amburger66's picture
LoRA fine-tune on RobotSmith real world task08 cutting
c8bcf44 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.05755395683453238,
"eval_steps": 50,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 5.755395683453237e-05,
"grad_norm": 19.742076873779297,
"learning_rate": 0.0,
"loss": 1.1824,
"step": 1
},
{
"epoch": 0.00011510791366906474,
"grad_norm": 14.090814590454102,
"learning_rate": 2.0000000000000002e-07,
"loss": 1.5161,
"step": 2
},
{
"epoch": 0.00017266187050359714,
"grad_norm": 9.585816383361816,
"learning_rate": 4.0000000000000003e-07,
"loss": 0.4535,
"step": 3
},
{
"epoch": 0.0002302158273381295,
"grad_norm": 13.000758171081543,
"learning_rate": 6.000000000000001e-07,
"loss": 1.1017,
"step": 4
},
{
"epoch": 0.00028776978417266187,
"grad_norm": 7.664703369140625,
"learning_rate": 8.000000000000001e-07,
"loss": 0.7945,
"step": 5
},
{
"epoch": 0.00034532374100719427,
"grad_norm": 12.25283145904541,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.1652,
"step": 6
},
{
"epoch": 0.0004028776978417266,
"grad_norm": 11.274255752563477,
"learning_rate": 1.2000000000000002e-06,
"loss": 0.719,
"step": 7
},
{
"epoch": 0.000460431654676259,
"grad_norm": 7.998445510864258,
"learning_rate": 1.4000000000000001e-06,
"loss": 0.6316,
"step": 8
},
{
"epoch": 0.0005179856115107913,
"grad_norm": 8.421154022216797,
"learning_rate": 1.6000000000000001e-06,
"loss": 0.8872,
"step": 9
},
{
"epoch": 0.0005755395683453237,
"grad_norm": 9.403829574584961,
"learning_rate": 1.8000000000000001e-06,
"loss": 0.7519,
"step": 10
},
{
"epoch": 0.0006330935251798561,
"grad_norm": 14.926182746887207,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.8313,
"step": 11
},
{
"epoch": 0.0006906474820143885,
"grad_norm": 5.170341491699219,
"learning_rate": 2.2e-06,
"loss": 0.4195,
"step": 12
},
{
"epoch": 0.0007482014388489208,
"grad_norm": 4.143105506896973,
"learning_rate": 2.4000000000000003e-06,
"loss": 0.3894,
"step": 13
},
{
"epoch": 0.0008057553956834532,
"grad_norm": 7.147073268890381,
"learning_rate": 2.6e-06,
"loss": 0.5323,
"step": 14
},
{
"epoch": 0.0008633093525179857,
"grad_norm": 5.619192600250244,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.3687,
"step": 15
},
{
"epoch": 0.000920863309352518,
"grad_norm": 6.790902614593506,
"learning_rate": 3e-06,
"loss": 0.8943,
"step": 16
},
{
"epoch": 0.0009784172661870504,
"grad_norm": 7.149299621582031,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.5624,
"step": 17
},
{
"epoch": 0.0010359712230215827,
"grad_norm": 6.0668439865112305,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.5257,
"step": 18
},
{
"epoch": 0.0010935251798561152,
"grad_norm": 7.927511692047119,
"learning_rate": 3.6000000000000003e-06,
"loss": 0.7319,
"step": 19
},
{
"epoch": 0.0011510791366906475,
"grad_norm": 3.352524757385254,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.3106,
"step": 20
},
{
"epoch": 0.0012086330935251798,
"grad_norm": 17.440523147583008,
"learning_rate": 4.000000000000001e-06,
"loss": 1.4364,
"step": 21
},
{
"epoch": 0.0012661870503597123,
"grad_norm": 3.9956955909729004,
"learning_rate": 4.2000000000000004e-06,
"loss": 0.4213,
"step": 22
},
{
"epoch": 0.0013237410071942446,
"grad_norm": 8.989081382751465,
"learning_rate": 4.4e-06,
"loss": 0.552,
"step": 23
},
{
"epoch": 0.001381294964028777,
"grad_norm": 8.471776008605957,
"learning_rate": 4.600000000000001e-06,
"loss": 0.8194,
"step": 24
},
{
"epoch": 0.0014388489208633094,
"grad_norm": 9.24136734008789,
"learning_rate": 4.800000000000001e-06,
"loss": 0.7052,
"step": 25
},
{
"epoch": 0.0014964028776978417,
"grad_norm": 4.3442158699035645,
"learning_rate": 5e-06,
"loss": 0.3331,
"step": 26
},
{
"epoch": 0.0015539568345323742,
"grad_norm": 9.657572746276855,
"learning_rate": 5.2e-06,
"loss": 0.5311,
"step": 27
},
{
"epoch": 0.0016115107913669065,
"grad_norm": 7.1384992599487305,
"learning_rate": 5.400000000000001e-06,
"loss": 0.3951,
"step": 28
},
{
"epoch": 0.0016690647482014388,
"grad_norm": 16.853891372680664,
"learning_rate": 5.600000000000001e-06,
"loss": 1.3869,
"step": 29
},
{
"epoch": 0.0017266187050359713,
"grad_norm": 9.793233871459961,
"learning_rate": 5.8e-06,
"loss": 0.8952,
"step": 30
},
{
"epoch": 0.0017841726618705036,
"grad_norm": 7.212555408477783,
"learning_rate": 6e-06,
"loss": 0.7198,
"step": 31
},
{
"epoch": 0.001841726618705036,
"grad_norm": 6.492921829223633,
"learning_rate": 6.200000000000001e-06,
"loss": 0.5619,
"step": 32
},
{
"epoch": 0.0018992805755395684,
"grad_norm": 6.3283915519714355,
"learning_rate": 6.4000000000000006e-06,
"loss": 1.0968,
"step": 33
},
{
"epoch": 0.0019568345323741007,
"grad_norm": 9.577136993408203,
"learning_rate": 6.600000000000001e-06,
"loss": 0.8173,
"step": 34
},
{
"epoch": 0.002014388489208633,
"grad_norm": 9.083664894104004,
"learning_rate": 6.800000000000001e-06,
"loss": 0.8457,
"step": 35
},
{
"epoch": 0.0020719424460431653,
"grad_norm": 4.176952362060547,
"learning_rate": 7e-06,
"loss": 0.3089,
"step": 36
},
{
"epoch": 0.002129496402877698,
"grad_norm": 5.284554958343506,
"learning_rate": 7.2000000000000005e-06,
"loss": 0.3452,
"step": 37
},
{
"epoch": 0.0021870503597122303,
"grad_norm": 12.007137298583984,
"learning_rate": 7.4e-06,
"loss": 1.4472,
"step": 38
},
{
"epoch": 0.0022446043165467626,
"grad_norm": 9.785158157348633,
"learning_rate": 7.600000000000001e-06,
"loss": 1.2897,
"step": 39
},
{
"epoch": 0.002302158273381295,
"grad_norm": 8.333564758300781,
"learning_rate": 7.800000000000002e-06,
"loss": 0.3833,
"step": 40
},
{
"epoch": 0.0023597122302158272,
"grad_norm": 5.145090103149414,
"learning_rate": 8.000000000000001e-06,
"loss": 0.2996,
"step": 41
},
{
"epoch": 0.0024172661870503595,
"grad_norm": 10.46713924407959,
"learning_rate": 8.2e-06,
"loss": 0.6667,
"step": 42
},
{
"epoch": 0.0024748201438848923,
"grad_norm": 4.508591175079346,
"learning_rate": 8.400000000000001e-06,
"loss": 0.4044,
"step": 43
},
{
"epoch": 0.0025323741007194246,
"grad_norm": 9.932887077331543,
"learning_rate": 8.6e-06,
"loss": 0.6707,
"step": 44
},
{
"epoch": 0.002589928057553957,
"grad_norm": 4.800012588500977,
"learning_rate": 8.8e-06,
"loss": 0.5579,
"step": 45
},
{
"epoch": 0.002647482014388489,
"grad_norm": 2.2920901775360107,
"learning_rate": 9e-06,
"loss": 0.2296,
"step": 46
},
{
"epoch": 0.0027050359712230214,
"grad_norm": 7.941588878631592,
"learning_rate": 9.200000000000002e-06,
"loss": 0.5172,
"step": 47
},
{
"epoch": 0.002762589928057554,
"grad_norm": 13.144063949584961,
"learning_rate": 9.4e-06,
"loss": 1.1074,
"step": 48
},
{
"epoch": 0.0028201438848920865,
"grad_norm": 4.377220630645752,
"learning_rate": 9.600000000000001e-06,
"loss": 0.2969,
"step": 49
},
{
"epoch": 0.0028776978417266188,
"grad_norm": 6.273579120635986,
"learning_rate": 9.800000000000001e-06,
"loss": 0.382,
"step": 50
},
{
"epoch": 0.002935251798561151,
"grad_norm": 6.432647705078125,
"learning_rate": 1e-05,
"loss": 0.7747,
"step": 51
},
{
"epoch": 0.0029928057553956834,
"grad_norm": 9.977968215942383,
"learning_rate": 1.02e-05,
"loss": 0.6861,
"step": 52
},
{
"epoch": 0.0030503597122302157,
"grad_norm": 6.838860034942627,
"learning_rate": 1.04e-05,
"loss": 0.4142,
"step": 53
},
{
"epoch": 0.0031079136690647484,
"grad_norm": 15.348876953125,
"learning_rate": 1.0600000000000002e-05,
"loss": 1.1043,
"step": 54
},
{
"epoch": 0.0031654676258992807,
"grad_norm": 4.6307291984558105,
"learning_rate": 1.0800000000000002e-05,
"loss": 0.5295,
"step": 55
},
{
"epoch": 0.003223021582733813,
"grad_norm": 13.125807762145996,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.9708,
"step": 56
},
{
"epoch": 0.0032805755395683453,
"grad_norm": 11.310308456420898,
"learning_rate": 1.1200000000000001e-05,
"loss": 1.0685,
"step": 57
},
{
"epoch": 0.0033381294964028776,
"grad_norm": 4.394148826599121,
"learning_rate": 1.14e-05,
"loss": 0.274,
"step": 58
},
{
"epoch": 0.00339568345323741,
"grad_norm": 2.017930030822754,
"learning_rate": 1.16e-05,
"loss": 0.1795,
"step": 59
},
{
"epoch": 0.0034532374100719426,
"grad_norm": 5.994359970092773,
"learning_rate": 1.18e-05,
"loss": 0.5362,
"step": 60
},
{
"epoch": 0.003510791366906475,
"grad_norm": 5.567146301269531,
"learning_rate": 1.2e-05,
"loss": 0.7449,
"step": 61
},
{
"epoch": 0.003568345323741007,
"grad_norm": 4.293890476226807,
"learning_rate": 1.22e-05,
"loss": 0.5452,
"step": 62
},
{
"epoch": 0.0036258992805755395,
"grad_norm": 8.327472686767578,
"learning_rate": 1.2400000000000002e-05,
"loss": 0.6844,
"step": 63
},
{
"epoch": 0.003683453237410072,
"grad_norm": 4.723773002624512,
"learning_rate": 1.2600000000000001e-05,
"loss": 0.3286,
"step": 64
},
{
"epoch": 0.0037410071942446045,
"grad_norm": 10.427044868469238,
"learning_rate": 1.2800000000000001e-05,
"loss": 0.7127,
"step": 65
},
{
"epoch": 0.003798561151079137,
"grad_norm": 3.913362741470337,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.4672,
"step": 66
},
{
"epoch": 0.003856115107913669,
"grad_norm": 4.887617588043213,
"learning_rate": 1.3200000000000002e-05,
"loss": 0.5162,
"step": 67
},
{
"epoch": 0.003913669064748201,
"grad_norm": 6.661849021911621,
"learning_rate": 1.3400000000000002e-05,
"loss": 0.6116,
"step": 68
},
{
"epoch": 0.003971223021582734,
"grad_norm": 7.069644451141357,
"learning_rate": 1.3600000000000002e-05,
"loss": 0.637,
"step": 69
},
{
"epoch": 0.004028776978417266,
"grad_norm": 4.68384313583374,
"learning_rate": 1.38e-05,
"loss": 0.4406,
"step": 70
},
{
"epoch": 0.004086330935251799,
"grad_norm": 5.5008134841918945,
"learning_rate": 1.4e-05,
"loss": 0.3483,
"step": 71
},
{
"epoch": 0.004143884892086331,
"grad_norm": 4.3354105949401855,
"learning_rate": 1.4200000000000001e-05,
"loss": 0.5613,
"step": 72
},
{
"epoch": 0.004201438848920863,
"grad_norm": 5.105152130126953,
"learning_rate": 1.4400000000000001e-05,
"loss": 0.3011,
"step": 73
},
{
"epoch": 0.004258992805755396,
"grad_norm": 2.7917816638946533,
"learning_rate": 1.46e-05,
"loss": 0.2686,
"step": 74
},
{
"epoch": 0.004316546762589928,
"grad_norm": 4.981333255767822,
"learning_rate": 1.48e-05,
"loss": 0.3631,
"step": 75
},
{
"epoch": 0.004374100719424461,
"grad_norm": 3.1867117881774902,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.4055,
"step": 76
},
{
"epoch": 0.0044316546762589925,
"grad_norm": 7.01395320892334,
"learning_rate": 1.5200000000000002e-05,
"loss": 0.5945,
"step": 77
},
{
"epoch": 0.004489208633093525,
"grad_norm": 7.518090724945068,
"learning_rate": 1.54e-05,
"loss": 0.9833,
"step": 78
},
{
"epoch": 0.004546762589928057,
"grad_norm": 3.8727071285247803,
"learning_rate": 1.5600000000000003e-05,
"loss": 0.3201,
"step": 79
},
{
"epoch": 0.00460431654676259,
"grad_norm": 3.5626542568206787,
"learning_rate": 1.58e-05,
"loss": 0.4482,
"step": 80
},
{
"epoch": 0.004661870503597123,
"grad_norm": 4.564299583435059,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.3674,
"step": 81
},
{
"epoch": 0.0047194244604316545,
"grad_norm": 3.150601387023926,
"learning_rate": 1.62e-05,
"loss": 0.4653,
"step": 82
},
{
"epoch": 0.004776978417266187,
"grad_norm": 4.487133502960205,
"learning_rate": 1.64e-05,
"loss": 0.3553,
"step": 83
},
{
"epoch": 0.004834532374100719,
"grad_norm": 9.726006507873535,
"learning_rate": 1.66e-05,
"loss": 0.7446,
"step": 84
},
{
"epoch": 0.004892086330935252,
"grad_norm": 4.138358116149902,
"learning_rate": 1.6800000000000002e-05,
"loss": 0.2264,
"step": 85
},
{
"epoch": 0.0049496402877697845,
"grad_norm": 4.193305015563965,
"learning_rate": 1.7e-05,
"loss": 0.3754,
"step": 86
},
{
"epoch": 0.005007194244604316,
"grad_norm": 5.699273109436035,
"learning_rate": 1.72e-05,
"loss": 0.2857,
"step": 87
},
{
"epoch": 0.005064748201438849,
"grad_norm": 3.635141134262085,
"learning_rate": 1.7400000000000003e-05,
"loss": 0.4759,
"step": 88
},
{
"epoch": 0.005122302158273381,
"grad_norm": 4.9962592124938965,
"learning_rate": 1.76e-05,
"loss": 0.3983,
"step": 89
},
{
"epoch": 0.005179856115107914,
"grad_norm": 9.600940704345703,
"learning_rate": 1.7800000000000002e-05,
"loss": 0.6997,
"step": 90
},
{
"epoch": 0.005237410071942446,
"grad_norm": 5.990379333496094,
"learning_rate": 1.8e-05,
"loss": 0.325,
"step": 91
},
{
"epoch": 0.005294964028776978,
"grad_norm": 3.9311363697052,
"learning_rate": 1.8200000000000002e-05,
"loss": 0.221,
"step": 92
},
{
"epoch": 0.005352517985611511,
"grad_norm": 4.358214378356934,
"learning_rate": 1.8400000000000003e-05,
"loss": 0.5484,
"step": 93
},
{
"epoch": 0.005410071942446043,
"grad_norm": 4.867093086242676,
"learning_rate": 1.86e-05,
"loss": 0.3255,
"step": 94
},
{
"epoch": 0.005467625899280576,
"grad_norm": 5.905974864959717,
"learning_rate": 1.88e-05,
"loss": 0.7437,
"step": 95
},
{
"epoch": 0.005525179856115108,
"grad_norm": 3.0502068996429443,
"learning_rate": 1.9e-05,
"loss": 0.3985,
"step": 96
},
{
"epoch": 0.00558273381294964,
"grad_norm": 3.0659244060516357,
"learning_rate": 1.9200000000000003e-05,
"loss": 0.4518,
"step": 97
},
{
"epoch": 0.005640287769784173,
"grad_norm": 5.97324800491333,
"learning_rate": 1.94e-05,
"loss": 0.5574,
"step": 98
},
{
"epoch": 0.005697841726618705,
"grad_norm": 5.410953998565674,
"learning_rate": 1.9600000000000002e-05,
"loss": 0.6145,
"step": 99
},
{
"epoch": 0.0057553956834532375,
"grad_norm": 3.2696614265441895,
"learning_rate": 1.98e-05,
"loss": 0.3616,
"step": 100
},
{
"epoch": 0.005812949640287769,
"grad_norm": 6.245858192443848,
"learning_rate": 2e-05,
"loss": 0.2664,
"step": 101
},
{
"epoch": 0.005870503597122302,
"grad_norm": 8.132523536682129,
"learning_rate": 1.9999939076577906e-05,
"loss": 0.4025,
"step": 102
},
{
"epoch": 0.005928057553956835,
"grad_norm": 11.821637153625488,
"learning_rate": 1.9999756307053947e-05,
"loss": 0.9175,
"step": 103
},
{
"epoch": 0.005985611510791367,
"grad_norm": 5.507436752319336,
"learning_rate": 1.9999451693655125e-05,
"loss": 0.5407,
"step": 104
},
{
"epoch": 0.0060431654676258995,
"grad_norm": 2.0903306007385254,
"learning_rate": 1.9999025240093045e-05,
"loss": 0.2037,
"step": 105
},
{
"epoch": 0.006100719424460431,
"grad_norm": 6.703708648681641,
"learning_rate": 1.9998476951563914e-05,
"loss": 0.3923,
"step": 106
},
{
"epoch": 0.006158273381294964,
"grad_norm": 7.479000091552734,
"learning_rate": 1.9997806834748455e-05,
"loss": 0.5959,
"step": 107
},
{
"epoch": 0.006215827338129497,
"grad_norm": 3.7436177730560303,
"learning_rate": 1.9997014897811834e-05,
"loss": 0.3848,
"step": 108
},
{
"epoch": 0.006273381294964029,
"grad_norm": 6.0655388832092285,
"learning_rate": 1.9996101150403543e-05,
"loss": 0.3867,
"step": 109
},
{
"epoch": 0.006330935251798561,
"grad_norm": 2.6236915588378906,
"learning_rate": 1.9995065603657317e-05,
"loss": 0.2235,
"step": 110
},
{
"epoch": 0.006388489208633093,
"grad_norm": 3.6318767070770264,
"learning_rate": 1.999390827019096e-05,
"loss": 0.2775,
"step": 111
},
{
"epoch": 0.006446043165467626,
"grad_norm": 3.6131348609924316,
"learning_rate": 1.999262916410621e-05,
"loss": 0.473,
"step": 112
},
{
"epoch": 0.006503597122302159,
"grad_norm": 5.235411167144775,
"learning_rate": 1.9991228300988586e-05,
"loss": 0.5327,
"step": 113
},
{
"epoch": 0.0065611510791366906,
"grad_norm": 4.830928802490234,
"learning_rate": 1.998970569790715e-05,
"loss": 0.3054,
"step": 114
},
{
"epoch": 0.006618705035971223,
"grad_norm": 3.149054527282715,
"learning_rate": 1.9988061373414342e-05,
"loss": 0.2049,
"step": 115
},
{
"epoch": 0.006676258992805755,
"grad_norm": 0.9228267669677734,
"learning_rate": 1.9986295347545738e-05,
"loss": 0.137,
"step": 116
},
{
"epoch": 0.006733812949640288,
"grad_norm": 1.2999602556228638,
"learning_rate": 1.9984407641819812e-05,
"loss": 0.1165,
"step": 117
},
{
"epoch": 0.00679136690647482,
"grad_norm": 5.368565082550049,
"learning_rate": 1.9982398279237657e-05,
"loss": 0.4314,
"step": 118
},
{
"epoch": 0.0068489208633093525,
"grad_norm": 11.419666290283203,
"learning_rate": 1.9980267284282718e-05,
"loss": 0.7708,
"step": 119
},
{
"epoch": 0.006906474820143885,
"grad_norm": 5.084501266479492,
"learning_rate": 1.9978014682920503e-05,
"loss": 0.2652,
"step": 120
},
{
"epoch": 0.006964028776978417,
"grad_norm": 5.778707027435303,
"learning_rate": 1.9975640502598243e-05,
"loss": 0.6306,
"step": 121
},
{
"epoch": 0.00702158273381295,
"grad_norm": 4.678982257843018,
"learning_rate": 1.997314477224458e-05,
"loss": 0.2769,
"step": 122
},
{
"epoch": 0.007079136690647482,
"grad_norm": 3.2127490043640137,
"learning_rate": 1.9970527522269204e-05,
"loss": 0.3237,
"step": 123
},
{
"epoch": 0.007136690647482014,
"grad_norm": 8.614684104919434,
"learning_rate": 1.9967788784562474e-05,
"loss": 0.7173,
"step": 124
},
{
"epoch": 0.007194244604316547,
"grad_norm": 7.342788219451904,
"learning_rate": 1.9964928592495046e-05,
"loss": 0.6905,
"step": 125
},
{
"epoch": 0.007251798561151079,
"grad_norm": 3.408464193344116,
"learning_rate": 1.9961946980917457e-05,
"loss": 0.2835,
"step": 126
},
{
"epoch": 0.007309352517985612,
"grad_norm": 4.440028667449951,
"learning_rate": 1.9958843986159705e-05,
"loss": 0.2777,
"step": 127
},
{
"epoch": 0.007366906474820144,
"grad_norm": 2.4154090881347656,
