Files
titulm-gemma-2-2b-v1.1/trainer_state.json
ModelHub XC 613be5fb35 初始化项目,由ModelHub XC社区提供模型
Model: hishab/titulm-gemma-2-2b-v1.1
Source: Original Platform
2026-05-28 23:11:03 +08:00

39432 lines
961 KiB
JSON

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9999555733262251,
"eval_steps": 500,
"global_step": 5627,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00017770669509973788,
"grad_norm": 612.283849944697,
"learning_rate": 7.017543859649123e-07,
"loss": 7.1545,
"step": 1
},
{
"epoch": 0.00035541339019947576,
"grad_norm": 585.8741226277815,
"learning_rate": 1.4035087719298246e-06,
"loss": 7.0683,
"step": 2
},
{
"epoch": 0.0005331200852992136,
"grad_norm": 584.2671161224255,
"learning_rate": 2.105263157894737e-06,
"loss": 6.9978,
"step": 3
},
{
"epoch": 0.0007108267803989515,
"grad_norm": 327.94342764569393,
"learning_rate": 2.8070175438596493e-06,
"loss": 5.4451,
"step": 4
},
{
"epoch": 0.0008885334754986894,
"grad_norm": 99.63405273271715,
"learning_rate": 3.5087719298245615e-06,
"loss": 4.2204,
"step": 5
},
{
"epoch": 0.0010662401705984273,
"grad_norm": 73.84910206671451,
"learning_rate": 4.210526315789474e-06,
"loss": 3.8457,
"step": 6
},
{
"epoch": 0.0012439468656981652,
"grad_norm": 41.576742844309614,
"learning_rate": 4.912280701754386e-06,
"loss": 3.2549,
"step": 7
},
{
"epoch": 0.001421653560797903,
"grad_norm": 33.433894388086735,
"learning_rate": 5.6140350877192985e-06,
"loss": 3.2042,
"step": 8
},
{
"epoch": 0.001599360255897641,
"grad_norm": 15.565174503791875,
"learning_rate": 6.31578947368421e-06,
"loss": 2.6104,
"step": 9
},
{
"epoch": 0.0017770669509973788,
"grad_norm": 7.892138776470482,
"learning_rate": 7.017543859649123e-06,
"loss": 2.4822,
"step": 10
},
{
"epoch": 0.0019547736460971167,
"grad_norm": 4.8222549439054525,
"learning_rate": 7.719298245614036e-06,
"loss": 2.6837,
"step": 11
},
{
"epoch": 0.0021324803411968546,
"grad_norm": 9.085072969885518,
"learning_rate": 8.421052631578948e-06,
"loss": 2.4258,
"step": 12
},
{
"epoch": 0.0023101870362965925,
"grad_norm": 6.676185210675449,
"learning_rate": 9.12280701754386e-06,
"loss": 2.463,
"step": 13
},
{
"epoch": 0.0024878937313963304,
"grad_norm": 4.788682687261237,
"learning_rate": 9.824561403508772e-06,
"loss": 2.3072,
"step": 14
},
{
"epoch": 0.0026656004264960682,
"grad_norm": 3.1955970240838956,
"learning_rate": 1.0526315789473684e-05,
"loss": 2.4683,
"step": 15
},
{
"epoch": 0.002843307121595806,
"grad_norm": 2.7370205513723915,
"learning_rate": 1.1228070175438597e-05,
"loss": 2.4925,
"step": 16
},
{
"epoch": 0.003021013816695544,
"grad_norm": 2.542497166094289,
"learning_rate": 1.192982456140351e-05,
"loss": 2.2503,
"step": 17
},
{
"epoch": 0.003198720511795282,
"grad_norm": 2.562112158635685,
"learning_rate": 1.263157894736842e-05,
"loss": 2.2602,
"step": 18
},
{
"epoch": 0.0033764272068950198,
"grad_norm": 2.5938565195147048,
"learning_rate": 1.3333333333333333e-05,
"loss": 2.3477,
"step": 19
},
{
"epoch": 0.0035541339019947576,
"grad_norm": 2.5189647278076333,
"learning_rate": 1.4035087719298246e-05,
"loss": 2.3348,
"step": 20
},
{
"epoch": 0.0037318405970944955,
"grad_norm": 2.810141421773279,
"learning_rate": 1.4736842105263159e-05,
"loss": 2.1845,
"step": 21
},
{
"epoch": 0.003909547292194233,
"grad_norm": 2.1664535875204542,
