad_norm": 60.25,
      "learning_rate": 0.0009582519568464971,
      "loss": 1.7787,
      "step": 1570
    },
    {
      "epoch": 0.13292156696844065,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009581986627424328,
      "loss": 1.76,
      "step": 1571
    },
    {
      "epoch": 0.1330061764954734,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009581453361272947,
      "loss": 1.2456,
      "step": 1572
    },
    {
      "epoch": 0.13309078602250612,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009580919770048667,
      "loss": 1.2971,
      "step": 1573
    },
    {
      "epoch": 0.1331753955495389,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0009580385853789346,
      "loss": 1.0194,
      "step": 1574
    },
    {
      "epoch": 0.13326000507657162,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009579851612532868,
      "loss": 1.6361,
      "step": 1575
    },
    {
      "epoch": 0.13334461460360436,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009579317046317142,
      "loss": 1.405,
      "step": 1576
    },
    {
      "epoch": 0.1334292241306371,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009578782155180097,
      "loss": 1.3932,
      "step": 1577
    },
    {
      "epoch": 0.13351383365766986,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009578246939159682,
      "loss": 1.9802,
      "step": 1578
    },
    {
      "epoch": 0.1335984431847026,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0009577711398293878,
      "loss": 1.1689,
      "step": 1579
    },
    {
      "epoch": 0.13368305271173533,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009577175532620681,
      "loss": 1.6548,
      "step": 1580
    },
    {
      "epoch": 0.1337676622387681,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009576639342178114,
      "loss": 1.2942,
      "step": 1581
    },
    {
      "epoch": 0.13385227176580083,
      "grad_norm": 1.8125,
      "learning_rate": 0.000957610282700422,
      "loss": 1.3007,
      "step": 1582
    },
    {
      "epoch": 0.13393688129283357,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0009575565987137069,
      "loss": 1.3788,
      "step": 1583
    },
    {
      "epoch": 0.1340214908198663,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0009575028822614751,
      "loss": 1.2827,
      "step": 1584
    },
    {
      "epoch": 0.13410610034689907,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009574491333475382,
      "loss": 1.4072,
      "step": 1585
    },
    {
      "epoch": 0.1341907098739318,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009573953519757095,
      "loss": 1.4955,
      "step": 1586
    },
    {
      "epoch": 0.13427531940096454,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009573415381498052,
      "loss": 1.5971,
      "step": 1587
    },
    {
      "epoch": 0.13435992892799728,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0009572876918736437,
      "loss": 1.1167,
      "step": 1588
    },
    {
      "epoch": 0.13444453845503004,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009572338131510458,
      "loss": 1.9549,
      "step": 1589
    },
    {
      "epoch": 0.13452914798206278,
      "grad_norm": 2.609375,
      "learning_rate": 0.000957179901985834,
      "loss": 1.6955,
      "step": 1590
    },
    {
      "epoch": 0.13461375750909552,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0009571259583818337,
      "loss": 1.7026,
      "step": 1591
    },
    {
      "epoch": 0.13469836703612828,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009570719823428724,
      "loss": 1.6718,
      "step": 1592
    },
    {
      "epoch": 0.13478297656316102,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00095701797387278,
      "loss": 1.2349,
      "step": 1593
    },
    {
      "epoch": 0.13486758609019375,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009569639329753885,
      "loss": 1.7962,
      "step": 1594
    },
    {
      "epoch": 0.1349521956172265,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0009569098596545325,
      "loss": 1.2815,
      "step": 1595
    },
    {
      "epoch": 0.13503680514425925,
      "grad_norm": 1.5,
      "learning_rate": 0.0009568557539140484,
      "loss": 1.3097,
      "step": 1596
    },
    {
      "epoch": 0.135121414671292,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009568016157577757,
      "loss": 1.3788,
      "step": 1597
    },
    {
      "epoch": 0.13520602419832473,
      "grad_norm": 1.515625,
      "learning_rate": 0.0009567474451895552,
      "loss": 1.3699,
      "step": 1598
    },
    {
      "epoch": 0.13529063372535746,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009566932422132309,
      "loss": 1.6911,
      "step": 1599
    },
    {
      "epoch": 0.13537524325239023,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009566390068326487,
      "loss": 1.5648,
      "step": 1600
    },
    {
      "epoch": 0.13545985277942296,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009565847390516567,
      "loss": 1.5729,
      "step": 1601
    },
    {
      "epoch": 0.1355444623064557,
      "grad_norm": 2.0,
      "learning_rate": 0.0009565304388741053,
      "loss": 1.349,
      "step": 1602
    },
    {
      "epoch": 0.13562907183348846,
      "grad_norm": 2.5,
      "learning_rate": 0.0009564761063038476,
      "loss": 1.7593,
      "step": 1603
    },
    {
      "epoch": 0.1357136813605212,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009564217413447385,
      "loss": 1.7255,
      "step": 1604
    },
    {
      "epoch": 0.13579829088755394,
      "grad_norm": 2.375,
      "learning_rate": 0.0009563673440006357,
      "loss": 1.4864,
      "step": 1605
    },
    {
      "epoch": 0.13588290041458667,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009563129142753985,
      "loss": 1.335,
      "step": 1606
    },
    {
      "epoch": 0.13596750994161944,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009562584521728892,
      "loss": 1.8015,
      "step": 1607
    },
    {
      "epoch": 0.13605211946865217,
      "grad_norm": 1.34375,
      "learning_rate": 0.000956203957696972,
      "loss": 1.3246,
      "step": 1608
    },
    {
      "epoch": 0.1361367289956849,
      "grad_norm": 1.875,
      "learning_rate": 0.0009561494308515138,
      "loss": 1.8571,
      "step": 1609
    },
    {
      "epoch": 0.13622133852271764,
      "grad_norm": 1.65625,
      "learning_rate": 0.000956094871640383,
      "loss": 1.5245,
      "step": 1610
    },
    {
      "epoch": 0.1363059480497504,
      "grad_norm": 1.546875,
      "learning_rate": 0.000956040280067451,
      "loss": 1.1897,
      "step": 1611
    },
    {
      "epoch": 0.13639055757678314,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009559856561365915,
      "loss": 1.3043,
      "step": 1612
    },
    {
      "epoch": 0.13647516710381588,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009559309998516801,
      "loss": 1.5654,
      "step": 1613
    },
    {
      "epoch": 0.13655977663084864,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000955876311216595,
      "loss": 1.6837,
      "step": 1614
    },
    {
      "epoch": 0.13664438615788138,
      "grad_norm": 1.71875,
      "learning_rate": 0.0009558215902352163,
      "loss": 1.8868,
      "step": 1615
    },
    {
      "epoch": 0.13672899568491412,
      "grad_norm": 2.515625,
      "learning_rate": 0.000955766836911427,
      "loss": 1.9476,
      "step": 1616
    },
    {
      "epoch": 0.13681360521194685,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0009557120512491121,
      "loss": 1.2054,
      "step": 1617
    },
    {
      "epoch": 0.13689821473897962,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009556572332521586,
      "loss": 1.9723,
      "step": 1618
    },
    {
      "epoch": 0.13698282426601235,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009556023829244562,
      "loss": 1.2909,
      "step": 1619
    },
    {
      "epoch": 0.1370674337930451,
      "grad_norm": 3.921875,
      "learning_rate": 0.0009555475002698967,
      "loss": 1.2966,
      "step": 1620
    },
    {
      "epoch": 0.13715204332007785,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009554925852923744,
      "loss": 1.2749,
      "step": 1621
    },
    {
      "epoch": 0.1372366528471106,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0009554376379957855,
      "loss": 1.2405,
      "step": 1622
    },
    {
      "epoch": 0.13732126237414333,
      "grad_norm": 1.9296875,
      "learning_rate": 0.000955382658384029,
      "loss": 1.5182,
      "step": 1623
    },
    {
      "epoch": 0.13740587190117606,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009553276464610058,
      "loss": 1.3612,
      "step": 1624
    },
    {
      "epoch": 0.13749048142820883,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0009552726022306192,
      "loss": 1.174,
      "step": 1625
    },
    {
      "epoch": 0.13757509095524156,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009552175256967748,
      "loss": 1.3092,
      "step": 1626
    },
    {
      "epoch": 0.1376597004822743,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009551624168633806,
      "loss": 1.7415,
      "step": 1627
    },
    {
      "epoch": 0.13774431000930704,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009551072757343467,
      "loss": 1.7089,
      "step": 1628
    },
    {
      "epoch": 0.1378289195363398,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009550521023135856,
      "loss": 1.3249,
      "step": 1629
    },
    {
      "epoch": 0.13791352906337254,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009549968966050122,
      "loss": 1.3095,
      "step": 1630
    },
    {
      "epoch": 0.13799813859040527,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009549416586125434,
      "loss": 1.8808,
      "step": 1631
    },
    {
      "epoch": 0.13808274811743804,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009548863883400986,
      "loss": 1.3694,
      "step": 1632
    },
    {
      "epoch": 0.13816735764447077,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009548310857915996,
      "loss": 1.3032,
      "step": 1633
    },
    {
      "epoch": 0.1382519671715035,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0009547757509709702,
      "loss": 1.1035,
      "step": 1634
    },
    {
      "epoch": 0.13833657669853625,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009547203838821367,
      "loss": 1.1446,
      "step": 1635
    },
    {
      "epoch": 0.138421186225569,
      "grad_norm": 1.53125,
      "learning_rate": 0.0009546649845290276,
      "loss": 1.1907,
      "step": 1636
    },
    {
      "epoch": 0.13850579575260175,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009546095529155736,
      "loss": 1.6539,
      "step": 1637
    },
    {
      "epoch": 0.13859040527963448,
      "grad_norm": 1.390625,
      "learning_rate": 0.000954554089045708,
      "loss": 1.0665,
      "step": 1638
    },
    {
      "epoch": 0.13867501480666722,
      "grad_norm": 1.96875,
      "learning_rate": 0.000954498592923366,
      "loss": 1.5894,
      "step": 1639
    },
    {
      "epoch": 0.13875962433369998,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009544430645524856,
      "loss": 1.6517,
      "step": 1640
    },
    {
      "epoch": 0.13884423386073272,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009543875039370065,
      "loss": 1.572,
      "step": 1641
    },
    {
      "epoch": 0.13892884338776545,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009543319110808708,
      "loss": 2.5625,
      "step": 1642
    },
    {
      "epoch": 0.13901345291479822,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009542762859880232,
      "loss": 1.1857,
      "step": 1643
    },
    {
      "epoch": 0.13909806244183096,
      "grad_norm": 1.71875,
      "learning_rate": 0.0009542206286624106,
      "loss": 1.14,
      "step": 1644
    },
    {
      "epoch": 0.1391826719688637,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0009541649391079823,
      "loss": 1.2491,
      "step": 1645
    },
    {
      "epoch": 0.13926728149589643,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009541092173286894,
      "loss": 1.6809,
      "step": 1646
    },
    {
      "epoch": 0.1393518910229292,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009540534633284854,
      "loss": 1.5627,
      "step": 1647
    },
    {
      "epoch": 0.13943650054996193,
      "grad_norm": 1.5625,
      "learning_rate": 0.0009539976771113268,
      "loss": 1.2508,
      "step": 1648
    },
    {
      "epoch": 0.13952111007699466,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0009539418586811716,
      "loss": 1.5665,
      "step": 1649
    },
    {
      "epoch": 0.1396057196040274,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009538860080419802,
      "loss": 1.8145,
      "step": 1650
    },
    {
      "epoch": 0.13969032913106016,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009538301251977157,
      "loss": 2.0336,
      "step": 1651
    },
    {
      "epoch": 0.1397749386580929,
      "grad_norm": 1.5,
      "learning_rate": 0.0009537742101523431,
      "loss": 1.2069,
      "step": 1652
    },
    {
      "epoch": 0.13985954818512564,
      "grad_norm": 2.125,
      "learning_rate": 0.0009537182629098297,
      "loss": 1.66,
      "step": 1653
    },
    {
      "epoch": 0.1399441577121584,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009536622834741454,
      "loss": 1.6608,
      "step": 1654
    },
    {
      "epoch": 0.14002876723919114,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009536062718492619,
      "loss": 1.3784,
      "step": 1655
    },
    {
      "epoch": 0.14011337676622387,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009535502280391538,
      "loss": 1.2864,
      "step": 1656
    },
    {
      "epoch": 0.1401979862932566,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009534941520477974,
      "loss": 1.3351,
      "step": 1657
    },
    {
      "epoch": 0.14028259582028937,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009534380438791716,
      "loss": 1.1172,
      "step": 1658
    },
    {
      "epoch": 0.1403672053473221,
      "grad_norm": 1.484375,
      "learning_rate": 0.0009533819035372573,
      "loss": 1.3158,
      "step": 1659
    },
    {
      "epoch": 0.14045181487435485,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0009533257310260382,
      "loss": 1.3607,
      "step": 1660
    },
    {
      "epoch": 0.14053642440138758,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0009532695263494998,
      "loss": 1.337,
      "step": 1661
    },
    {
      "epoch": 0.14062103392842035,
      "grad_norm": 3.625,
      "learning_rate": 0.0009532132895116301,
      "loss": 1.711,
      "step": 1662
    },
    {
      "epoch": 0.14070564345545308,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009531570205164192,
      "loss": 1.1428,
      "step": 1663
    },
    {
      "epoch": 0.14079025298248582,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009531007193678599,
      "loss": 1.7447,
      "step": 1664
    },
    {
      "epoch": 0.14087486250951858,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009530443860699468,
      "loss": 1.5005,
      "step": 1665
    },
    {
      "epoch": 0.14095947203655132,
      "grad_norm": 2.578125,
      "learning_rate": 0.000952988020626677,
      "loss": 1.8935,
      "step": 1666
    },
    {
      "epoch": 0.14104408156358406,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009529316230420499,
      "loss": 1.5742,
      "step": 1667
    },
    {
      "epoch": 0.1411286910906168,
      "grad_norm": 2.421875,
      "learning_rate": 0.000952875193320067,
      "loss": 1.5488,
      "step": 1668
    },
    {
      "epoch": 0.14121330061764956,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009528187314647324,
      "loss": 1.9287,
      "step": 1669
    },
    {
      "epoch": 0.1412979101446823,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009527622374800522,
      "loss": 1.3773,
      "step": 1670
    },
    {
      "epoch": 0.14138251967171503,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009527057113700349,
      "loss": 1.9271,
      "step": 1671
    },
    {
      "epoch": 0.14146712919874777,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009526491531386914,
      "loss": 1.8381,
      "step": 1672
    },
    {
      "epoch": 0.14155173872578053,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009525925627900345,
      "loss": 1.3332,
      "step": 1673
    },
    {
      "epoch": 0.14163634825281327,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009525359403280795,
      "loss": 1.4558,
      "step": 1674
    },
    {
      "epoch": 0.141720957779846,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009524792857568443,
      "loss": 1.738,
      "step": 1675
    },
    {
      "epoch": 0.14180556730687877,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009524225990803486,
      "loss": 1.4631,
      "step": 1676
    },
    {
      "epoch": 0.1418901768339115,
      "grad_norm": 1.515625,
      "learning_rate": 0.0009523658803026145,
      "loss": 1.3987,
      "step": 1677
    },
    {
      "epoch": 0.14197478636094424,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009523091294276664,
      "loss": 1.3746,
      "step": 1678
    },
    {
      "epoch": 0.14205939588797697,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009522523464595312,
      "loss": 1.5523,
      "step": 1679
    },
    {
      "epoch": 0.14214400541500974,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009521955314022377,
      "loss": 1.9243,
      "step": 1680
    },
    {
      "epoch": 0.14222861494204248,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009521386842598172,
      "loss": 1.4206,
      "step": 1681
    },
    {
      "epoch": 0.1423132244690752,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0009520818050363034,
      "loss": 1.2447,
      "step": 1682
    },
    {
      "epoch": 0.14239783399610798,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009520248937357319,
      "loss": 1.6907,
      "step": 1683
    },
    {
      "epoch": 0.1424824435231407,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009519679503621408,
      "loss": 1.4386,
      "step": 1684
    },
    {
      "epoch": 0.14256705305017345,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009519109749195707,
      "loss": 1.481,
      "step": 1685
    },
    {
      "epoch": 0.14265166257720618,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009518539674120639,
      "loss": 1.5913,
      "step": 1686
    },
    {
      "epoch": 0.14273627210423895,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009517969278436656,
      "loss": 2.0174,
      "step": 1687
    },
    {
      "epoch": 0.14282088163127168,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0009517398562184228,
      "loss": 1.2917,
      "step": 1688
    },
    {
      "epoch": 0.14290549115830442,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009516827525403852,
      "loss": 1.8857,
      "step": 1689
    },
    {
      "epoch": 0.14299010068533716,
      "grad_norm": 1.515625,
      "learning_rate": 0.0009516256168136043,
      "loss": 1.0799,
      "step": 1690
    },
    {
      "epoch": 0.14307471021236992,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009515684490421342,
      "loss": 1.7626,
      "step": 1691
    },
    {
      "epoch": 0.14315931973940266,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009515112492300313,
      "loss": 1.9574,
      "step": 1692
    },
    {
      "epoch": 0.1432439292664354,
      "grad_norm": 1.8125,
      "learning_rate": 0.000951454017381354,
      "loss": 1.2316,
      "step": 1693
    },
    {
      "epoch": 0.14332853879346816,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009513967535001631,
      "loss": 1.6454,
      "step": 1694
    },
    {
      "epoch": 0.1434131483205009,
      "grad_norm": 1.75,
      "learning_rate": 0.0009513394575905217,
      "loss": 1.6778,
      "step": 1695
    },
    {
      "epoch": 0.14349775784753363,
      "grad_norm": 1.515625,
      "learning_rate": 0.0009512821296564955,
      "loss": 1.2004,
      "step": 1696
    },
    {
      "epoch": 0.14358236737456637,
      "grad_norm": 1.375,
      "learning_rate": 0.0009512247697021519,
      "loss": 1.0097,
      "step": 1697
    },
    {
      "epoch": 0.14366697690159913,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009511673777315608,
      "loss": 1.8296,
      "step": 1698
    },
    {
      "epoch": 0.14375158642863187,
      "grad_norm": 1.40625,
      "learning_rate": 0.0009511099537487944,
      "loss": 1.3419,
      "step": 1699
    },
    {
      "epoch": 0.1438361959556646,
      "grad_norm": 1.515625,
      "learning_rate": 0.0009510524977579273,
      "loss": 1.1729,
      "step": 1700
    },
    {
      "epoch": 0.14392080548269734,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0009509950097630361,
      "loss": 1.3179,
      "step": 1701
    },
    {
      "epoch": 0.1440054150097301,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009509374897682,
      "loss": 1.7018,
      "step": 1702
    },
    {
      "epoch": 0.14409002453676284,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009508799377774999,
      "loss": 1.2689,
      "step": 1703
    },
    {
      "epoch": 0.14417463406379558,
      "grad_norm": 1.421875,
      "learning_rate": 0.0009508223537950198,
      "loss": 1.2131,
      "step": 1704
    },
    {
      "epoch": 0.14425924359082834,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009507647378248452,
      "loss": 1.2194,
      "step": 1705
    },
    {
      "epoch": 0.14434385311786108,
      "grad_norm": 1.625,
      "learning_rate": 0.0009507070898710642,
      "loss": 1.3081,
      "step": 1706
    },
    {
      "epoch": 0.1444284626448938,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009506494099377674,
      "loss": 1.4638,
      "step": 1707
    },
    {
      "epoch": 0.14451307217192655,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009505916980290473,
      "loss": 2.0841,
      "step": 1708
    },
    {
      "epoch": 0.1445976816989593,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009505339541489988,
      "loss": 2.1747,
      "step": 1709
    },
    {
      "epoch": 0.14468229122599205,
      "grad_norm": 6.0625,
      "learning_rate": 0.000950476178301719,
      "loss": 2.0709,
      "step": 1710
    },
    {
      "epoch": 0.14476690075302479,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009504183704913074,
      "loss": 1.538,
      "step": 1711
    },
    {
      "epoch": 0.14485151028005752,
      "grad_norm": 1.71875,
      "learning_rate": 0.0009503605307218657,
      "loss": 1.3979,
      "step": 1712
    },
    {
      "epoch": 0.14493611980709029,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009503026589974977,
      "loss": 2.109,
      "step": 1713
    },
    {
      "epoch": 0.14502072933412302,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00095024475532231,
      "loss": 1.3838,
      "step": 1714
    },
    {
      "epoch": 0.14510533886115576,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009501868197004106,
      "loss": 1.769,
      "step": 1715
    },
    {
      "epoch": 0.14518994838818852,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009501288521359108,
      "loss": 2.2404,
      "step": 1716
    },
    {
      "epoch": 0.14527455791522126,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0009500708526329234,
      "loss": 1.6782,
      "step": 1717
    },
    {
      "epoch": 0.145359167442254,
      "grad_norm": 1.75,
      "learning_rate": 0.0009500128211955637,
      "loss": 1.6564,
      "step": 1718
    },
    {
      "epoch": 0.14544377696928673,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0009499547578279493,
      "loss": 1.2327,
      "step": 1719
    },
    {
      "epoch": 0.1455283864963195,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009498966625341998,
      "loss": 1.0971,
      "step": 1720
    },
    {
      "epoch": 0.14561299602335223,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009498385353184375,
      "loss": 1.5503,
      "step": 1721
    },
    {
      "epoch": 0.14569760555038497,
      "grad_norm": 2.328125,
      "learning_rate": 0.000949780376184787,
      "loss": 1.5708,
      "step": 1722
    },
    {
      "epoch": 0.1457822150774177,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009497221851373745,
      "loss": 1.6788,
      "step": 1723
    },
    {
      "epoch": 0.14586682460445047,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0009496639621803292,
      "loss": 1.2563,
      "step": 1724
    },
    {
      "epoch": 0.1459514341314832,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009496057073177822,
      "loss": 1.1107,
      "step": 1725
    },
    {
      "epoch": 0.14603604365851594,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009495474205538668,
      "loss": 1.5067,
      "step": 1726
    },
    {
      "epoch": 0.1461206531855487,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009494891018927186,
      "loss": 1.4078,
      "step": 1727
    },
    {
      "epoch": 0.14620526271258144,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009494307513384759,
      "loss": 1.2332,
      "step": 1728
    },
    {
      "epoch": 0.14628987223961418,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009493723688952787,
      "loss": 1.1935,
      "step": 1729
    },
    {
      "epoch": 0.1463744817666469,
      "grad_norm": 1.796875,
      "learning_rate": 0.0009493139545672693,
      "loss": 1.6738,
      "step": 1730
    },
    {
      "epoch": 0.14645909129367968,
      "grad_norm": 1.484375,
      "learning_rate": 0.0009492555083585927,
      "loss": 1.7654,
      "step": 1731
    },
    {
      "epoch": 0.1465437008207124,
      "grad_norm": 2.15625,
      "learning_rate": 0.000949197030273396,
      "loss": 2.1815,
      "step": 1732
    },
    {
      "epoch": 0.14662831034774515,
      "grad_norm": 1.984375,
      "learning_rate": 0.000949138520315828,
      "loss": 1.4195,
      "step": 1733
    },
    {
      "epoch": 0.1467129198747779,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009490799784900408,
      "loss": 1.4145,
      "step": 1734
    },
    {
      "epoch": 0.14679752940181065,
      "grad_norm": 1.390625,
      "learning_rate": 0.0009490214048001877,
      "loss": 1.3291,
      "step": 1735
    },
    {
      "epoch": 0.1468821389288434,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0009489627992504249,
      "loss": 1.8131,
      "step": 1736
    },
    {
      "epoch": 0.14696674845587612,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009489041618449108,
      "loss": 1.4758,
      "step": 1737
    },
    {
      "epoch": 0.1470513579829089,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0009488454925878059,
      "loss": 1.1145,
      "step": 1738
    },
    {
      "epoch": 0.14713596750994162,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009487867914832729,
      "loss": 1.4255,
      "step": 1739
    },
    {
      "epoch": 0.14722057703697436,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009487280585354772,
      "loss": 1.2368,
      "step": 1740
    },
    {
      "epoch": 0.1473051865640071,
      "grad_norm": 25.375,
      "learning_rate": 0.0009486692937485858,
      "loss": 1.6871,
      "step": 1741
    },
    {
      "epoch": 0.14738979609103986,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0009486104971267684,
      "loss": 1.1487,
      "step": 1742
    },
    {
      "epoch": 0.1474744056180726,
      "grad_norm": 2.421875,
      "learning_rate": 0.000948551668674197,
      "loss": 1.804,
      "step": 1743
    },
    {