"learning_rate": 1.99556196460308e-05,
"loss": 0.205,
"step": 128
},
{
"epoch": 0.007424460431654676,
"grad_norm": 2.5792348384857178,
"learning_rate": 1.9952273999818312e-05,
"loss": 0.2947,
"step": 129
},
{
"epoch": 0.007482014388489209,
"grad_norm": 5.055807113647461,
"learning_rate": 1.9948807088287884e-05,
"loss": 0.4372,
"step": 130
},
{
"epoch": 0.007539568345323741,
"grad_norm": 5.54684591293335,
"learning_rate": 1.9945218953682736e-05,
"loss": 0.305,
"step": 131
},
{
"epoch": 0.007597122302158274,
"grad_norm": 5.550899028778076,
"learning_rate": 1.9941509639723155e-05,
"loss": 0.5005,
"step": 132
},
{
"epoch": 0.0076546762589928055,
"grad_norm": 6.129327297210693,
"learning_rate": 1.9937679191605964e-05,
"loss": 0.7251,
"step": 133
},
{
"epoch": 0.007712230215827338,
"grad_norm": 14.2527437210083,
"learning_rate": 1.9933727656003964e-05,
"loss": 0.6724,
"step": 134
},
{
"epoch": 0.00776978417266187,
"grad_norm": 5.460100173950195,
"learning_rate": 1.992965508106537e-05,
"loss": 0.3455,
"step": 135
},
{
"epoch": 0.007827338129496403,
"grad_norm": 6.3173828125,
"learning_rate": 1.9925461516413224e-05,
"loss": 0.5253,
"step": 136
},
{
"epoch": 0.007884892086330935,
"grad_norm": 2.2781403064727783,
"learning_rate": 1.9921147013144782e-05,
"loss": 0.284,
"step": 137
},
{
"epoch": 0.007942446043165468,
"grad_norm": 8.100839614868164,
"learning_rate": 1.9916711623830904e-05,
"loss": 0.561,
"step": 138
},
{
"epoch": 0.008,
"grad_norm": 2.185936212539673,
"learning_rate": 1.991215540251542e-05,
"loss": 0.2032,
"step": 139
},
{
"epoch": 0.008057553956834532,
"grad_norm": 3.2472143173217773,
"learning_rate": 1.9907478404714438e-05,
"loss": 0.3126,
"step": 140
},
{
"epoch": 0.008115107913669064,
"grad_norm": 2.253760576248169,
"learning_rate": 1.9902680687415704e-05,
"loss": 0.28,
"step": 141
},
{
"epoch": 0.008172661870503597,
"grad_norm": 2.6927177906036377,
"learning_rate": 1.989776230907789e-05,
"loss": 0.32,
"step": 142
},
{
"epoch": 0.00823021582733813,
"grad_norm": 6.3567681312561035,
"learning_rate": 1.9892723329629885e-05,
"loss": 0.4475,
"step": 143
},
{
"epoch": 0.008287769784172661,
"grad_norm": 2.409682512283325,
"learning_rate": 1.988756381047006e-05,
"loss": 0.1429,
"step": 144
},
{
"epoch": 0.008345323741007195,
"grad_norm": 1.8926228284835815,
"learning_rate": 1.988228381446553e-05,
"loss": 0.1483,
"step": 145
},
{
"epoch": 0.008402877697841727,
"grad_norm": 4.373330116271973,
"learning_rate": 1.9876883405951378e-05,
"loss": 0.2895,
"step": 146
},
{
"epoch": 0.008460431654676259,
"grad_norm": 5.756573677062988,
"learning_rate": 1.987136265072988e-05,
"loss": 0.3329,
"step": 147
},
{
"epoch": 0.008517985611510792,
"grad_norm": 4.497378826141357,
"learning_rate": 1.9865721616069695e-05,
"loss": 0.4046,
"step": 148
},
{
"epoch": 0.008575539568345324,
"grad_norm": 3.908770799636841,
"learning_rate": 1.985996037070505e-05,
"loss": 0.3289,
"step": 149
},
{
"epoch": 0.008633093525179856,
"grad_norm": 9.844822883605957,
"learning_rate": 1.9854078984834904e-05,
"loss": 0.4939,
"step": 150
},
{
"epoch": 0.008690647482014388,
"grad_norm": 8.726676940917969,
"learning_rate": 1.9848077530122083e-05,
"loss": 0.5217,
"step": 151
},
{
"epoch": 0.008748201438848921,
"grad_norm": 2.3976569175720215,
"learning_rate": 1.984195607969242e-05,
"loss": 0.0699,
"step": 152
},
{
"epoch": 0.008805755395683453,
"grad_norm": 5.398916244506836,
"learning_rate": 1.983571470813386e-05,
"loss": 0.4521,
"step": 153
},
{
"epoch": 0.008863309352517985,
"grad_norm": 6.7461137771606445,
"learning_rate": 1.9829353491495545e-05,
"loss": 0.3828,
"step": 154
},
{
"epoch": 0.008920863309352519,
"grad_norm": 4.338736057281494,
"learning_rate": 1.982287250728689e-05,
"loss": 0.3047,
"step": 155
},
{
"epoch": 0.00897841726618705,
"grad_norm": 6.214763164520264,
"learning_rate": 1.9816271834476642e-05,
"loss": 0.3477,
"step": 156
},
{
"epoch": 0.009035971223021582,
"grad_norm": 2.8336939811706543,
"learning_rate": 1.9809551553491918e-05,
"loss": 0.3321,
"step": 157
},
{
"epoch": 0.009093525179856114,
"grad_norm": 4.502871990203857,
"learning_rate": 1.9802711746217222e-05,
"loss": 0.4364,
"step": 158
},
{
"epoch": 0.009151079136690648,
"grad_norm": 8.72002124786377,
"learning_rate": 1.979575249599344e-05,
"loss": 0.4955,
"step": 159
},
{
"epoch": 0.00920863309352518,
"grad_norm": 6.201029300689697,
"learning_rate": 1.9788673887616852e-05,
"loss": 0.2945,
"step": 160
},
{
"epoch": 0.009266187050359712,
"grad_norm": 8.309669494628906,
"learning_rate": 1.9781476007338058e-05,
"loss": 0.3601,
"step": 161
},
{
"epoch": 0.009323741007194245,
"grad_norm": 2.303840398788452,
"learning_rate": 1.9774158942860962e-05,
"loss": 0.2439,
"step": 162
},
{
"epoch": 0.009381294964028777,
"grad_norm": 1.9197461605072021,
"learning_rate": 1.9766722783341682e-05,
"loss": 0.1073,
"step": 163
},
{
"epoch": 0.009438848920863309,
"grad_norm": 2.906818389892578,
"learning_rate": 1.9759167619387474e-05,
"loss": 0.3538,
"step": 164
},
{
"epoch": 0.009496402877697842,
"grad_norm": 6.95295524597168,
"learning_rate": 1.9751493543055634e-05,
"loss": 0.7104,
"step": 165
},
{
"epoch": 0.009553956834532374,
"grad_norm": 13.919471740722656,
"learning_rate": 1.9743700647852356e-05,
"loss": 0.5481,
"step": 166
},
{
"epoch": 0.009611510791366906,
"grad_norm": 5.098160743713379,
"learning_rate": 1.9735789028731603e-05,
"loss": 0.6219,
"step": 167
},
{
"epoch": 0.009669064748201438,
"grad_norm": 2.4541916847229004,
"learning_rate": 1.972775878209397e-05,
"loss": 0.1824,
"step": 168
},
{
"epoch": 0.009726618705035972,
"grad_norm": 2.5615649223327637,
"learning_rate": 1.9719610005785466e-05,
"loss": 0.2294,
"step": 169
},
{
"epoch": 0.009784172661870504,
"grad_norm": 5.56582498550415,
"learning_rate": 1.971134279909636e-05,
"loss": 0.4881,
"step": 170
},
{
"epoch": 0.009841726618705035,
"grad_norm": 3.3772428035736084,
"learning_rate": 1.9702957262759964e-05,
"loss": 0.2457,
"step": 171
},
{
"epoch": 0.009899280575539569,
"grad_norm": 3.7561800479888916,
"learning_rate": 1.9694453498951392e-05,
"loss": 0.2876,
"step": 172
},
{
"epoch": 0.009956834532374101,
"grad_norm": 4.44319486618042,
"learning_rate": 1.9685831611286312e-05,
"loss": 0.3955,
"step": 173
},
{
"epoch": 0.010014388489208633,
"grad_norm": 7.627588748931885,
"learning_rate": 1.9677091704819714e-05,
"loss": 0.5145,
"step": 174
},
{
"epoch": 0.010071942446043165,
"grad_norm": 2.9267642498016357,
"learning_rate": 1.9668233886044597e-05,
"loss": 0.2853,
"step": 175
},
{
"epoch": 0.010129496402877698,
"grad_norm": 2.577763319015503,
"learning_rate": 1.9659258262890683e-05,
"loss": 0.1479,
"step": 176
},
{
"epoch": 0.01018705035971223,
"grad_norm": 3.698162794113159,
"learning_rate": 1.9650164944723116e-05,
"loss": 0.3178,
"step": 177
},
{
"epoch": 0.010244604316546762,
"grad_norm": 3.1427218914031982,
"learning_rate": 1.96409540423411e-05,
"loss": 0.3621,
"step": 178
},
{
"epoch": 0.010302158273381296,
"grad_norm": 6.883255958557129,
"learning_rate": 1.9631625667976584e-05,
"loss": 0.6752,
"step": 179
},
{
"epoch": 0.010359712230215827,
"grad_norm": 5.5977044105529785,
"learning_rate": 1.9622179935292855e-05,
"loss": 0.2751,
"step": 180
},
{
"epoch": 0.01041726618705036,
"grad_norm": 6.071504592895508,
"learning_rate": 1.961261695938319e-05,
"loss": 0.2445,
"step": 181
},
{
"epoch": 0.010474820143884893,
"grad_norm": 3.8929355144500732,
"learning_rate": 1.9602936856769432e-05,
"loss": 0.3608,
"step": 182
},
{
"epoch": 0.010532374100719425,
"grad_norm": 3.836050033569336,
"learning_rate": 1.9593139745400575e-05,
"loss": 0.2327,
"step": 183
},
{
"epoch": 0.010589928057553957,
"grad_norm": 11.480710983276367,
"learning_rate": 1.9583225744651334e-05,
"loss": 0.6428,
"step": 184
},
{
"epoch": 0.010647482014388488,
"grad_norm": 4.123970985412598,
"learning_rate": 1.9573194975320672e-05,
"loss": 0.2694,
"step": 185
},
{
"epoch": 0.010705035971223022,
"grad_norm": 5.072369575500488,
"learning_rate": 1.9563047559630356e-05,
"loss": 0.3314,
"step": 186
},
{
"epoch": 0.010762589928057554,
"grad_norm": 6.05760383605957,
"learning_rate": 1.9552783621223437e-05,
"loss": 0.4093,
"step": 187
},
{
"epoch": 0.010820143884892086,
"grad_norm": 1.4690526723861694,
"learning_rate": 1.954240328516277e-05,
"loss": 0.1567,
"step": 188
},
{
"epoch": 0.01087769784172662,
"grad_norm": 4.700074672698975,
"learning_rate": 1.9531906677929472e-05,
"loss": 0.224,
"step": 189
},
{
"epoch": 0.010935251798561151,
"grad_norm": 7.230707168579102,
"learning_rate": 1.9521293927421388e-05,
"loss": 0.5614,
"step": 190
},
{
"epoch": 0.010992805755395683,
"grad_norm": 3.8735275268554688,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.2567,
"step": 191
},
{
"epoch": 0.011050359712230217,
"grad_norm": 1.882738471031189,
"learning_rate": 1.9499720515246524e-05,
"loss": 0.1471,
"step": 192
},
{
"epoch": 0.011107913669064749,
"grad_norm": 11.0597505569458,
"learning_rate": 1.9488760116444966e-05,
"loss": 0.6557,
"step": 193
},
{
"epoch": 0.01116546762589928,
"grad_norm": 8.127897262573242,
"learning_rate": 1.947768410009586e-05,
"loss": 0.6195,
"step": 194
},
{
"epoch": 0.011223021582733812,
"grad_norm": 10.546082496643066,
"learning_rate": 1.9466492601156964e-05,
"loss": 0.6428,
"step": 195
},
{
"epoch": 0.011280575539568346,
"grad_norm": 5.149999141693115,
"learning_rate": 1.945518575599317e-05,
"loss": 0.5332,
"step": 196
},
{
"epoch": 0.011338129496402878,
"grad_norm": 1.2085461616516113,
"learning_rate": 1.944376370237481e-05,
"loss": 0.0889,
"step": 197
},
{
"epoch": 0.01139568345323741,
"grad_norm": 4.670867919921875,
"learning_rate": 1.943222657947601e-05,
"loss": 0.3224,
"step": 198
},
{
"epoch": 0.011453237410071943,
"grad_norm": 10.440308570861816,
"learning_rate": 1.942057452787297e-05,
"loss": 0.5673,
"step": 199
},
{
"epoch": 0.011510791366906475,
"grad_norm": 2.601593017578125,
"learning_rate": 1.9408807689542257e-05,
"loss": 0.1987,
"step": 200
},
{
"epoch": 0.011568345323741007,
"grad_norm": 3.630814552307129,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.1889,
"step": 201
},
{
"epoch": 0.011625899280575539,
"grad_norm": 5.2653632164001465,
"learning_rate": 1.938493022759556e-05,
"loss": 0.5126,
"step": 202
},
{
"epoch": 0.011683453237410072,
"grad_norm": 5.171159744262695,
"learning_rate": 1.937281989491892e-05,
"loss": 0.4045,
"step": 203
},
{
"epoch": 0.011741007194244604,
"grad_norm": 4.25119686126709,
"learning_rate": 1.9360595357389735e-05,
"loss": 0.4102,
"step": 204
},
{
"epoch": 0.011798561151079136,
"grad_norm": 4.690221309661865,
"learning_rate": 1.9348256763960146e-05,
"loss": 0.1851,
"step": 205
},
{
"epoch": 0.01185611510791367,
"grad_norm": 2.5318408012390137,
"learning_rate": 1.9335804264972018e-05,
"loss": 0.1637,
"step": 206
},
{
"epoch": 0.011913669064748202,
"grad_norm": 3.1044423580169678,
"learning_rate": 1.9323238012155125e-05,
"loss": 0.2009,
"step": 207
},
{
"epoch": 0.011971223021582733,
"grad_norm": 3.260160207748413,
"learning_rate": 1.9310558158625286e-05,
"loss": 0.3784,
"step": 208
},
{
"epoch": 0.012028776978417267,
"grad_norm": 2.753958225250244,
"learning_rate": 1.9297764858882516e-05,
"loss": 0.2527,
"step": 209
},
{
"epoch": 0.012086330935251799,
"grad_norm": 2.3980395793914795,
"learning_rate": 1.9284858268809135e-05,
"loss": 0.2148,
"step": 210
},
{
"epoch": 0.01214388489208633,
"grad_norm": 7.618435382843018,
"learning_rate": 1.9271838545667876e-05,
"loss": 0.4407,
"step": 211
},
{
"epoch": 0.012201438848920863,
"grad_norm": 2.246246099472046,
"learning_rate": 1.925870584809995e-05,
"loss": 0.2267,
"step": 212
},
{
"epoch": 0.012258992805755396,
"grad_norm": 6.476871490478516,
"learning_rate": 1.9245460336123136e-05,
"loss": 0.3225,
"step": 213
},
{
"epoch": 0.012316546762589928,
"grad_norm": 3.0804295539855957,
"learning_rate": 1.923210217112981e-05,
"loss": 0.2715,
"step": 214
},
{
"epoch": 0.01237410071942446,
"grad_norm": 2.8532204627990723,
"learning_rate": 1.9218631515885007e-05,
"loss": 0.1833,
"step": 215
},
{
"epoch": 0.012431654676258994,
"grad_norm": 5.331664562225342,
"learning_rate": 1.9205048534524405e-05,
"loss": 0.2848,
"step": 216
},
{
"epoch": 0.012489208633093525,
"grad_norm": 6.048542022705078,
"learning_rate": 1.9191353392552346e-05,
"loss": 0.5696,
"step": 217
},
{
"epoch": 0.012546762589928057,
"grad_norm": 1.7071490287780762,
"learning_rate": 1.9177546256839814e-05,
"loss": 0.1342,
"step": 218
},
{
"epoch": 0.01260431654676259,
"grad_norm": 2.181581974029541,
"learning_rate": 1.9163627295622397e-05,
"loss": 0.123,
"step": 219
},
{
"epoch": 0.012661870503597123,
"grad_norm": 3.1651432514190674,
"learning_rate": 1.914959667849825e-05,
"loss": 0.1772,
"step": 220
},
{
"epoch": 0.012719424460431655,
"grad_norm": 5.666025638580322,
"learning_rate": 1.913545457642601e-05,
"loss": 0.3085,
"step": 221
},
{
"epoch": 0.012776978417266186,
"grad_norm": 2.379805088043213,
"learning_rate": 1.9121201161722732e-05,
"loss": 0.1892,
"step": 222
},
{
"epoch": 0.01283453237410072,
"grad_norm": 2.259615182876587,
"learning_rate": 1.910683660806177e-05,
"loss": 0.1345,
"step": 223
},
{
"epoch": 0.012892086330935252,
"grad_norm": 5.213752746582031,
"learning_rate": 1.9092361090470688e-05,
"loss": 0.3298,
"step": 224
},
{
"epoch": 0.012949640287769784,
"grad_norm": 4.78728723526001,
"learning_rate": 1.907777478532909e-05,
"loss": 0.2825,
"step": 225
},
{
"epoch": 0.013007194244604317,
"grad_norm": 0.8624676465988159,
"learning_rate": 1.9063077870366504e-05,
"loss": 0.109,
"step": 226
},
{
"epoch": 0.01306474820143885,
"grad_norm": 1.283377766609192,
"learning_rate": 1.9048270524660197e-05,
"loss": 0.1188,
"step": 227
},
{
"epoch": 0.013122302158273381,
"grad_norm": 9.221684455871582,
"learning_rate": 1.903335292863301e-05,
"loss": 0.5137,
"step": 228
},
{
"epoch": 0.013179856115107913,
"grad_norm": 11.753008842468262,
"learning_rate": 1.901832526405114e-05,
"loss": 0.4472,
"step": 229
},
{
"epoch": 0.013237410071942447,
"grad_norm": 1.6660693883895874,
"learning_rate": 1.9003187714021936e-05,
"loss": 0.1548,
"step": 230
},
{
"epoch": 0.013294964028776978,
"grad_norm": 0.9586846232414246,
"learning_rate": 1.8987940462991673e-05,
"loss": 0.0959,
"step": 231
},
{
"epoch": 0.01335251798561151,
"grad_norm": 0.9193116426467896,
"learning_rate": 1.8972583696743284e-05,
"loss": 0.0977,
"step": 232
},
{
"epoch": 0.013410071942446044,
"grad_norm": 0.9497302174568176,
"learning_rate": 1.895711760239413e-05,
"loss": 0.0993,
"step": 233
},
{
"epoch": 0.013467625899280576,
"grad_norm": 2.8770296573638916,
"learning_rate": 1.8941542368393683e-05,
"loss": 0.1175,
"step": 234
},
{
"epoch": 0.013525179856115108,
"grad_norm": 5.624111175537109,
"learning_rate": 1.892585818452126e-05,
"loss": 0.2421,
"step": 235
},
{
"epoch": 0.01358273381294964,
"grad_norm": 1.8553048372268677,
"learning_rate": 1.891006524188368e-05,
"loss": 0.1538,
"step": 236
},
{
"epoch": 0.013640287769784173,
"grad_norm": 6.076197624206543,
"learning_rate": 1.889416373291298e-05,
"loss": 0.2427,
"step": 237
},
{
"epoch": 0.013697841726618705,
"grad_norm": 3.465892791748047,
"learning_rate": 1.8878153851364013e-05,
"loss": 0.1723,
"step": 238
},
{
"epoch": 0.013755395683453237,
"grad_norm": 15.277708053588867,
"learning_rate": 1.8862035792312148e-05,
"loss": 0.5853,
"step": 239
},
{
"epoch": 0.01381294964028777,
"grad_norm": 4.846475124359131,
"learning_rate": 1.884580975215084e-05,
"loss": 0.2031,
"step": 240
},
{
"epoch": 0.013870503597122302,
"grad_norm": 3.0435233116149902,
"learning_rate": 1.8829475928589272e-05,
"loss": 0.2053,
"step": 241
},
{
"epoch": 0.013928057553956834,
"grad_norm": 6.824376583099365,
"learning_rate": 1.8813034520649923e-05,
"loss": 0.1688,
"step": 242
},
{
"epoch": 0.013985611510791368,
"grad_norm": 1.9684380292892456,
"learning_rate": 1.879648572866617e-05,
"loss": 0.1105,
"step": 243
},
{
"epoch": 0.0140431654676259,
"grad_norm": 7.668671131134033,
"learning_rate": 1.8779829754279806e-05,
"loss": 0.4672,
"step": 244
},
{
"epoch": 0.014100719424460431,
"grad_norm": 7.829854488372803,
"learning_rate": 1.8763066800438638e-05,
"loss": 0.229,
"step": 245
},
{
"epoch": 0.014158273381294963,
"grad_norm": 2.082008123397827,
"learning_rate": 1.874619707139396e-05,
"loss": 0.1406,
"step": 246
},
{
"epoch": 0.014215827338129497,
"grad_norm": 6.171285629272461,
"learning_rate": 1.8729220772698096e-05,
"loss": 0.1633,
"step": 247
},
{
"epoch": 0.014273381294964029,
"grad_norm": 2.4114866256713867,
"learning_rate": 1.8712138111201898e-05,
"loss": 0.1011,
"step": 248
},
{
"epoch": 0.01433093525179856,
"grad_norm": 6.919300079345703,
"learning_rate": 1.869494929505219e-05,
"loss": 0.2706,
"step": 249
},
{
"epoch": 0.014388489208633094,
"grad_norm": 2.967616558074951,
"learning_rate": 1.8677654533689287e-05,
"loss": 0.1721,
"step": 250
},
{
"epoch": 0.014446043165467626,
"grad_norm": 19.468122482299805,
"learning_rate": 1.866025403784439e-05,
"loss": 0.7281,
"step": 251
},
{
"epoch": 0.014503597122302158,
"grad_norm": 5.602455139160156,
"learning_rate": 1.864274801953705e-05,
"loss": 0.108,