"learning_rate": 1.543859649122807e-05,
"loss": 2.2383,
"step": 22
},
{
"epoch": 0.004087253987293971,
"grad_norm": 2.507979402576355,
"learning_rate": 1.6140350877192984e-05,
"loss": 2.2439,
"step": 23
},
{
"epoch": 0.004264960682393709,
"grad_norm": 2.614010607500675,
"learning_rate": 1.6842105263157896e-05,
"loss": 2.2131,
"step": 24
},
{
"epoch": 0.004442667377493447,
"grad_norm": 2.145926382510738,
"learning_rate": 1.754385964912281e-05,
"loss": 2.1535,
"step": 25
},
{
"epoch": 0.004620374072593185,
"grad_norm": 2.4569691774800555,
"learning_rate": 1.824561403508772e-05,
"loss": 2.1073,
"step": 26
},
{
"epoch": 0.004798080767692923,
"grad_norm": 2.419364996046518,
"learning_rate": 1.894736842105263e-05,
"loss": 2.0962,
"step": 27
},
{
"epoch": 0.004975787462792661,
"grad_norm": 2.073814752553096,
"learning_rate": 1.9649122807017544e-05,
"loss": 2.2948,
"step": 28
},
{
"epoch": 0.005153494157892399,
"grad_norm": 2.8355714823425573,
"learning_rate": 2.035087719298246e-05,
"loss": 2.1363,
"step": 29
},
{
"epoch": 0.0053312008529921365,
"grad_norm": 2.3021139996901994,
"learning_rate": 2.105263157894737e-05,
"loss": 2.1441,
"step": 30
},
{
"epoch": 0.005508907548091874,
"grad_norm": 2.3921665480709193,
"learning_rate": 2.1754385964912285e-05,
"loss": 1.9666,
"step": 31
},
{
"epoch": 0.005686614243191612,
"grad_norm": 2.62934709116376,
"learning_rate": 2.2456140350877194e-05,
"loss": 2.0112,
"step": 32
},
{
"epoch": 0.00586432093829135,
"grad_norm": 1.976992981651804,
"learning_rate": 2.3157894736842107e-05,
"loss": 2.0019,
"step": 33
},
{
"epoch": 0.006042027633391088,
"grad_norm": 2.154119569303339,
"learning_rate": 2.385964912280702e-05,
"loss": 2.007,
"step": 34
},
{
"epoch": 0.006219734328490826,
"grad_norm": 1.9982416850758185,
"learning_rate": 2.4561403508771932e-05,
"loss": 2.0439,
"step": 35
},
{
"epoch": 0.006397441023590564,
"grad_norm": 2.885321404653079,
"learning_rate": 2.526315789473684e-05,
"loss": 2.027,
"step": 36
},
{
"epoch": 0.006575147718690302,
"grad_norm": 1.8143642456636926,
"learning_rate": 2.5964912280701757e-05,
"loss": 2.006,
"step": 37
},
{
"epoch": 0.0067528544137900395,
"grad_norm": 2.3631243841229415,
"learning_rate": 2.6666666666666667e-05,
"loss": 1.9948,
"step": 38
},
{
"epoch": 0.006930561108889777,
"grad_norm": 2.5471042503188785,
"learning_rate": 2.7368421052631583e-05,
"loss": 2.0396,
"step": 39
},
{
"epoch": 0.007108267803989515,
"grad_norm": 2.1692192151870913,
"learning_rate": 2.8070175438596492e-05,
"loss": 1.968,
"step": 40
},
{
"epoch": 0.007285974499089253,
"grad_norm": 2.5071405305934786,
"learning_rate": 2.8771929824561408e-05,
"loss": 1.9611,
"step": 41
},
{
"epoch": 0.007463681194188991,
"grad_norm": 2.7104520924627504,
"learning_rate": 2.9473684210526317e-05,
"loss": 2.0122,
"step": 42
},
{
"epoch": 0.007641387889288729,
"grad_norm": 1.9274493926978122,
"learning_rate": 3.017543859649123e-05,
"loss": 2.0516,
"step": 43
},
{
"epoch": 0.007819094584388467,
"grad_norm": 2.433389962273218,
"learning_rate": 3.087719298245614e-05,
"loss": 1.8842,
"step": 44
},
{
"epoch": 0.007996801279488205,
"grad_norm": 1.9068072100178852,
"learning_rate": 3.157894736842106e-05,
"loss": 2.0397,
"step": 45
},
{
"epoch": 0.008174507974587943,
"grad_norm": 2.6624483126301315,
"learning_rate": 3.228070175438597e-05,
"loss": 2.007,
"step": 46
