eb_man_ablation_history_15 / trainer_state.json
Hanyang81's picture
Upload trainer_state.json with huggingface_hub
ba3e33e 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": 516.6194458007812,
"learning_rate": 3.846153846153847e-07,
"loss": 3.1419,
"step": 1
},
{
"epoch": 0.007936507936507936,
"grad_norm": 547.3426513671875,
"learning_rate": 7.692307692307694e-07,
"loss": 3.4394,
"step": 2
},
{
"epoch": 0.011904761904761904,
"grad_norm": 638.59521484375,
"learning_rate": 1.153846153846154e-06,
"loss": 3.5139,
"step": 3
},
{
"epoch": 0.015873015873015872,
"grad_norm": 697.9334106445312,
"learning_rate": 1.5384615384615387e-06,
"loss": 3.5107,
"step": 4
},
{
"epoch": 0.01984126984126984,
"grad_norm": 691.7098388671875,
"learning_rate": 1.9230769230769234e-06,
"loss": 3.4129,
"step": 5
},
{
"epoch": 0.023809523809523808,
"grad_norm": 639.7687377929688,
"learning_rate": 2.307692307692308e-06,
"loss": 3.1175,
"step": 6
},
{
"epoch": 0.027777777777777776,
"grad_norm": 599.4917602539062,
"learning_rate": 2.6923076923076923e-06,
"loss": 3.1014,
"step": 7
},
{
"epoch": 0.031746031746031744,
"grad_norm": 71.47279357910156,
"learning_rate": 3.0769230769230774e-06,
"loss": 2.5963,
"step": 8
},
{
"epoch": 0.03571428571428571,
"grad_norm": 73.69940185546875,
"learning_rate": 3.4615384615384617e-06,
"loss": 2.4571,
"step": 9
},
{
"epoch": 0.03968253968253968,
"grad_norm": 122.35843658447266,
"learning_rate": 3.846153846153847e-06,
"loss": 2.4406,
"step": 10
},
{
"epoch": 0.04365079365079365,
"grad_norm": 21.891984939575195,
"learning_rate": 4.230769230769231e-06,
"loss": 2.1359,
"step": 11
},
{
"epoch": 0.047619047619047616,
"grad_norm": 17.47292709350586,
"learning_rate": 4.615384615384616e-06,
"loss": 1.8713,
"step": 12
},
{
"epoch": 0.051587301587301584,
"grad_norm": 20.601985931396484,
"learning_rate": 5e-06,
"loss": 1.8935,
"step": 13
},
{
"epoch": 0.05555555555555555,
"grad_norm": 24.105579376220703,
"learning_rate": 5.384615384615385e-06,
"loss": 1.8085,
"step": 14
},
{
"epoch": 0.05952380952380952,
"grad_norm": 17.53595733642578,
"learning_rate": 5.769230769230769e-06,
"loss": 1.634,
"step": 15
},
{
"epoch": 0.06349206349206349,
"grad_norm": 17.780704498291016,
"learning_rate": 6.153846153846155e-06,
"loss": 1.5045,
"step": 16
},
{
"epoch": 0.06746031746031746,
"grad_norm": 16.221433639526367,
"learning_rate": 6.538461538461539e-06,
"loss": 1.423,
"step": 17
},
{
"epoch": 0.07142857142857142,
"grad_norm": 13.835611343383789,
"learning_rate": 6.923076923076923e-06,
"loss": 1.3365,
"step": 18
},
{
"epoch": 0.07539682539682539,
"grad_norm": 9.73637580871582,
"learning_rate": 7.307692307692308e-06,
"loss": 1.1891,
"step": 19
},
{
"epoch": 0.07936507936507936,
"grad_norm": 15.151634216308594,
"learning_rate": 7.692307692307694e-06,
"loss": 1.1896,
"step": 20
},
{
"epoch": 0.08333333333333333,
"grad_norm": 7.9667205810546875,
"learning_rate": 8.076923076923077e-06,
"loss": 1.0527,
"step": 21
},
{
"epoch": 0.0873015873015873,
"grad_norm": 5.5614237785339355,
"learning_rate": 8.461538461538462e-06,
"loss": 1.0847,
"step": 22
},
{
"epoch": 0.09126984126984126,
"grad_norm": 4.895917892456055,
"learning_rate": 8.846153846153847e-06,
"loss": 0.9499,
"step": 23
},
{
"epoch": 0.09523809523809523,
"grad_norm": 4.149717330932617,
"learning_rate": 9.230769230769232e-06,
"loss": 0.8891,
"step": 24
},
{
"epoch": 0.0992063492063492,
"grad_norm": 4.035318851470947,
"learning_rate": 9.615384615384616e-06,
"loss": 0.8819,
"step": 25
},
{
"epoch": 0.10317460317460317,
"grad_norm": 3.4925167560577393,
"learning_rate": 1e-05,
"loss": 0.9215,
"step": 26
},
{
"epoch": 0.10714285714285714,
"grad_norm": 2.9664218425750732,
"learning_rate": 9.999892010284378e-06,
"loss": 0.879,
"step": 27
},
{
"epoch": 0.1111111111111111,
"grad_norm": 2.8206546306610107,
"learning_rate": 9.999568045802216e-06,
"loss": 0.8166,
"step": 28
},
{
"epoch": 0.11507936507936507,
"grad_norm": 2.932323932647705,
"learning_rate": 9.999028120547456e-06,
"loss": 0.7508,
"step": 29
},
{
"epoch": 0.11904761904761904,
"grad_norm": 2.47392201423645,
"learning_rate": 9.99827225784264e-06,
"loss": 0.6898,
"step": 30
},
{
"epoch": 0.12301587301587301,
"grad_norm": 2.220811605453491,
"learning_rate": 9.99730049033793e-06,
"loss": 0.6827,
"step": 31
},
{
"epoch": 0.12698412698412698,
"grad_norm": 2.2195913791656494,
"learning_rate": 9.996112860009689e-06,
"loss": 0.6567,
"step": 32
},
{
"epoch": 0.13095238095238096,
"grad_norm": 2.149188995361328,
"learning_rate": 9.994709418158652e-06,
"loss": 0.6401,
"step": 33
},
{
"epoch": 0.1349206349206349,
"grad_norm": 2.0333609580993652,
"learning_rate": 9.993090225407743e-06,
"loss": 0.6152,
"step": 34
},
{
"epoch": 0.1388888888888889,
"grad_norm": 2.2172834873199463,
"learning_rate": 9.991255351699422e-06,
"loss": 0.6447,
"step": 35
},
{
"epoch": 0.14285714285714285,
"grad_norm": 5.394314289093018,
"learning_rate": 9.98920487629269e-06,
"loss": 0.6048,
"step": 36
},
{
"epoch": 0.14682539682539683,
"grad_norm": 2.1033711433410645,
"learning_rate": 9.986938887759643e-06,
"loss": 0.6502,
"step": 37
},
{
"epoch": 0.15079365079365079,
"grad_norm": 2.02223801612854,
"learning_rate": 9.98445748398167e-06,
"loss": 0.6132,
"step": 38
},
{
"epoch": 0.15476190476190477,
"grad_norm": 2.3250157833099365,
"learning_rate": 9.981760772145201e-06,
"loss": 0.6628,
"step": 39
},
{
"epoch": 0.15873015873015872,
"grad_norm": 2.1451687812805176,
"learning_rate": 9.978848868737099e-06,
"loss": 0.5913,
"step": 40
},
{
"epoch": 0.1626984126984127,
"grad_norm": 2.299617290496826,
"learning_rate": 9.975721899539607e-06,
"loss": 0.6559,
"step": 41
},
{
"epoch": 0.16666666666666666,
"grad_norm": 2.4550540447235107,
"learning_rate": 9.972379999624935e-06,
"loss": 0.609,
"step": 42
},
{
"epoch": 0.17063492063492064,
"grad_norm": 2.3630247116088867,
"learning_rate": 9.968823313349412e-06,
"loss": 0.5839,
"step": 43
},
{
"epoch": 0.1746031746031746,
"grad_norm": 2.4661998748779297,
"learning_rate": 9.96505199434725e-06,
"loss": 0.568,
"step": 44
},
{
"epoch": 0.17857142857142858,
"grad_norm": 2.534061908721924,
"learning_rate": 9.961066205523917e-06,
"loss": 0.5719,
"step": 45
},
{
"epoch": 0.18253968253968253,
"grad_norm": 2.6632919311523438,
"learning_rate": 9.956866119049095e-06,
"loss": 0.6086,
"step": 46
},
{
"epoch": 0.1865079365079365,
"grad_norm": 2.699021339416504,
"learning_rate": 9.952451916349242e-06,
"loss": 0.6159,
"step": 47
},
{
"epoch": 0.19047619047619047,
"grad_norm": 2.7833011150360107,
"learning_rate": 9.947823788099754e-06,
"loss": 0.5831,
"step": 48
},
{
"epoch": 0.19444444444444445,
"grad_norm": 2.669100284576416,
"learning_rate": 9.942981934216731e-06,
"loss": 0.5561,
"step": 49
},
{
"epoch": 0.1984126984126984,
"grad_norm": 2.9000258445739746,
"learning_rate": 9.937926563848345e-06,
"loss": 0.6182,
"step": 50
},
{
"epoch": 0.20238095238095238,
"grad_norm": 2.731212854385376,
"learning_rate": 9.9326578953658e-06,
"loss": 0.6935,
"step": 51
},
{
"epoch": 0.20634920634920634,
"grad_norm": 2.7918338775634766,
"learning_rate": 9.9271761563539e-06,
"loss": 0.6478,
"step": 52
},
{
"epoch": 0.21031746031746032,
"grad_norm": 2.8246572017669678,
"learning_rate": 9.921481583601218e-06,
"loss": 0.6283,
"step": 53
},
{
"epoch": 0.21428571428571427,
"grad_norm": 3.2301411628723145,
"learning_rate": 9.915574423089872e-06,
"loss": 0.6046,
"step": 54
},
{
"epoch": 0.21825396825396826,
"grad_norm": 2.781745672225952,
"learning_rate": 9.909454929984894e-06,
"loss": 0.5791,
"step": 55
},
{
"epoch": 0.2222222222222222,
"grad_norm": 2.5260448455810547,
"learning_rate": 9.903123368623216e-06,
"loss": 0.5277,
"step": 56
},
{
"epoch": 0.2261904761904762,
"grad_norm": 2.536945343017578,
"learning_rate": 9.896580012502238e-06,
"loss": 0.4975,
"step": 57
},
{
"epoch": 0.23015873015873015,
"grad_norm": 2.782017707824707,
"learning_rate": 9.889825144268029e-06,
"loss": 0.5542,
"step": 58
},
{
"epoch": 0.23412698412698413,
"grad_norm": 2.78955078125,
"learning_rate": 9.882859055703109e-06,
"loss": 0.5299,
"step": 59
},
{
"epoch": 0.23809523809523808,
"grad_norm": 2.9246230125427246,
"learning_rate": 9.875682047713847e-06,
"loss": 0.5175,
"step": 60
},
{
"epoch": 0.24206349206349206,
"grad_norm": 2.78791880607605,
"learning_rate": 9.868294430317464e-06,
"loss": 0.467,
"step": 61
},
{
"epoch": 0.24603174603174602,
"grad_norm": 3.097339630126953,
"learning_rate": 9.860696522628638e-06,
"loss": 0.525,
"step": 62
},
{
"epoch": 0.25,
"grad_norm": 3.0526726245880127,
"learning_rate": 9.852888652845729e-06,
"loss": 0.5027,
