eb_alfred_baseline_action_only / trainer_state.json
Hanyang81's picture
Upload trainer_state.json with huggingface_hub
036defa verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 1106,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0018083182640144665,
"grad_norm": 374.9632568359375,
"learning_rate": 1.7857142857142858e-07,
"loss": 8.897,
"step": 1
},
{
"epoch": 0.003616636528028933,
"grad_norm": 384.3598327636719,
"learning_rate": 3.5714285714285716e-07,
"loss": 8.9314,
"step": 2
},
{
"epoch": 0.0054249547920434,
"grad_norm": 379.2837829589844,
"learning_rate": 5.357142857142857e-07,
"loss": 8.8808,
"step": 3
},
{
"epoch": 0.007233273056057866,
"grad_norm": 373.4541320800781,
"learning_rate": 7.142857142857143e-07,
"loss": 8.8473,
"step": 4
},
{
"epoch": 0.009041591320072333,
"grad_norm": 382.3844909667969,
"learning_rate": 8.928571428571429e-07,
"loss": 8.8092,
"step": 5
},
{
"epoch": 0.0108499095840868,
"grad_norm": 397.5791320800781,
"learning_rate": 1.0714285714285714e-06,
"loss": 8.9407,
"step": 6
},
{
"epoch": 0.012658227848101266,
"grad_norm": 384.0470275878906,
"learning_rate": 1.25e-06,
"loss": 8.6481,
"step": 7
},
{
"epoch": 0.014466546112115732,
"grad_norm": 378.48675537109375,
"learning_rate": 1.4285714285714286e-06,
"loss": 8.7333,
"step": 8
},
{
"epoch": 0.0162748643761302,
"grad_norm": 328.6068115234375,
"learning_rate": 1.6071428571428574e-06,
"loss": 8.1644,
"step": 9
},
{
"epoch": 0.018083182640144666,
"grad_norm": 326.6074523925781,
"learning_rate": 1.7857142857142859e-06,
"loss": 7.943,
"step": 10
},
{
"epoch": 0.019891500904159132,
"grad_norm": 221.39767456054688,
"learning_rate": 1.9642857142857144e-06,
"loss": 6.6854,
"step": 11
},
{
"epoch": 0.0216998191681736,
"grad_norm": 206.39569091796875,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.5808,
"step": 12
},
{
"epoch": 0.023508137432188065,
"grad_norm": 190.39564514160156,
"learning_rate": 2.321428571428572e-06,
"loss": 6.35,
"step": 13
},
{
"epoch": 0.02531645569620253,
"grad_norm": 133.76596069335938,
"learning_rate": 2.5e-06,
"loss": 5.9181,
"step": 14
},
{
"epoch": 0.027124773960216998,
"grad_norm": 62.421669006347656,
"learning_rate": 2.6785714285714285e-06,
"loss": 5.4189,
"step": 15
},
{
"epoch": 0.028933092224231464,
"grad_norm": 174.518798828125,
"learning_rate": 2.8571428571428573e-06,
"loss": 4.9992,
"step": 16
},
{
"epoch": 0.03074141048824593,
"grad_norm": 60.2504997253418,
"learning_rate": 3.0357142857142856e-06,
"loss": 4.9165,
"step": 17
},
{
"epoch": 0.0325497287522604,
"grad_norm": 46.278438568115234,
"learning_rate": 3.2142857142857147e-06,
"loss": 4.8263,
"step": 18
},
{
"epoch": 0.034358047016274866,
"grad_norm": 40.315025329589844,
"learning_rate": 3.3928571428571435e-06,
"loss": 4.6813,
"step": 19
},
{
"epoch": 0.03616636528028933,
"grad_norm": 32.01236343383789,
"learning_rate": 3.5714285714285718e-06,
"loss": 4.4115,
"step": 20
},
{
"epoch": 0.0379746835443038,
"grad_norm": 27.191560745239258,
"learning_rate": 3.7500000000000005e-06,
"loss": 4.1287,
"step": 21
},
{
"epoch": 0.039783001808318265,
"grad_norm": 24.12420082092285,
"learning_rate": 3.928571428571429e-06,
"loss": 4.1339,
"step": 22
},
{
"epoch": 0.04159132007233273,
"grad_norm": 21.648895263671875,
"learning_rate": 4.107142857142857e-06,
"loss": 3.8526,
"step": 23
},
{
"epoch": 0.0433996383363472,
"grad_norm": 21.065244674682617,
"learning_rate": 4.2857142857142855e-06,
"loss": 3.8854,
"step": 24
},
{
"epoch": 0.045207956600361664,
"grad_norm": 20.010845184326172,
"learning_rate": 4.464285714285715e-06,
"loss": 3.875,
"step": 25
},
{
"epoch": 0.04701627486437613,
"grad_norm": 20.262365341186523,
"learning_rate": 4.642857142857144e-06,
"loss": 3.6755,
"step": 26
},
{
"epoch": 0.048824593128390596,
"grad_norm": 19.531538009643555,
"learning_rate": 4.821428571428572e-06,
"loss": 3.6935,
"step": 27
},
{
"epoch": 0.05063291139240506,
"grad_norm": 18.67901039123535,
"learning_rate": 5e-06,
"loss": 3.5598,
"step": 28
},
{
"epoch": 0.05244122965641953,
"grad_norm": 17.71997833251953,
"learning_rate": 5.1785714285714296e-06,
"loss": 3.5288,
"step": 29
},
{
"epoch": 0.054249547920433995,
"grad_norm": 18.615488052368164,
"learning_rate": 5.357142857142857e-06,
"loss": 3.561,
"step": 30
},
{
"epoch": 0.05605786618444846,
"grad_norm": 17.184545516967773,
"learning_rate": 5.535714285714286e-06,
"loss": 3.4718,
"step": 31
},
{
"epoch": 0.05786618444846293,
"grad_norm": 14.041234016418457,
"learning_rate": 5.7142857142857145e-06,
"loss": 3.2861,
"step": 32
},
{
"epoch": 0.059674502712477394,
"grad_norm": 12.927190780639648,
"learning_rate": 5.892857142857144e-06,
"loss": 3.1855,
"step": 33
},
{
"epoch": 0.06148282097649186,
"grad_norm": 13.494967460632324,
"learning_rate": 6.071428571428571e-06,
"loss": 3.1813,
"step": 34
},
{
"epoch": 0.06329113924050633,
"grad_norm": 20.254974365234375,
"learning_rate": 6.25e-06,
"loss": 3.1447,
"step": 35
},
{
"epoch": 0.0650994575045208,
"grad_norm": 11.912328720092773,
"learning_rate": 6.4285714285714295e-06,
"loss": 3.1702,
"step": 36
},
{
"epoch": 0.06690777576853527,
"grad_norm": 10.546191215515137,
"learning_rate": 6.607142857142858e-06,
"loss": 3.0884,
"step": 37
},
{
"epoch": 0.06871609403254973,
"grad_norm": 10.353287696838379,
"learning_rate": 6.785714285714287e-06,
"loss": 3.0961,
"step": 38
},
{
"epoch": 0.0705244122965642,
"grad_norm": 11.047510147094727,
"learning_rate": 6.964285714285714e-06,
"loss": 3.1163,
"step": 39
},
{
"epoch": 0.07233273056057866,
"grad_norm": 9.003999710083008,
"learning_rate": 7.1428571428571436e-06,
"loss": 3.0804,
"step": 40
},
{
"epoch": 0.07414104882459313,
"grad_norm": 14.103655815124512,
"learning_rate": 7.321428571428572e-06,
"loss": 3.0605,
"step": 41
},
{
"epoch": 0.0759493670886076,
"grad_norm": 14.631102561950684,
"learning_rate": 7.500000000000001e-06,
"loss": 3.0604,
"step": 42
},
{
"epoch": 0.07775768535262206,
"grad_norm": 9.40045166015625,
"learning_rate": 7.67857142857143e-06,
"loss": 3.0951,
"step": 43
},
{
"epoch": 0.07956600361663653,
"grad_norm": 8.937652587890625,
"learning_rate": 7.857142857142858e-06,
"loss": 3.0957,
"step": 44
},
{
"epoch": 0.081374321880651,
"grad_norm": 8.327485084533691,
"learning_rate": 8.035714285714286e-06,
"loss": 3.0625,
"step": 45
},
{
"epoch": 0.08318264014466546,
"grad_norm": 9.024964332580566,
"learning_rate": 8.214285714285714e-06,
"loss": 2.9555,
"step": 46
},
{
"epoch": 0.08499095840867993,
"grad_norm": 14.808930397033691,
"learning_rate": 8.392857142857144e-06,
"loss": 3.0339,
"step": 47
},
{
"epoch": 0.0867992766726944,
"grad_norm": 8.41634750366211,
"learning_rate": 8.571428571428571e-06,
"loss": 2.966,
"step": 48
},
{
"epoch": 0.08860759493670886,
"grad_norm": 8.490717887878418,
"learning_rate": 8.750000000000001e-06,
"loss": 2.9571,
"step": 49
},
{
"epoch": 0.09041591320072333,
"grad_norm": 8.499176025390625,
"learning_rate": 8.92857142857143e-06,
"loss": 2.8819,
"step": 50
},
{
"epoch": 0.0922242314647378,
"grad_norm": 15.260178565979004,
"learning_rate": 9.107142857142858e-06,
"loss": 2.9966,
"step": 51
},
{
"epoch": 0.09403254972875226,
"grad_norm": 17.47574234008789,
"learning_rate": 9.285714285714288e-06,
"loss": 2.9552,
"step": 52
},
{
"epoch": 0.09584086799276673,
"grad_norm": 15.631826400756836,
"learning_rate": 9.464285714285714e-06,
"loss": 2.9249,
"step": 53
},
{
"epoch": 0.09764918625678119,
"grad_norm": 14.44354248046875,
"learning_rate": 9.642857142857144e-06,
"loss": 2.893,
"step": 54
},
{
"epoch": 0.09945750452079566,
"grad_norm": 12.236231803894043,
"learning_rate": 9.821428571428573e-06,
"loss": 3.0185,
"step": 55
},
{
"epoch": 0.10126582278481013,
"grad_norm": 10.945856094360352,
"learning_rate": 1e-05,
"loss": 2.9784,
"step": 56
},
{
"epoch": 0.10307414104882459,
"grad_norm": 10.860036849975586,
"learning_rate": 9.999977619961366e-06,
"loss": 2.8739,
"step": 57
},
{
"epoch": 0.10488245931283906,
"grad_norm": 11.251140594482422,
"learning_rate": 9.999910480045805e-06,
"loss": 2.8649,
"step": 58
},
{
"epoch": 0.10669077757685352,
"grad_norm": 11.549039840698242,
"learning_rate": 9.999798580854356e-06,
"loss": 2.8204,
"step": 59
},
{
"epoch": 0.10849909584086799,
"grad_norm": 14.085471153259277,
"learning_rate": 9.999641923388745e-06,
"loss": 2.8975,
"step": 60
},
{
"epoch": 0.11030741410488246,
"grad_norm": 12.601762771606445,
"learning_rate": 9.999440509051367e-06,
"loss": 2.789,
"step": 61
},
{
"epoch": 0.11211573236889692,
"grad_norm": 13.548413276672363,
"learning_rate": 9.999194339645292e-06,
"loss": 2.8601,
"step": 62
},
{
"epoch": 0.11392405063291139,
"grad_norm": 14.29236125946045,
"learning_rate": 9.998903417374228e-06,
"loss": 2.8428,
"step": 63
},
{
"epoch": 0.11573236889692586,
"grad_norm": 14.931453704833984,
"learning_rate": 9.998567744842518e-06,
"loss": 2.8159,
"step": 64
},
{
"epoch": 0.11754068716094032,
"grad_norm": 15.400191307067871,
"learning_rate": 9.998187325055107e-06,
"loss": 2.7035,
"step": 65
},
{
"epoch": 0.11934900542495479,
"grad_norm": 16.511560440063477,
"learning_rate": 9.997762161417517e-06,
"loss": 2.7915,
"step": 66
},
{
"epoch": 0.12115732368896925,
"grad_norm": 16.48167610168457,
"learning_rate": 9.997292257735822e-06,
"loss": 2.6187,
"step": 67
},
{
"epoch": 0.12296564195298372,
"grad_norm": 17.774396896362305,
"learning_rate": 9.996777618216608e-06,
"loss": 2.7332,
"step": 68
},
{
"epoch": 0.12477396021699819,
"grad_norm": 18.196462631225586,
"learning_rate": 9.996218247466932e-06,
"loss": 2.6149,
"step": 69
},
{
"epoch": 0.12658227848101267,
"grad_norm": 19.114198684692383,
"learning_rate": 9.995614150494293e-06,
"loss": 2.6487,
"step": 70
},
{
"epoch": 0.12839059674502712,
"grad_norm": 19.922454833984375,
"learning_rate": 9.994965332706574e-06,
"loss": 2.5652,
"step": 71
},
{
"epoch": 0.1301989150090416,
"grad_norm": 20.709423065185547,
"learning_rate": 9.994271799912004e-06,
"loss": 2.5855,
"step": 72
},
{
"epoch": 0.13200723327305605,
"grad_norm": 22.094491958618164,
"learning_rate": 9.993533558319098e-06,
"loss": 2.6111,
"step": 73
},
{
"epoch": 0.13381555153707053,
"grad_norm": 21.925947189331055,
"learning_rate": 9.992750614536606e-06,
"loss": 2.4331,
"step": 74
},
{
"epoch": 0.13562386980108498,
"grad_norm": 23.11043357849121,
"learning_rate": 9.991922975573453e-06,
"loss": 2.4456,
"step": 75
},
{
"epoch": 0.13743218806509946,
"grad_norm": 23.701190948486328,
"learning_rate": 9.991050648838676e-06,
"loss": 2.4021,
"step": 76
},
{
"epoch": 0.13924050632911392,
"grad_norm": 24.970277786254883,
"learning_rate": 9.990133642141359e-06,
"loss": 2.4236,
"step": 77
},
{
"epoch": 0.1410488245931284,
"grad_norm": 25.31788444519043,
"learning_rate": 9.989171963690556e-06,
"loss": 2.3709,
"step": 78
},
{
"epoch": 0.14285714285714285,
"grad_norm": 26.034448623657227,
"learning_rate": 9.988165622095233e-06,
"loss": 2.3826,
"step": 79
},
{
"epoch": 0.14466546112115733,
"grad_norm": 25.721010208129883,
"learning_rate": 9.987114626364172e-06,
"loss": 2.2357,
"step": 80
},
{
"epoch": 0.14647377938517178,
"grad_norm": 25.75619888305664,
"learning_rate": 9.986018985905901e-06,
"loss": 2.2227,
"step": 81
},
{
"epoch": 0.14828209764918626,
"grad_norm": 25.732624053955078,
"learning_rate": 9.984878710528615e-06,
"loss": 2.1467,
"step": 82
},
{
"epoch": 0.15009041591320071,
"grad_norm": 25.783708572387695,
"learning_rate": 9.983693810440073e-06,
"loss": 2.1177,
"step": 83
},
{
"epoch": 0.1518987341772152,
"grad_norm": 25.99880599975586,
"learning_rate": 9.982464296247523e-06,
"loss": 2.1211,
"step": 84
},
{
"epoch": 0.15370705244122965,
"grad_norm": 25.63880157470703,
"learning_rate": 9.98119017895759e-06,
"loss": 2.0502,
"step": 85
},
{
"epoch": 0.15551537070524413,
"grad_norm": 25.79373550415039,
"learning_rate": 9.979871469976197e-06,
"loss": 2.0514,
"step": 86
},
{
"epoch": 0.15732368896925858,
"grad_norm": 25.356674194335938,
"learning_rate": 9.978508181108442e-06,
"loss": 1.9652,
"step": 87
},
{
"epoch": 0.15913200723327306,
"grad_norm": 24.95633316040039,
"learning_rate": 9.97710032455851e-06,
"loss": 1.9228,
"step": 88
},
{
"epoch": 0.1609403254972875,
"grad_norm": 24.74396324157715,
"learning_rate": 9.975647912929558e-06,
"loss": 1.9094,
"step": 89
},
{
"epoch": 0.162748643761302,
"grad_norm": 24.885238647460938,
"learning_rate": 9.974150959223591e-06,
"loss": 1.8552,
"step": 90
},
{
"epoch": 0.16455696202531644,
"grad_norm": 24.835893630981445,
"learning_rate": 9.972609476841368e-06,
"loss": 1.8147,
"step": 91
},
{
"epoch": 0.16636528028933092,
"grad_norm": 24.22718620300293,
"learning_rate": 9.971023479582258e-06,
"loss": 1.726,
"step": 92
},
{
"epoch": 0.16817359855334538,
"grad_norm": 24.750812530517578,
"learning_rate": 9.969392981644138e-06,
"loss": 1.7396,
"step": 93
},
{
"epoch": 0.16998191681735986,
"grad_norm": 24.80465316772461,
"learning_rate": 9.967717997623245e-06,
"loss": 1.716,
"step": 94
},
{
"epoch": 0.1717902350813743,
"grad_norm": 24.216758728027344,
"learning_rate": 9.965998542514066e-06,
"loss": 1.6297,
"step": 95
},
{
"epoch": 0.1735985533453888,
"grad_norm": 24.366065979003906,
"learning_rate": 9.964234631709188e-06,
"loss": 1.6014,
"step": 96
},
{
"epoch": 0.17540687160940324,
"grad_norm": 24.15585708618164,
"learning_rate": 9.962426280999168e-06,
"loss": 1.5522,
"step": 97
},
{
"epoch": 0.17721518987341772,
"grad_norm": 24.19598388671875,
"learning_rate": 9.960573506572391e-06,
"loss": 1.5223,
"step": 98
},
{
"epoch": 0.17902350813743217,
"grad_norm": 24.109111785888672,
"learning_rate": 9.95867632501492e-06,
"loss": 1.5041,
"step": 99
},
{
"epoch": 0.18083182640144665,
"grad_norm": 23.227954864501953,
"learning_rate": 9.956734753310355e-06,
"loss": 1.3882,
"step": 100
},
{
"epoch": 0.18264014466546113,
"grad_norm": 23.608726501464844,
"learning_rate": 9.954748808839675e-06,
"loss": 1.3996,
"step": 101
},
{
"epoch": 0.1844484629294756,
"grad_norm": 23.02730941772461,
"learning_rate": 9.952718509381086e-06,
"loss": 1.3708,
"step": 102
},
{
"epoch": 0.18625678119349007,
"grad_norm": 23.250051498413086,
"learning_rate": 9.950643873109861e-06,
"loss": 1.3515,
"step": 103
},
{
"epoch": 0.18806509945750452,
"grad_norm": 23.402767181396484,
"learning_rate": 9.948524918598175e-06,
"loss": 1.315,
"step": 104
},
{
"epoch": 0.189873417721519,
"grad_norm": 23.3504581451416,
"learning_rate": 9.946361664814942e-06,
"loss": 1.3215,
"step": 105
},
{
"epoch": 0.19168173598553345,
"grad_norm": 22.480173110961914,
"learning_rate": 9.944154131125643e-06,
"loss": 1.2024,
"step": 106
},
{
"epoch": 0.19349005424954793,
"grad_norm": 22.358362197875977,
"learning_rate": 9.941902337292156e-06,
"loss": 1.1616,
"step": 107
},
{
"epoch": 0.19529837251356238,
"grad_norm": 22.022979736328125,
"learning_rate": 9.93960630347257e-06,
"loss": 1.111,
"step": 108
},
{
"epoch": 0.19710669077757687,
"grad_norm": 21.303504943847656,
"learning_rate": 9.937266050221015e-06,
"loss": 1.0742,
"step": 109
},
{
"epoch": 0.19891500904159132,
"grad_norm": 21.511245727539062,
"learning_rate": 9.934881598487478e-06,
"loss": 1.0695,
"step": 110
},
{
"epoch": 0.2007233273056058,
"grad_norm": 21.223609924316406,
"learning_rate": 9.932452969617607e-06,
"loss": 0.9994,
"step": 111
},
{
"epoch": 0.20253164556962025,
"grad_norm": 20.472251892089844,
"learning_rate": 9.929980185352525e-06,
"loss": 0.9462,
"step": 112
},
{
"epoch": 0.20433996383363473,
"grad_norm": 20.204402923583984,
"learning_rate": 9.927463267828635e-06,
"loss": 0.9217,
"step": 113
},
{
"epoch": 0.20614828209764918,
"grad_norm": 19.72013282775879,
"learning_rate": 9.924902239577419e-06,
"loss": 0.8955,
"step": 114
},
{
"epoch": 0.20795660036166366,
"grad_norm": 19.025146484375,
"learning_rate": 9.922297123525244e-06,
"loss": 0.811,
"step": 115
},
{
"epoch": 0.20976491862567812,
"grad_norm": 18.705778121948242,
"learning_rate": 9.91964794299315e-06,
"loss": 0.8193,
"step": 116
},
{
"epoch": 0.2115732368896926,
"grad_norm": 17.645185470581055,
"learning_rate": 9.91695472169664e-06,
"loss": 0.76,
"step": 117
},
{
"epoch": 0.21338155515370705,
"grad_norm": 16.864002227783203,
"learning_rate": 9.914217483745472e-06,
"loss": 0.7138,
"step": 118
},
{
"epoch": 0.21518987341772153,
"grad_norm": 16.277565002441406,
"learning_rate": 9.911436253643445e-06,
"loss": 0.6984,
"step": 119
},
{
"epoch": 0.21699819168173598,
"grad_norm": 15.373967170715332,
"learning_rate": 9.90861105628817e-06,
"loss": 0.6673,
"step": 120
},
{
"epoch": 0.21880650994575046,
"grad_norm": 14.397501945495605,
"learning_rate": 9.905741916970863e-06,
"loss": 0.6259,
"step": 121
},
{
"epoch": 0.2206148282097649,
"grad_norm": 13.482257843017578,
"learning_rate": 9.902828861376101e-06,
"loss": 0.5967,
"step": 122
},
{
"epoch": 0.2224231464737794,
"grad_norm": 12.610859870910645,
"learning_rate": 9.8998719155816e-06,
"loss": 0.6098,
"step": 123
},
{
"epoch": 0.22423146473779385,
"grad_norm": 11.894183158874512,
"learning_rate": 9.896871106057989e-06,
"loss": 0.5804,
"step": 124
},
{
"epoch": 0.22603978300180833,
"grad_norm": 11.055013656616211,
"learning_rate": 9.89382645966856e-06,
"loss": 0.5729,
"step": 125
},
{
"epoch": 0.22784810126582278,
"grad_norm": 10.054003715515137,
"learning_rate": 9.890738003669029e-06,
"loss": 0.5485,
"step": 126
},
{
"epoch": 0.22965641952983726,
"grad_norm": 9.11440658569336,
"learning_rate": 9.887605765707309e-06,
"loss": 0.4833,
"step": 127
},
{
"epoch": 0.2314647377938517,
"grad_norm": 8.774530410766602,
"learning_rate": 9.884429773823238e-06,
"loss": 0.5453,
"step": 128
},
{
"epoch": 0.2332730560578662,
"grad_norm": 7.480116844177246,
"learning_rate": 9.88121005644835e-06,
"loss": 0.4593,
"step": 129
},
{
"epoch": 0.23508137432188064,
"grad_norm": 7.175760746002197,
"learning_rate": 9.877946642405598e-06,
"loss": 0.495,
"step": 130
},
{
"epoch": 0.23688969258589512,
"grad_norm": 6.368906021118164,
"learning_rate": 9.874639560909118e-06,
"loss": 0.4352,
"step": 131
},
{
"epoch": 0.23869801084990958,
"grad_norm": 5.852133274078369,
"learning_rate": 9.871288841563956e-06,
"loss": 0.4482,
"step": 132
},
{
"epoch": 0.24050632911392406,
"grad_norm": 6.125861167907715,
"learning_rate": 9.867894514365802e-06,
"loss": 0.4516,
"step": 133
},
{
"epoch": 0.2423146473779385,
"grad_norm": 8.970146179199219,
"learning_rate": 9.864456609700726e-06,
"loss": 0.4457,
"step": 134
},
{
"epoch": 0.244122965641953,
"grad_norm": 5.39919900894165,
"learning_rate": 9.860975158344902e-06,
"loss": 0.4423,
"step": 135
},
{
"epoch": 0.24593128390596744,
"grad_norm": 7.831792831420898,
"learning_rate": 9.857450191464337e-06,
"loss": 0.4337,
"step": 136
},
{
"epoch": 0.24773960216998192,
"grad_norm": 5.569422721862793,
"learning_rate": 9.853881740614591e-06,
"loss": 0.4486,
"step": 137
},
{
"epoch": 0.24954792043399637,
"grad_norm": 7.2985382080078125,
"learning_rate": 9.85026983774049e-06,
"loss": 0.4464,
"step": 138