"step": 252
},
{
"epoch": 0.01456115107913669,
"grad_norm": 16.94603729248047,
"learning_rate": 1.8625136692072577e-05,
"loss": 0.4322,
"step": 253
},
{
"epoch": 0.014618705035971223,
"grad_norm": 5.4511823654174805,
"learning_rate": 1.860742027003944e-05,
"loss": 0.2628,
"step": 254
},
{
"epoch": 0.014676258992805755,
"grad_norm": 3.1577935218811035,
"learning_rate": 1.8589598969306646e-05,
"loss": 0.1408,
"step": 255
},
{
"epoch": 0.014733812949640287,
"grad_norm": 0.9684641361236572,
"learning_rate": 1.8571673007021124e-05,
"loss": 0.0771,
"step": 256
},
{
"epoch": 0.01479136690647482,
"grad_norm": 10.628517150878906,
"learning_rate": 1.855364260160507e-05,
"loss": 0.4681,
"step": 257
},
{
"epoch": 0.014848920863309353,
"grad_norm": 1.4679758548736572,
"learning_rate": 1.8535507972753275e-05,
"loss": 0.1161,
"step": 258
},
{
"epoch": 0.014906474820143885,
"grad_norm": 2.098684549331665,
"learning_rate": 1.851726934143048e-05,
"loss": 0.1365,
"step": 259
},
{
"epoch": 0.014964028776978418,
"grad_norm": 1.1517246961593628,
"learning_rate": 1.849892692986864e-05,
"loss": 0.1261,
"step": 260
},
{
"epoch": 0.01502158273381295,
"grad_norm": 11.601617813110352,
"learning_rate": 1.848048096156426e-05,
"loss": 0.5597,
"step": 261
},
{
"epoch": 0.015079136690647482,
"grad_norm": 9.081189155578613,
"learning_rate": 1.8461931661275642e-05,
"loss": 0.3012,
"step": 262
},
{
"epoch": 0.015136690647482014,
"grad_norm": 1.1501121520996094,
"learning_rate": 1.8443279255020153e-05,
"loss": 0.1101,
"step": 263
},
{
"epoch": 0.015194244604316547,
"grad_norm": 1.6358212232589722,
"learning_rate": 1.842452397007148e-05,
"loss": 0.0907,
"step": 264
},
{
"epoch": 0.01525179856115108,
"grad_norm": 1.1782243251800537,
"learning_rate": 1.8405666034956842e-05,
"loss": 0.0821,
"step": 265
},
{
"epoch": 0.015309352517985611,
"grad_norm": 3.5769565105438232,
"learning_rate": 1.8386705679454243e-05,
"loss": 0.1347,
"step": 266
},
{
"epoch": 0.015366906474820145,
"grad_norm": 9.618369102478027,
"learning_rate": 1.836764313458962e-05,
"loss": 0.2999,
"step": 267
},
{
"epoch": 0.015424460431654676,
"grad_norm": 3.8770649433135986,
"learning_rate": 1.8348478632634067e-05,
"loss": 0.1558,
"step": 268
},
{
"epoch": 0.015482014388489208,
"grad_norm": 5.656462669372559,
"learning_rate": 1.8329212407100996e-05,
"loss": 0.2134,
"step": 269
},
{
"epoch": 0.01553956834532374,
"grad_norm": 7.180418491363525,
"learning_rate": 1.8309844692743283e-05,
"loss": 0.2609,
"step": 270
},
{
"epoch": 0.015597122302158274,
"grad_norm": 1.3020349740982056,
"learning_rate": 1.8290375725550417e-05,
"loss": 0.0988,
"step": 271
},
{
"epoch": 0.015654676258992806,
"grad_norm": 5.860103607177734,
"learning_rate": 1.827080574274562e-05,
"loss": 0.2128,
"step": 272
},
{
"epoch": 0.01571223021582734,
"grad_norm": 10.250383377075195,
"learning_rate": 1.8251134982782952e-05,
"loss": 0.2957,
"step": 273
},
{
"epoch": 0.01576978417266187,
"grad_norm": 0.8903178572654724,
"learning_rate": 1.8231363685344422e-05,
"loss": 0.1002,
"step": 274
},
{
"epoch": 0.015827338129496403,
"grad_norm": 9.552652359008789,
"learning_rate": 1.821149209133704e-05,
"loss": 0.2226,
"step": 275
},
{
"epoch": 0.015884892086330937,
"grad_norm": 8.22533893585205,
"learning_rate": 1.819152044288992e-05,
"loss": 0.4168,
"step": 276
},
{
"epoch": 0.015942446043165467,
"grad_norm": 3.7411088943481445,
"learning_rate": 1.8171448983351284e-05,
"loss": 0.133,
"step": 277
},
{
"epoch": 0.016,
"grad_norm": 6.830077171325684,
"learning_rate": 1.815127795728554e-05,
"loss": 0.1879,
"step": 278
},
{
"epoch": 0.016057553956834534,
"grad_norm": 6.609959125518799,
"learning_rate": 1.8131007610470278e-05,
"loss": 0.3989,
"step": 279
},
{
"epoch": 0.016115107913669064,
"grad_norm": 8.54151725769043,
"learning_rate": 1.8110638189893267e-05,
"loss": 0.2935,
"step": 280
},
{
"epoch": 0.016172661870503598,
"grad_norm": 2.277355909347534,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.0824,
"step": 281
},
{
"epoch": 0.016230215827338128,
"grad_norm": 0.6691151857376099,
"learning_rate": 1.806960312143802e-05,
"loss": 0.0981,
"step": 282
},
{
"epoch": 0.01628776978417266,
"grad_norm": 1.0405458211898804,
"learning_rate": 1.804893797355914e-05,
"loss": 0.1054,
"step": 283
},
{
"epoch": 0.016345323741007195,
"grad_norm": 3.0072195529937744,
"learning_rate": 1.8028174751911147e-05,
"loss": 0.1007,
"step": 284
},
{
"epoch": 0.016402877697841725,
"grad_norm": 5.276819229125977,
"learning_rate": 1.8007313709487334e-05,
"loss": 0.2197,
"step": 285
},
{
"epoch": 0.01646043165467626,
"grad_norm": 0.3860107362270355,
"learning_rate": 1.798635510047293e-05,
"loss": 0.0449,
"step": 286
},
{
"epoch": 0.016517985611510792,
"grad_norm": 7.217343330383301,
"learning_rate": 1.7965299180241963e-05,
"loss": 0.2553,
"step": 287
},
{
"epoch": 0.016575539568345322,
"grad_norm": 6.964923858642578,
"learning_rate": 1.7944146205354182e-05,
"loss": 0.3295,
"step": 288
},
{
"epoch": 0.016633093525179856,
"grad_norm": 6.25247859954834,
"learning_rate": 1.792289643355191e-05,
"loss": 0.2184,
"step": 289
},
{
"epoch": 0.01669064748201439,
"grad_norm": 2.1732687950134277,
"learning_rate": 1.7901550123756906e-05,
"loss": 0.0992,
"step": 290
},
{
"epoch": 0.01674820143884892,
"grad_norm": 4.60042142868042,
"learning_rate": 1.788010753606722e-05,
"loss": 0.241,
"step": 291
},
{
"epoch": 0.016805755395683453,
"grad_norm": 1.7903494834899902,
"learning_rate": 1.785856893175402e-05,
"loss": 0.1072,
"step": 292
},
{
"epoch": 0.016863309352517987,
"grad_norm": 9.696785926818848,
"learning_rate": 1.78369345732584e-05,
"loss": 0.2978,
"step": 293
},
{
"epoch": 0.016920863309352517,
"grad_norm": 9.798710823059082,
"learning_rate": 1.781520472418819e-05,
"loss": 0.3003,
"step": 294
},
{
"epoch": 0.01697841726618705,
"grad_norm": 1.50823175907135,
"learning_rate": 1.7793379649314743e-05,
"loss": 0.1167,
"step": 295
},
{
"epoch": 0.017035971223021584,
"grad_norm": 15.534778594970703,
"learning_rate": 1.777145961456971e-05,
"loss": 0.3782,
"step": 296
},
{
"epoch": 0.017093525179856114,
"grad_norm": 1.2554796934127808,
"learning_rate": 1.7749444887041797e-05,
"loss": 0.0898,
"step": 297
},
{
"epoch": 0.017151079136690648,
"grad_norm": 4.486696720123291,
"learning_rate": 1.7727335734973512e-05,
"loss": 0.1792,
"step": 298
},
{
"epoch": 0.017208633093525178,
"grad_norm": 4.518349647521973,
"learning_rate": 1.7705132427757895e-05,
"loss": 0.2248,
"step": 299
},
{
"epoch": 0.017266187050359712,
"grad_norm": 11.271574020385742,
"learning_rate": 1.7682835235935236e-05,
"loss": 0.7558,
"step": 300
},
{
"epoch": 0.017323741007194245,
"grad_norm": 7.780160427093506,
"learning_rate": 1.766044443118978e-05,
"loss": 0.2102,
"step": 301
},
{
"epoch": 0.017381294964028775,
"grad_norm": 4.762843132019043,
"learning_rate": 1.7637960286346423e-05,
"loss": 0.2039,
"step": 302
},
{
"epoch": 0.01743884892086331,
"grad_norm": 2.5707688331604004,
"learning_rate": 1.761538307536737e-05,
"loss": 0.1405,
"step": 303
},
{
"epoch": 0.017496402877697843,
"grad_norm": 1.7889554500579834,
"learning_rate": 1.759271307334881e-05,
"loss": 0.1155,
"step": 304
},
{
"epoch": 0.017553956834532373,
"grad_norm": 4.67526912689209,
"learning_rate": 1.7569950556517566e-05,
"loss": 0.2787,
"step": 305
},
{
"epoch": 0.017611510791366906,
"grad_norm": 3.5515453815460205,
"learning_rate": 1.7547095802227723e-05,
"loss": 0.1925,
"step": 306
},
{
"epoch": 0.01766906474820144,
"grad_norm": 5.906973838806152,
"learning_rate": 1.7524149088957244e-05,
"loss": 0.251,
"step": 307
},
{
"epoch": 0.01772661870503597,
"grad_norm": 11.664872169494629,
"learning_rate": 1.7501110696304598e-05,
"loss": 0.9332,
"step": 308
},
{
"epoch": 0.017784172661870504,
"grad_norm": 6.628945827484131,
"learning_rate": 1.747798090498532e-05,
"loss": 0.294,
"step": 309
},
{
"epoch": 0.017841726618705037,
"grad_norm": 1.0758421421051025,
"learning_rate": 1.7454759996828622e-05,
"loss": 0.1241,
"step": 310
},
{
"epoch": 0.017899280575539567,
"grad_norm": 5.263257026672363,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.2367,
"step": 311
},
{
"epoch": 0.0179568345323741,
"grad_norm": 1.0564324855804443,
"learning_rate": 1.74080459628675e-05,
"loss": 0.1437,
"step": 312
},
{
"epoch": 0.018014388489208635,
"grad_norm": 7.245335578918457,
"learning_rate": 1.7384553406258842e-05,
"loss": 0.2376,
"step": 313
},
{
"epoch": 0.018071942446043165,
"grad_norm": 1.8886218070983887,
"learning_rate": 1.7360970871197347e-05,
"loss": 0.1031,
"step": 314
},
{
"epoch": 0.0181294964028777,
"grad_norm": 16.046236038208008,
"learning_rate": 1.7337298645028764e-05,
"loss": 0.6099,
"step": 315
},
{
"epoch": 0.01818705035971223,
"grad_norm": 3.0381016731262207,
"learning_rate": 1.7313537016191706e-05,
"loss": 0.13,
"step": 316
},
{
"epoch": 0.018244604316546762,
"grad_norm": 4.647341251373291,
"learning_rate": 1.7289686274214116e-05,
"loss": 0.1707,
"step": 317
},
{
"epoch": 0.018302158273381296,
"grad_norm": 7.225495338439941,
"learning_rate": 1.7265746709709762e-05,
"loss": 0.3956,
"step": 318
},
{
"epoch": 0.018359712230215826,
"grad_norm": 4.374240875244141,
"learning_rate": 1.7241718614374678e-05,
"loss": 0.2331,
"step": 319
},
{
"epoch": 0.01841726618705036,
"grad_norm": 13.612607955932617,
"learning_rate": 1.7217602280983622e-05,
"loss": 0.645,
"step": 320
},
{
"epoch": 0.018474820143884893,
"grad_norm": 4.012824535369873,
"learning_rate": 1.7193398003386514e-05,
"loss": 0.2226,
"step": 321
},
{
"epoch": 0.018532374100719423,
"grad_norm": 1.8649362325668335,
"learning_rate": 1.716910607650483e-05,
"loss": 0.1594,
"step": 322
},
{
"epoch": 0.018589928057553957,
"grad_norm": 5.170797824859619,
"learning_rate": 1.7144726796328034e-05,
"loss": 0.4557,
"step": 323
},
{
"epoch": 0.01864748201438849,
"grad_norm": 7.867353916168213,
"learning_rate": 1.712026045990997e-05,
"loss": 0.325,
"step": 324
},
{
"epoch": 0.01870503597122302,
"grad_norm": 4.790438175201416,
"learning_rate": 1.709570736536521e-05,
"loss": 0.2865,
"step": 325
},
{
"epoch": 0.018762589928057554,
"grad_norm": 2.507946252822876,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.104,
"step": 326
},
{
"epoch": 0.018820143884892088,
"grad_norm": 3.4884064197540283,
"learning_rate": 1.7046342099635948e-05,
"loss": 0.1544,
"step": 327
},
{
"epoch": 0.018877697841726618,
"grad_norm": 4.94096565246582,
"learning_rate": 1.7021530529951627e-05,
"loss": 0.2202,
"step": 328
},
{
"epoch": 0.01893525179856115,
"grad_norm": 2.591435432434082,
"learning_rate": 1.6996633405133656e-05,
"loss": 0.1226,
"step": 329
},
{
"epoch": 0.018992805755395685,
"grad_norm": 1.8974847793579102,
"learning_rate": 1.697165102854565e-05,
"loss": 0.1559,
"step": 330
},
{
"epoch": 0.019050359712230215,
"grad_norm": 1.4150128364562988,
"learning_rate": 1.6946583704589973e-05,
"loss": 0.1162,
"step": 331
},
{
"epoch": 0.01910791366906475,
"grad_norm": 3.744464874267578,
"learning_rate": 1.692143173870407e-05,
"loss": 0.2291,
"step": 332
},
{
"epoch": 0.01916546762589928,
"grad_norm": 0.9393559694290161,
"learning_rate": 1.68961954373567e-05,
"loss": 0.1093,
"step": 333
},
{
"epoch": 0.019223021582733812,
"grad_norm": 4.989112377166748,
"learning_rate": 1.6870875108044233e-05,
"loss": 0.2268,
"step": 334
},
{
"epoch": 0.019280575539568346,
"grad_norm": 3.0673670768737793,
"learning_rate": 1.684547105928689e-05,
"loss": 0.1281,
"step": 335
},
{
"epoch": 0.019338129496402876,
"grad_norm": 7.947831153869629,
"learning_rate": 1.6819983600624986e-05,
"loss": 0.665,
"step": 336
},
{
"epoch": 0.01939568345323741,
"grad_norm": 6.416712284088135,
"learning_rate": 1.6794413042615168e-05,
"loss": 0.2137,
"step": 337
},
{
"epoch": 0.019453237410071943,
"grad_norm": 16.095256805419922,
"learning_rate": 1.6768759696826608e-05,
"loss": 0.3756,
"step": 338
},
{
"epoch": 0.019510791366906474,
"grad_norm": 2.4824581146240234,
"learning_rate": 1.6743023875837233e-05,
"loss": 0.117,
"step": 339
},
{
"epoch": 0.019568345323741007,
"grad_norm": 5.838296890258789,
"learning_rate": 1.6717205893229904e-05,
"loss": 0.3567,
"step": 340
},
{
"epoch": 0.01962589928057554,
"grad_norm": 5.100681781768799,
"learning_rate": 1.6691306063588583e-05,
"loss": 0.2389,
"step": 341
},
{
"epoch": 0.01968345323741007,
"grad_norm": 7.527962684631348,
"learning_rate": 1.6665324702494524e-05,
"loss": 0.4136,
"step": 342
},
{
"epoch": 0.019741007194244604,
"grad_norm": 2.3168222904205322,
"learning_rate": 1.6639262126522417e-05,
"loss": 0.158,
"step": 343
},
{
"epoch": 0.019798561151079138,
"grad_norm": 0.8575424551963806,
"learning_rate": 1.661311865323652e-05,
"loss": 0.0539,
"step": 344
},
{
"epoch": 0.019856115107913668,
"grad_norm": 3.8268239498138428,
"learning_rate": 1.6586894601186804e-05,
"loss": 0.2135,
"step": 345
},
{
"epoch": 0.019913669064748202,
"grad_norm": 2.2744338512420654,
"learning_rate": 1.6560590289905074e-05,
"loss": 0.1278,
"step": 346
},
{
"epoch": 0.019971223021582735,
"grad_norm": 5.595208168029785,
"learning_rate": 1.6534206039901057e-05,
"loss": 0.2693,
"step": 347
},
{
"epoch": 0.020028776978417265,
"grad_norm": 2.456190347671509,
"learning_rate": 1.650774217265851e-05,
"loss": 0.1437,
"step": 348
},
{
"epoch": 0.0200863309352518,
"grad_norm": 10.795378684997559,
"learning_rate": 1.6481199010631312e-05,
"loss": 0.7296,
"step": 349
},
{
"epoch": 0.02014388489208633,
"grad_norm": 0.6766440868377686,
"learning_rate": 1.645457687723951e-05,
"loss": 0.104,
"step": 350
},
{
"epoch": 0.020201438848920863,
"grad_norm": 2.074960708618164,
"learning_rate": 1.6427876096865394e-05,
"loss": 0.1433,
"step": 351
},
{
"epoch": 0.020258992805755396,
"grad_norm": 3.726234197616577,
"learning_rate": 1.6401096994849558e-05,
"loss": 0.1568,
"step": 352
},
{
"epoch": 0.020316546762589927,
"grad_norm": 1.0608885288238525,
"learning_rate": 1.63742398974869e-05,
"loss": 0.095,
"step": 353
},
{
"epoch": 0.02037410071942446,
"grad_norm": 3.429298162460327,
"learning_rate": 1.6347305132022677e-05,
"loss": 0.1117,
"step": 354
},
{
"epoch": 0.020431654676258994,
"grad_norm": 5.276256561279297,
"learning_rate": 1.632029302664851e-05,
"loss": 0.4022,
"step": 355
},
{
"epoch": 0.020489208633093524,
"grad_norm": 6.028234481811523,
"learning_rate": 1.6293203910498375e-05,
"loss": 0.1726,
"step": 356
},
{
"epoch": 0.020546762589928057,
"grad_norm": 7.822464942932129,
"learning_rate": 1.6266038113644605e-05,
"loss": 0.3716,
"step": 357
},
{
"epoch": 0.02060431654676259,
"grad_norm": 9.866429328918457,
"learning_rate": 1.6238795967093865e-05,
"loss": 0.383,
"step": 358
},
{
"epoch": 0.02066187050359712,
"grad_norm": 2.492753505706787,
"learning_rate": 1.6211477802783105e-05,
"loss": 0.1362,
"step": 359
},
{
"epoch": 0.020719424460431655,
"grad_norm": 5.538491249084473,
"learning_rate": 1.6184083953575543e-05,
"loss": 0.2055,
"step": 360
},
{
"epoch": 0.02077697841726619,
"grad_norm": 1.3676029443740845,
"learning_rate": 1.6156614753256583e-05,
"loss": 0.0914,
"step": 361
},
{
"epoch": 0.02083453237410072,
"grad_norm": 5.788948059082031,
"learning_rate": 1.6129070536529767e-05,
"loss": 0.195,
"step": 362
},
{
"epoch": 0.020892086330935252,
"grad_norm": 1.366302251815796,
"learning_rate": 1.610145163901268e-05,
"loss": 0.112,
"step": 363
},
{
"epoch": 0.020949640287769786,
"grad_norm": 2.1532888412475586,
"learning_rate": 1.607375839723287e-05,
"loss": 0.1321,
"step": 364
},
{
"epoch": 0.021007194244604316,
"grad_norm": 2.4224016666412354,
"learning_rate": 1.6045991148623752e-05,
"loss": 0.1307,
"step": 365
},
{
"epoch": 0.02106474820143885,
"grad_norm": 2.301203489303589,
"learning_rate": 1.6018150231520486e-05,
"loss": 0.1129,
"step": 366
},
{
"epoch": 0.02112230215827338,
"grad_norm": 2.732968807220459,
"learning_rate": 1.599023598515586e-05,
"loss": 0.1113,
"step": 367
},
{
"epoch": 0.021179856115107913,
"grad_norm": 4.585814952850342,
"learning_rate": 1.5962248749656158e-05,
"loss": 0.2737,
"step": 368
},
{
"epoch": 0.021237410071942447,
"grad_norm": 3.1472392082214355,
"learning_rate": 1.5934188866037017e-05,
"loss": 0.1172,
"step": 369
},
{
"epoch": 0.021294964028776977,
"grad_norm": 0.760041356086731,
"learning_rate": 1.5906056676199256e-05,
"loss": 0.0888,
"step": 370
},
{
"epoch": 0.02135251798561151,
"grad_norm": 1.1722584962844849,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.0718,
"step": 371
},
{
"epoch": 0.021410071942446044,
"grad_norm": 3.714165449142456,
"learning_rate": 1.584957674987216e-05,
"loss": 0.2098,
"step": 372
},
{
"epoch": 0.021467625899280574,
"grad_norm": 13.37151050567627,
"learning_rate": 1.5821229701572897e-05,
"loss": 0.6504,
"step": 373
},
{
"epoch": 0.021525179856115108,
"grad_norm": 7.64274787902832,
"learning_rate": 1.5792811723426787e-05,
"loss": 0.1673,
"step": 374
},
{
"epoch": 0.02158273381294964,
"grad_norm": 5.333349227905273,
"learning_rate": 1.5764323161697933e-05,
"loss": 0.1838,
"step": 375
},
{
"epoch": 0.02164028776978417,
"grad_norm": 5.268608570098877,
"learning_rate": 1.573576436351046e-05,
"loss": 0.201,
"step": 376
},
{
"epoch": 0.021697841726618705,
"grad_norm": 1.2459615468978882,
