eb_man_ablation_history_3 / trainer_state.json
Hanyang81's picture
Upload trainer_state.json with huggingface_hub
b489896 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 504,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003968253968253968,
"grad_norm": 235.90704345703125,
"learning_rate": 3.846153846153847e-07,
"loss": 1.767,
"step": 1
},
{
"epoch": 0.007936507936507936,
"grad_norm": 241.3156280517578,
"learning_rate": 7.692307692307694e-07,
"loss": 1.8234,
"step": 2
},
{
"epoch": 0.011904761904761904,
"grad_norm": 230.72515869140625,
"learning_rate": 1.153846153846154e-06,
"loss": 1.8479,
"step": 3
},
{
"epoch": 0.015873015873015872,
"grad_norm": 213.76596069335938,
"learning_rate": 1.5384615384615387e-06,
"loss": 1.8585,
"step": 4
},
{
"epoch": 0.01984126984126984,
"grad_norm": 214.55667114257812,
"learning_rate": 1.9230769230769234e-06,
"loss": 1.7188,
"step": 5
},
{
"epoch": 0.023809523809523808,
"grad_norm": 174.86378479003906,
"learning_rate": 2.307692307692308e-06,
"loss": 1.7824,
"step": 6
},
{
"epoch": 0.027777777777777776,
"grad_norm": 145.74945068359375,
"learning_rate": 2.6923076923076923e-06,
"loss": 1.751,
"step": 7
},
{
"epoch": 0.031746031746031744,
"grad_norm": 71.36531829833984,
"learning_rate": 3.0769230769230774e-06,
"loss": 1.5968,
"step": 8
},
{
"epoch": 0.03571428571428571,
"grad_norm": 72.0484619140625,
"learning_rate": 3.4615384615384617e-06,
"loss": 1.5851,
"step": 9
},
{
"epoch": 0.03968253968253968,
"grad_norm": 112.96443176269531,
"learning_rate": 3.846153846153847e-06,
"loss": 1.5741,
"step": 10
},
{
"epoch": 0.04365079365079365,
"grad_norm": 241.17657470703125,
"learning_rate": 4.230769230769231e-06,
"loss": 1.3939,
"step": 11
},
{
"epoch": 0.047619047619047616,
"grad_norm": 25.074222564697266,
"learning_rate": 4.615384615384616e-06,
"loss": 1.3286,
"step": 12
},
{
"epoch": 0.051587301587301584,
"grad_norm": 17.152912139892578,
"learning_rate": 5e-06,
"loss": 1.2566,
"step": 13
},
{
"epoch": 0.05555555555555555,
"grad_norm": 21.181367874145508,
"learning_rate": 5.384615384615385e-06,
"loss": 1.2396,
"step": 14
},
{
"epoch": 0.05952380952380952,
"grad_norm": 20.319046020507812,
"learning_rate": 5.769230769230769e-06,
"loss": 1.0925,
"step": 15
},
{
"epoch": 0.06349206349206349,
"grad_norm": 12.114007949829102,
"learning_rate": 6.153846153846155e-06,
"loss": 0.9436,
"step": 16
},
{
"epoch": 0.06746031746031746,
"grad_norm": 13.733194351196289,
"learning_rate": 6.538461538461539e-06,
"loss": 0.914,
"step": 17
},
{
"epoch": 0.07142857142857142,
"grad_norm": 8.807427406311035,
"learning_rate": 6.923076923076923e-06,
"loss": 0.8369,
"step": 18
},
{
"epoch": 0.07539682539682539,
"grad_norm": 6.129525661468506,
"learning_rate": 7.307692307692308e-06,
"loss": 0.8301,
"step": 19
},
{
"epoch": 0.07936507936507936,
"grad_norm": 21.240720748901367,
"learning_rate": 7.692307692307694e-06,
"loss": 0.8122,
"step": 20
},
{
"epoch": 0.08333333333333333,
"grad_norm": 4.987209796905518,
"learning_rate": 8.076923076923077e-06,
"loss": 0.7484,
"step": 21
},
{
"epoch": 0.0873015873015873,
"grad_norm": 5.195412635803223,
"learning_rate": 8.461538461538462e-06,
"loss": 1.014,
"step": 22
},
{
"epoch": 0.09126984126984126,
"grad_norm": 5.63572883605957,
"learning_rate": 8.846153846153847e-06,
"loss": 1.2039,
"step": 23
},
{
"epoch": 0.09523809523809523,
"grad_norm": 5.321966171264648,
"learning_rate": 9.230769230769232e-06,
"loss": 1.0996,
"step": 24
},
{
"epoch": 0.0992063492063492,
"grad_norm": 4.601699352264404,
"learning_rate": 9.615384615384616e-06,
"loss": 1.0137,
"step": 25
},
{
"epoch": 0.10317460317460317,
"grad_norm": 3.9707424640655518,
"learning_rate": 1e-05,
"loss": 0.7417,
"step": 26
},
{
"epoch": 0.10714285714285714,
"grad_norm": 3.230611801147461,
"learning_rate": 9.999892010284378e-06,
"loss": 0.7435,
"step": 27
},
{
"epoch": 0.1111111111111111,
"grad_norm": 2.600545644760132,
"learning_rate": 9.999568045802216e-06,
"loss": 0.6923,
"step": 28
},
{
"epoch": 0.11507936507936507,
"grad_norm": 2.814257860183716,
"learning_rate": 9.999028120547456e-06,
"loss": 0.6563,
"step": 29
},
{
"epoch": 0.11904761904761904,
"grad_norm": 2.679490804672241,
"learning_rate": 9.99827225784264e-06,
"loss": 0.724,
"step": 30
},
{
"epoch": 0.12301587301587301,
"grad_norm": 2.3675742149353027,
"learning_rate": 9.99730049033793e-06,
"loss": 0.6534,
"step": 31
},
{
"epoch": 0.12698412698412698,
"grad_norm": 2.4358465671539307,
"learning_rate": 9.996112860009689e-06,
"loss": 0.7047,
"step": 32
},
{
"epoch": 0.13095238095238096,
"grad_norm": 2.435880661010742,
"learning_rate": 9.994709418158652e-06,
"loss": 0.688,
"step": 33
},
{
"epoch": 0.1349206349206349,
"grad_norm": 1.9640806913375854,
"learning_rate": 9.993090225407743e-06,
"loss": 0.5918,
"step": 34
},
{
"epoch": 0.1388888888888889,
"grad_norm": 2.118821382522583,
"learning_rate": 9.991255351699422e-06,
"loss": 0.6037,
"step": 35
},
{
"epoch": 0.14285714285714285,
"grad_norm": 2.1266019344329834,
"learning_rate": 9.98920487629269e-06,
"loss": 0.6221,
"step": 36
},
{
"epoch": 0.14682539682539683,
"grad_norm": 2.0951614379882812,
"learning_rate": 9.986938887759643e-06,
"loss": 0.6003,
"step": 37
},
{
"epoch": 0.15079365079365079,
"grad_norm": 2.0986926555633545,
"learning_rate": 9.98445748398167e-06,
"loss": 0.543,
"step": 38
},
{
"epoch": 0.15476190476190477,
"grad_norm": 2.0133657455444336,
"learning_rate": 9.981760772145201e-06,
"loss": 0.5052,
"step": 39
},
{
"epoch": 0.15873015873015872,
"grad_norm": 2.1145994663238525,
"learning_rate": 9.978848868737099e-06,
"loss": 0.5225,
"step": 40
},
{
"epoch": 0.1626984126984127,
"grad_norm": 2.259349822998047,
"learning_rate": 9.975721899539607e-06,
"loss": 0.5375,
"step": 41
},
{
"epoch": 0.16666666666666666,
"grad_norm": 2.281940460205078,
"learning_rate": 9.972379999624935e-06,
"loss": 0.5372,
"step": 42
},
{
"epoch": 0.17063492063492064,
"grad_norm": 2.0918450355529785,
"learning_rate": 9.968823313349412e-06,
"loss": 0.4573,
"step": 43
},
{
"epoch": 0.1746031746031746,
"grad_norm": 2.2035586833953857,
"learning_rate": 9.96505199434725e-06,
"loss": 0.4736,
"step": 44
},
{
"epoch": 0.17857142857142858,
"grad_norm": 2.2186124324798584,
"learning_rate": 9.961066205523917e-06,
"loss": 0.4823,
"step": 45
},
{
"epoch": 0.18253968253968253,
"grad_norm": 2.360074520111084,
"learning_rate": 9.956866119049095e-06,
"loss": 0.552,
"step": 46
},
{
"epoch": 0.1865079365079365,
"grad_norm": 2.8934898376464844,
"learning_rate": 9.952451916349242e-06,
"loss": 0.6458,
"step": 47
},
{
"epoch": 0.19047619047619047,
"grad_norm": 2.722872734069824,
"learning_rate": 9.947823788099754e-06,
"loss": 0.5951,
"step": 48
},
{
"epoch": 0.19444444444444445,
"grad_norm": 2.631500482559204,
"learning_rate": 9.942981934216731e-06,
"loss": 0.5879,
"step": 49
},
{
"epoch": 0.1984126984126984,
"grad_norm": 3.297386884689331,
"learning_rate": 9.937926563848345e-06,
"loss": 0.6188,
"step": 50
},
{
"epoch": 0.20238095238095238,
"grad_norm": 2.9066505432128906,
"learning_rate": 9.9326578953658e-06,
"loss": 0.7093,
"step": 51
},
{
"epoch": 0.20634920634920634,
"grad_norm": 2.8666725158691406,
"learning_rate": 9.9271761563539e-06,
"loss": 0.5934,
"step": 52
},
{
"epoch": 0.21031746031746032,
"grad_norm": 2.732038736343384,
"learning_rate": 9.921481583601218e-06,
"loss": 0.5566,
"step": 53
},
{
"epoch": 0.21428571428571427,
"grad_norm": 2.783133029937744,
"learning_rate": 9.915574423089872e-06,
"loss": 0.5774,
"step": 54
},
{
"epoch": 0.21825396825396826,
"grad_norm": 2.9054157733917236,
"learning_rate": 9.909454929984894e-06,
"loss": 0.5537,
"step": 55
},
{
"epoch": 0.2222222222222222,
"grad_norm": 2.9825472831726074,
"learning_rate": 9.903123368623216e-06,
"loss": 0.5632,
"step": 56
},
{
"epoch": 0.2261904761904762,
"grad_norm": 3.146071434020996,
"learning_rate": 9.896580012502238e-06,
"loss": 0.5093,
"step": 57
},
{
"epoch": 0.23015873015873015,
"grad_norm": 3.2221028804779053,
"learning_rate": 9.889825144268029e-06,
"loss": 0.5587,
"step": 58
},
{
"epoch": 0.23412698412698413,
"grad_norm": 2.982947826385498,
"learning_rate": 9.882859055703109e-06,
"loss": 0.5353,
"step": 59
},
{
"epoch": 0.23809523809523808,
"grad_norm": 3.0619916915893555,
"learning_rate": 9.875682047713847e-06,
"loss": 0.5086,
"step": 60
},
{
"epoch": 0.24206349206349206,
"grad_norm": 3.0601282119750977,
"learning_rate": 9.868294430317464e-06,
"loss": 0.4684,
"step": 61
},
{
"epoch": 0.24603174603174602,
"grad_norm": 3.114269256591797,
"learning_rate": 9.860696522628638e-06,
"loss": 0.4727,
"step": 62
},
{
"epoch": 0.25,