},
{
"epoch": 0.2513562386980108,
"grad_norm": 5.201531887054443,
"learning_rate": 9.846614515175843e-06,
"loss": 0.3947,
"step": 139
},
{
"epoch": 0.25316455696202533,
"grad_norm": 5.368192672729492,
"learning_rate": 9.842915805643156e-06,
"loss": 0.4098,
"step": 140
},
{
"epoch": 0.2549728752260398,
"grad_norm": 5.139791011810303,
"learning_rate": 9.839173742253334e-06,
"loss": 0.3653,
"step": 141
},
{
"epoch": 0.25678119349005424,
"grad_norm": 5.5092244148254395,
"learning_rate": 9.835388358505383e-06,
"loss": 0.3791,
"step": 142
},
{
"epoch": 0.2585895117540687,
"grad_norm": 5.9908013343811035,
"learning_rate": 9.83155968828612e-06,
"loss": 0.3896,
"step": 143
},
{
"epoch": 0.2603978300180832,
"grad_norm": 6.346823215484619,
"learning_rate": 9.827687765869859e-06,
"loss": 0.3935,
"step": 144
},
{
"epoch": 0.26220614828209765,
"grad_norm": 8.417757987976074,
"learning_rate": 9.823772625918111e-06,
"loss": 0.3432,
"step": 145
},
{
"epoch": 0.2640144665461121,
"grad_norm": 6.771397113800049,
"learning_rate": 9.819814303479268e-06,
"loss": 0.3423,
"step": 146
},
{
"epoch": 0.26582278481012656,
"grad_norm": 7.216489315032959,
"learning_rate": 9.815812833988292e-06,
"loss": 0.3454,
"step": 147
},
{
"epoch": 0.26763110307414106,
"grad_norm": 7.83514928817749,
"learning_rate": 9.811768253266401e-06,
"loss": 0.3471,
"step": 148
},
{
"epoch": 0.2694394213381555,
"grad_norm": 7.734335899353027,
"learning_rate": 9.807680597520746e-06,
"loss": 0.3468,
"step": 149
},
{
"epoch": 0.27124773960216997,
"grad_norm": 8.087752342224121,
"learning_rate": 9.803549903344081e-06,
"loss": 0.3293,
"step": 150
},
{
"epoch": 0.2730560578661845,
"grad_norm": 8.151535034179688,
"learning_rate": 9.799376207714446e-06,
"loss": 0.3136,
"step": 151
},
{
"epoch": 0.27486437613019893,
"grad_norm": 8.482094764709473,
"learning_rate": 9.79515954799483e-06,
"loss": 0.3048,
"step": 152
},
{
"epoch": 0.2766726943942134,
"grad_norm": 8.386483192443848,
"learning_rate": 9.790899961932833e-06,
"loss": 0.3284,
"step": 153
},
{
"epoch": 0.27848101265822783,
"grad_norm": 8.386569023132324,
"learning_rate": 9.786597487660336e-06,
"loss": 0.2995,
"step": 154
},
{
"epoch": 0.28028933092224234,
"grad_norm": 8.845312118530273,
"learning_rate": 9.782252163693159e-06,
"loss": 0.2945,
"step": 155
},
{
"epoch": 0.2820976491862568,
"grad_norm": 8.939291954040527,
"learning_rate": 9.777864028930705e-06,
"loss": 0.3326,
"step": 156
},
{
"epoch": 0.28390596745027125,
"grad_norm": 8.173895835876465,
"learning_rate": 9.773433122655625e-06,
"loss": 0.2409,
"step": 157
},
{
"epoch": 0.2857142857142857,
"grad_norm": 8.183178901672363,
"learning_rate": 9.768959484533461e-06,
"loss": 0.2606,
"step": 158
},
{
"epoch": 0.2875226039783002,
"grad_norm": 8.397600173950195,
"learning_rate": 9.76444315461229e-06,
"loss": 0.2588,
"step": 159
},
{
"epoch": 0.28933092224231466,
"grad_norm": 8.846126556396484,
"learning_rate": 9.75988417332237e-06,
"loss": 0.2658,
"step": 160
},
{
"epoch": 0.2911392405063291,
"grad_norm": 7.795368194580078,
"learning_rate": 9.755282581475769e-06,
"loss": 0.2461,
"step": 161
},
{
"epoch": 0.29294755877034356,
"grad_norm": 7.818116664886475,
"learning_rate": 9.750638420266008e-06,
"loss": 0.2272,
"step": 162
},
{
"epoch": 0.29475587703435807,
"grad_norm": 7.697627544403076,
"learning_rate": 9.745951731267693e-06,
"loss": 0.2062,
"step": 163
},
{
"epoch": 0.2965641952983725,
"grad_norm": 7.596677303314209,
"learning_rate": 9.741222556436132e-06,
"loss": 0.2031,
"step": 164
},
{
"epoch": 0.298372513562387,
"grad_norm": 7.392519474029541,
"learning_rate": 9.736450938106976e-06,
"loss": 0.2077,
"step": 165
},
{
"epoch": 0.30018083182640143,
"grad_norm": 7.342133522033691,
"learning_rate": 9.731636918995821e-06,
"loss": 0.2019,
"step": 166
},
{
"epoch": 0.30198915009041594,
"grad_norm": 6.93854284286499,
"learning_rate": 9.726780542197845e-06,
"loss": 0.2101,
"step": 167
},
{
"epoch": 0.3037974683544304,
"grad_norm": 6.741715908050537,
"learning_rate": 9.721881851187406e-06,
"loss": 0.1662,
"step": 168
},
{
"epoch": 0.30560578661844484,
"grad_norm": 6.388901233673096,
"learning_rate": 9.716940889817662e-06,
"loss": 0.1538,
"step": 169
},
{
"epoch": 0.3074141048824593,
"grad_norm": 6.611469745635986,
"learning_rate": 9.711957702320176e-06,
"loss": 0.1847,
"step": 170
},
{
"epoch": 0.3092224231464738,
"grad_norm": 8.091357231140137,
"learning_rate": 9.706932333304518e-06,
"loss": 0.1753,
"step": 171
},
{
"epoch": 0.31103074141048825,
"grad_norm": 5.52731990814209,
"learning_rate": 9.701864827757868e-06,
"loss": 0.1137,
"step": 172
},
{
"epoch": 0.3128390596745027,
"grad_norm": 6.323424339294434,
"learning_rate": 9.696755231044618e-06,
"loss": 0.166,
"step": 173
},
{
"epoch": 0.31464737793851716,
"grad_norm": 5.795332908630371,
"learning_rate": 9.691603588905956e-06,
"loss": 0.1421,
"step": 174
},
{
"epoch": 0.31645569620253167,
"grad_norm": 5.6995978355407715,
"learning_rate": 9.68640994745946e-06,
"loss": 0.1484,
"step": 175
},
{
"epoch": 0.3182640144665461,
"grad_norm": 5.68313455581665,
"learning_rate": 9.681174353198687e-06,
"loss": 0.1372,
"step": 176
},
{
"epoch": 0.32007233273056057,
"grad_norm": 5.7383928298950195,
"learning_rate": 9.675896852992762e-06,
"loss": 0.1213,
"step": 177
},
{
"epoch": 0.321880650994575,
"grad_norm": 4.374049663543701,
"learning_rate": 9.670577494085945e-06,
"loss": 0.1064,
"step": 178
},
{
"epoch": 0.32368896925858953,
"grad_norm": 4.4554901123046875,
"learning_rate": 9.665216324097222e-06,
"loss": 0.1634,
"step": 179
},
{
"epoch": 0.325497287522604,
"grad_norm": 4.913774013519287,
"learning_rate": 9.659813391019867e-06,
"loss": 0.1237,
"step": 180
},
{
"epoch": 0.32730560578661844,
"grad_norm": 4.121984958648682,
"learning_rate": 9.654368743221022e-06,
"loss": 0.1291,
"step": 181
},
{
"epoch": 0.3291139240506329,
"grad_norm": 3.992001533508301,
"learning_rate": 9.648882429441258e-06,
"loss": 0.1014,
"step": 182
},
{
"epoch": 0.3309222423146474,
"grad_norm": 3.465792417526245,
"learning_rate": 9.643354498794139e-06,
"loss": 0.0807,
"step": 183
},
{
"epoch": 0.33273056057866185,
"grad_norm": 3.275437116622925,
"learning_rate": 9.637785000765789e-06,
"loss": 0.0791,
"step": 184
},
{
"epoch": 0.3345388788426763,
"grad_norm": 3.963881492614746,
"learning_rate": 9.632173985214438e-06,
"loss": 0.1265,
"step": 185
},
{
"epoch": 0.33634719710669075,
"grad_norm": 4.472430229187012,
"learning_rate": 9.626521502369984e-06,
"loss": 0.0948,
"step": 186
},
{
"epoch": 0.33815551537070526,
"grad_norm": 3.274829864501953,
"learning_rate": 9.620827602833542e-06,
"loss": 0.0971,
"step": 187
},
{
"epoch": 0.3399638336347197,
"grad_norm": 3.167621612548828,
"learning_rate": 9.615092337576987e-06,
"loss": 0.0999,
"step": 188
},
{
"epoch": 0.34177215189873417,
"grad_norm": 2.7695417404174805,
"learning_rate": 9.609315757942504e-06,
"loss": 0.1195,
"step": 189
},
{
"epoch": 0.3435804701627486,
"grad_norm": 2.588524103164673,
"learning_rate": 9.603497915642122e-06,
"loss": 0.0944,
"step": 190
},
{
"epoch": 0.3453887884267631,
"grad_norm": 2.3558359146118164,
"learning_rate": 9.597638862757255e-06,
"loss": 0.0695,
"step": 191
},
{
"epoch": 0.3471971066907776,
"grad_norm": 2.444850206375122,
"learning_rate": 9.591738651738235e-06,
"loss": 0.0905,
"step": 192
},
{
"epoch": 0.34900542495479203,
"grad_norm": 2.4169912338256836,
"learning_rate": 9.585797335403843e-06,
"loss": 0.0764,
"step": 193
},
{
"epoch": 0.3508137432188065,
"grad_norm": 2.25830078125,
"learning_rate": 9.579814966940833e-06,
"loss": 0.0571,
"step": 194
},
{
"epoch": 0.352622061482821,
"grad_norm": 2.6071224212646484,
"learning_rate": 9.573791599903463e-06,
"loss": 0.0578,
"step": 195
},
{
"epoch": 0.35443037974683544,
"grad_norm": 1.7793872356414795,
"learning_rate": 9.567727288213005e-06,
"loss": 0.047,
"step": 196
},
{
"epoch": 0.3562386980108499,
"grad_norm": 1.9689217805862427,
"learning_rate": 9.561622086157273e-06,
"loss": 0.047,
"step": 197
},
{
"epoch": 0.35804701627486435,
"grad_norm": 1.7441576719284058,
"learning_rate": 9.55547604839013e-06,
"loss": 0.063,
"step": 198
},
{
"epoch": 0.35985533453887886,
"grad_norm": 2.4873642921447754,
"learning_rate": 9.549289229930997e-06,
"loss": 0.0693,
"step": 199
},
{
"epoch": 0.3616636528028933,
"grad_norm": 1.4174587726593018,
"learning_rate": 9.543061686164374e-06,
"loss": 0.0424,
"step": 200
},
{
"epoch": 0.36347197106690776,
"grad_norm": 2.614192008972168,
"learning_rate": 9.536793472839325e-06,
"loss": 0.1087,
"step": 201
},
{
"epoch": 0.36528028933092227,
"grad_norm": 1.5075057744979858,
"learning_rate": 9.530484646068996e-06,
"loss": 0.056,
"step": 202
},
{
"epoch": 0.3670886075949367,
"grad_norm": 2.0895304679870605,
"learning_rate": 9.524135262330098e-06,
"loss": 0.0765,
"step": 203
},
{
"epoch": 0.3688969258589512,
"grad_norm": 4.0753397941589355,
"learning_rate": 9.517745378462417e-06,
"loss": 0.1199,
"step": 204
},
{
"epoch": 0.3707052441229656,
"grad_norm": 6.124849796295166,
"learning_rate": 9.511315051668287e-06,
"loss": 0.0626,
"step": 205
},
{
"epoch": 0.37251356238698013,
"grad_norm": 2.387824296951294,
"learning_rate": 9.504844339512096e-06,
"loss": 0.0949,
"step": 206
},
{
"epoch": 0.3743218806509946,
"grad_norm": 1.6911492347717285,
"learning_rate": 9.498333299919759e-06,
"loss": 0.0641,
"step": 207
},
{
"epoch": 0.37613019891500904,
"grad_norm": 1.5189487934112549,
"learning_rate": 9.491781991178203e-06,
"loss": 0.0585,
"step": 208
},
{
"epoch": 0.3779385171790235,
"grad_norm": 2.6246490478515625,
"learning_rate": 9.485190471934845e-06,
"loss": 0.0824,
"step": 209
},
{
"epoch": 0.379746835443038,
"grad_norm": 7.05039119720459,
"learning_rate": 9.478558801197065e-06,
"loss": 0.1881,
"step": 210
},
{
"epoch": 0.38155515370705245,
"grad_norm": 5.523131847381592,
"learning_rate": 9.471887038331686e-06,
"loss": 0.1386,
"step": 211
},
{
"epoch": 0.3833634719710669,
"grad_norm": 5.898447036743164,
"learning_rate": 9.465175243064428e-06,
"loss": 0.1115,
"step": 212
},
{
"epoch": 0.38517179023508136,
"grad_norm": 1.4749406576156616,
"learning_rate": 9.458423475479387e-06,
"loss": 0.0555,
"step": 213
},
{
"epoch": 0.38698010849909587,
"grad_norm": 1.5287400484085083,
"learning_rate": 9.451631796018495e-06,
"loss": 0.0552,
"step": 214
},
{
"epoch": 0.3887884267631103,
"grad_norm": 1.926813006401062,
"learning_rate": 9.444800265480968e-06,
"loss": 0.0618,
"step": 215
},
{
"epoch": 0.39059674502712477,
"grad_norm": 3.549189805984497,
"learning_rate": 9.437928945022772e-06,
"loss": 0.0595,
"step": 216
},
{
"epoch": 0.3924050632911392,
"grad_norm": 3.4709861278533936,
"learning_rate": 9.431017896156074e-06,
"loss": 0.0772,
"step": 217
},
{
"epoch": 0.39421338155515373,
"grad_norm": 1.84910249710083,
"learning_rate": 9.424067180748692e-06,
"loss": 0.0293,
"step": 218
},
{
"epoch": 0.3960216998191682,
"grad_norm": 1.111405849456787,
"learning_rate": 9.417076861023539e-06,
"loss": 0.0321,
"step": 219
},
{
"epoch": 0.39783001808318263,
"grad_norm": 1.045786738395691,
"learning_rate": 9.410046999558062e-06,
"loss": 0.0535,
"step": 220
},
{
"epoch": 0.3996383363471971,
"grad_norm": 2.857057809829712,
"learning_rate": 9.40297765928369e-06,
"loss": 0.0716,
"step": 221
},
{
"epoch": 0.4014466546112116,
"grad_norm": 3.469581365585327,
"learning_rate": 9.395868903485269e-06,
"loss": 0.0819,
"step": 222
},
{
"epoch": 0.40325497287522605,
"grad_norm": 4.196573257446289,
"learning_rate": 9.388720795800488e-06,
"loss": 0.1773,
"step": 223
},
{
"epoch": 0.4050632911392405,
"grad_norm": 3.5620219707489014,
"learning_rate": 9.381533400219319e-06,
"loss": 0.1018,
"step": 224
},
{
"epoch": 0.40687160940325495,
"grad_norm": 4.233828067779541,
"learning_rate": 9.374306781083437e-06,
"loss": 0.0745,
"step": 225
},
{
"epoch": 0.40867992766726946,
"grad_norm": 3.0582659244537354,
"learning_rate": 9.36704100308565e-06,
"loss": 0.0688,
"step": 226
},
{
"epoch": 0.4104882459312839,
"grad_norm": 3.3210911750793457,
"learning_rate": 9.359736131269312e-06,
"loss": 0.1041,
"step": 227
},
{
"epoch": 0.41229656419529837,
"grad_norm": 2.8717358112335205,
"learning_rate": 9.352392231027752e-06,
"loss": 0.0415,
"step": 228
},
{
"epoch": 0.4141048824593128,
"grad_norm": 1.4785810708999634,
"learning_rate": 9.345009368103677e-06,
"loss": 0.0705,
"step": 229
},
{
"epoch": 0.4159132007233273,
"grad_norm": 2.2191975116729736,
"learning_rate": 9.337587608588588e-06,
"loss": 0.0938,
"step": 230
},
{
"epoch": 0.4177215189873418,
"grad_norm": 1.842610478401184,
"learning_rate": 9.330127018922195e-06,
"loss": 0.0778,
"step": 231
},
{
"epoch": 0.41952983725135623,
"grad_norm": 1.9353058338165283,
"learning_rate": 9.322627665891807e-06,
"loss": 0.104,
"step": 232
},
{
"epoch": 0.4213381555153707,
"grad_norm": 0.9981403946876526,
"learning_rate": 9.315089616631752e-06,
"loss": 0.0183,
"step": 233
},
{
"epoch": 0.4231464737793852,
"grad_norm": 4.318927764892578,
"learning_rate": 9.307512938622762e-06,
"loss": 0.0639,
"step": 234
},
{
"epoch": 0.42495479204339964,
"grad_norm": 0.7978769540786743,
"learning_rate": 9.299897699691377e-06,
"loss": 0.0207,
"step": 235
},
{
"epoch": 0.4267631103074141,
"grad_norm": 1.1614809036254883,
"learning_rate": 9.292243968009332e-06,
"loss": 0.0591,
"step": 236
},
{
"epoch": 0.42857142857142855,
"grad_norm": 4.371250152587891,
"learning_rate": 9.28455181209295e-06,
"loss": 0.0769,
"step": 237
},
{
"epoch": 0.43037974683544306,
"grad_norm": 1.4584596157073975,
"learning_rate": 9.276821300802535e-06,
"loss": 0.0636,
"step": 238
},
{
"epoch": 0.4321880650994575,
"grad_norm": 1.551882028579712,
"learning_rate": 9.269052503341737e-06,
"loss": 0.0515,
"step": 239
},
{
"epoch": 0.43399638336347196,
"grad_norm": 2.8107869625091553,
"learning_rate": 9.261245489256956e-06,
"loss": 0.0685,
"step": 240
},
{
"epoch": 0.4358047016274864,
"grad_norm": 1.132842779159546,
"learning_rate": 9.253400328436699e-06,
"loss": 0.0418,
"step": 241
},
{
"epoch": 0.4376130198915009,
"grad_norm": 0.9092134833335876,
"learning_rate": 9.24551709111097e-06,
"loss": 0.0243,
"step": 242
},
{
"epoch": 0.4394213381555154,
"grad_norm": 1.61929452419281,
"learning_rate": 9.237595847850628e-06,
"loss": 0.0444,
"step": 243
},
{
"epoch": 0.4412296564195298,
"grad_norm": 12.186718940734863,
"learning_rate": 9.229636669566769e-06,
"loss": 0.0599,
"step": 244
},
{
"epoch": 0.4430379746835443,
"grad_norm": 5.799802780151367,
"learning_rate": 9.221639627510076e-06,
"loss": 0.1219,
"step": 245
},
{
"epoch": 0.4448462929475588,
"grad_norm": 1.5052590370178223,
"learning_rate": 9.213604793270196e-06,
"loss": 0.0611,
"step": 246
},
{
"epoch": 0.44665461121157324,
"grad_norm": 1.974016547203064,
"learning_rate": 9.205532238775089e-06,
"loss": 0.0454,
"step": 247
},
{
"epoch": 0.4484629294755877,
"grad_norm": 5.944695949554443,
"learning_rate": 9.197422036290386e-06,
"loss": 0.0522,
"step": 248
},
{
"epoch": 0.45027124773960214,
"grad_norm": 0.9125894904136658,
"learning_rate": 9.189274258418748e-06,
"loss": 0.0363,
"step": 249
},
{
"epoch": 0.45207956600361665,
"grad_norm": 1.2161636352539062,
"learning_rate": 9.181088978099203e-06,
"loss": 0.0276,
"step": 250
},
{
"epoch": 0.4538878842676311,
"grad_norm": 1.5363112688064575,
"learning_rate": 9.172866268606514e-06,
"loss": 0.0387,
"step": 251
},
{
"epoch": 0.45569620253164556,
"grad_norm": 1.7365622520446777,
"learning_rate": 9.164606203550498e-06,
"loss": 0.0869,
"step": 252
},
{
"epoch": 0.45750452079566006,
"grad_norm": 1.1523739099502563,
"learning_rate": 9.156308856875386e-06,
"loss": 0.0398,
"step": 253
},
{
"epoch": 0.4593128390596745,
"grad_norm": 2.569274663925171,
"learning_rate": 9.147974302859158e-06,
"loss": 0.0612,
"step": 254
},
{
"epoch": 0.46112115732368897,
"grad_norm": 3.165576934814453,
"learning_rate": 9.139602616112864e-06,
"loss": 0.0523,
"step": 255
},
{
"epoch": 0.4629294755877034,
"grad_norm": 3.1608726978302,
"learning_rate": 9.131193871579975e-06,
"loss": 0.0586,
"step": 256
},
{
"epoch": 0.46473779385171793,
"grad_norm": 4.367858409881592,
"learning_rate": 9.122748144535704e-06,
"loss": 0.0579,
"step": 257
},
{
"epoch": 0.4665461121157324,
"grad_norm": 0.6591166257858276,
"learning_rate": 9.114265510586329e-06,
"loss": 0.0218,
"step": 258
},
{
"epoch": 0.46835443037974683,
"grad_norm": 1.370559811592102,
"learning_rate": 9.10574604566852e-06,
"loss": 0.0565,
"step": 259
},
{
"epoch": 0.4701627486437613,
"grad_norm": 5.232730388641357,
"learning_rate": 9.09718982604866e-06,
"loss": 0.1028,
"step": 260
},
{
"epoch": 0.4719710669077758,
"grad_norm": 7.323339939117432,
"learning_rate": 9.088596928322158e-06,
"loss": 0.0323,
"step": 261
},
{
"epoch": 0.47377938517179025,
"grad_norm": 1.0591479539871216,
"learning_rate": 9.079967429412766e-06,
"loss": 0.0308,
"step": 262
},
{
"epoch": 0.4755877034358047,
"grad_norm": 4.1140923500061035,
"learning_rate": 9.071301406571893e-06,
"loss": 0.0437,
"step": 263
},
{
"epoch": 0.47739602169981915,
"grad_norm": 0.9397031664848328,
"learning_rate": 9.062598937377911e-06,
"loss": 0.0453,
"step": 264
},
{
"epoch": 0.47920433996383366,
"grad_norm": 1.8974077701568604,
"learning_rate": 9.053860099735455e-06,
"loss": 0.0501,
"step": 265
},
{
"epoch": 0.4810126582278481,
"grad_norm": 1.5855399370193481,
"learning_rate": 9.045084971874738e-06,
"loss": 0.0581,
"step": 266
},
{
"epoch": 0.48282097649186256,
"grad_norm": 0.7080597281455994,
"learning_rate": 9.036273632350839e-06,
"loss": 0.0289,
"step": 267
},
{
"epoch": 0.484629294755877,
"grad_norm": 1.3441932201385498,
"learning_rate": 9.027426160043005e-06,
"loss": 0.0841,
"step": 268
},
{
"epoch": 0.4864376130198915,
"grad_norm": 0.9173007011413574,
"learning_rate": 9.018542634153944e-06,
"loss": 0.0387,
"step": 269
},
{
"epoch": 0.488245931283906,
"grad_norm": 1.4931542873382568,
"learning_rate": 9.00962313420912e-06,
"loss": 0.0945,
"step": 270
},
{
"epoch": 0.49005424954792043,
"grad_norm": 0.8358903527259827,
"learning_rate": 9.000667740056033e-06,
"loss": 0.0389,
"step": 271
},
{
"epoch": 0.4918625678119349,
"grad_norm": 1.0547810792922974,
"learning_rate": 8.991676531863507e-06,
"loss": 0.0479,
"step": 272
},
{
"epoch": 0.4936708860759494,
"grad_norm": 1.5400824546813965,
"learning_rate": 8.982649590120982e-06,
"loss": 0.0584,
"step": 273
},
{
"epoch": 0.49547920433996384,
"grad_norm": 0.7854518890380859,
"learning_rate": 8.973586995637778e-06,
"loss": 0.0347,
"step": 274
},
{
"epoch": 0.4972875226039783,
"grad_norm": 0.7611461877822876,
"learning_rate": 8.964488829542377e-06,
"loss": 0.0501,
"step": 275
},
{
"epoch": 0.49909584086799275,
"grad_norm": 1.001150369644165,
"learning_rate": 8.955355173281709e-06,
"loss": 0.0376,
"step": 276
},
{
"epoch": 0.5009041591320073,
"grad_norm": 1.156732439994812,
"learning_rate": 8.946186108620397e-06,
"loss": 0.0618,
"step": 277
},
{