},
{
"epoch": 0.00835221466968768,
"grad_norm": 2.120544274974673,
"learning_rate": 3.298245614035088e-05,
"loss": 1.927,
"step": 47
},
{
"epoch": 0.008529921364787418,
"grad_norm": 2.6107441979047707,
"learning_rate": 3.368421052631579e-05,
"loss": 2.0038,
"step": 48
},
{
"epoch": 0.008707628059887156,
"grad_norm": 2.0242957807896325,
"learning_rate": 3.43859649122807e-05,
"loss": 1.8921,
"step": 49
},
{
"epoch": 0.008885334754986894,
"grad_norm": 2.6200138902118484,
"learning_rate": 3.508771929824562e-05,
"loss": 1.9108,
"step": 50
},
{
"epoch": 0.009063041450086632,
"grad_norm": 1.9236474346335766,
"learning_rate": 3.578947368421053e-05,
"loss": 1.9533,
"step": 51
},
{
"epoch": 0.00924074814518637,
"grad_norm": 1.9367370251150215,
"learning_rate": 3.649122807017544e-05,
"loss": 1.9123,
"step": 52
},
{
"epoch": 0.009418454840286108,
"grad_norm": 2.0823071422960435,
"learning_rate": 3.719298245614035e-05,
"loss": 1.9081,
"step": 53
},
{
"epoch": 0.009596161535385846,
"grad_norm": 2.2745764735878957,
"learning_rate": 3.789473684210526e-05,
"loss": 1.9663,
"step": 54
},
{
"epoch": 0.009773868230485584,
"grad_norm": 1.7316799153506028,
"learning_rate": 3.859649122807018e-05,
"loss": 1.8523,
"step": 55
},
{
"epoch": 0.009951574925585321,
"grad_norm": 1.9821925439845491,
"learning_rate": 3.929824561403509e-05,
"loss": 1.928,
"step": 56
},
{
"epoch": 0.01012928162068506,
"grad_norm": 2.395861938911586,
"learning_rate": 4e-05,
"loss": 1.8731,
"step": 57
},
{
"epoch": 0.010306988315784797,
"grad_norm": 1.856953270516331,
"learning_rate": 3.999999681881194e-05,
"loss": 1.8971,
"step": 58
},
{
"epoch": 0.010484695010884535,
"grad_norm": 2.3014641627472385,
"learning_rate": 3.9999987275248785e-05,
"loss": 1.9188,
"step": 59
},
{
"epoch": 0.010662401705984273,
"grad_norm": 2.2374354833389494,
"learning_rate": 3.999997136931355e-05,
"loss": 1.8498,
"step": 60
},
{
"epoch": 0.01084010840108401,
"grad_norm": 1.7376469900034501,
"learning_rate": 3.9999949101011305e-05,
"loss": 1.8885,
"step": 61
},
{
"epoch": 0.011017815096183749,
"grad_norm": 2.0312125568147246,
"learning_rate": 3.999992047034914e-05,
"loss": 1.8108,
"step": 62
},
{
"epoch": 0.011195521791283487,
"grad_norm": 1.4641327664772255,
"learning_rate": 3.9999885477336156e-05,
"loss": 1.882,
"step": 63
},
{
"epoch": 0.011373228486383224,
"grad_norm": 1.9141792528699655,
"learning_rate": 3.999984412198349e-05,
"loss": 1.7691,
"step": 64
},
{
"epoch": 0.011550935181482962,
"grad_norm": 2.0705569829409654,
"learning_rate": 3.9999796404304294e-05,
"loss": 1.8332,
"step": 65
},
{
"epoch": 0.0117286418765827,
"grad_norm": 1.6864132268773273,
"learning_rate": 3.999974232431375e-05,
"loss": 1.9104,
"step": 66
},
{
"epoch": 0.011906348571682438,
"grad_norm": 2.1225040926974135,
"learning_rate": 3.999968188202905e-05,
"loss": 1.8237,
"step": 67
},
{
"epoch": 0.012084055266782176,
"grad_norm": 1.5023718248739064,
"learning_rate": 3.999961507746944e-05,
"loss": 1.812,
"step": 68
},
{
"epoch": 0.012261761961881914,
"grad_norm": 2.2778850430851056,
"learning_rate": 3.999954191065617e-05,
"loss": 1.8748,
"step": 69
},
{
"epoch": 0.012439468656981652,
"grad_norm": 1.4153123369802831,
"learning_rate": 3.9999462381612505e-05,
"loss": 1.7363,
"step": 70
},
{
"epoch": 0.01261717535208139,
"grad_norm": 1.9949735683336705,