"grad_norm": 3.3955790996551514,
"learning_rate": 9.852888652845729e-06,
"loss": 0.5166,
"step": 63
},
{
"epoch": 0.25396825396825395,
"grad_norm": 3.3680572509765625,
"learning_rate": 9.84487115823659e-06,
"loss": 0.466,
"step": 64
},
{
"epoch": 0.25793650793650796,
"grad_norm": 3.2246522903442383,
"learning_rate": 9.836644385124006e-06,
"loss": 0.4282,
"step": 65
},
{
"epoch": 0.2619047619047619,
"grad_norm": 3.240110397338867,
"learning_rate": 9.828208688870736e-06,
"loss": 0.4553,
"step": 66
},
{
"epoch": 0.26587301587301587,
"grad_norm": 3.2070765495300293,
"learning_rate": 9.81956443386415e-06,
"loss": 0.4253,
"step": 67
},
{
"epoch": 0.2698412698412698,
"grad_norm": 3.3490705490112305,
"learning_rate": 9.810711993500506e-06,
"loss": 0.4316,
"step": 68
},
{
"epoch": 0.27380952380952384,
"grad_norm": 3.000161647796631,
"learning_rate": 9.801651750168815e-06,
"loss": 0.3579,
"step": 69
},
{
"epoch": 0.2777777777777778,
"grad_norm": 3.283475160598755,
"learning_rate": 9.792384095234312e-06,
"loss": 0.3907,
"step": 70
},
{
"epoch": 0.28174603174603174,
"grad_norm": 3.505194664001465,
"learning_rate": 9.782909429021568e-06,
"loss": 0.4034,
"step": 71
},
{
"epoch": 0.2857142857142857,
"grad_norm": 3.879484176635742,
"learning_rate": 9.773228160797187e-06,
"loss": 0.4659,
"step": 72
},
{
"epoch": 0.2896825396825397,
"grad_norm": 3.662142038345337,
"learning_rate": 9.76334070875213e-06,
"loss": 0.5106,
"step": 73
},
{
"epoch": 0.29365079365079366,
"grad_norm": 3.825267791748047,
"learning_rate": 9.753247499983649e-06,
"loss": 0.4363,
"step": 74
},
{
"epoch": 0.2976190476190476,
"grad_norm": 4.072417259216309,
"learning_rate": 9.742948970476845e-06,
"loss": 0.4333,
"step": 75
},
{
"epoch": 0.30158730158730157,
"grad_norm": 3.57492733001709,
"learning_rate": 9.732445565085823e-06,
"loss": 0.4891,
"step": 76
},
{
"epoch": 0.3055555555555556,
"grad_norm": 3.957214593887329,
"learning_rate": 9.721737737514492e-06,
"loss": 0.4559,
"step": 77
},
{
"epoch": 0.30952380952380953,
"grad_norm": 4.2466721534729,
"learning_rate": 9.71082595029695e-06,
"loss": 0.4509,
"step": 78
},
{
"epoch": 0.3134920634920635,
"grad_norm": 4.68859338760376,
"learning_rate": 9.699710674777519e-06,
"loss": 0.4541,
"step": 79
},
{
"epoch": 0.31746031746031744,
"grad_norm": 4.4894537925720215,
"learning_rate": 9.688392391090374e-06,
"loss": 0.4328,
"step": 80
},
{
"epoch": 0.32142857142857145,
"grad_norm": 4.291043758392334,
"learning_rate": 9.676871588138812e-06,
"loss": 0.4144,
"step": 81
},
{
"epoch": 0.3253968253968254,
"grad_norm": 4.541322708129883,
"learning_rate": 9.665148763574123e-06,
"loss": 0.3881,
"step": 82
},
{
"epoch": 0.32936507936507936,
"grad_norm": 4.119651794433594,
"learning_rate": 9.653224423774107e-06,
"loss": 0.3724,
"step": 83
},
{
"epoch": 0.3333333333333333,
"grad_norm": 4.078544616699219,
"learning_rate": 9.64109908382119e-06,
"loss": 0.3471,
"step": 84
},
{
"epoch": 0.3373015873015873,
"grad_norm": 3.9480912685394287,
"learning_rate": 9.628773267480177e-06,
"loss": 0.3607,
"step": 85
},
{
"epoch": 0.3412698412698413,
"grad_norm": 4.11284065246582,
"learning_rate": 9.616247507175624e-06,
"loss": 0.4117,
"step": 86
},
{
"epoch": 0.34523809523809523,
"grad_norm": 4.337547302246094,
"learning_rate": 9.603522343968852e-06,
"loss": 0.3804,
"step": 87
},
{
"epoch": 0.3492063492063492,
"grad_norm": 3.7461917400360107,
"learning_rate": 9.590598327534563e-06,
"loss": 0.3036,
"step": 88
},
{
"epoch": 0.3531746031746032,
"grad_norm": 3.533254861831665,
"learning_rate": 9.577476016137105e-06,
"loss": 0.29,
"step": 89
},
{
"epoch": 0.35714285714285715,
"grad_norm": 3.5489161014556885,
"learning_rate": 9.56415597660634e-06,
"loss": 0.2816,
"step": 90
},
{
"epoch": 0.3611111111111111,
"grad_norm": 3.6155216693878174,
"learning_rate": 9.550638784313187e-06,
"loss": 0.2905,
"step": 91
},
{
"epoch": 0.36507936507936506,
"grad_norm": 3.810203790664673,
"learning_rate": 9.536925023144742e-06,
"loss": 0.2886,
"step": 92
},
{
"epoch": 0.36904761904761907,
"grad_norm": 3.370607376098633,
"learning_rate": 9.523015285479076e-06,
"loss": 0.2496,
"step": 93
},
{
"epoch": 0.373015873015873,
"grad_norm": 3.1837635040283203,
"learning_rate": 9.508910172159635e-06,
"loss": 0.253,
"step": 94
},
{
"epoch": 0.376984126984127,
"grad_norm": 3.253502607345581,
"learning_rate": 9.494610292469287e-06,
"loss": 0.2581,
"step": 95
},
{
"epoch": 0.38095238095238093,
"grad_norm": 3.2101755142211914,
"learning_rate": 9.48011626410401e-06,
"loss": 0.2289,
"step": 96
},
{
"epoch": 0.38492063492063494,
"grad_norm": 3.2434983253479004,
"learning_rate": 9.465428713146206e-06,
"loss": 0.3129,
"step": 97
},
{
"epoch": 0.3888888888888889,
"grad_norm": 3.1097633838653564,
"learning_rate": 9.450548274037652e-06,
"loss": 0.3317,
"step": 98
},
{
"epoch": 0.39285714285714285,
"grad_norm": 3.181685447692871,
"learning_rate": 9.435475589552107e-06,
"loss": 0.3214,
"step": 99
},
{
"epoch": 0.3968253968253968,
"grad_norm": 3.569370746612549,
"learning_rate": 9.420211310767534e-06,
"loss": 0.3431,
"step": 100
},
{
"epoch": 0.4007936507936508,
"grad_norm": 3.0865094661712646,
"learning_rate": 9.40475609703798e-06,
"loss": 0.3767,
"step": 101
},
{
"epoch": 0.40476190476190477,
"grad_norm": 2.965320348739624,
"learning_rate": 9.389110615965102e-06,
"loss": 0.2968,
"step": 102
},
{
"epoch": 0.4087301587301587,
"grad_norm": 3.1394431591033936,
"learning_rate": 9.37327554336932e-06,
"loss": 0.3423,
"step": 103
},
{
"epoch": 0.4126984126984127,
"grad_norm": 2.97037410736084,
"learning_rate": 9.35725156326063e-06,
"loss": 0.2901,
"step": 104
},
{
"epoch": 0.4166666666666667,
"grad_norm": 2.8134193420410156,
"learning_rate": 9.341039367809056e-06,
"loss": 0.2411,
"step": 105
},
{
"epoch": 0.42063492063492064,
"grad_norm": 2.9861857891082764,
"learning_rate": 9.324639657314742e-06,
"loss": 0.2497,
"step": 106
},
{
"epoch": 0.4246031746031746,
"grad_norm": 2.8079447746276855,
"learning_rate": 9.308053140177722e-06,
"loss": 0.2793,
"step": 107
},
{
"epoch": 0.42857142857142855,
"grad_norm": 2.834700107574463,
"learning_rate": 9.291280532867301e-06,
"loss": 0.2178,
"step": 108
},
{
"epoch": 0.43253968253968256,
"grad_norm": 2.828073501586914,
"learning_rate": 9.27432255989112e-06,
"loss": 0.2033,
"step": 109
},
{
"epoch": 0.4365079365079365,
"grad_norm": 2.561239004135132,
"learning_rate": 9.257179953763846e-06,
"loss": 0.2015,
"step": 110
},
{
"epoch": 0.44047619047619047,
"grad_norm": 2.4958364963531494,
"learning_rate": 9.239853454975548e-06,
"loss": 0.2331,
"step": 111
},
{
"epoch": 0.4444444444444444,
"grad_norm": 2.190389633178711,
"learning_rate": 9.222343811959694e-06,
"loss": 0.2149,
"step": 112
},
{
"epoch": 0.44841269841269843,
"grad_norm": 2.0037808418273926,
"learning_rate": 9.204651781060832e-06,
"loss": 0.1764,
"step": 113
},
{
"epoch": 0.4523809523809524,
"grad_norm": 1.89824378490448,
"learning_rate": 9.186778126501916e-06,
"loss": 0.1862,
"step": 114
},
{
"epoch": 0.45634920634920634,
"grad_norm": 1.7148778438568115,
"learning_rate": 9.168723620351298e-06,
"loss": 0.1945,
"step": 115
},
{
"epoch": 0.4603174603174603,
"grad_norm": 1.7352793216705322,
"learning_rate": 9.150489042489368e-06,
"loss": 0.17,
"step": 116
},
{
"epoch": 0.4642857142857143,
"grad_norm": 1.4055451154708862,
"learning_rate": 9.13207518057488e-06,
"loss": 0.1574,
"step": 117
},
{
"epoch": 0.46825396825396826,
"grad_norm": 1.3583143949508667,
"learning_rate": 9.113482830010918e-06,
"loss": 0.1616,
"step": 118
},
{
"epoch": 0.4722222222222222,
"grad_norm": 1.2464125156402588,
"learning_rate": 9.094712793910541e-06,
"loss": 0.1484,
"step": 119
},
{
"epoch": 0.47619047619047616,
"grad_norm": 1.2419325113296509,
"learning_rate": 9.075765883062093e-06,
"loss": 0.1602,
"step": 120
},
{
"epoch": 0.4801587301587302,
"grad_norm": 1.2618135213851929,
"learning_rate": 9.056642915894182e-06,
"loss": 0.1918,
"step": 121
},
{
"epoch": 0.48412698412698413,
"grad_norm": 1.4067702293395996,
"learning_rate": 9.037344718440321e-06,
"loss": 0.235,
"step": 122
},
{
"epoch": 0.4880952380952381,
"grad_norm": 1.2719956636428833,
"learning_rate": 9.017872124303255e-06,
"loss": 0.2387,
"step": 123
},
{
"epoch": 0.49206349206349204,
"grad_norm": 1.3802380561828613,
"learning_rate": 8.99822597461894e-06,
"loss": 0.1926,
"step": 124
},
{
"epoch": 0.49603174603174605,
"grad_norm": 1.3154234886169434,
"learning_rate": 8.978407118020226e-06,
"loss": 0.1961,
"step": 125
},
{
"epoch": 0.5,
"grad_norm": 1.506628394126892,
"learning_rate": 8.958416410600188e-06,
"loss": 0.3073,
"step": 126
},
{
"epoch": 0.503968253968254,