"epoch": 0.5027124773960217,
"grad_norm": 1.0109257698059082,
"learning_rate": 8.936981717640061e-06,
"loss": 0.0486,
"step": 278
},
{
"epoch": 0.5045207956600362,
"grad_norm": 2.299504041671753,
"learning_rate": 8.927742082738542e-06,
"loss": 0.0562,
"step": 279
},
{
"epoch": 0.5063291139240507,
"grad_norm": 3.316723346710205,
"learning_rate": 8.9184672866292e-06,
"loss": 0.055,
"step": 280
},
{
"epoch": 0.5081374321880651,
"grad_norm": 1.1725521087646484,
"learning_rate": 8.90915741234015e-06,
"loss": 0.0523,
"step": 281
},
{
"epoch": 0.5099457504520796,
"grad_norm": 1.017571210861206,
"learning_rate": 8.899812543213532e-06,
"loss": 0.0635,
"step": 282
},
{
"epoch": 0.5117540687160941,
"grad_norm": 1.3900868892669678,
"learning_rate": 8.890432762904757e-06,
"loss": 0.029,
"step": 283
},
{
"epoch": 0.5135623869801085,
"grad_norm": 0.4764467179775238,
"learning_rate": 8.881018155381766e-06,
"loss": 0.0221,
"step": 284
},
{
"epoch": 0.515370705244123,
"grad_norm": 2.174265146255493,
"learning_rate": 8.871568804924269e-06,
"loss": 0.0945,
"step": 285
},
{
"epoch": 0.5171790235081374,
"grad_norm": 1.6411850452423096,
"learning_rate": 8.862084796122998e-06,
"loss": 0.0588,
"step": 286
},
{
"epoch": 0.5189873417721519,
"grad_norm": 1.3759667873382568,
"learning_rate": 8.852566213878947e-06,
"loss": 0.0841,
"step": 287
},
{
"epoch": 0.5207956600361664,
"grad_norm": 3.2979047298431396,
"learning_rate": 8.84301314340261e-06,
"loss": 0.0474,
"step": 288
},
{
"epoch": 0.5226039783001808,
"grad_norm": 3.1667160987854004,
"learning_rate": 8.833425670213223e-06,
"loss": 0.0761,
"step": 289
},
{
"epoch": 0.5244122965641953,
"grad_norm": 0.6941249370574951,
"learning_rate": 8.823803880137993e-06,
"loss": 0.0365,
"step": 290
},
{
"epoch": 0.5262206148282098,
"grad_norm": 0.6920735239982605,
"learning_rate": 8.814147859311333e-06,
"loss": 0.0531,
"step": 291
},
{
"epoch": 0.5280289330922242,
"grad_norm": 2.3563013076782227,
"learning_rate": 8.804457694174093e-06,
"loss": 0.0642,
"step": 292
},
{
"epoch": 0.5298372513562387,
"grad_norm": 1.030884861946106,
"learning_rate": 8.794733471472778e-06,
"loss": 0.0507,
"step": 293
},
{
"epoch": 0.5316455696202531,
"grad_norm": 1.0619593858718872,
"learning_rate": 8.784975278258783e-06,
"loss": 0.0424,
"step": 294
},
{
"epoch": 0.5334538878842676,
"grad_norm": 1.0486905574798584,
"learning_rate": 8.775183201887603e-06,
"loss": 0.0574,
"step": 295
},
{
"epoch": 0.5352622061482821,
"grad_norm": 1.174930453300476,
"learning_rate": 8.765357330018056e-06,
"loss": 0.0657,
"step": 296
},
{
"epoch": 0.5370705244122965,
"grad_norm": 1.3345396518707275,
"learning_rate": 8.755497750611498e-06,
"loss": 0.0682,
"step": 297
},
{
"epoch": 0.538878842676311,
"grad_norm": 0.8685816526412964,
"learning_rate": 8.745604551931042e-06,
"loss": 0.0501,
"step": 298
},
{
"epoch": 0.5406871609403255,
"grad_norm": 0.6714222431182861,
"learning_rate": 8.73567782254075e-06,
"loss": 0.0324,
"step": 299
},
{
"epoch": 0.5424954792043399,
"grad_norm": 0.6705530285835266,
"learning_rate": 8.725717651304856e-06,
"loss": 0.0271,
"step": 300
},
{
"epoch": 0.5443037974683544,
"grad_norm": 1.1572880744934082,
"learning_rate": 8.715724127386971e-06,
"loss": 0.044,
"step": 301
},
{
"epoch": 0.546112115732369,
"grad_norm": 2.1497600078582764,
"learning_rate": 8.705697340249275e-06,
"loss": 0.06,
"step": 302
},
{
"epoch": 0.5479204339963833,
"grad_norm": 1.4481446743011475,
"learning_rate": 8.695637379651719e-06,
"loss": 0.0509,
"step": 303
},
{
"epoch": 0.5497287522603979,
"grad_norm": 2.159526824951172,
"learning_rate": 8.685544335651226e-06,
"loss": 0.0666,
"step": 304
},
{
"epoch": 0.5515370705244123,
"grad_norm": 1.2908952236175537,
"learning_rate": 8.675418298600884e-06,
"loss": 0.0525,
"step": 305
},
{
"epoch": 0.5533453887884268,
"grad_norm": 1.292710304260254,
"learning_rate": 8.665259359149132e-06,
"loss": 0.0424,
"step": 306
},
{
"epoch": 0.5551537070524413,
"grad_norm": 1.0498162508010864,
"learning_rate": 8.655067608238953e-06,
"loss": 0.0561,
"step": 307
},
{
"epoch": 0.5569620253164557,
"grad_norm": 0.6912449598312378,
"learning_rate": 8.644843137107058e-06,
"loss": 0.0413,
"step": 308
},
{
"epoch": 0.5587703435804702,
"grad_norm": 1.1498081684112549,
"learning_rate": 8.634586037283072e-06,
"loss": 0.0531,
"step": 309
},
{
"epoch": 0.5605786618444847,
"grad_norm": 1.0166106224060059,
"learning_rate": 8.62429640058871e-06,
"loss": 0.0225,
"step": 310
},
{
"epoch": 0.5623869801084991,
"grad_norm": 2.0684378147125244,
"learning_rate": 8.613974319136959e-06,
"loss": 0.0528,
"step": 311
},
{
"epoch": 0.5641952983725136,
"grad_norm": 1.092981219291687,
"learning_rate": 8.603619885331251e-06,
"loss": 0.0798,
"step": 312
},
{
"epoch": 0.566003616636528,
"grad_norm": 1.361475944519043,
"learning_rate": 8.593233191864638e-06,
"loss": 0.0609,
"step": 313
},
{
"epoch": 0.5678119349005425,
"grad_norm": 1.1198203563690186,
"learning_rate": 8.582814331718961e-06,
"loss": 0.0751,
"step": 314
},
{
"epoch": 0.569620253164557,
"grad_norm": 0.8867238163948059,
"learning_rate": 8.572363398164017e-06,
"loss": 0.0311,
"step": 315
},
{
"epoch": 0.5714285714285714,
"grad_norm": 0.949080228805542,
"learning_rate": 8.561880484756726e-06,
"loss": 0.0521,
"step": 316
},
{
"epoch": 0.5732368896925859,
"grad_norm": 0.6268083453178406,
"learning_rate": 8.551365685340285e-06,
"loss": 0.021,
"step": 317
},
{
"epoch": 0.5750452079566004,
"grad_norm": 0.7100647687911987,
"learning_rate": 8.540819094043349e-06,
"loss": 0.0334,
"step": 318
},
{
"epoch": 0.5768535262206148,
"grad_norm": 0.8354303240776062,
"learning_rate": 8.530240805279159e-06,
"loss": 0.0526,
"step": 319
},
{
"epoch": 0.5786618444846293,
"grad_norm": 0.9385378956794739,
"learning_rate": 8.519630913744726e-06,
"loss": 0.0466,
"step": 320
},
{
"epoch": 0.5804701627486437,
"grad_norm": 1.5632243156433105,
"learning_rate": 8.508989514419959e-06,
"loss": 0.0346,
"step": 321
},
{
"epoch": 0.5822784810126582,
"grad_norm": 0.35631757974624634,
"learning_rate": 8.498316702566828e-06,
"loss": 0.0179,
"step": 322
},
{
"epoch": 0.5840867992766727,
"grad_norm": 0.5897761583328247,
"learning_rate": 8.487612573728513e-06,
"loss": 0.0228,
"step": 323
},
{
"epoch": 0.5858951175406871,
"grad_norm": 0.5373837947845459,
"learning_rate": 8.476877223728539e-06,
"loss": 0.028,
"step": 324
},
{
"epoch": 0.5877034358047016,
"grad_norm": 0.694993257522583,
"learning_rate": 8.466110748669926e-06,
"loss": 0.0371,
"step": 325
},
{
"epoch": 0.5895117540687161,
"grad_norm": 0.7707152962684631,
"learning_rate": 8.455313244934324e-06,
"loss": 0.0238,
"step": 326
},
{
"epoch": 0.5913200723327305,
"grad_norm": 1.1971808671951294,
"learning_rate": 8.444484809181155e-06,
"loss": 0.0577,
"step": 327
},
{
"epoch": 0.593128390596745,
"grad_norm": 0.45127764344215393,
"learning_rate": 8.433625538346742e-06,
"loss": 0.0126,
"step": 328
},
{
"epoch": 0.5949367088607594,
"grad_norm": 0.686957597732544,
"learning_rate": 8.422735529643445e-06,
"loss": 0.0453,
"step": 329
},
{
"epoch": 0.596745027124774,
"grad_norm": 2.691413402557373,
"learning_rate": 8.41181488055879e-06,
"loss": 0.093,
"step": 330
},
{
"epoch": 0.5985533453887885,
"grad_norm": 1.3400367498397827,
"learning_rate": 8.400863688854598e-06,
"loss": 0.0741,
"step": 331
},
{
"epoch": 0.6003616636528029,
"grad_norm": 1.2479829788208008,
"learning_rate": 8.389882052566106e-06,
"loss": 0.0619,
"step": 332
},
{
"epoch": 0.6021699819168174,
"grad_norm": 1.0524506568908691,
"learning_rate": 8.37887007000109e-06,
"loss": 0.0533,
"step": 333
},
{
"epoch": 0.6039783001808319,
"grad_norm": 3.235327959060669,
"learning_rate": 8.36782783973899e-06,
"loss": 0.0585,
"step": 334
},
{
"epoch": 0.6057866184448463,
"grad_norm": 0.6744747161865234,
"learning_rate": 8.35675546063002e-06,
"loss": 0.0381,
"step": 335
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.6971617341041565,
"learning_rate": 8.345653031794292e-06,
"loss": 0.029,
"step": 336
},
{
"epoch": 0.6094032549728752,
"grad_norm": 0.9673008322715759,
"learning_rate": 8.334520652620918e-06,
"loss": 0.0346,
"step": 337
},
{
"epoch": 0.6112115732368897,
"grad_norm": 0.3614397346973419,
"learning_rate": 8.32335842276713e-06,
"loss": 0.0119,
"step": 338
},
{
"epoch": 0.6130198915009042,
"grad_norm": 0.6786327362060547,
"learning_rate": 8.31216644215738e-06,
"loss": 0.042,
"step": 339
},
{
"epoch": 0.6148282097649186,
"grad_norm": 0.7057145833969116,
"learning_rate": 8.300944810982452e-06,
"loss": 0.0332,
"step": 340
},
{
"epoch": 0.6166365280289331,
"grad_norm": 1.17477285861969,
"learning_rate": 8.289693629698564e-06,
"loss": 0.0173,
"step": 341
},
{
"epoch": 0.6184448462929476,
"grad_norm": 0.5841035842895508,
"learning_rate": 8.278412999026462e-06,
"loss": 0.0345,
"step": 342
},
{
"epoch": 0.620253164556962,
"grad_norm": 0.8609682321548462,
"learning_rate": 8.267103019950529e-06,
"loss": 0.0333,
"step": 343
},
{
"epoch": 0.6220614828209765,
"grad_norm": 0.8763616681098938,
"learning_rate": 8.255763793717868e-06,
"loss": 0.0211,
"step": 344
},
{
"epoch": 0.6238698010849909,
"grad_norm": 1.5052502155303955,
"learning_rate": 8.244395421837412e-06,
"loss": 0.0248,
"step": 345
},
{
"epoch": 0.6256781193490054,
"grad_norm": 0.8762037754058838,
"learning_rate": 8.232998006078998e-06,
"loss": 0.0534,
"step": 346
},
{
"epoch": 0.6274864376130199,
"grad_norm": 2.080583095550537,
"learning_rate": 8.221571648472473e-06,
"loss": 0.0169,
"step": 347
},
{
"epoch": 0.6292947558770343,
"grad_norm": 1.3618680238723755,
"learning_rate": 8.210116451306762e-06,
"loss": 0.0422,
"step": 348
},
{
"epoch": 0.6311030741410488,
"grad_norm": 1.288297176361084,
"learning_rate": 8.198632517128968e-06,
"loss": 0.0648,
"step": 349
},
{
"epoch": 0.6329113924050633,
"grad_norm": 1.192158818244934,
"learning_rate": 8.18711994874345e-06,
"loss": 0.0251,
"step": 350
},
{
"epoch": 0.6347197106690777,
"grad_norm": 0.5660313367843628,
"learning_rate": 8.175578849210894e-06,
"loss": 0.0222,
"step": 351
},
{
"epoch": 0.6365280289330922,
"grad_norm": 0.5471470952033997,
"learning_rate": 8.164009321847405e-06,
"loss": 0.0357,
"step": 352
},
{
"epoch": 0.6383363471971067,
"grad_norm": 0.6500267386436462,
"learning_rate": 8.15241147022357e-06,
"loss": 0.0282,
"step": 353
},
{
"epoch": 0.6401446654611211,
"grad_norm": 1.1992710828781128,
"learning_rate": 8.140785398163535e-06,
"loss": 0.0658,
"step": 354
},
{
"epoch": 0.6419529837251357,
"grad_norm": 1.25344717502594,
"learning_rate": 8.129131209744075e-06,
"loss": 0.0468,
"step": 355
},
{
"epoch": 0.64376130198915,
"grad_norm": 1.5819299221038818,
"learning_rate": 8.117449009293668e-06,
"loss": 0.0856,
"step": 356
},
{
"epoch": 0.6455696202531646,
"grad_norm": 0.8980645537376404,
"learning_rate": 8.105738901391553e-06,
"loss": 0.0445,
"step": 357
},
{
"epoch": 0.6473779385171791,
"grad_norm": 0.584520697593689,
"learning_rate": 8.094000990866795e-06,
"loss": 0.0348,
"step": 358
},
{
"epoch": 0.6491862567811935,
"grad_norm": 0.8350209593772888,
"learning_rate": 8.08223538279735e-06,
"loss": 0.0221,
"step": 359
},
{
"epoch": 0.650994575045208,
"grad_norm": 1.2882463932037354,
"learning_rate": 8.070442182509127e-06,
"loss": 0.0853,
"step": 360
},
{
"epoch": 0.6528028933092225,
"grad_norm": 0.7137979865074158,
"learning_rate": 8.058621495575032e-06,
"loss": 0.0503,
"step": 361
},
{
"epoch": 0.6546112115732369,
"grad_norm": 0.6045083999633789,
"learning_rate": 8.046773427814043e-06,
"loss": 0.0254,
"step": 362
},
{
"epoch": 0.6564195298372514,
"grad_norm": 0.7745179533958435,
"learning_rate": 8.034898085290239e-06,
"loss": 0.0468,
"step": 363
},
{
"epoch": 0.6582278481012658,
"grad_norm": 1.1030985116958618,
"learning_rate": 8.022995574311876e-06,
"loss": 0.0658,
"step": 364
},
{
"epoch": 0.6600361663652803,
"grad_norm": 2.026185989379883,
"learning_rate": 8.011066001430412e-06,
"loss": 0.0453,
"step": 365
},
{
"epoch": 0.6618444846292948,
"grad_norm": 0.8097315430641174,
"learning_rate": 7.99910947343957e-06,
"loss": 0.0621,
"step": 366
},
{
"epoch": 0.6636528028933092,
"grad_norm": 1.414018154144287,
"learning_rate": 7.987126097374372e-06,
"loss": 0.0785,
"step": 367
},
{
"epoch": 0.6654611211573237,
"grad_norm": 0.659436047077179,
"learning_rate": 7.975115980510187e-06,
"loss": 0.0279,
"step": 368
},
{
"epoch": 0.6672694394213382,
"grad_norm": 0.9226219058036804,
"learning_rate": 7.963079230361765e-06,
"loss": 0.0286,
"step": 369
},
{
"epoch": 0.6690777576853526,
"grad_norm": 0.5210973024368286,
"learning_rate": 7.951015954682281e-06,
"loss": 0.0307,
"step": 370
},
{
"epoch": 0.6708860759493671,
"grad_norm": 0.7646634578704834,
"learning_rate": 7.938926261462366e-06,
"loss": 0.0399,
"step": 371
},
{
"epoch": 0.6726943942133815,
"grad_norm": 1.1255048513412476,
"learning_rate": 7.926810258929138e-06,
"loss": 0.041,
"step": 372
},
{
"epoch": 0.674502712477396,
"grad_norm": 0.4182261526584625,
"learning_rate": 7.914668055545242e-06,
"loss": 0.0183,
"step": 373
},
{
"epoch": 0.6763110307414105,
"grad_norm": 2.2014074325561523,
"learning_rate": 7.902499760007867e-06,
"loss": 0.0472,
"step": 374
},
{
"epoch": 0.6781193490054249,
"grad_norm": 0.47527626156806946,
"learning_rate": 7.890305481247786e-06,
"loss": 0.0245,
"step": 375
},
{
"epoch": 0.6799276672694394,
"grad_norm": 0.9921447038650513,
"learning_rate": 7.87808532842837e-06,
"loss": 0.0291,
"step": 376
},
{
"epoch": 0.6817359855334539,
"grad_norm": 1.0139051675796509,
"learning_rate": 7.865839410944613e-06,
"loss": 0.0603,
"step": 377
},
{
"epoch": 0.6835443037974683,
"grad_norm": 0.4706767201423645,
"learning_rate": 7.85356783842216e-06,
"loss": 0.0251,
"step": 378
},
{
"epoch": 0.6853526220614828,
"grad_norm": 1.011164665222168,
"learning_rate": 7.841270720716318e-06,
"loss": 0.0544,
"step": 379
},
{
"epoch": 0.6871609403254972,
"grad_norm": 0.5896030068397522,
"learning_rate": 7.828948167911073e-06,
"loss": 0.0366,
"step": 380
},
{
"epoch": 0.6889692585895117,
"grad_norm": 0.8391334414482117,
"learning_rate": 7.81660029031811e-06,
"loss": 0.0424,
"step": 381
},
{
"epoch": 0.6907775768535263,
"grad_norm": 1.242993712425232,
"learning_rate": 7.804227198475823e-06,
"loss": 0.0459,
"step": 382
},
{
"epoch": 0.6925858951175407,
"grad_norm": 0.9627134203910828,
"learning_rate": 7.791829003148313e-06,
"loss": 0.0386,
"step": 383
},
{
"epoch": 0.6943942133815552,
"grad_norm": 0.2087165117263794,
"learning_rate": 7.779405815324424e-06,
"loss": 0.0079,
"step": 384
},
{
"epoch": 0.6962025316455697,
"grad_norm": 0.7946087718009949,
"learning_rate": 7.76695774621672e-06,
"loss": 0.047,
"step": 385
},
{
"epoch": 0.6980108499095841,
"grad_norm": 1.766443133354187,
"learning_rate": 7.754484907260513e-06,
"loss": 0.0638,
"step": 386
},
{
"epoch": 0.6998191681735986,
"grad_norm": 0.8510230183601379,
"learning_rate": 7.741987410112847e-06,
"loss": 0.0705,
"step": 387
},
{
"epoch": 0.701627486437613,
"grad_norm": 0.7218502759933472,
"learning_rate": 7.72946536665151e-06,
"loss": 0.0293,
"step": 388
},
{
"epoch": 0.7034358047016275,
"grad_norm": 0.26688218116760254,
"learning_rate": 7.716918888974029e-06,
"loss": 0.005,
"step": 389
},
{
"epoch": 0.705244122965642,
"grad_norm": 0.9809504151344299,
"learning_rate": 7.704348089396667e-06,
"loss": 0.045,
"step": 390
},
{
"epoch": 0.7070524412296564,
"grad_norm": 2.1851389408111572,
"learning_rate": 7.691753080453413e-06,
"loss": 0.0601,
"step": 391
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.7249248623847961,
"learning_rate": 7.679133974894984e-06,
"loss": 0.0205,
"step": 392
},
{
"epoch": 0.7106690777576854,
"grad_norm": 0.8017813563346863,
"learning_rate": 7.666490885687807e-06,
"loss": 0.0243,
"step": 393
},
{
"epoch": 0.7124773960216998,
"grad_norm": 0.9082496762275696,
"learning_rate": 7.653823926013016e-06,
"loss": 0.0629,
"step": 394
},
{
"epoch": 0.7142857142857143,
"grad_norm": 0.7406293749809265,
"learning_rate": 7.641133209265423e-06,
"loss": 0.0235,
"step": 395
},
{
"epoch": 0.7160940325497287,
"grad_norm": 0.5935758948326111,
"learning_rate": 7.628418849052523e-06,
"loss": 0.0263,
"step": 396
},
{
"epoch": 0.7179023508137432,
"grad_norm": 0.35810256004333496,
"learning_rate": 7.615680959193469e-06,
"loss": 0.0136,
"step": 397
},
{
"epoch": 0.7197106690777577,
"grad_norm": 0.7280279397964478,
"learning_rate": 7.602919653718044e-06,
"loss": 0.0382,
"step": 398
},
{
"epoch": 0.7215189873417721,
"grad_norm": 0.6724569797515869,
"learning_rate": 7.590135046865652e-06,
"loss": 0.0363,
"step": 399
},
{
"epoch": 0.7233273056057866,
"grad_norm": 0.5071216821670532,
"learning_rate": 7.577327253084292e-06,
"loss": 0.0239,
"step": 400
},
{
"epoch": 0.7251356238698011,
"grad_norm": 0.674170196056366,
"learning_rate": 7.564496387029532e-06,
"loss": 0.0599,
"step": 401
},
{
"epoch": 0.7269439421338155,
"grad_norm": 0.7169482707977295,
"learning_rate": 7.551642563563481e-06,
"loss": 0.0253,
"step": 402
},
{
"epoch": 0.72875226039783,
"grad_norm": 0.7994451522827148,
"learning_rate": 7.5387658977537695e-06,
"loss": 0.0393,
"step": 403
},
{
"epoch": 0.7305605786618445,
"grad_norm": 0.6274312138557434,
"learning_rate": 7.5258665048725065e-06,
"loss": 0.0363,
"step": 404
},
{
"epoch": 0.7323688969258589,
"grad_norm": 0.9662621021270752,
"learning_rate": 7.512944500395255e-06,
"loss": 0.0603,
"step": 405
},
{
"epoch": 0.7341772151898734,
"grad_norm": 1.441202163696289,
"learning_rate": 7.500000000000001e-06,
"loss": 0.0439,
"step": 406
},
{
"epoch": 0.7359855334538878,
"grad_norm": 0.920659601688385,
"learning_rate": 7.48703311956611e-06,
"loss": 0.0525,
"step": 407
},
{
"epoch": 0.7377938517179023,
"grad_norm": 1.1080375909805298,
"learning_rate": 7.4740439751732994e-06,
"loss": 0.041,
"step": 408
},
{
"epoch": 0.7396021699819169,
"grad_norm": 0.6033267378807068,
"learning_rate": 7.461032683100587e-06,
"loss": 0.0313,
"step": 409
},
{
"epoch": 0.7414104882459313,
"grad_norm": 0.7940472960472107,
"learning_rate": 7.447999359825263e-06,
"loss": 0.0306,
"step": 410
},
{
"epoch": 0.7432188065099458,
"grad_norm": 0.5652586221694946,
"learning_rate": 7.434944122021837e-06,
"loss": 0.0213,
"step": 411
},
{
"epoch": 0.7450271247739603,
"grad_norm": 1.069399118423462,
"learning_rate": 7.421867086561001e-06,
"loss": 0.0503,
"step": 412
},
{
"epoch": 0.7468354430379747,
"grad_norm": 1.1518359184265137,
"learning_rate": 7.408768370508577e-06,
"loss": 0.0201,
"step": 413
},
{
"epoch": 0.7486437613019892,
"grad_norm": 0.407006174325943,
"learning_rate": 7.395648091124476e-06,
"loss": 0.0141,
"step": 414
},
{
"epoch": 0.7504520795660036,
"grad_norm": 1.1469006538391113,
"learning_rate": 7.382506365861639e-06,
"loss": 0.0442,
"step": 415
},
{
"epoch": 0.7522603978300181,
"grad_norm": 0.8655562400817871,
"learning_rate": 7.369343312364994e-06,
"loss": 0.0362,
"step": 416
},
{
"epoch": 0.7540687160940326,