"learning_rate": 1.570713567684432e-05,
"loss": 0.0831,
"step": 377
},
{
"epoch": 0.02175539568345324,
"grad_norm": 0.3400329053401947,
"learning_rate": 1.5678437450531014e-05,
"loss": 0.0705,
"step": 378
},
{
"epoch": 0.02181294964028777,
"grad_norm": 3.784626007080078,
"learning_rate": 1.564967003424938e-05,
"loss": 0.1281,
"step": 379
},
{
"epoch": 0.021870503597122302,
"grad_norm": 2.0622975826263428,
"learning_rate": 1.5620833778521306e-05,
"loss": 0.1074,
"step": 380
},
{
"epoch": 0.021928057553956836,
"grad_norm": 2.558227777481079,
"learning_rate": 1.5591929034707468e-05,
"loss": 0.1039,
"step": 381
},
{
"epoch": 0.021985611510791366,
"grad_norm": 1.8183974027633667,
"learning_rate": 1.556295615500305e-05,
"loss": 0.1491,
"step": 382
},
{
"epoch": 0.0220431654676259,
"grad_norm": 3.397756814956665,
"learning_rate": 1.553391549243344e-05,
"loss": 0.0963,
"step": 383
},
{
"epoch": 0.022100719424460433,
"grad_norm": 1.9542852640151978,
"learning_rate": 1.5504807400849957e-05,
"loss": 0.1328,
"step": 384
},
{
"epoch": 0.022158273381294964,
"grad_norm": 1.1182966232299805,
"learning_rate": 1.5475632234925505e-05,
"loss": 0.0329,
"step": 385
},
{
"epoch": 0.022215827338129497,
"grad_norm": 6.058446407318115,
"learning_rate": 1.5446390350150272e-05,
"loss": 0.2127,
"step": 386
},
{
"epoch": 0.022273381294964027,
"grad_norm": 4.640816688537598,
"learning_rate": 1.54170821028274e-05,
"loss": 0.1501,
"step": 387
},
{
"epoch": 0.02233093525179856,
"grad_norm": 12.117772102355957,
"learning_rate": 1.5387707850068633e-05,
"loss": 0.4365,
"step": 388
},
{
"epoch": 0.022388489208633094,
"grad_norm": 3.080798625946045,
"learning_rate": 1.5358267949789968e-05,
"loss": 0.1568,
"step": 389
},
{
"epoch": 0.022446043165467625,
"grad_norm": 9.55615520477295,
"learning_rate": 1.53287627607073e-05,
"loss": 0.2771,
"step": 390
},
{
"epoch": 0.022503597122302158,
"grad_norm": 0.7347004413604736,
"learning_rate": 1.529919264233205e-05,
"loss": 0.078,
"step": 391
},
{
"epoch": 0.022561151079136692,
"grad_norm": 2.543454647064209,
"learning_rate": 1.5269557954966777e-05,
"loss": 0.0838,
"step": 392
},
{
"epoch": 0.022618705035971222,
"grad_norm": 0.8127100467681885,
"learning_rate": 1.5239859059700794e-05,
"loss": 0.05,
"step": 393
},
{
"epoch": 0.022676258992805755,
"grad_norm": 15.900167465209961,
"learning_rate": 1.5210096318405768e-05,
"loss": 0.313,
"step": 394
},
{
"epoch": 0.02273381294964029,
"grad_norm": 3.528336524963379,
"learning_rate": 1.5180270093731305e-05,
"loss": 0.0843,
"step": 395
},
{
"epoch": 0.02279136690647482,
"grad_norm": 0.9315251708030701,
"learning_rate": 1.5150380749100545e-05,
"loss": 0.0912,
"step": 396
},
{
"epoch": 0.022848920863309353,
"grad_norm": 13.889352798461914,
"learning_rate": 1.5120428648705716e-05,
"loss": 0.3968,
"step": 397
},
{
"epoch": 0.022906474820143886,
"grad_norm": 4.333874225616455,
"learning_rate": 1.5090414157503715e-05,
"loss": 0.1176,
"step": 398
},
{
"epoch": 0.022964028776978417,
"grad_norm": 1.4791545867919922,
"learning_rate": 1.5060337641211637e-05,
"loss": 0.1132,
"step": 399
},
{
"epoch": 0.02302158273381295,
"grad_norm": 17.43948745727539,
"learning_rate": 1.5030199466302354e-05,
"loss": 0.3854,
"step": 400
},
{
"epoch": 0.023079136690647484,
"grad_norm": 1.4201478958129883,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.0633,
"step": 401
},
{
"epoch": 0.023136690647482014,
"grad_norm": 6.711745738983154,
"learning_rate": 1.4969739610275556e-05,
"loss": 0.1732,
"step": 402
},
{
"epoch": 0.023194244604316547,
"grad_norm": 1.3818823099136353,
"learning_rate": 1.493941866584231e-05,
"loss": 0.0769,
"step": 403
},
{
"epoch": 0.023251798561151078,
"grad_norm": 11.818349838256836,
"learning_rate": 1.490903753615141e-05,
"loss": 0.6309,
"step": 404
},
{
"epoch": 0.02330935251798561,
"grad_norm": 6.571465969085693,
"learning_rate": 1.4878596591387329e-05,
"loss": 0.1149,
"step": 405
},
{
"epoch": 0.023366906474820145,
"grad_norm": 8.349808692932129,
"learning_rate": 1.4848096202463373e-05,
"loss": 0.1896,
"step": 406
},
{
"epoch": 0.023424460431654675,
"grad_norm": 2.857787609100342,
"learning_rate": 1.4817536741017153e-05,
"loss": 0.0704,
"step": 407
},
{
"epoch": 0.02348201438848921,
"grad_norm": 8.498452186584473,
"learning_rate": 1.478691857940607e-05,
"loss": 0.2247,
"step": 408
},
{
"epoch": 0.023539568345323742,
"grad_norm": 0.468039333820343,
"learning_rate": 1.4756242090702756e-05,
"loss": 0.0615,
"step": 409
},
{
"epoch": 0.023597122302158272,
"grad_norm": 0.5978952646255493,
"learning_rate": 1.4725507648690542e-05,
"loss": 0.0501,
"step": 410
},
{
"epoch": 0.023654676258992806,
"grad_norm": 11.912870407104492,
"learning_rate": 1.469471562785891e-05,
"loss": 0.3164,
"step": 411
},
{
"epoch": 0.02371223021582734,
"grad_norm": 8.567428588867188,
"learning_rate": 1.4663866403398915e-05,
"loss": 0.2484,
"step": 412
},
{
"epoch": 0.02376978417266187,
"grad_norm": 5.721550941467285,
"learning_rate": 1.463296035119862e-05,
"loss": 0.1515,
"step": 413
},
{
"epoch": 0.023827338129496403,
"grad_norm": 11.112815856933594,
"learning_rate": 1.4601997847838518e-05,
"loss": 0.218,
"step": 414
},
{
"epoch": 0.023884892086330937,
"grad_norm": 0.5129345655441284,
"learning_rate": 1.4570979270586944e-05,
"loss": 0.0569,
"step": 415
},
{
"epoch": 0.023942446043165467,
"grad_norm": 7.7596259117126465,
"learning_rate": 1.4539904997395468e-05,
"loss": 0.1584,
"step": 416
},
{
"epoch": 0.024,
"grad_norm": 6.294882774353027,
"learning_rate": 1.4508775406894308e-05,
"loss": 0.0663,
"step": 417
},
{
"epoch": 0.024057553956834534,
"grad_norm": 10.160890579223633,
"learning_rate": 1.4477590878387697e-05,
"loss": 0.3199,
"step": 418
},
{
"epoch": 0.024115107913669064,
"grad_norm": 0.9614742398262024,
"learning_rate": 1.4446351791849276e-05,
"loss": 0.077,
"step": 419
},
{
"epoch": 0.024172661870503598,
"grad_norm": 3.362921953201294,
"learning_rate": 1.4415058527917454e-05,
"loss": 0.1266,
"step": 420
},
{
"epoch": 0.024230215827338128,
"grad_norm": 8.087349891662598,
"learning_rate": 1.4383711467890776e-05,
"loss": 0.3689,
"step": 421
},
{
"epoch": 0.02428776978417266,
"grad_norm": 8.182154655456543,
"learning_rate": 1.4352310993723277e-05,
"loss": 0.1767,
"step": 422
},
{
"epoch": 0.024345323741007195,
"grad_norm": 1.1635339260101318,
"learning_rate": 1.4320857488019826e-05,
"loss": 0.1257,
"step": 423
},
{
"epoch": 0.024402877697841725,
"grad_norm": 3.1351680755615234,
"learning_rate": 1.4289351334031461e-05,
"loss": 0.1368,
"step": 424
},
{
"epoch": 0.02446043165467626,
"grad_norm": 8.145259857177734,
"learning_rate": 1.4257792915650728e-05,
"loss": 0.2471,
"step": 425
},
{
"epoch": 0.024517985611510792,
"grad_norm": 0.6130021810531616,
"learning_rate": 1.4226182617406996e-05,
"loss": 0.0769,
"step": 426
},
{
"epoch": 0.024575539568345323,
"grad_norm": 30.53700065612793,
"learning_rate": 1.4194520824461773e-05,
"loss": 0.2786,
"step": 427
},
{
"epoch": 0.024633093525179856,
"grad_norm": 10.179473876953125,
"learning_rate": 1.4162807922604014e-05,
"loss": 0.22,
"step": 428
},
{
"epoch": 0.02469064748201439,
"grad_norm": 20.345251083374023,
"learning_rate": 1.413104429824542e-05,
"loss": 1.2058,
"step": 429
},
{
"epoch": 0.02474820143884892,
"grad_norm": 17.933361053466797,
"learning_rate": 1.4099230338415728e-05,
"loss": 0.5017,
"step": 430
},
{
"epoch": 0.024805755395683454,
"grad_norm": 16.122068405151367,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.4488,
"step": 431
},
{
"epoch": 0.024863309352517987,
"grad_norm": 3.910707712173462,
"learning_rate": 1.4035452963523903e-05,
"loss": 0.1475,
"step": 432
},
{
"epoch": 0.024920863309352517,
"grad_norm": 3.197533130645752,
"learning_rate": 1.4003490325568953e-05,
"loss": 0.1384,
"step": 433
},
{
"epoch": 0.02497841726618705,
"grad_norm": 1.7127132415771484,
"learning_rate": 1.3971478906347806e-05,
"loss": 0.0953,
"step": 434
},
{
"epoch": 0.025035971223021584,
"grad_norm": 0.6795534491539001,
"learning_rate": 1.3939419095909513e-05,
"loss": 0.084,
"step": 435
},
{
"epoch": 0.025093525179856115,
"grad_norm": 0.7117164731025696,
"learning_rate": 1.3907311284892737e-05,
"loss": 0.1028,
"step": 436
},
{
"epoch": 0.025151079136690648,
"grad_norm": 5.165547847747803,
"learning_rate": 1.3875155864521031e-05,
"loss": 0.1473,
"step": 437
},
{
"epoch": 0.02520863309352518,
"grad_norm": 1.277207374572754,
"learning_rate": 1.3842953226598036e-05,
"loss": 0.1054,
"step": 438
},
{
"epoch": 0.025266187050359712,
"grad_norm": 6.697636604309082,
"learning_rate": 1.3810703763502744e-05,
"loss": 0.2188,
"step": 439
},
{
"epoch": 0.025323741007194246,
"grad_norm": 0.8036256432533264,
"learning_rate": 1.3778407868184674e-05,
"loss": 0.0717,
"step": 440
},
{
"epoch": 0.025381294964028776,
"grad_norm": 14.702827453613281,
"learning_rate": 1.3746065934159123e-05,
"loss": 0.5481,
"step": 441
},
{
"epoch": 0.02543884892086331,
"grad_norm": 0.643853485584259,
"learning_rate": 1.371367835550235e-05,
"loss": 0.0649,
"step": 442
},
{
"epoch": 0.025496402877697843,
"grad_norm": 6.889122009277344,
"learning_rate": 1.3681245526846782e-05,
"loss": 0.1057,
"step": 443
},
{
"epoch": 0.025553956834532373,
"grad_norm": 3.3967247009277344,
"learning_rate": 1.3648767843376196e-05,
"loss": 0.1841,
"step": 444
},
{
"epoch": 0.025611510791366907,
"grad_norm": 4.166593074798584,
"learning_rate": 1.3616245700820922e-05,
"loss": 0.1322,
"step": 445
},
{
"epoch": 0.02566906474820144,
"grad_norm": 0.9764627814292908,
"learning_rate": 1.3583679495453e-05,
"loss": 0.0708,
"step": 446
},
{
"epoch": 0.02572661870503597,
"grad_norm": 8.78773307800293,
"learning_rate": 1.3551069624081372e-05,
"loss": 0.2444,
"step": 447
},
{
"epoch": 0.025784172661870504,
"grad_norm": 0.8957186341285706,
"learning_rate": 1.3518416484047018e-05,
"loss": 0.0952,
"step": 448
},
{
"epoch": 0.025841726618705037,
"grad_norm": 2.754924774169922,
"learning_rate": 1.3485720473218153e-05,
"loss": 0.1718,
"step": 449
},
{
"epoch": 0.025899280575539568,
"grad_norm": 9.26056957244873,
"learning_rate": 1.3452981989985347e-05,
"loss": 0.1871,
"step": 450
},
{
"epoch": 0.0259568345323741,
"grad_norm": 2.4848737716674805,
"learning_rate": 1.342020143325669e-05,
"loss": 0.0396,
"step": 451
},
{
"epoch": 0.026014388489208635,
"grad_norm": 3.2999818325042725,
"learning_rate": 1.3387379202452917e-05,
"loss": 0.1324,
"step": 452
},
{
"epoch": 0.026071942446043165,
"grad_norm": 13.430712699890137,
"learning_rate": 1.3354515697502552e-05,
"loss": 0.4404,
"step": 453
},
{
"epoch": 0.0261294964028777,
"grad_norm": 0.6919059753417969,
"learning_rate": 1.3321611318837033e-05,
"loss": 0.0747,
"step": 454
},
{
"epoch": 0.02618705035971223,
"grad_norm": 1.3426858186721802,
"learning_rate": 1.3288666467385834e-05,
"loss": 0.1142,
"step": 455
},
{
"epoch": 0.026244604316546762,
"grad_norm": 0.5844228863716125,
"learning_rate": 1.3255681544571568e-05,
"loss": 0.0768,
"step": 456
},
{
"epoch": 0.026302158273381296,
"grad_norm": 1.4332048892974854,
"learning_rate": 1.3222656952305113e-05,
"loss": 0.0896,
"step": 457
},
{
"epoch": 0.026359712230215826,
"grad_norm": 5.514071464538574,
"learning_rate": 1.3189593092980701e-05,
"loss": 0.2002,
"step": 458
},
{
"epoch": 0.02641726618705036,
"grad_norm": 10.462922096252441,
"learning_rate": 1.3156490369471026e-05,
"loss": 0.3964,
"step": 459
},
{
"epoch": 0.026474820143884893,
"grad_norm": 7.807409763336182,
"learning_rate": 1.3123349185122328e-05,
"loss": 0.2995,
"step": 460
},
{
"epoch": 0.026532374100719423,
"grad_norm": 2.9369232654571533,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.1574,
"step": 461
},
{
"epoch": 0.026589928057553957,
"grad_norm": 2.5351967811584473,
"learning_rate": 1.3056953049631059e-05,
"loss": 0.0965,
"step": 462
},
{
"epoch": 0.02664748201438849,
"grad_norm": 1.906512975692749,
"learning_rate": 1.3023698907504447e-05,
"loss": 0.1145,
"step": 463
},
{
"epoch": 0.02670503597122302,
"grad_norm": 9.540634155273438,
"learning_rate": 1.2990407922560869e-05,
"loss": 0.3393,
"step": 464
},
{
"epoch": 0.026762589928057554,
"grad_norm": 13.937700271606445,
"learning_rate": 1.2957080500440469e-05,
"loss": 0.4293,
"step": 465
},
{
"epoch": 0.026820143884892088,
"grad_norm": 2.0209853649139404,
"learning_rate": 1.2923717047227368e-05,
"loss": 0.049,
"step": 466
},
{
"epoch": 0.026877697841726618,
"grad_norm": 8.90571117401123,
"learning_rate": 1.2890317969444716e-05,
"loss": 0.3679,
"step": 467
},
{
"epoch": 0.02693525179856115,
"grad_norm": 1.0252469778060913,
"learning_rate": 1.2856883674049736e-05,
"loss": 0.0684,
"step": 468
},
{
"epoch": 0.026992805755395685,
"grad_norm": 4.494703769683838,
"learning_rate": 1.2823414568428767e-05,
"loss": 0.1366,
"step": 469
},
{
"epoch": 0.027050359712230215,
"grad_norm": 4.636646747589111,
"learning_rate": 1.2789911060392295e-05,
"loss": 0.1983,
"step": 470
},
{
"epoch": 0.02710791366906475,
"grad_norm": 1.6157455444335938,
"learning_rate": 1.2756373558169992e-05,
"loss": 0.1423,
"step": 471
},
{
"epoch": 0.02716546762589928,
"grad_norm": 1.4866031408309937,
"learning_rate": 1.2722802470405744e-05,
"loss": 0.1039,
"step": 472
},
{
"epoch": 0.027223021582733813,
"grad_norm": 0.6129380464553833,
"learning_rate": 1.2689198206152657e-05,
"loss": 0.0738,
"step": 473
},
{
"epoch": 0.027280575539568346,
"grad_norm": 8.018013954162598,
"learning_rate": 1.265556117486809e-05,
"loss": 0.1237,
"step": 474
},
{
"epoch": 0.027338129496402876,
"grad_norm": 8.371533393859863,
"learning_rate": 1.2621891786408648e-05,
"loss": 0.2127,
"step": 475
},
{
"epoch": 0.02739568345323741,
"grad_norm": 9.338693618774414,
"learning_rate": 1.2588190451025209e-05,
"loss": 0.313,
"step": 476
},
{
"epoch": 0.027453237410071944,
"grad_norm": 2.514599323272705,
"learning_rate": 1.2554457579357906e-05,
"loss": 0.1236,
"step": 477
},
{
"epoch": 0.027510791366906474,
"grad_norm": 36.57285690307617,
"learning_rate": 1.252069358243114e-05,
"loss": 0.8909,
"step": 478
},
{
"epoch": 0.027568345323741007,
"grad_norm": 8.022867202758789,
"learning_rate": 1.2486898871648552e-05,
"loss": 0.243,
"step": 479
},
{
"epoch": 0.02762589928057554,
"grad_norm": 3.656367063522339,
"learning_rate": 1.2453073858788027e-05,
"loss": 0.1721,
"step": 480
},
{
"epoch": 0.02768345323741007,
"grad_norm": 7.827700614929199,
"learning_rate": 1.2419218955996677e-05,
"loss": 0.1393,
"step": 481
},
{
"epoch": 0.027741007194244605,
"grad_norm": 0.9252501130104065,
"learning_rate": 1.238533457578581e-05,
"loss": 0.091,
"step": 482
},
{
"epoch": 0.027798561151079138,
"grad_norm": 1.7856941223144531,
"learning_rate": 1.23514211310259e-05,
"loss": 0.1275,
"step": 483
},
{
"epoch": 0.02785611510791367,
"grad_norm": 23.890789031982422,
"learning_rate": 1.2317479034941572e-05,
"loss": 0.3568,
"step": 484
},
{
"epoch": 0.027913669064748202,
"grad_norm": 2.6963653564453125,
"learning_rate": 1.2283508701106559e-05,
"loss": 0.1143,
"step": 485
},
{
"epoch": 0.027971223021582736,
"grad_norm": 5.141092777252197,
"learning_rate": 1.2249510543438652e-05,
"loss": 0.2145,
"step": 486
},
{
"epoch": 0.028028776978417266,
"grad_norm": 4.484059810638428,
"learning_rate": 1.2215484976194675e-05,
"loss": 0.134,
"step": 487
},
{
"epoch": 0.0280863309352518,
"grad_norm": 13.956079483032227,
"learning_rate": 1.2181432413965428e-05,
"loss": 0.2296,
"step": 488
},
{
"epoch": 0.02814388489208633,
"grad_norm": 57.16883087158203,
"learning_rate": 1.2147353271670634e-05,
"loss": 0.3778,
"step": 489
},
{
"epoch": 0.028201438848920863,
"grad_norm": 0.9326462745666504,
"learning_rate": 1.211324796455389e-05,
"loss": 0.0922,
"step": 490
},
{
"epoch": 0.028258992805755397,
"grad_norm": 0.6182975172996521,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.0514,
"step": 491
},
{
"epoch": 0.028316546762589927,
"grad_norm": 0.42364293336868286,
"learning_rate": 1.2044960518417902e-05,
"loss": 0.0181,
"step": 492
},
{
"epoch": 0.02837410071942446,
"grad_norm": 3.4781551361083984,
"learning_rate": 1.2010779211459649e-05,
"loss": 0.161,
"step": 493
},
{
"epoch": 0.028431654676258994,
"grad_norm": 2.329385280609131,
"learning_rate": 1.1976573403791263e-05,
"loss": 0.0878,
"step": 494
},
{
"epoch": 0.028489208633093524,
"grad_norm": 5.409578323364258,
"learning_rate": 1.194234351219972e-05,
"loss": 0.1646,
"step": 495
},
{
"epoch": 0.028546762589928058,
"grad_norm": 0.2559332847595215,
"learning_rate": 1.190808995376545e-05,
"loss": 0.0438,
"step": 496
},
{
"epoch": 0.02860431654676259,
"grad_norm": 7.3071088790893555,
"learning_rate": 1.187381314585725e-05,
"loss": 0.2373,
"step": 497
},
{
"epoch": 0.02866187050359712,
"grad_norm": 13.68780517578125,
"learning_rate": 1.1839513506127202e-05,
"loss": 0.5267,
"step": 498
},
{
"epoch": 0.028719424460431655,
"grad_norm": 14.798240661621094,
"learning_rate": 1.1805191452505602e-05,
"loss": 0.4246,
"step": 499
},
{
"epoch": 0.02877697841726619,
"grad_norm": 0.8802618980407715,
"learning_rate": 1.1770847403195836e-05,
"loss": 0.0126,
"step": 500
},
{
"epoch": 0.02883453237410072,
"grad_norm": 2.4958908557891846,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.0668,