"grad_norm": 1.3993816375732422,
"learning_rate": 8.938254715875152e-06,
"loss": 0.2744,
"step": 127
},
{
"epoch": 0.5079365079365079,
"grad_norm": 1.3298118114471436,
"learning_rate": 8.917922904747385e-06,
"loss": 0.2525,
"step": 128
},
{
"epoch": 0.5119047619047619,
"grad_norm": 1.1921308040618896,
"learning_rate": 8.897421855467491e-06,
"loss": 0.2615,
"step": 129
},
{
"epoch": 0.5158730158730159,
"grad_norm": 1.0855599641799927,
"learning_rate": 8.876752453596462e-06,
"loss": 0.2342,
"step": 130
},
{
"epoch": 0.5198412698412699,
"grad_norm": 1.1835241317749023,
"learning_rate": 8.85591559196743e-06,
"loss": 0.2069,
"step": 131
},
{
"epoch": 0.5238095238095238,
"grad_norm": 1.1165026426315308,
"learning_rate": 8.834912170647102e-06,
"loss": 0.1915,
"step": 132
},
{
"epoch": 0.5277777777777778,
"grad_norm": 1.0183204412460327,
"learning_rate": 8.813743096896872e-06,
"loss": 0.169,
"step": 133
},
{
"epoch": 0.5317460317460317,
"grad_norm": 1.1600120067596436,
"learning_rate": 8.792409285133644e-06,
"loss": 0.2266,
"step": 134
},
{
"epoch": 0.5357142857142857,
"grad_norm": 0.9856257438659668,
"learning_rate": 8.770911656890325e-06,
"loss": 0.1713,
"step": 135
},
{
"epoch": 0.5396825396825397,
"grad_norm": 1.117665410041809,
"learning_rate": 8.749251140776016e-06,
"loss": 0.1946,
"step": 136
},
{
"epoch": 0.5436507936507936,
"grad_norm": 1.1588704586029053,
"learning_rate": 8.727428672435911e-06,
"loss": 0.178,
"step": 137
},
{
"epoch": 0.5476190476190477,
"grad_norm": 1.0892035961151123,
"learning_rate": 8.705445194510868e-06,
"loss": 0.1804,
"step": 138
},
{
"epoch": 0.5515873015873016,
"grad_norm": 0.9470120668411255,
"learning_rate": 8.6833016565967e-06,
"loss": 0.1514,
"step": 139
},
{
"epoch": 0.5555555555555556,
"grad_norm": 1.0042625665664673,
"learning_rate": 8.660999015203152e-06,
"loss": 0.1564,
"step": 140
},
{
"epoch": 0.5595238095238095,
"grad_norm": 1.1161589622497559,
"learning_rate": 8.638538233712581e-06,
"loss": 0.1921,
"step": 141
},
{
"epoch": 0.5634920634920635,
"grad_norm": 1.0121771097183228,
"learning_rate": 8.615920282338355e-06,
"loss": 0.1536,
"step": 142
},
{
"epoch": 0.5674603174603174,
"grad_norm": 1.1314657926559448,
"learning_rate": 8.593146138082925e-06,
"loss": 0.1764,
"step": 143
},
{
"epoch": 0.5714285714285714,
"grad_norm": 1.0470616817474365,
"learning_rate": 8.570216784695637e-06,
"loss": 0.1478,
"step": 144
},
{
"epoch": 0.5753968253968254,
"grad_norm": 1.0094786882400513,
"learning_rate": 8.54713321263023e-06,
"loss": 0.1272,
"step": 145
},
{
"epoch": 0.5793650793650794,
"grad_norm": 1.117626667022705,
"learning_rate": 8.52389641900206e-06,
"loss": 0.142,
"step": 146
},
{
"epoch": 0.5833333333333334,
"grad_norm": 1.5540390014648438,
"learning_rate": 8.50050740754502e-06,
"loss": 0.2644,
"step": 147
},
{
"epoch": 0.5873015873015873,
"grad_norm": 1.523404836654663,
"learning_rate": 8.476967188568187e-06,
"loss": 0.2198,
"step": 148
},
{
"epoch": 0.5912698412698413,
"grad_norm": 1.3182144165039062,
"learning_rate": 8.453276778912186e-06,
"loss": 0.2263,
"step": 149
},
{
"epoch": 0.5952380952380952,
"grad_norm": 1.2298833131790161,
"learning_rate": 8.429437201905254e-06,
"loss": 0.1846,
"step": 150
},
{
"epoch": 0.5992063492063492,
"grad_norm": 1.573198676109314,
"learning_rate": 8.405449487319049e-06,
"loss": 0.3268,
"step": 151
},
{
"epoch": 0.6031746031746031,
"grad_norm": 1.4254463911056519,
"learning_rate": 8.38131467132416e-06,
"loss": 0.2043,
"step": 152
},
{
"epoch": 0.6071428571428571,
"grad_norm": 1.360297679901123,
"learning_rate": 8.357033796445356e-06,
"loss": 0.2364,
"step": 153
},
{
"epoch": 0.6111111111111112,
"grad_norm": 1.343138575553894,
"learning_rate": 8.332607911516545e-06,
"loss": 0.1968,
"step": 154
},
{
"epoch": 0.6150793650793651,
"grad_norm": 1.3606559038162231,
"learning_rate": 8.308038071635475e-06,
"loss": 0.1789,
"step": 155
},
{
"epoch": 0.6190476190476191,
"grad_norm": 1.4764060974121094,
"learning_rate": 8.283325338118154e-06,
"loss": 0.2078,
"step": 156
},
{
"epoch": 0.623015873015873,
"grad_norm": 1.4408818483352661,
"learning_rate": 8.258470778453005e-06,
"loss": 0.1745,
"step": 157
},
{
"epoch": 0.626984126984127,
"grad_norm": 1.435282588005066,
"learning_rate": 8.233475466254766e-06,
"loss": 0.1766,
"step": 158
},
{
"epoch": 0.6309523809523809,
"grad_norm": 1.4514151811599731,
"learning_rate": 8.208340481218094e-06,
"loss": 0.164,
"step": 159
},
{
"epoch": 0.6349206349206349,
"grad_norm": 1.4849510192871094,
"learning_rate": 8.183066909070946e-06,
"loss": 0.1577,
"step": 160
},
{
"epoch": 0.6388888888888888,
"grad_norm": 1.334287166595459,
"learning_rate": 8.15765584152767e-06,
"loss": 0.1657,
"step": 161
},
{
"epoch": 0.6428571428571429,
"grad_norm": 1.426286220550537,
"learning_rate": 8.132108376241849e-06,
"loss": 0.1869,
"step": 162
},
{
"epoch": 0.6468253968253969,
"grad_norm": 1.329660177230835,
"learning_rate": 8.106425616758886e-06,
"loss": 0.1543,
"step": 163
},
{
"epoch": 0.6507936507936508,
"grad_norm": 1.3017834424972534,
"learning_rate": 8.08060867246834e-06,
"loss": 0.1661,
"step": 164
},
{
"epoch": 0.6547619047619048,
"grad_norm": 1.2990241050720215,
"learning_rate": 8.054658658555998e-06,
"loss": 0.1362,
"step": 165
},
{
"epoch": 0.6587301587301587,
"grad_norm": 1.3630975484848022,
"learning_rate": 8.028576695955711e-06,
"loss": 0.167,
"step": 166
},
{
"epoch": 0.6626984126984127,
"grad_norm": 1.2524833679199219,
"learning_rate": 8.002363911300966e-06,
"loss": 0.1604,
"step": 167
},
{
"epoch": 0.6666666666666666,
"grad_norm": 1.1783356666564941,
"learning_rate": 7.976021436876232e-06,
"loss": 0.1176,
"step": 168
},
{
"epoch": 0.6706349206349206,
"grad_norm": 1.2020381689071655,
"learning_rate": 7.949550410568033e-06,
"loss": 0.1223,
"step": 169
},
{
"epoch": 0.6746031746031746,
"grad_norm": 1.2098067998886108,
"learning_rate": 7.92295197581581e-06,
"loss": 0.113,
"step": 170
},
{
"epoch": 0.6785714285714286,
"grad_norm": 1.314218282699585,
"learning_rate": 7.89622728156253e-06,
"loss": 0.1434,
"step": 171
},
{
"epoch": 0.6825396825396826,
"grad_norm": 1.7175462245941162,
"learning_rate": 7.869377482205042e-06,
"loss": 0.2399,
"step": 172
},
{
"epoch": 0.6865079365079365,
"grad_norm": 1.5188542604446411,
"learning_rate": 7.842403737544226e-06,
"loss": 0.1923,
"step": 173
},
{
"epoch": 0.6904761904761905,
"grad_norm": 1.2788623571395874,
"learning_rate": 7.815307212734888e-06,
"loss": 0.1511,
"step": 174
},
{
"epoch": 0.6944444444444444,
"grad_norm": 1.3863226175308228,
"learning_rate": 7.788089078235432e-06,
"loss": 0.1897,
"step": 175
},
{
"epoch": 0.6984126984126984,
"grad_norm": 1.6767783164978027,
"learning_rate": 7.7607505097573e-06,
"loss": 0.2528,
"step": 176
},
{
"epoch": 0.7023809523809523,
"grad_norm": 1.488150954246521,
"learning_rate": 7.733292688214182e-06,
"loss": 0.2046,
"step": 177
},
{
"epoch": 0.7063492063492064,
"grad_norm": 1.6084896326065063,
"learning_rate": 7.705716799671019e-06,
"loss": 0.2141,
"step": 178
},
{
"epoch": 0.7103174603174603,
"grad_norm": 1.359824776649475,
"learning_rate": 7.678024035292757e-06,
"loss": 0.2024,
"step": 179
},
{
"epoch": 0.7142857142857143,
"grad_norm": 1.3372007608413696,
"learning_rate": 7.650215591292888e-06,
"loss": 0.1824,
"step": 180
},
{
"epoch": 0.7182539682539683,
"grad_norm": 1.3681352138519287,
"learning_rate": 7.622292668881805e-06,
"loss": 0.1821,
"step": 181
},
{
"epoch": 0.7222222222222222,
"grad_norm": 1.3499565124511719,
"learning_rate": 7.594256474214883e-06,
"loss": 0.1455,
"step": 182
},
{
"epoch": 0.7261904761904762,
"grad_norm": 1.2687739133834839,
"learning_rate": 7.566108218340399e-06,
"loss": 0.1401,
"step": 183
},
{
"epoch": 0.7301587301587301,
"grad_norm": 1.3020589351654053,
"learning_rate": 7.537849117147212e-06,
"loss": 0.1164,
"step": 184
},
{
"epoch": 0.7341269841269841,
"grad_norm": 1.3293858766555786,
"learning_rate": 7.509480391312243e-06,
"loss": 0.1568,
"step": 185
},
{
"epoch": 0.7380952380952381,
"grad_norm": 1.1928348541259766,
"learning_rate": 7.481003266247745e-06,
"loss": 0.1394,
"step": 186
},
{
"epoch": 0.7420634920634921,
"grad_norm": 1.1654976606369019,
"learning_rate": 7.452418972048372e-06,
"loss": 0.1425,
"step": 187
},
{
"epoch": 0.746031746031746,
"grad_norm": 1.040673851966858,
"learning_rate": 7.4237287434380485e-06,
"loss": 0.115,
"step": 188
},
{
"epoch": 0.75,
"grad_norm": 1.0692291259765625,
"learning_rate": 7.394933819716625e-06,
"loss": 0.1135,
"step": 189
},
{
"epoch": 0.753968253968254,
"grad_norm": 1.0655049085617065,
"learning_rate": 7.366035444706346e-06,