"grad_norm": 0.7426546216011047,
"learning_rate": 7.356159048470402e-06,
"loss": 0.0315,
"step": 417
},
{
"epoch": 0.755877034358047,
"grad_norm": 0.6241861581802368,
"learning_rate": 7.342953692203594e-06,
"loss": 0.0274,
"step": 418
},
{
"epoch": 0.7576853526220615,
"grad_norm": 0.3736267685890198,
"learning_rate": 7.3297273617791246e-06,
"loss": 0.0237,
"step": 419
},
{
"epoch": 0.759493670886076,
"grad_norm": 0.3901918828487396,
"learning_rate": 7.31648017559931e-06,
"loss": 0.0102,
"step": 420
},
{
"epoch": 0.7613019891500904,
"grad_norm": 0.8695743680000305,
"learning_rate": 7.303212252253163e-06,
"loss": 0.0403,
"step": 421
},
{
"epoch": 0.7631103074141049,
"grad_norm": 0.09739478677511215,
"learning_rate": 7.289923710515338e-06,
"loss": 0.0028,
"step": 422
},
{
"epoch": 0.7649186256781193,
"grad_norm": 0.635344922542572,
"learning_rate": 7.276614669345069e-06,
"loss": 0.0325,
"step": 423
},
{
"epoch": 0.7667269439421338,
"grad_norm": 0.4680003225803375,
"learning_rate": 7.263285247885097e-06,
"loss": 0.0194,
"step": 424
},
{
"epoch": 0.7685352622061483,
"grad_norm": 0.6451188325881958,
"learning_rate": 7.249935565460606e-06,
"loss": 0.026,
"step": 425
},
{
"epoch": 0.7703435804701627,
"grad_norm": 0.2888726592063904,
"learning_rate": 7.236565741578163e-06,
"loss": 0.0094,
"step": 426
},
{
"epoch": 0.7721518987341772,
"grad_norm": 1.4941391944885254,
"learning_rate": 7.223175895924638e-06,
"loss": 0.0453,
"step": 427
},
{
"epoch": 0.7739602169981917,
"grad_norm": 1.1318025588989258,
"learning_rate": 7.2097661483661355e-06,
"loss": 0.0569,
"step": 428
},
{
"epoch": 0.7757685352622061,
"grad_norm": 0.849553108215332,
"learning_rate": 7.19633661894692e-06,
"loss": 0.0279,
"step": 429
},
{
"epoch": 0.7775768535262206,
"grad_norm": 0.7323757410049438,
"learning_rate": 7.182887427888351e-06,
"loss": 0.0559,
"step": 430
},
{
"epoch": 0.779385171790235,
"grad_norm": 0.6708653569221497,
"learning_rate": 7.169418695587791e-06,
"loss": 0.0223,
"step": 431
},
{
"epoch": 0.7811934900542495,
"grad_norm": 0.8238298296928406,
"learning_rate": 7.155930542617543e-06,
"loss": 0.0432,
"step": 432
},
{
"epoch": 0.783001808318264,
"grad_norm": 0.5019189715385437,
"learning_rate": 7.142423089723758e-06,
"loss": 0.0126,
"step": 433
},
{
"epoch": 0.7848101265822784,
"grad_norm": 0.6828831434249878,
"learning_rate": 7.128896457825364e-06,
"loss": 0.0304,
"step": 434
},
{
"epoch": 0.786618444846293,
"grad_norm": 1.728789210319519,
"learning_rate": 7.11535076801298e-06,
"loss": 0.0487,
"step": 435
},
{
"epoch": 0.7884267631103075,
"grad_norm": 0.9259436130523682,
"learning_rate": 7.101786141547829e-06,
"loss": 0.0657,
"step": 436
},
{
"epoch": 0.7902350813743219,
"grad_norm": 0.8076736927032471,
"learning_rate": 7.088202699860656e-06,
"loss": 0.0563,
"step": 437
},
{
"epoch": 0.7920433996383364,
"grad_norm": 0.7568901181221008,
"learning_rate": 7.074600564550643e-06,
"loss": 0.0292,
"step": 438
},
{
"epoch": 0.7938517179023508,
"grad_norm": 0.7115985155105591,
"learning_rate": 7.060979857384316e-06,
"loss": 0.042,
"step": 439
},
{
"epoch": 0.7956600361663653,
"grad_norm": 1.6664183139801025,
"learning_rate": 7.047340700294454e-06,
"loss": 0.0554,
"step": 440
},
{
"epoch": 0.7974683544303798,
"grad_norm": 0.7137376070022583,
"learning_rate": 7.033683215379002e-06,
"loss": 0.0531,
"step": 441
},
{
"epoch": 0.7992766726943942,
"grad_norm": 0.802460253238678,
"learning_rate": 7.020007524899976e-06,
"loss": 0.0513,
"step": 442
},
{
"epoch": 0.8010849909584087,
"grad_norm": 1.0362261533737183,
"learning_rate": 7.006313751282372e-06,
"loss": 0.0331,
"step": 443
},
{
"epoch": 0.8028933092224232,
"grad_norm": 0.7008037567138672,
"learning_rate": 6.992602017113058e-06,
"loss": 0.0196,
"step": 444
},
{
"epoch": 0.8047016274864376,
"grad_norm": 0.5573314428329468,
"learning_rate": 6.978872445139695e-06,
"loss": 0.0292,
"step": 445
},
{
"epoch": 0.8065099457504521,
"grad_norm": 0.48531728982925415,
"learning_rate": 6.965125158269619e-06,
"loss": 0.0354,
"step": 446
},
{
"epoch": 0.8083182640144665,
"grad_norm": 0.5280648469924927,
"learning_rate": 6.951360279568758e-06,
"loss": 0.0212,
"step": 447
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.762917697429657,
"learning_rate": 6.9375779322605154e-06,
"loss": 0.0573,
"step": 448
},
{
"epoch": 0.8119349005424955,
"grad_norm": 0.35672587156295776,
"learning_rate": 6.923778239724681e-06,
"loss": 0.0177,
"step": 449
},
{
"epoch": 0.8137432188065099,
"grad_norm": 0.3046236038208008,
"learning_rate": 6.909961325496312e-06,
"loss": 0.0125,
"step": 450
},
{
"epoch": 0.8155515370705244,
"grad_norm": 0.7308725118637085,
"learning_rate": 6.896127313264643e-06,
"loss": 0.0417,
"step": 451
},
{
"epoch": 0.8173598553345389,
"grad_norm": 0.5233821272850037,
"learning_rate": 6.88227632687196e-06,
"loss": 0.0287,
"step": 452
},
{
"epoch": 0.8191681735985533,
"grad_norm": 0.5598361492156982,
"learning_rate": 6.868408490312511e-06,
"loss": 0.0393,
"step": 453
},
{
"epoch": 0.8209764918625678,
"grad_norm": 0.47879722714424133,
"learning_rate": 6.854523927731383e-06,
"loss": 0.0309,
"step": 454
},
{
"epoch": 0.8227848101265823,
"grad_norm": 1.0283002853393555,
"learning_rate": 6.840622763423391e-06,
"loss": 0.0456,
"step": 455
},
{
"epoch": 0.8245931283905967,
"grad_norm": 0.5238077640533447,
"learning_rate": 6.8267051218319766e-06,
"loss": 0.0294,
"step": 456
},
{
"epoch": 0.8264014466546112,
"grad_norm": 0.6644604206085205,
"learning_rate": 6.8127711275480805e-06,
"loss": 0.0319,
"step": 457
},
{
"epoch": 0.8282097649186256,
"grad_norm": 0.49551424384117126,
"learning_rate": 6.798820905309036e-06,
"loss": 0.0282,
"step": 458
},
{
"epoch": 0.8300180831826401,
"grad_norm": 0.28509873151779175,
"learning_rate": 6.784854579997446e-06,
"loss": 0.01,
"step": 459
},
{
"epoch": 0.8318264014466547,
"grad_norm": 0.7082386016845703,
"learning_rate": 6.7708722766400745e-06,
"loss": 0.0354,
"step": 460
},
{
"epoch": 0.833634719710669,
"grad_norm": 0.805962324142456,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.0529,
"step": 461
},
{
"epoch": 0.8354430379746836,
"grad_norm": 0.6210870742797852,
"learning_rate": 6.7428602366090764e-06,
"loss": 0.0392,
"step": 462
},
{
"epoch": 0.8372513562386981,
"grad_norm": 0.6341171264648438,
"learning_rate": 6.728830750699667e-06,
"loss": 0.022,
"step": 463
},
{
"epoch": 0.8390596745027125,
"grad_norm": 0.6270669102668762,
"learning_rate": 6.714785788270658e-06,
"loss": 0.0318,
"step": 464
},
{
"epoch": 0.840867992766727,
"grad_norm": 0.6710631251335144,
"learning_rate": 6.700725475052773e-06,
"loss": 0.0482,
"step": 465
},
{
"epoch": 0.8426763110307414,
"grad_norm": 0.5528821349143982,
"learning_rate": 6.686649936914151e-06,
"loss": 0.0283,
"step": 466
},
{
"epoch": 0.8444846292947559,
"grad_norm": 0.3247365355491638,
"learning_rate": 6.672559299859228e-06,
"loss": 0.0169,
"step": 467
},
{
"epoch": 0.8462929475587704,
"grad_norm": 0.9084714651107788,
"learning_rate": 6.658453690027604e-06,
"loss": 0.0492,
"step": 468
},
{
"epoch": 0.8481012658227848,
"grad_norm": 0.7043787240982056,
"learning_rate": 6.644333233692917e-06,
"loss": 0.0447,
"step": 469
},
{
"epoch": 0.8499095840867993,
"grad_norm": 0.7011797428131104,
"learning_rate": 6.63019805726171e-06,
"loss": 0.0483,
"step": 470
},
{
"epoch": 0.8517179023508138,
"grad_norm": 1.2029949426651,
"learning_rate": 6.616048287272301e-06,
"loss": 0.0604,
"step": 471
},
{
"epoch": 0.8535262206148282,
"grad_norm": 0.9464662671089172,
"learning_rate": 6.601884050393649e-06,
"loss": 0.0285,
"step": 472
},
{
"epoch": 0.8553345388788427,
"grad_norm": 0.9252480864524841,
"learning_rate": 6.587705473424223e-06,
"loss": 0.0362,
"step": 473
},
{
"epoch": 0.8571428571428571,
"grad_norm": 0.4970252513885498,
"learning_rate": 6.57351268329086e-06,
"loss": 0.0344,
"step": 474
},
{
"epoch": 0.8589511754068716,
"grad_norm": 0.6941171288490295,
"learning_rate": 6.55930580704764e-06,
"loss": 0.0334,
"step": 475
},
{
"epoch": 0.8607594936708861,
"grad_norm": 0.6766089797019958,
"learning_rate": 6.545084971874738e-06,
"loss": 0.0263,
"step": 476
},
{
"epoch": 0.8625678119349005,
"grad_norm": 0.31466686725616455,
"learning_rate": 6.5308503050772884e-06,
"loss": 0.0143,
"step": 477
},
{
"epoch": 0.864376130198915,
"grad_norm": 0.7536757588386536,
"learning_rate": 6.51660193408425e-06,
"loss": 0.0297,
"step": 478
},
{
"epoch": 0.8661844484629295,
"grad_norm": 0.4816785156726837,
"learning_rate": 6.50233998644726e-06,
"loss": 0.0248,
"step": 479
},
{
"epoch": 0.8679927667269439,
"grad_norm": 0.4103059768676758,
"learning_rate": 6.4880645898394935e-06,
"loss": 0.0276,
"step": 480
},
{
"epoch": 0.8698010849909584,
"grad_norm": 0.6490195989608765,
"learning_rate": 6.473775872054522e-06,
"loss": 0.018,
"step": 481
},
{
"epoch": 0.8716094032549728,
"grad_norm": 0.4475712776184082,
"learning_rate": 6.459473961005168e-06,
"loss": 0.0209,
"step": 482
},
{
"epoch": 0.8734177215189873,
"grad_norm": 0.9809353947639465,
"learning_rate": 6.445158984722358e-06,
"loss": 0.0437,
"step": 483
},
{
"epoch": 0.8752260397830018,
"grad_norm": 0.2743368148803711,
"learning_rate": 6.4308310713539845e-06,
"loss": 0.0065,
"step": 484
},
{
"epoch": 0.8770343580470162,
"grad_norm": 0.6572988629341125,
"learning_rate": 6.4164903491637475e-06,
"loss": 0.0237,
"step": 485
},
{
"epoch": 0.8788426763110307,
"grad_norm": 0.6621780395507812,
"learning_rate": 6.402136946530014e-06,
"loss": 0.0557,
"step": 486
},
{
"epoch": 0.8806509945750453,
"grad_norm": 2.1460120677948,
"learning_rate": 6.387770991944667e-06,
"loss": 0.0292,
"step": 487
},
{
"epoch": 0.8824593128390597,
"grad_norm": 0.6116335988044739,
"learning_rate": 6.373392614011952e-06,
"loss": 0.0355,
"step": 488
},
{
"epoch": 0.8842676311030742,
"grad_norm": 0.701877236366272,
"learning_rate": 6.359001941447331e-06,
"loss": 0.0466,
"step": 489
},
{
"epoch": 0.8860759493670886,
"grad_norm": 0.9009343385696411,
"learning_rate": 6.344599103076329e-06,
"loss": 0.0305,
"step": 490
},
{
"epoch": 0.8878842676311031,
"grad_norm": 0.7518783211708069,
"learning_rate": 6.330184227833376e-06,
"loss": 0.0263,
"step": 491
},
{
"epoch": 0.8896925858951176,
"grad_norm": 0.5423868894577026,
"learning_rate": 6.315757444760659e-06,
"loss": 0.0317,
"step": 492
},
{
"epoch": 0.891500904159132,
"grad_norm": 0.28551843762397766,
"learning_rate": 6.301318883006962e-06,
"loss": 0.0071,
"step": 493
},
{
"epoch": 0.8933092224231465,
"grad_norm": 0.27974775433540344,
"learning_rate": 6.286868671826513e-06,
"loss": 0.0107,
"step": 494
},
{
"epoch": 0.895117540687161,
"grad_norm": 0.4651285409927368,
"learning_rate": 6.272406940577827e-06,
"loss": 0.019,
"step": 495
},
{
"epoch": 0.8969258589511754,
"grad_norm": 0.6871982216835022,
"learning_rate": 6.257933818722544e-06,
"loss": 0.0411,
"step": 496
},
{
"epoch": 0.8987341772151899,
"grad_norm": 0.5792664289474487,
"learning_rate": 6.243449435824276e-06,
"loss": 0.0279,
"step": 497
},
{
"epoch": 0.9005424954792043,
"grad_norm": 0.5038998126983643,
"learning_rate": 6.228953921547441e-06,
"loss": 0.0234,
"step": 498
},
{
"epoch": 0.9023508137432188,
"grad_norm": 0.6361072063446045,
"learning_rate": 6.2144474056561076e-06,
"loss": 0.0209,
"step": 499
},
{
"epoch": 0.9041591320072333,
"grad_norm": 0.8955969214439392,
"learning_rate": 6.19993001801283e-06,
"loss": 0.0272,
"step": 500
},
{
"epoch": 0.9059674502712477,
"grad_norm": 0.7934896945953369,
"learning_rate": 6.185401888577488e-06,
"loss": 0.049,
"step": 501
},
{
"epoch": 0.9077757685352622,
"grad_norm": 0.621372640132904,
"learning_rate": 6.17086314740612e-06,
"loss": 0.0294,
"step": 502
},
{
"epoch": 0.9095840867992767,
"grad_norm": 0.6436402201652527,
"learning_rate": 6.1563139246497615e-06,
"loss": 0.0564,
"step": 503
},
{
"epoch": 0.9113924050632911,
"grad_norm": 1.5344867706298828,
"learning_rate": 6.141754350553279e-06,
"loss": 0.035,
"step": 504
},
{
"epoch": 0.9132007233273056,
"grad_norm": 0.7393510341644287,
"learning_rate": 6.1271845554542065e-06,
"loss": 0.0358,
"step": 505
},
{
"epoch": 0.9150090415913201,
"grad_norm": 0.4561188519001007,
"learning_rate": 6.112604669781572e-06,
"loss": 0.0231,
"step": 506
},
{
"epoch": 0.9168173598553345,
"grad_norm": 0.702992856502533,
"learning_rate": 6.098014824054741e-06,
"loss": 0.0394,
"step": 507
},
{
"epoch": 0.918625678119349,
"grad_norm": 0.6905729174613953,
"learning_rate": 6.083415148882236e-06,
"loss": 0.0289,
"step": 508
},
{
"epoch": 0.9204339963833634,
"grad_norm": 0.8555745482444763,
"learning_rate": 6.068805774960574e-06,
"loss": 0.0526,
"step": 509
},
{
"epoch": 0.9222423146473779,
"grad_norm": 3.80324649810791,
"learning_rate": 6.054186833073096e-06,
"loss": 0.0447,
"step": 510
},
{
"epoch": 0.9240506329113924,
"grad_norm": 0.7473819851875305,
"learning_rate": 6.039558454088796e-06,
"loss": 0.032,
"step": 511
},
{
"epoch": 0.9258589511754068,
"grad_norm": 0.6965436339378357,
"learning_rate": 6.024920768961153e-06,
"loss": 0.0329,
"step": 512
},
{
"epoch": 0.9276672694394213,
"grad_norm": 0.4097282588481903,
"learning_rate": 6.010273908726944e-06,
"loss": 0.0109,
"step": 513
},
{
"epoch": 0.9294755877034359,
"grad_norm": 0.3203713297843933,
"learning_rate": 5.995618004505091e-06,
"loss": 0.0139,
"step": 514
},
{
"epoch": 0.9312839059674503,
"grad_norm": 0.6554650664329529,
"learning_rate": 5.980953187495476e-06,
"loss": 0.0306,
"step": 515
},
{
"epoch": 0.9330922242314648,
"grad_norm": 0.3709014058113098,
"learning_rate": 5.9662795889777666e-06,
"loss": 0.0253,
"step": 516
},
{
"epoch": 0.9349005424954792,
"grad_norm": 0.2361323982477188,
"learning_rate": 5.951597340310244e-06,
"loss": 0.013,
"step": 517
},
{
"epoch": 0.9367088607594937,
"grad_norm": 0.8397784233093262,
"learning_rate": 5.936906572928625e-06,
"loss": 0.0395,
"step": 518
},
{
"epoch": 0.9385171790235082,
"grad_norm": 0.5494319796562195,
"learning_rate": 5.922207418344884e-06,
"loss": 0.029,
"step": 519
},
{
"epoch": 0.9403254972875226,
"grad_norm": 0.9510102868080139,
"learning_rate": 5.907500008146082e-06,
"loss": 0.0395,
"step": 520
},
{
"epoch": 0.9421338155515371,
"grad_norm": 0.6437675356864929,
"learning_rate": 5.892784473993184e-06,
"loss": 0.0476,
"step": 521
},
{
"epoch": 0.9439421338155516,
"grad_norm": 0.6374077796936035,
"learning_rate": 5.878060947619877e-06,
"loss": 0.0306,
"step": 522
},
{
"epoch": 0.945750452079566,
"grad_norm": 0.8432037830352783,
"learning_rate": 5.863329560831397e-06,
"loss": 0.052,
"step": 523
},
{
"epoch": 0.9475587703435805,
"grad_norm": 0.5162370800971985,
"learning_rate": 5.848590445503345e-06,
"loss": 0.0285,
"step": 524
},
{
"epoch": 0.9493670886075949,
"grad_norm": 0.5389014482498169,
"learning_rate": 5.8338437335805124e-06,
"loss": 0.0326,
"step": 525
},
{
"epoch": 0.9511754068716094,
"grad_norm": 0.5449837446212769,
"learning_rate": 5.819089557075689e-06,
"loss": 0.0272,
"step": 526
},
{
"epoch": 0.9529837251356239,
"grad_norm": 0.47794368863105774,
"learning_rate": 5.8043280480684925e-06,
"loss": 0.0155,
"step": 527
},
{
"epoch": 0.9547920433996383,
"grad_norm": 0.42232149839401245,
"learning_rate": 5.78955933870418e-06,
"loss": 0.0135,
"step": 528
},
{
"epoch": 0.9566003616636528,
"grad_norm": 0.8405925631523132,
"learning_rate": 5.7747835611924655e-06,
"loss": 0.0685,
"step": 529
},
{
"epoch": 0.9584086799276673,
"grad_norm": 0.5138303637504578,
"learning_rate": 5.760000847806337e-06,
"loss": 0.0167,
"step": 530
},
{
"epoch": 0.9602169981916817,
"grad_norm": 0.5169946551322937,
"learning_rate": 5.745211330880872e-06,
"loss": 0.0265,
"step": 531
},
{
"epoch": 0.9620253164556962,
"grad_norm": 0.7010546922683716,
"learning_rate": 5.730415142812059e-06,
"loss": 0.0253,
"step": 532
},
{
"epoch": 0.9638336347197106,
"grad_norm": 0.6818461418151855,
"learning_rate": 5.7156124160555985e-06,
"loss": 0.0283,
"step": 533
},
{
"epoch": 0.9656419529837251,
"grad_norm": 0.7806791067123413,
"learning_rate": 5.70080328312573e-06,
"loss": 0.0415,
"step": 534
},
{
"epoch": 0.9674502712477396,
"grad_norm": 0.6845316290855408,
"learning_rate": 5.68598787659404e-06,
"loss": 0.0365,
"step": 535
},
{
"epoch": 0.969258589511754,
"grad_norm": 0.6731887459754944,
"learning_rate": 5.671166329088278e-06,
"loss": 0.0223,
"step": 536
},
{
"epoch": 0.9710669077757685,
"grad_norm": 0.42423033714294434,
"learning_rate": 5.656338773291165e-06,
"loss": 0.0156,
"step": 537
},
{
"epoch": 0.972875226039783,
"grad_norm": 0.4014493525028229,
"learning_rate": 5.641505341939212e-06,
"loss": 0.0254,
"step": 538
},
{
"epoch": 0.9746835443037974,
"grad_norm": 0.7907595634460449,
"learning_rate": 5.626666167821522e-06,
"loss": 0.0344,
"step": 539
},
{
"epoch": 0.976491862567812,
"grad_norm": 0.5525479912757874,
"learning_rate": 5.611821383778614e-06,
"loss": 0.0314,
"step": 540
},
{
"epoch": 0.9783001808318263,
"grad_norm": 0.8520960211753845,
"learning_rate": 5.596971122701221e-06,
"loss": 0.0498,
"step": 541
},
{
"epoch": 0.9801084990958409,
"grad_norm": 0.8205142021179199,
"learning_rate": 5.582115517529114e-06,
"loss": 0.0439,
"step": 542
},
{
"epoch": 0.9819168173598554,
"grad_norm": 0.447050541639328,
"learning_rate": 5.5672547012499e-06,
"loss": 0.0338,
"step": 543
},
{
"epoch": 0.9837251356238698,
"grad_norm": 0.392455130815506,
"learning_rate": 5.55238880689783e-06,
"loss": 0.014,
"step": 544
},
{
"epoch": 0.9855334538878843,
"grad_norm": 0.49793773889541626,
"learning_rate": 5.537517967552626e-06,
"loss": 0.0094,
"step": 545
},
{
"epoch": 0.9873417721518988,
"grad_norm": 0.8518772721290588,
"learning_rate": 5.522642316338268e-06,
"loss": 0.0413,
"step": 546
},
{
"epoch": 0.9891500904159132,
"grad_norm": 0.814775824546814,
"learning_rate": 5.507761986421818e-06,
"loss": 0.0515,
"step": 547
},
{
"epoch": 0.9909584086799277,
"grad_norm": 0.2275906503200531,
"learning_rate": 5.4928771110122185e-06,
"loss": 0.0048,
"step": 548
},
{
"epoch": 0.9927667269439421,
"grad_norm": 0.5201494693756104,
"learning_rate": 5.477987823359104e-06,
"loss": 0.0253,
"step": 549
},
{
"epoch": 0.9945750452079566,
"grad_norm": 0.4527904689311981,
"learning_rate": 5.463094256751608e-06,
"loss": 0.0188,
"step": 550
},
{
"epoch": 0.9963833634719711,
"grad_norm": 0.4992627501487732,
"learning_rate": 5.448196544517168e-06,
"loss": 0.0252,
"step": 551
},
{
"epoch": 0.9981916817359855,
"grad_norm": 0.5513089299201965,
"learning_rate": 5.433294820020335e-06,
"loss": 0.0263,
"step": 552
},
{
"epoch": 1.0,
"grad_norm": 0.49543046951293945,
"learning_rate": 5.41838921666158e-06,
"loss": 0.0217,
"step": 553
},
{
"epoch": 1.0018083182640145,
"grad_norm": 0.52801513671875,
"learning_rate": 5.403479867876087e-06,
"loss": 0.0246,
"step": 554
},
{
"epoch": 1.003616636528029,
"grad_norm": 0.46472352743148804,
"learning_rate": 5.388566907132583e-06,
"loss": 0.0181,
"step": 555
},
{
"epoch": 1.0054249547920433,
"grad_norm": 0.4677007794380188,