"step": 501
},
{
"epoch": 0.028892086330935252,
"grad_norm": 8.02234172821045,
"learning_rate": 1.1702094991660326e-05,
"loss": 0.2478,
"step": 502
},
{
"epoch": 0.028949640287769786,
"grad_norm": 6.714267730712891,
"learning_rate": 1.1667687467161025e-05,
"loss": 0.2143,
"step": 503
},
{
"epoch": 0.029007194244604316,
"grad_norm": 7.356170177459717,
"learning_rate": 1.1633259622416224e-05,
"loss": 0.2208,
"step": 504
},
{
"epoch": 0.02906474820143885,
"grad_norm": 4.378060340881348,
"learning_rate": 1.159881187691835e-05,
"loss": 0.1349,
"step": 505
},
{
"epoch": 0.02912230215827338,
"grad_norm": 9.72322940826416,
"learning_rate": 1.156434465040231e-05,
"loss": 0.277,
"step": 506
},
{
"epoch": 0.029179856115107913,
"grad_norm": 1.8536237478256226,
"learning_rate": 1.1529858362840383e-05,
"loss": 0.1448,
"step": 507
},
{
"epoch": 0.029237410071942447,
"grad_norm": 3.8134891986846924,
"learning_rate": 1.1495353434437098e-05,
"loss": 0.1221,
"step": 508
},
{
"epoch": 0.029294964028776977,
"grad_norm": 8.059671401977539,
"learning_rate": 1.1460830285624119e-05,
"loss": 0.3355,
"step": 509
},
{
"epoch": 0.02935251798561151,
"grad_norm": 10.443696022033691,
"learning_rate": 1.1426289337055119e-05,
"loss": 0.329,
"step": 510
},
{
"epoch": 0.029410071942446044,
"grad_norm": 1.0404391288757324,
"learning_rate": 1.1391731009600655e-05,
"loss": 0.0193,
"step": 511
},
{
"epoch": 0.029467625899280574,
"grad_norm": 9.676289558410645,
"learning_rate": 1.1357155724343046e-05,
"loss": 0.2689,
"step": 512
},
{
"epoch": 0.029525179856115108,
"grad_norm": 25.724231719970703,
"learning_rate": 1.1322563902571227e-05,
"loss": 0.7274,
"step": 513
},
{
"epoch": 0.02958273381294964,
"grad_norm": 3.297560214996338,
"learning_rate": 1.128795596577563e-05,
"loss": 0.1354,
"step": 514
},
{
"epoch": 0.02964028776978417,
"grad_norm": 7.687169551849365,
"learning_rate": 1.1253332335643043e-05,
"loss": 0.8631,
"step": 515
},
{
"epoch": 0.029697841726618705,
"grad_norm": 1.4026639461517334,
"learning_rate": 1.1218693434051475e-05,
"loss": 0.0832,
"step": 516
},
{
"epoch": 0.02975539568345324,
"grad_norm": 1.158055067062378,
"learning_rate": 1.1184039683065014e-05,
"loss": 0.1053,
"step": 517
},
{
"epoch": 0.02981294964028777,
"grad_norm": 1.325383186340332,
"learning_rate": 1.1149371504928667e-05,
"loss": 0.0601,
"step": 518
},
{
"epoch": 0.029870503597122303,
"grad_norm": 21.55183219909668,
"learning_rate": 1.1114689322063255e-05,
"loss": 0.4671,
"step": 519
},
{
"epoch": 0.029928057553956836,
"grad_norm": 5.480937480926514,
"learning_rate": 1.1079993557060228e-05,
"loss": 0.19,
"step": 520
},
{
"epoch": 0.029985611510791366,
"grad_norm": 7.440226078033447,
"learning_rate": 1.1045284632676535e-05,
"loss": 0.2281,
"step": 521
},
{
"epoch": 0.0300431654676259,
"grad_norm": 2.201996326446533,
"learning_rate": 1.1010562971829464e-05,
"loss": 0.1137,
"step": 522
},
{
"epoch": 0.03010071942446043,
"grad_norm": 2.157567024230957,
"learning_rate": 1.0975828997591496e-05,
"loss": 0.1234,
"step": 523
},
{
"epoch": 0.030158273381294964,
"grad_norm": 1.4759501218795776,
"learning_rate": 1.0941083133185146e-05,
"loss": 0.0717,
"step": 524
},
{
"epoch": 0.030215827338129497,
"grad_norm": 4.390008449554443,
"learning_rate": 1.0906325801977804e-05,
"loss": 0.1566,
"step": 525
},
{
"epoch": 0.030273381294964027,
"grad_norm": 2.1558635234832764,
"learning_rate": 1.0871557427476585e-05,
"loss": 0.1084,
"step": 526
},
{
"epoch": 0.03033093525179856,
"grad_norm": 5.547511577606201,
"learning_rate": 1.083677843332316e-05,
"loss": 0.2485,
"step": 527
},
{
"epoch": 0.030388489208633095,
"grad_norm": 1.3841583728790283,
"learning_rate": 1.0801989243288588e-05,
"loss": 0.0932,
"step": 528
},
{
"epoch": 0.030446043165467625,
"grad_norm": 3.9912872314453125,
"learning_rate": 1.0767190281268187e-05,
"loss": 0.1574,
"step": 529
},
{
"epoch": 0.03050359712230216,
"grad_norm": 0.8950588703155518,
"learning_rate": 1.0732381971276318e-05,
"loss": 0.0841,
"step": 530
},
{
"epoch": 0.030561151079136692,
"grad_norm": 1.7501403093338013,
"learning_rate": 1.0697564737441254e-05,
"loss": 0.1257,
"step": 531
},
{
"epoch": 0.030618705035971222,
"grad_norm": 3.171156406402588,
"learning_rate": 1.0662739004000005e-05,
"loss": 0.1247,
"step": 532
},
{
"epoch": 0.030676258992805756,
"grad_norm": 8.79008960723877,
"learning_rate": 1.0627905195293135e-05,
"loss": 0.4406,
"step": 533
},
{
"epoch": 0.03073381294964029,
"grad_norm": 0.8652887344360352,
"learning_rate": 1.0593063735759619e-05,
"loss": 0.0557,
"step": 534
},
{
"epoch": 0.03079136690647482,
"grad_norm": 2.49337100982666,
"learning_rate": 1.055821504993164e-05,
"loss": 0.0957,
"step": 535
},
{
"epoch": 0.030848920863309353,
"grad_norm": 0.9073213934898376,
"learning_rate": 1.0523359562429441e-05,
"loss": 0.081,
"step": 536
},
{
"epoch": 0.030906474820143887,
"grad_norm": 2.930410385131836,
"learning_rate": 1.0488497697956134e-05,
"loss": 0.1306,
"step": 537
},
{
"epoch": 0.030964028776978417,
"grad_norm": 15.254547119140625,
"learning_rate": 1.0453629881292537e-05,
"loss": 0.5873,
"step": 538
},
{
"epoch": 0.03102158273381295,
"grad_norm": 9.663077354431152,
"learning_rate": 1.0418756537291996e-05,
"loss": 0.255,
"step": 539
},
{
"epoch": 0.03107913669064748,
"grad_norm": 0.9969314932823181,
"learning_rate": 1.03838780908752e-05,
"loss": 0.0821,
"step": 540
},
{
"epoch": 0.031136690647482014,
"grad_norm": 8.922150611877441,
"learning_rate": 1.0348994967025012e-05,
"loss": 0.2661,
"step": 541
},
{
"epoch": 0.031194244604316548,
"grad_norm": 1.0743968486785889,
"learning_rate": 1.0314107590781284e-05,
"loss": 0.1089,
"step": 542
},
{
"epoch": 0.03125179856115108,
"grad_norm": 4.709200382232666,
"learning_rate": 1.0279216387235691e-05,
"loss": 0.158,
"step": 543
},
{
"epoch": 0.03130935251798561,
"grad_norm": 6.3187947273254395,
"learning_rate": 1.0244321781526533e-05,
"loss": 0.2659,
"step": 544
},
{
"epoch": 0.03136690647482014,
"grad_norm": 0.4948660433292389,
"learning_rate": 1.0209424198833571e-05,
"loss": 0.0685,
"step": 545
},
{
"epoch": 0.03142446043165468,
"grad_norm": 1.0766798257827759,
"learning_rate": 1.0174524064372837e-05,
"loss": 0.0989,
"step": 546
},
{
"epoch": 0.03148201438848921,
"grad_norm": 4.577028751373291,
"learning_rate": 1.0139621803391454e-05,
"loss": 0.177,
"step": 547
},
{
"epoch": 0.03153956834532374,
"grad_norm": 0.9121817946434021,
"learning_rate": 1.010471784116246e-05,
"loss": 0.086,
"step": 548
},
{
"epoch": 0.031597122302158276,
"grad_norm": 8.214136123657227,
"learning_rate": 1.0069812602979617e-05,
"loss": 0.2782,
"step": 549
},
{
"epoch": 0.031654676258992806,
"grad_norm": 0.6466065049171448,
"learning_rate": 1.0034906514152239e-05,
"loss": 0.0647,
"step": 550
},
{
"epoch": 0.031712230215827336,
"grad_norm": 2.8535683155059814,
"learning_rate": 1e-05,
"loss": 0.1092,
"step": 551
},
{
"epoch": 0.03176978417266187,
"grad_norm": 0.7769191861152649,
"learning_rate": 9.965093485847766e-06,
"loss": 0.0612,
"step": 552
},
{
"epoch": 0.0318273381294964,
"grad_norm": 1.8424514532089233,
"learning_rate": 9.930187397020385e-06,
"loss": 0.1221,
"step": 553
},
{
"epoch": 0.031884892086330933,
"grad_norm": 1.2432526350021362,
"learning_rate": 9.895282158837545e-06,
"loss": 0.0838,
"step": 554
},
{
"epoch": 0.03194244604316547,
"grad_norm": 4.903514862060547,
"learning_rate": 9.860378196608549e-06,
"loss": 0.1696,
"step": 555
},
{
"epoch": 0.032,
"grad_norm": 0.8313055038452148,
"learning_rate": 9.825475935627165e-06,
"loss": 0.0785,
"step": 556
},
{
"epoch": 0.03205755395683453,
"grad_norm": 0.8260055780410767,
"learning_rate": 9.790575801166432e-06,
"loss": 0.0783,
"step": 557
},
{
"epoch": 0.03211510791366907,
"grad_norm": 15.278909683227539,
"learning_rate": 9.75567821847347e-06,
"loss": 0.4247,
"step": 558
},
{
"epoch": 0.0321726618705036,
"grad_norm": 1.1578339338302612,
"learning_rate": 9.720783612764314e-06,
"loss": 0.0904,
"step": 559
},
{
"epoch": 0.03223021582733813,
"grad_norm": 2.438309669494629,
"learning_rate": 9.685892409218718e-06,
"loss": 0.1525,
"step": 560
},
{
"epoch": 0.032287769784172665,
"grad_norm": 7.4440693855285645,
"learning_rate": 9.651005032974994e-06,
"loss": 0.2666,
"step": 561
},
{
"epoch": 0.032345323741007195,
"grad_norm": 1.554640293121338,
"learning_rate": 9.616121909124801e-06,
"loss": 0.0693,
"step": 562
},
{
"epoch": 0.032402877697841725,
"grad_norm": 3.4629485607147217,
"learning_rate": 9.581243462708007e-06,
"loss": 0.1241,
"step": 563
},
{
"epoch": 0.032460431654676256,
"grad_norm": 1.3280606269836426,
"learning_rate": 9.546370118707463e-06,
"loss": 0.0894,
"step": 564
},
{
"epoch": 0.03251798561151079,
"grad_norm": 3.824627637863159,
"learning_rate": 9.511502302043867e-06,
"loss": 0.1586,
"step": 565
},
{
"epoch": 0.03257553956834532,
"grad_norm": 9.243664741516113,
"learning_rate": 9.476640437570562e-06,
"loss": 0.258,
"step": 566
},
{
"epoch": 0.03263309352517985,
"grad_norm": 3.4505608081817627,
"learning_rate": 9.441784950068362e-06,
"loss": 0.134,
"step": 567
},
{
"epoch": 0.03269064748201439,
"grad_norm": 13.31429672241211,
"learning_rate": 9.406936264240386e-06,
"loss": 0.4339,
"step": 568
},
{
"epoch": 0.03274820143884892,
"grad_norm": 0.4915928542613983,
"learning_rate": 9.372094804706867e-06,
"loss": 0.0526,
"step": 569
},
{
"epoch": 0.03280575539568345,
"grad_norm": 9.329924583435059,
"learning_rate": 9.337260996000002e-06,
"loss": 0.1747,
"step": 570
},
{
"epoch": 0.03286330935251799,
"grad_norm": 1.121419906616211,
"learning_rate": 9.302435262558748e-06,
"loss": 0.0665,
"step": 571
},
{
"epoch": 0.03292086330935252,
"grad_norm": 1.8707423210144043,
"learning_rate": 9.267618028723687e-06,
"loss": 0.1144,
"step": 572
},
{
"epoch": 0.03297841726618705,
"grad_norm": 0.3859129548072815,
"learning_rate": 9.232809718731815e-06,
"loss": 0.0301,
"step": 573
},
{
"epoch": 0.033035971223021585,
"grad_norm": 0.4926592707633972,
"learning_rate": 9.198010756711413e-06,
"loss": 0.0795,
"step": 574
},
{
"epoch": 0.033093525179856115,
"grad_norm": 7.28505802154541,
"learning_rate": 9.163221566676847e-06,
"loss": 0.1051,
"step": 575
},
{
"epoch": 0.033151079136690645,
"grad_norm": 10.257981300354004,
"learning_rate": 9.128442572523418e-06,
"loss": 0.4004,
"step": 576
},
{
"epoch": 0.03320863309352518,
"grad_norm": 5.706249713897705,
"learning_rate": 9.093674198022201e-06,
"loss": 0.2009,
"step": 577
},
{
"epoch": 0.03326618705035971,
"grad_norm": 3.457526922225952,
"learning_rate": 9.058916866814857e-06,
"loss": 0.1433,
"step": 578
},
{
"epoch": 0.03332374100719424,
"grad_norm": 6.841573715209961,
"learning_rate": 9.024171002408507e-06,
"loss": 0.5298,
"step": 579
},
{
"epoch": 0.03338129496402878,
"grad_norm": 1.4832878112792969,
"learning_rate": 8.989437028170537e-06,
"loss": 0.0853,
"step": 580
},
{
"epoch": 0.03343884892086331,
"grad_norm": 11.173535346984863,
"learning_rate": 8.954715367323468e-06,
"loss": 0.2538,
"step": 581
},
{
"epoch": 0.03349640287769784,
"grad_norm": 0.7396141290664673,
"learning_rate": 8.920006442939772e-06,
"loss": 0.025,
"step": 582
},
{
"epoch": 0.03355395683453238,
"grad_norm": 4.9781293869018555,
"learning_rate": 8.885310677936746e-06,
"loss": 0.1923,
"step": 583
},
{
"epoch": 0.03361151079136691,
"grad_norm": 0.557285726070404,
"learning_rate": 8.850628495071336e-06,
"loss": 0.0087,
"step": 584
},
{
"epoch": 0.03366906474820144,
"grad_norm": 11.92111873626709,
"learning_rate": 8.815960316934991e-06,
"loss": 0.2586,
"step": 585
},
{
"epoch": 0.033726618705035974,
"grad_norm": 1.8631489276885986,
"learning_rate": 8.781306565948528e-06,
"loss": 0.1197,
"step": 586
},
{
"epoch": 0.033784172661870504,
"grad_norm": 1.890995740890503,
"learning_rate": 8.746667664356957e-06,
"loss": 0.1011,
"step": 587
},
{
"epoch": 0.033841726618705034,
"grad_norm": 0.8469424247741699,
"learning_rate": 8.712044034224374e-06,
"loss": 0.0862,
"step": 588
},
{
"epoch": 0.03389928057553957,
"grad_norm": 0.5587536096572876,
"learning_rate": 8.677436097428775e-06,
"loss": 0.0414,
"step": 589
},
{
"epoch": 0.0339568345323741,
"grad_norm": 14.286009788513184,
"learning_rate": 8.642844275656957e-06,
"loss": 0.2033,
"step": 590
},
{
"epoch": 0.03401438848920863,
"grad_norm": 0.828925371170044,
"learning_rate": 8.60826899039935e-06,
"loss": 0.0879,
"step": 591
},
{
"epoch": 0.03407194244604317,
"grad_norm": 2.5899553298950195,
"learning_rate": 8.573710662944884e-06,
"loss": 0.1128,
"step": 592
},
{
"epoch": 0.0341294964028777,
"grad_norm": 0.7016512751579285,
"learning_rate": 8.539169714375885e-06,
"loss": 0.0559,
"step": 593
},
{
"epoch": 0.03418705035971223,
"grad_norm": 0.7938787937164307,
"learning_rate": 8.504646565562907e-06,
"loss": 0.0751,
"step": 594
},
{
"epoch": 0.034244604316546766,
"grad_norm": 4.789397716522217,
"learning_rate": 8.47014163715962e-06,
"loss": 0.1729,
"step": 595
},
{
"epoch": 0.034302158273381296,
"grad_norm": 8.780428886413574,
"learning_rate": 8.43565534959769e-06,
"loss": 0.2045,
"step": 596
},
{
"epoch": 0.034359712230215826,
"grad_norm": 1.7636146545410156,
"learning_rate": 8.401188123081653e-06,
"loss": 0.0915,
"step": 597
},
{
"epoch": 0.034417266187050356,
"grad_norm": 4.673388957977295,
"learning_rate": 8.366740377583781e-06,
"loss": 0.1396,
"step": 598
},
{
"epoch": 0.03447482014388489,
"grad_norm": 10.594704627990723,
"learning_rate": 8.332312532838978e-06,
"loss": 0.2385,
"step": 599
},
{
"epoch": 0.034532374100719423,
"grad_norm": 0.38678157329559326,
"learning_rate": 8.297905008339677e-06,
"loss": 0.0511,
"step": 600
},
{
"epoch": 0.034589928057553954,
"grad_norm": 2.4528400897979736,
"learning_rate": 8.263518223330698e-06,
"loss": 0.1191,
"step": 601
},
{
"epoch": 0.03464748201438849,
"grad_norm": 9.87134075164795,
"learning_rate": 8.22915259680417e-06,
"loss": 0.3747,
"step": 602
},
{
"epoch": 0.03470503597122302,
"grad_norm": 2.0023324489593506,
"learning_rate": 8.194808547494401e-06,
"loss": 0.0745,
"step": 603
},
{
"epoch": 0.03476258992805755,
"grad_norm": 14.806257247924805,
"learning_rate": 8.1604864938728e-06,
"loss": 0.5362,
"step": 604
},
{
"epoch": 0.03482014388489209,
"grad_norm": 4.73097562789917,
"learning_rate": 8.126186854142752e-06,
"loss": 0.1718,
"step": 605
},
{
"epoch": 0.03487769784172662,
"grad_norm": 1.0992658138275146,
"learning_rate": 8.091910046234552e-06,
"loss": 0.0868,
"step": 606
},
{
"epoch": 0.03493525179856115,
"grad_norm": 0.5213341116905212,
"learning_rate": 8.057656487800283e-06,
"loss": 0.0579,
"step": 607
},
{
"epoch": 0.034992805755395685,
"grad_norm": 0.8178665637969971,
"learning_rate": 8.023426596208739e-06,
"loss": 0.0724,
"step": 608
},
{
"epoch": 0.035050359712230215,
"grad_norm": 2.403118133544922,
"learning_rate": 7.989220788540356e-06,
"loss": 0.1376,
"step": 609
},
{
"epoch": 0.035107913669064746,
"grad_norm": 6.1761579513549805,
"learning_rate": 7.955039481582098e-06,
"loss": 0.2501,
"step": 610
},
{
"epoch": 0.03516546762589928,
"grad_norm": 1.1938904523849487,
"learning_rate": 7.92088309182241e-06,
"loss": 0.1253,
"step": 611
},
{
"epoch": 0.03522302158273381,
"grad_norm": 1.653713345527649,
"learning_rate": 7.886752035446116e-06,
"loss": 0.0934,
"step": 612
},
{
"epoch": 0.03528057553956834,
"grad_norm": 0.9503543972969055,
"learning_rate": 7.852646728329368e-06,
"loss": 0.0567,
"step": 613
},
{
"epoch": 0.03533812949640288,
"grad_norm": 4.422244548797607,
"learning_rate": 7.818567586034578e-06,
"loss": 0.1249,
"step": 614
},
{
"epoch": 0.03539568345323741,
"grad_norm": 3.623196840286255,
"learning_rate": 7.784515023805328e-06,
"loss": 0.1158,
"step": 615
},
{
"epoch": 0.03545323741007194,
"grad_norm": 9.297059059143066,
"learning_rate": 7.750489456561351e-06,
"loss": 0.2269,
"step": 616
},
{
"epoch": 0.03551079136690648,
"grad_norm": 4.102265357971191,
"learning_rate": 7.716491298893443e-06,
"loss": 0.1605,
"step": 617
},
{
"epoch": 0.03556834532374101,
"grad_norm": 1.0876208543777466,
"learning_rate": 7.68252096505843e-06,
"loss": 0.0712,
"step": 618
},
{
"epoch": 0.03562589928057554,
"grad_norm": 3.1616358757019043,
"learning_rate": 7.6485788689741e-06,
"loss": 0.1216,
"step": 619
},
{
"epoch": 0.035683453237410075,
"grad_norm": 1.6394915580749512,
"learning_rate": 7.6146654242141935e-06,
"loss": 0.1088,
"step": 620
},
{
"epoch": 0.035741007194244605,
"grad_norm": 0.6849850416183472,
"learning_rate": 7.580781044003324e-06,
"loss": 0.0639,
"step": 621
},
{
"epoch": 0.035798561151079135,
"grad_norm": 0.7542056441307068,
"learning_rate": 7.546926141211975e-06,
"loss": 0.0541,
"step": 622
},
{
"epoch": 0.03585611510791367,
"grad_norm": 1.4969114065170288,
"learning_rate": 7.513101128351454e-06,
"loss": 0.0822,
"step": 623
},
{
"epoch": 0.0359136690647482,
"grad_norm": 0.6035341620445251,
"learning_rate": 7.4793064175688635e-06,
"loss": 0.0831,
"step": 624
},
{
"epoch": 0.03597122302158273,
"grad_norm": 9.536357879638672,
"learning_rate": 7.445542420642097e-06,
"loss": 0.297,
"step": 625
},
{
"epoch": 0.03602877697841727,
"grad_norm": 10.490472793579102,
"learning_rate": 7.411809548974792e-06,