"loss": 0.1115,
"step": 190
},
{
"epoch": 0.7579365079365079,
"grad_norm": 1.08991539478302,
"learning_rate": 7.337034866698138e-06,
"loss": 0.1131,
"step": 191
},
{
"epoch": 0.7619047619047619,
"grad_norm": 1.0103273391723633,
"learning_rate": 7.307933338397667e-06,
"loss": 0.1198,
"step": 192
},
{
"epoch": 0.7658730158730159,
"grad_norm": 1.0840766429901123,
"learning_rate": 7.278732116871239e-06,
"loss": 0.1194,
"step": 193
},
{
"epoch": 0.7698412698412699,
"grad_norm": 1.0601056814193726,
"learning_rate": 7.249432463491498e-06,
"loss": 0.1162,
"step": 194
},
{
"epoch": 0.7738095238095238,
"grad_norm": 0.9770033359527588,
"learning_rate": 7.220035643882938e-06,
"loss": 0.1023,
"step": 195
},
{
"epoch": 0.7777777777777778,
"grad_norm": 0.8996371626853943,
"learning_rate": 7.190542927867234e-06,
"loss": 0.0917,
"step": 196
},
{
"epoch": 0.7817460317460317,
"grad_norm": 1.501573920249939,
"learning_rate": 7.160955589408395e-06,
"loss": 0.2398,
"step": 197
},
{
"epoch": 0.7857142857142857,
"grad_norm": 1.5477681159973145,
"learning_rate": 7.131274906557725e-06,
"loss": 0.2033,
"step": 198
},
{
"epoch": 0.7896825396825397,
"grad_norm": 31.689571380615234,
"learning_rate": 7.101502161398626e-06,
"loss": 0.1534,
"step": 199
},
{
"epoch": 0.7936507936507936,
"grad_norm": 1.0546965599060059,
"learning_rate": 7.0716386399912075e-06,
"loss": 0.1479,
"step": 200
},
{
"epoch": 0.7976190476190477,
"grad_norm": 1.2119866609573364,
"learning_rate": 7.041685632316748e-06,
"loss": 0.1662,
"step": 201
},
{
"epoch": 0.8015873015873016,
"grad_norm": 1.1996482610702515,
"learning_rate": 7.0116444322219575e-06,
"loss": 0.1621,
"step": 202
},
{
"epoch": 0.8055555555555556,
"grad_norm": 1.3382314443588257,
"learning_rate": 6.981516337363099e-06,
"loss": 0.1987,
"step": 203
},
{
"epoch": 0.8095238095238095,
"grad_norm": 1.1566039323806763,
"learning_rate": 6.95130264914993e-06,
"loss": 0.156,
"step": 204
},
{
"epoch": 0.8134920634920635,
"grad_norm": 1.0945398807525635,
"learning_rate": 6.9210046726894885e-06,
"loss": 0.1641,
"step": 205
},
{
"epoch": 0.8174603174603174,
"grad_norm": 1.185166358947754,
"learning_rate": 6.890623716729724e-06,
"loss": 0.1427,
"step": 206
},
{
"epoch": 0.8214285714285714,
"grad_norm": 1.0171687602996826,
"learning_rate": 6.860161093602949e-06,
"loss": 0.1352,
"step": 207
},
{
"epoch": 0.8253968253968254,
"grad_norm": 0.9693849682807922,
"learning_rate": 6.829618119169169e-06,
"loss": 0.1211,
"step": 208
},
{
"epoch": 0.8293650793650794,
"grad_norm": 0.9703529477119446,
"learning_rate": 6.798996112759233e-06,
"loss": 0.0989,
"step": 209
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.9441483616828918,
"learning_rate": 6.768296397117848e-06,
"loss": 0.0921,
"step": 210
},
{
"epoch": 0.8373015873015873,
"grad_norm": 1.0283149480819702,
"learning_rate": 6.737520298346438e-06,
"loss": 0.1479,
"step": 211
},
{
"epoch": 0.8412698412698413,
"grad_norm": 0.9409569501876831,
"learning_rate": 6.706669145845863e-06,
"loss": 0.1255,
"step": 212
},
{
"epoch": 0.8452380952380952,
"grad_norm": 0.9471866488456726,
"learning_rate": 6.6757442722590015e-06,
"loss": 0.1276,
"step": 213
},
{
"epoch": 0.8492063492063492,
"grad_norm": 0.7956596612930298,
"learning_rate": 6.6447470134131685e-06,
"loss": 0.0958,
"step": 214
},
{
"epoch": 0.8531746031746031,
"grad_norm": 0.8528507351875305,
"learning_rate": 6.613678708262439e-06,
"loss": 0.1138,
"step": 215
},
{
"epoch": 0.8571428571428571,
"grad_norm": 0.8008120059967041,
"learning_rate": 6.5825406988297815e-06,
"loss": 0.0878,
"step": 216
},
{
"epoch": 0.8611111111111112,
"grad_norm": 0.8399624228477478,
"learning_rate": 6.551334330149114e-06,
"loss": 0.102,
"step": 217
},
{
"epoch": 0.8650793650793651,
"grad_norm": 0.8966987133026123,
"learning_rate": 6.520060950207186e-06,
"loss": 0.1127,
"step": 218
},
{
"epoch": 0.8690476190476191,
"grad_norm": 0.7477026581764221,
"learning_rate": 6.488721909885359e-06,
"loss": 0.0962,
"step": 219
},
{
"epoch": 0.873015873015873,
"grad_norm": 0.8046629428863525,
"learning_rate": 6.457318562901257e-06,
"loss": 0.1015,
"step": 220
},
{
"epoch": 0.876984126984127,
"grad_norm": 1.112833857536316,
"learning_rate": 6.425852265750282e-06,
"loss": 0.1353,
"step": 221
},
{
"epoch": 0.8809523809523809,
"grad_norm": 1.528663158416748,
"learning_rate": 6.394324377647028e-06,
"loss": 0.1828,
"step": 222
},
{
"epoch": 0.8849206349206349,
"grad_norm": 1.0469210147857666,
"learning_rate": 6.362736260466561e-06,
"loss": 0.1331,
"step": 223
},
{
"epoch": 0.8888888888888888,
"grad_norm": 1.1281147003173828,
"learning_rate": 6.331089278685599e-06,
"loss": 0.2015,
"step": 224
},
{
"epoch": 0.8928571428571429,
"grad_norm": 1.0916060209274292,
"learning_rate": 6.299384799323568e-06,
"loss": 0.1267,
"step": 225
},
{
"epoch": 0.8968253968253969,
"grad_norm": 1.2098743915557861,
"learning_rate": 6.267624191883551e-06,
"loss": 0.2375,
"step": 226
},
{
"epoch": 0.9007936507936508,
"grad_norm": 0.9832237958908081,
"learning_rate": 6.235808828293135e-06,
"loss": 0.1488,
"step": 227
},
{
"epoch": 0.9047619047619048,
"grad_norm": 1.071043848991394,
"learning_rate": 6.203940082845144e-06,
"loss": 0.1418,
"step": 228
},
{
"epoch": 0.9087301587301587,
"grad_norm": 1.0467451810836792,
"learning_rate": 6.172019332138285e-06,
"loss": 0.1786,
"step": 229
},
{
"epoch": 0.9126984126984127,
"grad_norm": 1.033318281173706,
"learning_rate": 6.140047955017672e-06,
"loss": 0.1466,
"step": 230
},
{
"epoch": 0.9166666666666666,
"grad_norm": 1.0274657011032104,
"learning_rate": 6.108027332515276e-06,
"loss": 0.1502,
"step": 231
},
{
"epoch": 0.9206349206349206,
"grad_norm": 0.9021082520484924,
"learning_rate": 6.075958847790262e-06,
"loss": 0.1241,
"step": 232
},
{
"epoch": 0.9246031746031746,
"grad_norm": 0.9323355555534363,
"learning_rate": 6.043843886069251e-06,
"loss": 0.1201,
"step": 233
},
{
"epoch": 0.9285714285714286,
"grad_norm": 0.9611258506774902,
"learning_rate": 6.011683834586474e-06,
"loss": 0.1156,
"step": 234
},
{
"epoch": 0.9325396825396826,
"grad_norm": 0.9434248805046082,
"learning_rate": 5.979480082523858e-06,
"loss": 0.1053,
"step": 235
},
{
"epoch": 0.9365079365079365,
"grad_norm": 0.9511739611625671,
"learning_rate": 5.947234020951015e-06,
"loss": 0.1351,
"step": 236
},
{
"epoch": 0.9404761904761905,
"grad_norm": 0.9289358854293823,
"learning_rate": 5.914947042765149e-06,
"loss": 0.1231,
"step": 237
},
{
"epoch": 0.9444444444444444,
"grad_norm": 0.8261659145355225,
"learning_rate": 5.882620542630901e-06,
"loss": 0.1127,
"step": 238
},
{
"epoch": 0.9484126984126984,
"grad_norm": 0.8136823773384094,
"learning_rate": 5.850255916920093e-06,
"loss": 0.1065,
"step": 239
},
{
"epoch": 0.9523809523809523,
"grad_norm": 0.9295393824577332,
"learning_rate": 5.817854563651415e-06,
"loss": 0.0889,
"step": 240
},
{
"epoch": 0.9563492063492064,
"grad_norm": 0.8647883534431458,
"learning_rate": 5.785417882430035e-06,
"loss": 0.1097,
"step": 241
},
{
"epoch": 0.9603174603174603,
"grad_norm": 0.742449164390564,
"learning_rate": 5.752947274387147e-06,
"loss": 0.0908,
"step": 242
},
{
"epoch": 0.9642857142857143,
"grad_norm": 0.7407726645469666,
"learning_rate": 5.720444142119445e-06,
"loss": 0.0911,
"step": 243
},
{
"epoch": 0.9682539682539683,
"grad_norm": 0.754944384098053,
"learning_rate": 5.687909889628529e-06,
"loss": 0.0764,
"step": 244
},
{
"epoch": 0.9722222222222222,
"grad_norm": 0.7992216348648071,
"learning_rate": 5.6553459222602714e-06,
"loss": 0.0924,
"step": 245
},
{
"epoch": 0.9761904761904762,
"grad_norm": 0.768112063407898,
"learning_rate": 5.622753646644102e-06,
"loss": 0.0879,
"step": 246
},
{
"epoch": 0.9801587301587301,
"grad_norm": 1.0914860963821411,
"learning_rate": 5.59013447063225e-06,
"loss": 0.1729,
"step": 247
},
{
"epoch": 0.9841269841269841,
"grad_norm": 1.119153380393982,
"learning_rate": 5.557489803238934e-06,
"loss": 0.1343,
"step": 248
},
{
"epoch": 0.9880952380952381,
"grad_norm": 0.9458724856376648,
"learning_rate": 5.524821054579491e-06,
"loss": 0.1392,
"step": 249
},
{
"epoch": 0.9920634920634921,
"grad_norm": 0.8553447127342224,
"learning_rate": 5.492129635809473e-06,
"loss": 0.1224,
"step": 250
},
{
"epoch": 0.996031746031746,
"grad_norm": 0.9589707851409912,
"learning_rate": 5.459416959063688e-06,
"loss": 0.1294,
"step": 251
},
{
"epoch": 1.0,
"grad_norm": 0.890637993812561,
"learning_rate": 5.426684437395196e-06,
"loss": 0.161,
"step": 252
},
{
"epoch": 1.003968253968254,
"grad_norm": 1.0126850605010986,
"learning_rate": 5.393933484714284e-06,
"loss": 0.176,
"step": 253
},
{
"epoch": 1.007936507936508,