"learning_rate": 3.999937649036375e-05,
"loss": 1.7783,
"step": 71
},
{
"epoch": 0.012794882047181128,
"grad_norm": 1.6553996216299374,
"learning_rate": 3.999928423693723e-05,
"loss": 1.7838,
"step": 72
},
{
"epoch": 0.012972588742280865,
"grad_norm": 1.703880035270774,
"learning_rate": 3.999918562136229e-05,
"loss": 1.816,
"step": 73
},
{
"epoch": 0.013150295437380603,
"grad_norm": 1.5664697295932895,
"learning_rate": 3.999908064367029e-05,
"loss": 1.8286,
"step": 74
},
{
"epoch": 0.013328002132480341,
"grad_norm": 1.7789387986164191,
"learning_rate": 3.999896930389465e-05,
"loss": 1.8032,
"step": 75
},
{
"epoch": 0.013505708827580079,
"grad_norm": 1.6367240261479283,
"learning_rate": 3.9998851602070775e-05,
"loss": 1.8146,
"step": 76
},
{
"epoch": 0.013683415522679817,
"grad_norm": 1.745256779760865,
"learning_rate": 3.999872753823611e-05,
"loss": 1.825,
"step": 77
},
{
"epoch": 0.013861122217779555,
"grad_norm": 1.6615403039947148,
"learning_rate": 3.9998597112430124e-05,
"loss": 1.8412,
"step": 78
},
{
"epoch": 0.014038828912879293,
"grad_norm": 1.3569827083907198,
"learning_rate": 3.99984603246943e-05,
"loss": 1.7968,
"step": 79
},
{
"epoch": 0.01421653560797903,
"grad_norm": 1.4686741116039,
"learning_rate": 3.999831717507217e-05,
"loss": 1.7711,
"step": 80
},
{
"epoch": 0.014394242303078768,
"grad_norm": 1.5690804091870811,
"learning_rate": 3.999816766360925e-05,
"loss": 1.7927,
"step": 81
},
{
"epoch": 0.014571948998178506,
"grad_norm": 1.3583052979401566,
"learning_rate": 3.9998011790353117e-05,
"loss": 1.7488,
"step": 82
},
{
"epoch": 0.014749655693278244,
"grad_norm": 1.447772581411551,
"learning_rate": 3.9997849555353356e-05,
"loss": 1.8654,
"step": 83
},
{
"epoch": 0.014927362388377982,
"grad_norm": 1.475546059933149,
"learning_rate": 3.999768095866157e-05,
"loss": 1.7623,
"step": 84
},
{
"epoch": 0.01510506908347772,
"grad_norm": 1.495784429114487,
"learning_rate": 3.999750600033141e-05,
"loss": 1.7645,
"step": 85
},
{
"epoch": 0.015282775778577458,
"grad_norm": 1.3553651047108126,
"learning_rate": 3.9997324680418514e-05,
"loss": 1.7335,
"step": 86
},
{
"epoch": 0.015460482473677196,
"grad_norm": 1.387326095784698,
"learning_rate": 3.999713699898057e-05,
"loss": 1.7694,
"step": 87
},
{
"epoch": 0.015638189168776934,
"grad_norm": 1.544011162990764,
"learning_rate": 3.999694295607728e-05,
"loss": 1.7343,
"step": 88
},
{
"epoch": 0.01581589586387667,
"grad_norm": 1.518052328315435,
"learning_rate": 3.999674255177038e-05,
"loss": 1.8091,
"step": 89
},
{
"epoch": 0.01599360255897641,
"grad_norm": 1.33728239611186,
"learning_rate": 3.999653578612362e-05,
"loss": 1.7431,
"step": 90
},
{
"epoch": 0.016171309254076147,
"grad_norm": 1.369362178625371,
"learning_rate": 3.999632265920277e-05,
"loss": 1.7802,
"step": 91
},
{
"epoch": 0.016349015949175885,
"grad_norm": 1.5017477647037982,
"learning_rate": 3.999610317107564e-05,
"loss": 1.7826,
"step": 92
},
{
"epoch": 0.016526722644275623,
"grad_norm": 1.3483316612472427,
"learning_rate": 3.999587732181205e-05,
"loss": 1.7418,
"step": 93
},
{
"epoch": 0.01670442933937536,
"grad_norm": 1.3975311435775617,
"learning_rate": 3.999564511148384e-05,
"loss": 1.7564,
"step": 94
},
{
"epoch": 0.0168821360344751,
"grad_norm": 1.3880731779328874,
"learning_rate": 3.999540654016488e-05,
"loss": 1.7633,