      "epoch": 0.14755901514510533,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009484928083950457,
      "loss": 1.5622,
      "step": 1744
    },
    {
      "epoch": 0.1476436246721381,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009484339162934908,
      "loss": 1.5271,
      "step": 1745
    },
    {
      "epoch": 0.14772823419917083,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009483749923737109,
      "loss": 1.5016,
      "step": 1746
    },
    {
      "epoch": 0.14781284372620357,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009483160366398868,
      "loss": 1.2014,
      "step": 1747
    },
    {
      "epoch": 0.1478974532532363,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009482570490962019,
      "loss": 1.4114,
      "step": 1748
    },
    {
      "epoch": 0.14798206278026907,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009481980297468415,
      "loss": 1.3073,
      "step": 1749
    },
    {
      "epoch": 0.1480666723073018,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009481389785959933,
      "loss": 1.4749,
      "step": 1750
    },
    {
      "epoch": 0.14815128183433454,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009480798956478471,
      "loss": 1.3195,
      "step": 1751
    },
    {
      "epoch": 0.14823589136136728,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009480207809065953,
      "loss": 1.7721,
      "step": 1752
    },
    {
      "epoch": 0.14832050088840004,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000947961634376432,
      "loss": 1.0431,
      "step": 1753
    },
    {
      "epoch": 0.14840511041543278,
      "grad_norm": 1.4375,
      "learning_rate": 0.0009479024560615541,
      "loss": 1.2831,
      "step": 1754
    },
    {
      "epoch": 0.14848971994246551,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009478432459661607,
      "loss": 1.6614,
      "step": 1755
    },
    {
      "epoch": 0.14857432946949828,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009477840040944528,
      "loss": 1.2942,
      "step": 1756
    },
    {
      "epoch": 0.14865893899653101,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009477247304506338,
      "loss": 1.3369,
      "step": 1757
    },
    {
      "epoch": 0.14874354852356375,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009476654250389094,
      "loss": 1.2856,
      "step": 1758
    },
    {
      "epoch": 0.1488281580505965,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009476060878634878,
      "loss": 1.6847,
      "step": 1759
    },
    {
      "epoch": 0.14891276757762925,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0009475467189285791,
      "loss": 1.2582,
      "step": 1760
    },
    {
      "epoch": 0.148997377104662,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009474873182383956,
      "loss": 1.2678,
      "step": 1761
    },
    {
      "epoch": 0.14908198663169472,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0009474278857971521,
      "loss": 1.1557,
      "step": 1762
    },
    {
      "epoch": 0.14916659615872746,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009473684216090659,
      "loss": 1.4825,
      "step": 1763
    },
    {
      "epoch": 0.14925120568576022,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009473089256783557,
      "loss": 1.1629,
      "step": 1764
    },
    {
      "epoch": 0.14933581521279296,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009472493980092433,
      "loss": 1.6079,
      "step": 1765
    },
    {
      "epoch": 0.1494204247398257,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009471898386059526,
      "loss": 1.6401,
      "step": 1766
    },
    {
      "epoch": 0.14950503426685846,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000947130247472709,
      "loss": 1.7783,
      "step": 1767
    },
    {
      "epoch": 0.1495896437938912,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009470706246137413,
      "loss": 1.4931,
      "step": 1768
    },
    {
      "epoch": 0.14967425332092393,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0009470109700332797,
      "loss": 1.5906,
      "step": 1769
    },
    {
      "epoch": 0.14975886284795667,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009469512837355572,
      "loss": 1.448,
      "step": 1770
    },
    {
      "epoch": 0.14984347237498943,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0009468915657248083,
      "loss": 1.877,
      "step": 1771
    },
    {
      "epoch": 0.14992808190202217,
      "grad_norm": 1.4375,
      "learning_rate": 0.0009468318160052707,
      "loss": 1.1507,
      "step": 1772
    },
    {
      "epoch": 0.1500126914290549,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009467720345811837,
      "loss": 1.3734,
      "step": 1773
    },
    {
      "epoch": 0.15009730095608764,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009467122214567892,
      "loss": 1.4645,
      "step": 1774
    },
    {
      "epoch": 0.1501819104831204,
      "grad_norm": 1.609375,
      "learning_rate": 0.0009466523766363308,
      "loss": 1.6265,
      "step": 1775
    },
    {
      "epoch": 0.15026652001015314,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0009465925001240551,
      "loss": 1.0501,
      "step": 1776
    },
    {
      "epoch": 0.15035112953718588,
      "grad_norm": 1.5625,
      "learning_rate": 0.0009465325919242107,
      "loss": 1.1689,
      "step": 1777
    },
    {
      "epoch": 0.15043573906421864,
      "grad_norm": 1.515625,
      "learning_rate": 0.0009464726520410479,
      "loss": 1.404,
      "step": 1778
    },
    {
      "epoch": 0.15052034859125138,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009464126804788198,
      "loss": 1.0053,
      "step": 1779
    },
    {
      "epoch": 0.15060495811828412,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009463526772417819,
      "loss": 1.7042,
      "step": 1780
    },
    {
      "epoch": 0.15068956764531685,
      "grad_norm": 1.484375,
      "learning_rate": 0.0009462926423341913,
      "loss": 1.2346,
      "step": 1781
    },
    {
      "epoch": 0.15077417717234962,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0009462325757603081,
      "loss": 1.1178,
      "step": 1782
    },
    {
      "epoch": 0.15085878669938235,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009461724775243938,
      "loss": 1.5126,
      "step": 1783
    },
    {
      "epoch": 0.1509433962264151,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009461123476307132,
      "loss": 1.6105,
      "step": 1784
    },
    {
      "epoch": 0.15102800575344782,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009460521860835322,
      "loss": 1.4037,
      "step": 1785
    },
    {
      "epoch": 0.1511126152804806,
      "grad_norm": 2.25,
      "learning_rate": 0.00094599199288712,
      "loss": 1.8476,
      "step": 1786
    },
    {
      "epoch": 0.15119722480751333,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009459317680457472,
      "loss": 1.6835,
      "step": 1787
    },
    {
      "epoch": 0.15128183433454606,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009458715115636871,
      "loss": 1.0112,
      "step": 1788
    },
    {
      "epoch": 0.15136644386157883,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009458112234452153,
      "loss": 1.4347,
      "step": 1789
    },
    {
      "epoch": 0.15145105338861156,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009457509036946092,
      "loss": 1.6033,
      "step": 1790
    },
    {
      "epoch": 0.1515356629156443,
      "grad_norm": 2.109375,
      "learning_rate": 0.000945690552316149,
      "loss": 2.2909,
      "step": 1791
    },
    {
      "epoch": 0.15162027244267703,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009456301693141167,
      "loss": 1.2648,
      "step": 1792
    },
    {
      "epoch": 0.1517048819697098,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009455697546927968,
      "loss": 1.6219,
      "step": 1793
    },
    {
      "epoch": 0.15178949149674253,
      "grad_norm": 2.125,
      "learning_rate": 0.0009455093084564759,
      "loss": 1.9848,
      "step": 1794
    },
    {
      "epoch": 0.15187410102377527,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009454488306094431,
      "loss": 2.2436,
      "step": 1795
    },
    {
      "epoch": 0.151958710550808,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009453883211559893,
      "loss": 1.5176,
      "step": 1796
    },
    {
      "epoch": 0.15204332007784077,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009453277801004081,
      "loss": 1.2729,
      "step": 1797
    },
    {
      "epoch": 0.1521279296048735,
      "grad_norm": 2.328125,
      "learning_rate": 0.000945267207446995,
      "loss": 1.7422,
      "step": 1798
    },
    {
      "epoch": 0.15221253913190624,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009452066032000479,
      "loss": 1.463,
      "step": 1799
    },
    {
      "epoch": 0.152297148658939,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000945145967363867,
      "loss": 1.1533,
      "step": 1800
    },
    {
      "epoch": 0.15238175818597174,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009450852999427544,
      "loss": 1.1459,
      "step": 1801
    },
    {
      "epoch": 0.15246636771300448,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009450246009410151,
      "loss": 1.5402,
      "step": 1802
    },
    {
      "epoch": 0.15255097724003722,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009449638703629557,
      "loss": 1.6568,
      "step": 1803
    },
    {
      "epoch": 0.15263558676706998,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009449031082128854,
      "loss": 1.4453,
      "step": 1804
    },
    {
      "epoch": 0.15272019629410272,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009448423144951153,
      "loss": 1.4113,
      "step": 1805
    },
    {
      "epoch": 0.15280480582113545,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009447814892139593,
      "loss": 1.3675,
      "step": 1806
    },
    {
      "epoch": 0.15288941534816822,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009447206323737327,
      "loss": 1.1828,
      "step": 1807
    },
    {
      "epoch": 0.15297402487520095,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009446597439787542,
      "loss": 1.2599,
      "step": 1808
    },
    {
      "epoch": 0.1530586344022337,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009445988240333437,
      "loss": 1.6488,
      "step": 1809
    },
    {
      "epoch": 0.15314324392926643,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0009445378725418236,
      "loss": 1.5707,
      "step": 1810
    },
    {
      "epoch": 0.1532278534562992,
      "grad_norm": 2.203125,
      "learning_rate": 0.000944476889508519,
      "loss": 1.8129,
      "step": 1811
    },
    {
      "epoch": 0.15331246298333193,
      "grad_norm": 1.609375,
      "learning_rate": 0.0009444158749377565,
      "loss": 1.3808,
      "step": 1812
    },
    {
      "epoch": 0.15339707251036466,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0009443548288338657,
      "loss": 1.2372,
      "step": 1813
    },
    {
      "epoch": 0.1534816820373974,
      "grad_norm": 5.875,
      "learning_rate": 0.0009442937512011779,
      "loss": 1.1614,
      "step": 1814
    },
    {
      "epoch": 0.15356629156443016,
      "grad_norm": 1.53125,
      "learning_rate": 0.0009442326420440268,
      "loss": 1.2138,
      "step": 1815
    },
    {
      "epoch": 0.1536509010914629,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009441715013667484,
      "loss": 1.5862,
      "step": 1816
    },
    {
      "epoch": 0.15373551061849564,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009441103291736809,
      "loss": 1.1423,
      "step": 1817
    },
    {
      "epoch": 0.1538201201455284,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009440491254691647,
      "loss": 1.1956,
      "step": 1818
    },
    {
      "epoch": 0.15390472967256114,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0009439878902575422,
      "loss": 1.6562,
      "step": 1819
    },
    {
      "epoch": 0.15398933919959387,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009439266235431588,
      "loss": 1.8061,
      "step": 1820
    },
    {
      "epoch": 0.1540739487266266,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009438653253303613,
      "loss": 1.232,
      "step": 1821
    },
    {
      "epoch": 0.15415855825365937,
      "grad_norm": 3.140625,
      "learning_rate": 0.0009438039956234989,
      "loss": 2.5967,
      "step": 1822
    },
    {
      "epoch": 0.1542431677806921,
      "grad_norm": 1.578125,
      "learning_rate": 0.0009437426344269237,
      "loss": 1.1262,
      "step": 1823
    },
    {
      "epoch": 0.15432777730772484,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009436812417449891,
      "loss": 1.4386,
      "step": 1824
    },
    {
      "epoch": 0.15441238683475758,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009436198175820511,
      "loss": 1.4278,
      "step": 1825
    },
    {
      "epoch": 0.15449699636179035,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009435583619424683,
      "loss": 1.1076,
      "step": 1826
    },
    {
      "epoch": 0.15458160588882308,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009434968748306013,
      "loss": 1.9931,
      "step": 1827
    },
    {
      "epoch": 0.15466621541585582,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009434353562508126,
      "loss": 1.5573,
      "step": 1828
    },
    {
      "epoch": 0.15475082494288858,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009433738062074672,
      "loss": 1.5508,
      "step": 1829
    },
    {
      "epoch": 0.15483543446992132,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0009433122247049324,
      "loss": 1.1859,
      "step": 1830
    },
    {
      "epoch": 0.15492004399695405,
      "grad_norm": 2.0,
      "learning_rate": 0.0009432506117475777,
      "loss": 1.5653,
      "step": 1831
    },
    {
      "epoch": 0.1550046535239868,
      "grad_norm": 1.875,
      "learning_rate": 0.0009431889673397747,
      "loss": 1.1332,
      "step": 1832
    },
    {
      "epoch": 0.15508926305101955,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009431272914858974,
      "loss": 1.6936,
      "step": 1833
    },
    {
      "epoch": 0.1551738725780523,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000943065584190322,
      "loss": 1.3503,
      "step": 1834
    },
    {
      "epoch": 0.15525848210508503,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0009430038454574267,
      "loss": 1.1929,
      "step": 1835
    },
    {
      "epoch": 0.15534309163211776,
      "grad_norm": 1.625,
      "learning_rate": 0.0009429420752915923,
      "loss": 1.2787,
      "step": 1836
    },
    {
      "epoch": 0.15542770115915053,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009428802736972015,
      "loss": 1.5332,
      "step": 1837
    },
    {
      "epoch": 0.15551231068618326,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009428184406786395,
      "loss": 1.7973,
      "step": 1838
    },
    {
      "epoch": 0.155596920213216,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009427565762402937,
      "loss": 1.2278,
      "step": 1839
    },
    {
      "epoch": 0.15568152974024876,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009426946803865533,
      "loss": 1.5101,
      "step": 1840
    },
    {
      "epoch": 0.1557661392672815,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009426327531218104,
      "loss": 1.8662,
      "step": 1841
    },
    {
      "epoch": 0.15585074879431424,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0009425707944504588,
      "loss": 1.0973,
      "step": 1842
    },
    {
      "epoch": 0.15593535832134697,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0009425088043768948,
      "loss": 0.9126,
      "step": 1843
    },
    {
      "epoch": 0.15601996784837974,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009424467829055168,
      "loss": 1.6611,
      "step": 1844
    },
    {
      "epoch": 0.15610457737541247,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009423847300407255,
      "loss": 1.6741,
      "step": 1845
    },
    {
      "epoch": 0.1561891869024452,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009423226457869238,
      "loss": 1.2816,
      "step": 1846
    },
    {
      "epoch": 0.15627379642947795,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009422605301485169,
      "loss": 2.3079,
      "step": 1847
    },
    {
      "epoch": 0.1563584059565107,
      "grad_norm": 1.78125,
      "learning_rate": 0.000942198383129912,
      "loss": 1.2403,
      "step": 1848
    },
    {
      "epoch": 0.15644301548354345,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009421362047355189,
      "loss": 1.3749,
      "step": 1849
    },
    {
      "epoch": 0.15652762501057618,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009420739949697493,
      "loss": 1.2933,
      "step": 1850
    },
    {
      "epoch": 0.15661223453760895,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009420117538370172,
      "loss": 1.2073,
      "step": 1851
    },
    {
      "epoch": 0.15669684406464168,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009419494813417389,
      "loss": 1.4225,
      "step": 1852
    },
    {
      "epoch": 0.15678145359167442,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009418871774883328,
      "loss": 1.4623,
      "step": 1853
    },
    {
      "epoch": 0.15686606311870716,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0009418248422812198,
      "loss": 1.7863,
      "step": 1854
    },
    {
      "epoch": 0.15695067264573992,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009417624757248228,
      "loss": 1.6152,
      "step": 1855
    },
    {
      "epoch": 0.15703528217277266,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009417000778235668,
      "loss": 1.3535,
      "step": 1856
    },
    {
      "epoch": 0.1571198916998054,
      "grad_norm": 1.796875,
      "learning_rate": 0.0009416376485818792,
      "loss": 1.8126,
      "step": 1857
    },
    {
      "epoch": 0.15720450122683813,
      "grad_norm": 1.640625,
      "learning_rate": 0.00094157518800419,
      "loss": 1.3791,
      "step": 1858
    },
    {
      "epoch": 0.1572891107538709,
      "grad_norm": 2.375,
      "learning_rate": 0.0009415126960949305,
      "loss": 1.8918,
      "step": 1859
    },
    {
      "epoch": 0.15737372028090363,
      "grad_norm": 2.65625,
      "learning_rate": 0.000941450172858535,
      "loss": 1.2439,
      "step": 1860
    },
    {
      "epoch": 0.15745832980793636,
      "grad_norm": 2.125,
      "learning_rate": 0.0009413876182994399,
      "loss": 1.3569,
      "step": 1861
    },
    {
      "epoch": 0.15754293933496913,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009413250324220835,
      "loss": 1.9869,
      "step": 1862
    },
    {
      "epoch": 0.15762754886200186,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0009412624152309066,
      "loss": 1.2524,
      "step": 1863
    },
    {
      "epoch": 0.1577121583890346,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009411997667303523,
      "loss": 1.0561,
      "step": 1864
    },
    {
      "epoch": 0.15779676791606734,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009411370869248654,
      "loss": 1.6306,
      "step": 1865
    },
    {
      "epoch": 0.1578813774431001,
      "grad_norm": 1.375,
      "learning_rate": 0.0009410743758188936,
      "loss": 1.1667,
      "step": 1866
    },
    {
      "epoch": 0.15796598697013284,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009410116334168864,
      "loss": 1.6263,
      "step": 1867
    },
    {
      "epoch": 0.15805059649716557,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009409488597232955,
      "loss": 1.8814,
      "step": 1868
    },
    {
      "epoch": 0.15813520602419834,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0009408860547425754,
      "loss": 1.4415,
      "step": 1869
    },
    {
      "epoch": 0.15821981555123107,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009408232184791818,
      "loss": 1.8425,
      "step": 1870
    },
    {
      "epoch": 0.1583044250782638,
      "grad_norm": 3.625,
      "learning_rate": 0.0009407603509375737,
      "loss": 2.0668,
      "step": 1871
    },
    {
      "epoch": 0.15838903460529655,
      "grad_norm": 2.5,
      "learning_rate": 0.0009406974521222114,
      "loss": 1.2793,
      "step": 1872
    },
    {
      "epoch": 0.1584736441323293,
      "grad_norm": 1.390625,
      "learning_rate": 0.000940634522037558,
      "loss": 1.2061,
      "step": 1873
    },
    {
      "epoch": 0.15855825365936205,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009405715606880787,
      "loss": 1.673,
      "step": 1874
    },
    {
      "epoch": 0.15864286318639478,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009405085680782409,
      "loss": 1.4777,
      "step": 1875
    },
    {
      "epoch": 0.15872747271342752,
      "grad_norm": 1.8984375,
      "learning_rate": 0.000940445544212514,
      "loss": 1.7374,
      "step": 1876
    },
    {
      "epoch": 0.15881208224046028,
      "grad_norm": 9.6875,
      "learning_rate": 0.0009403824890953697,
      "loss": 1.0062,
      "step": 1877
    },
    {
      "epoch": 0.15889669176749302,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009403194027312825,
      "loss": 1.4799,
      "step": 1878
    },
    {
      "epoch": 0.15898130129452576,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0009402562851247283,
      "loss": 1.492,
      "step": 1879
    },
    {
      "epoch": 0.15906591082155852,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009401931362801855,
      "loss": 1.1135,
      "step": 1880
    },
    {
      "epoch": 0.15915052034859126,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009401299562021348,
      "loss": 1.6576,
      "step": 1881
    },
    {
      "epoch": 0.159235129875624,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009400667448950593,
      "loss": 1.3681,
      "step": 1882
    },
    {
      "epoch": 0.15931973940265673,
      "grad_norm": 2.0,
      "learning_rate": 0.0009400035023634439,
      "loss": 1.3583,
      "step": 1883
    },
    {
      "epoch": 0.1594043489296895,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009399402286117761,
      "loss": 1.2424,
      "step": 1884
    },
    {
      "epoch": 0.15948895845672223,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009398769236445453,
      "loss": 1.7771,
      "step": 1885
    },
    {
      "epoch": 0.15957356798375497,
      "grad_norm": 2.046875,
      "learning_rate": 0.000939813587466243,
      "loss": 1.9298,
      "step": 1886
    },
    {
      "epoch": 0.1596581775107877,
      "grad_norm": 1.625,
      "learning_rate": 0.0009397502200813637,
      "loss": 1.2473,
      "step": 1887
    },
    {
      "epoch": 0.15974278703782047,
      "grad_norm": 1.5625,
      "learning_rate": 0.0009396868214944032,
      "loss": 1.4653,
      "step": 1888
    },
    {
      "epoch": 0.1598273965648532,
      "grad_norm": 1.453125,
      "learning_rate": 0.00093962339170986,
      "loss": 1.3443,
      "step": 1889
    },
    {
      "epoch": 0.15991200609188594,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009395599307322346,
      "loss": 1.6403,
      "step": 1890
    },
    {
      "epoch": 0.1599966156189187,
      "grad_norm": 1.421875,
      "learning_rate": 0.00093949643856603,
      "loss": 1.2417,
      "step": 1891
    },
    {
      "epoch": 0.16008122514595144,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000939432915215751,
      "loss": 1.165,
      "step": 1892
    },
    {
      "epoch": 0.16016583467298418,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009393693606859052,
      "loss": 1.9597,
      "step": 1893
    },
    {
      "epoch": 0.1602504442000169,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009393057749810017,
      "loss": 1.3675,
      "step": 1894
    },
    {
      "epoch": 0.16033505372704968,
      "grad_norm": 1.40625,
      "learning_rate": 0.0009392421581055524,
      "loss": 1.3526,
      "step": 1895
    },
    {
      "epoch": 0.1604196632540824,
      "grad_norm": 1.578125,
      "learning_rate": 0.0009391785100640709,
      "loss": 1.5649,
      "step": 1896
    },
    {
      "epoch": 0.16050427278111515,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009391148308610735,
      "loss": 1.2963,
      "step": 1897
    },
    {
      "epoch": 0.16058888230814788,
      "grad_norm": 1.5625,
      "learning_rate": 0.0009390511205010785,
      "loss": 1.4875,
      "step": 1898
    },
    {
      "epoch": 0.16067349183518065,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009389873789886063,
      "loss": 1.464,
      "step": 1899
    },
    {
      "epoch": 0.16075810136221338,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0009389236063281799,
      "loss": 1.1415,
      "step": 1900
    },
    {
      "epoch": 0.16084271088924612,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009388598025243238,
      "loss": 1.5802,
      "step": 1901
    },
    {
      "epoch": 0.16092732041627889,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009387959675815655,
      "loss": 1.4158,
      "step": 1902
    },
    {
      "epoch": 0.16101192994331162,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009387321015044343,
      "loss": 1.8283,
      "step": 1903
    },
    {
      "epoch": 0.16109653947034436,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009386682042974615,
      "loss": 1.4637,
      "step": 1904
    },
    {
      "epoch": 0.1611811489973771,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009386042759651812,
      "loss": 1.3931,
      "step": 1905
    },
    {
      "epoch": 0.16126575852440986,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0009385403165121291,
      "loss": 1.2017,
      "step": 1906
    },
    {
      "epoch": 0.1613503680514426,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0009384763259428435,
      "loss": 1.7418,
      "step": 1907
    },
    {
      "epoch": 0.16143497757847533,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009384123042618648,
      "loss": 1.6229,
      "step": 1908
    },
    {
      "epoch": 0.16151958710550807,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0009383482514737358,
      "loss": 1.055,
      "step": 1909
    },
    {
      "epoch": 0.16160419663254083,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009382841675830009,
      "loss": 2.5533,
      "step": 1910
    },
    {
      "epoch": 0.16168880615957357,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009382200525942076,
      "loss": 1.4937,
      "step": 1911
    },
    {
      "epoch": 0.1617734156866063,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009381559065119045,
      "loss": 1.4919,
      "step": 1912
    },
    {
      "epoch": 0.16185802521363907,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009380917293406436,
      "loss": 1.7114,
      "step": 1913
    },
    {
      "epoch": 0.1619426347406718,
      "grad_norm": 1.875,
      "learning_rate": 0.0009380275210849782,
      "loss": 1.356,
      "step": 1914
    },
    {
      "epoch": 0.16202724426770454,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009379632817494644,
      "loss": 1.6278,
      "step": 1915
    },
    {
      "epoch": 0.16211185379473728,
      "grad_norm": 1.625,
      "learning_rate": 0.00093789901133866,
      "loss": 1.3118,
      "step": 1916
    },
    {
      "epoch": 0.16219646332177004,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009378347098571254,
      "loss": 1.8237,
      "step": 1917
    },
    {
      "epoch": 0.16228107284880278,