"step": 63
},
{
"epoch": 0.25396825396825395,
"grad_norm": 3.2078192234039307,
"learning_rate": 9.84487115823659e-06,
"loss": 0.4953,
"step": 64
},
{
"epoch": 0.25793650793650796,
"grad_norm": 3.360300064086914,
"learning_rate": 9.836644385124006e-06,
"loss": 0.5221,
"step": 65
},
{
"epoch": 0.2619047619047619,
"grad_norm": 3.363569974899292,
"learning_rate": 9.828208688870736e-06,
"loss": 0.4947,
"step": 66
},
{
"epoch": 0.26587301587301587,
"grad_norm": 3.6224539279937744,
"learning_rate": 9.81956443386415e-06,
"loss": 0.5421,
"step": 67
},
{
"epoch": 0.2698412698412698,
"grad_norm": 3.6316075325012207,
"learning_rate": 9.810711993500506e-06,
"loss": 0.4698,
"step": 68
},
{
"epoch": 0.27380952380952384,
"grad_norm": 3.8527607917785645,
"learning_rate": 9.801651750168815e-06,
"loss": 0.4784,
"step": 69
},
{
"epoch": 0.2777777777777778,
"grad_norm": 3.931734085083008,
"learning_rate": 9.792384095234312e-06,
"loss": 0.4699,
"step": 70
},
{
"epoch": 0.28174603174603174,
"grad_norm": 3.845679998397827,
"learning_rate": 9.782909429021568e-06,
"loss": 0.4503,
"step": 71
},
{
"epoch": 0.2857142857142857,
"grad_norm": 3.8892247676849365,
"learning_rate": 9.773228160797187e-06,
"loss": 0.4548,
"step": 72
},
{
"epoch": 0.2896825396825397,
"grad_norm": 4.180114269256592,
"learning_rate": 9.76334070875213e-06,
"loss": 0.4459,
"step": 73
},
{
"epoch": 0.29365079365079366,
"grad_norm": 4.126506328582764,
"learning_rate": 9.753247499983649e-06,
"loss": 0.4501,
"step": 74
},
{
"epoch": 0.2976190476190476,
"grad_norm": 4.344264984130859,
"learning_rate": 9.742948970476845e-06,
"loss": 0.4374,
"step": 75
},
{
"epoch": 0.30158730158730157,
"grad_norm": 3.8102173805236816,
"learning_rate": 9.732445565085823e-06,
"loss": 0.5537,
"step": 76
},
{
"epoch": 0.3055555555555556,
"grad_norm": 4.063323974609375,
"learning_rate": 9.721737737514492e-06,
"loss": 0.5051,
"step": 77
},
{
"epoch": 0.30952380952380953,
"grad_norm": 4.042793273925781,
"learning_rate": 9.71082595029695e-06,
"loss": 0.4795,
"step": 78
},
{
"epoch": 0.3134920634920635,
"grad_norm": 4.043231964111328,
"learning_rate": 9.699710674777519e-06,
"loss": 0.4432,
"step": 79
},
{
"epoch": 0.31746031746031744,
"grad_norm": 3.8575873374938965,
"learning_rate": 9.688392391090374e-06,
"loss": 0.431,
"step": 80
},
{
"epoch": 0.32142857142857145,
"grad_norm": 3.8616952896118164,
"learning_rate": 9.676871588138812e-06,
"loss": 0.3887,
"step": 81
},
{
"epoch": 0.3253968253968254,
"grad_norm": 3.711002826690674,
"learning_rate": 9.665148763574123e-06,
"loss": 0.3882,
"step": 82
},
{
"epoch": 0.32936507936507936,
"grad_norm": 3.8032736778259277,
"learning_rate": 9.653224423774107e-06,
"loss": 0.3502,
"step": 83
},
{
"epoch": 0.3333333333333333,
"grad_norm": 3.8471624851226807,
"learning_rate": 9.64109908382119e-06,
"loss": 0.3883,
"step": 84
},
{
"epoch": 0.3373015873015873,
"grad_norm": 3.822849750518799,
"learning_rate": 9.628773267480177e-06,
"loss": 0.3725,
"step": 85
},
{
"epoch": 0.3412698412698413,
"grad_norm": 3.840853452682495,
"learning_rate": 9.616247507175624e-06,
"loss": 0.3746,
"step": 86
},
{
"epoch": 0.34523809523809523,
"grad_norm": 3.8119332790374756,
"learning_rate": 9.603522343968852e-06,
"loss": 0.3716,
"step": 87
},
{
"epoch": 0.3492063492063492,
"grad_norm": 3.8732521533966064,
"learning_rate": 9.590598327534563e-06,
"loss": 0.3797,
"step": 88
},
{
"epoch": 0.3531746031746032,
"grad_norm": 3.7719802856445312,
"learning_rate": 9.577476016137105e-06,
"loss": 0.3485,
"step": 89
},
{
"epoch": 0.35714285714285715,
"grad_norm": 3.8533449172973633,
"learning_rate": 9.56415597660634e-06,
"loss": 0.3506,
"step": 90
},
{
"epoch": 0.3611111111111111,
"grad_norm": 3.943777322769165,
"learning_rate": 9.550638784313187e-06,
"loss": 0.3451,
"step": 91
},
{
"epoch": 0.36507936507936506,
"grad_norm": 3.823855400085449,
"learning_rate": 9.536925023144742e-06,
"loss": 0.3288,
"step": 92
},
{
"epoch": 0.36904761904761907,
"grad_norm": 3.940521240234375,
"learning_rate": 9.523015285479076e-06,
"loss": 0.3393,
"step": 93
},
{
"epoch": 0.373015873015873,
"grad_norm": 3.901230573654175,
"learning_rate": 9.508910172159635e-06,
"loss": 0.3467,
"step": 94
},
{
"epoch": 0.376984126984127,
"grad_norm": 3.6534783840179443,
"learning_rate": 9.494610292469287e-06,
"loss": 0.3097,
"step": 95
},
{
"epoch": 0.38095238095238093,
"grad_norm": 3.7912635803222656,
"learning_rate": 9.48011626410401e-06,
"loss": 0.3364,
"step": 96
},
{
"epoch": 0.38492063492063494,
"grad_norm": 3.9513766765594482,
"learning_rate": 9.465428713146206e-06,
"loss": 0.3216,
"step": 97
},
{
"epoch": 0.3888888888888889,
"grad_norm": 3.6714870929718018,
"learning_rate": 9.450548274037652e-06,
"loss": 0.312,
"step": 98
},
{
"epoch": 0.39285714285714285,
"grad_norm": 3.496143102645874,
"learning_rate": 9.435475589552107e-06,
"loss": 0.2938,
"step": 99
},
{
"epoch": 0.3968253968253968,
"grad_norm": 3.759218692779541,
"learning_rate": 9.420211310767534e-06,
"loss": 0.3045,
"step": 100
},
{
"epoch": 0.4007936507936508,
"grad_norm": 3.053839921951294,
"learning_rate": 9.40475609703798e-06,
"loss": 0.4447,
"step": 101
},
{
"epoch": 0.40476190476190477,
"grad_norm": 3.100240707397461,
"learning_rate": 9.389110615965102e-06,
"loss": 0.381,
"step": 102
},
{
"epoch": 0.4087301587301587,
"grad_norm": 3.1526577472686768,
"learning_rate": 9.37327554336932e-06,
"loss": 0.2965,
"step": 103
},
{
"epoch": 0.4126984126984127,
"grad_norm": 3.0102667808532715,
"learning_rate": 9.35725156326063e-06,
"loss": 0.3458,
"step": 104
},
{
"epoch": 0.4166666666666667,
"grad_norm": 3.083493709564209,
"learning_rate": 9.341039367809056e-06,
"loss": 0.3076,
"step": 105
},
{
"epoch": 0.42063492063492064,
"grad_norm": 2.747546672821045,
"learning_rate": 9.324639657314742e-06,
"loss": 0.2691,
"step": 106
},
{
"epoch": 0.4246031746031746,
"grad_norm": 2.5286612510681152,
"learning_rate": 9.308053140177722e-06,
"loss": 0.2749,
"step": 107
},
{
"epoch": 0.42857142857142855,
"grad_norm": 2.553208351135254,
"learning_rate": 9.291280532867301e-06,
"loss": 0.2665,
"step": 108
},
{
"epoch": 0.43253968253968256,
"grad_norm": 2.4933829307556152,
"learning_rate": 9.27432255989112e-06,
"loss": 0.2422,
"step": 109
},
{
"epoch": 0.4365079365079365,
"grad_norm": 2.3846466541290283,
"learning_rate": 9.257179953763846e-06,
"loss": 0.2475,
"step": 110
},
{
"epoch": 0.44047619047619047,
"grad_norm": 2.2483131885528564,
"learning_rate": 9.239853454975548e-06,
"loss": 0.279,
"step": 111
},
{
"epoch": 0.4444444444444444,
"grad_norm": 2.259126901626587,
"learning_rate": 9.222343811959694e-06,
"loss": 0.2473,
"step": 112
},
{
"epoch": 0.44841269841269843,
"grad_norm": 2.121239185333252,
"learning_rate": 9.204651781060832e-06,
"loss": 0.2334,
"step": 113
},
{
"epoch": 0.4523809523809524,
"grad_norm": 1.9731487035751343,
"learning_rate": 9.186778126501916e-06,
"loss": 0.2397,
"step": 114
},
{
"epoch": 0.45634920634920634,
"grad_norm": 1.9351516962051392,
"learning_rate": 9.168723620351298e-06,
"loss": 0.2164,
"step": 115
},
{
"epoch": 0.4603174603174603,
"grad_norm": 1.8169538974761963,
"learning_rate": 9.150489042489368e-06,
"loss": 0.2263,
"step": 116
},
{
"epoch": 0.4642857142857143,
"grad_norm": 1.7139441967010498,
"learning_rate": 9.13207518057488e-06,
"loss": 0.2154,
"step": 117
},
{
"epoch": 0.46825396825396826,
"grad_norm": 1.6937518119812012,
"learning_rate": 9.113482830010918e-06,
"loss": 0.2203,
"step": 118
},
{
"epoch": 0.4722222222222222,
"grad_norm": 1.705714225769043,
"learning_rate": 9.094712793910541e-06,
"loss": 0.2388,
"step": 119
},
{
"epoch": 0.47619047619047616,
"grad_norm": 1.521938443183899,
"learning_rate": 9.075765883062093e-06,
"loss": 0.2304,
"step": 120
},
{
"epoch": 0.4801587301587302,
"grad_norm": 1.5313669443130493,
"learning_rate": 9.056642915894182e-06,
"loss": 0.2502,
"step": 121
},
{
"epoch": 0.48412698412698413,
"grad_norm": 1.5910950899124146,
"learning_rate": 9.037344718440321e-06,
"loss": 0.2571,
"step": 122
},
{
"epoch": 0.4880952380952381,
"grad_norm": 1.4521363973617554,
"learning_rate": 9.017872124303255e-06,
"loss": 0.2127,
"step": 123
},
{
"epoch": 0.49206349206349204,
"grad_norm": 1.4084386825561523,
"learning_rate": 8.99822597461894e-06,
"loss": 0.1819,
"step": 124
},
{
"epoch": 0.49603174603174605,
"grad_norm": 1.143968939781189,
"learning_rate": 8.978407118020226e-06,
"loss": 0.1803,
"step": 125
},
{
"epoch": 0.5,
"grad_norm": 1.7648898363113403,
"learning_rate": 8.958416410600188e-06,
"loss": 0.3738,
"step": 126
},
{
"epoch": 0.503968253968254,
"grad_norm": 1.6367061138153076,