"learning_rate": 5.373650467932122e-06,
"loss": 0.0264,
"step": 556
},
{
"epoch": 1.0072332730560578,
"grad_norm": 0.873718798160553,
"learning_rate": 5.358730683806897e-06,
"loss": 0.0448,
"step": 557
},
{
"epoch": 1.0090415913200723,
"grad_norm": 0.5990148186683655,
"learning_rate": 5.343807688319047e-06,
"loss": 0.0356,
"step": 558
},
{
"epoch": 1.0108499095840868,
"grad_norm": 0.6841799020767212,
"learning_rate": 5.328881615059458e-06,
"loss": 0.0526,
"step": 559
},
{
"epoch": 1.0126582278481013,
"grad_norm": 0.47151437401771545,
"learning_rate": 5.3139525976465675e-06,
"loss": 0.0273,
"step": 560
},
{
"epoch": 1.0144665461121158,
"grad_norm": 0.4732193648815155,
"learning_rate": 5.299020769725172e-06,
"loss": 0.0114,
"step": 561
},
{
"epoch": 1.0162748643761301,
"grad_norm": 0.5929259657859802,
"learning_rate": 5.284086264965224e-06,
"loss": 0.0237,
"step": 562
},
{
"epoch": 1.0180831826401446,
"grad_norm": 0.726356565952301,
"learning_rate": 5.269149217060642e-06,
"loss": 0.0421,
"step": 563
},
{
"epoch": 1.0198915009041591,
"grad_norm": 0.6595012545585632,
"learning_rate": 5.2542097597281095e-06,
"loss": 0.0124,
"step": 564
},
{
"epoch": 1.0216998191681737,
"grad_norm": 0.49609890580177307,
"learning_rate": 5.239268026705878e-06,
"loss": 0.0213,
"step": 565
},
{
"epoch": 1.0235081374321882,
"grad_norm": 0.3672851622104645,
"learning_rate": 5.224324151752575e-06,
"loss": 0.0241,
"step": 566
},
{
"epoch": 1.0253164556962024,
"grad_norm": 0.575946569442749,
"learning_rate": 5.209378268645998e-06,
"loss": 0.0336,
"step": 567
},
{
"epoch": 1.027124773960217,
"grad_norm": 0.37768223881721497,
"learning_rate": 5.194430511181925e-06,
"loss": 0.028,
"step": 568
},
{
"epoch": 1.0289330922242315,
"grad_norm": 0.5540050268173218,
"learning_rate": 5.179481013172912e-06,
"loss": 0.0417,
"step": 569
},
{
"epoch": 1.030741410488246,
"grad_norm": 0.2120361328125,
"learning_rate": 5.1645299084470936e-06,
"loss": 0.0057,
"step": 570
},
{
"epoch": 1.0325497287522605,
"grad_norm": 1.054090142250061,
"learning_rate": 5.1495773308469935e-06,
"loss": 0.0399,
"step": 571
},
{
"epoch": 1.0343580470162748,
"grad_norm": 0.6278601884841919,
"learning_rate": 5.134623414228315e-06,
"loss": 0.0339,
"step": 572
},
{
"epoch": 1.0361663652802893,
"grad_norm": 0.660631537437439,
"learning_rate": 5.119668292458751e-06,
"loss": 0.0271,
"step": 573
},
{
"epoch": 1.0379746835443038,
"grad_norm": 0.39478397369384766,
"learning_rate": 5.1047120994167855e-06,
"loss": 0.0162,
"step": 574
},
{
"epoch": 1.0397830018083183,
"grad_norm": 0.3979175090789795,
"learning_rate": 5.0897549689904865e-06,
"loss": 0.0289,
"step": 575
},
{
"epoch": 1.0415913200723328,
"grad_norm": 0.6313123106956482,
"learning_rate": 5.074797035076319e-06,
"loss": 0.0358,
"step": 576
},
{
"epoch": 1.0433996383363473,
"grad_norm": 0.4100612699985504,
"learning_rate": 5.059838431577937e-06,
"loss": 0.0137,
"step": 577
},
{
"epoch": 1.0452079566003616,
"grad_norm": 0.3369003236293793,
"learning_rate": 5.04487929240499e-06,
"loss": 0.0118,
"step": 578
},
{
"epoch": 1.047016274864376,
"grad_norm": 0.41355615854263306,
"learning_rate": 5.029919751471925e-06,
"loss": 0.0083,
"step": 579
},
{
"epoch": 1.0488245931283906,
"grad_norm": 0.6382481455802917,
"learning_rate": 5.014959942696782e-06,
"loss": 0.0184,
"step": 580
},
{
"epoch": 1.0506329113924051,
"grad_norm": 0.5917233824729919,
"learning_rate": 5e-06,
"loss": 0.0245,
"step": 581
},
{
"epoch": 1.0524412296564196,
"grad_norm": 0.7864559888839722,
"learning_rate": 4.98504005730322e-06,
"loss": 0.0292,
"step": 582
},
{
"epoch": 1.054249547920434,
"grad_norm": 0.7384347915649414,
"learning_rate": 4.970080248528077e-06,
"loss": 0.0346,
"step": 583
},
{
"epoch": 1.0560578661844484,
"grad_norm": 0.45373573899269104,
"learning_rate": 4.955120707595011e-06,
"loss": 0.0219,
"step": 584
},
{
"epoch": 1.057866184448463,
"grad_norm": 0.8453214168548584,
"learning_rate": 4.940161568422065e-06,
"loss": 0.0407,
"step": 585
},
{
"epoch": 1.0596745027124774,
"grad_norm": 0.9157655835151672,
"learning_rate": 4.9252029649236835e-06,
"loss": 0.0347,
"step": 586
},
{
"epoch": 1.061482820976492,
"grad_norm": 0.6446845531463623,
"learning_rate": 4.910245031009515e-06,
"loss": 0.0315,
"step": 587
},
{
"epoch": 1.0632911392405062,
"grad_norm": 0.18100093305110931,
"learning_rate": 4.895287900583216e-06,
"loss": 0.0054,
"step": 588
},
{
"epoch": 1.0650994575045207,
"grad_norm": 0.680534839630127,
"learning_rate": 4.88033170754125e-06,
"loss": 0.0294,
"step": 589
},
{
"epoch": 1.0669077757685352,
"grad_norm": 0.8065388798713684,
"learning_rate": 4.865376585771687e-06,
"loss": 0.0346,
"step": 590
},
{
"epoch": 1.0687160940325497,
"grad_norm": 0.45650598406791687,
"learning_rate": 4.850422669153009e-06,
"loss": 0.0123,
"step": 591
},
{
"epoch": 1.0705244122965643,
"grad_norm": 0.559516191482544,
"learning_rate": 4.835470091552906e-06,
"loss": 0.0326,
"step": 592
},
{
"epoch": 1.0723327305605788,
"grad_norm": 0.48694777488708496,
"learning_rate": 4.8205189868270894e-06,
"loss": 0.0313,
"step": 593
},
{
"epoch": 1.074141048824593,
"grad_norm": 0.2869080603122711,
"learning_rate": 4.805569488818077e-06,
"loss": 0.0071,
"step": 594
},
{
"epoch": 1.0759493670886076,
"grad_norm": 0.681262731552124,
"learning_rate": 4.7906217313540035e-06,
"loss": 0.033,
"step": 595
},
{
"epoch": 1.077757685352622,
"grad_norm": 0.3760509192943573,
"learning_rate": 4.775675848247427e-06,
"loss": 0.02,
"step": 596
},
{
"epoch": 1.0795660036166366,
"grad_norm": 0.6613526940345764,
"learning_rate": 4.760731973294122e-06,
"loss": 0.0197,
"step": 597
},
{
"epoch": 1.081374321880651,
"grad_norm": 0.3093162477016449,
"learning_rate": 4.745790240271892e-06,
"loss": 0.0131,
"step": 598
},
{
"epoch": 1.0831826401446654,
"grad_norm": 0.23930813372135162,
"learning_rate": 4.73085078293936e-06,
"loss": 0.0151,
"step": 599
},
{
"epoch": 1.0849909584086799,
"grad_norm": 0.21967868506908417,
"learning_rate": 4.715913735034779e-06,
"loss": 0.0065,
"step": 600
},
{
"epoch": 1.0867992766726944,
"grad_norm": 0.32417380809783936,
"learning_rate": 4.700979230274829e-06,
"loss": 0.012,
"step": 601
},
{
"epoch": 1.0886075949367089,
"grad_norm": 0.4261423647403717,
"learning_rate": 4.686047402353433e-06,
"loss": 0.0168,
"step": 602
},
{
"epoch": 1.0904159132007234,
"grad_norm": 0.31165367364883423,
"learning_rate": 4.6711183849405435e-06,
"loss": 0.0117,
"step": 603
},
{
"epoch": 1.092224231464738,
"grad_norm": 0.4939417243003845,
"learning_rate": 4.6561923116809545e-06,
"loss": 0.0222,
"step": 604
},
{
"epoch": 1.0940325497287522,
"grad_norm": 0.49815595149993896,
"learning_rate": 4.641269316193104e-06,
"loss": 0.0314,
"step": 605
},
{
"epoch": 1.0958408679927667,
"grad_norm": 0.4158225357532501,
"learning_rate": 4.626349532067879e-06,
"loss": 0.0146,
"step": 606
},
{
"epoch": 1.0976491862567812,
"grad_norm": 0.37719643115997314,
"learning_rate": 4.611433092867417e-06,
"loss": 0.0202,
"step": 607
},
{
"epoch": 1.0994575045207957,
"grad_norm": 0.48115360736846924,
"learning_rate": 4.596520132123915e-06,
"loss": 0.0181,
"step": 608
},
{
"epoch": 1.1012658227848102,
"grad_norm": 0.8638368248939514,
"learning_rate": 4.581610783338424e-06,
"loss": 0.035,
"step": 609
},
{
"epoch": 1.1030741410488245,
"grad_norm": 0.6068063974380493,
"learning_rate": 4.566705179979665e-06,
"loss": 0.0207,
"step": 610
},
{
"epoch": 1.104882459312839,
"grad_norm": 0.4472132921218872,
"learning_rate": 4.551803455482833e-06,
"loss": 0.0204,
"step": 611
},
{
"epoch": 1.1066907775768535,
"grad_norm": 1.0753980875015259,
"learning_rate": 4.536905743248394e-06,
"loss": 0.0485,
"step": 612
},
{
"epoch": 1.108499095840868,
"grad_norm": 0.45651987195014954,
"learning_rate": 4.522012176640898e-06,
"loss": 0.0296,
"step": 613
},
{
"epoch": 1.1103074141048825,
"grad_norm": 0.5501109957695007,
"learning_rate": 4.507122888987782e-06,
"loss": 0.0278,
"step": 614
},
{
"epoch": 1.1121157323688968,
"grad_norm": 0.40733832120895386,
"learning_rate": 4.4922380135781835e-06,
"loss": 0.0222,
"step": 615
},
{
"epoch": 1.1139240506329113,
"grad_norm": 0.9684047102928162,
"learning_rate": 4.477357683661734e-06,
"loss": 0.0367,
"step": 616
},
{
"epoch": 1.1157323688969258,
"grad_norm": 0.46146684885025024,
"learning_rate": 4.462482032447377e-06,
"loss": 0.0179,
"step": 617
},
{
"epoch": 1.1175406871609403,
"grad_norm": 0.4387267529964447,
"learning_rate": 4.447611193102171e-06,
"loss": 0.0184,
"step": 618
},
{
"epoch": 1.1193490054249549,
"grad_norm": 1.1471061706542969,
"learning_rate": 4.432745298750102e-06,
"loss": 0.0253,
"step": 619
},
{
"epoch": 1.1211573236889691,
"grad_norm": 0.4193446636199951,
"learning_rate": 4.417884482470887e-06,
"loss": 0.0113,
"step": 620
},
{
"epoch": 1.1229656419529837,
"grad_norm": 0.34244582056999207,
"learning_rate": 4.4030288772987795e-06,
"loss": 0.0069,
"step": 621
},
{
"epoch": 1.1247739602169982,
"grad_norm": 0.7945204973220825,
"learning_rate": 4.388178616221389e-06,
"loss": 0.0452,
"step": 622
},
{
"epoch": 1.1265822784810127,
"grad_norm": 0.5046751499176025,
"learning_rate": 4.373333832178478e-06,
"loss": 0.0303,
"step": 623
},
{
"epoch": 1.1283905967450272,
"grad_norm": 0.14286679029464722,
"learning_rate": 4.35849465806079e-06,
"loss": 0.0037,
"step": 624
},
{
"epoch": 1.1301989150090417,
"grad_norm": 0.3736821413040161,
"learning_rate": 4.3436612267088355e-06,
"loss": 0.0077,
"step": 625
},
{
"epoch": 1.132007233273056,
"grad_norm": 0.4808909595012665,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.0185,
"step": 626
},
{
"epoch": 1.1338155515370705,
"grad_norm": 0.4397783875465393,
"learning_rate": 4.314012123405961e-06,
"loss": 0.0205,
"step": 627
},
{
"epoch": 1.135623869801085,
"grad_norm": 0.6529375314712524,
"learning_rate": 4.299196716874271e-06,
"loss": 0.0339,
"step": 628
},
{
"epoch": 1.1374321880650995,
"grad_norm": 0.5052587985992432,
"learning_rate": 4.284387583944403e-06,
"loss": 0.0203,
"step": 629
},
{
"epoch": 1.139240506329114,
"grad_norm": 0.2759120464324951,
"learning_rate": 4.269584857187942e-06,
"loss": 0.0161,
"step": 630
},
{
"epoch": 1.1410488245931285,
"grad_norm": 0.6620070338249207,
"learning_rate": 4.254788669119127e-06,
"loss": 0.0289,
"step": 631
},
{
"epoch": 1.1428571428571428,
"grad_norm": 0.5193249583244324,
"learning_rate": 4.239999152193664e-06,
"loss": 0.0301,
"step": 632
},
{
"epoch": 1.1446654611211573,
"grad_norm": 0.28195878863334656,
"learning_rate": 4.225216438807536e-06,
"loss": 0.0083,
"step": 633
},
{
"epoch": 1.1464737793851718,
"grad_norm": 0.7462310791015625,
"learning_rate": 4.2104406612958216e-06,
"loss": 0.0498,
"step": 634
},
{
"epoch": 1.1482820976491863,
"grad_norm": 0.4901484251022339,
"learning_rate": 4.195671951931509e-06,
"loss": 0.0278,
"step": 635
},
{
"epoch": 1.1500904159132008,
"grad_norm": 0.3175576627254486,
"learning_rate": 4.180910442924312e-06,
"loss": 0.0152,
"step": 636
},
{
"epoch": 1.1518987341772151,
"grad_norm": 0.4822959899902344,
"learning_rate": 4.166156266419489e-06,
"loss": 0.0155,
"step": 637
},
{
"epoch": 1.1537070524412296,
"grad_norm": 0.5162426829338074,
"learning_rate": 4.1514095544966556e-06,
"loss": 0.019,
"step": 638
},
{
"epoch": 1.1555153707052441,
"grad_norm": 0.66053706407547,
"learning_rate": 4.136670439168605e-06,
"loss": 0.0221,
"step": 639
},
{
"epoch": 1.1573236889692586,
"grad_norm": 0.4949704706668854,
"learning_rate": 4.121939052380125e-06,
"loss": 0.0239,
"step": 640
},
{
"epoch": 1.1591320072332731,
"grad_norm": 1.0032978057861328,
"learning_rate": 4.107215526006818e-06,
"loss": 0.0171,
"step": 641
},
{
"epoch": 1.1609403254972874,
"grad_norm": 0.6254719495773315,
"learning_rate": 4.092499991853919e-06,
"loss": 0.0409,
"step": 642
},
{
"epoch": 1.162748643761302,
"grad_norm": 0.9865067005157471,
"learning_rate": 4.0777925816551175e-06,
"loss": 0.0549,
"step": 643
},
{
"epoch": 1.1645569620253164,
"grad_norm": 0.41432106494903564,
"learning_rate": 4.063093427071376e-06,
"loss": 0.0156,
"step": 644
},
{
"epoch": 1.166365280289331,
"grad_norm": 0.5362163782119751,
"learning_rate": 4.048402659689757e-06,
"loss": 0.0267,
"step": 645
},
{
"epoch": 1.1681735985533455,
"grad_norm": 0.5830875039100647,
"learning_rate": 4.033720411022235e-06,
"loss": 0.031,
"step": 646
},
{
"epoch": 1.1699819168173597,
"grad_norm": 0.6136839389801025,
"learning_rate": 4.019046812504526e-06,
"loss": 0.0429,
"step": 647
},
{
"epoch": 1.1717902350813743,
"grad_norm": 0.40732574462890625,
"learning_rate": 4.0043819954949105e-06,
"loss": 0.0259,
"step": 648
},
{
"epoch": 1.1735985533453888,
"grad_norm": 0.24956393241882324,
"learning_rate": 3.989726091273056e-06,
"loss": 0.0067,
"step": 649
},
{
"epoch": 1.1754068716094033,
"grad_norm": 0.4917099177837372,
"learning_rate": 3.975079231038848e-06,
"loss": 0.0146,
"step": 650
},
{
"epoch": 1.1772151898734178,
"grad_norm": 0.5232261419296265,
"learning_rate": 3.960441545911205e-06,
"loss": 0.0189,
"step": 651
},
{
"epoch": 1.179023508137432,
"grad_norm": 0.343420147895813,
"learning_rate": 3.9458131669269066e-06,
"loss": 0.0099,
"step": 652
},
{
"epoch": 1.1808318264014466,
"grad_norm": 0.4799347221851349,
"learning_rate": 3.931194225039427e-06,
"loss": 0.018,
"step": 653
},
{
"epoch": 1.182640144665461,
"grad_norm": 0.2552357017993927,
"learning_rate": 3.916584851117766e-06,
"loss": 0.0115,
"step": 654
},
{
"epoch": 1.1844484629294756,
"grad_norm": 0.22517123818397522,
"learning_rate": 3.90198517594526e-06,
"loss": 0.0098,
"step": 655
},
{
"epoch": 1.18625678119349,
"grad_norm": 0.23926715552806854,
"learning_rate": 3.887395330218429e-06,
"loss": 0.0055,
"step": 656
},
{
"epoch": 1.1880650994575046,
"grad_norm": 0.6009202003479004,
"learning_rate": 3.872815444545794e-06,
"loss": 0.029,
"step": 657
},
{
"epoch": 1.189873417721519,
"grad_norm": 0.5351443886756897,
"learning_rate": 3.8582456494467214e-06,
"loss": 0.0202,
"step": 658
},
{
"epoch": 1.1916817359855334,
"grad_norm": 0.6078570485115051,
"learning_rate": 3.843686075350239e-06,
"loss": 0.0317,
"step": 659
},
{
"epoch": 1.193490054249548,
"grad_norm": 0.36958634853363037,
"learning_rate": 3.829136852593881e-06,
"loss": 0.0182,
"step": 660
},
{
"epoch": 1.1952983725135624,
"grad_norm": 0.7101830244064331,
"learning_rate": 3.8145981114225135e-06,
"loss": 0.0452,
"step": 661
},
{
"epoch": 1.197106690777577,
"grad_norm": 0.5995045900344849,
"learning_rate": 3.8000699819871704e-06,
"loss": 0.0304,
"step": 662
},
{
"epoch": 1.1989150090415914,
"grad_norm": 0.7885802388191223,
"learning_rate": 3.785552594343894e-06,
"loss": 0.0313,
"step": 663
},
{
"epoch": 1.2007233273056057,
"grad_norm": 0.4450237452983856,
"learning_rate": 3.7710460784525617e-06,
"loss": 0.0118,
"step": 664
},
{
"epoch": 1.2025316455696202,
"grad_norm": 0.26192954182624817,
"learning_rate": 3.756550564175727e-06,
"loss": 0.0079,
"step": 665
},
{
"epoch": 1.2043399638336347,
"grad_norm": 0.5684515833854675,
"learning_rate": 3.7420661812774577e-06,
"loss": 0.024,
"step": 666
},
{
"epoch": 1.2061482820976492,
"grad_norm": 0.6785829663276672,
"learning_rate": 3.7275930594221752e-06,
"loss": 0.0355,
"step": 667
},
{
"epoch": 1.2079566003616637,
"grad_norm": 0.4799443781375885,
"learning_rate": 3.7131313281734895e-06,
"loss": 0.0297,
"step": 668
},
{
"epoch": 1.209764918625678,
"grad_norm": 0.3397943675518036,
"learning_rate": 3.69868111699304e-06,
"loss": 0.0155,
"step": 669
},
{
"epoch": 1.2115732368896925,
"grad_norm": 0.8505467772483826,
"learning_rate": 3.6842425552393424e-06,
"loss": 0.0299,
"step": 670
},
{
"epoch": 1.213381555153707,
"grad_norm": 0.6407603025436401,
"learning_rate": 3.669815772166625e-06,
"loss": 0.0245,
"step": 671
},
{
"epoch": 1.2151898734177216,
"grad_norm": 0.4153575003147125,
"learning_rate": 3.655400896923672e-06,
"loss": 0.0184,
"step": 672
},
{
"epoch": 1.216998191681736,
"grad_norm": 0.33345305919647217,
"learning_rate": 3.6409980585526707e-06,
"loss": 0.0144,
"step": 673
},
{
"epoch": 1.2188065099457503,
"grad_norm": 0.12314804643392563,
"learning_rate": 3.62660738598805e-06,
"loss": 0.0037,
"step": 674
},
{
"epoch": 1.2206148282097649,
"grad_norm": 0.18858946859836578,
"learning_rate": 3.6122290080553334e-06,
"loss": 0.0049,
"step": 675
},
{
"epoch": 1.2224231464737794,
"grad_norm": 0.7088531851768494,
"learning_rate": 3.5978630534699873e-06,
"loss": 0.0437,
"step": 676
},
{
"epoch": 1.2242314647377939,
"grad_norm": 0.9003376960754395,
"learning_rate": 3.583509650836254e-06,
"loss": 0.0282,
"step": 677
},
{
"epoch": 1.2260397830018084,
"grad_norm": 1.0504286289215088,
"learning_rate": 3.5691689286460172e-06,
"loss": 0.0576,
"step": 678
},
{
"epoch": 1.2278481012658227,
"grad_norm": 0.20466826856136322,
"learning_rate": 3.5548410152776414e-06,
"loss": 0.0065,
"step": 679
},
{
"epoch": 1.2296564195298372,
"grad_norm": 0.2663809061050415,
"learning_rate": 3.540526038994834e-06,
"loss": 0.0066,
"step": 680
},
{
"epoch": 1.2314647377938517,
"grad_norm": 0.6395544409751892,
"learning_rate": 3.526224127945479e-06,
"loss": 0.0266,
"step": 681
},
{
"epoch": 1.2332730560578662,
"grad_norm": 0.45467811822891235,
"learning_rate": 3.5119354101605086e-06,
"loss": 0.0223,
"step": 682
},
{
"epoch": 1.2350813743218807,
"grad_norm": 0.5682668089866638,
"learning_rate": 3.4976600135527403e-06,
"loss": 0.0236,
"step": 683
},
{
"epoch": 1.2368896925858952,
"grad_norm": 0.7187696695327759,
"learning_rate": 3.4833980659157507e-06,
"loss": 0.0377,
"step": 684
},
{
"epoch": 1.2386980108499095,
"grad_norm": 0.42689749598503113,
"learning_rate": 3.469149694922712e-06,
"loss": 0.029,
"step": 685
},
{
"epoch": 1.240506329113924,
"grad_norm": 0.16523726284503937,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.0065,
"step": 686
},
{
"epoch": 1.2423146473779385,
"grad_norm": 0.37031322717666626,
"learning_rate": 3.4406941929523607e-06,
"loss": 0.0152,
"step": 687
},
{
"epoch": 1.244122965641953,
"grad_norm": 0.4927110970020294,
"learning_rate": 3.4264873167091405e-06,
"loss": 0.034,
"step": 688
},
{
"epoch": 1.2459312839059675,
"grad_norm": 0.7410340309143066,
"learning_rate": 3.412294526575779e-06,
"loss": 0.0169,
"step": 689
},
{
"epoch": 1.247739602169982,
"grad_norm": 1.0170695781707764,
"learning_rate": 3.398115949606352e-06,
"loss": 0.04,
"step": 690
},
{
"epoch": 1.2495479204339963,
"grad_norm": 0.6794872879981995,
"learning_rate": 3.383951712727701e-06,
"loss": 0.0253,
"step": 691
},
{
"epoch": 1.2513562386980108,
"grad_norm": 0.7767112255096436,
"learning_rate": 3.3698019427382912e-06,
"loss": 0.0434,
"step": 692
},
{
"epoch": 1.2531645569620253,
"grad_norm": 0.7374671697616577,
"learning_rate": 3.355666766307084e-06,
"loss": 0.041,
"step": 693
},
{
"epoch": 1.2549728752260398,
"grad_norm": 0.5744323134422302,
"learning_rate": 3.341546309972398e-06,
"loss": 0.017,
"step": 694
},
{
"epoch": 1.2567811934900543,