"loss": 0.2122,
"step": 626
},
{
"epoch": 0.0360863309352518,
"grad_norm": 2.7282614707946777,
"learning_rate": 7.378108213591355e-06,
"loss": 0.0946,
"step": 627
},
{
"epoch": 0.03614388489208633,
"grad_norm": 4.524618625640869,
"learning_rate": 7.344438825131912e-06,
"loss": 0.1423,
"step": 628
},
{
"epoch": 0.03620143884892087,
"grad_norm": 7.327822685241699,
"learning_rate": 7.310801793847344e-06,
"loss": 0.2327,
"step": 629
},
{
"epoch": 0.0362589928057554,
"grad_norm": 2.4597315788269043,
"learning_rate": 7.277197529594257e-06,
"loss": 0.1116,
"step": 630
},
{
"epoch": 0.03631654676258993,
"grad_norm": 8.071950912475586,
"learning_rate": 7.243626441830009e-06,
"loss": 0.1441,
"step": 631
},
{
"epoch": 0.03637410071942446,
"grad_norm": 0.2197929471731186,
"learning_rate": 7.210088939607709e-06,
"loss": 0.0039,
"step": 632
},
{
"epoch": 0.036431654676258994,
"grad_norm": 1.8408621549606323,
"learning_rate": 7.176585431571235e-06,
"loss": 0.0781,
"step": 633
},
{
"epoch": 0.036489208633093524,
"grad_norm": 9.146245002746582,
"learning_rate": 7.143116325950266e-06,
"loss": 0.2037,
"step": 634
},
{
"epoch": 0.036546762589928054,
"grad_norm": 4.6287922859191895,
"learning_rate": 7.109682030555283e-06,
"loss": 0.1195,
"step": 635
},
{
"epoch": 0.03660431654676259,
"grad_norm": 1.386813998222351,
"learning_rate": 7.076282952772634e-06,
"loss": 0.124,
"step": 636
},
{
"epoch": 0.03666187050359712,
"grad_norm": 0.7986965775489807,
"learning_rate": 7.042919499559538e-06,
"loss": 0.0724,
"step": 637
},
{
"epoch": 0.03671942446043165,
"grad_norm": 2.6151773929595947,
"learning_rate": 7.009592077439135e-06,
"loss": 0.0991,
"step": 638
},
{
"epoch": 0.03677697841726619,
"grad_norm": 9.886137962341309,
"learning_rate": 6.976301092495556e-06,
"loss": 0.2132,
"step": 639
},
{
"epoch": 0.03683453237410072,
"grad_norm": 1.0213325023651123,
"learning_rate": 6.943046950368944e-06,
"loss": 0.0824,
"step": 640
},
{
"epoch": 0.03689208633093525,
"grad_norm": 5.319864273071289,
"learning_rate": 6.909830056250527e-06,
"loss": 0.1426,
"step": 641
},
{
"epoch": 0.036949640287769786,
"grad_norm": 11.022806167602539,
"learning_rate": 6.876650814877675e-06,
"loss": 0.3286,
"step": 642
},
{
"epoch": 0.037007194244604316,
"grad_norm": 7.383906841278076,
"learning_rate": 6.843509630528977e-06,
"loss": 0.1446,
"step": 643
},
{
"epoch": 0.037064748201438846,
"grad_norm": 1.5146830081939697,
"learning_rate": 6.8104069070193e-06,
"loss": 0.0942,
"step": 644
},
{
"epoch": 0.03712230215827338,
"grad_norm": 18.990802764892578,
"learning_rate": 6.777343047694891e-06,
"loss": 0.2854,
"step": 645
},
{
"epoch": 0.037179856115107913,
"grad_norm": 2.318795680999756,
"learning_rate": 6.744318455428436e-06,
"loss": 0.0937,
"step": 646
},
{
"epoch": 0.037237410071942444,
"grad_norm": 0.372890830039978,
"learning_rate": 6.711333532614168e-06,
"loss": 0.0548,
"step": 647
},
{
"epoch": 0.03729496402877698,
"grad_norm": 2.1778178215026855,
"learning_rate": 6.67838868116297e-06,
"loss": 0.0613,
"step": 648
},
{
"epoch": 0.03735251798561151,
"grad_norm": 7.633350372314453,
"learning_rate": 6.645484302497452e-06,
"loss": 0.2113,
"step": 649
},
{
"epoch": 0.03741007194244604,
"grad_norm": 2.2798593044281006,
"learning_rate": 6.612620797547087e-06,
"loss": 0.1007,
"step": 650
},
{
"epoch": 0.03746762589928058,
"grad_norm": 1.3645232915878296,
"learning_rate": 6.579798566743314e-06,
"loss": 0.1054,
"step": 651
},
{
"epoch": 0.03752517985611511,
"grad_norm": 3.271883249282837,
"learning_rate": 6.547018010014654e-06,
"loss": 0.1515,
"step": 652
},
{
"epoch": 0.03758273381294964,
"grad_norm": 10.690248489379883,
"learning_rate": 6.5142795267818505e-06,
"loss": 0.2535,
"step": 653
},
{
"epoch": 0.037640287769784175,
"grad_norm": 7.398156642913818,
"learning_rate": 6.481583515952983e-06,
"loss": 0.1515,
"step": 654
},
{
"epoch": 0.037697841726618705,
"grad_norm": 5.868314266204834,
"learning_rate": 6.448930375918632e-06,
"loss": 0.114,
"step": 655
},
{
"epoch": 0.037755395683453236,
"grad_norm": 7.998626708984375,
"learning_rate": 6.4163205045469975e-06,
"loss": 0.1563,
"step": 656
},
{
"epoch": 0.03781294964028777,
"grad_norm": 13.154156684875488,
"learning_rate": 6.383754299179079e-06,
"loss": 0.5158,
"step": 657
},
{
"epoch": 0.0378705035971223,
"grad_norm": 1.3445124626159668,
"learning_rate": 6.351232156623803e-06,
"loss": 0.1183,
"step": 658
},
{
"epoch": 0.03792805755395683,
"grad_norm": 1.747786521911621,
"learning_rate": 6.318754473153221e-06,
"loss": 0.0859,
"step": 659
},
{
"epoch": 0.03798561151079137,
"grad_norm": 0.8909703493118286,
"learning_rate": 6.286321644497655e-06,
"loss": 0.0959,
"step": 660
},
{
"epoch": 0.0380431654676259,
"grad_norm": 2.6019179821014404,
"learning_rate": 6.25393406584088e-06,
"loss": 0.1215,
"step": 661
},
{
"epoch": 0.03810071942446043,
"grad_norm": 0.6137844324111938,
"learning_rate": 6.22159213181533e-06,
"loss": 0.0674,
"step": 662
},
{
"epoch": 0.03815827338129497,
"grad_norm": 5.27604341506958,
"learning_rate": 6.18929623649726e-06,
"loss": 0.1627,
"step": 663
},
{
"epoch": 0.0382158273381295,
"grad_norm": 1.312740445137024,
"learning_rate": 6.157046773401964e-06,
"loss": 0.0954,
"step": 664
},
{
"epoch": 0.03827338129496403,
"grad_norm": 10.17847728729248,
"learning_rate": 6.124844135478971e-06,
"loss": 0.2274,
"step": 665
},
{
"epoch": 0.03833093525179856,
"grad_norm": 2.691664457321167,
"learning_rate": 6.092688715107265e-06,
"loss": 0.0243,
"step": 666
},
{
"epoch": 0.038388489208633095,
"grad_norm": 5.8143229484558105,
"learning_rate": 6.06058090409049e-06,
"loss": 0.1523,
"step": 667
},
{
"epoch": 0.038446043165467625,
"grad_norm": 23.19382095336914,
"learning_rate": 6.028521093652195e-06,
"loss": 0.4124,
"step": 668
},
{
"epoch": 0.038503597122302155,
"grad_norm": 9.377325057983398,
"learning_rate": 5.996509674431053e-06,
"loss": 0.2113,
"step": 669
},
{
"epoch": 0.03856115107913669,
"grad_norm": 12.410694122314453,
"learning_rate": 5.9645470364761e-06,
"loss": 0.2557,
"step": 670
},
{
"epoch": 0.03861870503597122,
"grad_norm": 7.3218865394592285,
"learning_rate": 5.932633569242e-06,
"loss": 0.1555,
"step": 671
},
{
"epoch": 0.03867625899280575,
"grad_norm": 5.622396945953369,
"learning_rate": 5.900769661584273e-06,
"loss": 0.0826,
"step": 672
},
{
"epoch": 0.03873381294964029,
"grad_norm": 5.398272514343262,
"learning_rate": 5.868955701754584e-06,
"loss": 0.1456,
"step": 673
},
{
"epoch": 0.03879136690647482,
"grad_norm": 0.751691460609436,
"learning_rate": 5.83719207739599e-06,
"loss": 0.0772,
"step": 674
},
{
"epoch": 0.03884892086330935,
"grad_norm": 0.6720765233039856,
"learning_rate": 5.8054791755382286e-06,
"loss": 0.0869,
"step": 675
},
{
"epoch": 0.03890647482014389,
"grad_norm": 1.0747441053390503,
"learning_rate": 5.773817382593008e-06,
"loss": 0.072,
"step": 676
},
{
"epoch": 0.03896402877697842,
"grad_norm": 10.877379417419434,
"learning_rate": 5.742207084349274e-06,
"loss": 0.2319,
"step": 677
},
{
"epoch": 0.03902158273381295,
"grad_norm": 1.0769011974334717,
"learning_rate": 5.710648665968543e-06,
"loss": 0.072,
"step": 678
},
{
"epoch": 0.039079136690647484,
"grad_norm": 17.446186065673828,
"learning_rate": 5.679142511980176e-06,
"loss": 0.3885,
"step": 679
},
{
"epoch": 0.039136690647482014,
"grad_norm": 0.5651721358299255,
"learning_rate": 5.647689006276727e-06,
"loss": 0.0675,
"step": 680
},
{
"epoch": 0.039194244604316544,
"grad_norm": 12.628124237060547,
"learning_rate": 5.616288532109225e-06,
"loss": 0.5707,
"step": 681
},
{
"epoch": 0.03925179856115108,
"grad_norm": 1.9146260023117065,
"learning_rate": 5.584941472082549e-06,
"loss": 0.0913,
"step": 682
},
{
"epoch": 0.03930935251798561,
"grad_norm": 13.525870323181152,
"learning_rate": 5.553648208150728e-06,
"loss": 0.4603,
"step": 683
},
{
"epoch": 0.03936690647482014,
"grad_norm": 15.693315505981445,
"learning_rate": 5.522409121612304e-06,
"loss": 0.1253,
"step": 684
},
{
"epoch": 0.03942446043165468,
"grad_norm": 2.555248975753784,
"learning_rate": 5.491224593105695e-06,
"loss": 0.065,
"step": 685
},
{
"epoch": 0.03948201438848921,
"grad_norm": 13.866741180419922,
"learning_rate": 5.460095002604533e-06,
"loss": 0.2646,
"step": 686
},
{
"epoch": 0.03953956834532374,
"grad_norm": 1.9787272214889526,
"learning_rate": 5.429020729413062e-06,
"loss": 0.0937,
"step": 687
},
{
"epoch": 0.039597122302158276,
"grad_norm": 0.8634059429168701,
"learning_rate": 5.398002152161484e-06,
"loss": 0.0827,
"step": 688
},
{
"epoch": 0.039654676258992806,
"grad_norm": 7.688801288604736,
"learning_rate": 5.367039648801386e-06,
"loss": 0.187,
"step": 689
},
{
"epoch": 0.039712230215827336,
"grad_norm": 3.1192588806152344,
"learning_rate": 5.336133596601089e-06,
"loss": 0.1103,
"step": 690
},
{
"epoch": 0.03976978417266187,
"grad_norm": 0.4575349688529968,
"learning_rate": 5.305284372141095e-06,
"loss": 0.0628,
"step": 691
},
{
"epoch": 0.039827338129496404,
"grad_norm": 3.0432817935943604,
"learning_rate": 5.274492351309462e-06,
"loss": 0.0893,
"step": 692
},
{
"epoch": 0.039884892086330934,
"grad_norm": 7.722862243652344,
"learning_rate": 5.243757909297247e-06,
"loss": 0.1858,
"step": 693
},
{
"epoch": 0.03994244604316547,
"grad_norm": 1.459063172340393,
"learning_rate": 5.213081420593933e-06,
"loss": 0.0857,
"step": 694
},
{
"epoch": 0.04,
"grad_norm": 2.9894986152648926,
"learning_rate": 5.1824632589828465e-06,
"loss": 0.1136,
"step": 695
},
{
"epoch": 0.04005755395683453,
"grad_norm": 0.5634929537773132,
"learning_rate": 5.151903797536631e-06,
"loss": 0.0685,
"step": 696
},
{
"epoch": 0.04011510791366907,
"grad_norm": 4.808594703674316,
"learning_rate": 5.121403408612672e-06,
"loss": 0.1148,
"step": 697
},
{
"epoch": 0.0401726618705036,
"grad_norm": 0.5084381103515625,
"learning_rate": 5.090962463848592e-06,
"loss": 0.0665,
"step": 698
},
{
"epoch": 0.04023021582733813,
"grad_norm": 11.494372367858887,
"learning_rate": 5.060581334157693e-06,
"loss": 0.212,
"step": 699
},
{
"epoch": 0.04028776978417266,
"grad_norm": 13.936042785644531,
"learning_rate": 5.030260389724447e-06,
"loss": 0.403,
"step": 700
},
{
"epoch": 0.040345323741007195,
"grad_norm": 4.563905239105225,
"learning_rate": 5.000000000000003e-06,
"loss": 0.2219,
"step": 701
},
{
"epoch": 0.040402877697841726,
"grad_norm": 1.8324753046035767,
"learning_rate": 4.96980053369765e-06,
"loss": 0.0575,
"step": 702
},
{
"epoch": 0.040460431654676256,
"grad_norm": 0.6643631458282471,
"learning_rate": 4.939662358788364e-06,
"loss": 0.0747,
"step": 703
},
{
"epoch": 0.04051798561151079,
"grad_norm": 0.9275362491607666,
"learning_rate": 4.909585842496287e-06,
"loss": 0.0876,
"step": 704
},
{
"epoch": 0.04057553956834532,
"grad_norm": 13.703351020812988,
"learning_rate": 4.879571351294287e-06,
"loss": 0.3226,
"step": 705
},
{
"epoch": 0.04063309352517985,
"grad_norm": 0.6237985491752625,
"learning_rate": 4.849619250899458e-06,
"loss": 0.0687,
"step": 706
},
{
"epoch": 0.04069064748201439,
"grad_norm": 0.9500293135643005,
"learning_rate": 4.8197299062687e-06,
"loss": 0.0092,
"step": 707
},
{
"epoch": 0.04074820143884892,
"grad_norm": 31.131847381591797,
"learning_rate": 4.78990368159424e-06,
"loss": 0.7294,
"step": 708
},
{
"epoch": 0.04080575539568345,
"grad_norm": 2.998800277709961,
"learning_rate": 4.76014094029921e-06,
"loss": 0.0865,
"step": 709
},
{
"epoch": 0.04086330935251799,
"grad_norm": 2.464053153991699,
"learning_rate": 4.7304420450332244e-06,
"loss": 0.0947,
"step": 710
},
{
"epoch": 0.04092086330935252,
"grad_norm": 5.868266582489014,
"learning_rate": 4.700807357667953e-06,
"loss": 0.1665,
"step": 711
},
{
"epoch": 0.04097841726618705,
"grad_norm": 0.5691702365875244,
"learning_rate": 4.671237239292699e-06,
"loss": 0.0655,
"step": 712
},
{
"epoch": 0.041035971223021585,
"grad_norm": 2.4520723819732666,
"learning_rate": 4.641732050210032e-06,
"loss": 0.1552,
"step": 713
},
{
"epoch": 0.041093525179856115,
"grad_norm": 0.49852254986763,
"learning_rate": 4.612292149931369e-06,
"loss": 0.0324,
"step": 714
},
{
"epoch": 0.041151079136690645,
"grad_norm": 12.27933120727539,
"learning_rate": 4.582917897172603e-06,
"loss": 0.1786,
"step": 715
},
{
"epoch": 0.04120863309352518,
"grad_norm": 5.295310020446777,
"learning_rate": 4.5536096498497295e-06,
"loss": 0.0978,
"step": 716
},
{
"epoch": 0.04126618705035971,
"grad_norm": 0.554686427116394,
"learning_rate": 4.524367765074499e-06,
"loss": 0.0519,
"step": 717
},
{
"epoch": 0.04132374100719424,
"grad_norm": 0.8725594282150269,
"learning_rate": 4.495192599150045e-06,
"loss": 0.037,
"step": 718
},
{
"epoch": 0.04138129496402878,
"grad_norm": 1.3279333114624023,
"learning_rate": 4.46608450756656e-06,
"loss": 0.0527,
"step": 719
},
{
"epoch": 0.04143884892086331,
"grad_norm": 7.157567501068115,
"learning_rate": 4.437043844996952e-06,
"loss": 0.1266,
"step": 720
},
{
"epoch": 0.04149640287769784,
"grad_norm": 1.4394761323928833,
"learning_rate": 4.408070965292534e-06,
"loss": 0.1102,
"step": 721
},
{
"epoch": 0.04155395683453238,
"grad_norm": 1.952061653137207,
"learning_rate": 4.379166221478697e-06,
"loss": 0.1155,
"step": 722
},
{
"epoch": 0.04161151079136691,
"grad_norm": 3.076018810272217,
"learning_rate": 4.350329965750622e-06,
"loss": 0.0957,
"step": 723
},
{
"epoch": 0.04166906474820144,
"grad_norm": 5.298429012298584,
"learning_rate": 4.321562549468991e-06,
"loss": 0.0811,
"step": 724
},
{
"epoch": 0.041726618705035974,
"grad_norm": 1.0459011793136597,
"learning_rate": 4.292864323155684e-06,
"loss": 0.0744,
"step": 725
},
{
"epoch": 0.041784172661870504,
"grad_norm": 3.071275234222412,
"learning_rate": 4.264235636489542e-06,
"loss": 0.1764,
"step": 726
},
{
"epoch": 0.041841726618705034,
"grad_norm": 0.6313245296478271,
"learning_rate": 4.235676838302069e-06,
"loss": 0.0659,
"step": 727
},
{
"epoch": 0.04189928057553957,
"grad_norm": 0.4360724687576294,
"learning_rate": 4.207188276573214e-06,
"loss": 0.0432,
"step": 728
},
{
"epoch": 0.0419568345323741,
"grad_norm": 1.621596097946167,
"learning_rate": 4.178770298427107e-06,
"loss": 0.0963,
"step": 729
},
{
"epoch": 0.04201438848920863,
"grad_norm": 7.333670139312744,
"learning_rate": 4.150423250127846e-06,
"loss": 0.1117,
"step": 730
},
{
"epoch": 0.04207194244604317,
"grad_norm": 9.378557205200195,
"learning_rate": 4.12214747707527e-06,
"loss": 0.1928,
"step": 731
},
{
"epoch": 0.0421294964028777,
"grad_norm": 7.514365196228027,
"learning_rate": 4.093943323800746e-06,
"loss": 0.1798,
"step": 732
},
{
"epoch": 0.04218705035971223,
"grad_norm": 3.0120790004730225,
"learning_rate": 4.065811133962987e-06,
"loss": 0.1246,
"step": 733
},
{
"epoch": 0.04224460431654676,
"grad_norm": 1.4734376668930054,
"learning_rate": 4.037751250343841e-06,
"loss": 0.0872,
"step": 734
},
{
"epoch": 0.042302158273381296,
"grad_norm": 0.6325392723083496,
"learning_rate": 4.009764014844143e-06,
"loss": 0.0912,
"step": 735
},
{
"epoch": 0.042359712230215826,
"grad_norm": 2.082659959793091,
"learning_rate": 3.981849768479516e-06,
"loss": 0.1113,
"step": 736
},
{
"epoch": 0.042417266187050356,
"grad_norm": 3.347140312194824,
"learning_rate": 3.954008851376252e-06,
"loss": 0.1088,
"step": 737
},
{
"epoch": 0.042474820143884894,
"grad_norm": 0.7007344365119934,
"learning_rate": 3.9262416027671354e-06,
"loss": 0.0753,
"step": 738
},
{
"epoch": 0.042532374100719424,
"grad_norm": 6.338494300842285,
"learning_rate": 3.898548360987325e-06,
"loss": 0.1733,
"step": 739
},
{
"epoch": 0.042589928057553954,
"grad_norm": 7.861074924468994,
"learning_rate": 3.8709294634702374e-06,
"loss": 0.1722,
"step": 740
},
{
"epoch": 0.04264748201438849,
"grad_norm": 1.3125121593475342,
"learning_rate": 3.8433852467434175e-06,
"loss": 0.0739,
"step": 741
},
{
"epoch": 0.04270503597122302,
"grad_norm": 2.1111292839050293,
"learning_rate": 3.81591604642446e-06,
"loss": 0.1041,
"step": 742
},
{
"epoch": 0.04276258992805755,
"grad_norm": 2.2051801681518555,
"learning_rate": 3.7885221972168974e-06,
"loss": 0.0826,
"step": 743
},
{
"epoch": 0.04282014388489209,
"grad_norm": 4.703516960144043,
"learning_rate": 3.7612040329061405e-06,
"loss": 0.1218,
"step": 744
},
{
"epoch": 0.04287769784172662,
"grad_norm": 0.9855811595916748,
"learning_rate": 3.7339618863553983e-06,
"loss": 0.0765,
"step": 745
},
{
"epoch": 0.04293525179856115,
"grad_norm": 0.4522510766983032,
"learning_rate": 3.7067960895016277e-06,
"loss": 0.0639,
"step": 746
},
{
"epoch": 0.042992805755395685,
"grad_norm": 0.5402107834815979,
"learning_rate": 3.679706973351491e-06,
"loss": 0.0463,
"step": 747
},
{
"epoch": 0.043050359712230216,
"grad_norm": 8.379833221435547,
"learning_rate": 3.6526948679773256e-06,
"loss": 0.2167,
"step": 748
},
{
"epoch": 0.043107913669064746,
"grad_norm": 0.9145704507827759,
"learning_rate": 3.625760102513103e-06,
"loss": 0.098,
"step": 749
},
{
"epoch": 0.04316546762589928,
"grad_norm": 1.1772428750991821,
"learning_rate": 3.598903005150444e-06,
"loss": 0.0908,
"step": 750
},
{
"epoch": 0.04322302158273381,
"grad_norm": 1.089911699295044,
"learning_rate": 3.5721239031346067e-06,
"loss": 0.0803,
"step": 751
},
{
"epoch": 0.04328057553956834,