"learning_rate": 8.938254715875152e-06,
"loss": 0.4007,
"step": 127
},
{
"epoch": 0.5079365079365079,
"grad_norm": 1.444301962852478,
"learning_rate": 8.917922904747385e-06,
"loss": 0.2889,
"step": 128
},
{
"epoch": 0.5119047619047619,
"grad_norm": 1.352399468421936,
"learning_rate": 8.897421855467491e-06,
"loss": 0.2692,
"step": 129
},
{
"epoch": 0.5158730158730159,
"grad_norm": 1.3992383480072021,
"learning_rate": 8.876752453596462e-06,
"loss": 0.2664,
"step": 130
},
{
"epoch": 0.5198412698412699,
"grad_norm": 1.4774725437164307,
"learning_rate": 8.85591559196743e-06,
"loss": 0.2673,
"step": 131
},
{
"epoch": 0.5238095238095238,
"grad_norm": 1.3351188898086548,
"learning_rate": 8.834912170647102e-06,
"loss": 0.2466,
"step": 132
},
{
"epoch": 0.5277777777777778,
"grad_norm": 1.1079375743865967,
"learning_rate": 8.813743096896872e-06,
"loss": 0.2255,
"step": 133
},
{
"epoch": 0.5317460317460317,
"grad_norm": 1.1617735624313354,
"learning_rate": 8.792409285133644e-06,
"loss": 0.2117,
"step": 134
},
{
"epoch": 0.5357142857142857,
"grad_norm": 1.2714942693710327,
"learning_rate": 8.770911656890325e-06,
"loss": 0.2108,
"step": 135
},
{
"epoch": 0.5396825396825397,
"grad_norm": 1.4111723899841309,
"learning_rate": 8.749251140776016e-06,
"loss": 0.2385,
"step": 136
},
{
"epoch": 0.5436507936507936,
"grad_norm": 1.192548394203186,
"learning_rate": 8.727428672435911e-06,
"loss": 0.2159,
"step": 137
},
{
"epoch": 0.5476190476190477,
"grad_norm": 1.2401196956634521,
"learning_rate": 8.705445194510868e-06,
"loss": 0.1971,
"step": 138
},
{
"epoch": 0.5515873015873016,
"grad_norm": 1.2631113529205322,
"learning_rate": 8.6833016565967e-06,
"loss": 0.2578,
"step": 139
},
{
"epoch": 0.5555555555555556,
"grad_norm": 1.2014082670211792,
"learning_rate": 8.660999015203152e-06,
"loss": 0.2443,
"step": 140
},
{
"epoch": 0.5595238095238095,
"grad_norm": 1.2346869707107544,
"learning_rate": 8.638538233712581e-06,
"loss": 0.2019,
"step": 141
},
{
"epoch": 0.5634920634920635,
"grad_norm": 1.2435442209243774,
"learning_rate": 8.615920282338355e-06,
"loss": 0.2015,
"step": 142
},
{
"epoch": 0.5674603174603174,
"grad_norm": 1.073967456817627,
"learning_rate": 8.593146138082925e-06,
"loss": 0.1844,
"step": 143
},
{
"epoch": 0.5714285714285714,
"grad_norm": 1.468886137008667,
"learning_rate": 8.570216784695637e-06,
"loss": 0.2024,
"step": 144
},
{
"epoch": 0.5753968253968254,
"grad_norm": 1.246264934539795,
"learning_rate": 8.54713321263023e-06,
"loss": 0.208,
"step": 145
},
{
"epoch": 0.5793650793650794,
"grad_norm": 1.1872477531433105,
"learning_rate": 8.52389641900206e-06,
"loss": 0.2103,
"step": 146
},
{
"epoch": 0.5833333333333334,
"grad_norm": 1.1925252676010132,
"learning_rate": 8.50050740754502e-06,
"loss": 0.2102,
"step": 147
},
{
"epoch": 0.5873015873015873,
"grad_norm": 1.3745256662368774,
"learning_rate": 8.476967188568187e-06,
"loss": 0.2278,
"step": 148
},
{
"epoch": 0.5912698412698413,
"grad_norm": 1.1875501871109009,
"learning_rate": 8.453276778912186e-06,
"loss": 0.1906,
"step": 149
},
{
"epoch": 0.5952380952380952,
"grad_norm": 1.1388797760009766,
"learning_rate": 8.429437201905254e-06,
"loss": 0.162,
"step": 150
},
{
"epoch": 0.5992063492063492,
"grad_norm": 1.775430679321289,
"learning_rate": 8.405449487319049e-06,
"loss": 0.3291,
"step": 151
},
{
"epoch": 0.6031746031746031,
"grad_norm": 1.6980030536651611,
"learning_rate": 8.38131467132416e-06,
"loss": 0.2782,
"step": 152
},
{
"epoch": 0.6071428571428571,
"grad_norm": 1.6253050565719604,
"learning_rate": 8.357033796445356e-06,
"loss": 0.3221,
"step": 153
},
{
"epoch": 0.6111111111111112,
"grad_norm": 1.5341050624847412,
"learning_rate": 8.332607911516545e-06,
"loss": 0.253,
"step": 154
},
{
"epoch": 0.6150793650793651,
"grad_norm": 1.4683576822280884,
"learning_rate": 8.308038071635475e-06,
"loss": 0.2119,
"step": 155
},
{
"epoch": 0.6190476190476191,
"grad_norm": 1.49379563331604,
"learning_rate": 8.283325338118154e-06,
"loss": 0.2651,
"step": 156
},
{
"epoch": 0.623015873015873,
"grad_norm": 1.5718587636947632,
"learning_rate": 8.258470778453005e-06,
"loss": 0.2494,
"step": 157
},
{
"epoch": 0.626984126984127,
"grad_norm": 1.4596575498580933,
"learning_rate": 8.233475466254766e-06,
"loss": 0.2334,
"step": 158
},
{
"epoch": 0.6309523809523809,
"grad_norm": 1.405685305595398,
"learning_rate": 8.208340481218094e-06,
"loss": 0.2396,
"step": 159
},
{
"epoch": 0.6349206349206349,
"grad_norm": 1.3298189640045166,
"learning_rate": 8.183066909070946e-06,
"loss": 0.1985,
"step": 160
},
{
"epoch": 0.6388888888888888,
"grad_norm": 1.3141931295394897,
"learning_rate": 8.15765584152767e-06,
"loss": 0.1953,
"step": 161
},
{
"epoch": 0.6428571428571429,
"grad_norm": 1.4742801189422607,
"learning_rate": 8.132108376241849e-06,
"loss": 0.2168,
"step": 162
},
{
"epoch": 0.6468253968253969,
"grad_norm": 1.3559314012527466,
"learning_rate": 8.106425616758886e-06,
"loss": 0.2008,
"step": 163
},
{
"epoch": 0.6507936507936508,
"grad_norm": 1.388225793838501,
"learning_rate": 8.08060867246834e-06,
"loss": 0.1964,
"step": 164
},
{
"epoch": 0.6547619047619048,
"grad_norm": 1.452958345413208,
"learning_rate": 8.054658658555998e-06,
"loss": 0.1794,
"step": 165
},
{
"epoch": 0.6587301587301587,
"grad_norm": 1.5204930305480957,
"learning_rate": 8.028576695955711e-06,
"loss": 0.1941,
"step": 166
},
{
"epoch": 0.6626984126984127,
"grad_norm": 1.3194262981414795,
"learning_rate": 8.002363911300966e-06,
"loss": 0.1862,
"step": 167
},
{
"epoch": 0.6666666666666666,
"grad_norm": 1.3155059814453125,
"learning_rate": 7.976021436876232e-06,
"loss": 0.1805,
"step": 168
},
{
"epoch": 0.6706349206349206,
"grad_norm": 1.5197312831878662,
"learning_rate": 7.949550410568033e-06,
"loss": 0.1926,
"step": 169
},
{
"epoch": 0.6746031746031746,
"grad_norm": 1.3390307426452637,
"learning_rate": 7.92295197581581e-06,
"loss": 0.1739,
"step": 170
},
{
"epoch": 0.6785714285714286,
"grad_norm": 1.3677082061767578,
"learning_rate": 7.89622728156253e-06,
"loss": 0.1755,
"step": 171
},
{
"epoch": 0.6825396825396826,
"grad_norm": 1.37470281124115,
"learning_rate": 7.869377482205042e-06,
"loss": 0.168,
"step": 172
},
{
"epoch": 0.6865079365079365,
"grad_norm": 1.4983570575714111,
"learning_rate": 7.842403737544226e-06,
"loss": 0.1807,
"step": 173
},
{
"epoch": 0.6904761904761905,
"grad_norm": 1.2006269693374634,
"learning_rate": 7.815307212734888e-06,
"loss": 0.1373,
"step": 174
},
{
"epoch": 0.6944444444444444,
"grad_norm": 1.5238745212554932,
"learning_rate": 7.788089078235432e-06,
"loss": 0.1855,
"step": 175
},
{
"epoch": 0.6984126984126984,
"grad_norm": 1.7328555583953857,
"learning_rate": 7.7607505097573e-06,
"loss": 0.2623,
"step": 176
},
{
"epoch": 0.7023809523809523,
"grad_norm": 1.7647768259048462,
"learning_rate": 7.733292688214182e-06,
"loss": 0.2974,
"step": 177
},
{
"epoch": 0.7063492063492064,
"grad_norm": 1.4417369365692139,
"learning_rate": 7.705716799671019e-06,
"loss": 0.2656,
"step": 178
},
{
"epoch": 0.7103174603174603,
"grad_norm": 1.3809551000595093,
"learning_rate": 7.678024035292757e-06,
"loss": 0.1914,
"step": 179
},
{
"epoch": 0.7142857142857143,
"grad_norm": 1.4950612783432007,
"learning_rate": 7.650215591292888e-06,
"loss": 0.2296,
"step": 180
},
{
"epoch": 0.7182539682539683,
"grad_norm": 1.3621734380722046,
"learning_rate": 7.622292668881805e-06,
"loss": 0.2122,
"step": 181
},
{
"epoch": 0.7222222222222222,
"grad_norm": 1.4526594877243042,
"learning_rate": 7.594256474214883e-06,
"loss": 0.1913,
"step": 182
},
{
"epoch": 0.7261904761904762,
"grad_norm": 1.220705509185791,
"learning_rate": 7.566108218340399e-06,
"loss": 0.1684,
"step": 183
},
{
"epoch": 0.7301587301587301,
"grad_norm": 1.299600601196289,
"learning_rate": 7.537849117147212e-06,
"loss": 0.1776,
"step": 184
},
{
"epoch": 0.7341269841269841,
"grad_norm": 1.3342280387878418,
"learning_rate": 7.509480391312243e-06,
"loss": 0.1812,
"step": 185
},
{
"epoch": 0.7380952380952381,
"grad_norm": 1.298844337463379,
"learning_rate": 7.481003266247745e-06,
"loss": 0.1756,
"step": 186
},
{
"epoch": 0.7420634920634921,
"grad_norm": 1.3185352087020874,
"learning_rate": 7.452418972048372e-06,
"loss": 0.1669,
"step": 187
},
{
"epoch": 0.746031746031746,
"grad_norm": 1.390432357788086,
"learning_rate": 7.4237287434380485e-06,
"loss": 0.1866,
"step": 188
},
{
"epoch": 0.75,
"grad_norm": 1.325130581855774,
"learning_rate": 7.394933819716625e-06,
"loss": 0.1567,
"step": 189
},
{
"epoch": 0.753968253968254,
"grad_norm": 1.3810374736785889,
"learning_rate": 7.366035444706346e-06,
"loss": 0.2268,
"step": 190
},