      "grad_norm": 1.421875,
      "learning_rate": 0.0009377703773094228,
      "loss": 1.3333,
      "step": 1918
    },
    {
      "epoch": 0.1623656823758355,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009377060137001173,
      "loss": 1.8687,
      "step": 1919
    },
    {
      "epoch": 0.16245029190286828,
      "grad_norm": 1.53125,
      "learning_rate": 0.0009376416190337754,
      "loss": 1.7222,
      "step": 1920
    },
    {
      "epoch": 0.162534901429901,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009375771933149663,
      "loss": 1.6239,
      "step": 1921
    },
    {
      "epoch": 0.16261951095693375,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0009375127365482613,
      "loss": 1.0085,
      "step": 1922
    },
    {
      "epoch": 0.16270412048396649,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009374482487382338,
      "loss": 1.468,
      "step": 1923
    },
    {
      "epoch": 0.16278873001099925,
      "grad_norm": 1.4375,
      "learning_rate": 0.0009373837298894594,
      "loss": 1.255,
      "step": 1924
    },
    {
      "epoch": 0.16287333953803199,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009373191800065161,
      "loss": 1.6439,
      "step": 1925
    },
    {
      "epoch": 0.16295794906506472,
      "grad_norm": 1.859375,
      "learning_rate": 0.000937254599093984,
      "loss": 1.7379,
      "step": 1926
    },
    {
      "epoch": 0.16304255859209746,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009371899871564454,
      "loss": 1.2702,
      "step": 1927
    },
    {
      "epoch": 0.16312716811913022,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0009371253441984847,
      "loss": 1.0493,
      "step": 1928
    },
    {
      "epoch": 0.16321177764616296,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009370606702246887,
      "loss": 1.3762,
      "step": 1929
    },
    {
      "epoch": 0.1632963871731957,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000936995965239646,
      "loss": 1.2363,
      "step": 1930
    },
    {
      "epoch": 0.16338099670022846,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009369312292479478,
      "loss": 1.4387,
      "step": 1931
    },
    {
      "epoch": 0.1634656062272612,
      "grad_norm": 1.578125,
      "learning_rate": 0.0009368664622541876,
      "loss": 1.2653,
      "step": 1932
    },
    {
      "epoch": 0.16355021575429393,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009368016642629606,
      "loss": 2.1522,
      "step": 1933
    },
    {
      "epoch": 0.16363482528132667,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009367368352788648,
      "loss": 2.1567,
      "step": 1934
    },
    {
      "epoch": 0.16371943480835943,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009366719753064999,
      "loss": 1.3402,
      "step": 1935
    },
    {
      "epoch": 0.16380404433539217,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0009366070843504678,
      "loss": 1.223,
      "step": 1936
    },
    {
      "epoch": 0.1638886538624249,
      "grad_norm": 1.75,
      "learning_rate": 0.0009365421624153732,
      "loss": 1.5243,
      "step": 1937
    },
    {
      "epoch": 0.16397326338945764,
      "grad_norm": 1.796875,
      "learning_rate": 0.0009364772095058221,
      "loss": 1.7981,
      "step": 1938
    },
    {
      "epoch": 0.1640578729164904,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009364122256264235,
      "loss": 2.1738,
      "step": 1939
    },
    {
      "epoch": 0.16414248244352314,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0009363472107817882,
      "loss": 1.8093,
      "step": 1940
    },
    {
      "epoch": 0.16422709197055588,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0009362821649765292,
      "loss": 1.1975,
      "step": 1941
    },
    {
      "epoch": 0.16431170149758864,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009362170882152618,
      "loss": 1.6203,
      "step": 1942
    },
    {
      "epoch": 0.16439631102462138,
      "grad_norm": 1.625,
      "learning_rate": 0.0009361519805026038,
      "loss": 1.3249,
      "step": 1943
    },
    {
      "epoch": 0.16448092055165411,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009360868418431743,
      "loss": 1.5663,
      "step": 1944
    },
    {
      "epoch": 0.16456553007868685,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009360216722415953,
      "loss": 1.1818,
      "step": 1945
    },
    {
      "epoch": 0.16465013960571961,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009359564717024911,
      "loss": 1.611,
      "step": 1946
    },
    {
      "epoch": 0.16473474913275235,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009358912402304877,
      "loss": 1.769,
      "step": 1947
    },
    {
      "epoch": 0.1648193586597851,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0009358259778302137,
      "loss": 1.1229,
      "step": 1948
    },
    {
      "epoch": 0.16490396818681782,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009357606845062996,
      "loss": 1.4662,
      "step": 1949
    },
    {
      "epoch": 0.1649885777138506,
      "grad_norm": 1.5,
      "learning_rate": 0.0009356953602633784,
      "loss": 1.4891,
      "step": 1950
    },
    {
      "epoch": 0.16507318724088332,
      "grad_norm": 2.015625,
      "learning_rate": 0.000935630005106085,
      "loss": 2.1075,
      "step": 1951
    },
    {
      "epoch": 0.16515779676791606,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009355646190390565,
      "loss": 1.5036,
      "step": 1952
    },
    {
      "epoch": 0.16524240629494882,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009354992020669326,
      "loss": 1.2627,
      "step": 1953
    },
    {
      "epoch": 0.16532701582198156,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0009354337541943547,
      "loss": 1.2213,
      "step": 1954
    },
    {
      "epoch": 0.1654116253490143,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0009353682754259667,
      "loss": 1.4819,
      "step": 1955
    },
    {
      "epoch": 0.16549623487604703,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009353027657664146,
      "loss": 1.3307,
      "step": 1956
    },
    {
      "epoch": 0.1655808444030798,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009352372252203464,
      "loss": 1.4658,
      "step": 1957
    },
    {
      "epoch": 0.16566545393011253,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0009351716537924126,
      "loss": 1.1424,
      "step": 1958
    },
    {
      "epoch": 0.16575006345714527,
      "grad_norm": 1.765625,
      "learning_rate": 0.0009351060514872657,
      "loss": 1.2496,
      "step": 1959
    },
    {
      "epoch": 0.165834672984178,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009350404183095605,
      "loss": 1.4414,
      "step": 1960
    },
    {
      "epoch": 0.16591928251121077,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009349747542639539,
      "loss": 1.5162,
      "step": 1961
    },
    {
      "epoch": 0.1660038920382435,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0009349090593551053,
      "loss": 1.1895,
      "step": 1962
    },
    {
      "epoch": 0.16608850156527624,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009348433335876756,
      "loss": 1.4449,
      "step": 1963
    },
    {
      "epoch": 0.166173111092309,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009347775769663286,
      "loss": 1.6672,
      "step": 1964
    },
    {
      "epoch": 0.16625772061934174,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009347117894957298,
      "loss": 1.6937,
      "step": 1965
    },
    {
      "epoch": 0.16634233014637448,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009346459711805472,
      "loss": 1.658,
      "step": 1966
    },
    {
      "epoch": 0.16642693967340721,
      "grad_norm": 1.703125,
      "learning_rate": 0.000934580122025451,
      "loss": 1.461,
      "step": 1967
    },
    {
      "epoch": 0.16651154920043998,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009345142420351134,
      "loss": 1.5498,
      "step": 1968
    },
    {
      "epoch": 0.16659615872747272,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009344483312142087,
      "loss": 2.1397,
      "step": 1969
    },
    {
      "epoch": 0.16668076825450545,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0009343823895674135,
      "loss": 1.5917,
      "step": 1970
    },
    {
      "epoch": 0.1667653777815382,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009343164170994069,
      "loss": 1.6527,
      "step": 1971
    },
    {
      "epoch": 0.16684998730857095,
      "grad_norm": 1.609375,
      "learning_rate": 0.0009342504138148699,
      "loss": 1.1609,
      "step": 1972
    },
    {
      "epoch": 0.1669345968356037,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009341843797184855,
      "loss": 1.6447,
      "step": 1973
    },
    {
      "epoch": 0.16701920636263642,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009341183148149393,
      "loss": 1.4625,
      "step": 1974
    },
    {
      "epoch": 0.1671038158896692,
      "grad_norm": 1.796875,
      "learning_rate": 0.0009340522191089186,
      "loss": 1.754,
      "step": 1975
    },
    {
      "epoch": 0.16718842541670192,
      "grad_norm": 1.875,
      "learning_rate": 0.0009339860926051134,
      "loss": 1.6788,
      "step": 1976
    },
    {
      "epoch": 0.16727303494373466,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009339199353082155,
      "loss": 1.6167,
      "step": 1977
    },
    {
      "epoch": 0.1673576444707674,
      "grad_norm": 1.515625,
      "learning_rate": 0.0009338537472229192,
      "loss": 1.2873,
      "step": 1978
    },
    {
      "epoch": 0.16744225399780016,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009337875283539208,
      "loss": 1.3842,
      "step": 1979
    },
    {
      "epoch": 0.1675268635248329,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009337212787059185,
      "loss": 1.5902,
      "step": 1980
    },
    {
      "epoch": 0.16761147305186563,
      "grad_norm": 2.125,
      "learning_rate": 0.0009336549982836133,
      "loss": 1.3081,
      "step": 1981
    },
    {
      "epoch": 0.1676960825788984,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009335886870917079,
      "loss": 1.4388,
      "step": 1982
    },
    {
      "epoch": 0.16778069210593113,
      "grad_norm": 1.421875,
      "learning_rate": 0.0009335223451349075,
      "loss": 1.1311,
      "step": 1983
    },
    {
      "epoch": 0.16786530163296387,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0009334559724179192,
      "loss": 1.2062,
      "step": 1984
    },
    {
      "epoch": 0.1679499111599966,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009333895689454526,
      "loss": 1.9901,
      "step": 1985
    },
    {
      "epoch": 0.16803452068702937,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009333231347222191,
      "loss": 1.6975,
      "step": 1986
    },
    {
      "epoch": 0.1681191302140621,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009332566697529325,
      "loss": 1.8019,
      "step": 1987
    },
    {
      "epoch": 0.16820373974109484,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000933190174042309,
      "loss": 1.2028,
      "step": 1988
    },
    {
      "epoch": 0.16828834926812758,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0009331236475950664,
      "loss": 1.6539,
      "step": 1989
    },
    {
      "epoch": 0.16837295879516034,
      "grad_norm": 1.875,
      "learning_rate": 0.0009330570904159251,
      "loss": 1.6933,
      "step": 1990
    },
    {
      "epoch": 0.16845756832219308,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0009329905025096078,
      "loss": 1.1401,
      "step": 1991
    },
    {
      "epoch": 0.16854217784922582,
      "grad_norm": 1.5,
      "learning_rate": 0.0009329238838808392,
      "loss": 1.3752,
      "step": 1992
    },
    {
      "epoch": 0.16862678737625858,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009328572345343458,
      "loss": 1.4965,
      "step": 1993
    },
    {
      "epoch": 0.16871139690329132,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009327905544748569,
      "loss": 2.2287,
      "step": 1994
    },
    {
      "epoch": 0.16879600643032405,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0009327238437071038,
      "loss": 1.4141,
      "step": 1995
    },
    {
      "epoch": 0.1688806159573568,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0009326571022358198,
      "loss": 1.1512,
      "step": 1996
    },
    {
      "epoch": 0.16896522548438955,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0009325903300657404,
      "loss": 1.5299,
      "step": 1997
    },
    {
      "epoch": 0.1690498350114223,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0009325235272016035,
      "loss": 1.4144,
      "step": 1998
    },
    {
      "epoch": 0.16913444453845503,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009324566936481491,
      "loss": 1.5284,
      "step": 1999
    },
    {
      "epoch": 0.16921905406548776,
      "grad_norm": 1.46875,
      "learning_rate": 0.0009323898294101191,
      "loss": 1.5575,
      "step": 2000
    },
    {
      "epoch": 0.16930366359252053,
      "grad_norm": 5.75,
      "learning_rate": 0.0009323229344922578,
      "loss": 1.1132,
      "step": 2001
    },
    {
      "epoch": 0.16938827311955326,
      "grad_norm": 2.75,
      "learning_rate": 0.000932256008899312,
      "loss": 2.0506,
      "step": 2002
    },
    {
      "epoch": 0.169472882646586,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0009321890526360298,
      "loss": 1.7991,
      "step": 2003
    },
    {
      "epoch": 0.16955749217361876,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0009321220657071625,
      "loss": 1.358,
      "step": 2004
    },
    {
      "epoch": 0.1696421017006515,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009320550481174628,
      "loss": 2.42,
      "step": 2005
    },
    {
      "epoch": 0.16972671122768423,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009319879998716861,
      "loss": 1.2698,
      "step": 2006
    },
    {
      "epoch": 0.16981132075471697,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009319209209745897,
      "loss": 1.8028,
      "step": 2007
    },
    {
      "epoch": 0.16989593028174974,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009318538114309329,
      "loss": 1.6793,
      "step": 2008
    },
    {
      "epoch": 0.16998053980878247,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009317866712454777,
      "loss": 1.9235,
      "step": 2009
    },
    {
      "epoch": 0.1700651493358152,
      "grad_norm": 1.4375,
      "learning_rate": 0.0009317195004229879,
      "loss": 1.1186,
      "step": 2010
    },
    {
      "epoch": 0.17014975886284794,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009316522989682292,
      "loss": 1.3719,
      "step": 2011
    },
    {
      "epoch": 0.1702343683898807,
      "grad_norm": 1.5,
      "learning_rate": 0.0009315850668859704,
      "loss": 1.1396,
      "step": 2012
    },
    {
      "epoch": 0.17031897791691344,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009315178041809814,
      "loss": 1.1984,
      "step": 2013
    },
    {
      "epoch": 0.17040358744394618,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009314505108580351,
      "loss": 1.3112,
      "step": 2014
    },
    {
      "epoch": 0.17048819697097894,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009313831869219059,
      "loss": 2.1961,
      "step": 2015
    },
    {
      "epoch": 0.17057280649801168,
      "grad_norm": 1.46875,
      "learning_rate": 0.0009313158323773711,
      "loss": 1.1527,
      "step": 2016
    },
    {
      "epoch": 0.17065741602504442,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0009312484472292095,
      "loss": 1.1945,
      "step": 2017
    },
    {
      "epoch": 0.17074202555207715,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009311810314822024,
      "loss": 1.1694,
      "step": 2018
    },
    {
      "epoch": 0.17082663507910992,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009311135851411332,
      "loss": 1.7358,
      "step": 2019
    },
    {
      "epoch": 0.17091124460614265,
      "grad_norm": 1.15625,
      "learning_rate": 0.0009310461082107877,
      "loss": 1.0897,
      "step": 2020
    },
    {
      "epoch": 0.1709958541331754,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0009309786006959535,
      "loss": 1.1786,
      "step": 2021
    },
    {
      "epoch": 0.17108046366020813,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0009309110626014204,
      "loss": 1.2865,
      "step": 2022
    },
    {
      "epoch": 0.1711650731872409,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009308434939319807,
      "loss": 2.0812,
      "step": 2023
    },
    {
      "epoch": 0.17124968271427363,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009307758946924289,
      "loss": 1.6138,
      "step": 2024
    },
    {
      "epoch": 0.17133429224130636,
      "grad_norm": 1.421875,
      "learning_rate": 0.0009307082648875609,
      "loss": 1.3349,
      "step": 2025
    },
    {
      "epoch": 0.17141890176833913,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0009306406045221755,
      "loss": 1.4204,
      "step": 2026
    },
    {
      "epoch": 0.17150351129537186,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009305729136010739,
      "loss": 1.3412,
      "step": 2027
    },
    {
      "epoch": 0.1715881208224046,
      "grad_norm": 1.5,
      "learning_rate": 0.0009305051921290586,
      "loss": 1.1893,
      "step": 2028
    },
    {
      "epoch": 0.17167273034943734,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009304374401109348,
      "loss": 1.89,
      "step": 2029
    },
    {
      "epoch": 0.1717573398764701,
      "grad_norm": 1.5625,
      "learning_rate": 0.0009303696575515097,
      "loss": 1.5132,
      "step": 2030
    },
    {
      "epoch": 0.17184194940350284,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009303018444555929,
      "loss": 1.7629,
      "step": 2031
    },
    {
      "epoch": 0.17192655893053557,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009302340008279961,
      "loss": 1.8209,
      "step": 2032
    },
    {
      "epoch": 0.1720111684575683,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009301661266735328,
      "loss": 1.3558,
      "step": 2033
    },
    {
      "epoch": 0.17209577798460107,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009300982219970194,
      "loss": 1.3058,
      "step": 2034
    },
    {
      "epoch": 0.1721803875116338,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009300302868032735,
      "loss": 1.128,
      "step": 2035
    },
    {
      "epoch": 0.17226499703866655,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009299623210971157,
      "loss": 1.1239,
      "step": 2036
    },
    {
      "epoch": 0.1723496065656993,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0009298943248833683,
      "loss": 1.298,
      "step": 2037
    },
    {
      "epoch": 0.17243421609273205,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0009298262981668559,
      "loss": 1.1881,
      "step": 2038
    },
    {
      "epoch": 0.17251882561976478,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009297582409524054,
      "loss": 1.358,
      "step": 2039
    },
    {
      "epoch": 0.17260343514679752,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009296901532448459,
      "loss": 1.5932,
      "step": 2040
    },
    {
      "epoch": 0.17268804467383028,
      "grad_norm": 1.90625,
      "learning_rate": 0.000929622035049008,
      "loss": 1.4426,
      "step": 2041
    },
    {
      "epoch": 0.17277265420086302,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009295538863697254,
      "loss": 1.5209,
      "step": 2042
    },
    {
      "epoch": 0.17285726372789575,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0009294857072118332,
      "loss": 1.35,
      "step": 2043
    },
    {
      "epoch": 0.17294187325492852,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009294174975801693,
      "loss": 1.5191,
      "step": 2044
    },
    {
      "epoch": 0.17302648278196125,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0009293492574795734,
      "loss": 0.9924,
      "step": 2045
    },
    {
      "epoch": 0.173111092308994,
      "grad_norm": 1.40625,
      "learning_rate": 0.0009292809869148873,
      "loss": 1.2354,
      "step": 2046
    },
    {
      "epoch": 0.17319570183602673,
      "grad_norm": 1.4375,
      "learning_rate": 0.0009292126858909548,
      "loss": 1.0762,
      "step": 2047
    },
    {
      "epoch": 0.1732803113630595,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009291443544126228,
      "loss": 2.0232,
      "step": 2048
    },
    {
      "epoch": 0.17336492089009223,
      "grad_norm": 1.625,
      "learning_rate": 0.0009290759924847393,
      "loss": 1.1327,
      "step": 2049
    },
    {
      "epoch": 0.17344953041712496,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009290076001121548,
      "loss": 1.0436,
      "step": 2050
    },
    {
      "epoch": 0.1735341399441577,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009289391772997223,
      "loss": 1.5653,
      "step": 2051
    },
    {
      "epoch": 0.17361874947119046,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009288707240522962,
      "loss": 1.4876,
      "step": 2052
    },
    {
      "epoch": 0.1737033589982232,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009288022403747342,
      "loss": 1.2523,
      "step": 2053
    },
    {
      "epoch": 0.17378796852525594,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0009287337262718949,
      "loss": 1.4522,
      "step": 2054
    },
    {
      "epoch": 0.1738725780522887,
      "grad_norm": 2.0625,
      "learning_rate": 0.00092866518174864,
      "loss": 1.5914,
      "step": 2055
    },
    {
      "epoch": 0.17395718757932144,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009285966068098328,
      "loss": 1.7319,
      "step": 2056
    },
    {
      "epoch": 0.17404179710635417,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009285280014603392,
      "loss": 1.4253,
      "step": 2057
    },
    {
      "epoch": 0.1741264066333869,
      "grad_norm": 1.609375,
      "learning_rate": 0.0009284593657050269,
      "loss": 1.9715,
      "step": 2058
    },
    {
      "epoch": 0.17421101616041967,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009283906995487659,
      "loss": 1.4717,
      "step": 2059
    },
    {
      "epoch": 0.1742956256874524,
      "grad_norm": 1.484375,
      "learning_rate": 0.0009283220029964283,
      "loss": 1.5017,
      "step": 2060
    },
    {
      "epoch": 0.17438023521448515,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009282532760528884,
      "loss": 1.4316,
      "step": 2061
    },
    {
      "epoch": 0.17446484474151788,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009281845187230228,
      "loss": 1.3302,
      "step": 2062
    },
    {
      "epoch": 0.17454945426855065,
      "grad_norm": 1.5,
      "learning_rate": 0.0009281157310117101,
      "loss": 1.507,
      "step": 2063
    },
    {
      "epoch": 0.17463406379558338,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009280469129238309,
      "loss": 1.5604,
      "step": 2064
    },
    {
      "epoch": 0.17471867332261612,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009279780644642682,
      "loss": 1.864,
      "step": 2065
    },
    {
      "epoch": 0.17480328284964888,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009279091856379072,
      "loss": 1.5318,
      "step": 2066
    },
    {
      "epoch": 0.17488789237668162,
      "grad_norm": 2.90625,
      "learning_rate": 0.000927840276449635,
      "loss": 1.8271,
      "step": 2067
    },
    {
      "epoch": 0.17497250190371436,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000927771336904341,
      "loss": 1.1934,
      "step": 2068
    },
    {
      "epoch": 0.1750571114307471,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009277023670069167,
      "loss": 1.2772,
      "step": 2069
    },
    {
      "epoch": 0.17514172095777986,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009276333667622561,
      "loss": 1.5809,
      "step": 2070
    },
    {
      "epoch": 0.1752263304848126,
      "grad_norm": 1.53125,
      "learning_rate": 0.0009275643361752546,
      "loss": 1.2883,
      "step": 2071
    },
    {
      "epoch": 0.17531094001184533,
      "grad_norm": 1.96875,
      "learning_rate": 0.0009274952752508104,
      "loss": 1.2511,
      "step": 2072
    },
    {
      "epoch": 0.17539554953887806,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009274261839938237,
      "loss": 1.3826,
      "step": 2073
    },
    {
      "epoch": 0.17548015906591083,
      "grad_norm": 1.875,
      "learning_rate": 0.0009273570624091969,
      "loss": 1.6178,
      "step": 2074
    },
    {
      "epoch": 0.17556476859294357,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009272879105018342,
      "loss": 1.1386,
      "step": 2075
    },
    {
      "epoch": 0.1756493781199763,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009272187282766425,
      "loss": 1.3303,
      "step": 2076
    },
    {
      "epoch": 0.17573398764700907,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009271495157385304,
      "loss": 1.7576,
      "step": 2077
    },
    {
      "epoch": 0.1758185971740418,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009270802728924087,
      "loss": 1.4598,
      "step": 2078
    },
    {
      "epoch": 0.17590320670107454,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009270109997431906,
      "loss": 1.4973,
      "step": 2079
    },
    {
      "epoch": 0.17598781622810727,
      "grad_norm": 1.796875,
      "learning_rate": 0.0009269416962957913,
      "loss": 1.6244,
      "step": 2080
    },
    {
      "epoch": 0.17607242575514004,
      "grad_norm": 1.4375,
      "learning_rate": 0.0009268723625551283,
      "loss": 0.9554,
      "step": 2081
    },
    {
      "epoch": 0.17615703528217277,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0009268029985261211,
      "loss": 1.1307,
      "step": 2082
    },
    {
      "epoch": 0.1762416448092055,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009267336042136909,
      "loss": 1.5077,
      "step": 2083
    },
    {
      "epoch": 0.17632625433623825,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009266641796227622,
      "loss": 1.4737,
      "step": 2084
    },
    {
      "epoch": 0.176410863863271,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009265947247582603,
      "loss": 1.5799,
      "step": 2085
    },
    {
      "epoch": 0.17649547339030375,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009265252396251138,
      "loss": 1.4372,
      "step": 2086
    },
    {
      "epoch": 0.17658008291733648,
      "grad_norm": 4.375,
      "learning_rate": 0.0009264557242282527,
      "loss": 1.6497,
      "step": 2087
    },
    {
      "epoch": 0.17666469244436925,
      "grad_norm": 2.125,
      "learning_rate": 0.0009263861785726097,
      "loss": 1.6416,
      "step": 2088
    },
    {
      "epoch": 0.17674930197140198,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009263166026631189,
      "loss": 1.392,
      "step": 2089
    },
    {
      "epoch": 0.17683391149843472,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0009262469965047173,
      "loss": 1.3979,
      "step": 2090
    },
    {
      "epoch": 0.17691852102546746,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009261773601023438,
      "loss": 1.7699,
      "step": 2091
    },
    {