"grad_norm": 0.9073449373245239,
"learning_rate": 5.361165515727374e-06,
"loss": 0.1484,
"step": 254
},
{
"epoch": 1.0119047619047619,
"grad_norm": 0.9067085981369019,
"learning_rate": 5.328381945875933e-06,
"loss": 0.1318,
"step": 255
},
{
"epoch": 1.0158730158730158,
"grad_norm": 0.8945247530937195,
"learning_rate": 5.295584191275308e-06,
"loss": 0.1213,
"step": 256
},
{
"epoch": 1.0198412698412698,
"grad_norm": 1.1722592115402222,
"learning_rate": 5.26277366865358e-06,
"loss": 0.1207,
"step": 257
},
{
"epoch": 1.0238095238095237,
"grad_norm": 0.8221971988677979,
"learning_rate": 5.229951795290353e-06,
"loss": 0.0904,
"step": 258
},
{
"epoch": 1.0277777777777777,
"grad_norm": 0.8209845423698425,
"learning_rate": 5.197119988955534e-06,
"loss": 0.0943,
"step": 259
},
{
"epoch": 1.0317460317460316,
"grad_norm": 0.7519062161445618,
"learning_rate": 5.164279667848094e-06,
"loss": 0.086,
"step": 260
},
{
"epoch": 1.0357142857142858,
"grad_norm": 0.9061475992202759,
"learning_rate": 5.131432250534809e-06,
"loss": 0.1044,
"step": 261
},
{
"epoch": 1.0396825396825398,
"grad_norm": 0.7555239200592041,
"learning_rate": 5.0985791558889785e-06,
"loss": 0.0739,
"step": 262
},
{
"epoch": 1.0436507936507937,
"grad_norm": 0.8547952771186829,
"learning_rate": 5.065721803029146e-06,
"loss": 0.1092,
"step": 263
},
{
"epoch": 1.0476190476190477,
"grad_norm": 0.8634958267211914,
"learning_rate": 5.032861611257783e-06,
"loss": 0.1024,
"step": 264
},
{
"epoch": 1.0515873015873016,
"grad_norm": 0.8663954734802246,
"learning_rate": 5e-06,
"loss": 0.0924,
"step": 265
},
{
"epoch": 1.0555555555555556,
"grad_norm": 0.7999219298362732,
"learning_rate": 4.967138388742218e-06,
"loss": 0.0771,
"step": 266
},
{
"epoch": 1.0595238095238095,
"grad_norm": 0.8035377264022827,
"learning_rate": 4.934278196970857e-06,
"loss": 0.079,
"step": 267
},
{
"epoch": 1.0634920634920635,
"grad_norm": 0.9056430459022522,
"learning_rate": 4.9014208441110215e-06,
"loss": 0.0816,
"step": 268
},
{
"epoch": 1.0674603174603174,
"grad_norm": 0.7253872752189636,
"learning_rate": 4.868567749465192e-06,
"loss": 0.0714,
"step": 269
},
{
"epoch": 1.0714285714285714,
"grad_norm": 0.6427790522575378,
"learning_rate": 4.835720332151907e-06,
"loss": 0.0585,
"step": 270
},
{
"epoch": 1.0753968253968254,
"grad_norm": 0.8090056777000427,
"learning_rate": 4.802880011044467e-06,
"loss": 0.0828,
"step": 271
},
{
"epoch": 1.0793650793650793,
"grad_norm": 0.7762559652328491,
"learning_rate": 4.770048204709648e-06,
"loss": 0.0631,
"step": 272
},
{
"epoch": 1.0833333333333333,
"grad_norm": 0.8392686247825623,
"learning_rate": 4.73722633134642e-06,
"loss": 0.0738,
"step": 273
},
{
"epoch": 1.0873015873015872,
"grad_norm": 0.9419091939926147,
"learning_rate": 4.7044158087246926e-06,
"loss": 0.1084,
"step": 274
},
{
"epoch": 1.0912698412698412,
"grad_norm": 1.1349728107452393,
"learning_rate": 4.67161805412407e-06,
"loss": 0.1172,
"step": 275
},
{
"epoch": 1.0952380952380953,
"grad_norm": 0.8157318234443665,
"learning_rate": 4.6388344842726266e-06,
"loss": 0.0938,
"step": 276
},
{
"epoch": 1.0992063492063493,
"grad_norm": 0.8122566938400269,
"learning_rate": 4.606066515285719e-06,
"loss": 0.1079,
"step": 277
},
{
"epoch": 1.1031746031746033,
"grad_norm": 1.0092262029647827,
"learning_rate": 4.573315562604804e-06,
"loss": 0.1485,
"step": 278
},
{
"epoch": 1.1071428571428572,
"grad_norm": 1.0518463850021362,
"learning_rate": 4.540583040936313e-06,
"loss": 0.1039,
"step": 279
},
{
"epoch": 1.1111111111111112,
"grad_norm": 0.9932454228401184,
"learning_rate": 4.5078703641905275e-06,
"loss": 0.1004,
"step": 280
},
{
"epoch": 1.1150793650793651,
"grad_norm": 0.9040830731391907,
"learning_rate": 4.4751789454205105e-06,
"loss": 0.0907,
"step": 281
},
{
"epoch": 1.119047619047619,
"grad_norm": 0.8703064322471619,
"learning_rate": 4.442510196761068e-06,
"loss": 0.0929,
"step": 282
},
{
"epoch": 1.123015873015873,
"grad_norm": 0.8670096397399902,
"learning_rate": 4.409865529367751e-06,
"loss": 0.0956,
"step": 283
},
{
"epoch": 1.126984126984127,
"grad_norm": 0.8028353452682495,
"learning_rate": 4.377246353355899e-06,
"loss": 0.0943,
"step": 284
},
{
"epoch": 1.130952380952381,
"grad_norm": 0.7226769328117371,
"learning_rate": 4.34465407773973e-06,
"loss": 0.0798,
"step": 285
},
{
"epoch": 1.1349206349206349,
"grad_norm": 0.7797760367393494,
"learning_rate": 4.312090110371473e-06,
"loss": 0.0891,
"step": 286
},
{
"epoch": 1.1388888888888888,
"grad_norm": 0.7644324898719788,
"learning_rate": 4.279555857880558e-06,
"loss": 0.0745,
"step": 287
},
{
"epoch": 1.1428571428571428,
"grad_norm": 0.7222065329551697,
"learning_rate": 4.247052725612853e-06,
"loss": 0.0779,
"step": 288
},
{
"epoch": 1.1468253968253967,
"grad_norm": 0.7465067505836487,
"learning_rate": 4.214582117569966e-06,
"loss": 0.0811,
"step": 289
},
{
"epoch": 1.1507936507936507,
"grad_norm": 0.7537140846252441,
"learning_rate": 4.182145436348587e-06,
"loss": 0.0735,
"step": 290
},
{
"epoch": 1.1547619047619047,
"grad_norm": 0.7015092968940735,
"learning_rate": 4.1497440830799084e-06,
"loss": 0.0722,
"step": 291
},
{
"epoch": 1.1587301587301586,
"grad_norm": 0.7781518697738647,
"learning_rate": 4.1173794573691e-06,
"loss": 0.0748,
"step": 292
},
{
"epoch": 1.1626984126984128,
"grad_norm": 0.7920653223991394,
"learning_rate": 4.0850529572348505e-06,
"loss": 0.0729,
"step": 293
},
{
"epoch": 1.1666666666666667,
"grad_norm": 0.7699657082557678,
"learning_rate": 4.052765979048986e-06,
"loss": 0.0784,
"step": 294
},
{
"epoch": 1.1706349206349207,
"grad_norm": 0.7221378684043884,
"learning_rate": 4.0205199174761435e-06,
"loss": 0.0587,
"step": 295
},
{
"epoch": 1.1746031746031746,
"grad_norm": 0.7864474058151245,
"learning_rate": 3.988316165413528e-06,
"loss": 0.0737,
"step": 296
},
{
"epoch": 1.1785714285714286,
"grad_norm": 0.7506036162376404,
"learning_rate": 3.956156113930752e-06,
"loss": 0.0607,
"step": 297
},
{
"epoch": 1.1825396825396826,
"grad_norm": 0.7952970862388611,
"learning_rate": 3.924041152209739e-06,
"loss": 0.1052,
"step": 298
},
{
"epoch": 1.1865079365079365,
"grad_norm": 0.98213130235672,
"learning_rate": 3.891972667484726e-06,
"loss": 0.1121,
"step": 299
},
{
"epoch": 1.1904761904761905,
"grad_norm": 0.8967873454093933,
"learning_rate": 3.859952044982329e-06,
"loss": 0.1146,
"step": 300
},
{
"epoch": 1.1944444444444444,
"grad_norm": 0.8294995427131653,
"learning_rate": 3.827980667861716e-06,
"loss": 0.1108,
"step": 301
},
{
"epoch": 1.1984126984126984,
"grad_norm": 0.6658321022987366,
"learning_rate": 3.7960599171548572e-06,
"loss": 0.0837,
"step": 302
},
{
"epoch": 1.2023809523809523,
"grad_norm": 1.0642445087432861,
"learning_rate": 3.764191171706867e-06,
"loss": 0.1586,
"step": 303
},
{
"epoch": 1.2063492063492063,
"grad_norm": 0.9127349257469177,
"learning_rate": 3.732375808116451e-06,
"loss": 0.0986,
"step": 304
},
{
"epoch": 1.2103174603174602,
"grad_norm": 0.8787649273872375,
"learning_rate": 3.7006152006764336e-06,
"loss": 0.0999,
"step": 305
},
{
"epoch": 1.2142857142857142,
"grad_norm": 0.9173398613929749,
"learning_rate": 3.6689107213144025e-06,
"loss": 0.103,
"step": 306
},
{
"epoch": 1.2182539682539684,
"grad_norm": 0.7624230980873108,
"learning_rate": 3.6372637395334416e-06,
"loss": 0.0874,
"step": 307
},
{
"epoch": 1.2222222222222223,
"grad_norm": 0.9817637801170349,
"learning_rate": 3.6056756223529734e-06,
"loss": 0.0955,
"step": 308
},
{
"epoch": 1.2261904761904763,
"grad_norm": 0.930548906326294,
"learning_rate": 3.574147734249719e-06,
"loss": 0.0911,
"step": 309
},
{
"epoch": 1.2301587301587302,
"grad_norm": 0.7714835405349731,
"learning_rate": 3.542681437098745e-06,
"loss": 0.0729,
"step": 310
},
{
"epoch": 1.2341269841269842,
"grad_norm": 0.7281912565231323,
"learning_rate": 3.5112780901146426e-06,
"loss": 0.0657,
"step": 311
},
{
"epoch": 1.2380952380952381,
"grad_norm": 0.639687180519104,
"learning_rate": 3.479939049792817e-06,
"loss": 0.0641,
"step": 312
},
{
"epoch": 1.242063492063492,
"grad_norm": 0.7349096536636353,
"learning_rate": 3.448665669850888e-06,
"loss": 0.0789,
"step": 313
},
{
"epoch": 1.246031746031746,
"grad_norm": 0.9019551873207092,
"learning_rate": 3.4174593011702197e-06,
"loss": 0.1181,
"step": 314
},
{
"epoch": 1.25,
"grad_norm": 0.8876196146011353,
"learning_rate": 3.386321291737563e-06,
"loss": 0.1097,
"step": 315
},
{
"epoch": 1.253968253968254,
"grad_norm": 0.7250246405601501,
"learning_rate": 3.3552529865868323e-06,
"loss": 0.079,
"step": 316
},
{
"epoch": 1.257936507936508,
"grad_norm": 0.6744477152824402,
"learning_rate": 3.3242557277410015e-06,