{
"epoch": 0.7579365079365079,
"grad_norm": 1.3227951526641846,
"learning_rate": 7.337034866698138e-06,
"loss": 0.1604,
"step": 191
},
{
"epoch": 0.7619047619047619,
"grad_norm": 1.3274725675582886,
"learning_rate": 7.307933338397667e-06,
"loss": 0.1752,
"step": 192
},
{
"epoch": 0.7658730158730159,
"grad_norm": 1.2616709470748901,
"learning_rate": 7.278732116871239e-06,
"loss": 0.1672,
"step": 193
},
{
"epoch": 0.7698412698412699,
"grad_norm": 1.355488896369934,
"learning_rate": 7.249432463491498e-06,
"loss": 0.162,
"step": 194
},
{
"epoch": 0.7738095238095238,
"grad_norm": 1.260221004486084,
"learning_rate": 7.220035643882938e-06,
"loss": 0.1778,
"step": 195
},
{
"epoch": 0.7777777777777778,
"grad_norm": 1.3204224109649658,
"learning_rate": 7.190542927867234e-06,
"loss": 0.1792,
"step": 196
},
{
"epoch": 0.7817460317460317,
"grad_norm": 1.0790374279022217,
"learning_rate": 7.160955589408395e-06,
"loss": 0.145,
"step": 197
},
{
"epoch": 0.7857142857142857,
"grad_norm": 1.1473934650421143,
"learning_rate": 7.131274906557725e-06,
"loss": 0.1408,
"step": 198
},
{
"epoch": 0.7896825396825397,
"grad_norm": 1.106053352355957,
"learning_rate": 7.101502161398626e-06,
"loss": 0.145,
"step": 199
},
{
"epoch": 0.7936507936507936,
"grad_norm": 1.3413796424865723,
"learning_rate": 7.0716386399912075e-06,
"loss": 0.1348,
"step": 200
},
{
"epoch": 0.7976190476190477,
"grad_norm": 1.3990546464920044,
"learning_rate": 7.041685632316748e-06,
"loss": 0.2515,
"step": 201
},
{
"epoch": 0.8015873015873016,
"grad_norm": 1.3854154348373413,
"learning_rate": 7.0116444322219575e-06,
"loss": 0.229,
"step": 202
},
{
"epoch": 0.8055555555555556,
"grad_norm": 1.3249114751815796,
"learning_rate": 6.981516337363099e-06,
"loss": 0.2038,
"step": 203
},
{
"epoch": 0.8095238095238095,
"grad_norm": 1.279220461845398,
"learning_rate": 6.95130264914993e-06,
"loss": 0.177,
"step": 204
},
{
"epoch": 0.8134920634920635,
"grad_norm": 1.2281877994537354,
"learning_rate": 6.9210046726894885e-06,
"loss": 0.212,
"step": 205
},
{
"epoch": 0.8174603174603174,
"grad_norm": 1.0341265201568604,
"learning_rate": 6.890623716729724e-06,
"loss": 0.166,
"step": 206
},
{
"epoch": 0.8214285714285714,
"grad_norm": 1.0794569253921509,
"learning_rate": 6.860161093602949e-06,
"loss": 0.1619,
"step": 207
},
{
"epoch": 0.8253968253968254,
"grad_norm": 1.0289993286132812,
"learning_rate": 6.829618119169169e-06,
"loss": 0.1648,
"step": 208
},
{
"epoch": 0.8293650793650794,
"grad_norm": 1.1661471128463745,
"learning_rate": 6.798996112759233e-06,
"loss": 0.1707,
"step": 209
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.998158872127533,
"learning_rate": 6.768296397117848e-06,
"loss": 0.1506,
"step": 210
},
{
"epoch": 0.8373015873015873,
"grad_norm": 1.1756287813186646,
"learning_rate": 6.737520298346438e-06,
"loss": 0.165,
"step": 211
},
{
"epoch": 0.8412698412698413,
"grad_norm": 1.271363377571106,
"learning_rate": 6.706669145845863e-06,
"loss": 0.1943,
"step": 212
},
{
"epoch": 0.8452380952380952,
"grad_norm": 1.1092045307159424,
"learning_rate": 6.6757442722590015e-06,
"loss": 0.1655,
"step": 213
},
{
"epoch": 0.8492063492063492,
"grad_norm": 1.0941603183746338,
"learning_rate": 6.6447470134131685e-06,
"loss": 0.1406,
"step": 214
},
{
"epoch": 0.8531746031746031,
"grad_norm": 1.034894585609436,
"learning_rate": 6.613678708262439e-06,
"loss": 0.1429,
"step": 215
},
{
"epoch": 0.8571428571428571,
"grad_norm": 1.0365397930145264,
"learning_rate": 6.5825406988297815e-06,
"loss": 0.1401,
"step": 216
},
{
"epoch": 0.8611111111111112,
"grad_norm": 0.977988064289093,
"learning_rate": 6.551334330149114e-06,
"loss": 0.1281,
"step": 217
},
{
"epoch": 0.8650793650793651,
"grad_norm": 1.0525593757629395,
"learning_rate": 6.520060950207186e-06,
"loss": 0.141,
"step": 218
},
{
"epoch": 0.8690476190476191,
"grad_norm": 1.0969583988189697,
"learning_rate": 6.488721909885359e-06,
"loss": 0.1504,
"step": 219
},
{
"epoch": 0.873015873015873,
"grad_norm": 1.094177484512329,
"learning_rate": 6.457318562901257e-06,
"loss": 0.1458,
"step": 220
},
{
"epoch": 0.876984126984127,
"grad_norm": 1.2514578104019165,
"learning_rate": 6.425852265750282e-06,
"loss": 0.1701,
"step": 221
},
{
"epoch": 0.8809523809523809,
"grad_norm": 0.9712778329849243,
"learning_rate": 6.394324377647028e-06,
"loss": 0.1305,
"step": 222
},
{
"epoch": 0.8849206349206349,
"grad_norm": 1.1288741827011108,
"learning_rate": 6.362736260466561e-06,
"loss": 0.1666,
"step": 223
},
{
"epoch": 0.8888888888888888,
"grad_norm": 1.0433393716812134,
"learning_rate": 6.331089278685599e-06,
"loss": 0.1607,
"step": 224
},
{
"epoch": 0.8928571428571429,
"grad_norm": 0.9745596647262573,
"learning_rate": 6.299384799323568e-06,
"loss": 0.1291,
"step": 225
},
{
"epoch": 0.8968253968253969,
"grad_norm": 1.2714403867721558,
"learning_rate": 6.267624191883551e-06,
"loss": 0.2571,
"step": 226
},
{
"epoch": 0.9007936507936508,
"grad_norm": 1.2934309244155884,
"learning_rate": 6.235808828293135e-06,
"loss": 0.2654,
"step": 227
},
{
"epoch": 0.9047619047619048,
"grad_norm": 1.3339515924453735,
"learning_rate": 6.203940082845144e-06,
"loss": 0.2246,
"step": 228
},
{
"epoch": 0.9087301587301587,
"grad_norm": 1.2278695106506348,
"learning_rate": 6.172019332138285e-06,
"loss": 0.1706,
"step": 229
},
{
"epoch": 0.9126984126984127,
"grad_norm": 1.012449026107788,
"learning_rate": 6.140047955017672e-06,
"loss": 0.1502,
"step": 230
},
{
"epoch": 0.9166666666666666,
"grad_norm": 1.0517585277557373,
"learning_rate": 6.108027332515276e-06,
"loss": 0.1572,
"step": 231
},
{
"epoch": 0.9206349206349206,
"grad_norm": 1.264817714691162,
"learning_rate": 6.075958847790262e-06,
"loss": 0.2073,
"step": 232
},
{
"epoch": 0.9246031746031746,
"grad_norm": 1.0246200561523438,
"learning_rate": 6.043843886069251e-06,
"loss": 0.1412,
"step": 233
},
{
"epoch": 0.9285714285714286,
"grad_norm": 1.0280753374099731,
"learning_rate": 6.011683834586474e-06,
"loss": 0.1544,
"step": 234
},
{
"epoch": 0.9325396825396826,
"grad_norm": 0.9970900416374207,
"learning_rate": 5.979480082523858e-06,
"loss": 0.1428,
"step": 235
},
{
"epoch": 0.9365079365079365,
"grad_norm": 1.0177240371704102,
"learning_rate": 5.947234020951015e-06,
"loss": 0.1457,
"step": 236
},
{
"epoch": 0.9404761904761905,
"grad_norm": 0.8569260239601135,
"learning_rate": 5.914947042765149e-06,
"loss": 0.1157,
"step": 237
},
{
"epoch": 0.9444444444444444,
"grad_norm": 0.9930649995803833,
"learning_rate": 5.882620542630901e-06,
"loss": 0.1498,
"step": 238
},
{
"epoch": 0.9484126984126984,
"grad_norm": 1.0476784706115723,
"learning_rate": 5.850255916920093e-06,
"loss": 0.1493,
"step": 239
},
{
"epoch": 0.9523809523809523,
"grad_norm": 1.173066258430481,
"learning_rate": 5.817854563651415e-06,
"loss": 0.1602,
"step": 240
},
{
"epoch": 0.9563492063492064,
"grad_norm": 1.2765151262283325,
"learning_rate": 5.785417882430035e-06,
"loss": 0.1681,
"step": 241
},
{
"epoch": 0.9603174603174603,
"grad_norm": 1.1309196949005127,
"learning_rate": 5.752947274387147e-06,
"loss": 0.158,
"step": 242
},
{
"epoch": 0.9642857142857143,
"grad_norm": 1.1185320615768433,
"learning_rate": 5.720444142119445e-06,
"loss": 0.1463,
"step": 243
},
{
"epoch": 0.9682539682539683,
"grad_norm": 0.9425800442695618,
"learning_rate": 5.687909889628529e-06,
"loss": 0.1474,
"step": 244
},
{
"epoch": 0.9722222222222222,
"grad_norm": 1.032344937324524,
"learning_rate": 5.6553459222602714e-06,
"loss": 0.143,
"step": 245
},
{
"epoch": 0.9761904761904762,
"grad_norm": 0.937047004699707,
"learning_rate": 5.622753646644102e-06,
"loss": 0.1235,
"step": 246
},
{
"epoch": 0.9801587301587301,
"grad_norm": 0.9617612957954407,
"learning_rate": 5.59013447063225e-06,
"loss": 0.1412,
"step": 247
},
{
"epoch": 0.9841269841269841,
"grad_norm": 0.9130176305770874,
"learning_rate": 5.557489803238934e-06,
"loss": 0.1181,
"step": 248
},
{
"epoch": 0.9880952380952381,
"grad_norm": 0.8578863143920898,
"learning_rate": 5.524821054579491e-06,
"loss": 0.1301,
"step": 249
},
{
"epoch": 0.9920634920634921,
"grad_norm": 0.863932728767395,
"learning_rate": 5.492129635809473e-06,
"loss": 0.111,
"step": 250
},
{
"epoch": 0.996031746031746,
"grad_norm": 1.1702795028686523,
"learning_rate": 5.459416959063688e-06,
"loss": 0.1775,
"step": 251
},
{
"epoch": 1.0,
"grad_norm": 1.1006587743759155,
"learning_rate": 5.426684437395196e-06,
"loss": 0.2497,
"step": 252
},
{
"epoch": 1.003968253968254,
"grad_norm": 1.1237730979919434,
"learning_rate": 5.393933484714284e-06,
"loss": 0.2066,
"step": 253
},
{
"epoch": 1.007936507936508,