      "epoch": 0.17700313055250022,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009261076934609391,
      "loss": 2.2941,
      "step": 2092
    },
    {
      "epoch": 0.17708774007953296,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009260379965854464,
      "loss": 1.786,
      "step": 2093
    },
    {
      "epoch": 0.1771723496065657,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009259682694808113,
      "loss": 1.4906,
      "step": 2094
    },
    {
      "epoch": 0.17725695913359843,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009258985121519808,
      "loss": 1.3156,
      "step": 2095
    },
    {
      "epoch": 0.1773415686606312,
      "grad_norm": 2.0,
      "learning_rate": 0.0009258287246039047,
      "loss": 1.6285,
      "step": 2096
    },
    {
      "epoch": 0.17742617818766393,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009257589068415349,
      "loss": 1.8433,
      "step": 2097
    },
    {
      "epoch": 0.17751078771469667,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009256890588698248,
      "loss": 1.4639,
      "step": 2098
    },
    {
      "epoch": 0.17759539724172943,
      "grad_norm": 1.390625,
      "learning_rate": 0.0009256191806937306,
      "loss": 1.2467,
      "step": 2099
    },
    {
      "epoch": 0.17768000676876217,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009255492723182106,
      "loss": 1.7057,
      "step": 2100
    },
    {
      "epoch": 0.1777646162957949,
      "grad_norm": 2.234375,
      "learning_rate": 0.000925479333748225,
      "loss": 1.6582,
      "step": 2101
    },
    {
      "epoch": 0.17784922582282764,
      "grad_norm": 1.796875,
      "learning_rate": 0.000925409364988736,
      "loss": 1.4644,
      "step": 2102
    },
    {
      "epoch": 0.1779338353498604,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0009253393660447086,
      "loss": 1.4866,
      "step": 2103
    },
    {
      "epoch": 0.17801844487689314,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009252693369211091,
      "loss": 1.2202,
      "step": 2104
    },
    {
      "epoch": 0.17810305440392588,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009251992776229066,
      "loss": 1.8845,
      "step": 2105
    },
    {
      "epoch": 0.17818766393095864,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000925129188155072,
      "loss": 1.1619,
      "step": 2106
    },
    {
      "epoch": 0.17827227345799138,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0009250590685225783,
      "loss": 1.3261,
      "step": 2107
    },
    {
      "epoch": 0.1783568829850241,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0009249889187304013,
      "loss": 1.1612,
      "step": 2108
    },
    {
      "epoch": 0.17844149251205685,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009249187387835178,
      "loss": 1.1502,
      "step": 2109
    },
    {
      "epoch": 0.1785261020390896,
      "grad_norm": 2.0,
      "learning_rate": 0.0009248485286869077,
      "loss": 1.2041,
      "step": 2110
    },
    {
      "epoch": 0.17861071156612235,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009247782884455524,
      "loss": 2.0468,
      "step": 2111
    },
    {
      "epoch": 0.17869532109315509,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000924708018064436,
      "loss": 1.8602,
      "step": 2112
    },
    {
      "epoch": 0.17877993062018782,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009246377175485445,
      "loss": 1.3335,
      "step": 2113
    },
    {
      "epoch": 0.17886454014722059,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009245673869028659,
      "loss": 1.5982,
      "step": 2114
    },
    {
      "epoch": 0.17894914967425332,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009244970261323903,
      "loss": 1.4255,
      "step": 2115
    },
    {
      "epoch": 0.17903375920128606,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009244266352421105,
      "loss": 1.392,
      "step": 2116
    },
    {
      "epoch": 0.17911836872831882,
      "grad_norm": 3.5,
      "learning_rate": 0.0009243562142370208,
      "loss": 1.4956,
      "step": 2117
    },
    {
      "epoch": 0.17920297825535156,
      "grad_norm": 2.5,
      "learning_rate": 0.0009242857631221175,
      "loss": 1.671,
      "step": 2118
    },
    {
      "epoch": 0.1792875877823843,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009242152819023999,
      "loss": 1.8055,
      "step": 2119
    },
    {
      "epoch": 0.17937219730941703,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009241447705828687,
      "loss": 1.1456,
      "step": 2120
    },
    {
      "epoch": 0.1794568068364498,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009240742291685271,
      "loss": 2.1941,
      "step": 2121
    },
    {
      "epoch": 0.17954141636348253,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009240036576643803,
      "loss": 1.3052,
      "step": 2122
    },
    {
      "epoch": 0.17962602589051527,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009239330560754354,
      "loss": 1.6161,
      "step": 2123
    },
    {
      "epoch": 0.179710635417548,
      "grad_norm": 2.78125,
      "learning_rate": 0.000923862424406702,
      "loss": 1.4737,
      "step": 2124
    },
    {
      "epoch": 0.17979524494458077,
      "grad_norm": 5.0,
      "learning_rate": 0.0009237917626631918,
      "loss": 1.5241,
      "step": 2125
    },
    {
      "epoch": 0.1798798544716135,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009237210708499184,
      "loss": 2.414,
      "step": 2126
    },
    {
      "epoch": 0.17996446399864624,
      "grad_norm": 2.125,
      "learning_rate": 0.000923650348971898,
      "loss": 1.9227,
      "step": 2127
    },
    {
      "epoch": 0.180049073525679,
      "grad_norm": 2.03125,
      "learning_rate": 0.000923579597034148,
      "loss": 1.7176,
      "step": 2128
    },
    {
      "epoch": 0.18013368305271174,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009235088150416892,
      "loss": 1.8486,
      "step": 2129
    },
    {
      "epoch": 0.18021829257974448,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009234380029995435,
      "loss": 1.5925,
      "step": 2130
    },
    {
      "epoch": 0.1803029021067772,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0009233671609127352,
      "loss": 1.3302,
      "step": 2131
    },
    {
      "epoch": 0.18038751163380998,
      "grad_norm": 2.75,
      "learning_rate": 0.0009232962887862912,
      "loss": 1.8536,
      "step": 2132
    },
    {
      "epoch": 0.1804721211608427,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0009232253866252399,
      "loss": 1.1746,
      "step": 2133
    },
    {
      "epoch": 0.18055673068787545,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009231544544346123,
      "loss": 1.2852,
      "step": 2134
    },
    {
      "epoch": 0.18064134021490819,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009230834922194411,
      "loss": 2.2466,
      "step": 2135
    },
    {
      "epoch": 0.18072594974194095,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009230124999847615,
      "loss": 1.1943,
      "step": 2136
    },
    {
      "epoch": 0.1808105592689737,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009229414777356106,
      "loss": 1.3942,
      "step": 2137
    },
    {
      "epoch": 0.18089516879600642,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009228704254770279,
      "loss": 1.5476,
      "step": 2138
    },
    {
      "epoch": 0.1809797783230392,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009227993432140546,
      "loss": 1.5338,
      "step": 2139
    },
    {
      "epoch": 0.18106438785007192,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009227282309517345,
      "loss": 1.5624,
      "step": 2140
    },
    {
      "epoch": 0.18114899737710466,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009226570886951131,
      "loss": 1.4732,
      "step": 2141
    },
    {
      "epoch": 0.1812336069041374,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009225859164492384,
      "loss": 1.6608,
      "step": 2142
    },
    {
      "epoch": 0.18131821643117016,
      "grad_norm": 1.484375,
      "learning_rate": 0.0009225147142191603,
      "loss": 1.3427,
      "step": 2143
    },
    {
      "epoch": 0.1814028259582029,
      "grad_norm": 1.8515625,
      "learning_rate": 0.000922443482009931,
      "loss": 1.2796,
      "step": 2144
    },
    {
      "epoch": 0.18148743548523563,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0009223722198266046,
      "loss": 1.7606,
      "step": 2145
    },
    {
      "epoch": 0.18157204501226837,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009223009276742374,
      "loss": 1.4295,
      "step": 2146
    },
    {
      "epoch": 0.18165665453930113,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009222296055578879,
      "loss": 1.2771,
      "step": 2147
    },
    {
      "epoch": 0.18174126406633387,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0009221582534826167,
      "loss": 1.0887,
      "step": 2148
    },
    {
      "epoch": 0.1818258735933666,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009220868714534867,
      "loss": 1.5068,
      "step": 2149
    },
    {
      "epoch": 0.18191048312039937,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0009220154594755625,
      "loss": 1.5045,
      "step": 2150
    },
    {
      "epoch": 0.1819950926474321,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009219440175539112,
      "loss": 1.8064,
      "step": 2151
    },
    {
      "epoch": 0.18207970217446484,
      "grad_norm": 1.046875,
      "learning_rate": 0.0009218725456936019,
      "loss": 0.9804,
      "step": 2152
    },
    {
      "epoch": 0.18216431170149758,
      "grad_norm": 1.578125,
      "learning_rate": 0.0009218010438997058,
      "loss": 1.4257,
      "step": 2153
    },
    {
      "epoch": 0.18224892122853034,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009217295121772962,
      "loss": 1.5343,
      "step": 2154
    },
    {
      "epoch": 0.18233353075556308,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009216579505314487,
      "loss": 1.1877,
      "step": 2155
    },
    {
      "epoch": 0.18241814028259581,
      "grad_norm": 1.578125,
      "learning_rate": 0.0009215863589672407,
      "loss": 1.1004,
      "step": 2156
    },
    {
      "epoch": 0.18250274980962855,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0009215147374897519,
      "loss": 1.4609,
      "step": 2157
    },
    {
      "epoch": 0.18258735933666131,
      "grad_norm": 1.71875,
      "learning_rate": 0.0009214430861040644,
      "loss": 1.1618,
      "step": 2158
    },
    {
      "epoch": 0.18267196886369405,
      "grad_norm": 1.640625,
      "learning_rate": 0.000921371404815262,
      "loss": 1.584,
      "step": 2159
    },
    {
      "epoch": 0.1827565783907268,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009212996936284306,
      "loss": 1.4214,
      "step": 2160
    },
    {
      "epoch": 0.18284118791775955,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0009212279525486589,
      "loss": 1.6114,
      "step": 2161
    },
    {
      "epoch": 0.1829257974447923,
      "grad_norm": 1.53125,
      "learning_rate": 0.0009211561815810368,
      "loss": 1.1864,
      "step": 2162
    },
    {
      "epoch": 0.18301040697182502,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009210843807306567,
      "loss": 1.6037,
      "step": 2163
    },
    {
      "epoch": 0.18309501649885776,
      "grad_norm": 1.875,
      "learning_rate": 0.0009210125500026135,
      "loss": 1.5924,
      "step": 2164
    },
    {
      "epoch": 0.18317962602589052,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009209406894020036,
      "loss": 1.7134,
      "step": 2165
    },
    {
      "epoch": 0.18326423555292326,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0009208687989339261,
      "loss": 1.0128,
      "step": 2166
    },
    {
      "epoch": 0.183348845079956,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0009207968786034815,
      "loss": 1.1552,
      "step": 2167
    },
    {
      "epoch": 0.18343345460698876,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009207249284157732,
      "loss": 1.3322,
      "step": 2168
    },
    {
      "epoch": 0.1835180641340215,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0009206529483759063,
      "loss": 1.1785,
      "step": 2169
    },
    {
      "epoch": 0.18360267366105423,
      "grad_norm": 2.1875,
      "learning_rate": 0.000920580938488988,
      "loss": 1.2554,
      "step": 2170
    },
    {
      "epoch": 0.18368728318808697,
      "grad_norm": 7.59375,
      "learning_rate": 0.0009205088987601275,
      "loss": 1.3092,
      "step": 2171
    },
    {
      "epoch": 0.18377189271511973,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009204368291944368,
      "loss": 1.4937,
      "step": 2172
    },
    {
      "epoch": 0.18385650224215247,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009203647297970292,
      "loss": 1.39,
      "step": 2173
    },
    {
      "epoch": 0.1839411117691852,
      "grad_norm": 1.375,
      "learning_rate": 0.0009202926005730205,
      "loss": 1.0359,
      "step": 2174
    },
    {
      "epoch": 0.18402572129621794,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0009202204415275286,
      "loss": 1.3274,
      "step": 2175
    },
    {
      "epoch": 0.1841103308232507,
      "grad_norm": 1.515625,
      "learning_rate": 0.0009201482526656734,
      "loss": 1.4137,
      "step": 2176
    },
    {
      "epoch": 0.18419494035028344,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009200760339925771,
      "loss": 1.5318,
      "step": 2177
    },
    {
      "epoch": 0.18427954987731618,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009200037855133638,
      "loss": 1.3045,
      "step": 2178
    },
    {
      "epoch": 0.18436415940434894,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009199315072331599,
      "loss": 1.282,
      "step": 2179
    },
    {
      "epoch": 0.18444876893138168,
      "grad_norm": 1.75,
      "learning_rate": 0.0009198591991570938,
      "loss": 1.7231,
      "step": 2180
    },
    {
      "epoch": 0.18453337845841442,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009197868612902962,
      "loss": 1.4484,
      "step": 2181
    },
    {
      "epoch": 0.18461798798544715,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009197144936378996,
      "loss": 1.5176,
      "step": 2182
    },
    {
      "epoch": 0.18470259751247992,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0009196420962050386,
      "loss": 1.2198,
      "step": 2183
    },
    {
      "epoch": 0.18478720703951265,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009195696689968507,
      "loss": 1.5066,
      "step": 2184
    },
    {
      "epoch": 0.1848718165665454,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009194972120184744,
      "loss": 1.7683,
      "step": 2185
    },
    {
      "epoch": 0.18495642609357812,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009194247252750509,
      "loss": 1.5985,
      "step": 2186
    },
    {
      "epoch": 0.1850410356206109,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009193522087717234,
      "loss": 1.2666,
      "step": 2187
    },
    {
      "epoch": 0.18512564514764362,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009192796625136374,
      "loss": 1.7045,
      "step": 2188
    },
    {
      "epoch": 0.18521025467467636,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009192070865059406,
      "loss": 1.3567,
      "step": 2189
    },
    {
      "epoch": 0.18529486420170913,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009191344807537818,
      "loss": 1.3978,
      "step": 2190
    },
    {
      "epoch": 0.18537947372874186,
      "grad_norm": 1.34375,
      "learning_rate": 0.0009190618452623134,
      "loss": 1.3272,
      "step": 2191
    },
    {
      "epoch": 0.1854640832557746,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0009189891800366887,
      "loss": 1.1084,
      "step": 2192
    },
    {
      "epoch": 0.18554869278280733,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009189164850820639,
      "loss": 1.3813,
      "step": 2193
    },
    {
      "epoch": 0.1856333023098401,
      "grad_norm": 2.0,
      "learning_rate": 0.000918843760403597,
      "loss": 1.2002,
      "step": 2194
    },
    {
      "epoch": 0.18571791183687283,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009187710060064478,
      "loss": 1.3292,
      "step": 2195
    },
    {
      "epoch": 0.18580252136390557,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0009186982218957789,
      "loss": 1.5023,
      "step": 2196
    },
    {
      "epoch": 0.1858871308909383,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0009186254080767546,
      "loss": 1.1386,
      "step": 2197
    },
    {
      "epoch": 0.18597174041797107,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000918552564554541,
      "loss": 1.3189,
      "step": 2198
    },
    {
      "epoch": 0.1860563499450038,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000918479691334307,
      "loss": 1.4116,
      "step": 2199
    },
    {
      "epoch": 0.18614095947203654,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000918406788421223,
      "loss": 1.254,
      "step": 2200
    },
    {
      "epoch": 0.1862255689990693,
      "grad_norm": 1.375,
      "learning_rate": 0.000918333855820462,
      "loss": 1.2948,
      "step": 2201
    },
    {
      "epoch": 0.18631017852610204,
      "grad_norm": 1.296875,
      "learning_rate": 0.0009182608935371987,
      "loss": 1.1104,
      "step": 2202
    },
    {
      "epoch": 0.18639478805313478,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009181879015766101,
      "loss": 1.6329,
      "step": 2203
    },
    {
      "epoch": 0.18647939758016752,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009181148799438753,
      "loss": 1.7026,
      "step": 2204
    },
    {
      "epoch": 0.18656400710720028,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009180418286441756,
      "loss": 1.207,
      "step": 2205
    },
    {
      "epoch": 0.18664861663423302,
      "grad_norm": 1.875,
      "learning_rate": 0.000917968747682694,
      "loss": 1.4727,
      "step": 2206
    },
    {
      "epoch": 0.18673322616126575,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009178956370646161,
      "loss": 1.1654,
      "step": 2207
    },
    {
      "epoch": 0.1868178356882985,
      "grad_norm": 1.875,
      "learning_rate": 0.0009178224967951294,
      "loss": 1.755,
      "step": 2208
    },
    {
      "epoch": 0.18690244521533125,
      "grad_norm": 1.578125,
      "learning_rate": 0.0009177493268794235,
      "loss": 1.4767,
      "step": 2209
    },
    {
      "epoch": 0.186987054742364,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0009176761273226901,
      "loss": 1.3751,
      "step": 2210
    },
    {
      "epoch": 0.18707166426939673,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009176028981301229,
      "loss": 1.1781,
      "step": 2211
    },
    {
      "epoch": 0.1871562737964295,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009175296393069179,
      "loss": 1.3688,
      "step": 2212
    },
    {
      "epoch": 0.18724088332346223,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0009174563508582731,
      "loss": 1.3899,
      "step": 2213
    },
    {
      "epoch": 0.18732549285049496,
      "grad_norm": 1.484375,
      "learning_rate": 0.0009173830327893886,
      "loss": 1.1216,
      "step": 2214
    },
    {
      "epoch": 0.1874101023775277,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009173096851054667,
      "loss": 1.6513,
      "step": 2215
    },
    {
      "epoch": 0.18749471190456046,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009172363078117117,
      "loss": 1.3796,
      "step": 2216
    },
    {
      "epoch": 0.1875793214315932,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009171629009133297,
      "loss": 1.5884,
      "step": 2217
    },
    {
      "epoch": 0.18766393095862594,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009170894644155299,
      "loss": 1.5429,
      "step": 2218
    },
    {
      "epoch": 0.18774854048565867,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009170159983235222,
      "loss": 1.3887,
      "step": 2219
    },
    {
      "epoch": 0.18783315001269144,
      "grad_norm": 1.5625,
      "learning_rate": 0.0009169425026425197,
      "loss": 1.2938,
      "step": 2220
    },
    {
      "epoch": 0.18791775953972417,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009168689773777373,
      "loss": 1.3356,
      "step": 2221
    },
    {
      "epoch": 0.1880023690667569,
      "grad_norm": 1.609375,
      "learning_rate": 0.0009167954225343918,
      "loss": 1.2224,
      "step": 2222
    },
    {
      "epoch": 0.18808697859378967,
      "grad_norm": 1.5625,
      "learning_rate": 0.000916721838117702,
      "loss": 1.167,
      "step": 2223
    },
    {
      "epoch": 0.1881715881208224,
      "grad_norm": 1.765625,
      "learning_rate": 0.0009166482241328893,
      "loss": 1.3221,
      "step": 2224
    },
    {
      "epoch": 0.18825619764785514,
      "grad_norm": 1.46875,
      "learning_rate": 0.0009165745805851768,
      "loss": 1.2626,
      "step": 2225
    },
    {
      "epoch": 0.18834080717488788,
      "grad_norm": 1.5625,
      "learning_rate": 0.0009165009074797899,
      "loss": 1.1417,
      "step": 2226
    },
    {
      "epoch": 0.18842541670192064,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009164272048219558,
      "loss": 1.8275,
      "step": 2227
    },
    {
      "epoch": 0.18851002622895338,
      "grad_norm": 1.5625,
      "learning_rate": 0.0009163534726169042,
      "loss": 1.1833,
      "step": 2228
    },
    {
      "epoch": 0.18859463575598612,
      "grad_norm": 1.1875,
      "learning_rate": 0.0009162797108698669,
      "loss": 0.9852,
      "step": 2229
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 1.96875,
      "learning_rate": 0.000916205919586077,
      "loss": 1.7494,
      "step": 2230
    },
    {
      "epoch": 0.18876385481005162,
      "grad_norm": 1.484375,
      "learning_rate": 0.0009161320987707708,
      "loss": 1.1872,
      "step": 2231
    },
    {
      "epoch": 0.18884846433708435,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009160582484291862,
      "loss": 1.4756,
      "step": 2232
    },
    {
      "epoch": 0.1889330738641171,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0009159843685665628,
      "loss": 1.1538,
      "step": 2233
    },
    {
      "epoch": 0.18901768339114985,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009159104591881428,
      "loss": 1.3384,
      "step": 2234
    },
    {
      "epoch": 0.1891022929181826,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0009158365202991708,
      "loss": 1.2659,
      "step": 2235
    },
    {
      "epoch": 0.18918690244521533,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009157625519048925,
      "loss": 1.2394,
      "step": 2236
    },
    {
      "epoch": 0.18927151197224806,
      "grad_norm": 1.75,
      "learning_rate": 0.0009156885540105566,
      "loss": 1.9432,
      "step": 2237
    },
    {
      "epoch": 0.18935612149928083,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009156145266214135,
      "loss": 1.1162,
      "step": 2238
    },
    {
      "epoch": 0.18944073102631356,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009155404697427157,
      "loss": 1.5853,
      "step": 2239
    },
    {
      "epoch": 0.1895253405533463,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0009154663833797177,
      "loss": 1.2452,
      "step": 2240
    },
    {
      "epoch": 0.18960995008037906,
      "grad_norm": 1.40625,
      "learning_rate": 0.0009153922675376765,
      "loss": 1.4376,
      "step": 2241
    },
    {
      "epoch": 0.1896945596074118,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0009153181222218507,
      "loss": 0.9616,
      "step": 2242
    },
    {
      "epoch": 0.18977916913444454,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009152439474375014,
      "loss": 1.3223,
      "step": 2243
    },
    {
      "epoch": 0.18986377866147727,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0009151697431898913,
      "loss": 1.4978,
      "step": 2244
    },
    {
      "epoch": 0.18994838818851004,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000915095509484286,
      "loss": 1.107,
      "step": 2245
    },
    {
      "epoch": 0.19003299771554277,
      "grad_norm": 1.75,
      "learning_rate": 0.0009150212463259522,
      "loss": 1.6537,
      "step": 2246
    },
    {
      "epoch": 0.1901176072425755,
      "grad_norm": 1.453125,
      "learning_rate": 0.0009149469537201595,
      "loss": 1.5124,
      "step": 2247
    },
    {
      "epoch": 0.19020221676960825,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009148726316721791,
      "loss": 1.3403,
      "step": 2248
    },
    {
      "epoch": 0.190286826296641,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0009147982801872845,
      "loss": 1.0746,
      "step": 2249
    },
    {
      "epoch": 0.19037143582367375,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009147238992707512,
      "loss": 1.4484,
      "step": 2250
    },
    {
      "epoch": 0.19045604535070648,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009146494889278568,
      "loss": 1.9725,
      "step": 2251
    },
    {
      "epoch": 0.19054065487773925,
      "grad_norm": 1.75,
      "learning_rate": 0.0009145750491638811,
      "loss": 1.4135,
      "step": 2252
    },
    {
      "epoch": 0.19062526440477198,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009145005799841059,
      "loss": 1.7837,
      "step": 2253
    },
    {
      "epoch": 0.19070987393180472,
      "grad_norm": 1.890625,
      "learning_rate": 0.0009144260813938153,
      "loss": 1.3222,
      "step": 2254
    },
    {
      "epoch": 0.19079448345883745,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000914351553398295,
      "loss": 1.2527,
      "step": 2255
    },
    {
      "epoch": 0.19087909298587022,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0009142769960028327,
      "loss": 1.0412,
      "step": 2256
    },
    {
      "epoch": 0.19096370251290296,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009142024092127195,
      "loss": 1.4463,
      "step": 2257
    },
    {
      "epoch": 0.1910483120399357,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0009141277930332469,
      "loss": 1.0304,
      "step": 2258
    },
    {
      "epoch": 0.19113292156696843,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009140531474697094,
      "loss": 1.1613,
      "step": 2259
    },
    {
      "epoch": 0.1912175310940012,
      "grad_norm": 1.484375,
      "learning_rate": 0.0009139784725274037,
      "loss": 1.5207,
      "step": 2260
    },
    {
      "epoch": 0.19130214062103393,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009139037682116278,
      "loss": 1.1336,
      "step": 2261
    },
    {
      "epoch": 0.19138675014806666,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009138290345276825,
      "loss": 1.6996,
      "step": 2262
    },
    {
      "epoch": 0.19147135967509943,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009137542714808707,
      "loss": 1.3926,
      "step": 2263
    },
    {
      "epoch": 0.19155596920213216,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009136794790764967,
      "loss": 1.2748,
      "step": 2264
    },
    {
      "epoch": 0.1916405787291649,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0009136046573198677,
      "loss": 1.5361,
      "step": 2265
    },
    {