"grad_norm": 0.3930213749408722,
"learning_rate": 3.327440700140774e-06,
"loss": 0.0216,
"step": 695
},
{
"epoch": 1.2585895117540686,
"grad_norm": 0.30312588810920715,
"learning_rate": 3.3133500630858507e-06,
"loss": 0.0124,
"step": 696
},
{
"epoch": 1.2603978300180831,
"grad_norm": 0.5649741888046265,
"learning_rate": 3.299274524947229e-06,
"loss": 0.0224,
"step": 697
},
{
"epoch": 1.2622061482820977,
"grad_norm": 0.6171383261680603,
"learning_rate": 3.2852142117293435e-06,
"loss": 0.0196,
"step": 698
},
{
"epoch": 1.2640144665461122,
"grad_norm": 0.5959765911102295,
"learning_rate": 3.2711692493003357e-06,
"loss": 0.0217,
"step": 699
},
{
"epoch": 1.2658227848101267,
"grad_norm": 0.44579416513442993,
"learning_rate": 3.2571397633909252e-06,
"loss": 0.0209,
"step": 700
},
{
"epoch": 1.267631103074141,
"grad_norm": 0.7416233420372009,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.0439,
"step": 701
},
{
"epoch": 1.2694394213381555,
"grad_norm": 0.31468456983566284,
"learning_rate": 3.229127723359927e-06,
"loss": 0.0104,
"step": 702
},
{
"epoch": 1.27124773960217,
"grad_norm": 0.2581050395965576,
"learning_rate": 3.215145420002555e-06,
"loss": 0.0075,
"step": 703
},
{
"epoch": 1.2730560578661845,
"grad_norm": 0.4038104712963104,
"learning_rate": 3.2011790946909673e-06,
"loss": 0.0147,
"step": 704
},
{
"epoch": 1.274864376130199,
"grad_norm": 0.513434112071991,
"learning_rate": 3.1872288724519207e-06,
"loss": 0.0274,
"step": 705
},
{
"epoch": 1.2766726943942133,
"grad_norm": 1.2186568975448608,
"learning_rate": 3.173294878168025e-06,
"loss": 0.0339,
"step": 706
},
{
"epoch": 1.2784810126582278,
"grad_norm": 0.5063746571540833,
"learning_rate": 3.1593772365766107e-06,
"loss": 0.0289,
"step": 707
},
{
"epoch": 1.2802893309222423,
"grad_norm": 0.3350663185119629,
"learning_rate": 3.1454760722686206e-06,
"loss": 0.019,
"step": 708
},
{
"epoch": 1.2820976491862568,
"grad_norm": 0.6459165215492249,
"learning_rate": 3.1315915096874894e-06,
"loss": 0.0148,
"step": 709
},
{
"epoch": 1.2839059674502713,
"grad_norm": 0.8068355321884155,
"learning_rate": 3.11772367312804e-06,
"loss": 0.0523,
"step": 710
},
{
"epoch": 1.2857142857142856,
"grad_norm": 0.5673767328262329,
"learning_rate": 3.1038726867353587e-06,
"loss": 0.0292,
"step": 711
},
{
"epoch": 1.2875226039783003,
"grad_norm": 0.6073427200317383,
"learning_rate": 3.090038674503688e-06,
"loss": 0.02,
"step": 712
},
{
"epoch": 1.2893309222423146,
"grad_norm": 0.7443808317184448,
"learning_rate": 3.076221760275321e-06,
"loss": 0.0527,
"step": 713
},
{
"epoch": 1.2911392405063291,
"grad_norm": 0.6055053472518921,
"learning_rate": 3.0624220677394854e-06,
"loss": 0.0255,
"step": 714
},
{
"epoch": 1.2929475587703436,
"grad_norm": 0.19690106809139252,
"learning_rate": 3.048639720431244e-06,
"loss": 0.0042,
"step": 715
},
{
"epoch": 1.2947558770343581,
"grad_norm": 0.5978537201881409,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.0276,
"step": 716
},
{
"epoch": 1.2965641952983726,
"grad_norm": 0.34526073932647705,
"learning_rate": 3.0211275548603076e-06,
"loss": 0.0093,
"step": 717
},
{
"epoch": 1.298372513562387,
"grad_norm": 0.3630492687225342,
"learning_rate": 3.007397982886942e-06,
"loss": 0.0142,
"step": 718
},
{
"epoch": 1.3001808318264014,
"grad_norm": 0.6456695795059204,
"learning_rate": 2.9936862487176295e-06,
"loss": 0.0216,
"step": 719
},
{
"epoch": 1.301989150090416,
"grad_norm": 0.14279302954673767,
"learning_rate": 2.979992475100024e-06,
"loss": 0.0038,
"step": 720
},
{
"epoch": 1.3037974683544304,
"grad_norm": 0.49573206901550293,
"learning_rate": 2.966316784621e-06,
"loss": 0.0231,
"step": 721
},
{
"epoch": 1.305605786618445,
"grad_norm": 0.4423304498195648,
"learning_rate": 2.9526592997055488e-06,
"loss": 0.0262,
"step": 722
},
{
"epoch": 1.3074141048824592,
"grad_norm": 0.15511740744113922,
"learning_rate": 2.9390201426156855e-06,
"loss": 0.0031,
"step": 723
},
{
"epoch": 1.3092224231464737,
"grad_norm": 0.34155717492103577,
"learning_rate": 2.9253994354493575e-06,
"loss": 0.0087,
"step": 724
},
{
"epoch": 1.3110307414104883,
"grad_norm": 0.4530353248119354,
"learning_rate": 2.911797300139345e-06,
"loss": 0.022,
"step": 725
},
{
"epoch": 1.3128390596745028,
"grad_norm": 0.7730844616889954,
"learning_rate": 2.8982138584521734e-06,
"loss": 0.0448,
"step": 726
},
{
"epoch": 1.3146473779385173,
"grad_norm": 0.3586406409740448,
"learning_rate": 2.884649231987021e-06,
"loss": 0.0102,
"step": 727
},
{
"epoch": 1.3164556962025316,
"grad_norm": 0.4699344336986542,
"learning_rate": 2.871103542174637e-06,
"loss": 0.0297,
"step": 728
},
{
"epoch": 1.318264014466546,
"grad_norm": 0.5743511319160461,
"learning_rate": 2.857576910276243e-06,
"loss": 0.0281,
"step": 729
},
{
"epoch": 1.3200723327305606,
"grad_norm": 0.2710094749927521,
"learning_rate": 2.844069457382459e-06,
"loss": 0.0066,
"step": 730
},
{
"epoch": 1.321880650994575,
"grad_norm": 0.4629327952861786,
"learning_rate": 2.83058130441221e-06,
"loss": 0.0264,
"step": 731
},
{
"epoch": 1.3236889692585896,
"grad_norm": 0.3657383620738983,
"learning_rate": 2.817112572111651e-06,
"loss": 0.0215,
"step": 732
},
{
"epoch": 1.3254972875226039,
"grad_norm": 0.38295358419418335,
"learning_rate": 2.803663381053081e-06,
"loss": 0.0152,
"step": 733
},
{
"epoch": 1.3273056057866184,
"grad_norm": 0.5653232336044312,
"learning_rate": 2.790233851633868e-06,
"loss": 0.0308,
"step": 734
},
{
"epoch": 1.3291139240506329,
"grad_norm": 0.8704638481140137,
"learning_rate": 2.776824104075364e-06,
"loss": 0.0365,
"step": 735
},
{
"epoch": 1.3309222423146474,
"grad_norm": 0.42136698961257935,
"learning_rate": 2.7634342584218364e-06,
"loss": 0.0219,
"step": 736
},
{
"epoch": 1.332730560578662,
"grad_norm": 0.40870401263237,
"learning_rate": 2.7500644345393945e-06,
"loss": 0.0237,
"step": 737
},
{
"epoch": 1.3345388788426762,
"grad_norm": 0.3043314516544342,
"learning_rate": 2.7367147521149052e-06,
"loss": 0.0095,
"step": 738
},
{
"epoch": 1.3363471971066907,
"grad_norm": 0.5929886698722839,
"learning_rate": 2.723385330654933e-06,
"loss": 0.0318,
"step": 739
},
{
"epoch": 1.3381555153707052,
"grad_norm": 0.7208443284034729,
"learning_rate": 2.7100762894846633e-06,
"loss": 0.0487,
"step": 740
},
{
"epoch": 1.3399638336347197,
"grad_norm": 4.5319671630859375,
"learning_rate": 2.6967877477468394e-06,
"loss": 0.0248,
"step": 741
},
{
"epoch": 1.3417721518987342,
"grad_norm": 0.5836402773857117,
"learning_rate": 2.683519824400693e-06,
"loss": 0.0338,
"step": 742
},
{
"epoch": 1.3435804701627485,
"grad_norm": 0.10844586789608002,
"learning_rate": 2.6702726382208775e-06,
"loss": 0.0027,
"step": 743
},
{
"epoch": 1.3453887884267632,
"grad_norm": 0.4178289473056793,
"learning_rate": 2.657046307796407e-06,
"loss": 0.0228,
"step": 744
},
{
"epoch": 1.3471971066907775,
"grad_norm": 0.47189414501190186,
"learning_rate": 2.6438409515295997e-06,
"loss": 0.0265,
"step": 745
},
{
"epoch": 1.349005424954792,
"grad_norm": 0.4005834758281708,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.019,
"step": 746
},
{
"epoch": 1.3508137432188065,
"grad_norm": 0.3166942596435547,
"learning_rate": 2.617493634138363e-06,
"loss": 0.0061,
"step": 747
},
{
"epoch": 1.352622061482821,
"grad_norm": 0.22715604305267334,
"learning_rate": 2.6043519088755263e-06,
"loss": 0.0038,
"step": 748
},
{
"epoch": 1.3544303797468356,
"grad_norm": 0.7108785510063171,
"learning_rate": 2.5912316294914232e-06,
"loss": 0.037,
"step": 749
},
{
"epoch": 1.3562386980108498,
"grad_norm": 0.45363155007362366,
"learning_rate": 2.578132913439e-06,
"loss": 0.0245,
"step": 750
},
{
"epoch": 1.3580470162748643,
"grad_norm": 0.2980150282382965,
"learning_rate": 2.5650558779781635e-06,
"loss": 0.0114,
"step": 751
},
{
"epoch": 1.3598553345388789,
"grad_norm": 0.5413243770599365,
"learning_rate": 2.55200064017474e-06,
"loss": 0.0326,
"step": 752
},
{
"epoch": 1.3616636528028934,
"grad_norm": 0.39414241909980774,
"learning_rate": 2.538967316899414e-06,
"loss": 0.0163,
"step": 753
},
{
"epoch": 1.3634719710669079,
"grad_norm": 0.18338921666145325,
"learning_rate": 2.5259560248267022e-06,
"loss": 0.009,
"step": 754
},
{
"epoch": 1.3652802893309222,
"grad_norm": 0.5580337643623352,
"learning_rate": 2.512966880433891e-06,
"loss": 0.0202,
"step": 755
},
{
"epoch": 1.3670886075949367,
"grad_norm": 0.5050540566444397,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.0159,
"step": 756
},
{
"epoch": 1.3688969258589512,
"grad_norm": 0.4926493167877197,
"learning_rate": 2.4870554996047454e-06,
"loss": 0.0263,
"step": 757
},
{
"epoch": 1.3707052441229657,
"grad_norm": 0.40215957164764404,
"learning_rate": 2.4741334951274948e-06,
"loss": 0.0137,
"step": 758
},
{
"epoch": 1.3725135623869802,
"grad_norm": 0.6963564157485962,
"learning_rate": 2.461234102246231e-06,
"loss": 0.0396,
"step": 759
},
{
"epoch": 1.3743218806509945,
"grad_norm": 0.4896616041660309,
"learning_rate": 2.448357436436519e-06,
"loss": 0.0238,
"step": 760
},
{
"epoch": 1.376130198915009,
"grad_norm": 0.36359885334968567,
"learning_rate": 2.43550361297047e-06,
"loss": 0.0183,
"step": 761
},
{
"epoch": 1.3779385171790235,
"grad_norm": 0.4107934832572937,
"learning_rate": 2.4226727469157097e-06,
"loss": 0.0245,
"step": 762
},
{
"epoch": 1.379746835443038,
"grad_norm": 0.598741352558136,
"learning_rate": 2.40986495313435e-06,
"loss": 0.0368,
"step": 763
},
{
"epoch": 1.3815551537070525,
"grad_norm": 0.4001891613006592,
"learning_rate": 2.3970803462819586e-06,
"loss": 0.0154,
"step": 764
},
{
"epoch": 1.3833634719710668,
"grad_norm": 0.3861198425292969,
"learning_rate": 2.384319040806533e-06,
"loss": 0.008,
"step": 765
},
{
"epoch": 1.3851717902350813,
"grad_norm": 0.6267303228378296,
"learning_rate": 2.371581150947476e-06,
"loss": 0.0202,
"step": 766
},
{
"epoch": 1.3869801084990958,
"grad_norm": 0.41174575686454773,
"learning_rate": 2.3588667907345787e-06,
"loss": 0.0291,
"step": 767
},
{
"epoch": 1.3887884267631103,
"grad_norm": 0.16415934264659882,
"learning_rate": 2.3461760739869865e-06,
"loss": 0.0078,
"step": 768
},
{
"epoch": 1.3905967450271248,
"grad_norm": 0.6067925691604614,
"learning_rate": 2.3335091143121935e-06,
"loss": 0.0286,
"step": 769
},
{
"epoch": 1.3924050632911391,
"grad_norm": 0.25584307312965393,
"learning_rate": 2.320866025105016e-06,
"loss": 0.0056,
"step": 770
},
{
"epoch": 1.3942133815551538,
"grad_norm": 0.41166993975639343,
"learning_rate": 2.3082469195465893e-06,
"loss": 0.0333,
"step": 771
},
{
"epoch": 1.3960216998191681,
"grad_norm": 0.5305166244506836,
"learning_rate": 2.2956519106033366e-06,
"loss": 0.0244,
"step": 772
},
{
"epoch": 1.3978300180831826,
"grad_norm": 0.4027380347251892,
"learning_rate": 2.283081111025973e-06,
"loss": 0.0138,
"step": 773
},
{
"epoch": 1.3996383363471971,
"grad_norm": 0.3015824854373932,
"learning_rate": 2.2705346333484925e-06,
"loss": 0.0179,
"step": 774
},
{
"epoch": 1.4014466546112117,
"grad_norm": 0.4979518949985504,
"learning_rate": 2.258012589887154e-06,
"loss": 0.023,
"step": 775
},
{
"epoch": 1.4032549728752262,
"grad_norm": 0.5164641737937927,
"learning_rate": 2.245515092739488e-06,
"loss": 0.0218,
"step": 776
},
{
"epoch": 1.4050632911392404,
"grad_norm": 0.34080126881599426,
"learning_rate": 2.23304225378328e-06,
"loss": 0.0135,
"step": 777
},
{
"epoch": 1.406871609403255,
"grad_norm": 0.3776486814022064,
"learning_rate": 2.2205941846755787e-06,
"loss": 0.0136,
"step": 778
},
{
"epoch": 1.4086799276672695,
"grad_norm": 0.36694374680519104,
"learning_rate": 2.2081709968516867e-06,
"loss": 0.012,
"step": 779
},
{
"epoch": 1.410488245931284,
"grad_norm": 0.43783846497535706,
"learning_rate": 2.1957728015241793e-06,
"loss": 0.0281,
"step": 780
},
{
"epoch": 1.4122965641952985,
"grad_norm": 0.3449253737926483,
"learning_rate": 2.1833997096818897e-06,
"loss": 0.0154,
"step": 781
},
{
"epoch": 1.4141048824593128,
"grad_norm": 0.5510876774787903,
"learning_rate": 2.171051832088928e-06,
"loss": 0.0289,
"step": 782
},
{
"epoch": 1.4159132007233273,
"grad_norm": 1.227265477180481,
"learning_rate": 2.158729279283684e-06,
"loss": 0.0344,
"step": 783
},
{
"epoch": 1.4177215189873418,
"grad_norm": 0.20464493334293365,
"learning_rate": 2.146432161577842e-06,
"loss": 0.0042,
"step": 784
},
{
"epoch": 1.4195298372513563,
"grad_norm": 0.419443279504776,
"learning_rate": 2.1341605890553895e-06,
"loss": 0.0252,
"step": 785
},
{
"epoch": 1.4213381555153708,
"grad_norm": 0.5845422148704529,
"learning_rate": 2.1219146715716332e-06,
"loss": 0.0371,
"step": 786
},
{
"epoch": 1.423146473779385,
"grad_norm": 0.4962576925754547,
"learning_rate": 2.109694518752216e-06,
"loss": 0.0269,
"step": 787
},
{
"epoch": 1.4249547920433996,
"grad_norm": 0.7850181460380554,
"learning_rate": 2.097500239992132e-06,
"loss": 0.0591,
"step": 788
},
{
"epoch": 1.426763110307414,
"grad_norm": 0.24771703779697418,
"learning_rate": 2.085331944454759e-06,
"loss": 0.0117,
"step": 789
},
{
"epoch": 1.4285714285714286,
"grad_norm": 0.46636396646499634,
"learning_rate": 2.0731897410708618e-06,
"loss": 0.0224,
"step": 790
},
{
"epoch": 1.4303797468354431,
"grad_norm": 0.5206124782562256,
"learning_rate": 2.061073738537635e-06,
"loss": 0.0248,
"step": 791
},
{
"epoch": 1.4321880650994574,
"grad_norm": 0.5815186500549316,
"learning_rate": 2.0489840453177198e-06,
"loss": 0.0308,
"step": 792
},
{
"epoch": 1.433996383363472,
"grad_norm": 0.5012145042419434,
"learning_rate": 2.0369207696382366e-06,
"loss": 0.0157,
"step": 793
},
{
"epoch": 1.4358047016274864,
"grad_norm": 0.5018612742424011,
"learning_rate": 2.0248840194898155e-06,
"loss": 0.0274,
"step": 794
},
{
"epoch": 1.437613019891501,
"grad_norm": 0.31762170791625977,
"learning_rate": 2.0128739026256306e-06,
"loss": 0.0071,
"step": 795
},
{
"epoch": 1.4394213381555154,
"grad_norm": 0.37913939356803894,
"learning_rate": 2.0008905265604316e-06,
"loss": 0.0151,
"step": 796
},
{
"epoch": 1.4412296564195297,
"grad_norm": 0.46139174699783325,
"learning_rate": 1.9889339985695894e-06,
"loss": 0.0175,
"step": 797
},
{
"epoch": 1.4430379746835442,
"grad_norm": 0.29624858498573303,
"learning_rate": 1.977004425688126e-06,
"loss": 0.0084,
"step": 798
},
{
"epoch": 1.4448462929475587,
"grad_norm": 0.4915001094341278,
"learning_rate": 1.9651019147097624e-06,
"loss": 0.0126,
"step": 799
},
{
"epoch": 1.4466546112115732,
"grad_norm": 0.5287671685218811,
"learning_rate": 1.95322657218596e-06,
"loss": 0.0368,
"step": 800
},
{
"epoch": 1.4484629294755877,
"grad_norm": 0.21899214386940002,
"learning_rate": 1.941378504424968e-06,
"loss": 0.0091,
"step": 801
},
{
"epoch": 1.450271247739602,
"grad_norm": 0.44369742274284363,
"learning_rate": 1.929557817490874e-06,
"loss": 0.0151,
"step": 802
},
{
"epoch": 1.4520795660036168,
"grad_norm": 0.18337507545948029,
"learning_rate": 1.9177646172026513e-06,
"loss": 0.006,
"step": 803
},
{
"epoch": 1.453887884267631,
"grad_norm": 0.38527706265449524,
"learning_rate": 1.9059990091332082e-06,
"loss": 0.0143,
"step": 804
},
{
"epoch": 1.4556962025316456,
"grad_norm": 0.5627532005310059,
"learning_rate": 1.8942610986084487e-06,
"loss": 0.0212,
"step": 805
},
{
"epoch": 1.45750452079566,
"grad_norm": 0.7829427123069763,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.035,
"step": 806
},
{
"epoch": 1.4593128390596746,
"grad_norm": 0.5543667078018188,
"learning_rate": 1.8708687902559264e-06,
"loss": 0.0219,
"step": 807
},
{
"epoch": 1.461121157323689,
"grad_norm": 0.3493656814098358,
"learning_rate": 1.8592146018364682e-06,
"loss": 0.0108,
"step": 808
},
{
"epoch": 1.4629294755877034,
"grad_norm": 0.6314903497695923,
"learning_rate": 1.8475885297764307e-06,
"loss": 0.0375,
"step": 809
},
{
"epoch": 1.4647377938517179,
"grad_norm": 0.8084059953689575,
"learning_rate": 1.8359906781525955e-06,
"loss": 0.039,
"step": 810
},
{
"epoch": 1.4665461121157324,
"grad_norm": 0.4033972918987274,
"learning_rate": 1.8244211507891064e-06,
"loss": 0.0093,
"step": 811
},
{
"epoch": 1.4683544303797469,
"grad_norm": 0.48802632093429565,
"learning_rate": 1.8128800512565514e-06,
"loss": 0.0418,
"step": 812
},
{
"epoch": 1.4701627486437614,
"grad_norm": 0.6012276411056519,
"learning_rate": 1.8013674828710326e-06,
"loss": 0.0283,
"step": 813
},
{
"epoch": 1.4719710669077757,
"grad_norm": 0.5648048520088196,
"learning_rate": 1.7898835486932398e-06,
"loss": 0.0215,
"step": 814
},
{
"epoch": 1.4737793851717902,
"grad_norm": 0.6136791706085205,
"learning_rate": 1.7784283515275292e-06,
"loss": 0.0341,
"step": 815
},
{
"epoch": 1.4755877034358047,
"grad_norm": 0.2797410190105438,
"learning_rate": 1.7670019939210025e-06,
"loss": 0.0097,
"step": 816
},
{
"epoch": 1.4773960216998192,
"grad_norm": 0.4214756488800049,
"learning_rate": 1.7556045781625902e-06,
"loss": 0.0187,
"step": 817
},
{
"epoch": 1.4792043399638337,
"grad_norm": 0.5396080017089844,
"learning_rate": 1.7442362062821323e-06,
"loss": 0.0211,
"step": 818
},
{
"epoch": 1.481012658227848,
"grad_norm": 0.5082389116287231,
"learning_rate": 1.7328969800494727e-06,
"loss": 0.0298,
"step": 819
},
{
"epoch": 1.4828209764918625,
"grad_norm": 0.4891066253185272,
"learning_rate": 1.7215870009735386e-06,
"loss": 0.0305,
"step": 820
},
{
"epoch": 1.484629294755877,
"grad_norm": 0.5944884419441223,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.0225,
"step": 821
},
{
"epoch": 1.4864376130198915,
"grad_norm": 0.6353393793106079,
"learning_rate": 1.6990551890175488e-06,
"loss": 0.0335,
"step": 822
},
{
"epoch": 1.488245931283906,
"grad_norm": 0.6281875967979431,
"learning_rate": 1.6878335578426225e-06,
"loss": 0.02,
"step": 823
},
{
"epoch": 1.4900542495479203,
"grad_norm": 0.422478049993515,
"learning_rate": 1.6766415772328732e-06,
"loss": 0.0163,
"step": 824
},
{
"epoch": 1.4918625678119348,
"grad_norm": 0.1902221441268921,
"learning_rate": 1.6654793473790842e-06,
"loss": 0.0038,
"step": 825
},
{
"epoch": 1.4936708860759493,
"grad_norm": 0.4058833718299866,
"learning_rate": 1.6543469682057105e-06,
"loss": 0.0223,
"step": 826
},
{
"epoch": 1.4954792043399638,
"grad_norm": 0.5484557747840881,
"learning_rate": 1.6432445393699803e-06,
"loss": 0.0255,
"step": 827
},
{
"epoch": 1.4972875226039783,
"grad_norm": 0.16963370144367218,
"learning_rate": 1.632172160261012e-06,
"loss": 0.0048,
"step": 828
},
{
"epoch": 1.4990958408679926,
"grad_norm": 0.5578981637954712,
"learning_rate": 1.621129929998912e-06,
"loss": 0.0263,
"step": 829
},
{
"epoch": 1.5009041591320074,
"grad_norm": 0.6207785606384277,
"learning_rate": 1.610117947433897e-06,
"loss": 0.0236,
"step": 830
},
{
"epoch": 1.5027124773960217,
"grad_norm": 0.3062937557697296,
"learning_rate": 1.5991363111454023e-06,
"loss": 0.0129,
"step": 831
},
{
"epoch": 1.5045207956600362,
"grad_norm": 0.48795732855796814,
"learning_rate": 1.5881851194412106e-06,
"loss": 0.022,
"step": 832
},
{
"epoch": 1.5063291139240507,