"grad_norm": 1.2142930030822754,
"learning_rate": 5.361165515727374e-06,
"loss": 0.2216,
"step": 254
},
{
"epoch": 1.0119047619047619,
"grad_norm": 1.1424957513809204,
"learning_rate": 5.328381945875933e-06,
"loss": 0.1806,
"step": 255
},
{
"epoch": 1.0158730158730158,
"grad_norm": 0.9865773320198059,
"learning_rate": 5.295584191275308e-06,
"loss": 0.1464,
"step": 256
},
{
"epoch": 1.0198412698412698,
"grad_norm": 0.9836987257003784,
"learning_rate": 5.26277366865358e-06,
"loss": 0.1193,
"step": 257
},
{
"epoch": 1.0238095238095237,
"grad_norm": 0.930524468421936,
"learning_rate": 5.229951795290353e-06,
"loss": 0.144,
"step": 258
},
{
"epoch": 1.0277777777777777,
"grad_norm": 1.0486645698547363,
"learning_rate": 5.197119988955534e-06,
"loss": 0.136,
"step": 259
},
{
"epoch": 1.0317460317460316,
"grad_norm": 0.8506413698196411,
"learning_rate": 5.164279667848094e-06,
"loss": 0.1267,
"step": 260
},
{
"epoch": 1.0357142857142858,
"grad_norm": 0.8831179738044739,
"learning_rate": 5.131432250534809e-06,
"loss": 0.1247,
"step": 261
},
{
"epoch": 1.0396825396825398,
"grad_norm": 0.95271897315979,
"learning_rate": 5.0985791558889785e-06,
"loss": 0.1363,
"step": 262
},
{
"epoch": 1.0436507936507937,
"grad_norm": 1.0076528787612915,
"learning_rate": 5.065721803029146e-06,
"loss": 0.1184,
"step": 263
},
{
"epoch": 1.0476190476190477,
"grad_norm": 0.9802455306053162,
"learning_rate": 5.032861611257783e-06,
"loss": 0.1362,
"step": 264
},
{
"epoch": 1.0515873015873016,
"grad_norm": 0.9055737853050232,
"learning_rate": 5e-06,
"loss": 0.1361,
"step": 265
},
{
"epoch": 1.0555555555555556,
"grad_norm": 0.9631403088569641,
"learning_rate": 4.967138388742218e-06,
"loss": 0.1182,
"step": 266
},
{
"epoch": 1.0595238095238095,
"grad_norm": 0.9077386856079102,
"learning_rate": 4.934278196970857e-06,
"loss": 0.0995,
"step": 267
},
{
"epoch": 1.0634920634920635,
"grad_norm": 0.9423613548278809,
"learning_rate": 4.9014208441110215e-06,
"loss": 0.105,
"step": 268
},
{
"epoch": 1.0674603174603174,
"grad_norm": 0.8865099549293518,
"learning_rate": 4.868567749465192e-06,
"loss": 0.1063,
"step": 269
},
{
"epoch": 1.0714285714285714,
"grad_norm": 0.8101741075515747,
"learning_rate": 4.835720332151907e-06,
"loss": 0.1019,
"step": 270
},
{
"epoch": 1.0753968253968254,
"grad_norm": 0.8372849822044373,
"learning_rate": 4.802880011044467e-06,
"loss": 0.0992,
"step": 271
},
{
"epoch": 1.0793650793650793,
"grad_norm": 1.1325459480285645,
"learning_rate": 4.770048204709648e-06,
"loss": 0.115,
"step": 272
},
{
"epoch": 1.0833333333333333,
"grad_norm": 0.8753513097763062,
"learning_rate": 4.73722633134642e-06,
"loss": 0.099,
"step": 273
},
{
"epoch": 1.0873015873015872,
"grad_norm": 0.9488489627838135,
"learning_rate": 4.7044158087246926e-06,
"loss": 0.1128,
"step": 274
},
{
"epoch": 1.0912698412698412,
"grad_norm": 1.040241003036499,
"learning_rate": 4.67161805412407e-06,
"loss": 0.1107,
"step": 275
},
{
"epoch": 1.0952380952380953,
"grad_norm": 0.8925001621246338,
"learning_rate": 4.6388344842726266e-06,
"loss": 0.0947,
"step": 276
},
{
"epoch": 1.0992063492063493,
"grad_norm": 0.8387725949287415,
"learning_rate": 4.606066515285719e-06,
"loss": 0.0936,
"step": 277
},
{
"epoch": 1.1031746031746033,
"grad_norm": 1.1508458852767944,
"learning_rate": 4.573315562604804e-06,
"loss": 0.158,
"step": 278
},
{
"epoch": 1.1071428571428572,
"grad_norm": 1.1206495761871338,
"learning_rate": 4.540583040936313e-06,
"loss": 0.1592,
"step": 279
},
{
"epoch": 1.1111111111111112,
"grad_norm": 1.0632399320602417,
"learning_rate": 4.5078703641905275e-06,
"loss": 0.144,
"step": 280
},
{
"epoch": 1.1150793650793651,
"grad_norm": 1.1150530576705933,
"learning_rate": 4.4751789454205105e-06,
"loss": 0.1513,
"step": 281
},
{
"epoch": 1.119047619047619,
"grad_norm": 0.8953108191490173,
"learning_rate": 4.442510196761068e-06,
"loss": 0.1266,
"step": 282
},
{
"epoch": 1.123015873015873,
"grad_norm": 0.9804735779762268,
"learning_rate": 4.409865529367751e-06,
"loss": 0.1315,
"step": 283
},
{
"epoch": 1.126984126984127,
"grad_norm": 1.0136425495147705,
"learning_rate": 4.377246353355899e-06,
"loss": 0.1335,
"step": 284
},
{
"epoch": 1.130952380952381,
"grad_norm": 1.0654220581054688,
"learning_rate": 4.34465407773973e-06,
"loss": 0.1424,
"step": 285
},
{
"epoch": 1.1349206349206349,
"grad_norm": 0.8981422185897827,
"learning_rate": 4.312090110371473e-06,
"loss": 0.1119,
"step": 286
},
{
"epoch": 1.1388888888888888,
"grad_norm": 0.7868387699127197,
"learning_rate": 4.279555857880558e-06,
"loss": 0.1033,
"step": 287
},
{
"epoch": 1.1428571428571428,
"grad_norm": 0.9177349805831909,
"learning_rate": 4.247052725612853e-06,
"loss": 0.1063,
"step": 288
},
{
"epoch": 1.1468253968253967,
"grad_norm": 0.9164119362831116,
"learning_rate": 4.214582117569966e-06,
"loss": 0.1252,
"step": 289
},
{
"epoch": 1.1507936507936507,
"grad_norm": 0.9463087320327759,
"learning_rate": 4.182145436348587e-06,
"loss": 0.1175,
"step": 290
},
{
"epoch": 1.1547619047619047,
"grad_norm": 0.8865921497344971,
"learning_rate": 4.1497440830799084e-06,
"loss": 0.0941,
"step": 291
},
{
"epoch": 1.1587301587301586,
"grad_norm": 1.1493929624557495,
"learning_rate": 4.1173794573691e-06,
"loss": 0.1351,
"step": 292
},
{
"epoch": 1.1626984126984128,
"grad_norm": 0.9012559652328491,
"learning_rate": 4.0850529572348505e-06,
"loss": 0.1036,
"step": 293
},
{
"epoch": 1.1666666666666667,
"grad_norm": 0.9597694277763367,
"learning_rate": 4.052765979048986e-06,
"loss": 0.0981,
"step": 294
},
{
"epoch": 1.1706349206349207,
"grad_norm": 0.9837350249290466,
"learning_rate": 4.0205199174761435e-06,
"loss": 0.1107,
"step": 295
},
{
"epoch": 1.1746031746031746,
"grad_norm": 0.8875190615653992,
"learning_rate": 3.988316165413528e-06,
"loss": 0.1017,
"step": 296
},
{
"epoch": 1.1785714285714286,
"grad_norm": 0.9989994168281555,
"learning_rate": 3.956156113930752e-06,
"loss": 0.1173,
"step": 297
},
{
"epoch": 1.1825396825396826,
"grad_norm": 0.9141669869422913,
"learning_rate": 3.924041152209739e-06,
"loss": 0.1015,
"step": 298
},
{
"epoch": 1.1865079365079365,
"grad_norm": 0.9252498149871826,
"learning_rate": 3.891972667484726e-06,
"loss": 0.1139,
"step": 299
},
{
"epoch": 1.1904761904761905,
"grad_norm": 0.9392082691192627,
"learning_rate": 3.859952044982329e-06,
"loss": 0.0936,
"step": 300
},
{
"epoch": 1.1944444444444444,
"grad_norm": 0.7951925992965698,
"learning_rate": 3.827980667861716e-06,
"loss": 0.0768,
"step": 301
},
{
"epoch": 1.1984126984126984,
"grad_norm": 0.7657447457313538,
"learning_rate": 3.7960599171548572e-06,
"loss": 0.0738,
"step": 302
},
{
"epoch": 1.2023809523809523,
"grad_norm": 1.2552180290222168,
"learning_rate": 3.764191171706867e-06,
"loss": 0.1651,
"step": 303
},
{
"epoch": 1.2063492063492063,
"grad_norm": 1.119985580444336,
"learning_rate": 3.732375808116451e-06,
"loss": 0.1637,
"step": 304
},
{
"epoch": 1.2103174603174602,
"grad_norm": 1.2039250135421753,
"learning_rate": 3.7006152006764336e-06,
"loss": 0.1737,
"step": 305
},
{
"epoch": 1.2142857142857142,
"grad_norm": 1.1200075149536133,
"learning_rate": 3.6689107213144025e-06,
"loss": 0.137,
"step": 306
},
{
"epoch": 1.2182539682539684,
"grad_norm": 0.9427171945571899,
"learning_rate": 3.6372637395334416e-06,
"loss": 0.1207,
"step": 307
},
{
"epoch": 1.2222222222222223,
"grad_norm": 0.9506517052650452,
"learning_rate": 3.6056756223529734e-06,
"loss": 0.132,
"step": 308
},
{
"epoch": 1.2261904761904763,
"grad_norm": 1.032836675643921,
"learning_rate": 3.574147734249719e-06,
"loss": 0.1136,
"step": 309
},
{
"epoch": 1.2301587301587302,
"grad_norm": 0.9958872199058533,
"learning_rate": 3.542681437098745e-06,
"loss": 0.1159,
"step": 310
},
{
"epoch": 1.2341269841269842,
"grad_norm": 0.9964204430580139,
"learning_rate": 3.5112780901146426e-06,
"loss": 0.1305,
"step": 311
},
{
"epoch": 1.2380952380952381,
"grad_norm": 0.918283224105835,
"learning_rate": 3.479939049792817e-06,
"loss": 0.1265,
"step": 312
},
{
"epoch": 1.242063492063492,
"grad_norm": 0.9123579859733582,
"learning_rate": 3.448665669850888e-06,
"loss": 0.114,
"step": 313
},
{
"epoch": 1.246031746031746,
"grad_norm": 0.8953122496604919,
"learning_rate": 3.4174593011702197e-06,
"loss": 0.1205,
"step": 314
},
{
"epoch": 1.25,
"grad_norm": 0.9500359296798706,
"learning_rate": 3.386321291737563e-06,
"loss": 0.1221,
"step": 315
},
{
"epoch": 1.253968253968254,
"grad_norm": 0.84867924451828,
"learning_rate": 3.3552529865868323e-06,
"loss": 0.1186,
"step": 316
},
{
"epoch": 1.257936507936508,
"grad_norm": 0.9007363319396973,
"learning_rate": 3.3242557277410015e-06,