      "epoch": 0.19172518825619764,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009135298062162925,
      "loss": 1.8665,
      "step": 2266
    },
    {
      "epoch": 0.1918097977832304,
      "grad_norm": 1.859375,
      "learning_rate": 0.0009134549257710819,
      "loss": 1.736,
      "step": 2267
    },
    {
      "epoch": 0.19189440731026314,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009133800159895493,
      "loss": 1.7561,
      "step": 2268
    },
    {
      "epoch": 0.19197901683729587,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009133050768770096,
      "loss": 1.887,
      "step": 2269
    },
    {
      "epoch": 0.1920636263643286,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009132301084387802,
      "loss": 1.4751,
      "step": 2270
    },
    {
      "epoch": 0.19214823589136137,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009131551106801803,
      "loss": 2.0235,
      "step": 2271
    },
    {
      "epoch": 0.1922328454183941,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0009130800836065313,
      "loss": 1.135,
      "step": 2272
    },
    {
      "epoch": 0.19231745494542685,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009130050272231568,
      "loss": 1.3973,
      "step": 2273
    },
    {
      "epoch": 0.1924020644724596,
      "grad_norm": 1.359375,
      "learning_rate": 0.0009129299415353823,
      "loss": 0.998,
      "step": 2274
    },
    {
      "epoch": 0.19248667399949235,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009128548265485355,
      "loss": 1.2169,
      "step": 2275
    },
    {
      "epoch": 0.19257128352652508,
      "grad_norm": 1.453125,
      "learning_rate": 0.0009127796822679459,
      "loss": 1.475,
      "step": 2276
    },
    {
      "epoch": 0.19265589305355782,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009127045086989453,
      "loss": 1.6614,
      "step": 2277
    },
    {
      "epoch": 0.19274050258059058,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000912629305846868,
      "loss": 1.6123,
      "step": 2278
    },
    {
      "epoch": 0.19282511210762332,
      "grad_norm": 1.625,
      "learning_rate": 0.0009125540737170495,
      "loss": 1.4699,
      "step": 2279
    },
    {
      "epoch": 0.19290972163465606,
      "grad_norm": 1.671875,
      "learning_rate": 0.000912478812314828,
      "loss": 1.2954,
      "step": 2280
    },
    {
      "epoch": 0.1929943311616888,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009124035216455435,
      "loss": 1.7954,
      "step": 2281
    },
    {
      "epoch": 0.19307894068872156,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0009123282017145384,
      "loss": 1.0851,
      "step": 2282
    },
    {
      "epoch": 0.1931635502157543,
      "grad_norm": 1.484375,
      "learning_rate": 0.0009122528525271567,
      "loss": 1.1904,
      "step": 2283
    },
    {
      "epoch": 0.19324815974278703,
      "grad_norm": 1.734375,
      "learning_rate": 0.000912177474088745,
      "loss": 1.602,
      "step": 2284
    },
    {
      "epoch": 0.1933327692698198,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009121020664046516,
      "loss": 1.3934,
      "step": 2285
    },
    {
      "epoch": 0.19341737879685253,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009120266294802269,
      "loss": 1.5267,
      "step": 2286
    },
    {
      "epoch": 0.19350198832388527,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009119511633208234,
      "loss": 2.046,
      "step": 2287
    },
    {
      "epoch": 0.193586597850918,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009118756679317958,
      "loss": 1.4614,
      "step": 2288
    },
    {
      "epoch": 0.19367120737795077,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009118001433185011,
      "loss": 2.3825,
      "step": 2289
    },
    {
      "epoch": 0.1937558169049835,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0009117245894862978,
      "loss": 1.0952,
      "step": 2290
    },
    {
      "epoch": 0.19384042643201624,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009116490064405467,
      "loss": 1.3006,
      "step": 2291
    },
    {
      "epoch": 0.193925035959049,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009115733941866108,
      "loss": 1.9692,
      "step": 2292
    },
    {
      "epoch": 0.19400964548608174,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0009114977527298552,
      "loss": 1.1803,
      "step": 2293
    },
    {
      "epoch": 0.19409425501311448,
      "grad_norm": 1.53125,
      "learning_rate": 0.0009114220820756468,
      "loss": 1.2199,
      "step": 2294
    },
    {
      "epoch": 0.1941788645401472,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009113463822293549,
      "loss": 1.6762,
      "step": 2295
    },
    {
      "epoch": 0.19426347406717998,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009112706531963506,
      "loss": 1.5845,
      "step": 2296
    },
    {
      "epoch": 0.1943480835942127,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0009111948949820073,
      "loss": 1.7383,
      "step": 2297
    },
    {
      "epoch": 0.19443269312124545,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009111191075917003,
      "loss": 1.8204,
      "step": 2298
    },
    {
      "epoch": 0.19451730264827818,
      "grad_norm": 1.71875,
      "learning_rate": 0.0009110432910308068,
      "loss": 1.1609,
      "step": 2299
    },
    {
      "epoch": 0.19460191217531095,
      "grad_norm": 1.453125,
      "learning_rate": 0.0009109674453047068,
      "loss": 1.1249,
      "step": 2300
    },
    {
      "epoch": 0.19468652170234368,
      "grad_norm": 1.5,
      "learning_rate": 0.0009108915704187814,
      "loss": 1.2179,
      "step": 2301
    },
    {
      "epoch": 0.19477113122937642,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009108156663784144,
      "loss": 1.7145,
      "step": 2302
    },
    {
      "epoch": 0.19485574075640918,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0009107397331889915,
      "loss": 1.0961,
      "step": 2303
    },
    {
      "epoch": 0.19494035028344192,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009106637708559006,
      "loss": 1.1866,
      "step": 2304
    },
    {
      "epoch": 0.19502495981047466,
      "grad_norm": 1.296875,
      "learning_rate": 0.0009105877793845314,
      "loss": 1.2544,
      "step": 2305
    },
    {
      "epoch": 0.1951095693375074,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0009105117587802758,
      "loss": 1.1294,
      "step": 2306
    },
    {
      "epoch": 0.19519417886454016,
      "grad_norm": 7.40625,
      "learning_rate": 0.0009104357090485278,
      "loss": 1.5854,
      "step": 2307
    },
    {
      "epoch": 0.1952787883915729,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009103596301946833,
      "loss": 1.2407,
      "step": 2308
    },
    {
      "epoch": 0.19536339791860563,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009102835222241408,
      "loss": 1.5886,
      "step": 2309
    },
    {
      "epoch": 0.19544800744563837,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009102073851423001,
      "loss": 1.5726,
      "step": 2310
    },
    {
      "epoch": 0.19553261697267113,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009101312189545635,
      "loss": 1.0301,
      "step": 2311
    },
    {
      "epoch": 0.19561722649970387,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009100550236663354,
      "loss": 1.3398,
      "step": 2312
    },
    {
      "epoch": 0.1957018360267366,
      "grad_norm": 1.609375,
      "learning_rate": 0.0009099787992830222,
      "loss": 1.185,
      "step": 2313
    },
    {
      "epoch": 0.19578644555376937,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009099025458100321,
      "loss": 1.3896,
      "step": 2314
    },
    {
      "epoch": 0.1958710550808021,
      "grad_norm": 2.28125,
      "learning_rate": 0.000909826263252776,
      "loss": 1.402,
      "step": 2315
    },
    {
      "epoch": 0.19595566460783484,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009097499516166661,
      "loss": 1.1302,
      "step": 2316
    },
    {
      "epoch": 0.19604027413486758,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000909673610907117,
      "loss": 1.0956,
      "step": 2317
    },
    {
      "epoch": 0.19612488366190034,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009095972411295457,
      "loss": 1.7005,
      "step": 2318
    },
    {
      "epoch": 0.19620949318893308,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0009095208422893708,
      "loss": 1.1225,
      "step": 2319
    },
    {
      "epoch": 0.1962941027159658,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0009094444143920131,
      "loss": 1.3376,
      "step": 2320
    },
    {
      "epoch": 0.19637871224299855,
      "grad_norm": 1.328125,
      "learning_rate": 0.0009093679574428954,
      "loss": 1.1678,
      "step": 2321
    },
    {
      "epoch": 0.1964633217700313,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009092914714474427,
      "loss": 1.9783,
      "step": 2322
    },
    {
      "epoch": 0.19654793129706405,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009092149564110818,
      "loss": 1.3787,
      "step": 2323
    },
    {
      "epoch": 0.19663254082409679,
      "grad_norm": 1.359375,
      "learning_rate": 0.0009091384123392424,
      "loss": 1.1309,
      "step": 2324
    },
    {
      "epoch": 0.19671715035112955,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009090618392373548,
      "loss": 1.3368,
      "step": 2325
    },
    {
      "epoch": 0.19680175987816229,
      "grad_norm": 1.421875,
      "learning_rate": 0.0009089852371108527,
      "loss": 1.0897,
      "step": 2326
    },
    {
      "epoch": 0.19688636940519502,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000908908605965171,
      "loss": 1.3994,
      "step": 2327
    },
    {
      "epoch": 0.19697097893222776,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0009088319458057473,
      "loss": 1.257,
      "step": 2328
    },
    {
      "epoch": 0.19705558845926052,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0009087552566380209,
      "loss": 1.1568,
      "step": 2329
    },
    {
      "epoch": 0.19714019798629326,
      "grad_norm": 1.484375,
      "learning_rate": 0.000908678538467433,
      "loss": 1.1905,
      "step": 2330
    },
    {
      "epoch": 0.197224807513326,
      "grad_norm": 1.390625,
      "learning_rate": 0.0009086017912994271,
      "loss": 1.2623,
      "step": 2331
    },
    {
      "epoch": 0.19730941704035873,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000908525015139449,
      "loss": 1.0003,
      "step": 2332
    },
    {
      "epoch": 0.1973940265673915,
      "grad_norm": 1.796875,
      "learning_rate": 0.0009084482099929461,
      "loss": 1.1562,
      "step": 2333
    },
    {
      "epoch": 0.19747863609442423,
      "grad_norm": 2.25,
      "learning_rate": 0.000908371375865368,
      "loss": 1.5529,
      "step": 2334
    },
    {
      "epoch": 0.19756324562145697,
      "grad_norm": 1.46875,
      "learning_rate": 0.0009082945127621664,
      "loss": 1.1999,
      "step": 2335
    },
    {
      "epoch": 0.19764785514848973,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0009082176206887951,
      "loss": 1.5201,
      "step": 2336
    },
    {
      "epoch": 0.19773246467552247,
      "grad_norm": 1.421875,
      "learning_rate": 0.00090814069965071,
      "loss": 1.4584,
      "step": 2337
    },
    {
      "epoch": 0.1978170742025552,
      "grad_norm": 1.375,
      "learning_rate": 0.0009080637496533688,
      "loss": 1.1343,
      "step": 2338
    },
    {
      "epoch": 0.19790168372958794,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009079867707022316,
      "loss": 1.6996,
      "step": 2339
    },
    {
      "epoch": 0.1979862932566207,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009079097628027601,
      "loss": 1.6344,
      "step": 2340
    },
    {
      "epoch": 0.19807090278365344,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0009078327259604186,
      "loss": 1.3483,
      "step": 2341
    },
    {
      "epoch": 0.19815551231068618,
      "grad_norm": 2.0,
      "learning_rate": 0.000907755660180673,
      "loss": 1.6691,
      "step": 2342
    },
    {
      "epoch": 0.1982401218377189,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009076785654689915,
      "loss": 1.2387,
      "step": 2343
    },
    {
      "epoch": 0.19832473136475168,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009076014418308442,
      "loss": 1.8789,
      "step": 2344
    },
    {
      "epoch": 0.1984093408917844,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009075242892717038,
      "loss": 1.1345,
      "step": 2345
    },
    {
      "epoch": 0.19849395041881715,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0009074471077970438,
      "loss": 1.3652,
      "step": 2346
    },
    {
      "epoch": 0.19857855994584991,
      "grad_norm": 1.8125,
      "learning_rate": 0.000907369897412341,
      "loss": 1.1288,
      "step": 2347
    },
    {
      "epoch": 0.19866316947288265,
      "grad_norm": 1.625,
      "learning_rate": 0.0009072926581230738,
      "loss": 1.4516,
      "step": 2348
    },
    {
      "epoch": 0.1987477789999154,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0009072153899347227,
      "loss": 1.4141,
      "step": 2349
    },
    {
      "epoch": 0.19883238852694812,
      "grad_norm": 1.84375,
      "learning_rate": 0.00090713809285277,
      "loss": 1.2366,
      "step": 2350
    },
    {
      "epoch": 0.1989169980539809,
      "grad_norm": 1.84375,
      "learning_rate": 0.0009070607668827004,
      "loss": 1.2034,
      "step": 2351
    },
    {
      "epoch": 0.19900160758101362,
      "grad_norm": 1.375,
      "learning_rate": 0.0009069834120300004,
      "loss": 1.0135,
      "step": 2352
    },
    {
      "epoch": 0.19908621710804636,
      "grad_norm": 1.625,
      "learning_rate": 0.0009069060283001586,
      "loss": 1.5767,
      "step": 2353
    },
    {
      "epoch": 0.19917082663507912,
      "grad_norm": 1.3125,
      "learning_rate": 0.0009068286156986658,
      "loss": 1.2334,
      "step": 2354
    },
    {
      "epoch": 0.19925543616211186,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0009067511742310146,
      "loss": 1.4696,
      "step": 2355
    },
    {
      "epoch": 0.1993400456891446,
      "grad_norm": 1.671875,
      "learning_rate": 0.0009066737039027001,
      "loss": 1.364,
      "step": 2356
    },
    {
      "epoch": 0.19942465521617733,
      "grad_norm": 1.53125,
      "learning_rate": 0.0009065962047192189,
      "loss": 1.2737,
      "step": 2357
    },
    {
      "epoch": 0.1995092647432101,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009065186766860698,
      "loss": 1.9808,
      "step": 2358
    },
    {
      "epoch": 0.19959387427024283,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0009064411198087539,
      "loss": 1.4105,
      "step": 2359
    },
    {
      "epoch": 0.19967848379727557,
      "grad_norm": 1.078125,
      "learning_rate": 0.000906363534092774,
      "loss": 0.9741,
      "step": 2360
    },
    {
      "epoch": 0.1997630933243083,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009062859195436355,
      "loss": 1.4164,
      "step": 2361
    },
    {
      "epoch": 0.19984770285134107,
      "grad_norm": 5.625,
      "learning_rate": 0.0009062082761668451,
      "loss": 1.327,
      "step": 2362
    },
    {
      "epoch": 0.1999323123783738,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000906130603967912,
      "loss": 1.5411,
      "step": 2363
    },
    {
      "epoch": 0.20001692190540654,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0009060529029523474,
      "loss": 1.4187,
      "step": 2364
    },
    {
      "epoch": 0.2001015314324393,
      "grad_norm": 2.0,
      "learning_rate": 0.0009059751731256646,
      "loss": 1.5101,
      "step": 2365
    },
    {
      "epoch": 0.20018614095947204,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009058974144933787,
      "loss": 1.767,
      "step": 2366
    },
    {
      "epoch": 0.20027075048650478,
      "grad_norm": 2.0625,
      "learning_rate": 0.000905819627061007,
      "loss": 1.4327,
      "step": 2367
    },
    {
      "epoch": 0.20035536001353751,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009057418108340689,
      "loss": 1.5096,
      "step": 2368
    },
    {
      "epoch": 0.20043996954057028,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009056639658180859,
      "loss": 1.2079,
      "step": 2369
    },
    {
      "epoch": 0.20052457906760301,
      "grad_norm": 1.21875,
      "learning_rate": 0.0009055860920185811,
      "loss": 0.9657,
      "step": 2370
    },
    {
      "epoch": 0.20060918859463575,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0009055081894410802,
      "loss": 1.193,
      "step": 2371
    },
    {
      "epoch": 0.2006937981216685,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0009054302580911107,
      "loss": 1.249,
      "step": 2372
    },
    {
      "epoch": 0.20077840764870125,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0009053522979742022,
      "loss": 0.9413,
      "step": 2373
    },
    {
      "epoch": 0.200863017175734,
      "grad_norm": 1.671875,
      "learning_rate": 0.000905274309095886,
      "loss": 1.3488,
      "step": 2374
    },
    {
      "epoch": 0.20094762670276672,
      "grad_norm": 1.5,
      "learning_rate": 0.0009051962914616961,
      "loss": 1.1056,
      "step": 2375
    },
    {
      "epoch": 0.2010322362297995,
      "grad_norm": 1.4375,
      "learning_rate": 0.0009051182450771679,
      "loss": 1.1472,
      "step": 2376
    },
    {
      "epoch": 0.20111684575683222,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009050401699478392,
      "loss": 1.6678,
      "step": 2377
    },
    {
      "epoch": 0.20120145528386496,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0009049620660792499,
      "loss": 1.3897,
      "step": 2378
    },
    {
      "epoch": 0.2012860648108977,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009048839334769415,
      "loss": 1.8433,
      "step": 2379
    },
    {
      "epoch": 0.20137067433793046,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009048057721464581,
      "loss": 2.0182,
      "step": 2380
    },
    {
      "epoch": 0.2014552838649632,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009047275820933453,
      "loss": 1.1403,
      "step": 2381
    },
    {
      "epoch": 0.20153989339199593,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0009046493633231514,
      "loss": 1.2296,
      "step": 2382
    },
    {
      "epoch": 0.20162450291902867,
      "grad_norm": 1.46875,
      "learning_rate": 0.0009045711158414258,
      "loss": 1.1582,
      "step": 2383
    },
    {
      "epoch": 0.20170911244606143,
      "grad_norm": 2.046875,
      "learning_rate": 0.000904492839653721,
      "loss": 1.3214,
      "step": 2384
    },
    {
      "epoch": 0.20179372197309417,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009044145347655907,
      "loss": 1.2736,
      "step": 2385
    },
    {
      "epoch": 0.2018783315001269,
      "grad_norm": 2.703125,
      "learning_rate": 0.000904336201182591,
      "loss": 1.4438,
      "step": 2386
    },
    {
      "epoch": 0.20196294102715967,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0009042578389102801,
      "loss": 1.2284,
      "step": 2387
    },
    {
      "epoch": 0.2020475505541924,
      "grad_norm": 1.3125,
      "learning_rate": 0.0009041794479542181,
      "loss": 1.0408,
      "step": 2388
    },
    {
      "epoch": 0.20213216008122514,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009041010283199671,
      "loss": 2.1843,
      "step": 2389
    },
    {
      "epoch": 0.20221676960825788,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009040225800130913,
      "loss": 1.5271,
      "step": 2390
    },
    {
      "epoch": 0.20230137913529064,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009039441030391569,
      "loss": 1.7187,
      "step": 2391
    },
    {
      "epoch": 0.20238598866232338,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0009038655974037322,
      "loss": 1.4733,
      "step": 2392
    },
    {
      "epoch": 0.20247059818935612,
      "grad_norm": 1.578125,
      "learning_rate": 0.0009037870631123875,
      "loss": 0.9981,
      "step": 2393
    },
    {
      "epoch": 0.20255520771638885,
      "grad_norm": 2.140625,
      "learning_rate": 0.000903708500170695,
      "loss": 1.8024,
      "step": 2394
    },
    {
      "epoch": 0.20263981724342162,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0009036299085842294,
      "loss": 1.3167,
      "step": 2395
    },
    {
      "epoch": 0.20272442677045435,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0009035512883585668,
      "loss": 1.3702,
      "step": 2396
    },
    {
      "epoch": 0.2028090362974871,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009034726394992858,
      "loss": 1.3869,
      "step": 2397
    },
    {
      "epoch": 0.20289364582451985,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009033939620119668,
      "loss": 1.4452,
      "step": 2398
    },
    {
      "epoch": 0.2029782553515526,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009033152559021923,
      "loss": 1.3606,
      "step": 2399
    },
    {
      "epoch": 0.20306286487858533,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009032365211755465,
      "loss": 1.7065,
      "step": 2400
    },
    {
      "epoch": 0.20314747440561806,
      "grad_norm": 1.71875,
      "learning_rate": 0.0009031577578376165,
      "loss": 1.2851,
      "step": 2401
    },
    {
      "epoch": 0.20323208393265083,
      "grad_norm": 5.96875,
      "learning_rate": 0.0009030789658939904,
      "loss": 1.1013,
      "step": 2402
    },
    {
      "epoch": 0.20331669345968356,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009030001453502592,
      "loss": 1.013,
      "step": 2403
    },
    {
      "epoch": 0.2034013029867163,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009029212962120154,
      "loss": 1.3218,
      "step": 2404
    },
    {
      "epoch": 0.20348591251374903,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009028424184848536,
      "loss": 1.5475,
      "step": 2405
    },
    {
      "epoch": 0.2035705220407818,
      "grad_norm": 1.265625,
      "learning_rate": 0.0009027635121743705,
      "loss": 0.9247,
      "step": 2406
    },
    {
      "epoch": 0.20365513156781453,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009026845772861649,
      "loss": 1.6982,
      "step": 2407
    },
    {
      "epoch": 0.20373974109484727,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009026056138258375,
      "loss": 1.606,
      "step": 2408
    },
    {
      "epoch": 0.20382435062188003,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009025266217989912,
      "loss": 1.5454,
      "step": 2409
    },
    {
      "epoch": 0.20390896014891277,
      "grad_norm": 1.625,
      "learning_rate": 0.0009024476012112306,
      "loss": 0.9943,
      "step": 2410
    },
    {
      "epoch": 0.2039935696759455,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009023685520681626,
      "loss": 1.9137,
      "step": 2411
    },
    {
      "epoch": 0.20407817920297824,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0009022894743753962,
      "loss": 1.2983,
      "step": 2412
    },
    {
      "epoch": 0.204162788730011,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009022103681385421,
      "loss": 1.4534,
      "step": 2413
    },
    {
      "epoch": 0.20424739825704374,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009021312333632133,
      "loss": 1.778,
      "step": 2414
    },
    {
      "epoch": 0.20433200778407648,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009020520700550249,
      "loss": 1.1399,
      "step": 2415
    },
    {
      "epoch": 0.20441661731110924,
      "grad_norm": 1.5625,
      "learning_rate": 0.0009019728782195936,
      "loss": 1.4993,
      "step": 2416
    },
    {
      "epoch": 0.20450122683814198,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0009018936578625385,
      "loss": 1.0006,
      "step": 2417
    },
    {
      "epoch": 0.20458583636517472,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009018144089894807,
      "loss": 1.7064,
      "step": 2418
    },
    {
      "epoch": 0.20467044589220745,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0009017351316060429,
      "loss": 1.0797,
      "step": 2419
    },
    {
      "epoch": 0.20475505541924022,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009016558257178506,
      "loss": 1.4292,
      "step": 2420
    },
    {
      "epoch": 0.20483966494627295,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0009015764913305305,
      "loss": 1.441,
      "step": 2421
    },
    {
      "epoch": 0.2049242744733057,
      "grad_norm": 1.8515625,
      "learning_rate": 0.000901497128449712,
      "loss": 1.2225,
      "step": 2422
    },
    {
      "epoch": 0.20500888400033843,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009014177370810259,
      "loss": 1.3433,
      "step": 2423
    },
    {
      "epoch": 0.2050934935273712,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009013383172301057,
      "loss": 1.5261,
      "step": 2424
    },
    {
      "epoch": 0.20517810305440393,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009012588689025862,
      "loss": 1.3596,
      "step": 2425
    },
    {
      "epoch": 0.20526271258143666,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009011793921041049,
      "loss": 1.9291,
      "step": 2426
    },
    {
      "epoch": 0.20534732210846943,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009010998868403011,
      "loss": 1.3361,
      "step": 2427
    },
    {
      "epoch": 0.20543193163550216,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0009010203531168155,
      "loss": 1.4515,
      "step": 2428
    },
    {
      "epoch": 0.2055165411625349,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009009407909392917,
      "loss": 1.7527,
      "step": 2429
    },
    {
      "epoch": 0.20560115068956764,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009008612003133749,
      "loss": 1.3654,
      "step": 2430
    },
    {
      "epoch": 0.2056857602166004,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0009007815812447125,
      "loss": 1.5151,
      "step": 2431
    },
    {
      "epoch": 0.20577036974363314,
      "grad_norm": 1.59375,
      "learning_rate": 0.0009007019337389537,
      "loss": 1.9208,
      "step": 2432
    },
    {
      "epoch": 0.20585497927066587,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0009006222578017498,
      "loss": 1.4966,
      "step": 2433
    },
    {
      "epoch": 0.2059395887976986,
      "grad_norm": 1.421875,
      "learning_rate": 0.000900542553438754,
      "loss": 1.5673,
      "step": 2434
    },
    {
      "epoch": 0.20602419832473137,
      "grad_norm": 1.703125,
      "learning_rate": 0.0009004628206556219,
      "loss": 1.3673,
      "step": 2435
    },
    {
      "epoch": 0.2061088078517641,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009003830594580107,
      "loss": 1.4469,
      "step": 2436
    },
    {
      "epoch": 0.20619341737879684,
      "grad_norm": 2.203125,
      "learning_rate": 0.00090030326985158,
      "loss": 2.2375,
      "step": 2437
    },
    {
      "epoch": 0.2062780269058296,
      "grad_norm": 1.6875,
      "learning_rate": 0.0009002234518419909,
      "loss": 1.8203,
      "step": 2438
    },
    {
      "epoch": 0.20636263643286235,
      "grad_norm": 1.421875,
      "learning_rate": 0.000900143605434907,
      "loss": 1.0731,
      "step": 2439