"grad_norm": 0.725675642490387,
"learning_rate": 3.545423122760493e-06,
"loss": 0.0763,
"step": 752
},
{
"epoch": 0.04333812949640288,
"grad_norm": 1.3648688793182373,
"learning_rate": 3.5188009893686916e-06,
"loss": 0.0807,
"step": 753
},
{
"epoch": 0.04339568345323741,
"grad_norm": 5.510678768157959,
"learning_rate": 3.492257827341492e-06,
"loss": 0.1315,
"step": 754
},
{
"epoch": 0.04345323741007194,
"grad_norm": 1.459609031677246,
"learning_rate": 3.4657939600989453e-06,
"loss": 0.1219,
"step": 755
},
{
"epoch": 0.04351079136690648,
"grad_norm": 8.776249885559082,
"learning_rate": 3.4394097100949286e-06,
"loss": 0.1991,
"step": 756
},
{
"epoch": 0.04356834532374101,
"grad_norm": 2.417057514190674,
"learning_rate": 3.4131053988131947e-06,
"loss": 0.0998,
"step": 757
},
{
"epoch": 0.04362589928057554,
"grad_norm": 12.036551475524902,
"learning_rate": 3.3868813467634833e-06,
"loss": 0.2524,
"step": 758
},
{
"epoch": 0.043683453237410075,
"grad_norm": 14.389498710632324,
"learning_rate": 3.360737873477584e-06,
"loss": 0.3022,
"step": 759
},
{
"epoch": 0.043741007194244605,
"grad_norm": 12.136941909790039,
"learning_rate": 3.3346752975054763e-06,
"loss": 0.2324,
"step": 760
},
{
"epoch": 0.043798561151079135,
"grad_norm": 5.893527507781982,
"learning_rate": 3.308693936411421e-06,
"loss": 0.1189,
"step": 761
},
{
"epoch": 0.04385611510791367,
"grad_norm": 1.1870757341384888,
"learning_rate": 3.2827941067700996e-06,
"loss": 0.0794,
"step": 762
},
{
"epoch": 0.0439136690647482,
"grad_norm": 0.639008641242981,
"learning_rate": 3.2569761241627694e-06,
"loss": 0.0606,
"step": 763
},
{
"epoch": 0.04397122302158273,
"grad_norm": 2.8850464820861816,
"learning_rate": 3.2312403031733943e-06,
"loss": 0.1593,
"step": 764
},
{
"epoch": 0.04402877697841727,
"grad_norm": 15.170232772827148,
"learning_rate": 3.2055869573848374e-06,
"loss": 0.2961,
"step": 765
},
{
"epoch": 0.0440863309352518,
"grad_norm": 1.2211562395095825,
"learning_rate": 3.1800163993750166e-06,
"loss": 0.1089,
"step": 766
},
{
"epoch": 0.04414388489208633,
"grad_norm": 0.4650379717350006,
"learning_rate": 3.1545289407131128e-06,
"loss": 0.0683,
"step": 767
},
{
"epoch": 0.04420143884892087,
"grad_norm": 1.6909816265106201,
"learning_rate": 3.1291248919557717e-06,
"loss": 0.123,
"step": 768
},
{
"epoch": 0.0442589928057554,
"grad_norm": 0.07863793522119522,
"learning_rate": 3.103804562643302e-06,
"loss": 0.0008,
"step": 769
},
{
"epoch": 0.04431654676258993,
"grad_norm": 8.237005233764648,
"learning_rate": 3.0785682612959334e-06,
"loss": 0.1946,
"step": 770
},
{
"epoch": 0.04437410071942446,
"grad_norm": 1.3069466352462769,
"learning_rate": 3.0534162954100264e-06,
"loss": 0.0659,
"step": 771
},
{
"epoch": 0.044431654676258994,
"grad_norm": 0.9894955158233643,
"learning_rate": 3.028348971454356e-06,
"loss": 0.0554,
"step": 772
},
{
"epoch": 0.044489208633093524,
"grad_norm": 1.0656694173812866,
"learning_rate": 3.003366594866345e-06,
"loss": 0.0713,
"step": 773
},
{
"epoch": 0.044546762589928054,
"grad_norm": 2.17790150642395,
"learning_rate": 2.978469470048376e-06,
"loss": 0.1119,
"step": 774
},
{
"epoch": 0.04460431654676259,
"grad_norm": 1.1125558614730835,
"learning_rate": 2.953657900364053e-06,
"loss": 0.0799,
"step": 775
},
{
"epoch": 0.04466187050359712,
"grad_norm": 0.7158892154693604,
"learning_rate": 2.9289321881345257e-06,
"loss": 0.0711,
"step": 776
},
{
"epoch": 0.04471942446043165,
"grad_norm": 1.9382431507110596,
"learning_rate": 2.9042926346347932e-06,
"loss": 0.0626,
"step": 777
},
{
"epoch": 0.04477697841726619,
"grad_norm": 3.6688735485076904,
"learning_rate": 2.8797395400900362e-06,
"loss": 0.1775,
"step": 778
},
{
"epoch": 0.04483453237410072,
"grad_norm": 0.9889999032020569,
"learning_rate": 2.855273203671969e-06,
"loss": 0.0765,
"step": 779
},
{
"epoch": 0.04489208633093525,
"grad_norm": 14.821270942687988,
"learning_rate": 2.830893923495173e-06,
"loss": 0.5594,
"step": 780
},
{
"epoch": 0.044949640287769786,
"grad_norm": 2.756847620010376,
"learning_rate": 2.8066019966134907e-06,
"loss": 0.0805,
"step": 781
},
{
"epoch": 0.045007194244604316,
"grad_norm": 2.8205065727233887,
"learning_rate": 2.7823977190163788e-06,
"loss": 0.1263,
"step": 782
},
{
"epoch": 0.045064748201438846,
"grad_norm": 12.99278736114502,
"learning_rate": 2.7582813856253276e-06,
"loss": 0.4152,
"step": 783
},
{
"epoch": 0.045122302158273384,
"grad_norm": 2.810309410095215,
"learning_rate": 2.7342532902902418e-06,
"loss": 0.0943,
"step": 784
},
{
"epoch": 0.045179856115107914,
"grad_norm": 1.1653300523757935,
"learning_rate": 2.7103137257858867e-06,
"loss": 0.0877,
"step": 785
},
{
"epoch": 0.045237410071942444,
"grad_norm": 5.321506023406982,
"learning_rate": 2.6864629838082957e-06,
"loss": 0.1012,
"step": 786
},
{
"epoch": 0.04529496402877698,
"grad_norm": 0.6735225915908813,
"learning_rate": 2.6627013549712355e-06,
"loss": 0.066,
"step": 787
},
{
"epoch": 0.04535251798561151,
"grad_norm": 0.7565364241600037,
"learning_rate": 2.639029128802657e-06,
"loss": 0.0728,
"step": 788
},
{
"epoch": 0.04541007194244604,
"grad_norm": 1.2969605922698975,
"learning_rate": 2.615446593741161e-06,
"loss": 0.0699,
"step": 789
},
{
"epoch": 0.04546762589928058,
"grad_norm": 1.2415449619293213,
"learning_rate": 2.5919540371325005e-06,
"loss": 0.0802,
"step": 790
},
{
"epoch": 0.04552517985611511,
"grad_norm": 3.416377067565918,
"learning_rate": 2.5685517452260566e-06,
"loss": 0.0328,
"step": 791
},
{
"epoch": 0.04558273381294964,
"grad_norm": 3.912259340286255,
"learning_rate": 2.5452400031713786e-06,
"loss": 0.0989,
"step": 792
},
{
"epoch": 0.045640287769784176,
"grad_norm": 4.516401767730713,
"learning_rate": 2.522019095014683e-06,
"loss": 0.1032,
"step": 793
},
{
"epoch": 0.045697841726618706,
"grad_norm": 1.5598896741867065,
"learning_rate": 2.4988893036954045e-06,
"loss": 0.0709,
"step": 794
},
{
"epoch": 0.045755395683453236,
"grad_norm": 4.611602306365967,
"learning_rate": 2.4758509110427576e-06,
"loss": 0.1643,
"step": 795
},
{
"epoch": 0.04581294964028777,
"grad_norm": 5.493204593658447,
"learning_rate": 2.45290419777228e-06,
"loss": 0.0297,
"step": 796
},
{
"epoch": 0.0458705035971223,
"grad_norm": 0.6497253775596619,
"learning_rate": 2.4300494434824373e-06,
"loss": 0.0486,
"step": 797
},
{
"epoch": 0.04592805755395683,
"grad_norm": 1.2920628786087036,
"learning_rate": 2.407286926651192e-06,
"loss": 0.0787,
"step": 798
},
{
"epoch": 0.04598561151079137,
"grad_norm": 2.440001964569092,
"learning_rate": 2.3846169246326345e-06,
"loss": 0.1111,
"step": 799
},
{
"epoch": 0.0460431654676259,
"grad_norm": 1.9451384544372559,
"learning_rate": 2.362039713653581e-06,
"loss": 0.0868,
"step": 800
},
{
"epoch": 0.04610071942446043,
"grad_norm": 2.6515021324157715,
"learning_rate": 2.339555568810221e-06,
"loss": 0.1016,
"step": 801
},
{
"epoch": 0.04615827338129497,
"grad_norm": 2.210442304611206,
"learning_rate": 2.317164764064769e-06,
"loss": 0.0648,
"step": 802
},
{
"epoch": 0.0462158273381295,
"grad_norm": 1.4524075984954834,
"learning_rate": 2.2948675722421086e-06,
"loss": 0.0752,
"step": 803
},
{
"epoch": 0.04627338129496403,
"grad_norm": 16.323625564575195,
"learning_rate": 2.27266426502649e-06,
"loss": 0.2994,
"step": 804
},
{
"epoch": 0.04633093525179856,
"grad_norm": 0.39062535762786865,
"learning_rate": 2.2505551129582047e-06,
"loss": 0.0519,
"step": 805
},
{
"epoch": 0.046388489208633095,
"grad_norm": 1.672765851020813,
"learning_rate": 2.2285403854302912e-06,
"loss": 0.064,
"step": 806
},
{
"epoch": 0.046446043165467625,
"grad_norm": 8.528423309326172,
"learning_rate": 2.206620350685257e-06,
"loss": 0.1726,
"step": 807
},
{
"epoch": 0.046503597122302155,
"grad_norm": 2.00990629196167,
"learning_rate": 2.1847952758118118e-06,
"loss": 0.0689,
"step": 808
},
{
"epoch": 0.04656115107913669,
"grad_norm": 1.1098554134368896,
"learning_rate": 2.163065426741603e-06,
"loss": 0.0746,
"step": 809
},
{
"epoch": 0.04661870503597122,
"grad_norm": 6.2346014976501465,
"learning_rate": 2.1414310682459805e-06,
"loss": 0.128,
"step": 810
},
{
"epoch": 0.04667625899280575,
"grad_norm": 20.44989776611328,
"learning_rate": 2.119892463932781e-06,
"loss": 0.2095,
"step": 811
},
{
"epoch": 0.04673381294964029,
"grad_norm": 6.084557056427002,
"learning_rate": 2.098449876243096e-06,
"loss": 0.1533,
"step": 812
},
{
"epoch": 0.04679136690647482,
"grad_norm": 18.029020309448242,
"learning_rate": 2.0771035664480944e-06,
"loss": 0.2013,
"step": 813
},
{
"epoch": 0.04684892086330935,
"grad_norm": 2.0399179458618164,
"learning_rate": 2.0558537946458177e-06,
"loss": 0.0732,
"step": 814
},
{
"epoch": 0.04690647482014389,
"grad_norm": 0.7458027005195618,
"learning_rate": 2.0347008197580376e-06,
"loss": 0.0815,
"step": 815
},
{
"epoch": 0.04696402877697842,
"grad_norm": 1.1791083812713623,
"learning_rate": 2.013644899527074e-06,
"loss": 0.0646,
"step": 816
},
{
"epoch": 0.04702158273381295,
"grad_norm": 1.5898396968841553,
"learning_rate": 1.9926862905126663e-06,
"loss": 0.0786,
"step": 817
},
{
"epoch": 0.047079136690647484,
"grad_norm": 0.5349504947662354,
"learning_rate": 1.9718252480888567e-06,
"loss": 0.0573,
"step": 818
},
{
"epoch": 0.047136690647482014,
"grad_norm": 1.0246250629425049,
"learning_rate": 1.95106202644086e-06,
"loss": 0.0941,
"step": 819
},
{
"epoch": 0.047194244604316545,
"grad_norm": 2.1197612285614014,
"learning_rate": 1.930396878561983e-06,
"loss": 0.0905,
"step": 820
},
{
"epoch": 0.04725179856115108,
"grad_norm": 4.380585670471191,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.1047,
"step": 821
},
{
"epoch": 0.04730935251798561,
"grad_norm": 0.4877808094024658,
"learning_rate": 1.8893618101067357e-06,
"loss": 0.0526,
"step": 822
},
{
"epoch": 0.04736690647482014,
"grad_norm": 5.169129848480225,
"learning_rate": 1.8689923895297247e-06,
"loss": 0.1075,
"step": 823
},
{
"epoch": 0.04742446043165468,
"grad_norm": 0.528556227684021,
"learning_rate": 1.848722042714457e-06,
"loss": 0.0662,
"step": 824
},
{
"epoch": 0.04748201438848921,
"grad_norm": 1.4885525703430176,
"learning_rate": 1.8285510166487154e-06,
"loss": 0.0809,
"step": 825
},
{
"epoch": 0.04753956834532374,
"grad_norm": 0.9235038161277771,
"learning_rate": 1.808479557110081e-06,
"loss": 0.049,
"step": 826
},
{
"epoch": 0.047597122302158276,
"grad_norm": 3.030466079711914,
"learning_rate": 1.7885079086629598e-06,
"loss": 0.0674,
"step": 827
},
{
"epoch": 0.047654676258992806,
"grad_norm": 0.3380258083343506,
"learning_rate": 1.7686363146555807e-06,
"loss": 0.0407,
"step": 828
},
{
"epoch": 0.047712230215827336,
"grad_norm": 0.3373538553714752,
"learning_rate": 1.7488650172170496e-06,
"loss": 0.0557,
"step": 829
},
{
"epoch": 0.047769784172661874,
"grad_norm": 8.12345027923584,
"learning_rate": 1.7291942572543806e-06,
"loss": 0.1242,
"step": 830
},
{
"epoch": 0.047827338129496404,
"grad_norm": 18.61945915222168,
"learning_rate": 1.709624274449584e-06,
"loss": 0.1312,
"step": 831
},
{
"epoch": 0.047884892086330934,
"grad_norm": 1.0393412113189697,
"learning_rate": 1.6901553072567189e-06,
"loss": 0.0669,
"step": 832
},
{
"epoch": 0.04794244604316547,
"grad_norm": 0.405079185962677,
"learning_rate": 1.6707875928990059e-06,
"loss": 0.0455,
"step": 833
},
{
"epoch": 0.048,
"grad_norm": 1.6890630722045898,
"learning_rate": 1.651521367365936e-06,
"loss": 0.0977,
"step": 834
},
{
"epoch": 0.04805755395683453,
"grad_norm": 2.578721523284912,
"learning_rate": 1.6323568654103838e-06,
"loss": 0.0692,
"step": 835
},
{
"epoch": 0.04811510791366907,
"grad_norm": 1.0500586032867432,
"learning_rate": 1.6132943205457607e-06,
"loss": 0.1101,
"step": 836
},
{
"epoch": 0.0481726618705036,
"grad_norm": 25.174041748046875,
"learning_rate": 1.5943339650431578e-06,
"loss": 0.341,
"step": 837
},
{
"epoch": 0.04823021582733813,
"grad_norm": 0.7596455216407776,
"learning_rate": 1.5754760299285255e-06,
"loss": 0.0793,
"step": 838
},
{
"epoch": 0.04828776978417266,
"grad_norm": 0.9107263684272766,
"learning_rate": 1.5567207449798517e-06,
"loss": 0.079,
"step": 839
},
{
"epoch": 0.048345323741007196,
"grad_norm": 15.38805866241455,
"learning_rate": 1.538068338724361e-06,
"loss": 0.4072,
"step": 840
},
{
"epoch": 0.048402877697841726,
"grad_norm": 0.3658386766910553,
"learning_rate": 1.5195190384357405e-06,
"loss": 0.0568,
"step": 841
},
{
"epoch": 0.048460431654676256,
"grad_norm": 5.78350305557251,
"learning_rate": 1.5010730701313626e-06,
"loss": 0.078,
"step": 842
},
{
"epoch": 0.04851798561151079,
"grad_norm": 0.8679736256599426,
"learning_rate": 1.4827306585695234e-06,
"loss": 0.0967,
"step": 843
},
{
"epoch": 0.04857553956834532,
"grad_norm": 11.486621856689453,
"learning_rate": 1.4644920272467245e-06,
"loss": 0.1852,
"step": 844
},
{
"epoch": 0.04863309352517985,
"grad_norm": 7.43509578704834,
"learning_rate": 1.446357398394934e-06,
"loss": 0.1715,
"step": 845
},
{
"epoch": 0.04869064748201439,
"grad_norm": 0.7901409864425659,
"learning_rate": 1.4283269929788779e-06,
"loss": 0.0405,
"step": 846
},
{
"epoch": 0.04874820143884892,
"grad_norm": 1.712028980255127,
"learning_rate": 1.4104010306933558e-06,
"loss": 0.088,
"step": 847
},
{
"epoch": 0.04880575539568345,
"grad_norm": 1.992058277130127,
"learning_rate": 1.3925797299605649e-06,
"loss": 0.1193,
"step": 848
},
{
"epoch": 0.04886330935251799,
"grad_norm": 0.7998857498168945,
"learning_rate": 1.3748633079274254e-06,
"loss": 0.0473,
"step": 849
},
{
"epoch": 0.04892086330935252,
"grad_norm": 0.4824025332927704,
"learning_rate": 1.3572519804629537e-06,
"loss": 0.0652,
"step": 850
},
{
"epoch": 0.04897841726618705,
"grad_norm": 0.8086461424827576,
"learning_rate": 1.339745962155613e-06,
"loss": 0.0645,
"step": 851
},
{
"epoch": 0.049035971223021585,
"grad_norm": 27.238941192626953,
"learning_rate": 1.322345466310717e-06,
"loss": 0.4208,
"step": 852
},
{
"epoch": 0.049093525179856115,
"grad_norm": 1.820080041885376,
"learning_rate": 1.30505070494781e-06,
"loss": 0.1134,
"step": 853
},
{
"epoch": 0.049151079136690645,
"grad_norm": 0.9083067178726196,
"learning_rate": 1.2878618887981064e-06,
"loss": 0.0782,
"step": 854
},
{
"epoch": 0.04920863309352518,
"grad_norm": 0.7896227240562439,
"learning_rate": 1.2707792273019049e-06,
"loss": 0.0586,
"step": 855
},
{
"epoch": 0.04926618705035971,
"grad_norm": 0.8508791923522949,
"learning_rate": 1.2538029286060428e-06,
"loss": 0.059,
"step": 856
},
{
"epoch": 0.04932374100719424,
"grad_norm": 1.0554744005203247,
"learning_rate": 1.2369331995613664e-06,
"loss": 0.0841,
"step": 857
},
{
"epoch": 0.04938129496402878,
"grad_norm": 1.7418127059936523,
"learning_rate": 1.2201702457201948e-06,
"loss": 0.0688,
"step": 858
},
{
"epoch": 0.04943884892086331,
"grad_norm": 3.3948047161102295,
"learning_rate": 1.2035142713338366e-06,
"loss": 0.0954,
"step": 859
},
{
"epoch": 0.04949640287769784,
"grad_norm": 1.1235921382904053,
"learning_rate": 1.1869654793500784e-06,
"loss": 0.0726,
"step": 860
},
{
"epoch": 0.04955395683453238,
"grad_norm": 4.876044750213623,
"learning_rate": 1.1705240714107301e-06,
"loss": 0.1062,
"step": 861
},
{
"epoch": 0.04961151079136691,
"grad_norm": 6.03566837310791,
"learning_rate": 1.1541902478491607e-06,
"loss": 0.1015,
"step": 862
},
{
"epoch": 0.04966906474820144,
"grad_norm": 0.6971962451934814,
"learning_rate": 1.1379642076878528e-06,
"loss": 0.0514,
"step": 863
},
{
"epoch": 0.049726618705035974,
"grad_norm": 11.001822471618652,
"learning_rate": 1.1218461486359878e-06,
"loss": 0.467,
"step": 864
},
{
"epoch": 0.049784172661870504,
"grad_norm": 1.605398178100586,
"learning_rate": 1.1058362670870248e-06,
"loss": 0.1081,
"step": 865
},
{
"epoch": 0.049841726618705035,
"grad_norm": 17.996734619140625,
"learning_rate": 1.0899347581163222e-06,
"loss": 0.3179,
"step": 866
},
{
"epoch": 0.04989928057553957,
"grad_norm": 0.2730165123939514,
"learning_rate": 1.0741418154787443e-06,
"loss": 0.0261,
"step": 867
},
{
"epoch": 0.0499568345323741,
"grad_norm": 5.960131645202637,
"learning_rate": 1.058457631606319e-06,
"loss": 0.1051,
"step": 868
},
{
"epoch": 0.05001438848920863,
"grad_norm": 2.7779898643493652,
"learning_rate": 1.042882397605871e-06,
"loss": 0.0692,
"step": 869
},
{
"epoch": 0.05007194244604317,
"grad_norm": 0.33091625571250916,
"learning_rate": 1.0274163032567165e-06,
"loss": 0.0534,
"step": 870
},
{
"epoch": 0.0501294964028777,
"grad_norm": 0.6273438334465027,
"learning_rate": 1.012059537008332e-06,
"loss": 0.0617,
"step": 871
},
{
"epoch": 0.05018705035971223,
"grad_norm": 0.8654516935348511,
"learning_rate": 9.968122859780648e-07,
"loss": 0.0568,
"step": 872
},
{
"epoch": 0.05024460431654676,
"grad_norm": 13.922266960144043,
"learning_rate": 9.816747359488632e-07,
"loss": 0.1749,
"step": 873
},
{
"epoch": 0.050302158273381296,
"grad_norm": 16.455718994140625,
"learning_rate": 9.666470713669918e-07,
"loss": 0.2171,
"step": 874
},
{
"epoch": 0.050359712230215826,
"grad_norm": 11.806230545043945,
"learning_rate": 9.517294753398066e-07,
"loss": 0.25,
"step": 875
},
{
"epoch": 0.05041726618705036,
"grad_norm": 0.3645191192626953,
"learning_rate": 9.369221296335007e-07,
"loss": 0.05,