"loss": 0.0665,
"step": 317
},
{
"epoch": 1.2619047619047619,
"grad_norm": 0.7410127520561218,
"learning_rate": 3.2933308541541365e-06,
"loss": 0.0802,
"step": 318
},
{
"epoch": 1.2658730158730158,
"grad_norm": 0.7229401469230652,
"learning_rate": 3.2624797016535626e-06,
"loss": 0.0636,
"step": 319
},
{
"epoch": 1.2698412698412698,
"grad_norm": 0.6742044687271118,
"learning_rate": 3.2317036028821523e-06,
"loss": 0.0698,
"step": 320
},
{
"epoch": 1.2738095238095237,
"grad_norm": 0.7387793064117432,
"learning_rate": 3.201003887240768e-06,
"loss": 0.075,
"step": 321
},
{
"epoch": 1.2777777777777777,
"grad_norm": 0.7309768199920654,
"learning_rate": 3.1703818808308327e-06,
"loss": 0.0605,
"step": 322
},
{
"epoch": 1.2817460317460316,
"grad_norm": 0.6845470666885376,
"learning_rate": 3.1398389063970512e-06,
"loss": 0.0614,
"step": 323
},
{
"epoch": 1.2857142857142856,
"grad_norm": 0.8162021040916443,
"learning_rate": 3.1093762832702775e-06,
"loss": 0.0998,
"step": 324
},
{
"epoch": 1.2896825396825398,
"grad_norm": 0.8767663240432739,
"learning_rate": 3.0789953273105123e-06,
"loss": 0.1087,
"step": 325
},
{
"epoch": 1.2936507936507937,
"grad_norm": 0.9251488447189331,
"learning_rate": 3.048697350850073e-06,
"loss": 0.1319,
"step": 326
},
{
"epoch": 1.2976190476190477,
"grad_norm": 0.9319996237754822,
"learning_rate": 3.0184836626369034e-06,
"loss": 0.1148,
"step": 327
},
{
"epoch": 1.3015873015873016,
"grad_norm": 0.9127830862998962,
"learning_rate": 2.988355567778043e-06,
"loss": 0.1343,
"step": 328
},
{
"epoch": 1.3055555555555556,
"grad_norm": 0.9537873864173889,
"learning_rate": 2.9583143676832526e-06,
"loss": 0.1296,
"step": 329
},
{
"epoch": 1.3095238095238095,
"grad_norm": 0.9344713091850281,
"learning_rate": 2.9283613600087933e-06,
"loss": 0.1186,
"step": 330
},
{
"epoch": 1.3134920634920635,
"grad_norm": 0.8103954792022705,
"learning_rate": 2.8984978386013767e-06,
"loss": 0.0969,
"step": 331
},
{
"epoch": 1.3174603174603174,
"grad_norm": 0.7412911653518677,
"learning_rate": 2.8687250934422774e-06,
"loss": 0.0901,
"step": 332
},
{
"epoch": 1.3214285714285714,
"grad_norm": 0.7849720120429993,
"learning_rate": 2.839044410591606e-06,
"loss": 0.0899,
"step": 333
},
{
"epoch": 1.3253968253968254,
"grad_norm": 0.8258337378501892,
"learning_rate": 2.809457072132766e-06,
"loss": 0.0837,
"step": 334
},
{
"epoch": 1.3293650793650793,
"grad_norm": 0.7033872008323669,
"learning_rate": 2.779964356117063e-06,
"loss": 0.0686,
"step": 335
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.862842321395874,
"learning_rate": 2.750567536508504e-06,
"loss": 0.0759,
"step": 336
},
{
"epoch": 1.3373015873015874,
"grad_norm": 0.7889394760131836,
"learning_rate": 2.7212678831287627e-06,
"loss": 0.0854,
"step": 337
},
{
"epoch": 1.3412698412698414,
"grad_norm": 0.785755455493927,
"learning_rate": 2.692066661602333e-06,
"loss": 0.0788,
"step": 338
},
{
"epoch": 1.3452380952380953,
"grad_norm": 0.7505653500556946,
"learning_rate": 2.662965133301862e-06,
"loss": 0.0758,
"step": 339
},
{
"epoch": 1.3492063492063493,
"grad_norm": 0.6643927693367004,
"learning_rate": 2.633964555293654e-06,
"loss": 0.0701,
"step": 340
},
{
"epoch": 1.3531746031746033,
"grad_norm": 0.6269500851631165,
"learning_rate": 2.605066180283378e-06,
"loss": 0.061,
"step": 341
},
{
"epoch": 1.3571428571428572,
"grad_norm": 0.6118724346160889,
"learning_rate": 2.576271256561953e-06,
"loss": 0.0575,
"step": 342
},
{
"epoch": 1.3611111111111112,
"grad_norm": 0.8483335375785828,
"learning_rate": 2.5475810279516287e-06,
"loss": 0.0714,
"step": 343
},
{
"epoch": 1.3650793650793651,
"grad_norm": 0.6641015410423279,
"learning_rate": 2.5189967337522574e-06,
"loss": 0.0724,
"step": 344
},
{
"epoch": 1.369047619047619,
"grad_norm": 0.6798058748245239,
"learning_rate": 2.49051960868776e-06,
"loss": 0.0607,
"step": 345
},
{
"epoch": 1.373015873015873,
"grad_norm": 0.6718233227729797,
"learning_rate": 2.46215088285279e-06,
"loss": 0.0615,
"step": 346
},
{
"epoch": 1.376984126984127,
"grad_norm": 0.7738707661628723,
"learning_rate": 2.433891781659603e-06,
"loss": 0.0677,
"step": 347
},
{
"epoch": 1.380952380952381,
"grad_norm": 0.8537704348564148,
"learning_rate": 2.4057435257851173e-06,
"loss": 0.1193,
"step": 348
},
{
"epoch": 1.3849206349206349,
"grad_norm": 0.8057063817977905,
"learning_rate": 2.377707331118196e-06,
"loss": 0.1052,
"step": 349
},
{
"epoch": 1.3888888888888888,
"grad_norm": 0.6823853254318237,
"learning_rate": 2.349784408707112e-06,
"loss": 0.0933,
"step": 350
},
{
"epoch": 1.3928571428571428,
"grad_norm": 0.7286165952682495,
"learning_rate": 2.3219759647072467e-06,
"loss": 0.0808,
"step": 351
},
{
"epoch": 1.3968253968253967,
"grad_norm": 0.7728666663169861,
"learning_rate": 2.2942832003289823e-06,
"loss": 0.0908,
"step": 352
},
{
"epoch": 1.4007936507936507,
"grad_norm": 0.8903089761734009,
"learning_rate": 2.2667073117858185e-06,
"loss": 0.1235,
"step": 353
},
{
"epoch": 1.4047619047619047,
"grad_norm": 0.8520140647888184,
"learning_rate": 2.2392494902427027e-06,
"loss": 0.1086,
"step": 354
},
{
"epoch": 1.4087301587301586,
"grad_norm": 0.8345479965209961,
"learning_rate": 2.2119109217645697e-06,
"loss": 0.1069,
"step": 355
},
{
"epoch": 1.4126984126984126,
"grad_norm": 0.9587746858596802,
"learning_rate": 2.1846927872651135e-06,
"loss": 0.1391,
"step": 356
},
{
"epoch": 1.4166666666666667,
"grad_norm": 0.879552960395813,
"learning_rate": 2.1575962624557754e-06,
"loss": 0.1098,
"step": 357
},
{
"epoch": 1.4206349206349207,
"grad_norm": 0.951420783996582,
"learning_rate": 2.1306225177949584e-06,
"loss": 0.1109,
"step": 358
},
{
"epoch": 1.4246031746031746,
"grad_norm": 0.7829410433769226,
"learning_rate": 2.1037727184374705e-06,
"loss": 0.0749,
"step": 359
},
{
"epoch": 1.4285714285714286,
"grad_norm": 0.7655155062675476,
"learning_rate": 2.07704802418419e-06,
"loss": 0.0781,
"step": 360
},
{
"epoch": 1.4325396825396826,
"grad_norm": 0.6791927218437195,
"learning_rate": 2.050449589431969e-06,
"loss": 0.0599,
"step": 361
},
{
"epoch": 1.4365079365079365,
"grad_norm": 0.8351665735244751,
"learning_rate": 2.023978563123771e-06,
"loss": 0.0846,
"step": 362
},
{
"epoch": 1.4404761904761905,
"grad_norm": 0.7846753597259521,
"learning_rate": 1.997636088699035e-06,
"loss": 0.08,
"step": 363
},
{
"epoch": 1.4444444444444444,
"grad_norm": 0.9299310445785522,
"learning_rate": 1.9714233040442915e-06,
"loss": 0.1063,
"step": 364
},
{
"epoch": 1.4484126984126984,
"grad_norm": 0.8038491010665894,
"learning_rate": 1.9453413414440043e-06,
"loss": 0.0893,
"step": 365
},
{
"epoch": 1.4523809523809523,
"grad_norm": 0.7234587669372559,
"learning_rate": 1.919391327531663e-06,
"loss": 0.0758,
"step": 366
},
{
"epoch": 1.4563492063492063,
"grad_norm": 0.6610151529312134,
"learning_rate": 1.8935743832411163e-06,
"loss": 0.0703,
"step": 367
},
{
"epoch": 1.4603174603174602,
"grad_norm": 0.7390925288200378,
"learning_rate": 1.8678916237581524e-06,
"loss": 0.057,
"step": 368
},
{
"epoch": 1.4642857142857144,
"grad_norm": 0.6798797845840454,
"learning_rate": 1.8423441584723312e-06,
"loss": 0.0621,
"step": 369
},
{
"epoch": 1.4682539682539684,
"grad_norm": 0.8608136177062988,
"learning_rate": 1.816933090929055e-06,
"loss": 0.0683,
"step": 370
},
{
"epoch": 1.4722222222222223,
"grad_norm": 0.7216015458106995,
"learning_rate": 1.791659518781908e-06,
"loss": 0.0652,
"step": 371
},
{
"epoch": 1.4761904761904763,
"grad_norm": 0.6137444376945496,
"learning_rate": 1.7665245337452368e-06,
"loss": 0.054,
"step": 372
},
{
"epoch": 1.4801587301587302,
"grad_norm": 0.660344660282135,
"learning_rate": 1.7415292215469948e-06,
"loss": 0.0614,
"step": 373
},
{
"epoch": 1.4841269841269842,
"grad_norm": 0.7364965677261353,
"learning_rate": 1.716674661881848e-06,
"loss": 0.0809,
"step": 374
},
{
"epoch": 1.4880952380952381,
"grad_norm": 1.0031238794326782,
"learning_rate": 1.6919619283645262e-06,
"loss": 0.1081,
"step": 375
},
{
"epoch": 1.492063492063492,
"grad_norm": 0.765551745891571,
"learning_rate": 1.667392088483456e-06,
"loss": 0.0912,
"step": 376
},
{
"epoch": 1.496031746031746,
"grad_norm": 0.8344995975494385,
"learning_rate": 1.6429662035546451e-06,
"loss": 0.0869,
"step": 377
},
{
"epoch": 1.5,
"grad_norm": 0.844972550868988,
"learning_rate": 1.6186853286758397e-06,
"loss": 0.1095,
"step": 378
},
{
"epoch": 1.503968253968254,
"grad_norm": 0.8978922367095947,
"learning_rate": 1.5945505126809524e-06,
"loss": 0.1265,
"step": 379
},
{
"epoch": 1.507936507936508,
"grad_norm": 0.9027635455131531,
"learning_rate": 1.570562798094747e-06,
"loss": 0.1104,