    },
    {
      "epoch": 0.20644724595989508,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0009000637306359938,
      "loss": 1.1951,
      "step": 2440
    },
    {
      "epoch": 0.20653185548692782,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0008999838274509185,
      "loss": 1.1173,
      "step": 2441
    },
    {
      "epoch": 0.20661646501396058,
      "grad_norm": 1.546875,
      "learning_rate": 0.0008999038958853507,
      "loss": 1.7688,
      "step": 2442
    },
    {
      "epoch": 0.20670107454099332,
      "grad_norm": 1.640625,
      "learning_rate": 0.0008998239359449618,
      "loss": 1.3929,
      "step": 2443
    },
    {
      "epoch": 0.20678568406802605,
      "grad_norm": 1.890625,
      "learning_rate": 0.0008997439476354255,
      "loss": 1.383,
      "step": 2444
    },
    {
      "epoch": 0.2068702935950588,
      "grad_norm": 1.453125,
      "learning_rate": 0.0008996639309624171,
      "loss": 1.4021,
      "step": 2445
    },
    {
      "epoch": 0.20695490312209155,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0008995838859316141,
      "loss": 1.8299,
      "step": 2446
    },
    {
      "epoch": 0.2070395126491243,
      "grad_norm": 1.375,
      "learning_rate": 0.0008995038125486961,
      "loss": 1.2027,
      "step": 2447
    },
    {
      "epoch": 0.20712412217615703,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0008994237108193448,
      "loss": 1.5898,
      "step": 2448
    },
    {
      "epoch": 0.2072087317031898,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0008993435807492433,
      "loss": 1.8602,
      "step": 2449
    },
    {
      "epoch": 0.20729334123022253,
      "grad_norm": 1.921875,
      "learning_rate": 0.0008992634223440777,
      "loss": 2.0188,
      "step": 2450
    },
    {
      "epoch": 0.20737795075725526,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000899183235609535,
      "loss": 1.2205,
      "step": 2451
    },
    {
      "epoch": 0.207462560284288,
      "grad_norm": 2.140625,
      "learning_rate": 0.0008991030205513053,
      "loss": 2.3748,
      "step": 2452
    },
    {
      "epoch": 0.20754716981132076,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0008990227771750799,
      "loss": 1.396,
      "step": 2453
    },
    {
      "epoch": 0.2076317793383535,
      "grad_norm": 1.84375,
      "learning_rate": 0.0008989425054865524,
      "loss": 1.8429,
      "step": 2454
    },
    {
      "epoch": 0.20771638886538624,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0008988622054914184,
      "loss": 1.21,
      "step": 2455
    },
    {
      "epoch": 0.20780099839241897,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0008987818771953758,
      "loss": 1.128,
      "step": 2456
    },
    {
      "epoch": 0.20788560791945174,
      "grad_norm": 2.53125,
      "learning_rate": 0.0008987015206041237,
      "loss": 1.3079,
      "step": 2457
    },
    {
      "epoch": 0.20797021744648447,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0008986211357233642,
      "loss": 1.2274,
      "step": 2458
    },
    {
      "epoch": 0.2080548269735172,
      "grad_norm": 2.046875,
      "learning_rate": 0.0008985407225588007,
      "loss": 1.5793,
      "step": 2459
    },
    {
      "epoch": 0.20813943650054997,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0008984602811161388,
      "loss": 1.2449,
      "step": 2460
    },
    {
      "epoch": 0.2082240460275827,
      "grad_norm": 1.65625,
      "learning_rate": 0.0008983798114010864,
      "loss": 1.1036,
      "step": 2461
    },
    {
      "epoch": 0.20830865555461545,
      "grad_norm": 1.578125,
      "learning_rate": 0.0008982993134193532,
      "loss": 1.1015,
      "step": 2462
    },
    {
      "epoch": 0.20839326508164818,
      "grad_norm": 1.890625,
      "learning_rate": 0.0008982187871766503,
      "loss": 1.313,
      "step": 2463
    },
    {
      "epoch": 0.20847787460868095,
      "grad_norm": 2.359375,
      "learning_rate": 0.000898138232678692,
      "loss": 1.7213,
      "step": 2464
    },
    {
      "epoch": 0.20856248413571368,
      "grad_norm": 3.8125,
      "learning_rate": 0.0008980576499311936,
      "loss": 1.7602,
      "step": 2465
    },
    {
      "epoch": 0.20864709366274642,
      "grad_norm": 2.4375,
      "learning_rate": 0.0008979770389398728,
      "loss": 1.6117,
      "step": 2466
    },
    {
      "epoch": 0.20873170318977916,
      "grad_norm": 1.78125,
      "learning_rate": 0.0008978963997104494,
      "loss": 1.198,
      "step": 2467
    },
    {
      "epoch": 0.20881631271681192,
      "grad_norm": 4.03125,
      "learning_rate": 0.0008978157322486452,
      "loss": 1.1206,
      "step": 2468
    },
    {
      "epoch": 0.20890092224384466,
      "grad_norm": 3.984375,
      "learning_rate": 0.0008977350365601838,
      "loss": 1.0643,
      "step": 2469
    },
    {
      "epoch": 0.2089855317708774,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0008976543126507906,
      "loss": 1.1602,
      "step": 2470
    },
    {
      "epoch": 0.20907014129791016,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0008975735605261935,
      "loss": 1.3832,
      "step": 2471
    },
    {
      "epoch": 0.2091547508249429,
      "grad_norm": 1.875,
      "learning_rate": 0.0008974927801921223,
      "loss": 1.4801,
      "step": 2472
    },
    {
      "epoch": 0.20923936035197563,
      "grad_norm": 1.640625,
      "learning_rate": 0.0008974119716543086,
      "loss": 1.05,
      "step": 2473
    },
    {
      "epoch": 0.20932396987900836,
      "grad_norm": 2.25,
      "learning_rate": 0.0008973311349184862,
      "loss": 1.4596,
      "step": 2474
    },
    {
      "epoch": 0.20940857940604113,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0008972502699903907,
      "loss": 1.4191,
      "step": 2475
    },
    {
      "epoch": 0.20949318893307387,
      "grad_norm": 2.0,
      "learning_rate": 0.0008971693768757597,
      "loss": 1.1295,
      "step": 2476
    },
    {
      "epoch": 0.2095777984601066,
      "grad_norm": 1.4375,
      "learning_rate": 0.000897088455580333,
      "loss": 1.293,
      "step": 2477
    },
    {
      "epoch": 0.20966240798713937,
      "grad_norm": 2.125,
      "learning_rate": 0.0008970075061098524,
      "loss": 1.6866,
      "step": 2478
    },
    {
      "epoch": 0.2097470175141721,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0008969265284700614,
      "loss": 1.5655,
      "step": 2479
    },
    {
      "epoch": 0.20983162704120484,
      "grad_norm": 2.0,
      "learning_rate": 0.0008968455226667059,
      "loss": 1.8487,
      "step": 2480
    },
    {
      "epoch": 0.20991623656823757,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0008967644887055333,
      "loss": 1.3777,
      "step": 2481
    },
    {
      "epoch": 0.21000084609527034,
      "grad_norm": 2.640625,
      "learning_rate": 0.0008966834265922936,
      "loss": 1.3945,
      "step": 2482
    },
    {
      "epoch": 0.21008545562230307,
      "grad_norm": 2.140625,
      "learning_rate": 0.0008966023363327385,
      "loss": 1.4827,
      "step": 2483
    },
    {
      "epoch": 0.2101700651493358,
      "grad_norm": 1.578125,
      "learning_rate": 0.0008965212179326214,
      "loss": 1.6236,
      "step": 2484
    },
    {
      "epoch": 0.21025467467636855,
      "grad_norm": 2.171875,
      "learning_rate": 0.0008964400713976982,
      "loss": 1.652,
      "step": 2485
    },
    {
      "epoch": 0.2103392842034013,
      "grad_norm": 1.296875,
      "learning_rate": 0.0008963588967337264,
      "loss": 1.0395,
      "step": 2486
    },
    {
      "epoch": 0.21042389373043405,
      "grad_norm": 2.203125,
      "learning_rate": 0.0008962776939464658,
      "loss": 2.1581,
      "step": 2487
    },
    {
      "epoch": 0.21050850325746678,
      "grad_norm": 2.296875,
      "learning_rate": 0.0008961964630416782,
      "loss": 1.9719,
      "step": 2488
    },
    {
      "epoch": 0.21059311278449955,
      "grad_norm": 1.65625,
      "learning_rate": 0.000896115204025127,
      "loss": 1.2211,
      "step": 2489
    },
    {
      "epoch": 0.21067772231153228,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000896033916902578,
      "loss": 1.1039,
      "step": 2490
    },
    {
      "epoch": 0.21076233183856502,
      "grad_norm": 1.8828125,
      "learning_rate": 0.000895952601679799,
      "loss": 1.3436,
      "step": 2491
    },
    {
      "epoch": 0.21084694136559776,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0008958712583625592,
      "loss": 1.2605,
      "step": 2492
    },
    {
      "epoch": 0.21093155089263052,
      "grad_norm": 1.453125,
      "learning_rate": 0.0008957898869566308,
      "loss": 1.039,
      "step": 2493
    },
    {
      "epoch": 0.21101616041966326,
      "grad_norm": 1.640625,
      "learning_rate": 0.0008957084874677871,
      "loss": 1.3151,
      "step": 2494
    },
    {
      "epoch": 0.211100769946696,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0008956270599018038,
      "loss": 1.4182,
      "step": 2495
    },
    {
      "epoch": 0.21118537947372873,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0008955456042644585,
      "loss": 1.0842,
      "step": 2496
    },
    {
      "epoch": 0.2112699890007615,
      "grad_norm": 3.34375,
      "learning_rate": 0.0008954641205615308,
      "loss": 2.4689,
      "step": 2497
    },
    {
      "epoch": 0.21135459852779423,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0008953826087988023,
      "loss": 1.5552,
      "step": 2498
    },
    {
      "epoch": 0.21143920805482697,
      "grad_norm": 1.65625,
      "learning_rate": 0.0008953010689820567,
      "loss": 1.168,
      "step": 2499
    },
    {
      "epoch": 0.21152381758185973,
      "grad_norm": 1.71875,
      "learning_rate": 0.0008952195011170794,
      "loss": 1.6484,
      "step": 2500
    },
    {
      "epoch": 0.21160842710889247,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0008951379052096582,
      "loss": 1.45,
      "step": 2501
    },
    {
      "epoch": 0.2116930366359252,
      "grad_norm": 2.25,
      "learning_rate": 0.0008950562812655825,
      "loss": 1.5187,
      "step": 2502
    },
    {
      "epoch": 0.21177764616295794,
      "grad_norm": 1.671875,
      "learning_rate": 0.0008949746292906439,
      "loss": 1.4828,
      "step": 2503
    },
    {
      "epoch": 0.2118622556899907,
      "grad_norm": 2.0625,
      "learning_rate": 0.0008948929492906359,
      "loss": 1.8354,
      "step": 2504
    },
    {
      "epoch": 0.21194686521702344,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0008948112412713542,
      "loss": 1.0898,
      "step": 2505
    },
    {
      "epoch": 0.21203147474405618,
      "grad_norm": 1.3125,
      "learning_rate": 0.000894729505238596,
      "loss": 1.2541,
      "step": 2506
    },
    {
      "epoch": 0.2121160842710889,
      "grad_norm": 1.4375,
      "learning_rate": 0.000894647741198161,
      "loss": 1.4018,
      "step": 2507
    },
    {
      "epoch": 0.21220069379812168,
      "grad_norm": 1.328125,
      "learning_rate": 0.0008945659491558509,
      "loss": 1.1878,
      "step": 2508
    },
    {
      "epoch": 0.2122853033251544,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0008944841291174689,
      "loss": 1.1348,
      "step": 2509
    },
    {
      "epoch": 0.21236991285218715,
      "grad_norm": 1.71875,
      "learning_rate": 0.0008944022810888205,
      "loss": 1.6959,
      "step": 2510
    },
    {
      "epoch": 0.2124545223792199,
      "grad_norm": 5.5,
      "learning_rate": 0.0008943204050757132,
      "loss": 1.5868,
      "step": 2511
    },
    {
      "epoch": 0.21253913190625265,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0008942385010839564,
      "loss": 1.4721,
      "step": 2512
    },
    {
      "epoch": 0.21262374143328538,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0008941565691193616,
      "loss": 1.5138,
      "step": 2513
    },
    {
      "epoch": 0.21270835096031812,
      "grad_norm": 14.4375,
      "learning_rate": 0.000894074609187742,
      "loss": 1.3912,
      "step": 2514
    },
    {
      "epoch": 0.21279296048735089,
      "grad_norm": 1.96875,
      "learning_rate": 0.0008939926212949134,
      "loss": 1.7743,
      "step": 2515
    },
    {
      "epoch": 0.21287757001438362,
      "grad_norm": 1.9140625,
      "learning_rate": 0.000893910605446693,
      "loss": 1.8886,
      "step": 2516
    },
    {
      "epoch": 0.21296217954141636,
      "grad_norm": 2.0,
      "learning_rate": 0.0008938285616489,
      "loss": 1.0949,
      "step": 2517
    },
    {
      "epoch": 0.2130467890684491,
      "grad_norm": 1.5625,
      "learning_rate": 0.0008937464899073556,
      "loss": 1.2388,
      "step": 2518
    },
    {
      "epoch": 0.21313139859548186,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0008936643902278836,
      "loss": 1.246,
      "step": 2519
    },
    {
      "epoch": 0.2132160081225146,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000893582262616309,
      "loss": 1.7066,
      "step": 2520
    },
    {
      "epoch": 0.21330061764954733,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0008935001070784591,
      "loss": 1.6349,
      "step": 2521
    },
    {
      "epoch": 0.2133852271765801,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0008934179236201634,
      "loss": 1.3974,
      "step": 2522
    },
    {
      "epoch": 0.21346983670361283,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0008933357122472529,
      "loss": 1.752,
      "step": 2523
    },
    {
      "epoch": 0.21355444623064557,
      "grad_norm": 1.875,
      "learning_rate": 0.000893253472965561,
      "loss": 1.3237,
      "step": 2524
    },
    {
      "epoch": 0.2136390557576783,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0008931712057809228,
      "loss": 1.1395,
      "step": 2525
    },
    {
      "epoch": 0.21372366528471107,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0008930889106991756,
      "loss": 1.2093,
      "step": 2526
    },
    {
      "epoch": 0.2138082748117438,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0008930065877261584,
      "loss": 1.1918,
      "step": 2527
    },
    {
      "epoch": 0.21389288433877654,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0008929242368677124,
      "loss": 1.375,
      "step": 2528
    },
    {
      "epoch": 0.2139774938658093,
      "grad_norm": 1.21875,
      "learning_rate": 0.0008928418581296809,
      "loss": 1.137,
      "step": 2529
    },
    {
      "epoch": 0.21406210339284204,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0008927594515179089,
      "loss": 1.5825,
      "step": 2530
    },
    {
      "epoch": 0.21414671291987478,
      "grad_norm": 1.21875,
      "learning_rate": 0.0008926770170382435,
      "loss": 1.0552,
      "step": 2531
    },
    {
      "epoch": 0.2142313224469075,
      "grad_norm": 1.984375,
      "learning_rate": 0.0008925945546965337,
      "loss": 1.6677,
      "step": 2532
    },
    {
      "epoch": 0.21431593197394028,
      "grad_norm": 1.84375,
      "learning_rate": 0.0008925120644986306,
      "loss": 1.2906,
      "step": 2533
    },
    {
      "epoch": 0.214400541500973,
      "grad_norm": 1.40625,
      "learning_rate": 0.0008924295464503872,
      "loss": 1.2665,
      "step": 2534
    },
    {
      "epoch": 0.21448515102800575,
      "grad_norm": 1.75,
      "learning_rate": 0.0008923470005576586,
      "loss": 1.1124,
      "step": 2535
    },
    {
      "epoch": 0.21456976055503849,
      "grad_norm": 1.65625,
      "learning_rate": 0.0008922644268263016,
      "loss": 1.3918,
      "step": 2536
    },
    {
      "epoch": 0.21465437008207125,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0008921818252621751,
      "loss": 1.1893,
      "step": 2537
    },
    {
      "epoch": 0.21473897960910399,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0008920991958711403,
      "loss": 1.6657,
      "step": 2538
    },
    {
      "epoch": 0.21482358913613672,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00089201653865906,
      "loss": 1.5966,
      "step": 2539
    },
    {
      "epoch": 0.2149081986631695,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0008919338536317988,
      "loss": 1.5345,
      "step": 2540
    },
    {
      "epoch": 0.21499280819020222,
      "grad_norm": 1.859375,
      "learning_rate": 0.0008918511407952239,
      "loss": 1.5964,
      "step": 2541
    },
    {
      "epoch": 0.21507741771723496,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0008917684001552038,
      "loss": 1.861,
      "step": 2542
    },
    {
      "epoch": 0.2151620272442677,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0008916856317176097,
      "loss": 1.567,
      "step": 2543
    },
    {
      "epoch": 0.21524663677130046,
      "grad_norm": 1.953125,
      "learning_rate": 0.0008916028354883138,
      "loss": 1.3746,
      "step": 2544
    },
    {
      "epoch": 0.2153312462983332,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0008915200114731911,
      "loss": 1.8498,
      "step": 2545
    },
    {
      "epoch": 0.21541585582536593,
      "grad_norm": 1.546875,
      "learning_rate": 0.0008914371596781186,
      "loss": 1.2825,
      "step": 2546
    },
    {
      "epoch": 0.21550046535239867,
      "grad_norm": 1.5625,
      "learning_rate": 0.0008913542801089747,
      "loss": 1.2897,
      "step": 2547
    },
    {
      "epoch": 0.21558507487943143,
      "grad_norm": 1.359375,
      "learning_rate": 0.00089127137277164,
      "loss": 1.2085,
      "step": 2548
    },
    {
      "epoch": 0.21566968440646417,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0008911884376719972,
      "loss": 1.5559,
      "step": 2549
    },
    {
      "epoch": 0.2157542939334969,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0008911054748159307,
      "loss": 1.7889,
      "step": 2550
    },
    {
      "epoch": 0.21583890346052967,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0008910224842093274,
      "loss": 1.2688,
      "step": 2551
    },
    {
      "epoch": 0.2159235129875624,
      "grad_norm": 1.46875,
      "learning_rate": 0.0008909394658580756,
      "loss": 1.33,
      "step": 2552
    },
    {
      "epoch": 0.21600812251459514,
      "grad_norm": 1.546875,
      "learning_rate": 0.0008908564197680658,
      "loss": 1.4977,
      "step": 2553
    },
    {
      "epoch": 0.21609273204162788,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0008907733459451905,
      "loss": 1.427,
      "step": 2554
    },
    {
      "epoch": 0.21617734156866064,
      "grad_norm": 1.59375,
      "learning_rate": 0.0008906902443953442,
      "loss": 1.4393,
      "step": 2555
    },
    {
      "epoch": 0.21626195109569338,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0008906071151244231,
      "loss": 1.4365,
      "step": 2556
    },
    {
      "epoch": 0.21634656062272611,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0008905239581383257,
      "loss": 1.0837,
      "step": 2557
    },
    {
      "epoch": 0.21643117014975885,
      "grad_norm": 1.515625,
      "learning_rate": 0.0008904407734429523,
      "loss": 1.9066,
      "step": 2558
    },
    {
      "epoch": 0.21651577967679161,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0008903575610442052,
      "loss": 1.5318,
      "step": 2559
    },
    {
      "epoch": 0.21660038920382435,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0008902743209479887,
      "loss": 1.1657,
      "step": 2560
    },
    {
      "epoch": 0.2166849987308571,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000890191053160209,
      "loss": 1.3481,
      "step": 2561
    },
    {
      "epoch": 0.21676960825788985,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0008901077576867742,
      "loss": 0.9893,
      "step": 2562
    },
    {
      "epoch": 0.2168542177849226,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0008900244345335947,
      "loss": 1.0684,
      "step": 2563
    },
    {
      "epoch": 0.21693882731195532,
      "grad_norm": 1.734375,
      "learning_rate": 0.0008899410837065824,
      "loss": 1.3021,
      "step": 2564
    },
    {
      "epoch": 0.21702343683898806,
      "grad_norm": 1.71875,
      "learning_rate": 0.0008898577052116513,
      "loss": 1.4724,
      "step": 2565
    },
    {
      "epoch": 0.21710804636602082,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0008897742990547177,
      "loss": 1.3151,
      "step": 2566
    },
    {
      "epoch": 0.21719265589305356,
      "grad_norm": 2.140625,
      "learning_rate": 0.0008896908652416995,
      "loss": 1.6081,
      "step": 2567
    },
    {
      "epoch": 0.2172772654200863,
      "grad_norm": 2.0,
      "learning_rate": 0.0008896074037785166,
      "loss": 1.5427,
      "step": 2568
    },
    {
      "epoch": 0.21736187494711903,
      "grad_norm": 1.578125,
      "learning_rate": 0.0008895239146710911,
      "loss": 1.1243,
      "step": 2569
    },
    {
      "epoch": 0.2174464844741518,
      "grad_norm": 1.765625,
      "learning_rate": 0.0008894403979253467,
      "loss": 1.4176,
      "step": 2570
    },
    {
      "epoch": 0.21753109400118453,
      "grad_norm": 1.484375,
      "learning_rate": 0.0008893568535472094,
      "loss": 1.3226,
      "step": 2571
    },
    {
      "epoch": 0.21761570352821727,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0008892732815426069,
      "loss": 1.2629,
      "step": 2572
    },
    {
      "epoch": 0.21770031305525003,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0008891896819174691,
      "loss": 1.5304,
      "step": 2573
    },
    {
      "epoch": 0.21778492258228277,
      "grad_norm": 1.5,
      "learning_rate": 0.0008891060546777277,
      "loss": 1.1629,
      "step": 2574
    },
    {
      "epoch": 0.2178695321093155,
      "grad_norm": 1.546875,
      "learning_rate": 0.0008890223998293163,
      "loss": 1.3848,
      "step": 2575
    },
    {
      "epoch": 0.21795414163634824,
      "grad_norm": 1.515625,
      "learning_rate": 0.0008889387173781707,
      "loss": 1.3257,
      "step": 2576
    },
    {
      "epoch": 0.218038751163381,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0008888550073302282,
      "loss": 2.0771,
      "step": 2577
    },
    {
      "epoch": 0.21812336069041374,
      "grad_norm": 35.75,
      "learning_rate": 0.0008887712696914289,
      "loss": 1.0835,
      "step": 2578
    },
    {
      "epoch": 0.21820797021744648,
      "grad_norm": 1.5,
      "learning_rate": 0.000888687504467714,
      "loss": 1.581,
      "step": 2579
    },
    {
      "epoch": 0.21829257974447921,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000888603711665027,
      "loss": 1.2986,
      "step": 2580
    },
    {
      "epoch": 0.21837718927151198,
      "grad_norm": 2.15625,
      "learning_rate": 0.0008885198912893133,
      "loss": 1.7609,
      "step": 2581
    },
    {
      "epoch": 0.21846179879854472,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0008884360433465204,
      "loss": 1.3864,
      "step": 2582
    },
    {
      "epoch": 0.21854640832557745,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0008883521678425977,
      "loss": 1.3401,
      "step": 2583
    },
    {
      "epoch": 0.21863101785261022,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0008882682647834963,
      "loss": 1.2742,
      "step": 2584
    },
    {
      "epoch": 0.21871562737964295,
      "grad_norm": 2.15625,
      "learning_rate": 0.0008881843341751696,
      "loss": 1.0519,
      "step": 2585
    },
    {
      "epoch": 0.2188002369066757,
      "grad_norm": 1.921875,
      "learning_rate": 0.0008881003760235728,
      "loss": 1.3839,
      "step": 2586
    },
    {
      "epoch": 0.21888484643370842,
      "grad_norm": 1.6875,
      "learning_rate": 0.0008880163903346633,
      "loss": 1.4174,
      "step": 2587
    },
    {
      "epoch": 0.2189694559607412,
      "grad_norm": 19.75,
      "learning_rate": 0.0008879323771143998,
      "loss": 1.3053,
      "step": 2588
    },
    {
      "epoch": 0.21905406548777392,
      "grad_norm": 2.046875,
      "learning_rate": 0.0008878483363687437,
      "loss": 1.4272,
      "step": 2589
    },
    {
      "epoch": 0.21913867501480666,
      "grad_norm": 2.21875,
      "learning_rate": 0.0008877642681036581,
      "loss": 1.0243,
      "step": 2590
    },
    {
      "epoch": 0.21922328454183942,
      "grad_norm": 3.03125,
      "learning_rate": 0.0008876801723251076,
      "loss": 1.1029,
      "step": 2591
    },
    {
      "epoch": 0.21930789406887216,
      "grad_norm": 2.375,
      "learning_rate": 0.0008875960490390595,
      "loss": 2.2641,
      "step": 2592
    },
    {
      "epoch": 0.2193925035959049,
      "grad_norm": 68.5,
      "learning_rate": 0.0008875118982514825,
      "loss": 1.3603,
      "step": 2593
    },
    {
      "epoch": 0.21947711312293763,
      "grad_norm": 1512.0,
      "learning_rate": 0.0008874277199683476,
      "loss": 1.2561,
      "step": 2594
    },
    {
      "epoch": 0.2195617226499704,
      "grad_norm": 1408.0,
      "learning_rate": 0.0008873435141956274,
      "loss": 1.4461,
      "step": 2595
    },
    {
      "epoch": 0.21964633217700313,
      "grad_norm": 25.375,
      "learning_rate": 0.0008872592809392967,
      "loss": 2.0389,
      "step": 2596
    },
    {
      "epoch": 0.21973094170403587,
      "grad_norm": 473088.0,
      "learning_rate": 0.0008871750202053323,
      "loss": 1.8174,
      "step": 2597
    },
    {
      "epoch": 0.2198155512310686,
      "grad_norm": 1466368.0,
      "learning_rate": 0.0008870907319997128,
      "loss": 1.3063,
      "step": 2598
    },
    {
      "epoch": 0.21990016075810137,
      "grad_norm": 1458176.0,
      "learning_rate": 0.0008870064163284189,
      "loss": 1.4864,
      "step": 2599
    },
    {
      "epoch": 0.2199847702851341,
      "grad_norm": 1122304.0,
      "learning_rate": 0.0008869220731974328,
      "loss": 1.4685,
      "step": 2600
    },
    {
      "epoch": 0.22006937981216684,
      "grad_norm": 217088.0,
      "learning_rate": 0.0008868377026127393,
      "loss": 1.2127,
      "step": 2601
    },
    {
      "epoch": 0.2201539893391996,
      "grad_norm": 505856.0,
      "learning_rate": 0.0008867533045803249,
      "loss": 1.2835,
      "step": 2602
    },
    {
      "epoch": 0.22023859886623234,
      "grad_norm": 219136.0,
      "learning_rate": 0.0008866688791061777,
      "loss": 1.5861,
      "step": 2603
    },
    {
      "epoch": 0.22032320839326508,
      "grad_norm": 41728.0,
      "learning_rate": 0.000886584426196288,
      "loss": 1.6669,
      "step": 2604
    },
    {
      "epoch": 0.22040781792029782,
      "grad_norm": 22144.0,
      "learning_rate": 0.0008864999458566485,
      "loss": 1.6118,
      "step": 2605
    },
    {
      "epoch": 0.22049242744733058,
      "grad_norm": 5472.0,
      "learning_rate": 0.0008864154380932531,
      "loss": 1.6802,
      "step": 2606
    },
    {
      "epoch": 0.22057703697436332,
      "grad_norm": 516.0,
      "learning_rate": 0.000886330902912098,
      "loss": 1.8683,
      "step": 2607
    },
    {
      "epoch": 0.22066164650139605,
      "grad_norm": 76.0,
      "learning_rate": 0.0008862463403191814,
      "loss": 1.391,
      "step": 2608
    },
    {
      "epoch": 0.2207462560284288,
      "grad_norm": 3.0,
      "learning_rate": 0.0008861617503205034,
      "loss": 1.4949,
      "step": 2609
    },
    {
      "epoch": 0.22083086555546155,
      "grad_norm": 2.40625,
      "learning_rate": 0.0008860771329220658,
      "loss": 1.6207,
      "step": 2610
    },
    {
      "epoch": 0.2209154750824943,
      "grad_norm": 2.859375,
      "learning_rate": 0.0008859924881298728,
      "loss": 1.9426,
      "step": 2611
    },
    {
      "epoch": 0.22100008460952703,
      "grad_norm": 3.0,
      "learning_rate": 0.0008859078159499302,
      "loss": 1.257,
      "step": 2612
    },