"loss": 0.0972,
"step": 317
},
{
"epoch": 1.2619047619047619,
"grad_norm": 0.835606038570404,
"learning_rate": 3.2933308541541365e-06,
"loss": 0.1063,
"step": 318
},
{
"epoch": 1.2658730158730158,
"grad_norm": 0.9490189552307129,
"learning_rate": 3.2624797016535626e-06,
"loss": 0.12,
"step": 319
},
{
"epoch": 1.2698412698412698,
"grad_norm": 0.9719697833061218,
"learning_rate": 3.2317036028821523e-06,
"loss": 0.1262,
"step": 320
},
{
"epoch": 1.2738095238095237,
"grad_norm": 0.9631256461143494,
"learning_rate": 3.201003887240768e-06,
"loss": 0.119,
"step": 321
},
{
"epoch": 1.2777777777777777,
"grad_norm": 0.786014199256897,
"learning_rate": 3.1703818808308327e-06,
"loss": 0.1074,
"step": 322
},
{
"epoch": 1.2817460317460316,
"grad_norm": 1.019045352935791,
"learning_rate": 3.1398389063970512e-06,
"loss": 0.1376,
"step": 323
},
{
"epoch": 1.2857142857142856,
"grad_norm": 0.9258590340614319,
"learning_rate": 3.1093762832702775e-06,
"loss": 0.0912,
"step": 324
},
{
"epoch": 1.2896825396825398,
"grad_norm": 0.8826021552085876,
"learning_rate": 3.0789953273105123e-06,
"loss": 0.0927,
"step": 325
},
{
"epoch": 1.2936507936507937,
"grad_norm": 0.8910852670669556,
"learning_rate": 3.048697350850073e-06,
"loss": 0.1084,
"step": 326
},
{
"epoch": 1.2976190476190477,
"grad_norm": 0.9962997436523438,
"learning_rate": 3.0184836626369034e-06,
"loss": 0.1091,
"step": 327
},
{
"epoch": 1.3015873015873016,
"grad_norm": 1.1140745878219604,
"learning_rate": 2.988355567778043e-06,
"loss": 0.1818,
"step": 328
},
{
"epoch": 1.3055555555555556,
"grad_norm": 1.0912885665893555,
"learning_rate": 2.9583143676832526e-06,
"loss": 0.1765,
"step": 329
},
{
"epoch": 1.3095238095238095,
"grad_norm": 1.108447551727295,
"learning_rate": 2.9283613600087933e-06,
"loss": 0.1636,
"step": 330
},
{
"epoch": 1.3134920634920635,
"grad_norm": 1.046484351158142,
"learning_rate": 2.8984978386013767e-06,
"loss": 0.1465,
"step": 331
},
{
"epoch": 1.3174603174603174,
"grad_norm": 0.944888710975647,
"learning_rate": 2.8687250934422774e-06,
"loss": 0.1142,
"step": 332
},
{
"epoch": 1.3214285714285714,
"grad_norm": 0.9226338267326355,
"learning_rate": 2.839044410591606e-06,
"loss": 0.1095,
"step": 333
},
{
"epoch": 1.3253968253968254,
"grad_norm": 0.922489583492279,
"learning_rate": 2.809457072132766e-06,
"loss": 0.1145,
"step": 334
},
{
"epoch": 1.3293650793650793,
"grad_norm": 0.9101172685623169,
"learning_rate": 2.779964356117063e-06,
"loss": 0.1062,
"step": 335
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.953028678894043,
"learning_rate": 2.750567536508504e-06,
"loss": 0.1061,
"step": 336
},
{
"epoch": 1.3373015873015874,
"grad_norm": 0.980619490146637,
"learning_rate": 2.7212678831287627e-06,
"loss": 0.1212,
"step": 337
},
{
"epoch": 1.3412698412698414,
"grad_norm": 0.9573182463645935,
"learning_rate": 2.692066661602333e-06,
"loss": 0.116,
"step": 338
},
{
"epoch": 1.3452380952380953,
"grad_norm": 0.9239686131477356,
"learning_rate": 2.662965133301862e-06,
"loss": 0.1197,
"step": 339
},
{
"epoch": 1.3492063492063493,
"grad_norm": 0.8666567802429199,
"learning_rate": 2.633964555293654e-06,
"loss": 0.1095,
"step": 340
},
{
"epoch": 1.3531746031746033,
"grad_norm": 0.8586628437042236,
"learning_rate": 2.605066180283378e-06,
"loss": 0.1045,
"step": 341
},
{
"epoch": 1.3571428571428572,
"grad_norm": 0.8865628242492676,
"learning_rate": 2.576271256561953e-06,
"loss": 0.1057,
"step": 342
},
{
"epoch": 1.3611111111111112,
"grad_norm": 0.9049112796783447,
"learning_rate": 2.5475810279516287e-06,
"loss": 0.117,
"step": 343
},
{
"epoch": 1.3650793650793651,
"grad_norm": 0.8492547869682312,
"learning_rate": 2.5189967337522574e-06,
"loss": 0.0904,
"step": 344
},
{
"epoch": 1.369047619047619,
"grad_norm": 0.8558246493339539,
"learning_rate": 2.49051960868776e-06,
"loss": 0.1047,
"step": 345
},
{
"epoch": 1.373015873015873,
"grad_norm": 0.8904597163200378,
"learning_rate": 2.46215088285279e-06,
"loss": 0.1112,
"step": 346
},
{
"epoch": 1.376984126984127,
"grad_norm": 0.7617968916893005,
"learning_rate": 2.433891781659603e-06,
"loss": 0.0944,
"step": 347
},
{
"epoch": 1.380952380952381,
"grad_norm": 0.8228709697723389,
"learning_rate": 2.4057435257851173e-06,
"loss": 0.0975,
"step": 348
},
{
"epoch": 1.3849206349206349,
"grad_norm": 0.8048103451728821,
"learning_rate": 2.377707331118196e-06,
"loss": 0.0862,
"step": 349
},
{
"epoch": 1.3888888888888888,
"grad_norm": 0.9368796348571777,
"learning_rate": 2.349784408707112e-06,
"loss": 0.0841,
"step": 350
},
{
"epoch": 1.3928571428571428,
"grad_norm": 0.7275282144546509,
"learning_rate": 2.3219759647072467e-06,
"loss": 0.0725,
"step": 351
},
{
"epoch": 1.3968253968253967,
"grad_norm": 0.8309359550476074,
"learning_rate": 2.2942832003289823e-06,
"loss": 0.079,
"step": 352
},
{
"epoch": 1.4007936507936507,
"grad_norm": 1.1370983123779297,
"learning_rate": 2.2667073117858185e-06,
"loss": 0.1893,
"step": 353
},
{
"epoch": 1.4047619047619047,
"grad_norm": 1.1589655876159668,
"learning_rate": 2.2392494902427027e-06,
"loss": 0.1664,
"step": 354
},
{
"epoch": 1.4087301587301586,
"grad_norm": 1.1058390140533447,
"learning_rate": 2.2119109217645697e-06,
"loss": 0.1582,
"step": 355
},
{
"epoch": 1.4126984126984126,
"grad_norm": 1.0430413484573364,
"learning_rate": 2.1846927872651135e-06,
"loss": 0.1333,
"step": 356
},
{
"epoch": 1.4166666666666667,
"grad_norm": 1.0792104005813599,
"learning_rate": 2.1575962624557754e-06,
"loss": 0.1334,
"step": 357
},
{
"epoch": 1.4206349206349207,
"grad_norm": 0.8786894679069519,
"learning_rate": 2.1306225177949584e-06,
"loss": 0.1282,
"step": 358
},
{
"epoch": 1.4246031746031746,
"grad_norm": 0.9336158633232117,
"learning_rate": 2.1037727184374705e-06,
"loss": 0.1313,
"step": 359
},
{
"epoch": 1.4285714285714286,
"grad_norm": 0.8968210220336914,
"learning_rate": 2.07704802418419e-06,
"loss": 0.1116,
"step": 360
},
{
"epoch": 1.4325396825396826,
"grad_norm": 1.072925090789795,
"learning_rate": 2.050449589431969e-06,
"loss": 0.1267,
"step": 361
},
{
"epoch": 1.4365079365079365,
"grad_norm": 0.9037191867828369,
"learning_rate": 2.023978563123771e-06,
"loss": 0.1176,
"step": 362
},
{
"epoch": 1.4404761904761905,
"grad_norm": 0.791119396686554,
"learning_rate": 1.997636088699035e-06,
"loss": 0.1121,
"step": 363
},
{
"epoch": 1.4444444444444444,
"grad_norm": 0.8636593818664551,
"learning_rate": 1.9714233040442915e-06,
"loss": 0.1124,
"step": 364
},
{
"epoch": 1.4484126984126984,
"grad_norm": 0.8613457083702087,
"learning_rate": 1.9453413414440043e-06,
"loss": 0.0996,
"step": 365
},
{
"epoch": 1.4523809523809523,
"grad_norm": 0.8780179619789124,
"learning_rate": 1.919391327531663e-06,
"loss": 0.0992,
"step": 366
},
{
"epoch": 1.4563492063492063,
"grad_norm": 0.9861639738082886,
"learning_rate": 1.8935743832411163e-06,
"loss": 0.1003,
"step": 367
},
{
"epoch": 1.4603174603174602,
"grad_norm": 0.8164539933204651,
"learning_rate": 1.8678916237581524e-06,
"loss": 0.0945,
"step": 368
},
{
"epoch": 1.4642857142857144,
"grad_norm": 0.9254552125930786,
"learning_rate": 1.8423441584723312e-06,
"loss": 0.0922,
"step": 369
},
{
"epoch": 1.4682539682539684,
"grad_norm": 0.9601562023162842,
"learning_rate": 1.816933090929055e-06,
"loss": 0.1142,
"step": 370
},
{
"epoch": 1.4722222222222223,
"grad_norm": 1.0283352136611938,
"learning_rate": 1.791659518781908e-06,
"loss": 0.1133,
"step": 371
},
{
"epoch": 1.4761904761904763,
"grad_norm": 0.9066034555435181,
"learning_rate": 1.7665245337452368e-06,
"loss": 0.1064,
"step": 372
},
{
"epoch": 1.4801587301587302,
"grad_norm": 0.8869957327842712,
"learning_rate": 1.7415292215469948e-06,
"loss": 0.103,
"step": 373
},
{
"epoch": 1.4841269841269842,
"grad_norm": 0.9691331386566162,
"learning_rate": 1.716674661881848e-06,
"loss": 0.1055,
"step": 374
},
{
"epoch": 1.4880952380952381,
"grad_norm": 0.9107316136360168,
"learning_rate": 1.6919619283645262e-06,
"loss": 0.0947,
"step": 375
},
{
"epoch": 1.492063492063492,
"grad_norm": 0.8729169368743896,
"learning_rate": 1.667392088483456e-06,
"loss": 0.091,
"step": 376
},
{
"epoch": 1.496031746031746,
"grad_norm": 0.8033152222633362,
"learning_rate": 1.6429662035546451e-06,
"loss": 0.0757,
"step": 377
},
{
"epoch": 1.5,
"grad_norm": 1.2133307456970215,
"learning_rate": 1.6186853286758397e-06,
"loss": 0.1571,
"step": 378
},
{
"epoch": 1.503968253968254,
"grad_norm": 1.0417262315750122,
"learning_rate": 1.5945505126809524e-06,
"loss": 0.1756,
"step": 379
},
{
"epoch": 1.507936507936508,
"grad_norm": 0.9838248491287231,
"learning_rate": 1.570562798094747e-06,
"loss": 0.141,