"step": 380
},
{
"epoch": 1.5119047619047619,
"grad_norm": 0.8563842177391052,
"learning_rate": 1.5467232210878153e-06,
"loss": 0.0868,
"step": 381
},
{
"epoch": 1.5158730158730158,
"grad_norm": 0.7753051519393921,
"learning_rate": 1.5230328114318127e-06,
"loss": 0.0843,
"step": 382
},
{
"epoch": 1.5198412698412698,
"grad_norm": 0.8066627383232117,
"learning_rate": 1.4994925924549797e-06,
"loss": 0.0838,
"step": 383
},
{
"epoch": 1.5238095238095237,
"grad_norm": 0.6520328521728516,
"learning_rate": 1.4761035809979395e-06,
"loss": 0.0581,
"step": 384
},
{
"epoch": 1.5277777777777777,
"grad_norm": 0.8568812608718872,
"learning_rate": 1.452866787369771e-06,
"loss": 0.0933,
"step": 385
},
{
"epoch": 1.5317460317460316,
"grad_norm": 0.745733380317688,
"learning_rate": 1.4297832153043657e-06,
"loss": 0.0661,
"step": 386
},
{
"epoch": 1.5357142857142856,
"grad_norm": 0.7340918779373169,
"learning_rate": 1.4068538619170763e-06,
"loss": 0.0666,
"step": 387
},
{
"epoch": 1.5396825396825395,
"grad_norm": 0.7992677688598633,
"learning_rate": 1.3840797176616467e-06,
"loss": 0.09,
"step": 388
},
{
"epoch": 1.5436507936507935,
"grad_norm": 0.9505175352096558,
"learning_rate": 1.3614617662874197e-06,
"loss": 0.0787,
"step": 389
},
{
"epoch": 1.5476190476190477,
"grad_norm": 0.8059825897216797,
"learning_rate": 1.3390009847968505e-06,
"loss": 0.0678,
"step": 390
},
{
"epoch": 1.5515873015873016,
"grad_norm": 0.5820846557617188,
"learning_rate": 1.316698343403302e-06,
"loss": 0.0515,
"step": 391
},
{
"epoch": 1.5555555555555556,
"grad_norm": 0.6411604881286621,
"learning_rate": 1.2945548054891322e-06,
"loss": 0.057,
"step": 392
},
{
"epoch": 1.5595238095238095,
"grad_norm": 0.7719141840934753,
"learning_rate": 1.27257132756409e-06,
"loss": 0.0597,
"step": 393
},
{
"epoch": 1.5634920634920635,
"grad_norm": 0.7522517442703247,
"learning_rate": 1.2507488592239848e-06,
"loss": 0.0736,
"step": 394
},
{
"epoch": 1.5674603174603174,
"grad_norm": 0.6727685928344727,
"learning_rate": 1.2290883431096778e-06,
"loss": 0.0563,
"step": 395
},
{
"epoch": 1.5714285714285714,
"grad_norm": 0.6070297956466675,
"learning_rate": 1.2075907148663579e-06,
"loss": 0.051,
"step": 396
},
{
"epoch": 1.5753968253968254,
"grad_norm": 0.6297793984413147,
"learning_rate": 1.186256903103129e-06,
"loss": 0.0492,
"step": 397
},
{
"epoch": 1.5793650793650795,
"grad_norm": 0.790673553943634,
"learning_rate": 1.1650878293528994e-06,
"loss": 0.0792,
"step": 398
},
{
"epoch": 1.5833333333333335,
"grad_norm": 0.7880085110664368,
"learning_rate": 1.1440844080325703e-06,
"loss": 0.0991,
"step": 399
},
{
"epoch": 1.5873015873015874,
"grad_norm": 0.7657025456428528,
"learning_rate": 1.1232475464035386e-06,
"loss": 0.0957,
"step": 400
},
{
"epoch": 1.5912698412698414,
"grad_norm": 0.6656928658485413,
"learning_rate": 1.10257814453251e-06,
"loss": 0.0668,
"step": 401
},
{
"epoch": 1.5952380952380953,
"grad_norm": 0.8083186745643616,
"learning_rate": 1.0820770952526155e-06,
"loss": 0.0981,
"step": 402
},
{
"epoch": 1.5992063492063493,
"grad_norm": 0.9110124111175537,
"learning_rate": 1.0617452841248494e-06,
"loss": 0.1395,
"step": 403
},
{
"epoch": 1.6031746031746033,
"grad_norm": 0.8878102898597717,
"learning_rate": 1.0415835893998116e-06,
"loss": 0.1071,
"step": 404
},
{
"epoch": 1.6071428571428572,
"grad_norm": 0.8467666506767273,
"learning_rate": 1.0215928819797744e-06,
"loss": 0.0912,
"step": 405
},
{
"epoch": 1.6111111111111112,
"grad_norm": 0.9737541675567627,
"learning_rate": 1.0017740253810608e-06,
"loss": 0.1405,
"step": 406
},
{
"epoch": 1.6150793650793651,
"grad_norm": 0.7120548486709595,
"learning_rate": 9.821278756967467e-07,
"loss": 0.083,
"step": 407
},
{
"epoch": 1.619047619047619,
"grad_norm": 0.8963211178779602,
"learning_rate": 9.62655281559679e-07,
"loss": 0.1146,
"step": 408
},
{
"epoch": 1.623015873015873,
"grad_norm": 0.7156561613082886,
"learning_rate": 9.433570841058187e-07,
"loss": 0.0843,
"step": 409
},
{
"epoch": 1.626984126984127,
"grad_norm": 0.7529141902923584,
"learning_rate": 9.242341169379077e-07,
"loss": 0.081,
"step": 410
},
{
"epoch": 1.630952380952381,
"grad_norm": 0.711972177028656,
"learning_rate": 9.052872060894613e-07,
"loss": 0.0706,
"step": 411
},
{
"epoch": 1.6349206349206349,
"grad_norm": 0.7309200763702393,
"learning_rate": 8.865171699890835e-07,
"loss": 0.0681,
"step": 412
},
{
"epoch": 1.6388888888888888,
"grad_norm": 0.6559284329414368,
"learning_rate": 8.679248194251211e-07,
"loss": 0.0608,
"step": 413
},
{
"epoch": 1.6428571428571428,
"grad_norm": 0.770182192325592,
"learning_rate": 8.495109575106331e-07,
"loss": 0.074,
"step": 414
},
{
"epoch": 1.6468253968253967,
"grad_norm": 0.9094224572181702,
"learning_rate": 8.312763796487038e-07,
"loss": 0.0871,
"step": 415
},
{
"epoch": 1.6507936507936507,
"grad_norm": 0.7239559888839722,
"learning_rate": 8.132218734980852e-07,
"loss": 0.0568,
"step": 416
},
{
"epoch": 1.6547619047619047,
"grad_norm": 0.6560147404670715,
"learning_rate": 7.953482189391687e-07,
"loss": 0.057,
"step": 417
},
{
"epoch": 1.6587301587301586,
"grad_norm": 0.6419571042060852,
"learning_rate": 7.776561880403072e-07,
"loss": 0.052,
"step": 418
},
{
"epoch": 1.6626984126984126,
"grad_norm": 0.7225457429885864,
"learning_rate": 7.601465450244528e-07,
"loss": 0.0671,
"step": 419
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.7042800784111023,
"learning_rate": 7.42820046236154e-07,
"loss": 0.072,
"step": 420
},
{
"epoch": 1.6706349206349205,
"grad_norm": 0.6065309047698975,
"learning_rate": 7.256774401088817e-07,
"loss": 0.0533,
"step": 421
},
{
"epoch": 1.6746031746031746,
"grad_norm": 0.7448777556419373,
"learning_rate": 7.087194671326986e-07,
"loss": 0.0623,
"step": 422
},
{
"epoch": 1.6785714285714286,
"grad_norm": 0.7497352957725525,
"learning_rate": 6.91946859822279e-07,
"loss": 0.0551,
"step": 423
},
{
"epoch": 1.6825396825396826,
"grad_norm": 0.9579291939735413,
"learning_rate": 6.753603426852589e-07,
"loss": 0.1308,
"step": 424
},
{
"epoch": 1.6865079365079365,
"grad_norm": 0.7281100749969482,
"learning_rate": 6.589606321909464e-07,
"loss": 0.0884,
"step": 425
},
{
"epoch": 1.6904761904761905,
"grad_norm": 0.8299382328987122,
"learning_rate": 6.427484367393699e-07,
"loss": 0.1126,
"step": 426
},
{
"epoch": 1.6944444444444444,
"grad_norm": 0.7345311045646667,
"learning_rate": 6.267244566306801e-07,
"loss": 0.0839,
"step": 427
},
{
"epoch": 1.6984126984126984,
"grad_norm": 1.0190365314483643,
"learning_rate": 6.108893840348995e-07,
"loss": 0.1816,
"step": 428
},
{
"epoch": 1.7023809523809523,
"grad_norm": 0.9469826817512512,
"learning_rate": 5.952439029620222e-07,
"loss": 0.1314,
"step": 429
},
{
"epoch": 1.7063492063492065,
"grad_norm": 0.7321232557296753,
"learning_rate": 5.797886892324695e-07,
"loss": 0.0836,
"step": 430
},
{
"epoch": 1.7103174603174605,
"grad_norm": 0.8809959292411804,
"learning_rate": 5.645244104478947e-07,
"loss": 0.0996,
"step": 431
},
{
"epoch": 1.7142857142857144,
"grad_norm": 0.6729466319084167,
"learning_rate": 5.494517259623478e-07,
"loss": 0.0687,
"step": 432
},
{
"epoch": 1.7182539682539684,
"grad_norm": 0.8467239737510681,
"learning_rate": 5.34571286853795e-07,
"loss": 0.0985,
"step": 433
},
{
"epoch": 1.7222222222222223,
"grad_norm": 0.9383617043495178,
"learning_rate": 5.198837358959901e-07,
"loss": 0.0863,
"step": 434
},
{
"epoch": 1.7261904761904763,
"grad_norm": 0.6347936391830444,
"learning_rate": 5.05389707530714e-07,
"loss": 0.0588,
"step": 435
},
{
"epoch": 1.7301587301587302,
"grad_norm": 0.763221800327301,
"learning_rate": 4.91089827840367e-07,
"loss": 0.0704,
"step": 436
},
{
"epoch": 1.7341269841269842,
"grad_norm": 0.6416807770729065,
"learning_rate": 4.769847145209244e-07,
"loss": 0.054,
"step": 437
},
{
"epoch": 1.7380952380952381,
"grad_norm": 0.6379119157791138,
"learning_rate": 4.6307497685525894e-07,
"loss": 0.059,
"step": 438
},
{
"epoch": 1.742063492063492,
"grad_norm": 0.8275465369224548,
"learning_rate": 4.4936121568681546e-07,
"loss": 0.087,
"step": 439
},
{
"epoch": 1.746031746031746,
"grad_norm": 0.7025172710418701,
"learning_rate": 4.3584402339366174e-07,
"loss": 0.0681,
"step": 440
},
{
"epoch": 1.75,
"grad_norm": 0.7534537315368652,
"learning_rate": 4.225239838628981e-07,
"loss": 0.0816,
"step": 441
},
{
"epoch": 1.753968253968254,
"grad_norm": 0.6074185371398926,
"learning_rate": 4.0940167246543595e-07,
"loss": 0.0534,
"step": 442
},
{
"epoch": 1.757936507936508,
"grad_norm": 0.6256877779960632,
"learning_rate": 3.964776560311484e-07,
"loss": 0.0567,
"step": 443
},
{
"epoch": 1.7619047619047619,