    {
      "epoch": 0.2210846941365598,
      "grad_norm": 2.046875,
      "learning_rate": 0.0008858231163882457,
      "loss": 1.9186,
      "step": 2613
    },
    {
      "epoch": 0.22116930366359253,
      "grad_norm": 2.3125,
      "learning_rate": 0.0008857383894508294,
      "loss": 1.2892,
      "step": 2614
    },
    {
      "epoch": 0.22125391319062526,
      "grad_norm": 2.265625,
      "learning_rate": 0.0008856536351436927,
      "loss": 1.3198,
      "step": 2615
    },
    {
      "epoch": 0.221338522717658,
      "grad_norm": 2.15625,
      "learning_rate": 0.0008855688534728495,
      "loss": 1.3635,
      "step": 2616
    },
    {
      "epoch": 0.22142313224469076,
      "grad_norm": 1.8125,
      "learning_rate": 0.0008854840444443151,
      "loss": 1.2477,
      "step": 2617
    },
    {
      "epoch": 0.2215077417717235,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008853992080641073,
      "loss": 1.5673,
      "step": 2618
    },
    {
      "epoch": 0.22159235129875623,
      "grad_norm": 2.609375,
      "learning_rate": 0.0008853143443382456,
      "loss": 1.6482,
      "step": 2619
    },
    {
      "epoch": 0.22167696082578897,
      "grad_norm": 1.828125,
      "learning_rate": 0.0008852294532727514,
      "loss": 1.3679,
      "step": 2620
    },
    {
      "epoch": 0.22176157035282174,
      "grad_norm": 1.7890625,
      "learning_rate": 0.000885144534873648,
      "loss": 1.2207,
      "step": 2621
    },
    {
      "epoch": 0.22184617987985447,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0008850595891469608,
      "loss": 1.3827,
      "step": 2622
    },
    {
      "epoch": 0.2219307894068872,
      "grad_norm": 1.78125,
      "learning_rate": 0.0008849746160987169,
      "loss": 1.5024,
      "step": 2623
    },
    {
      "epoch": 0.22201539893391997,
      "grad_norm": 5.96875,
      "learning_rate": 0.0008848896157349455,
      "loss": 1.1484,
      "step": 2624
    },
    {
      "epoch": 0.2221000084609527,
      "grad_norm": 2.046875,
      "learning_rate": 0.000884804588061678,
      "loss": 1.9569,
      "step": 2625
    },
    {
      "epoch": 0.22218461798798544,
      "grad_norm": 2.109375,
      "learning_rate": 0.0008847195330849471,
      "loss": 1.5809,
      "step": 2626
    },
    {
      "epoch": 0.22226922751501818,
      "grad_norm": 2.078125,
      "learning_rate": 0.0008846344508107881,
      "loss": 1.129,
      "step": 2627
    },
    {
      "epoch": 0.22235383704205094,
      "grad_norm": 2.09375,
      "learning_rate": 0.0008845493412452378,
      "loss": 1.3943,
      "step": 2628
    },
    {
      "epoch": 0.22243844656908368,
      "grad_norm": 1.609375,
      "learning_rate": 0.000884464204394335,
      "loss": 1.2059,
      "step": 2629
    },
    {
      "epoch": 0.22252305609611642,
      "grad_norm": 1.8125,
      "learning_rate": 0.0008843790402641205,
      "loss": 1.4547,
      "step": 2630
    },
    {
      "epoch": 0.22260766562314915,
      "grad_norm": 1.875,
      "learning_rate": 0.0008842938488606374,
      "loss": 1.3521,
      "step": 2631
    },
    {
      "epoch": 0.22269227515018192,
      "grad_norm": 1.546875,
      "learning_rate": 0.0008842086301899298,
      "loss": 1.2176,
      "step": 2632
    },
    {
      "epoch": 0.22277688467721465,
      "grad_norm": 2.171875,
      "learning_rate": 0.0008841233842580448,
      "loss": 1.5455,
      "step": 2633
    },
    {
      "epoch": 0.2228614942042474,
      "grad_norm": 1.84375,
      "learning_rate": 0.0008840381110710307,
      "loss": 1.4032,
      "step": 2634
    },
    {
      "epoch": 0.22294610373128015,
      "grad_norm": 1.7578125,
      "learning_rate": 0.000883952810634938,
      "loss": 1.2338,
      "step": 2635
    },
    {
      "epoch": 0.2230307132583129,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0008838674829558193,
      "loss": 1.3313,
      "step": 2636
    },
    {
      "epoch": 0.22311532278534563,
      "grad_norm": 1.40625,
      "learning_rate": 0.0008837821280397287,
      "loss": 1.2865,
      "step": 2637
    },
    {
      "epoch": 0.22319993231237836,
      "grad_norm": 1.625,
      "learning_rate": 0.0008836967458927226,
      "loss": 1.1735,
      "step": 2638
    },
    {
      "epoch": 0.22328454183941113,
      "grad_norm": 1.828125,
      "learning_rate": 0.0008836113365208593,
      "loss": 1.1633,
      "step": 2639
    },
    {
      "epoch": 0.22336915136644386,
      "grad_norm": 2.71875,
      "learning_rate": 0.0008835258999301987,
      "loss": 1.5114,
      "step": 2640
    },
    {
      "epoch": 0.2234537608934766,
      "grad_norm": 1.875,
      "learning_rate": 0.0008834404361268032,
      "loss": 1.3958,
      "step": 2641
    },
    {
      "epoch": 0.22353837042050934,
      "grad_norm": 2.0,
      "learning_rate": 0.0008833549451167366,
      "loss": 1.554,
      "step": 2642
    },
    {
      "epoch": 0.2236229799475421,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0008832694269060649,
      "loss": 1.2976,
      "step": 2643
    },
    {
      "epoch": 0.22370758947457484,
      "grad_norm": 1.90625,
      "learning_rate": 0.0008831838815008558,
      "loss": 1.8131,
      "step": 2644
    },
    {
      "epoch": 0.22379219900160757,
      "grad_norm": 1.5,
      "learning_rate": 0.0008830983089071794,
      "loss": 1.1211,
      "step": 2645
    },
    {
      "epoch": 0.22387680852864034,
      "grad_norm": 1.53125,
      "learning_rate": 0.0008830127091311071,
      "loss": 1.3068,
      "step": 2646
    },
    {
      "epoch": 0.22396141805567307,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0008829270821787128,
      "loss": 1.6667,
      "step": 2647
    },
    {
      "epoch": 0.2240460275827058,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000882841428056072,
      "loss": 1.3173,
      "step": 2648
    },
    {
      "epoch": 0.22413063710973855,
      "grad_norm": 1.90625,
      "learning_rate": 0.0008827557467692621,
      "loss": 1.332,
      "step": 2649
    },
    {
      "epoch": 0.2242152466367713,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000882670038324363,
      "loss": 1.9832,
      "step": 2650
    },
    {
      "epoch": 0.22429985616380405,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0008825843027274554,
      "loss": 1.3166,
      "step": 2651
    },
    {
      "epoch": 0.22438446569083678,
      "grad_norm": 1.9609375,
      "learning_rate": 0.000882498539984623,
      "loss": 1.9933,
      "step": 2652
    },
    {
      "epoch": 0.22446907521786955,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000882412750101951,
      "loss": 1.336,
      "step": 2653
    },
    {
      "epoch": 0.22455368474490228,
      "grad_norm": 1.9375,
      "learning_rate": 0.0008823269330855264,
      "loss": 1.7634,
      "step": 2654
    },
    {
      "epoch": 0.22463829427193502,
      "grad_norm": 1.3125,
      "learning_rate": 0.0008822410889414384,
      "loss": 1.0534,
      "step": 2655
    },
    {
      "epoch": 0.22472290379896775,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0008821552176757779,
      "loss": 1.6947,
      "step": 2656
    },
    {
      "epoch": 0.22480751332600052,
      "grad_norm": 2.09375,
      "learning_rate": 0.0008820693192946379,
      "loss": 2.2214,
      "step": 2657
    },
    {
      "epoch": 0.22489212285303326,
      "grad_norm": 1.375,
      "learning_rate": 0.000881983393804113,
      "loss": 1.3105,
      "step": 2658
    },
    {
      "epoch": 0.224976732380066,
      "grad_norm": 1.484375,
      "learning_rate": 0.0008818974412103005,
      "loss": 1.2491,
      "step": 2659
    },
    {
      "epoch": 0.22506134190709873,
      "grad_norm": 1.640625,
      "learning_rate": 0.0008818114615192985,
      "loss": 1.1721,
      "step": 2660
    },
    {
      "epoch": 0.2251459514341315,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000881725454737208,
      "loss": 1.1451,
      "step": 2661
    },
    {
      "epoch": 0.22523056096116423,
      "grad_norm": 1.640625,
      "learning_rate": 0.0008816394208701313,
      "loss": 1.4411,
      "step": 2662
    },
    {
      "epoch": 0.22531517048819696,
      "grad_norm": 1.53125,
      "learning_rate": 0.0008815533599241729,
      "loss": 1.4658,
      "step": 2663
    },
    {
      "epoch": 0.22539978001522973,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0008814672719054394,
      "loss": 1.4821,
      "step": 2664
    },
    {
      "epoch": 0.22548438954226246,
      "grad_norm": 1.59375,
      "learning_rate": 0.0008813811568200388,
      "loss": 1.6654,
      "step": 2665
    },
    {
      "epoch": 0.2255689990692952,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0008812950146740815,
      "loss": 1.2943,
      "step": 2666
    },
    {
      "epoch": 0.22565360859632794,
      "grad_norm": 1.53125,
      "learning_rate": 0.0008812088454736798,
      "loss": 1.6562,
      "step": 2667
    },
    {
      "epoch": 0.2257382181233607,
      "grad_norm": 1.453125,
      "learning_rate": 0.0008811226492249473,
      "loss": 1.1574,
      "step": 2668
    },
    {
      "epoch": 0.22582282765039344,
      "grad_norm": 1.796875,
      "learning_rate": 0.0008810364259340002,
      "loss": 1.7134,
      "step": 2669
    },
    {
      "epoch": 0.22590743717742617,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0008809501756069564,
      "loss": 2.0919,
      "step": 2670
    },
    {
      "epoch": 0.2259920467044589,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0008808638982499359,
      "loss": 2.0838,
      "step": 2671
    },
    {
      "epoch": 0.22607665623149167,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0008807775938690601,
      "loss": 1.1516,
      "step": 2672
    },
    {
      "epoch": 0.2261612657585244,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0008806912624704529,
      "loss": 0.9851,
      "step": 2673
    },
    {
      "epoch": 0.22624587528555715,
      "grad_norm": 1.609375,
      "learning_rate": 0.0008806049040602398,
      "loss": 1.3182,
      "step": 2674
    },
    {
      "epoch": 0.2263304848125899,
      "grad_norm": 2.015625,
      "learning_rate": 0.0008805185186445482,
      "loss": 1.714,
      "step": 2675
    },
    {
      "epoch": 0.22641509433962265,
      "grad_norm": 1.34375,
      "learning_rate": 0.0008804321062295075,
      "loss": 1.1657,
      "step": 2676
    },
    {
      "epoch": 0.22649970386665538,
      "grad_norm": 1.328125,
      "learning_rate": 0.0008803456668212492,
      "loss": 1.0647,
      "step": 2677
    },
    {
      "epoch": 0.22658431339368812,
      "grad_norm": 1.390625,
      "learning_rate": 0.0008802592004259063,
      "loss": 1.2013,
      "step": 2678
    },
    {
      "epoch": 0.22666892292072088,
      "grad_norm": 1.953125,
      "learning_rate": 0.0008801727070496142,
      "loss": 1.9727,
      "step": 2679
    },
    {
      "epoch": 0.22675353244775362,
      "grad_norm": 1.515625,
      "learning_rate": 0.0008800861866985097,
      "loss": 1.1203,
      "step": 2680
    },
    {
      "epoch": 0.22683814197478636,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0008799996393787319,
      "loss": 1.5046,
      "step": 2681
    },
    {
      "epoch": 0.2269227515018191,
      "grad_norm": 2.4375,
      "learning_rate": 0.0008799130650964218,
      "loss": 1.6036,
      "step": 2682
    },
    {
      "epoch": 0.22700736102885186,
      "grad_norm": 1.40625,
      "learning_rate": 0.0008798264638577218,
      "loss": 1.0213,
      "step": 2683
    },
    {
      "epoch": 0.2270919705558846,
      "grad_norm": 1.625,
      "learning_rate": 0.0008797398356687772,
      "loss": 1.9584,
      "step": 2684
    },
    {
      "epoch": 0.22717658008291733,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0008796531805357341,
      "loss": 1.103,
      "step": 2685
    },
    {
      "epoch": 0.2272611896099501,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0008795664984647413,
      "loss": 1.4827,
      "step": 2686
    },
    {
      "epoch": 0.22734579913698283,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0008794797894619493,
      "loss": 1.2598,
      "step": 2687
    },
    {
      "epoch": 0.22743040866401557,
      "grad_norm": 3.5,
      "learning_rate": 0.0008793930535335102,
      "loss": 1.6202,
      "step": 2688
    },
    {
      "epoch": 0.2275150181910483,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0008793062906855785,
      "loss": 2.1636,
      "step": 2689
    },
    {
      "epoch": 0.22759962771808107,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0008792195009243104,
      "loss": 1.1157,
      "step": 2690
    },
    {
      "epoch": 0.2276842372451138,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0008791326842558638,
      "loss": 1.1193,
      "step": 2691
    },
    {
      "epoch": 0.22776884677214654,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0008790458406863988,
      "loss": 1.2044,
      "step": 2692
    },
    {
      "epoch": 0.22785345629917927,
      "grad_norm": 5.28125,
      "learning_rate": 0.0008789589702220773,
      "loss": 1.6874,
      "step": 2693
    },
    {
      "epoch": 0.22793806582621204,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0008788720728690632,
      "loss": 1.2665,
      "step": 2694
    },
    {
      "epoch": 0.22802267535324477,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0008787851486335219,
      "loss": 1.4513,
      "step": 2695
    },
    {
      "epoch": 0.2281072848802775,
      "grad_norm": 1.578125,
      "learning_rate": 0.0008786981975216215,
      "loss": 1.4304,
      "step": 2696
    },
    {
      "epoch": 0.22819189440731028,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0008786112195395311,
      "loss": 1.0832,
      "step": 2697
    },
    {
      "epoch": 0.228276503934343,
      "grad_norm": 2.109375,
      "learning_rate": 0.0008785242146934225,
      "loss": 2.1895,
      "step": 2698
    },
    {
      "epoch": 0.22836111346137575,
      "grad_norm": 1.21875,
      "learning_rate": 0.0008784371829894691,
      "loss": 1.5515,
      "step": 2699
    },
    {
      "epoch": 0.22844572298840848,
      "grad_norm": 1.609375,
      "learning_rate": 0.0008783501244338457,
      "loss": 1.4098,
      "step": 2700
    },
    {
      "epoch": 0.22853033251544125,
      "grad_norm": 1.59375,
      "learning_rate": 0.0008782630390327297,
      "loss": 1.4978,
      "step": 2701
    },
    {
      "epoch": 0.22861494204247398,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0008781759267923003,
      "loss": 1.1874,
      "step": 2702
    },
    {
      "epoch": 0.22869955156950672,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0008780887877187384,
      "loss": 1.1372,
      "step": 2703
    },
    {
      "epoch": 0.22878416109653946,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0008780016218182268,
      "loss": 0.9058,
      "step": 2704
    },
    {
      "epoch": 0.22886877062357222,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0008779144290969501,
      "loss": 1.6349,
      "step": 2705
    },
    {
      "epoch": 0.22895338015060496,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0008778272095610956,
      "loss": 1.5519,
      "step": 2706
    },
    {
      "epoch": 0.2290379896776377,
      "grad_norm": 1.7734375,
      "learning_rate": 0.000877739963216851,
      "loss": 1.1712,
      "step": 2707
    },
    {
      "epoch": 0.22912259920467046,
      "grad_norm": 1.671875,
      "learning_rate": 0.0008776526900704076,
      "loss": 1.4494,
      "step": 2708
    },
    {
      "epoch": 0.2292072087317032,
      "grad_norm": 2.40625,
      "learning_rate": 0.0008775653901279574,
      "loss": 1.3978,
      "step": 2709
    },
    {
      "epoch": 0.22929181825873593,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0008774780633956947,
      "loss": 1.0627,
      "step": 2710
    },
    {
      "epoch": 0.22937642778576867,
      "grad_norm": 1.421875,
      "learning_rate": 0.0008773907098798158,
      "loss": 1.0296,
      "step": 2711
    },
    {
      "epoch": 0.22946103731280143,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0008773033295865185,
      "loss": 1.0328,
      "step": 2712
    },
    {
      "epoch": 0.22954564683983417,
      "grad_norm": 1.890625,
      "learning_rate": 0.0008772159225220031,
      "loss": 1.6422,
      "step": 2713
    },
    {
      "epoch": 0.2296302563668669,
      "grad_norm": 1.53125,
      "learning_rate": 0.0008771284886924715,
      "loss": 1.1939,
      "step": 2714
    },
    {
      "epoch": 0.22971486589389967,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0008770410281041274,
      "loss": 1.027,
      "step": 2715
    },
    {
      "epoch": 0.2297994754209324,
      "grad_norm": 2.046875,
      "learning_rate": 0.0008769535407631764,
      "loss": 1.4467,
      "step": 2716
    },
    {
      "epoch": 0.22988408494796514,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0008768660266758262,
      "loss": 1.5729,
      "step": 2717
    },
    {
      "epoch": 0.22996869447499788,
      "grad_norm": 1.359375,
      "learning_rate": 0.0008767784858482861,
      "loss": 1.4839,
      "step": 2718
    },
    {
      "epoch": 0.23005330400203064,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0008766909182867678,
      "loss": 1.4801,
      "step": 2719
    },
    {
      "epoch": 0.23013791352906338,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0008766033239974843,
      "loss": 1.2452,
      "step": 2720
    },
    {
      "epoch": 0.2302225230560961,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000876515702986651,
      "loss": 1.8348,
      "step": 2721
    },
    {
      "epoch": 0.23030713258312885,
      "grad_norm": 2.390625,
      "learning_rate": 0.0008764280552604847,
      "loss": 1.2546,
      "step": 2722
    },
    {
      "epoch": 0.2303917421101616,
      "grad_norm": 1.375,
      "learning_rate": 0.0008763403808252046,
      "loss": 1.1628,
      "step": 2723
    },
    {
      "epoch": 0.23047635163719435,
      "grad_norm": 2.078125,
      "learning_rate": 0.0008762526796870314,
      "loss": 1.7299,
      "step": 2724
    },
    {
      "epoch": 0.23056096116422709,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0008761649518521879,
      "loss": 1.8137,
      "step": 2725
    },
    {
      "epoch": 0.23064557069125985,
      "grad_norm": 1.921875,
      "learning_rate": 0.0008760771973268989,
      "loss": 2.2917,
      "step": 2726
    },
    {
      "epoch": 0.23073018021829259,
      "grad_norm": 1.71875,
      "learning_rate": 0.000875989416117391,
      "loss": 1.5087,
      "step": 2727
    },
    {
      "epoch": 0.23081478974532532,
      "grad_norm": 1.359375,
      "learning_rate": 0.000875901608229892,
      "loss": 1.05,
      "step": 2728
    },
    {
      "epoch": 0.23089939927235806,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0008758137736706329,
      "loss": 1.0356,
      "step": 2729
    },
    {
      "epoch": 0.23098400879939082,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0008757259124458459,
      "loss": 1.5113,
      "step": 2730
    },
    {
      "epoch": 0.23106861832642356,
      "grad_norm": 1.734375,
      "learning_rate": 0.0008756380245617646,
      "loss": 1.3844,
      "step": 2731
    },
    {
      "epoch": 0.2311532278534563,
      "grad_norm": 1.734375,
      "learning_rate": 0.0008755501100246255,
      "loss": 1.5566,
      "step": 2732
    },
    {
      "epoch": 0.23123783738048903,
      "grad_norm": 1.359375,
      "learning_rate": 0.0008754621688406663,
      "loss": 1.2449,
      "step": 2733
    },
    {
      "epoch": 0.2313224469075218,
      "grad_norm": 2.03125,
      "learning_rate": 0.0008753742010161268,
      "loss": 1.1149,
      "step": 2734
    },
    {
      "epoch": 0.23140705643455453,
      "grad_norm": 2.078125,
      "learning_rate": 0.0008752862065572487,
      "loss": 1.7033,
      "step": 2735
    },
    {
      "epoch": 0.23149166596158727,
      "grad_norm": 1.84375,
      "learning_rate": 0.0008751981854702755,
      "loss": 1.3811,
      "step": 2736
    },
    {
      "epoch": 0.23157627548862003,
      "grad_norm": 2.1875,
      "learning_rate": 0.0008751101377614528,
      "loss": 2.261,
      "step": 2737
    },
    {
      "epoch": 0.23166088501565277,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0008750220634370278,
      "loss": 1.1927,
      "step": 2738
    },
    {
      "epoch": 0.2317454945426855,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0008749339625032499,
      "loss": 1.1146,
      "step": 2739
    },
    {
      "epoch": 0.23183010406971824,
      "grad_norm": 1.90625,
      "learning_rate": 0.00087484583496637,
      "loss": 1.8976,
      "step": 2740
    },
    {
      "epoch": 0.231914713596751,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0008747576808326413,
      "loss": 1.2385,
      "step": 2741
    },
    {
      "epoch": 0.23199932312378374,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0008746695001083187,
      "loss": 1.79,
      "step": 2742
    },
    {
      "epoch": 0.23208393265081648,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0008745812927996587,
      "loss": 1.6459,
      "step": 2743
    },
    {
      "epoch": 0.2321685421778492,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0008744930589129203,
      "loss": 1.6154,
      "step": 2744
    },
    {
      "epoch": 0.23225315170488198,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0008744047984543639,
      "loss": 1.3885,
      "step": 2745
    },
    {
      "epoch": 0.2323377612319147,
      "grad_norm": 1.140625,
      "learning_rate": 0.0008743165114302521,
      "loss": 0.985,
      "step": 2746
    },
    {
      "epoch": 0.23242237075894745,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0008742281978468489,
      "loss": 1.4935,
      "step": 2747
    },
    {
      "epoch": 0.2325069802859802,
      "grad_norm": 1.6875,
      "learning_rate": 0.0008741398577104208,
      "loss": 2.1737,
      "step": 2748
    },
    {
      "epoch": 0.23259158981301295,
      "grad_norm": 1.84375,
      "learning_rate": 0.0008740514910272359,
      "loss": 1.4344,
      "step": 2749
    },
    {
      "epoch": 0.2326761993400457,
      "grad_norm": 1.4375,
      "learning_rate": 0.000873963097803564,
      "loss": 1.2181,
      "step": 2750
    },
    {
      "epoch": 0.23276080886707842,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000873874678045677,
      "loss": 1.4556,
      "step": 2751
    },
    {
      "epoch": 0.2328454183941112,
      "grad_norm": 1.671875,
      "learning_rate": 0.0008737862317598488,
      "loss": 1.181,
      "step": 2752
    },
    {
      "epoch": 0.23293002792114392,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0008736977589523548,
      "loss": 1.2506,
      "step": 2753
    },
    {
      "epoch": 0.23301463744817666,
      "grad_norm": 2.0625,
      "learning_rate": 0.0008736092596294727,
      "loss": 1.764,
      "step": 2754
    },
    {
      "epoch": 0.2330992469752094,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0008735207337974819,
      "loss": 1.3002,
      "step": 2755
    },
    {
      "epoch": 0.23318385650224216,
      "grad_norm": 1.5,
      "learning_rate": 0.0008734321814626634,
      "loss": 1.2612,
      "step": 2756
    },
    {
      "epoch": 0.2332684660292749,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008733436026313006,
      "loss": 1.4061,
      "step": 2757
    },
    {
      "epoch": 0.23335307555630763,
      "grad_norm": 1.703125,
      "learning_rate": 0.0008732549973096786,
      "loss": 1.2828,
      "step": 2758
    },
    {
      "epoch": 0.2334376850833404,
      "grad_norm": 1.453125,
      "learning_rate": 0.0008731663655040841,
      "loss": 1.2475,
      "step": 2759
    },
    {
      "epoch": 0.23352229461037313,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000873077707220806,
      "loss": 1.0972,
      "step": 2760
    },
    {
      "epoch": 0.23360690413740587,
      "grad_norm": 1.7578125,
      "learning_rate": 0.000872989022466135,
      "loss": 1.1219,
      "step": 2761
    },
    {
      "epoch": 0.2336915136644386,
      "grad_norm": 1.171875,
      "learning_rate": 0.0008729003112463633,
      "loss": 1.0636,
      "step": 2762
    },
    {
      "epoch": 0.23377612319147137,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000872811573567786,
      "loss": 1.4327,
      "step": 2763
    },
    {
      "epoch": 0.2338607327185041,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0008727228094366988,
      "loss": 0.8678,
      "step": 2764
    },
    {
      "epoch": 0.23394534224553684,
      "grad_norm": 2.203125,
      "learning_rate": 0.0008726340188594001,
      "loss": 2.2619,
      "step": 2765
    },
    {
      "epoch": 0.23402995177256958,
      "grad_norm": 1.4375,
      "learning_rate": 0.00087254520184219,
      "loss": 1.2703,
      "step": 2766
    },
    {
      "epoch": 0.23411456129960234,
      "grad_norm": 2.0625,
      "learning_rate": 0.0008724563583913703,
      "loss": 1.5079,
      "step": 2767
    },
    {
      "epoch": 0.23419917082663508,
      "grad_norm": 2.109375,
      "learning_rate": 0.0008723674885132449,
      "loss": 2.199,
      "step": 2768
    },
    {
      "epoch": 0.23428378035366781,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0008722785922141196,
      "loss": 1.8848,
      "step": 2769
    },
    {
      "epoch": 0.23436838988070058,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0008721896695003018,
      "loss": 1.2944,
      "step": 2770
    },
    {
      "epoch": 0.23445299940773331,
      "grad_norm": 1.78125,
      "learning_rate": 0.0008721007203781008,
      "loss": 1.918,
      "step": 2771
    },
    {
      "epoch": 0.23453760893476605,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000872011744853828,
      "loss": 1.1977,
      "step": 2772
    },
    {
      "epoch": 0.2346222184617988,
      "grad_norm": 2.0,
      "learning_rate": 0.0008719227429337971,
      "loss": 1.0736,
      "step": 2773
    },
    {
      "epoch": 0.23470682798883155,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0008718337146243223,
      "loss": 1.146,
      "step": 2774
    },
    {
      "epoch": 0.2347914375158643,
      "grad_norm": 1.9296875,
      "learning_rate": 0.000871744659931721,
      "loss": 1.4926,
      "step": 2775
    },
    {
      "epoch": 0.23487604704289702,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0008716555788623119,
      "loss": 1.2434,
      "step": 2776
    },
    {
      "epoch": 0.2349606565699298,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0008715664714224156,
      "loss": 1.4593,
      "step": 2777
    },
    {
      "epoch": 0.23504526609696252,
      "grad_norm": 2.1875,
      "learning_rate": 0.0008714773376183549,
      "loss": 1.3533,
      "step": 2778
    },
    {
      "epoch": 0.23512987562399526,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0008713881774564539,
      "loss": 1.4273,
      "step": 2779
    },
    {
      "epoch": 0.235214485151028,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0008712989909430391,
      "loss": 2.0826,
      "step": 2780
    },
    {
      "epoch": 0.23529909467806076,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0008712097780844384,
      "loss": 1.7682,
      "step": 2781
    },
    {
      "epoch": 0.2353837042050935,
      "grad_norm": 1.71875,
      "learning_rate": 0.0008711205388869821,
      "loss": 1.2716,
      "step": 2782
    },
    {
      "epoch": 0.23546831373212623,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0008710312733570019,
      "loss": 1.029,
      "step": 2783
    },
    {
      "epoch": 0.23555292325915897,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0008709419815008317,
      "loss": 1.4865,
      "step": 2784
    },
    {
      "epoch": 0.23563753278619173,
      "grad_norm": 1.890625,
      "learning_rate": 0.0008708526633248071,
      "loss": 1.2381,