"step": 380
},
{
"epoch": 1.5119047619047619,
"grad_norm": 0.8970847129821777,
"learning_rate": 1.5467232210878153e-06,
"loss": 0.1054,
"step": 381
},
{
"epoch": 1.5158730158730158,
"grad_norm": 0.8847159147262573,
"learning_rate": 1.5230328114318127e-06,
"loss": 0.1137,
"step": 382
},
{
"epoch": 1.5198412698412698,
"grad_norm": 0.9331052899360657,
"learning_rate": 1.4994925924549797e-06,
"loss": 0.1072,
"step": 383
},
{
"epoch": 1.5238095238095237,
"grad_norm": 0.9449731707572937,
"learning_rate": 1.4761035809979395e-06,
"loss": 0.1221,
"step": 384
},
{
"epoch": 1.5277777777777777,
"grad_norm": 0.9156303405761719,
"learning_rate": 1.452866787369771e-06,
"loss": 0.1073,
"step": 385
},
{
"epoch": 1.5317460317460316,
"grad_norm": 0.7507684826850891,
"learning_rate": 1.4297832153043657e-06,
"loss": 0.0828,
"step": 386
},
{
"epoch": 1.5357142857142856,
"grad_norm": 0.8908015489578247,
"learning_rate": 1.4068538619170763e-06,
"loss": 0.1083,
"step": 387
},
{
"epoch": 1.5396825396825395,
"grad_norm": 0.9291139841079712,
"learning_rate": 1.3840797176616467e-06,
"loss": 0.11,
"step": 388
},
{
"epoch": 1.5436507936507935,
"grad_norm": 0.8711128234863281,
"learning_rate": 1.3614617662874197e-06,
"loss": 0.1018,
"step": 389
},
{
"epoch": 1.5476190476190477,
"grad_norm": 0.8244633674621582,
"learning_rate": 1.3390009847968505e-06,
"loss": 0.0898,
"step": 390
},
{
"epoch": 1.5515873015873016,
"grad_norm": 0.9386397004127502,
"learning_rate": 1.316698343403302e-06,
"loss": 0.1202,
"step": 391
},
{
"epoch": 1.5555555555555556,
"grad_norm": 0.9043394923210144,
"learning_rate": 1.2945548054891322e-06,
"loss": 0.1103,
"step": 392
},
{
"epoch": 1.5595238095238095,
"grad_norm": 0.9116290807723999,
"learning_rate": 1.27257132756409e-06,
"loss": 0.1053,
"step": 393
},
{
"epoch": 1.5634920634920635,
"grad_norm": 0.8217816352844238,
"learning_rate": 1.2507488592239848e-06,
"loss": 0.1032,
"step": 394
},
{
"epoch": 1.5674603174603174,
"grad_norm": 0.9115804433822632,
"learning_rate": 1.2290883431096778e-06,
"loss": 0.1085,
"step": 395
},
{
"epoch": 1.5714285714285714,
"grad_norm": 0.8311465382575989,
"learning_rate": 1.2075907148663579e-06,
"loss": 0.0799,
"step": 396
},
{
"epoch": 1.5753968253968254,
"grad_norm": 0.8430618643760681,
"learning_rate": 1.186256903103129e-06,
"loss": 0.0881,
"step": 397
},
{
"epoch": 1.5793650793650795,
"grad_norm": 0.8516837358474731,
"learning_rate": 1.1650878293528994e-06,
"loss": 0.0877,
"step": 398
},
{
"epoch": 1.5833333333333335,
"grad_norm": 0.7429177165031433,
"learning_rate": 1.1440844080325703e-06,
"loss": 0.0762,
"step": 399
},
{
"epoch": 1.5873015873015874,
"grad_norm": 1.0127265453338623,
"learning_rate": 1.1232475464035386e-06,
"loss": 0.0839,
"step": 400
},
{
"epoch": 1.5912698412698414,
"grad_norm": 0.7672459483146667,
"learning_rate": 1.10257814453251e-06,
"loss": 0.0884,
"step": 401
},
{
"epoch": 1.5952380952380953,
"grad_norm": 0.8456502556800842,
"learning_rate": 1.0820770952526155e-06,
"loss": 0.0851,
"step": 402
},
{
"epoch": 1.5992063492063493,
"grad_norm": 1.1015710830688477,
"learning_rate": 1.0617452841248494e-06,
"loss": 0.1795,
"step": 403
},
{
"epoch": 1.6031746031746033,
"grad_norm": 1.2175111770629883,
"learning_rate": 1.0415835893998116e-06,
"loss": 0.1939,
"step": 404
},
{
"epoch": 1.6071428571428572,
"grad_norm": 1.1487531661987305,
"learning_rate": 1.0215928819797744e-06,
"loss": 0.15,
"step": 405
},
{
"epoch": 1.6111111111111112,
"grad_norm": 0.9514380097389221,
"learning_rate": 1.0017740253810608e-06,
"loss": 0.1154,
"step": 406
},
{
"epoch": 1.6150793650793651,
"grad_norm": 0.9456499218940735,
"learning_rate": 9.821278756967467e-07,
"loss": 0.1306,
"step": 407
},
{
"epoch": 1.619047619047619,
"grad_norm": 0.9060970544815063,
"learning_rate": 9.62655281559679e-07,
"loss": 0.1218,
"step": 408
},
{
"epoch": 1.623015873015873,
"grad_norm": 0.9095820784568787,
"learning_rate": 9.433570841058187e-07,
"loss": 0.113,
"step": 409
},
{
"epoch": 1.626984126984127,
"grad_norm": 0.9578836560249329,
"learning_rate": 9.242341169379077e-07,
"loss": 0.1292,
"step": 410
},
{
"epoch": 1.630952380952381,
"grad_norm": 0.956503689289093,
"learning_rate": 9.052872060894613e-07,
"loss": 0.128,
"step": 411
},
{
"epoch": 1.6349206349206349,
"grad_norm": 0.9277134537696838,
"learning_rate": 8.865171699890835e-07,
"loss": 0.1172,
"step": 412
},
{
"epoch": 1.6388888888888888,
"grad_norm": 0.90636146068573,
"learning_rate": 8.679248194251211e-07,
"loss": 0.1091,
"step": 413
},
{
"epoch": 1.6428571428571428,
"grad_norm": 0.9789716005325317,
"learning_rate": 8.495109575106331e-07,
"loss": 0.1107,
"step": 414
},
{
"epoch": 1.6468253968253967,
"grad_norm": 0.7881717085838318,
"learning_rate": 8.312763796487038e-07,
"loss": 0.11,
"step": 415
},
{
"epoch": 1.6507936507936507,
"grad_norm": 0.8127108216285706,
"learning_rate": 8.132218734980852e-07,
"loss": 0.095,
"step": 416
},
{
"epoch": 1.6547619047619047,
"grad_norm": 0.751276969909668,
"learning_rate": 7.953482189391687e-07,
"loss": 0.0859,
"step": 417
},
{
"epoch": 1.6587301587301586,
"grad_norm": 0.8322916030883789,
"learning_rate": 7.776561880403072e-07,
"loss": 0.0917,
"step": 418
},
{
"epoch": 1.6626984126984126,
"grad_norm": 0.7949820160865784,
"learning_rate": 7.601465450244528e-07,
"loss": 0.0799,
"step": 419
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.8409376740455627,
"learning_rate": 7.42820046236154e-07,
"loss": 0.1,
"step": 420
},
{
"epoch": 1.6706349206349205,
"grad_norm": 0.9577579498291016,
"learning_rate": 7.256774401088817e-07,
"loss": 0.1028,
"step": 421
},
{
"epoch": 1.6746031746031746,
"grad_norm": 0.8095564842224121,
"learning_rate": 7.087194671326986e-07,
"loss": 0.0959,
"step": 422
},
{
"epoch": 1.6785714285714286,
"grad_norm": 0.9481910467147827,
"learning_rate": 6.91946859822279e-07,
"loss": 0.1187,
"step": 423
},
{
"epoch": 1.6825396825396826,
"grad_norm": 0.9387165904045105,
"learning_rate": 6.753603426852589e-07,
"loss": 0.1025,
"step": 424
},
{
"epoch": 1.6865079365079365,
"grad_norm": 0.8484113216400146,
"learning_rate": 6.589606321909464e-07,
"loss": 0.085,
"step": 425
},
{
"epoch": 1.6904761904761905,
"grad_norm": 0.9083000421524048,
"learning_rate": 6.427484367393699e-07,
"loss": 0.0969,
"step": 426
},
{
"epoch": 1.6944444444444444,
"grad_norm": 0.8922500610351562,
"learning_rate": 6.267244566306801e-07,
"loss": 0.0928,
"step": 427
},
{
"epoch": 1.6984126984126984,
"grad_norm": 1.056872844696045,
"learning_rate": 6.108893840348995e-07,
"loss": 0.1882,
"step": 428
},
{
"epoch": 1.7023809523809523,
"grad_norm": 1.1732290983200073,
"learning_rate": 5.952439029620222e-07,
"loss": 0.1576,
"step": 429
},
{
"epoch": 1.7063492063492065,
"grad_norm": 1.025284767150879,
"learning_rate": 5.797886892324695e-07,
"loss": 0.1504,
"step": 430
},
{
"epoch": 1.7103174603174605,
"grad_norm": 0.8623697757720947,
"learning_rate": 5.645244104478947e-07,
"loss": 0.1138,
"step": 431
},
{
"epoch": 1.7142857142857144,
"grad_norm": 0.8793927431106567,
"learning_rate": 5.494517259623478e-07,
"loss": 0.1097,
"step": 432
},
{
"epoch": 1.7182539682539684,
"grad_norm": 0.876624345779419,
"learning_rate": 5.34571286853795e-07,
"loss": 0.1071,
"step": 433
},
{
"epoch": 1.7222222222222223,
"grad_norm": 0.8404582142829895,
"learning_rate": 5.198837358959901e-07,
"loss": 0.0927,
"step": 434
},
{
"epoch": 1.7261904761904763,
"grad_norm": 0.83519446849823,
"learning_rate": 5.05389707530714e-07,
"loss": 0.105,
"step": 435
},
{
"epoch": 1.7301587301587302,
"grad_norm": 0.9712224006652832,
"learning_rate": 4.91089827840367e-07,
"loss": 0.1206,
"step": 436
},
{
"epoch": 1.7341269841269842,
"grad_norm": 0.8240790367126465,
"learning_rate": 4.769847145209244e-07,
"loss": 0.1035,
"step": 437
},
{
"epoch": 1.7380952380952381,
"grad_norm": 0.722660481929779,
"learning_rate": 4.6307497685525894e-07,
"loss": 0.0835,
"step": 438
},
{
"epoch": 1.742063492063492,
"grad_norm": 0.8676173686981201,
"learning_rate": 4.4936121568681546e-07,
"loss": 0.1001,
"step": 439
},
{
"epoch": 1.746031746031746,
"grad_norm": 0.9424171447753906,
"learning_rate": 4.3584402339366174e-07,
"loss": 0.0978,
"step": 440
},
{
"epoch": 1.75,
"grad_norm": 0.8304186463356018,
"learning_rate": 4.225239838628981e-07,
"loss": 0.0875,
"step": 441
},
{
"epoch": 1.753968253968254,
"grad_norm": 0.9304208755493164,
"learning_rate": 4.0940167246543595e-07,
"loss": 0.112,
"step": 442
},
{
"epoch": 1.757936507936508,
"grad_norm": 0.8676320314407349,
"learning_rate": 3.964776560311484e-07,
"loss": 0.1042,
"step": 443
},
{
"epoch": 1.7619047619047619,