"grad_norm": 0.6839221715927124,
"learning_rate": 3.8375249282437743e-07,
"loss": 0.0685,
"step": 444
},
{
"epoch": 1.7658730158730158,
"grad_norm": 0.7311641573905945,
"learning_rate": 3.71226732519826e-07,
"loss": 0.0646,
"step": 445
},
{
"epoch": 1.7698412698412698,
"grad_norm": 0.7500193119049072,
"learning_rate": 3.589009161788104e-07,
"loss": 0.0716,
"step": 446
},
{
"epoch": 1.7738095238095237,
"grad_norm": 0.5751092433929443,
"learning_rate": 3.4677557622589175e-07,
"loss": 0.0497,
"step": 447
},
{
"epoch": 1.7777777777777777,
"grad_norm": 0.72957444190979,
"learning_rate": 3.3485123642587657e-07,
"loss": 0.0564,
"step": 448
},
{
"epoch": 1.7817460317460316,
"grad_norm": 0.7068134546279907,
"learning_rate": 3.2312841186118937e-07,
"loss": 0.0718,
"step": 449
},
{
"epoch": 1.7857142857142856,
"grad_norm": 0.8162062168121338,
"learning_rate": 3.116076089096265e-07,
"loss": 0.0923,
"step": 450
},
{
"epoch": 1.7896825396825395,
"grad_norm": 0.7251884937286377,
"learning_rate": 3.0028932522248256e-07,
"loss": 0.0913,
"step": 451
},
{
"epoch": 1.7936507936507935,
"grad_norm": 0.728812575340271,
"learning_rate": 2.8917404970305096e-07,
"loss": 0.0783,
"step": 452
},
{
"epoch": 1.7976190476190477,
"grad_norm": 0.847225546836853,
"learning_rate": 2.782622624855097e-07,
"loss": 0.1182,
"step": 453
},
{
"epoch": 1.8015873015873016,
"grad_norm": 1.0258768796920776,
"learning_rate": 2.6755443491417786e-07,
"loss": 0.1321,
"step": 454
},
{
"epoch": 1.8055555555555556,
"grad_norm": 0.8574934005737305,
"learning_rate": 2.570510295231571e-07,
"loss": 0.0988,
"step": 455
},
{
"epoch": 1.8095238095238095,
"grad_norm": 0.8893408179283142,
"learning_rate": 2.467525000163523e-07,
"loss": 0.1102,
"step": 456
},
{
"epoch": 1.8134920634920635,
"grad_norm": 0.8654549717903137,
"learning_rate": 2.36659291247871e-07,
"loss": 0.1067,
"step": 457
},
{
"epoch": 1.8174603174603174,
"grad_norm": 0.9400739669799805,
"learning_rate": 2.2677183920281342e-07,
"loss": 0.1177,
"step": 458
},
{
"epoch": 1.8214285714285714,
"grad_norm": 0.7756461501121521,
"learning_rate": 2.1709057097843266e-07,
"loss": 0.0843,
"step": 459
},
{
"epoch": 1.8253968253968254,
"grad_norm": 0.7438918352127075,
"learning_rate": 2.0761590476568893e-07,
"loss": 0.0614,
"step": 460
},
{
"epoch": 1.8293650793650795,
"grad_norm": 0.694450855255127,
"learning_rate": 1.9834824983118673e-07,
"loss": 0.0611,
"step": 461
},
{
"epoch": 1.8333333333333335,
"grad_norm": 0.7721295952796936,
"learning_rate": 1.892880064994934e-07,
"loss": 0.0717,
"step": 462
},
{
"epoch": 1.8373015873015874,
"grad_norm": 0.6843567490577698,
"learning_rate": 1.8043556613585143e-07,
"loss": 0.0491,
"step": 463
},
{
"epoch": 1.8412698412698414,
"grad_norm": 0.826382577419281,
"learning_rate": 1.7179131112926628e-07,
"loss": 0.0847,
"step": 464
},
{
"epoch": 1.8452380952380953,
"grad_norm": 0.7417656183242798,
"learning_rate": 1.6335561487599406e-07,
"loss": 0.0738,
"step": 465
},
{
"epoch": 1.8492063492063493,
"grad_norm": 0.6591057181358337,
"learning_rate": 1.551288417634106e-07,
"loss": 0.0606,
"step": 466
},
{
"epoch": 1.8531746031746033,
"grad_norm": 0.5959115624427795,
"learning_rate": 1.471113471542712e-07,
"loss": 0.0505,
"step": 467
},
{
"epoch": 1.8571428571428572,
"grad_norm": 0.6840454339981079,
"learning_rate": 1.3930347737136195e-07,
"loss": 0.0564,
"step": 468
},
{
"epoch": 1.8611111111111112,
"grad_norm": 0.7096229195594788,
"learning_rate": 1.3170556968253756e-07,
"loss": 0.0611,
"step": 469
},
{
"epoch": 1.8650793650793651,
"grad_norm": 0.5797507762908936,
"learning_rate": 1.2431795228615372e-07,
"loss": 0.0496,
"step": 470
},
{
"epoch": 1.869047619047619,
"grad_norm": 0.6354245543479919,
"learning_rate": 1.1714094429689127e-07,
"loss": 0.0535,
"step": 471
},
{
"epoch": 1.873015873015873,
"grad_norm": 0.6456458568572998,
"learning_rate": 1.1017485573197151e-07,
"loss": 0.0432,
"step": 472
},
{
"epoch": 1.876984126984127,
"grad_norm": 0.6524754762649536,
"learning_rate": 1.0341998749776316e-07,
"loss": 0.0562,
"step": 473
},
{
"epoch": 1.880952380952381,
"grad_norm": 0.8275721669197083,
"learning_rate": 9.687663137678605e-08,
"loss": 0.0928,
"step": 474
},
{
"epoch": 1.8849206349206349,
"grad_norm": 0.8986220955848694,
"learning_rate": 9.054507001510727e-08,
"loss": 0.114,
"step": 475
},
{
"epoch": 1.8888888888888888,
"grad_norm": 0.6796401143074036,
"learning_rate": 8.442557691013042e-08,
"loss": 0.086,
"step": 476
},
{
"epoch": 1.8928571428571428,
"grad_norm": 0.8908485174179077,
"learning_rate": 7.851841639878399e-08,
"loss": 0.088,
"step": 477
},
{
"epoch": 1.8968253968253967,
"grad_norm": 0.9731160998344421,
"learning_rate": 7.282384364610207e-08,
"loss": 0.1339,
"step": 478
},
{
"epoch": 1.9007936507936507,
"grad_norm": 0.8356693983078003,
"learning_rate": 6.734210463420099e-08,
"loss": 0.0914,
"step": 479
},
{
"epoch": 1.9047619047619047,
"grad_norm": 0.790658175945282,
"learning_rate": 6.207343615165562e-08,
"loss": 0.0786,
"step": 480
},
{
"epoch": 1.9087301587301586,
"grad_norm": 0.8288230895996094,
"learning_rate": 5.701806578327029e-08,
"loss": 0.0895,
"step": 481
},
{
"epoch": 1.9126984126984126,
"grad_norm": 0.7168080806732178,
"learning_rate": 5.21762119002478e-08,
"loss": 0.0773,
"step": 482
},
{
"epoch": 1.9166666666666665,
"grad_norm": 0.7145270705223083,
"learning_rate": 4.7548083650759134e-08,
"loss": 0.0699,
"step": 483
},
{
"epoch": 1.9206349206349205,
"grad_norm": 0.7667180895805359,
"learning_rate": 4.31338809509052e-08,
"loss": 0.08,
"step": 484
},
{
"epoch": 1.9246031746031746,
"grad_norm": 0.6273404955863953,
"learning_rate": 3.8933794476083143e-08,
"loss": 0.0566,
"step": 485
},
{
"epoch": 1.9285714285714286,
"grad_norm": 0.7412590980529785,
"learning_rate": 3.494800565275125e-08,
"loss": 0.0546,
"step": 486
},
{
"epoch": 1.9325396825396826,
"grad_norm": 0.6933059096336365,
"learning_rate": 3.1176686650589147e-08,
"loss": 0.0641,
"step": 487
},
{
"epoch": 1.9365079365079365,
"grad_norm": 0.7381360530853271,
"learning_rate": 2.7620000375064848e-08,
"loss": 0.073,
"step": 488
},
{
"epoch": 1.9404761904761905,
"grad_norm": 0.7223060131072998,
"learning_rate": 2.4278100460393138e-08,
"loss": 0.057,
"step": 489
},
{
"epoch": 1.9444444444444444,
"grad_norm": 0.6583513021469116,
"learning_rate": 2.115113126290258e-08,
"loss": 0.0565,
"step": 490
},
{
"epoch": 1.9484126984126984,
"grad_norm": 0.7108239531517029,
"learning_rate": 1.8239227854799368e-08,
"loss": 0.0595,
"step": 491
},
{
"epoch": 1.9523809523809523,
"grad_norm": 0.6006252765655518,
"learning_rate": 1.554251601833201e-08,
"loss": 0.0485,
"step": 492
},
{
"epoch": 1.9563492063492065,
"grad_norm": 0.6265171766281128,
"learning_rate": 1.3061112240357887e-08,
"loss": 0.0559,
"step": 493
},
{
"epoch": 1.9603174603174605,
"grad_norm": 0.5632744431495667,
"learning_rate": 1.0795123707312283e-08,
"loss": 0.0491,
"step": 494
},
{
"epoch": 1.9642857142857144,
"grad_norm": 0.6409162282943726,
"learning_rate": 8.744648300578196e-09,
"loss": 0.054,
"step": 495
},
{
"epoch": 1.9682539682539684,
"grad_norm": 0.6455246806144714,
"learning_rate": 6.9097745922580564e-09,
"loss": 0.0574,
"step": 496
},
{
"epoch": 1.9722222222222223,
"grad_norm": 0.584445595741272,
"learning_rate": 5.2905818413478975e-09,
"loss": 0.0522,
"step": 497
},
{
"epoch": 1.9761904761904763,
"grad_norm": 0.7335327863693237,
"learning_rate": 3.887139990313427e-09,
"loss": 0.0798,
"step": 498
},
{
"epoch": 1.9801587301587302,
"grad_norm": 0.8302753567695618,
"learning_rate": 2.699509662069666e-09,
"loss": 0.1092,
"step": 499
},
{
"epoch": 1.9841269841269842,
"grad_norm": 0.6258664727210999,
"learning_rate": 1.7277421573608234e-09,
"loss": 0.0769,
"step": 500
},
{
"epoch": 1.9880952380952381,
"grad_norm": 0.7064999938011169,
"learning_rate": 9.71879452545399e-10,
"loss": 0.0905,
"step": 501
},
{
"epoch": 1.992063492063492,
"grad_norm": 0.8868645429611206,
"learning_rate": 4.3195419778319095e-10,
"loss": 0.114,
"step": 502
},
{
"epoch": 1.996031746031746,
"grad_norm": 0.7165728211402893,
"learning_rate": 1.0798971562364647e-10,
"loss": 0.0612,
"step": 503
},
{
"epoch": 2.0,
"grad_norm": 0.706081748008728,
"learning_rate": 0.0,
"loss": 0.0763,
"step": 504
},
{
"epoch": 2.0,
"step": 504,
"total_flos": 78280651210752.0,
"train_loss": 0.23003725075562084,
"train_runtime": 2992.2373,
"train_samples_per_second": 2.689,
"train_steps_per_second": 0.168
}
],
"logging_steps": 1.0,
"max_steps": 504,
"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": 78280651210752.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}