"step": 876
},
{
"epoch": 0.050474820143884894,
"grad_norm": 0.4936099648475647,
"learning_rate": 9.222252146709143e-07,
"loss": 0.0519,
"step": 877
},
{
"epoch": 0.050532374100719424,
"grad_norm": 13.692974090576172,
"learning_rate": 9.076389095293148e-07,
"loss": 0.2554,
"step": 878
},
{
"epoch": 0.050589928057553954,
"grad_norm": 0.6284224987030029,
"learning_rate": 8.931633919382299e-07,
"loss": 0.0659,
"step": 879
},
{
"epoch": 0.05064748201438849,
"grad_norm": 0.7923092842102051,
"learning_rate": 8.787988382772705e-07,
"loss": 0.0568,
"step": 880
},
{
"epoch": 0.05070503597122302,
"grad_norm": 22.334707260131836,
"learning_rate": 8.645454235739903e-07,
"loss": 0.2308,
"step": 881
},
{
"epoch": 0.05076258992805755,
"grad_norm": 4.818454742431641,
"learning_rate": 8.504033215017527e-07,
"loss": 0.1232,
"step": 882
},
{
"epoch": 0.05082014388489209,
"grad_norm": 1.418079137802124,
"learning_rate": 8.363727043776037e-07,
"loss": 0.0688,
"step": 883
},
{
"epoch": 0.05087769784172662,
"grad_norm": 0.7452939748764038,
"learning_rate": 8.224537431601886e-07,
"loss": 0.0562,
"step": 884
},
{
"epoch": 0.05093525179856115,
"grad_norm": 1.132005214691162,
"learning_rate": 8.086466074476562e-07,
"loss": 0.0804,
"step": 885
},
{
"epoch": 0.050992805755395686,
"grad_norm": 1.1662274599075317,
"learning_rate": 7.949514654755963e-07,
"loss": 0.0778,
"step": 886
},
{
"epoch": 0.051050359712230216,
"grad_norm": 0.6592736840248108,
"learning_rate": 7.81368484114996e-07,
"loss": 0.0662,
"step": 887
},
{
"epoch": 0.051107913669064746,
"grad_norm": 9.107586860656738,
"learning_rate": 7.678978288701911e-07,
"loss": 0.1515,
"step": 888
},
{
"epoch": 0.05116546762589928,
"grad_norm": 11.478974342346191,
"learning_rate": 7.545396638768698e-07,
"loss": 0.496,
"step": 889
},
{
"epoch": 0.05122302158273381,
"grad_norm": 0.9263907670974731,
"learning_rate": 7.412941519000527e-07,
"loss": 0.0756,
"step": 890
},
{
"epoch": 0.05128057553956834,
"grad_norm": 20.043466567993164,
"learning_rate": 7.281614543321269e-07,
"loss": 0.5612,
"step": 891
},
{
"epoch": 0.05133812949640288,
"grad_norm": 9.131746292114258,
"learning_rate": 7.151417311908648e-07,
"loss": 0.1701,
"step": 892
},
{
"epoch": 0.05139568345323741,
"grad_norm": 2.0880978107452393,
"learning_rate": 7.022351411174866e-07,
"loss": 0.1408,
"step": 893
},
{
"epoch": 0.05145323741007194,
"grad_norm": 0.7419898509979248,
"learning_rate": 6.894418413747183e-07,
"loss": 0.0609,
"step": 894
},
{
"epoch": 0.05151079136690648,
"grad_norm": 18.692054748535156,
"learning_rate": 6.767619878448783e-07,
"loss": 0.3365,
"step": 895
},
{
"epoch": 0.05156834532374101,
"grad_norm": 1.1873613595962524,
"learning_rate": 6.641957350279838e-07,
"loss": 0.0662,
"step": 896
},
{
"epoch": 0.05162589928057554,
"grad_norm": 1.0309622287750244,
"learning_rate": 6.517432360398556e-07,
"loss": 0.062,
"step": 897
},
{
"epoch": 0.051683453237410075,
"grad_norm": 1.502915620803833,
"learning_rate": 6.394046426102673e-07,
"loss": 0.0679,
"step": 898
},
{
"epoch": 0.051741007194244605,
"grad_norm": 14.485411643981934,
"learning_rate": 6.271801050810856e-07,
"loss": 0.2849,
"step": 899
},
{
"epoch": 0.051798561151079135,
"grad_norm": 0.6515581607818604,
"learning_rate": 6.150697724044407e-07,
"loss": 0.0497,
"step": 900
},
{
"epoch": 0.05185611510791367,
"grad_norm": 1.2064189910888672,
"learning_rate": 6.030737921409169e-07,
"loss": 0.0801,
"step": 901
},
{
"epoch": 0.0519136690647482,
"grad_norm": 9.05174446105957,
"learning_rate": 5.911923104577455e-07,
"loss": 0.1349,
"step": 902
},
{
"epoch": 0.05197122302158273,
"grad_norm": 0.8396304845809937,
"learning_rate": 5.794254721270331e-07,
"loss": 0.0493,
"step": 903
},
{
"epoch": 0.05202877697841727,
"grad_norm": 22.91259002685547,
"learning_rate": 5.677734205239904e-07,
"loss": 0.4679,
"step": 904
},
{
"epoch": 0.0520863309352518,
"grad_norm": 0.31920236349105835,
"learning_rate": 5.562362976251901e-07,
"loss": 0.0485,
"step": 905
},
{
"epoch": 0.05214388489208633,
"grad_norm": 0.7108873724937439,
"learning_rate": 5.448142440068316e-07,
"loss": 0.0509,
"step": 906
},
{
"epoch": 0.05220143884892086,
"grad_norm": 1.8079339265823364,
"learning_rate": 5.335073988430373e-07,
"loss": 0.0969,
"step": 907
},
{
"epoch": 0.0522589928057554,
"grad_norm": 10.256385803222656,
"learning_rate": 5.223158999041444e-07,
"loss": 0.2056,
"step": 908
},
{
"epoch": 0.05231654676258993,
"grad_norm": 0.8384624123573303,
"learning_rate": 5.112398835550348e-07,
"loss": 0.0781,
"step": 909
},
{
"epoch": 0.05237410071942446,
"grad_norm": 0.9950568079948425,
"learning_rate": 5.002794847534765e-07,
"loss": 0.0569,
"step": 910
},
{
"epoch": 0.052431654676258994,
"grad_norm": 0.7972750067710876,
"learning_rate": 4.894348370484648e-07,
"loss": 0.072,
"step": 911
},
{
"epoch": 0.052489208633093525,
"grad_norm": 8.564375877380371,
"learning_rate": 4.787060725786141e-07,
"loss": 0.1262,
"step": 912
},
{
"epoch": 0.052546762589928055,
"grad_norm": 0.7976115942001343,
"learning_rate": 4.6809332207053083e-07,
"loss": 0.0669,
"step": 913
},
{
"epoch": 0.05260431654676259,
"grad_norm": 0.6580243706703186,
"learning_rate": 4.575967148372318e-07,
"loss": 0.071,
"step": 914
},
{
"epoch": 0.05266187050359712,
"grad_norm": 0.8931045532226562,
"learning_rate": 4.4721637877656377e-07,
"loss": 0.0581,
"step": 915
},
{
"epoch": 0.05271942446043165,
"grad_norm": 0.581693172454834,
"learning_rate": 4.3695244036964567e-07,
"loss": 0.0569,
"step": 916
},
{
"epoch": 0.05277697841726619,
"grad_norm": 14.89986801147461,
"learning_rate": 4.268050246793276e-07,
"loss": 0.3469,
"step": 917
},
{
"epoch": 0.05283453237410072,
"grad_norm": 4.767395496368408,
"learning_rate": 4.167742553486676e-07,
"loss": 0.0892,
"step": 918
},
{
"epoch": 0.05289208633093525,
"grad_norm": 1.3745571374893188,
"learning_rate": 4.068602545994249e-07,
"loss": 0.0874,
"step": 919
},
{
"epoch": 0.052949640287769786,
"grad_norm": 7.833735466003418,
"learning_rate": 3.9706314323056936e-07,
"loss": 0.1356,
"step": 920
},
{
"epoch": 0.053007194244604317,
"grad_norm": 1.1918085813522339,
"learning_rate": 3.8738304061681107e-07,
"loss": 0.0653,
"step": 921
},
{
"epoch": 0.05306474820143885,
"grad_norm": 0.6890537142753601,
"learning_rate": 3.7782006470714614e-07,
"loss": 0.0692,
"step": 922
},
{
"epoch": 0.053122302158273384,
"grad_norm": 0.6393564939498901,
"learning_rate": 3.68374332023419e-07,
"loss": 0.0206,
"step": 923
},
{
"epoch": 0.053179856115107914,
"grad_norm": 1.7775517702102661,
"learning_rate": 3.590459576589e-07,
"loss": 0.1072,
"step": 924
},
{
"epoch": 0.053237410071942444,
"grad_norm": 0.5203324556350708,
"learning_rate": 3.498350552768859e-07,
"loss": 0.0651,
"step": 925
},
{
"epoch": 0.05329496402877698,
"grad_norm": 0.6219776272773743,
"learning_rate": 3.4074173710931804e-07,
"loss": 0.0653,
"step": 926
},
{
"epoch": 0.05335251798561151,
"grad_norm": 1.8375036716461182,
"learning_rate": 3.3176611395540625e-07,
"loss": 0.0614,
"step": 927
},
{
"epoch": 0.05341007194244604,
"grad_norm": 2.563394784927368,
"learning_rate": 3.2290829518028867e-07,
"loss": 0.1311,
"step": 928
},
{
"epoch": 0.05346762589928058,
"grad_norm": 17.644411087036133,
"learning_rate": 3.1416838871368925e-07,
"loss": 0.3422,
"step": 929
},
{
"epoch": 0.05352517985611511,
"grad_norm": 0.7539731860160828,
"learning_rate": 3.0554650104861137e-07,
"loss": 0.0837,
"step": 930
},
{
"epoch": 0.05358273381294964,
"grad_norm": 5.77402400970459,
"learning_rate": 2.970427372400353e-07,
"loss": 0.1455,
"step": 931
},
{
"epoch": 0.053640287769784176,
"grad_norm": 1.0641944408416748,
"learning_rate": 2.8865720090364037e-07,
"loss": 0.1101,
"step": 932
},
{
"epoch": 0.053697841726618706,
"grad_norm": 1.1459119319915771,
"learning_rate": 2.8038999421453827e-07,
"loss": 0.0608,
"step": 933
},
{
"epoch": 0.053755395683453236,
"grad_norm": 12.172619819641113,
"learning_rate": 2.7224121790603517e-07,
"loss": 0.3217,
"step": 934
},
{
"epoch": 0.05381294964028777,
"grad_norm": 11.910560607910156,
"learning_rate": 2.6421097126839714e-07,
"loss": 0.5355,
"step": 935
},
{
"epoch": 0.0538705035971223,
"grad_norm": 2.0426604747772217,
"learning_rate": 2.5629935214764866e-07,
"loss": 0.1398,
"step": 936
},
{
"epoch": 0.05392805755395683,
"grad_norm": 1.1188652515411377,
"learning_rate": 2.4850645694436736e-07,
"loss": 0.1328,
"step": 937
},
{
"epoch": 0.05398561151079137,
"grad_norm": 2.1541991233825684,
"learning_rate": 2.4083238061252565e-07,
"loss": 0.1437,
"step": 938
},
{
"epoch": 0.0540431654676259,
"grad_norm": 0.8203990459442139,
"learning_rate": 2.332772166583208e-07,
"loss": 0.0617,
"step": 939
},
{
"epoch": 0.05410071942446043,
"grad_norm": 14.203892707824707,
"learning_rate": 2.2584105713904126e-07,
"loss": 0.151,
"step": 940
},
{
"epoch": 0.05415827338129496,
"grad_norm": 0.3919861912727356,
"learning_rate": 2.1852399266194312e-07,
"loss": 0.0057,
"step": 941
},
{
"epoch": 0.0542158273381295,
"grad_norm": 3.8022513389587402,
"learning_rate": 2.1132611238315004e-07,
"loss": 0.118,
"step": 942
},
{
"epoch": 0.05427338129496403,
"grad_norm": 15.274145126342773,
"learning_rate": 2.0424750400655947e-07,
"loss": 0.2713,
"step": 943
},
{
"epoch": 0.05433093525179856,
"grad_norm": 1.0683073997497559,
"learning_rate": 1.9728825378278248e-07,
"loss": 0.0581,
"step": 944
},
{
"epoch": 0.054388489208633095,
"grad_norm": 13.709763526916504,
"learning_rate": 1.9044844650808468e-07,
"loss": 0.2315,
"step": 945
},
{
"epoch": 0.054446043165467625,
"grad_norm": 2.2819840908050537,
"learning_rate": 1.8372816552336025e-07,
"loss": 0.1143,
"step": 946
},
{
"epoch": 0.054503597122302155,
"grad_norm": 1.1235344409942627,
"learning_rate": 1.7712749271311392e-07,
"loss": 0.0717,
"step": 947
},
{
"epoch": 0.05456115107913669,
"grad_norm": 1.0503443479537964,
"learning_rate": 1.706465085044584e-07,
"loss": 0.0961,
"step": 948
},
{
"epoch": 0.05461870503597122,
"grad_norm": 0.7241622805595398,
"learning_rate": 1.6428529186614195e-07,
"loss": 0.078,
"step": 949
},
{
"epoch": 0.05467625899280575,
"grad_norm": 1.2095136642456055,
"learning_rate": 1.580439203075812e-07,
"loss": 0.0911,
"step": 950
},
{
"epoch": 0.05473381294964029,
"grad_norm": 4.268467426300049,
"learning_rate": 1.519224698779198e-07,
"loss": 0.0737,
"step": 951
},
{
"epoch": 0.05479136690647482,
"grad_norm": 0.36049625277519226,
"learning_rate": 1.4592101516509916e-07,
"loss": 0.0585,
"step": 952
},
{
"epoch": 0.05484892086330935,
"grad_norm": 2.7988123893737793,
"learning_rate": 1.400396292949513e-07,
"loss": 0.0986,
"step": 953
},
{
"epoch": 0.05490647482014389,
"grad_norm": 16.90985870361328,
"learning_rate": 1.3427838393030634e-07,
"loss": 0.5819,
"step": 954
},
{
"epoch": 0.05496402877697842,
"grad_norm": 0.7780059576034546,
"learning_rate": 1.2863734927012094e-07,
"loss": 0.0765,
"step": 955
},
{
"epoch": 0.05502158273381295,
"grad_norm": 1.1509917974472046,
"learning_rate": 1.231165940486234e-07,
"loss": 0.0709,
"step": 956
},
{
"epoch": 0.055079136690647484,
"grad_norm": 1.6221928596496582,
"learning_rate": 1.1771618553447217e-07,
"loss": 0.0934,
"step": 957
},
{
"epoch": 0.055136690647482015,
"grad_norm": 1.4992139339447021,
"learning_rate": 1.1243618952994195e-07,
"loss": 0.0649,
"step": 958
},
{
"epoch": 0.055194244604316545,
"grad_norm": 19.954071044921875,
"learning_rate": 1.0727667037011668e-07,
"loss": 0.4697,
"step": 959
},
{
"epoch": 0.05525179856115108,
"grad_norm": 1.0766866207122803,
"learning_rate": 1.0223769092211012e-07,
"loss": 0.0877,
"step": 960
},
{
"epoch": 0.05530935251798561,
"grad_norm": 5.830136299133301,
"learning_rate": 9.731931258429638e-08,
"loss": 0.1061,
"step": 961
},
{
"epoch": 0.05536690647482014,
"grad_norm": 0.7689708471298218,
"learning_rate": 9.252159528556404e-08,
"loss": 0.0681,
"step": 962
},
{
"epoch": 0.05542446043165468,
"grad_norm": 2.8470983505249023,
"learning_rate": 8.784459748458318e-08,
"loss": 0.1102,
"step": 963
},
{
"epoch": 0.05548201438848921,
"grad_norm": 13.154084205627441,
"learning_rate": 8.328837616909612e-08,
"loss": 0.3053,
"step": 964
},
{
"epoch": 0.05553956834532374,
"grad_norm": 3.423447370529175,
"learning_rate": 7.885298685522235e-08,
"loss": 0.071,
"step": 965
},
{
"epoch": 0.055597122302158276,
"grad_norm": 0.8389829397201538,
"learning_rate": 7.453848358678018e-08,
"loss": 0.0703,
"step": 966
},
{
"epoch": 0.055654676258992807,
"grad_norm": 1.1691104173660278,
"learning_rate": 7.034491893463059e-08,
"loss": 0.0365,
"step": 967
},
{
"epoch": 0.05571223021582734,
"grad_norm": 0.6704636216163635,
"learning_rate": 6.627234399603554e-08,
"loss": 0.0514,
"step": 968
},
{
"epoch": 0.055769784172661874,
"grad_norm": 1.450444221496582,
"learning_rate": 6.232080839403631e-08,
"loss": 0.0943,
"step": 969
},
{
"epoch": 0.055827338129496404,
"grad_norm": 1.0474416017532349,
"learning_rate": 5.849036027684607e-08,
"loss": 0.1006,
"step": 970
},
{
"epoch": 0.055884892086330934,
"grad_norm": 0.4784161150455475,
"learning_rate": 5.4781046317267103e-08,
"loss": 0.0408,
"step": 971
},
{
"epoch": 0.05594244604316547,
"grad_norm": 16.443510055541992,
"learning_rate": 5.119291171211793e-08,
"loss": 0.1611,
"step": 972
},
{
"epoch": 0.056,
"grad_norm": 0.6649782657623291,
"learning_rate": 4.772600018168816e-08,
"loss": 0.0723,
"step": 973
},
{
"epoch": 0.05605755395683453,
"grad_norm": 11.152069091796875,
"learning_rate": 4.438035396920004e-08,
"loss": 0.7074,
"step": 974
},
{
"epoch": 0.05611510791366906,
"grad_norm": 0.5316632390022278,
"learning_rate": 4.115601384029666e-08,
"loss": 0.0642,
"step": 975
},
{
"epoch": 0.0561726618705036,
"grad_norm": 2.629650354385376,
"learning_rate": 3.805301908254455e-08,
"loss": 0.1437,
"step": 976
},
{
"epoch": 0.05623021582733813,
"grad_norm": 1.2182719707489014,
"learning_rate": 3.50714075049563e-08,
"loss": 0.094,
"step": 977
},
{
"epoch": 0.05628776978417266,
"grad_norm": 1.433416485786438,
"learning_rate": 3.22112154375287e-08,
"loss": 0.1248,
"step": 978
},
{
"epoch": 0.056345323741007196,
"grad_norm": 0.4791356027126312,
"learning_rate": 2.947247773079753e-08,
"loss": 0.0573,
"step": 979
},
{
"epoch": 0.056402877697841726,
"grad_norm": 0.43710970878601074,
"learning_rate": 2.6855227755419046e-08,
"loss": 0.0632,
"step": 980
},
{
"epoch": 0.056460431654676256,
"grad_norm": 1.2036529779434204,
"learning_rate": 2.4359497401758026e-08,
"loss": 0.0758,
"step": 981
},
{
"epoch": 0.05651798561151079,
"grad_norm": 1.0879039764404297,
"learning_rate": 2.1985317079500358e-08,
"loss": 0.0864,
"step": 982
},
{
"epoch": 0.05657553956834532,
"grad_norm": 2.7998852729797363,
"learning_rate": 1.973271571728441e-08,
"loss": 0.0904,
"step": 983
},
{
"epoch": 0.05663309352517985,
"grad_norm": 0.8842798471450806,
"learning_rate": 1.7601720762346895e-08,
"loss": 0.072,
"step": 984
},
{
"epoch": 0.05669064748201439,
"grad_norm": 30.26481819152832,
"learning_rate": 1.5592358180189782e-08,
"loss": 0.3083,
"step": 985
},
{
"epoch": 0.05674820143884892,
"grad_norm": 1.791191577911377,
"learning_rate": 1.370465245426167e-08,
"loss": 0.0949,
"step": 986
},
{
"epoch": 0.05680575539568345,
"grad_norm": 1.102792501449585,
"learning_rate": 1.1938626585660252e-08,
"loss": 0.078,
"step": 987
},
{
"epoch": 0.05686330935251799,
"grad_norm": 1.305518627166748,
"learning_rate": 1.0294302092853647e-08,
"loss": 0.1071,
"step": 988
},
{
"epoch": 0.05692086330935252,
"grad_norm": 1.006156325340271,
"learning_rate": 8.771699011416169e-09,
"loss": 0.0736,
"step": 989
},
{
"epoch": 0.05697841726618705,
"grad_norm": 0.8825615048408508,
"learning_rate": 7.370835893788508e-09,
"loss": 0.0776,
"step": 990
},
{
"epoch": 0.057035971223021585,
"grad_norm": 1.2495946884155273,
"learning_rate": 6.091729809042379e-09,
"loss": 0.0883,
"step": 991
},
{
"epoch": 0.057093525179856115,
"grad_norm": 4.397529125213623,
"learning_rate": 4.9343963426840006e-09,
"loss": 0.1277,
"step": 992
},
{
"epoch": 0.057151079136690645,
"grad_norm": 20.961580276489258,
"learning_rate": 3.898849596456477e-09,
"loss": 0.3655,
"step": 993
},
{
"epoch": 0.05720863309352518,
"grad_norm": 0.7759275436401367,
"learning_rate": 2.9851021881688314e-09,
"loss": 0.0783,
"step": 994
},
{
"epoch": 0.05726618705035971,
"grad_norm": 0.40887534618377686,
"learning_rate": 2.193165251545004e-09,
"loss": 0.059,
"step": 995
},
{
"epoch": 0.05732374100719424,
"grad_norm": 1.2675024271011353,
"learning_rate": 1.5230484360873043e-09,
"loss": 0.0846,
"step": 996
},
{
"epoch": 0.05738129496402878,
"grad_norm": 0.7930789589881897,
"learning_rate": 9.74759906957612e-10,
"loss": 0.0724,
"step": 997
},
{
"epoch": 0.05743884892086331,
"grad_norm": 13.144028663635254,
"learning_rate": 5.483063448785686e-10,
"loss": 0.2582,
"step": 998
},
{
"epoch": 0.05749640287769784,
"grad_norm": 2.8434152603149414,
"learning_rate": 2.436929460525317e-10,
"loss": 0.0955,
"step": 999
},
{
"epoch": 0.05755395683453238,
"grad_norm": 0.9370110630989075,
"learning_rate": 6.092342209607083e-11,
"loss": 0.0814,
"step": 1000
}
],
"logging_steps": 1,
"max_steps": 1000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}