"grad_norm": 0.8660047650337219,
"learning_rate": 3.8375249282437743e-07,
"loss": 0.102,
"step": 444
},
{
"epoch": 1.7658730158730158,
"grad_norm": 0.8883973956108093,
"learning_rate": 3.71226732519826e-07,
"loss": 0.1001,
"step": 445
},
{
"epoch": 1.7698412698412698,
"grad_norm": 0.8700767755508423,
"learning_rate": 3.589009161788104e-07,
"loss": 0.1084,
"step": 446
},
{
"epoch": 1.7738095238095237,
"grad_norm": 0.8301666975021362,
"learning_rate": 3.4677557622589175e-07,
"loss": 0.0869,
"step": 447
},
{
"epoch": 1.7777777777777777,
"grad_norm": 0.7481169700622559,
"learning_rate": 3.3485123642587657e-07,
"loss": 0.0816,
"step": 448
},
{
"epoch": 1.7817460317460316,
"grad_norm": 0.8663469552993774,
"learning_rate": 3.2312841186118937e-07,
"loss": 0.092,
"step": 449
},
{
"epoch": 1.7857142857142856,
"grad_norm": 0.8146552443504333,
"learning_rate": 3.116076089096265e-07,
"loss": 0.0848,
"step": 450
},
{
"epoch": 1.7896825396825395,
"grad_norm": 0.9174822568893433,
"learning_rate": 3.0028932522248256e-07,
"loss": 0.0894,
"step": 451
},
{
"epoch": 1.7936507936507935,
"grad_norm": 0.8255031108856201,
"learning_rate": 2.8917404970305096e-07,
"loss": 0.0736,
"step": 452
},
{
"epoch": 1.7976190476190477,
"grad_norm": 1.1608794927597046,
"learning_rate": 2.782622624855097e-07,
"loss": 0.1821,
"step": 453
},
{
"epoch": 1.8015873015873016,
"grad_norm": 1.2320818901062012,
"learning_rate": 2.6755443491417786e-07,
"loss": 0.1867,
"step": 454
},
{
"epoch": 1.8055555555555556,
"grad_norm": 0.911833643913269,
"learning_rate": 2.570510295231571e-07,
"loss": 0.1347,
"step": 455
},
{
"epoch": 1.8095238095238095,
"grad_norm": 0.9706260561943054,
"learning_rate": 2.467525000163523e-07,
"loss": 0.1302,
"step": 456
},
{
"epoch": 1.8134920634920635,
"grad_norm": 0.9432376623153687,
"learning_rate": 2.36659291247871e-07,
"loss": 0.1141,
"step": 457
},
{
"epoch": 1.8174603174603174,
"grad_norm": 0.9442083239555359,
"learning_rate": 2.2677183920281342e-07,
"loss": 0.1321,
"step": 458
},
{
"epoch": 1.8214285714285714,
"grad_norm": 0.9746333360671997,
"learning_rate": 2.1709057097843266e-07,
"loss": 0.1271,
"step": 459
},
{
"epoch": 1.8253968253968254,
"grad_norm": 0.7531862258911133,
"learning_rate": 2.0761590476568893e-07,
"loss": 0.0944,
"step": 460
},
{
"epoch": 1.8293650793650795,
"grad_norm": 0.8026720881462097,
"learning_rate": 1.9834824983118673e-07,
"loss": 0.0977,
"step": 461
},
{
"epoch": 1.8333333333333335,
"grad_norm": 0.927125096321106,
"learning_rate": 1.892880064994934e-07,
"loss": 0.1169,
"step": 462
},
{
"epoch": 1.8373015873015874,
"grad_norm": 0.8158795237541199,
"learning_rate": 1.8043556613585143e-07,
"loss": 0.096,
"step": 463
},
{
"epoch": 1.8412698412698414,
"grad_norm": 0.8917427659034729,
"learning_rate": 1.7179131112926628e-07,
"loss": 0.1023,
"step": 464
},
{
"epoch": 1.8452380952380953,
"grad_norm": 0.9492359161376953,
"learning_rate": 1.6335561487599406e-07,
"loss": 0.1125,
"step": 465
},
{
"epoch": 1.8492063492063493,
"grad_norm": 0.8159032464027405,
"learning_rate": 1.551288417634106e-07,
"loss": 0.0826,
"step": 466
},
{
"epoch": 1.8531746031746033,
"grad_norm": 0.9698415398597717,
"learning_rate": 1.471113471542712e-07,
"loss": 0.1191,
"step": 467
},
{
"epoch": 1.8571428571428572,
"grad_norm": 0.8887535929679871,
"learning_rate": 1.3930347737136195e-07,
"loss": 0.0943,
"step": 468
},
{
"epoch": 1.8611111111111112,
"grad_norm": 0.8581032156944275,
"learning_rate": 1.3170556968253756e-07,
"loss": 0.0943,
"step": 469
},
{
"epoch": 1.8650793650793651,
"grad_norm": 0.7531769275665283,
"learning_rate": 1.2431795228615372e-07,
"loss": 0.0797,
"step": 470
},
{
"epoch": 1.869047619047619,
"grad_norm": 0.8543042540550232,
"learning_rate": 1.1714094429689127e-07,
"loss": 0.0912,
"step": 471
},
{
"epoch": 1.873015873015873,
"grad_norm": 0.8516489863395691,
"learning_rate": 1.1017485573197151e-07,
"loss": 0.0919,
"step": 472
},
{
"epoch": 1.876984126984127,
"grad_norm": 0.9233003854751587,
"learning_rate": 1.0341998749776316e-07,
"loss": 0.0999,
"step": 473
},
{
"epoch": 1.880952380952381,
"grad_norm": 0.8185317516326904,
"learning_rate": 9.687663137678605e-08,
"loss": 0.0847,
"step": 474
},
{
"epoch": 1.8849206349206349,
"grad_norm": 0.8176765441894531,
"learning_rate": 9.054507001510727e-08,
"loss": 0.0863,
"step": 475
},
{
"epoch": 1.8888888888888888,
"grad_norm": 0.8034681081771851,
"learning_rate": 8.442557691013042e-08,
"loss": 0.0776,
"step": 476
},
{
"epoch": 1.8928571428571428,
"grad_norm": 0.9487660527229309,
"learning_rate": 7.851841639878399e-08,
"loss": 0.0918,
"step": 477
},
{
"epoch": 1.8968253968253967,
"grad_norm": 1.2069216966629028,
"learning_rate": 7.282384364610207e-08,
"loss": 0.199,
"step": 478
},
{
"epoch": 1.9007936507936507,
"grad_norm": 1.0059871673583984,
"learning_rate": 6.734210463420099e-08,
"loss": 0.147,
"step": 479
},
{
"epoch": 1.9047619047619047,
"grad_norm": 1.0063914060592651,
"learning_rate": 6.207343615165562e-08,
"loss": 0.1137,
"step": 480
},
{
"epoch": 1.9087301587301586,
"grad_norm": 0.8561007976531982,
"learning_rate": 5.701806578327029e-08,
"loss": 0.1036,
"step": 481
},
{
"epoch": 1.9126984126984126,
"grad_norm": 0.7949555516242981,
"learning_rate": 5.21762119002478e-08,
"loss": 0.0926,
"step": 482
},
{
"epoch": 1.9166666666666665,
"grad_norm": 0.8420392274856567,
"learning_rate": 4.7548083650759134e-08,
"loss": 0.0997,
"step": 483
},
{
"epoch": 1.9206349206349205,
"grad_norm": 0.8691868782043457,
"learning_rate": 4.31338809509052e-08,
"loss": 0.1006,
"step": 484
},
{
"epoch": 1.9246031746031746,
"grad_norm": 0.9431844353675842,
"learning_rate": 3.8933794476083143e-08,
"loss": 0.1206,
"step": 485
},
{
"epoch": 1.9285714285714286,
"grad_norm": 0.8648985624313354,
"learning_rate": 3.494800565275125e-08,
"loss": 0.0997,
"step": 486
},
{
"epoch": 1.9325396825396826,
"grad_norm": 0.8204662799835205,
"learning_rate": 3.1176686650589147e-08,
"loss": 0.0937,
"step": 487
},
{
"epoch": 1.9365079365079365,
"grad_norm": 0.8854976296424866,
"learning_rate": 2.7620000375064848e-08,
"loss": 0.102,
"step": 488
},
{
"epoch": 1.9404761904761905,
"grad_norm": 0.9284180998802185,
"learning_rate": 2.4278100460393138e-08,
"loss": 0.1102,
"step": 489
},
{
"epoch": 1.9444444444444444,
"grad_norm": 0.7911974787712097,
"learning_rate": 2.115113126290258e-08,
"loss": 0.1053,
"step": 490
},
{
"epoch": 1.9484126984126984,
"grad_norm": 0.6992430090904236,
"learning_rate": 1.8239227854799368e-08,
"loss": 0.0747,
"step": 491
},
{
"epoch": 1.9523809523809523,
"grad_norm": 0.9787194728851318,
"learning_rate": 1.554251601833201e-08,
"loss": 0.1128,
"step": 492
},
{
"epoch": 1.9563492063492065,
"grad_norm": 0.7918096780776978,
"learning_rate": 1.3061112240357887e-08,
"loss": 0.0862,
"step": 493
},
{
"epoch": 1.9603174603174605,
"grad_norm": 0.8645848035812378,
"learning_rate": 1.0795123707312283e-08,
"loss": 0.0821,
"step": 494
},
{
"epoch": 1.9642857142857144,
"grad_norm": 0.8377680778503418,
"learning_rate": 8.744648300578196e-09,
"loss": 0.0917,
"step": 495
},
{
"epoch": 1.9682539682539684,
"grad_norm": 0.7998278141021729,
"learning_rate": 6.9097745922580564e-09,
"loss": 0.0819,
"step": 496
},
{
"epoch": 1.9722222222222223,
"grad_norm": 0.827128529548645,
"learning_rate": 5.2905818413478975e-09,
"loss": 0.0828,
"step": 497
},
{
"epoch": 1.9761904761904763,
"grad_norm": 0.8182300329208374,
"learning_rate": 3.887139990313427e-09,
"loss": 0.078,
"step": 498
},
{
"epoch": 1.9801587301587302,
"grad_norm": 0.8508040308952332,
"learning_rate": 2.699509662069666e-09,
"loss": 0.0844,
"step": 499
},
{
"epoch": 1.9841269841269842,
"grad_norm": 0.8178525567054749,
"learning_rate": 1.7277421573608234e-09,
"loss": 0.0891,
"step": 500
},
{
"epoch": 1.9880952380952381,
"grad_norm": 0.8030312657356262,
"learning_rate": 9.71879452545399e-10,
"loss": 0.0793,
"step": 501
},
{
"epoch": 1.992063492063492,
"grad_norm": 0.9470478892326355,
"learning_rate": 4.3195419778319095e-10,
"loss": 0.1077,
"step": 502
},
{
"epoch": 1.996031746031746,
"grad_norm": 0.8880834579467773,
"learning_rate": 1.0798971562364647e-10,
"loss": 0.1103,
"step": 503
},
{
"epoch": 2.0,
"grad_norm": 1.0543088912963867,
"learning_rate": 0.0,
"loss": 0.1278,
"step": 504
},
{
"epoch": 2.0,
"step": 504,
"total_flos": 51259110653952.0,
"train_loss": 0.29744523939573103,
"train_runtime": 2850.7356,
"train_samples_per_second": 2.822,
"train_steps_per_second": 0.177
}
],
"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": 51259110653952.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}