"grad_norm": 0.8056619763374329,
"learning_rate": 1.5772644703565564e-06,
"loss": 0.0414,
"step": 833
},
{
"epoch": 1.508137432188065,
"grad_norm": 0.48942360281944275,
"learning_rate": 1.5663744616532612e-06,
"loss": 0.0279,
"step": 834
},
{
"epoch": 1.5099457504520797,
"grad_norm": 0.5277988314628601,
"learning_rate": 1.5555151908188465e-06,
"loss": 0.0186,
"step": 835
},
{
"epoch": 1.511754068716094,
"grad_norm": 0.6291388869285583,
"learning_rate": 1.544686755065677e-06,
"loss": 0.0332,
"step": 836
},
{
"epoch": 1.5135623869801085,
"grad_norm": 0.39659711718559265,
"learning_rate": 1.5338892513300757e-06,
"loss": 0.0153,
"step": 837
},
{
"epoch": 1.515370705244123,
"grad_norm": 0.5414936542510986,
"learning_rate": 1.523122776271463e-06,
"loss": 0.0333,
"step": 838
},
{
"epoch": 1.5171790235081373,
"grad_norm": 0.680552065372467,
"learning_rate": 1.5123874262714893e-06,
"loss": 0.0211,
"step": 839
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.8744088411331177,
"learning_rate": 1.5016832974331725e-06,
"loss": 0.0324,
"step": 840
},
{
"epoch": 1.5207956600361663,
"grad_norm": 0.17419835925102234,
"learning_rate": 1.4910104855800429e-06,
"loss": 0.0055,
"step": 841
},
{
"epoch": 1.5226039783001808,
"grad_norm": 0.5196138024330139,
"learning_rate": 1.4803690862552755e-06,
"loss": 0.0175,
"step": 842
},
{
"epoch": 1.5244122965641953,
"grad_norm": 1.12484872341156,
"learning_rate": 1.4697591947208412e-06,
"loss": 0.0395,
"step": 843
},
{
"epoch": 1.5262206148282098,
"grad_norm": 0.2727562487125397,
"learning_rate": 1.459180905956653e-06,
"loss": 0.0115,
"step": 844
},
{
"epoch": 1.5280289330922243,
"grad_norm": 0.3629195988178253,
"learning_rate": 1.4486343146597154e-06,
"loss": 0.0166,
"step": 845
},
{
"epoch": 1.5298372513562386,
"grad_norm": 0.23602592945098877,
"learning_rate": 1.438119515243277e-06,
"loss": 0.006,
"step": 846
},
{
"epoch": 1.5316455696202531,
"grad_norm": 0.5272895693778992,
"learning_rate": 1.4276366018359845e-06,
"loss": 0.0231,
"step": 847
},
{
"epoch": 1.5334538878842676,
"grad_norm": 0.4020542502403259,
"learning_rate": 1.4171856682810386e-06,
"loss": 0.013,
"step": 848
},
{
"epoch": 1.5352622061482821,
"grad_norm": 0.5817469358444214,
"learning_rate": 1.4067668081353625e-06,
"loss": 0.0301,
"step": 849
},
{
"epoch": 1.5370705244122966,
"grad_norm": 0.1675567626953125,
"learning_rate": 1.39638011466875e-06,
"loss": 0.0056,
"step": 850
},
{
"epoch": 1.538878842676311,
"grad_norm": 0.21379594504833221,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.0062,
"step": 851
},
{
"epoch": 1.5406871609403257,
"grad_norm": 0.6673656702041626,
"learning_rate": 1.3757035994112915e-06,
"loss": 0.022,
"step": 852
},
{
"epoch": 1.54249547920434,
"grad_norm": 0.4816035330295563,
"learning_rate": 1.3654139627169282e-06,
"loss": 0.0153,
"step": 853
},
{
"epoch": 1.5443037974683544,
"grad_norm": 0.35747525095939636,
"learning_rate": 1.3551568628929434e-06,
"loss": 0.0124,
"step": 854
},
{
"epoch": 1.546112115732369,
"grad_norm": 0.5122990012168884,
"learning_rate": 1.344932391761049e-06,
"loss": 0.0275,
"step": 855
},
{
"epoch": 1.5479204339963832,
"grad_norm": 0.6455609202384949,
"learning_rate": 1.3347406408508695e-06,
"loss": 0.0368,
"step": 856
},
{
"epoch": 1.549728752260398,
"grad_norm": 0.6057823896408081,
"learning_rate": 1.3245817013991164e-06,
"loss": 0.0302,
"step": 857
},
{
"epoch": 1.5515370705244123,
"grad_norm": 0.35086557269096375,
"learning_rate": 1.3144556643487743e-06,
"loss": 0.0243,
"step": 858
},
{
"epoch": 1.5533453887884268,
"grad_norm": 0.3916558623313904,
"learning_rate": 1.3043626203482823e-06,
"loss": 0.0196,
"step": 859
},
{
"epoch": 1.5551537070524413,
"grad_norm": 0.5181978344917297,
"learning_rate": 1.2943026597507268e-06,
"loss": 0.027,
"step": 860
},
{
"epoch": 1.5569620253164556,
"grad_norm": 0.408285915851593,
"learning_rate": 1.2842758726130283e-06,
"loss": 0.0157,
"step": 861
},
{
"epoch": 1.5587703435804703,
"grad_norm": 0.17891541123390198,
"learning_rate": 1.2742823486951434e-06,
"loss": 0.0068,
"step": 862
},
{
"epoch": 1.5605786618444846,
"grad_norm": 0.8443412184715271,
"learning_rate": 1.2643221774592517e-06,
"loss": 0.0218,
"step": 863
},
{
"epoch": 1.562386980108499,
"grad_norm": 0.8202517628669739,
"learning_rate": 1.254395448068959e-06,
"loss": 0.0504,
"step": 864
},
{
"epoch": 1.5641952983725136,
"grad_norm": 0.7840216755867004,
"learning_rate": 1.2445022493885017e-06,
"loss": 0.0252,
"step": 865
},
{
"epoch": 1.5660036166365279,
"grad_norm": 0.13034725189208984,
"learning_rate": 1.234642669981946e-06,
"loss": 0.0039,
"step": 866
},
{
"epoch": 1.5678119349005426,
"grad_norm": 0.515975296497345,
"learning_rate": 1.2248167981123992e-06,
"loss": 0.0222,
"step": 867
},
{
"epoch": 1.5696202531645569,
"grad_norm": 0.4934346377849579,
"learning_rate": 1.2150247217412186e-06,
"loss": 0.0293,
"step": 868
},
{
"epoch": 1.5714285714285714,
"grad_norm": 0.48679468035697937,
"learning_rate": 1.205266528527223e-06,
"loss": 0.0216,
"step": 869
},
{
"epoch": 1.573236889692586,
"grad_norm": 0.4664691984653473,
"learning_rate": 1.195542305825908e-06,
"loss": 0.0241,
"step": 870
},
{
"epoch": 1.5750452079566004,
"grad_norm": 0.2800479531288147,
"learning_rate": 1.1858521406886674e-06,
"loss": 0.017,
"step": 871
},
{
"epoch": 1.576853526220615,
"grad_norm": 0.46079421043395996,
"learning_rate": 1.1761961198620081e-06,
"loss": 0.0167,
"step": 872
},
{
"epoch": 1.5786618444846292,
"grad_norm": 0.14544257521629333,
"learning_rate": 1.1665743297867781e-06,
"loss": 0.006,
"step": 873
},
{
"epoch": 1.5804701627486437,
"grad_norm": 0.5996628999710083,
"learning_rate": 1.1569868565973912e-06,
"loss": 0.0588,
"step": 874
},
{
"epoch": 1.5822784810126582,
"grad_norm": 0.2130950689315796,
"learning_rate": 1.1474337861210543e-06,
"loss": 0.005,
"step": 875
},
{
"epoch": 1.5840867992766727,
"grad_norm": 0.47326400876045227,
"learning_rate": 1.137915203877003e-06,
"loss": 0.0241,
"step": 876
},
{
"epoch": 1.5858951175406872,
"grad_norm": 0.29357361793518066,
"learning_rate": 1.1284311950757326e-06,
"loss": 0.0091,
"step": 877
},
{
"epoch": 1.5877034358047015,
"grad_norm": 0.2985672056674957,
"learning_rate": 1.118981844618236e-06,
"loss": 0.01,
"step": 878
},
{
"epoch": 1.5895117540687163,
"grad_norm": 0.5034512877464294,
"learning_rate": 1.1095672370952431e-06,
"loss": 0.025,
"step": 879
},
{
"epoch": 1.5913200723327305,
"grad_norm": 0.12862065434455872,
"learning_rate": 1.1001874567864696e-06,
"loss": 0.0024,
"step": 880
},
{
"epoch": 1.593128390596745,
"grad_norm": 0.4813363552093506,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.0212,
"step": 881
},
{
"epoch": 1.5949367088607596,
"grad_norm": 0.2282480001449585,
"learning_rate": 1.0815327133708015e-06,
"loss": 0.0102,
"step": 882
},
{
"epoch": 1.5967450271247738,
"grad_norm": 0.32203638553619385,
"learning_rate": 1.0722579172614577e-06,
"loss": 0.0107,
"step": 883
},
{
"epoch": 1.5985533453887886,
"grad_norm": 1.0874191522598267,
"learning_rate": 1.06301828235994e-06,
"loss": 0.0497,
"step": 884
},
{
"epoch": 1.6003616636528029,
"grad_norm": 0.639910876750946,
"learning_rate": 1.0538138913796032e-06,
"loss": 0.0268,
"step": 885
},
{
"epoch": 1.6021699819168174,
"grad_norm": 0.24462899565696716,
"learning_rate": 1.044644826718295e-06,
"loss": 0.0068,
"step": 886
},
{
"epoch": 1.6039783001808319,
"grad_norm": 0.2610553503036499,
"learning_rate": 1.0355111704576237e-06,
"loss": 0.0088,
"step": 887
},
{
"epoch": 1.6057866184448462,
"grad_norm": 0.29826587438583374,
"learning_rate": 1.0264130043622245e-06,
"loss": 0.0051,
"step": 888
},
{
"epoch": 1.6075949367088609,
"grad_norm": 0.3923262059688568,
"learning_rate": 1.0173504098790188e-06,
"loss": 0.0172,
"step": 889
},
{
"epoch": 1.6094032549728752,
"grad_norm": 0.47321057319641113,
"learning_rate": 1.0083234681364934e-06,
"loss": 0.0173,
"step": 890
},
{
"epoch": 1.6112115732368897,
"grad_norm": 0.5533871054649353,
"learning_rate": 9.993322599439692e-07,
"loss": 0.0307,
"step": 891
},
{
"epoch": 1.6130198915009042,
"grad_norm": 0.34101057052612305,
"learning_rate": 9.903768657908803e-07,
"loss": 0.0072,
"step": 892
},
{
"epoch": 1.6148282097649185,
"grad_norm": 0.2183762788772583,
"learning_rate": 9.814573658460564e-07,
"loss": 0.0093,
"step": 893
},
{
"epoch": 1.6166365280289332,
"grad_norm": 0.6712782382965088,
"learning_rate": 9.725738399569968e-07,
"loss": 0.0405,
"step": 894
},
{
"epoch": 1.6184448462929475,
"grad_norm": 0.6365363597869873,
"learning_rate": 9.637263676491627e-07,
"loss": 0.0399,
"step": 895
},
{
"epoch": 1.620253164556962,
"grad_norm": 0.3082372546195984,
"learning_rate": 9.549150281252633e-07,
"loss": 0.0101,
"step": 896
},
{
"epoch": 1.6220614828209765,
"grad_norm": 0.29272228479385376,
"learning_rate": 9.46139900264546e-07,
"loss": 0.0089,
"step": 897
},
{
"epoch": 1.6238698010849908,
"grad_norm": 0.1736106127500534,
"learning_rate": 9.374010626220908e-07,
"loss": 0.0039,
"step": 898
},
{
"epoch": 1.6256781193490055,
"grad_norm": 0.4352719187736511,
"learning_rate": 9.286985934281079e-07,
"loss": 0.0204,
"step": 899
},
{
"epoch": 1.6274864376130198,
"grad_norm": 0.15794239938259125,
"learning_rate": 9.200325705872342e-07,
"loss": 0.0051,
"step": 900
},
{
"epoch": 1.6292947558770343,
"grad_norm": 0.179263174533844,
"learning_rate": 9.114030716778433e-07,
"loss": 0.0033,
"step": 901
},
{
"epoch": 1.6311030741410488,
"grad_norm": 0.4568769931793213,
"learning_rate": 9.028101739513406e-07,
"loss": 0.0163,
"step": 902
},
{
"epoch": 1.6329113924050633,
"grad_norm": 0.3016704320907593,
"learning_rate": 8.942539543314799e-07,
"loss": 0.0113,
"step": 903
},
{
"epoch": 1.6347197106690778,
"grad_norm": 0.38421815633773804,
"learning_rate": 8.857344894136715e-07,
"loss": 0.0165,
"step": 904
},
{
"epoch": 1.6365280289330921,
"grad_norm": 0.31550708413124084,
"learning_rate": 8.772518554642973e-07,
"loss": 0.0261,
"step": 905
},
{
"epoch": 1.6383363471971069,
"grad_norm": 0.147488072514534,
"learning_rate": 8.688061284200266e-07,
"loss": 0.0033,
"step": 906
},
{
"epoch": 1.6401446654611211,
"grad_norm": 0.28551799058914185,
"learning_rate": 8.603973838871388e-07,
"loss": 0.0079,
"step": 907
},
{
"epoch": 1.6419529837251357,
"grad_norm": 0.5802464485168457,
"learning_rate": 8.520256971408453e-07,
"loss": 0.0322,
"step": 908
},
{
"epoch": 1.6437613019891502,
"grad_norm": 0.2847616374492645,
"learning_rate": 8.436911431246136e-07,
"loss": 0.0151,
"step": 909
},
{
"epoch": 1.6455696202531644,
"grad_norm": 0.263936847448349,
"learning_rate": 8.353937964495029e-07,
"loss": 0.012,
"step": 910
},
{
"epoch": 1.6473779385171792,
"grad_norm": 0.38852736353874207,
"learning_rate": 8.271337313934869e-07,
"loss": 0.0092,
"step": 911
},
{
"epoch": 1.6491862567811935,
"grad_norm": 0.4846401810646057,
"learning_rate": 8.189110219007967e-07,
"loss": 0.0191,
"step": 912
},
{
"epoch": 1.650994575045208,
"grad_norm": 0.3546898066997528,
"learning_rate": 8.107257415812526e-07,
"loss": 0.0123,
"step": 913
},
{
"epoch": 1.6528028933092225,
"grad_norm": 0.7781378030776978,
"learning_rate": 8.025779637096138e-07,
"loss": 0.0358,
"step": 914
},
{
"epoch": 1.6546112115732368,
"grad_norm": 0.5066570043563843,
"learning_rate": 7.944677612249113e-07,
"loss": 0.0213,
"step": 915
},
{
"epoch": 1.6564195298372515,
"grad_norm": 0.35833096504211426,
"learning_rate": 7.863952067298042e-07,
"loss": 0.0144,
"step": 916
},
{
"epoch": 1.6582278481012658,
"grad_norm": 0.8536073565483093,
"learning_rate": 7.783603724899258e-07,
"loss": 0.0297,
"step": 917
},
{
"epoch": 1.6600361663652803,
"grad_norm": 0.5022109150886536,
"learning_rate": 7.70363330433233e-07,
"loss": 0.019,
"step": 918
},
{
"epoch": 1.6618444846292948,
"grad_norm": 0.3407527506351471,
"learning_rate": 7.624041521493735e-07,
"loss": 0.0284,
"step": 919
},
{
"epoch": 1.663652802893309,
"grad_norm": 0.561324954032898,
"learning_rate": 7.544829088890326e-07,
"loss": 0.0344,
"step": 920
},
{
"epoch": 1.6654611211573238,
"grad_norm": 0.24968159198760986,
"learning_rate": 7.465996715633028e-07,
"loss": 0.0084,
"step": 921
},
{
"epoch": 1.667269439421338,
"grad_norm": 0.7005723118782043,
"learning_rate": 7.387545107430455e-07,
"loss": 0.0379,
"step": 922
},
{
"epoch": 1.6690777576853526,
"grad_norm": 0.5622426271438599,
"learning_rate": 7.309474966582636e-07,
"loss": 0.036,
"step": 923
},
{
"epoch": 1.6708860759493671,
"grad_norm": 0.4123494029045105,
"learning_rate": 7.23178699197467e-07,
"loss": 0.0138,
"step": 924
},
{
"epoch": 1.6726943942133814,
"grad_norm": 0.08860176056623459,
"learning_rate": 7.154481879070502e-07,
"loss": 0.0019,
"step": 925
},
{
"epoch": 1.6745027124773961,
"grad_norm": 0.40314149856567383,
"learning_rate": 7.077560319906696e-07,
"loss": 0.019,
"step": 926
},
{
"epoch": 1.6763110307414104,
"grad_norm": 0.34653234481811523,
"learning_rate": 7.001023003086243e-07,
"loss": 0.0147,
"step": 927
},
{
"epoch": 1.678119349005425,
"grad_norm": 0.30449631810188293,
"learning_rate": 6.924870613772388e-07,
"loss": 0.0123,
"step": 928
},
{
"epoch": 1.6799276672694394,
"grad_norm": 0.29880601167678833,
"learning_rate": 6.849103833682491e-07,
"loss": 0.0149,
"step": 929
},
{
"epoch": 1.681735985533454,
"grad_norm": 0.43195775151252747,
"learning_rate": 6.773723341081945e-07,
"loss": 0.0232,
"step": 930
},
{
"epoch": 1.6835443037974684,
"grad_norm": 0.34601637721061707,
"learning_rate": 6.698729810778065e-07,
"loss": 0.0096,
"step": 931
},
{
"epoch": 1.6853526220614827,
"grad_norm": 0.48838481307029724,
"learning_rate": 6.624123914114122e-07,
"loss": 0.0241,
"step": 932
},
{
"epoch": 1.6871609403254972,
"grad_norm": 0.4325457215309143,
"learning_rate": 6.549906318963245e-07,
"loss": 0.0219,
"step": 933
},
{
"epoch": 1.6889692585895117,
"grad_norm": 0.5117323398590088,
"learning_rate": 6.476077689722487e-07,
"loss": 0.0413,
"step": 934
},
{
"epoch": 1.6907775768535263,
"grad_norm": 0.5409876704216003,
"learning_rate": 6.402638687306872e-07,
"loss": 0.0157,
"step": 935
},
{
"epoch": 1.6925858951175408,
"grad_norm": 0.20510420203208923,
"learning_rate": 6.329589969143518e-07,
"loss": 0.0047,
"step": 936
},
{
"epoch": 1.694394213381555,
"grad_norm": 0.16492347419261932,
"learning_rate": 6.256932189165644e-07,
"loss": 0.0048,
"step": 937
},
{
"epoch": 1.6962025316455698,
"grad_norm": 0.7954071164131165,
"learning_rate": 6.184665997806832e-07,
"loss": 0.0339,
"step": 938
},
{
"epoch": 1.698010849909584,
"grad_norm": 0.8786876201629639,
"learning_rate": 6.112792041995125e-07,
"loss": 0.0283,
"step": 939
},
{
"epoch": 1.6998191681735986,
"grad_norm": 0.8798893690109253,
"learning_rate": 6.041310965147318e-07,
"loss": 0.0427,
"step": 940
},
{
"epoch": 1.701627486437613,
"grad_norm": 0.26145556569099426,
"learning_rate": 5.9702234071631e-07,
"loss": 0.0131,
"step": 941
},
{
"epoch": 1.7034358047016274,
"grad_norm": 0.34640127420425415,
"learning_rate": 5.899530004419396e-07,
"loss": 0.0104,
"step": 942
},
{
"epoch": 1.705244122965642,
"grad_norm": 0.3936959207057953,
"learning_rate": 5.829231389764628e-07,
"loss": 0.0154,
"step": 943
},
{
"epoch": 1.7070524412296564,
"grad_norm": 0.4007095694541931,
"learning_rate": 5.759328192513075e-07,
"loss": 0.0178,
"step": 944
},
{
"epoch": 1.7088607594936709,
"grad_norm": 0.4842199981212616,
"learning_rate": 5.689821038439264e-07,
"loss": 0.0222,
"step": 945
},
{
"epoch": 1.7106690777576854,
"grad_norm": 0.2961626648902893,
"learning_rate": 5.620710549772295e-07,
"loss": 0.0103,
"step": 946
},
{
"epoch": 1.7124773960216997,
"grad_norm": 0.44128212332725525,
"learning_rate": 5.55199734519034e-07,
"loss": 0.022,
"step": 947
},
{
"epoch": 1.7142857142857144,
"grad_norm": 0.7379593849182129,
"learning_rate": 5.483682039815059e-07,
"loss": 0.0609,
"step": 948
},
{
"epoch": 1.7160940325497287,
"grad_norm": 0.11172860115766525,
"learning_rate": 5.415765245206128e-07,
"loss": 0.0024,
"step": 949
},
{
"epoch": 1.7179023508137432,
"grad_norm": 0.5290942788124084,
"learning_rate": 5.348247569355736e-07,
"loss": 0.0412,
"step": 950
},
{
"epoch": 1.7197106690777577,
"grad_norm": 0.2841649651527405,
"learning_rate": 5.281129616683167e-07,
"loss": 0.0126,
"step": 951
},
{
"epoch": 1.721518987341772,
"grad_norm": 1.0218212604522705,
"learning_rate": 5.214411988029355e-07,
"loss": 0.0193,
"step": 952
},
{
"epoch": 1.7233273056057867,
"grad_norm": 0.7516573667526245,
"learning_rate": 5.148095280651566e-07,
"loss": 0.0328,
"step": 953
},
{
"epoch": 1.725135623869801,
"grad_norm": 0.32349979877471924,
"learning_rate": 5.082180088217981e-07,
"loss": 0.0129,
"step": 954
},
{
"epoch": 1.7269439421338155,
"grad_norm": 0.43182599544525146,
"learning_rate": 5.016667000802417e-07,
"loss": 0.0227,
"step": 955
},
{
"epoch": 1.72875226039783,
"grad_norm": 0.21323497593402863,
"learning_rate": 4.951556604879049e-07,
"loss": 0.0044,
"step": 956
},
{
"epoch": 1.7305605786618445,
"grad_norm": 0.494210809469223,
"learning_rate": 4.886849483317141e-07,
"loss": 0.0155,
"step": 957
},
{
"epoch": 1.732368896925859,
"grad_norm": 0.6554979085922241,
"learning_rate": 4.822546215375851e-07,
"loss": 0.0223,
"step": 958
},
{
"epoch": 1.7341772151898733,
"grad_norm": 0.8195722699165344,
"learning_rate": 4.758647376699033e-07,
"loss": 0.0279,
"step": 959
},
{
"epoch": 1.7359855334538878,
"grad_norm": 0.6029171943664551,
"learning_rate": 4.6951535393100654e-07,
"loss": 0.035,
"step": 960
},
{
"epoch": 1.7377938517179023,
"grad_norm": 0.881205141544342,
"learning_rate": 4.632065271606756e-07,
"loss": 0.0448,
"step": 961
},
{
"epoch": 1.7396021699819169,
"grad_norm": 0.34877485036849976,
"learning_rate": 4.569383138356276e-07,
"loss": 0.0121,
"step": 962
},
{
"epoch": 1.7414104882459314,
"grad_norm": 0.42677101492881775,
"learning_rate": 4.507107700690044e-07,
"loss": 0.0175,
"step": 963
},
{
"epoch": 1.7432188065099457,
"grad_norm": 0.6321128606796265,
"learning_rate": 4.4452395160987314e-07,
"loss": 0.0389,
"step": 964
},
{
"epoch": 1.7450271247739604,
"grad_norm": 1.192665457725525,
"learning_rate": 4.383779138427274e-07,
"loss": 0.0695,
"step": 965
},
{
"epoch": 1.7468354430379747,
"grad_norm": 0.3836626410484314,
"learning_rate": 4.322727117869951e-07,
"loss": 0.0189,
"step": 966
},
{
"epoch": 1.7486437613019892,
"grad_norm": 0.3103764057159424,
"learning_rate": 4.2620840009653827e-07,
"loss": 0.0109,
"step": 967
},
{
"epoch": 1.7504520795660037,
"grad_norm": 0.787624716758728,
"learning_rate": 4.201850330591678e-07,
"loss": 0.047,
"step": 968
},
{
"epoch": 1.752260397830018,
"grad_norm": 0.5515686869621277,
"learning_rate": 4.1420266459615944e-07,
"loss": 0.0159,
"step": 969
},
{
"epoch": 1.7540687160940327,
"grad_norm": 0.3365809917449951,
"learning_rate": 4.082613482617664e-07,
"loss": 0.0108,
"step": 970
},
{
"epoch": 1.755877034358047,
"grad_norm": 0.7576748728752136,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.0545,