      "step": 2785
    },
    {
      "epoch": 0.23572214231322447,
      "grad_norm": 1.59375,
      "learning_rate": 0.0008707633188352652,
      "loss": 1.3446,
      "step": 2786
    },
    {
      "epoch": 0.2358067518402572,
      "grad_norm": 1.46875,
      "learning_rate": 0.000870673948038546,
      "loss": 1.4228,
      "step": 2787
    },
    {
      "epoch": 0.23589136136728997,
      "grad_norm": 1.578125,
      "learning_rate": 0.0008705845509409903,
      "loss": 1.1726,
      "step": 2788
    },
    {
      "epoch": 0.2359759708943227,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0008704951275489412,
      "loss": 1.1583,
      "step": 2789
    },
    {
      "epoch": 0.23606058042135544,
      "grad_norm": 1.265625,
      "learning_rate": 0.0008704056778687439,
      "loss": 1.0434,
      "step": 2790
    },
    {
      "epoch": 0.23614518994838818,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0008703162019067451,
      "loss": 1.6138,
      "step": 2791
    },
    {
      "epoch": 0.23622979947542094,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0008702266996692933,
      "loss": 0.9944,
      "step": 2792
    },
    {
      "epoch": 0.23631440900245368,
      "grad_norm": 2.015625,
      "learning_rate": 0.0008701371711627395,
      "loss": 1.5954,
      "step": 2793
    },
    {
      "epoch": 0.23639901852948642,
      "grad_norm": 2.234375,
      "learning_rate": 0.0008700476163934356,
      "loss": 1.5631,
      "step": 2794
    },
    {
      "epoch": 0.23648362805651915,
      "grad_norm": 1.78125,
      "learning_rate": 0.0008699580353677361,
      "loss": 1.5404,
      "step": 2795
    },
    {
      "epoch": 0.23656823758355192,
      "grad_norm": 1.53125,
      "learning_rate": 0.0008698684280919973,
      "loss": 1.1929,
      "step": 2796
    },
    {
      "epoch": 0.23665284711058465,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0008697787945725769,
      "loss": 1.1729,
      "step": 2797
    },
    {
      "epoch": 0.2367374566376174,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000869689134815835,
      "loss": 1.0673,
      "step": 2798
    },
    {
      "epoch": 0.23682206616465015,
      "grad_norm": 1.875,
      "learning_rate": 0.0008695994488281332,
      "loss": 1.5636,
      "step": 2799
    },
    {
      "epoch": 0.2369066756916829,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0008695097366158351,
      "loss": 1.4929,
      "step": 2800
    },
    {
      "epoch": 0.23699128521871562,
      "grad_norm": 2.171875,
      "learning_rate": 0.0008694199981853064,
      "loss": 1.5947,
      "step": 2801
    },
    {
      "epoch": 0.23707589474574836,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000869330233542914,
      "loss": 1.409,
      "step": 2802
    },
    {
      "epoch": 0.23716050427278113,
      "grad_norm": 1.515625,
      "learning_rate": 0.0008692404426950273,
      "loss": 1.252,
      "step": 2803
    },
    {
      "epoch": 0.23724511379981386,
      "grad_norm": 1.96875,
      "learning_rate": 0.0008691506256480173,
      "loss": 1.2529,
      "step": 2804
    },
    {
      "epoch": 0.2373297233268466,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0008690607824082568,
      "loss": 1.206,
      "step": 2805
    },
    {
      "epoch": 0.23741433285387933,
      "grad_norm": 1.3125,
      "learning_rate": 0.0008689709129821207,
      "loss": 1.1905,
      "step": 2806
    },
    {
      "epoch": 0.2374989423809121,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0008688810173759856,
      "loss": 1.8923,
      "step": 2807
    },
    {
      "epoch": 0.23758355190794483,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0008687910955962298,
      "loss": 1.438,
      "step": 2808
    },
    {
      "epoch": 0.23766816143497757,
      "grad_norm": 1.453125,
      "learning_rate": 0.0008687011476492338,
      "loss": 1.2032,
      "step": 2809
    },
    {
      "epoch": 0.23775277096201033,
      "grad_norm": 1.46875,
      "learning_rate": 0.0008686111735413795,
      "loss": 1.2185,
      "step": 2810
    },
    {
      "epoch": 0.23783738048904307,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0008685211732790512,
      "loss": 1.0866,
      "step": 2811
    },
    {
      "epoch": 0.2379219900160758,
      "grad_norm": 1.453125,
      "learning_rate": 0.0008684311468686348,
      "loss": 1.3376,
      "step": 2812
    },
    {
      "epoch": 0.23800659954310854,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0008683410943165179,
      "loss": 1.7454,
      "step": 2813
    },
    {
      "epoch": 0.2380912090701413,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0008682510156290901,
      "loss": 1.6194,
      "step": 2814
    },
    {
      "epoch": 0.23817581859717404,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000868160910812743,
      "loss": 1.1913,
      "step": 2815
    },
    {
      "epoch": 0.23826042812420678,
      "grad_norm": 1.78125,
      "learning_rate": 0.0008680707798738699,
      "loss": 1.4462,
      "step": 2816
    },
    {
      "epoch": 0.23834503765123952,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0008679806228188656,
      "loss": 1.4198,
      "step": 2817
    },
    {
      "epoch": 0.23842964717827228,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0008678904396541274,
      "loss": 1.0139,
      "step": 2818
    },
    {
      "epoch": 0.23851425670530502,
      "grad_norm": 2.046875,
      "learning_rate": 0.0008678002303860544,
      "loss": 1.2707,
      "step": 2819
    },
    {
      "epoch": 0.23859886623233775,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0008677099950210467,
      "loss": 1.7213,
      "step": 2820
    },
    {
      "epoch": 0.23868347575937052,
      "grad_norm": 1.5,
      "learning_rate": 0.0008676197335655075,
      "loss": 1.2266,
      "step": 2821
    },
    {
      "epoch": 0.23876808528640325,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000867529446025841,
      "loss": 1.3053,
      "step": 2822
    },
    {
      "epoch": 0.238852694813436,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0008674391324084531,
      "loss": 1.071,
      "step": 2823
    },
    {
      "epoch": 0.23893730434046873,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0008673487927197526,
      "loss": 1.7749,
      "step": 2824
    },
    {
      "epoch": 0.2390219138675015,
      "grad_norm": 1.6875,
      "learning_rate": 0.0008672584269661488,
      "loss": 1.362,
      "step": 2825
    },
    {
      "epoch": 0.23910652339453423,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0008671680351540541,
      "loss": 1.6291,
      "step": 2826
    },
    {
      "epoch": 0.23919113292156696,
      "grad_norm": 1.546875,
      "learning_rate": 0.0008670776172898818,
      "loss": 1.5056,
      "step": 2827
    },
    {
      "epoch": 0.2392757424485997,
      "grad_norm": 1.421875,
      "learning_rate": 0.0008669871733800474,
      "loss": 1.117,
      "step": 2828
    },
    {
      "epoch": 0.23936035197563246,
      "grad_norm": 1.1875,
      "learning_rate": 0.0008668967034309685,
      "loss": 0.9995,
      "step": 2829
    },
    {
      "epoch": 0.2394449615026652,
      "grad_norm": 1.59375,
      "learning_rate": 0.0008668062074490641,
      "loss": 1.6379,
      "step": 2830
    },
    {
      "epoch": 0.23952957102969794,
      "grad_norm": 1.84375,
      "learning_rate": 0.0008667156854407555,
      "loss": 1.4909,
      "step": 2831
    },
    {
      "epoch": 0.2396141805567307,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0008666251374124654,
      "loss": 1.3236,
      "step": 2832
    },
    {
      "epoch": 0.23969879008376344,
      "grad_norm": 1.8125,
      "learning_rate": 0.0008665345633706188,
      "loss": 1.8935,
      "step": 2833
    },
    {
      "epoch": 0.23978339961079617,
      "grad_norm": 1.546875,
      "learning_rate": 0.0008664439633216419,
      "loss": 1.0412,
      "step": 2834
    },
    {
      "epoch": 0.2398680091378289,
      "grad_norm": 1.203125,
      "learning_rate": 0.0008663533372719634,
      "loss": 1.0011,
      "step": 2835
    },
    {
      "epoch": 0.23995261866486167,
      "grad_norm": 1.390625,
      "learning_rate": 0.0008662626852280137,
      "loss": 1.0947,
      "step": 2836
    },
    {
      "epoch": 0.2400372281918944,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0008661720071962246,
      "loss": 1.4713,
      "step": 2837
    },
    {
      "epoch": 0.24012183771892714,
      "grad_norm": 1.515625,
      "learning_rate": 0.0008660813031830303,
      "loss": 1.3567,
      "step": 2838
    },
    {
      "epoch": 0.2402064472459599,
      "grad_norm": 1.484375,
      "learning_rate": 0.0008659905731948668,
      "loss": 1.5275,
      "step": 2839
    },
    {
      "epoch": 0.24029105677299264,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0008658998172381714,
      "loss": 1.6217,
      "step": 2840
    },
    {
      "epoch": 0.24037566630002538,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0008658090353193837,
      "loss": 1.4286,
      "step": 2841
    },
    {
      "epoch": 0.24046027582705812,
      "grad_norm": 2.203125,
      "learning_rate": 0.0008657182274449453,
      "loss": 1.1282,
      "step": 2842
    },
    {
      "epoch": 0.24054488535409088,
      "grad_norm": 1.765625,
      "learning_rate": 0.0008656273936212991,
      "loss": 1.2086,
      "step": 2843
    },
    {
      "epoch": 0.24062949488112362,
      "grad_norm": 1.796875,
      "learning_rate": 0.0008655365338548902,
      "loss": 1.2861,
      "step": 2844
    },
    {
      "epoch": 0.24071410440815635,
      "grad_norm": 1.984375,
      "learning_rate": 0.0008654456481521656,
      "loss": 1.4803,
      "step": 2845
    },
    {
      "epoch": 0.2407987139351891,
      "grad_norm": 1.0,
      "learning_rate": 0.0008653547365195739,
      "loss": 0.9144,
      "step": 2846
    },
    {
      "epoch": 0.24088332346222185,
      "grad_norm": 1.4375,
      "learning_rate": 0.0008652637989635657,
      "loss": 1.1321,
      "step": 2847
    },
    {
      "epoch": 0.2409679329892546,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0008651728354905933,
      "loss": 0.876,
      "step": 2848
    },
    {
      "epoch": 0.24105254251628733,
      "grad_norm": 1.953125,
      "learning_rate": 0.000865081846107111,
      "loss": 1.8349,
      "step": 2849
    },
    {
      "epoch": 0.2411371520433201,
      "grad_norm": 1.578125,
      "learning_rate": 0.0008649908308195749,
      "loss": 1.6385,
      "step": 2850
    },
    {
      "epoch": 0.24122176157035283,
      "grad_norm": 1.5,
      "learning_rate": 0.0008648997896344428,
      "loss": 0.9535,
      "step": 2851
    },
    {
      "epoch": 0.24130637109738556,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0008648087225581745,
      "loss": 1.2518,
      "step": 2852
    },
    {
      "epoch": 0.2413909806244183,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0008647176295972318,
      "loss": 1.618,
      "step": 2853
    },
    {
      "epoch": 0.24147559015145106,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0008646265107580777,
      "loss": 1.1742,
      "step": 2854
    },
    {
      "epoch": 0.2415601996784838,
      "grad_norm": 1.359375,
      "learning_rate": 0.000864535366047178,
      "loss": 1.1411,
      "step": 2855
    },
    {
      "epoch": 0.24164480920551654,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0008644441954709992,
      "loss": 1.6697,
      "step": 2856
    },
    {
      "epoch": 0.24172941873254927,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0008643529990360106,
      "loss": 1.9578,
      "step": 2857
    },
    {
      "epoch": 0.24181402825958204,
      "grad_norm": 1.875,
      "learning_rate": 0.0008642617767486828,
      "loss": 1.317,
      "step": 2858
    },
    {
      "epoch": 0.24189863778661477,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0008641705286154884,
      "loss": 1.8751,
      "step": 2859
    },
    {
      "epoch": 0.2419832473136475,
      "grad_norm": 1.796875,
      "learning_rate": 0.0008640792546429019,
      "loss": 1.374,
      "step": 2860
    },
    {
      "epoch": 0.24206785684068027,
      "grad_norm": 2.15625,
      "learning_rate": 0.0008639879548373997,
      "loss": 1.8236,
      "step": 2861
    },
    {
      "epoch": 0.242152466367713,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0008638966292054597,
      "loss": 1.5013,
      "step": 2862
    },
    {
      "epoch": 0.24223707589474575,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0008638052777535617,
      "loss": 1.0258,
      "step": 2863
    },
    {
      "epoch": 0.24232168542177848,
      "grad_norm": 1.671875,
      "learning_rate": 0.0008637139004881878,
      "loss": 1.6959,
      "step": 2864
    },
    {
      "epoch": 0.24240629494881125,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0008636224974158214,
      "loss": 1.1078,
      "step": 2865
    },
    {
      "epoch": 0.24249090447584398,
      "grad_norm": 1.640625,
      "learning_rate": 0.000863531068542948,
      "loss": 1.4919,
      "step": 2866
    },
    {
      "epoch": 0.24257551400287672,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0008634396138760546,
      "loss": 1.1175,
      "step": 2867
    },
    {
      "epoch": 0.24266012352990945,
      "grad_norm": 1.84375,
      "learning_rate": 0.0008633481334216306,
      "loss": 1.5178,
      "step": 2868
    },
    {
      "epoch": 0.24274473305694222,
      "grad_norm": 1.5625,
      "learning_rate": 0.0008632566271861668,
      "loss": 1.7322,
      "step": 2869
    },
    {
      "epoch": 0.24282934258397496,
      "grad_norm": 2.125,
      "learning_rate": 0.000863165095176156,
      "loss": 1.3605,
      "step": 2870
    },
    {
      "epoch": 0.2429139521110077,
      "grad_norm": 1.890625,
      "learning_rate": 0.0008630735373980925,
      "loss": 1.3433,
      "step": 2871
    },
    {
      "epoch": 0.24299856163804046,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0008629819538584731,
      "loss": 1.3504,
      "step": 2872
    },
    {
      "epoch": 0.2430831711650732,
      "grad_norm": 2.203125,
      "learning_rate": 0.0008628903445637956,
      "loss": 1.2478,
      "step": 2873
    },
    {
      "epoch": 0.24316778069210593,
      "grad_norm": 1.8125,
      "learning_rate": 0.0008627987095205606,
      "loss": 1.2672,
      "step": 2874
    },
    {
      "epoch": 0.24325239021913866,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0008627070487352695,
      "loss": 1.2092,
      "step": 2875
    },
    {
      "epoch": 0.24333699974617143,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0008626153622144263,
      "loss": 1.2462,
      "step": 2876
    },
    {
      "epoch": 0.24342160927320416,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0008625236499645362,
      "loss": 1.2356,
      "step": 2877
    },
    {
      "epoch": 0.2435062188002369,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0008624319119921069,
      "loss": 1.0251,
      "step": 2878
    },
    {
      "epoch": 0.24359082832726964,
      "grad_norm": 1.890625,
      "learning_rate": 0.0008623401483036475,
      "loss": 1.3778,
      "step": 2879
    },
    {
      "epoch": 0.2436754378543024,
      "grad_norm": 1.59375,
      "learning_rate": 0.0008622483589056689,
      "loss": 1.1657,
      "step": 2880
    },
    {
      "epoch": 0.24376004738133514,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0008621565438046841,
      "loss": 1.4953,
      "step": 2881
    },
    {
      "epoch": 0.24384465690836787,
      "grad_norm": 1.546875,
      "learning_rate": 0.0008620647030072075,
      "loss": 1.3632,
      "step": 2882
    },
    {
      "epoch": 0.24392926643540064,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0008619728365197559,
      "loss": 1.2188,
      "step": 2883
    },
    {
      "epoch": 0.24401387596243337,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0008618809443488474,
      "loss": 1.4056,
      "step": 2884
    },
    {
      "epoch": 0.2440984854894661,
      "grad_norm": 1.515625,
      "learning_rate": 0.0008617890265010022,
      "loss": 1.2619,
      "step": 2885
    },
    {
      "epoch": 0.24418309501649885,
      "grad_norm": 2.09375,
      "learning_rate": 0.0008616970829827422,
      "loss": 2.0393,
      "step": 2886
    },
    {
      "epoch": 0.2442677045435316,
      "grad_norm": 1.8125,
      "learning_rate": 0.0008616051138005911,
      "loss": 1.6146,
      "step": 2887
    },
    {
      "epoch": 0.24435231407056435,
      "grad_norm": 1.28125,
      "learning_rate": 0.0008615131189610747,
      "loss": 1.0625,
      "step": 2888
    },
    {
      "epoch": 0.24443692359759708,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0008614210984707202,
      "loss": 1.14,
      "step": 2889
    },
    {
      "epoch": 0.24452153312462982,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0008613290523360571,
      "loss": 1.6951,
      "step": 2890
    },
    {
      "epoch": 0.24460614265166258,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000861236980563616,
      "loss": 1.0132,
      "step": 2891
    },
    {
      "epoch": 0.24469075217869532,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0008611448831599303,
      "loss": 1.1349,
      "step": 2892
    },
    {
      "epoch": 0.24477536170572806,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0008610527601315343,
      "loss": 1.5404,
      "step": 2893
    },
    {
      "epoch": 0.24485997123276082,
      "grad_norm": 1.84375,
      "learning_rate": 0.0008609606114849646,
      "loss": 1.683,
      "step": 2894
    },
    {
      "epoch": 0.24494458075979356,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0008608684372267597,
      "loss": 1.4211,
      "step": 2895
    },
    {
      "epoch": 0.2450291902868263,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0008607762373634596,
      "loss": 1.4385,
      "step": 2896
    },
    {
      "epoch": 0.24511379981385903,
      "grad_norm": 1.6875,
      "learning_rate": 0.0008606840119016061,
      "loss": 1.6186,
      "step": 2897
    },
    {
      "epoch": 0.2451984093408918,
      "grad_norm": 1.890625,
      "learning_rate": 0.0008605917608477433,
      "loss": 1.9468,
      "step": 2898
    },
    {
      "epoch": 0.24528301886792453,
      "grad_norm": 1.984375,
      "learning_rate": 0.0008604994842084166,
      "loss": 1.2648,
      "step": 2899
    },
    {
      "epoch": 0.24536762839495727,
      "grad_norm": 1.4375,
      "learning_rate": 0.0008604071819901734,
      "loss": 1.0198,
      "step": 2900
    },
    {
      "epoch": 0.24545223792199003,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0008603148541995629,
      "loss": 1.302,
      "step": 2901
    },
    {
      "epoch": 0.24553684744902277,
      "grad_norm": 1.546875,
      "learning_rate": 0.0008602225008431363,
      "loss": 1.9302,
      "step": 2902
    },
    {
      "epoch": 0.2456214569760555,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0008601301219274462,
      "loss": 1.9246,
      "step": 2903
    },
    {
      "epoch": 0.24570606650308824,
      "grad_norm": 1.609375,
      "learning_rate": 0.0008600377174590475,
      "loss": 2.0845,
      "step": 2904
    },
    {
      "epoch": 0.245790676030121,
      "grad_norm": 1.546875,
      "learning_rate": 0.0008599452874444967,
      "loss": 1.1711,
      "step": 2905
    },
    {
      "epoch": 0.24587528555715374,
      "grad_norm": 2.484375,
      "learning_rate": 0.0008598528318903517,
      "loss": 1.8542,
      "step": 2906
    },
    {
      "epoch": 0.24595989508418648,
      "grad_norm": 1.40625,
      "learning_rate": 0.0008597603508031731,
      "loss": 1.1227,
      "step": 2907
    },
    {
      "epoch": 0.2460445046112192,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0008596678441895224,
      "loss": 1.5422,
      "step": 2908
    },
    {
      "epoch": 0.24612911413825198,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0008595753120559635,
      "loss": 1.4239,
      "step": 2909
    },
    {
      "epoch": 0.2462137236652847,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0008594827544090621,
      "loss": 1.1886,
      "step": 2910
    },
    {
      "epoch": 0.24629833319231745,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0008593901712553853,
      "loss": 1.0686,
      "step": 2911
    },
    {
      "epoch": 0.2463829427193502,
      "grad_norm": 1.75,
      "learning_rate": 0.0008592975626015024,
      "loss": 1.0961,
      "step": 2912
    },
    {
      "epoch": 0.24646755224638295,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0008592049284539843,
      "loss": 1.3186,
      "step": 2913
    },
    {
      "epoch": 0.24655216177341568,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0008591122688194037,
      "loss": 1.0713,
      "step": 2914
    },
    {
      "epoch": 0.24663677130044842,
      "grad_norm": 2.21875,
      "learning_rate": 0.0008590195837043352,
      "loss": 1.5938,
      "step": 2915
    },
    {
      "epoch": 0.24672138082748118,
      "grad_norm": 8.25,
      "learning_rate": 0.0008589268731153554,
      "loss": 1.6384,
      "step": 2916
    },
    {
      "epoch": 0.24680599035451392,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0008588341370590422,
      "loss": 1.1262,
      "step": 2917
    },
    {
      "epoch": 0.24689059988154666,
      "grad_norm": 2.125,
      "learning_rate": 0.0008587413755419758,
      "loss": 1.8849,
      "step": 2918
    },
    {
      "epoch": 0.2469752094085794,
      "grad_norm": 1.484375,
      "learning_rate": 0.0008586485885707379,
      "loss": 0.9423,
      "step": 2919
    },
    {
      "epoch": 0.24705981893561216,
      "grad_norm": 2.109375,
      "learning_rate": 0.0008585557761519123,
      "loss": 1.833,
      "step": 2920
    },
    {
      "epoch": 0.2471444284626449,
      "grad_norm": 1.609375,
      "learning_rate": 0.0008584629382920843,
      "loss": 1.2429,
      "step": 2921
    },
    {
      "epoch": 0.24722903798967763,
      "grad_norm": 1.453125,
      "learning_rate": 0.000858370074997841,
      "loss": 1.1259,
      "step": 2922
    },
    {
      "epoch": 0.2473136475167104,
      "grad_norm": 1.421875,
      "learning_rate": 0.0008582771862757715,
      "loss": 1.0183,
      "step": 2923
    },
    {
      "epoch": 0.24739825704374313,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0008581842721324667,
      "loss": 1.1199,
      "step": 2924
    },
    {
      "epoch": 0.24748286657077587,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0008580913325745194,
      "loss": 1.288,
      "step": 2925
    },
    {
      "epoch": 0.2475674760978086,
      "grad_norm": 1.5,
      "learning_rate": 0.0008579983676085237,
      "loss": 1.3318,
      "step": 2926
    },
    {
      "epoch": 0.24765208562484137,
      "grad_norm": 1.453125,
      "learning_rate": 0.0008579053772410761,
      "loss": 1.3879,
      "step": 2927
    },
    {
      "epoch": 0.2477366951518741,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0008578123614787745,
      "loss": 1.2309,
      "step": 2928
    },
    {
      "epoch": 0.24782130467890684,
      "grad_norm": 1.859375,
      "learning_rate": 0.0008577193203282189,
      "loss": 1.748,
      "step": 2929
    },
    {
      "epoch": 0.24790591420593958,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0008576262537960107,
      "loss": 1.4697,
      "step": 2930
    },
    {
      "epoch": 0.24799052373297234,
      "grad_norm": 1.4375,
      "learning_rate": 0.0008575331618887537,
      "loss": 1.1713,
      "step": 2931
    },
    {
      "epoch": 0.24807513326000508,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0008574400446130528,
      "loss": 1.3693,
      "step": 2932
    },
    {
      "epoch": 0.2481597427870378,
      "grad_norm": 1.40625,
      "learning_rate": 0.0008573469019755154,
      "loss": 1.2322,
      "step": 2933
    },
    {
      "epoch": 0.24824435231407058,
      "grad_norm": 1.828125,
      "learning_rate": 0.00085725373398275,
      "loss": 1.2347,
      "step": 2934
    },
    {
      "epoch": 0.2483289618411033,
      "grad_norm": 1.65625,
      "learning_rate": 0.0008571605406413674,
      "loss": 1.4775,
      "step": 2935
    },
    {
      "epoch": 0.24841357136813605,
      "grad_norm": 1.359375,
      "learning_rate": 0.0008570673219579802,
      "loss": 1.3701,
      "step": 2936
    },
    {
      "epoch": 0.24849818089516879,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0008569740779392024,
      "loss": 1.9535,
      "step": 2937
    },
    {
      "epoch": 0.24858279042220155,
      "grad_norm": 1.46875,
      "learning_rate": 0.0008568808085916503,
      "loss": 1.3873,
      "step": 2938
    },
    {
      "epoch": 0.24866739994923429,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0008567875139219416,
      "loss": 1.775,
      "step": 2939
    },
    {
      "epoch": 0.24875200947626702,
      "grad_norm": 1.578125,
      "learning_rate": 0.0008566941939366959,
      "loss": 1.3409,
      "step": 2940
    },
    {
      "epoch": 0.24883661900329976,
      "grad_norm": 1.28125,
      "learning_rate": 0.0008566008486425345,
      "loss": 1.1273,
      "step": 2941
    },
    {
      "epoch": 0.24892122853033252,
      "grad_norm": 1.203125,
      "learning_rate": 0.0008565074780460811,
      "loss": 1.1493,
      "step": 2942
    },
    {
      "epoch": 0.24900583805736526,
      "grad_norm": 2.078125,
      "learning_rate": 0.0008564140821539603,
      "loss": 1.844,
      "step": 2943
    },
    {
      "epoch": 0.249090447584398,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0008563206609727991,
      "loss": 1.4655,
      "step": 2944
    },
    {
      "epoch": 0.24917505711143076,
      "grad_norm": 2.78125,
      "learning_rate": 0.000856227214509226,
      "loss": 1.8247,
      "step": 2945
    },
    {
      "epoch": 0.2492596666384635,
      "grad_norm": 1.375,
      "learning_rate": 0.0008561337427698717,
      "loss": 1.1893,
      "step": 2946
    },
    {
      "epoch": 0.24934427616549623,
      "grad_norm": 1.375,
      "learning_rate": 0.0008560402457613678,
      "loss": 1.6554,
      "step": 2947
    },
    {
      "epoch": 0.24942888569252897,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0008559467234903491,
      "loss": 1.199,
      "step": 2948
    },
    {
      "epoch": 0.24951349521956173,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0008558531759634507,
      "loss": 1.1253,
      "step": 2949
    },
    {
      "epoch": 0.24959810474659447,
      "grad_norm": 1.484375,
      "learning_rate": 0.0008557596031873106,
      "loss": 1.1262,
      "step": 2950
    },
    {
      "epoch": 0.2496827142736272,
      "grad_norm": 1.875,
      "learning_rate": 0.0008556660051685679,
      "loss": 1.6585,
      "step": 2951
    },
    {
      "epoch": 0.24976732380065994,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0008555723819138641,
      "loss": 1.2208,
      "step": 2952
    },
    {
      "epoch": 0.2498519333276927,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0008554787334298417,
      "loss": 1.184,
      "step": 2953
    },
    {
      "epoch": 0.24993654285472544,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0008553850597231459,
      "loss": 1.4987,
      "step": 2954
    },
    {
      "epoch": 0.2500211523817582,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0008552913608004228,
      "loss": 1.5925,
      "step": 2955
    },
    {
      "epoch": 0.2500211523817582,
      "eval_loss": 1.3909703493118286,
      "eval_runtime": 387.925,
      "eval_samples_per_second": 48.749,
      "eval_steps_per_second": 6.094,
      "step": 2955
    }
  ],
  "logging_steps": 1,
  "max_steps": 11819,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2955,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.881990609215488e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}