"step": 971
},
{
"epoch": 1.7576853526220615,
"grad_norm": 0.7700674533843994,
"learning_rate": 3.965020843578804e-07,
"loss": 0.0141,
"step": 972
},
{
"epoch": 1.759493670886076,
"grad_norm": 0.30188339948654175,
"learning_rate": 3.90684242057498e-07,
"loss": 0.01,
"step": 973
},
{
"epoch": 1.7613019891500903,
"grad_norm": 0.4622746706008911,
"learning_rate": 3.8490766242301356e-07,
"loss": 0.0268,
"step": 974
},
{
"epoch": 1.763110307414105,
"grad_norm": 0.450285404920578,
"learning_rate": 3.7917239716645906e-07,
"loss": 0.03,
"step": 975
},
{
"epoch": 1.7649186256781193,
"grad_norm": 0.46768561005592346,
"learning_rate": 3.734784976300165e-07,
"loss": 0.0184,
"step": 976
},
{
"epoch": 1.7667269439421338,
"grad_norm": 0.2854677736759186,
"learning_rate": 3.678260147855628e-07,
"loss": 0.0091,
"step": 977
},
{
"epoch": 1.7685352622061483,
"grad_norm": 0.5644873976707458,
"learning_rate": 3.6221499923421164e-07,
"loss": 0.0165,
"step": 978
},
{
"epoch": 1.7703435804701626,
"grad_norm": 0.33191636204719543,
"learning_rate": 3.566455012058612e-07,
"loss": 0.0144,
"step": 979
},
{
"epoch": 1.7721518987341773,
"grad_norm": 0.45694851875305176,
"learning_rate": 3.511175705587433e-07,
"loss": 0.0161,
"step": 980
},
{
"epoch": 1.7739602169981916,
"grad_norm": 0.35751280188560486,
"learning_rate": 3.4563125677897936e-07,
"loss": 0.0141,
"step": 981
},
{
"epoch": 1.7757685352622061,
"grad_norm": 0.43770647048950195,
"learning_rate": 3.4018660898013423e-07,
"loss": 0.0289,
"step": 982
},
{
"epoch": 1.7775768535262206,
"grad_norm": 0.5489205121994019,
"learning_rate": 3.347836759027789e-07,
"loss": 0.0301,
"step": 983
},
{
"epoch": 1.779385171790235,
"grad_norm": 0.44735196232795715,
"learning_rate": 3.2942250591405546e-07,
"loss": 0.0211,
"step": 984
},
{
"epoch": 1.7811934900542497,
"grad_norm": 0.6316576600074768,
"learning_rate": 3.241031470072398e-07,
"loss": 0.0216,
"step": 985
},
{
"epoch": 1.783001808318264,
"grad_norm": 0.424728125333786,
"learning_rate": 3.18825646801314e-07,
"loss": 0.026,
"step": 986
},
{
"epoch": 1.7848101265822784,
"grad_norm": 0.337449848651886,
"learning_rate": 3.135900525405428e-07,
"loss": 0.0108,
"step": 987
},
{
"epoch": 1.786618444846293,
"grad_norm": 0.5442370772361755,
"learning_rate": 3.0839641109404627e-07,
"loss": 0.0316,
"step": 988
},
{
"epoch": 1.7884267631103075,
"grad_norm": 0.3423631191253662,
"learning_rate": 3.03244768955383e-07,
"loss": 0.0151,
"step": 989
},
{
"epoch": 1.790235081374322,
"grad_norm": 0.25401565432548523,
"learning_rate": 2.9813517224213274e-07,
"loss": 0.012,
"step": 990
},
{
"epoch": 1.7920433996383363,
"grad_norm": 0.5420613288879395,
"learning_rate": 2.930676666954846e-07,
"loss": 0.0193,
"step": 991
},
{
"epoch": 1.7938517179023508,
"grad_norm": 0.5779260993003845,
"learning_rate": 2.8804229767982637e-07,
"loss": 0.0334,
"step": 992
},
{
"epoch": 1.7956600361663653,
"grad_norm": 0.42337730526924133,
"learning_rate": 2.8305911018233935e-07,
"loss": 0.0118,
"step": 993
},
{
"epoch": 1.7974683544303798,
"grad_norm": 0.30050355195999146,
"learning_rate": 2.7811814881259503e-07,
"loss": 0.0091,
"step": 994
},
{
"epoch": 1.7992766726943943,
"grad_norm": 0.5542883276939392,
"learning_rate": 2.7321945780215576e-07,
"loss": 0.0318,
"step": 995
},
{
"epoch": 1.8010849909584086,
"grad_norm": 0.6068146824836731,
"learning_rate": 2.6836308100417874e-07,
"loss": 0.0389,
"step": 996
},
{
"epoch": 1.8028933092224233,
"grad_norm": 0.28464120626449585,
"learning_rate": 2.6354906189302534e-07,
"loss": 0.0091,
"step": 997
},
{
"epoch": 1.8047016274864376,
"grad_norm": 0.40251314640045166,
"learning_rate": 2.587774435638679e-07,
"loss": 0.0275,
"step": 998
},
{
"epoch": 1.806509945750452,
"grad_norm": 0.5314449071884155,
"learning_rate": 2.5404826873230926e-07,
"loss": 0.0143,
"step": 999
},
{
"epoch": 1.8083182640144666,
"grad_norm": 0.5563151836395264,
"learning_rate": 2.4936157973399266e-07,
"loss": 0.0325,
"step": 1000
},
{
"epoch": 1.810126582278481,
"grad_norm": 0.095887690782547,
"learning_rate": 2.447174185242324e-07,
"loss": 0.0024,
"step": 1001
},
{
"epoch": 1.8119349005424956,
"grad_norm": 0.36221036314964294,
"learning_rate": 2.40115826677631e-07,
"loss": 0.0145,
"step": 1002
},
{
"epoch": 1.81374321880651,
"grad_norm": 0.4978484511375427,
"learning_rate": 2.3555684538770995e-07,
"loss": 0.0209,
"step": 1003
},
{
"epoch": 1.8155515370705244,
"grad_norm": 0.3918057084083557,
"learning_rate": 2.3104051546654016e-07,
"loss": 0.0148,
"step": 1004
},
{
"epoch": 1.817359855334539,
"grad_norm": 0.2561500668525696,
"learning_rate": 2.2656687734437577e-07,
"loss": 0.009,
"step": 1005
},
{
"epoch": 1.8191681735985532,
"grad_norm": 0.4532489478588104,
"learning_rate": 2.2213597106929608e-07,
"loss": 0.0287,
"step": 1006
},
{
"epoch": 1.820976491862568,
"grad_norm": 0.3167125880718231,
"learning_rate": 2.177478363068425e-07,
"loss": 0.0093,
"step": 1007
},
{
"epoch": 1.8227848101265822,
"grad_norm": 0.7253750562667847,
"learning_rate": 2.134025123396638e-07,
"loss": 0.0232,
"step": 1008
},
{
"epoch": 1.8245931283905967,
"grad_norm": 0.30931657552719116,
"learning_rate": 2.0910003806716817e-07,
"loss": 0.0125,
"step": 1009
},
{
"epoch": 1.8264014466546112,
"grad_norm": 0.34001412987709045,
"learning_rate": 2.0484045200517222e-07,
"loss": 0.0229,
"step": 1010
},
{
"epoch": 1.8282097649186255,
"grad_norm": 0.396189421415329,
"learning_rate": 2.006237922855553e-07,
"loss": 0.0194,
"step": 1011
},
{
"epoch": 1.8300180831826403,
"grad_norm": 0.6848304271697998,
"learning_rate": 1.9645009665592073e-07,
"loss": 0.0177,
"step": 1012
},
{
"epoch": 1.8318264014466545,
"grad_norm": 0.40357571840286255,
"learning_rate": 1.9231940247925573e-07,
"loss": 0.0251,
"step": 1013
},
{
"epoch": 1.833634719710669,
"grad_norm": 0.46881258487701416,
"learning_rate": 1.882317467335998e-07,
"loss": 0.0234,
"step": 1014
},
{
"epoch": 1.8354430379746836,
"grad_norm": 0.7889057993888855,
"learning_rate": 1.841871660117095e-07,
"loss": 0.055,
"step": 1015
},
{
"epoch": 1.837251356238698,
"grad_norm": 0.29039299488067627,
"learning_rate": 1.801856965207338e-07,
"loss": 0.0085,
"step": 1016
},
{
"epoch": 1.8390596745027126,
"grad_norm": 0.9131592512130737,
"learning_rate": 1.7622737408188984e-07,
"loss": 0.024,
"step": 1017
},
{
"epoch": 1.8408679927667269,
"grad_norm": 0.3902417719364166,
"learning_rate": 1.7231223413014086e-07,
"loss": 0.0148,
"step": 1018
},
{
"epoch": 1.8426763110307414,
"grad_norm": 0.3746405839920044,
"learning_rate": 1.6844031171388054e-07,
"loss": 0.0146,
"step": 1019
},
{
"epoch": 1.8444846292947559,
"grad_norm": 0.4381090998649597,
"learning_rate": 1.6461164149461805e-07,
"loss": 0.0144,
"step": 1020
},
{
"epoch": 1.8462929475587704,
"grad_norm": 0.9630091786384583,
"learning_rate": 1.6082625774666793e-07,
"loss": 0.0431,
"step": 1021
},
{
"epoch": 1.8481012658227849,
"grad_norm": 0.3410443961620331,
"learning_rate": 1.5708419435684463e-07,
"loss": 0.0213,
"step": 1022
},
{
"epoch": 1.8499095840867992,
"grad_norm": 0.43488380312919617,
"learning_rate": 1.5338548482415726e-07,
"loss": 0.0134,
"step": 1023
},
{
"epoch": 1.851717902350814,
"grad_norm": 0.40657466650009155,
"learning_rate": 1.4973016225951097e-07,
"loss": 0.0303,
"step": 1024
},
{
"epoch": 1.8535262206148282,
"grad_norm": 0.3977767825126648,
"learning_rate": 1.4611825938540936e-07,
"loss": 0.0212,
"step": 1025
},
{
"epoch": 1.8553345388788427,
"grad_norm": 0.6467922925949097,
"learning_rate": 1.4254980853566248e-07,
"loss": 0.0268,
"step": 1026
},
{
"epoch": 1.8571428571428572,
"grad_norm": 0.8614939451217651,
"learning_rate": 1.3902484165509877e-07,
"loss": 0.0466,
"step": 1027
},
{
"epoch": 1.8589511754068715,
"grad_norm": 0.573024570941925,
"learning_rate": 1.3554339029927532e-07,
"loss": 0.0335,
"step": 1028
},
{
"epoch": 1.8607594936708862,
"grad_norm": 0.45933103561401367,
"learning_rate": 1.3210548563419857e-07,
"loss": 0.0224,
"step": 1029
},
{
"epoch": 1.8625678119349005,
"grad_norm": 0.5910979509353638,
"learning_rate": 1.2871115843604508e-07,
"loss": 0.0285,
"step": 1030
},
{
"epoch": 1.864376130198915,
"grad_norm": 0.808844804763794,
"learning_rate": 1.253604390908819e-07,
"loss": 0.0264,
"step": 1031
},
{
"epoch": 1.8661844484629295,
"grad_norm": 0.3397260904312134,
"learning_rate": 1.220533575944033e-07,
"loss": 0.0113,
"step": 1032
},
{
"epoch": 1.8679927667269438,
"grad_norm": 0.6284553408622742,
"learning_rate": 1.1878994355165207e-07,
"loss": 0.0363,
"step": 1033
},
{
"epoch": 1.8698010849909585,
"grad_norm": 0.36261695623397827,
"learning_rate": 1.1557022617676217e-07,
"loss": 0.0207,
"step": 1034
},
{
"epoch": 1.8716094032549728,
"grad_norm": 0.3067341148853302,
"learning_rate": 1.1239423429269203e-07,
"loss": 0.0138,
"step": 1035
},
{
"epoch": 1.8734177215189873,
"grad_norm": 0.24892014265060425,
"learning_rate": 1.0926199633097156e-07,
"loss": 0.0111,
"step": 1036
},
{
"epoch": 1.8752260397830018,
"grad_norm": 0.46843257546424866,
"learning_rate": 1.0617354033144289e-07,
"loss": 0.0317,
"step": 1037
},
{
"epoch": 1.8770343580470161,
"grad_norm": 0.3763027489185333,
"learning_rate": 1.031288939420122e-07,
"loss": 0.0211,
"step": 1038
},
{
"epoch": 1.8788426763110309,
"grad_norm": 0.3539106547832489,
"learning_rate": 1.0012808441839994e-07,
"loss": 0.0161,
"step": 1039
},
{
"epoch": 1.8806509945750451,
"grad_norm": 0.4532868564128876,
"learning_rate": 9.717113862389993e-08,
"loss": 0.0205,
"step": 1040
},
{
"epoch": 1.8824593128390597,
"grad_norm": 0.6230972409248352,
"learning_rate": 9.42580830291373e-08,
"loss": 0.0284,
"step": 1041
},
{
"epoch": 1.8842676311030742,
"grad_norm": 0.6785014867782593,
"learning_rate": 9.138894371182983e-08,
"loss": 0.0348,
"step": 1042
},
{
"epoch": 1.8860759493670884,
"grad_norm": 0.4285714328289032,
"learning_rate": 8.856374635655696e-08,
"loss": 0.0245,
"step": 1043
},
{
"epoch": 1.8878842676311032,
"grad_norm": 0.43516093492507935,
"learning_rate": 8.57825162545295e-08,
"loss": 0.0119,
"step": 1044
},
{
"epoch": 1.8896925858951175,
"grad_norm": 0.6266226768493652,
"learning_rate": 8.304527830336251e-08,
"loss": 0.043,
"step": 1045
},
{
"epoch": 1.891500904159132,
"grad_norm": 0.44804707169532776,
"learning_rate": 8.035205700685167e-08,
"loss": 0.0242,
"step": 1046
},
{
"epoch": 1.8933092224231465,
"grad_norm": 0.19278812408447266,
"learning_rate": 7.770287647475672e-08,
"loss": 0.0046,
"step": 1047
},
{
"epoch": 1.895117540687161,
"grad_norm": 0.35671189427375793,
"learning_rate": 7.509776042258166e-08,
"loss": 0.0138,
"step": 1048
},
{
"epoch": 1.8969258589511755,
"grad_norm": 0.7324717044830322,
"learning_rate": 7.253673217136659e-08,
"loss": 0.0406,
"step": 1049
},
{
"epoch": 1.8987341772151898,
"grad_norm": 0.22765561938285828,
"learning_rate": 7.001981464747565e-08,
"loss": 0.007,
"step": 1050
},
{
"epoch": 1.9005424954792043,
"grad_norm": 0.23720940947532654,
"learning_rate": 6.75470303823933e-08,
"loss": 0.0051,
"step": 1051
},
{
"epoch": 1.9023508137432188,
"grad_norm": 0.47292348742485046,
"learning_rate": 6.511840151252169e-08,
"loss": 0.0164,
"step": 1052
},
{
"epoch": 1.9041591320072333,
"grad_norm": 0.41059109568595886,
"learning_rate": 6.273394977898528e-08,
"loss": 0.0175,
"step": 1053
},
{
"epoch": 1.9059674502712478,
"grad_norm": 0.3481151759624481,
"learning_rate": 6.039369652743266e-08,
"loss": 0.0161,
"step": 1054
},
{
"epoch": 1.907775768535262,
"grad_norm": 0.3793577253818512,
"learning_rate": 5.809766270784667e-08,
"loss": 0.0173,
"step": 1055
},
{
"epoch": 1.9095840867992768,
"grad_norm": 0.5948733687400818,
"learning_rate": 5.584586887435739e-08,
"loss": 0.0255,
"step": 1056
},
{
"epoch": 1.9113924050632911,
"grad_norm": 0.4178677201271057,
"learning_rate": 5.363833518505834e-08,
"loss": 0.0226,
"step": 1057
},
{
"epoch": 1.9132007233273056,
"grad_norm": 0.6221929788589478,
"learning_rate": 5.1475081401825553e-08,
"loss": 0.019,
"step": 1058
},
{
"epoch": 1.9150090415913201,
"grad_norm": 0.526636004447937,
"learning_rate": 4.9356126890139356e-08,
"loss": 0.0192,
"step": 1059
},
{
"epoch": 1.9168173598553344,
"grad_norm": 0.7365792989730835,
"learning_rate": 4.7281490618914516e-08,
"loss": 0.036,
"step": 1060
},
{
"epoch": 1.9186256781193491,
"grad_norm": 0.6622296571731567,
"learning_rate": 4.52511911603265e-08,
"loss": 0.0292,
"step": 1061
},
{
"epoch": 1.9204339963833634,
"grad_norm": 0.5662434101104736,
"learning_rate": 4.32652466896466e-08,
"loss": 0.0265,
"step": 1062
},
{
"epoch": 1.922242314647378,
"grad_norm": 0.25796136260032654,
"learning_rate": 4.13236749850815e-08,
"loss": 0.007,
"step": 1063
},
{
"epoch": 1.9240506329113924,
"grad_norm": 0.8616387248039246,
"learning_rate": 3.9426493427611177e-08,
"loss": 0.0598,
"step": 1064
},
{
"epoch": 1.9258589511754067,
"grad_norm": 0.46054112911224365,
"learning_rate": 3.7573719000832954e-08,
"loss": 0.012,
"step": 1065
},
{
"epoch": 1.9276672694394215,
"grad_norm": 0.4435920715332031,
"learning_rate": 3.576536829081323e-08,
"loss": 0.0207,
"step": 1066
},
{
"epoch": 1.9294755877034357,
"grad_norm": 0.7578052282333374,
"learning_rate": 3.400145748593542e-08,
"loss": 0.0531,
"step": 1067
},
{
"epoch": 1.9312839059674503,
"grad_norm": 0.4140692353248596,
"learning_rate": 3.2282002376756163e-08,
"loss": 0.0201,
"step": 1068
},
{
"epoch": 1.9330922242314648,
"grad_norm": 0.5629138350486755,
"learning_rate": 3.0607018355864326e-08,
"loss": 0.0281,
"step": 1069
},
{
"epoch": 1.934900542495479,
"grad_norm": 0.49671483039855957,
"learning_rate": 2.8976520417742794e-08,
"loss": 0.0406,
"step": 1070
},
{
"epoch": 1.9367088607594938,
"grad_norm": 0.43548744916915894,
"learning_rate": 2.7390523158633552e-08,
"loss": 0.0237,
"step": 1071
},
{
"epoch": 1.938517179023508,
"grad_norm": 0.4529683589935303,
"learning_rate": 2.584904077640893e-08,
"loss": 0.0266,
"step": 1072
},
{
"epoch": 1.9403254972875226,
"grad_norm": 0.5266997814178467,
"learning_rate": 2.43520870704439e-08,
"loss": 0.0245,
"step": 1073
},
{
"epoch": 1.942133815551537,
"grad_norm": 1.1091150045394897,
"learning_rate": 2.2899675441490078e-08,
"loss": 0.0251,
"step": 1074
},
{
"epoch": 1.9439421338155516,
"grad_norm": 0.39236384630203247,
"learning_rate": 2.1491818891559156e-08,
"loss": 0.0286,
"step": 1075
},
{
"epoch": 1.945750452079566,
"grad_norm": 0.43060103058815,
"learning_rate": 2.012853002380466e-08,
"loss": 0.0178,
"step": 1076
},
{
"epoch": 1.9475587703435804,
"grad_norm": 0.6426709890365601,
"learning_rate": 1.8809821042410358e-08,
"loss": 0.0279,
"step": 1077
},
{
"epoch": 1.9493670886075949,
"grad_norm": 0.20855635404586792,
"learning_rate": 1.753570375247815e-08,
"loss": 0.0053,
"step": 1078
},
{
"epoch": 1.9511754068716094,
"grad_norm": 0.5739432573318481,
"learning_rate": 1.630618955992702e-08,
"loss": 0.0218,
"step": 1079
},
{
"epoch": 1.952983725135624,
"grad_norm": 0.3275408446788788,
"learning_rate": 1.5121289471385915e-08,
"loss": 0.0156,
"step": 1080
},
{
"epoch": 1.9547920433996384,
"grad_norm": 0.5218201279640198,
"learning_rate": 1.3981014094099354e-08,
"loss": 0.0159,
"step": 1081
},
{
"epoch": 1.9566003616636527,
"grad_norm": 0.4116020202636719,
"learning_rate": 1.2885373635829756e-08,
"loss": 0.0136,
"step": 1082
},
{
"epoch": 1.9584086799276674,
"grad_norm": 0.2346084713935852,
"learning_rate": 1.1834377904768046e-08,
"loss": 0.0108,
"step": 1083
},
{
"epoch": 1.9602169981916817,
"grad_norm": 0.6031449437141418,
"learning_rate": 1.0828036309443735e-08,
"loss": 0.025,
"step": 1084
},
{
"epoch": 1.9620253164556962,
"grad_norm": 0.39427825808525085,
"learning_rate": 9.866357858642206e-09,
"loss": 0.0246,
"step": 1085
},
{
"epoch": 1.9638336347197107,
"grad_norm": 0.4683573544025421,
"learning_rate": 8.949351161324227e-09,
"loss": 0.0217,
"step": 1086
},
{
"epoch": 1.965641952983725,
"grad_norm": 0.5145902633666992,
"learning_rate": 8.077024426547675e-09,
"loss": 0.0206,
"step": 1087
},
{
"epoch": 1.9674502712477397,
"grad_norm": 0.5439916253089905,
"learning_rate": 7.249385463395375e-09,
"loss": 0.0158,
"step": 1088
},
{
"epoch": 1.969258589511754,
"grad_norm": 0.5073897242546082,
"learning_rate": 6.46644168090349e-09,
"loss": 0.0381,
"step": 1089
},
{
"epoch": 1.9710669077757685,
"grad_norm": 0.5142285823822021,
"learning_rate": 5.728200087997126e-09,
"loss": 0.0171,
"step": 1090
},
{
"epoch": 1.972875226039783,
"grad_norm": 0.3138543963432312,
"learning_rate": 5.034667293427053e-09,
"loss": 0.0133,
"step": 1091
},
{
"epoch": 1.9746835443037973,
"grad_norm": 0.3880334198474884,
"learning_rate": 4.385849505708084e-09,
"loss": 0.0227,
"step": 1092
},
{
"epoch": 1.976491862567812,
"grad_norm": 0.4833851456642151,
"learning_rate": 3.781752533068561e-09,
"loss": 0.0339,
"step": 1093
},
{
"epoch": 1.9783001808318263,
"grad_norm": 0.4207029640674591,
"learning_rate": 3.2223817833931803e-09,
"loss": 0.015,
"step": 1094
},
{
"epoch": 1.9801084990958409,
"grad_norm": 0.2624194025993347,
"learning_rate": 2.707742264178026e-09,
"loss": 0.0101,
"step": 1095
},
{
"epoch": 1.9819168173598554,
"grad_norm": 0.5078034996986389,
"learning_rate": 2.237838582483387e-09,
"loss": 0.0313,
"step": 1096
},
{
"epoch": 1.9837251356238697,
"grad_norm": 0.981736958026886,
"learning_rate": 1.8126749448943437e-09,
"loss": 0.0183,
"step": 1097
},
{
"epoch": 1.9855334538878844,
"grad_norm": 0.4022759199142456,
"learning_rate": 1.4322551574830202e-09,
"loss": 0.016,
"step": 1098
},
{
"epoch": 1.9873417721518987,
"grad_norm": 0.6016702055931091,
"learning_rate": 1.096582625772502e-09,
"loss": 0.0365,
"step": 1099
},
{
"epoch": 1.9891500904159132,
"grad_norm": 0.2897029221057892,
"learning_rate": 8.056603547090813e-10,
"loss": 0.0145,
"step": 1100
},
{
"epoch": 1.9909584086799277,
"grad_norm": 0.6789715886116028,
"learning_rate": 5.594909486328348e-10,
"loss": 0.0432,
"step": 1101
},
{
"epoch": 1.992766726943942,
"grad_norm": 0.4283156991004944,
"learning_rate": 3.580766112565304e-10,
"loss": 0.0216,
"step": 1102
},
{
"epoch": 1.9945750452079567,
"grad_norm": 0.42002472281455994,
"learning_rate": 2.0141914564453247e-10,
"loss": 0.0156,
"step": 1103
},
{
"epoch": 1.996383363471971,
"grad_norm": 0.34305939078330994,
"learning_rate": 8.951995419614889e-11,
"loss": 0.0117,
"step": 1104
},
{
"epoch": 1.9981916817359855,
"grad_norm": 0.256599485874176,
"learning_rate": 2.2380038635638758e-11,
"loss": 0.0086,
"step": 1105
},
{
"epoch": 2.0,
"grad_norm": 0.444455623626709,
"learning_rate": 0.0,
"loss": 0.0202,
"step": 1106
},
{
"epoch": 2.0,
"step": 1106,
"total_flos": 304377445122048.0,
"train_loss": 0.3876409401353381,
"train_runtime": 7641.6086,
"train_samples_per_second": 2.313,
"train_steps_per_second": 0.145
}
],
"logging_steps": 1.0,
"max_steps": 1106,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 304377445122048.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}