"step": 95
},
{
"epoch": 0.017059842729574837,
"grad_norm": 1.2985714651622657,
"learning_rate": 3.999516160793107e-05,
"loss": 1.7022,
"step": 96
},
{
"epoch": 0.017237549424674575,
"grad_norm": 1.8138128580573312,
"learning_rate": 3.9994910314860334e-05,
"loss": 1.7699,
"step": 97
},
{
"epoch": 0.017415256119774312,
"grad_norm": 1.219661038803346,
"learning_rate": 3.99946526610326e-05,
"loss": 1.7267,
"step": 98
},
{
"epoch": 0.01759296281487405,
"grad_norm": 1.164339868756061,
"learning_rate": 3.999438864652984e-05,
"loss": 1.7269,
"step": 99
},
{
"epoch": 0.017770669509973788,
"grad_norm": 1.3005449657712336,
"learning_rate": 3.999411827143604e-05,
"loss": 1.7254,
"step": 100
},
{
"epoch": 0.017948376205073526,
"grad_norm": 1.2267016874782017,
"learning_rate": 3.999384153583721e-05,
"loss": 1.7174,
"step": 101
},
{
"epoch": 0.018126082900173264,
"grad_norm": 1.3415199309846184,
"learning_rate": 3.999355843982139e-05,
"loss": 1.7588,
"step": 102
},
{
"epoch": 0.018303789595273002,
"grad_norm": 1.46439299105892,
"learning_rate": 3.999326898347863e-05,
"loss": 1.7692,
"step": 103
},
{
"epoch": 0.01848149629037274,
"grad_norm": 1.0940830919273599,
"learning_rate": 3.9992973166901026e-05,
"loss": 1.7229,
"step": 104
},
{
"epoch": 0.018659202985472478,
"grad_norm": 1.1122868183876007,
"learning_rate": 3.9992670990182666e-05,
"loss": 1.716,
"step": 105
},
{
"epoch": 0.018836909680572216,
"grad_norm": 1.2832399399935959,
"learning_rate": 3.999236245341968e-05,
"loss": 1.7099,
"step": 106
},
{
"epoch": 0.019014616375671953,
"grad_norm": 1.2052290699013377,
"learning_rate": 3.999204755671023e-05,
"loss": 1.7334,
"step": 107
},
{
"epoch": 0.01919232307077169,
"grad_norm": 1.3501909564343306,
"learning_rate": 3.999172630015448e-05,
"loss": 1.6963,
"step": 108
},
{
"epoch": 0.01937002976587143,
"grad_norm": 1.337947207792547,
"learning_rate": 3.999139868385464e-05,
"loss": 1.728,
"step": 109
},
{
"epoch": 0.019547736460971167,
"grad_norm": 1.1408689991726877,
"learning_rate": 3.999106470791492e-05,
"loss": 1.7138,
"step": 110
},
{
"epoch": 0.019725443156070905,
"grad_norm": 1.2461654801645532,
"learning_rate": 3.999072437244157e-05,
"loss": 1.7108,
"step": 111
},
{
"epoch": 0.019903149851170643,
"grad_norm": 1.059952165226307,
"learning_rate": 3.999037767754285e-05,
"loss": 1.6981,
"step": 112
},
{
"epoch": 0.02008085654627038,
"grad_norm": 1.2743303850708365,
"learning_rate": 3.999002462332905e-05,
"loss": 1.6938,
"step": 113
},
{
"epoch": 0.02025856324137012,
"grad_norm": 1.7941370078975247,
"learning_rate": 3.99896652099125e-05,
"loss": 1.6945,
"step": 114
},
{
"epoch": 0.020436269936469856,
"grad_norm": 1.1504631468231157,
"learning_rate": 3.998929943740752e-05,
"loss": 1.7146,
"step": 115
},
{
"epoch": 0.020613976631569594,
"grad_norm": 1.5912126603349892,
"learning_rate": 3.998892730593047e-05,
"loss": 1.6832,
"step": 116
},
{
"epoch": 0.020791683326669332,
"grad_norm": 1.0953860874009314,
"learning_rate": 3.998854881559974e-05,
"loss": 1.6759,
"step": 117
},
{
"epoch": 0.02096939002176907,
"grad_norm": 1.4139767435569135,
"learning_rate": 3.998816396653573e-05,
"loss": 1.7151,
"step": 118
},
{
"epoch": 0.021147096716868808,
"grad_norm": 1.0977514739279752,
"learning_rate": 3.998777275886086e-05,
"loss": 1.7496,
"step": 119
},
{
"epoch": 0.021324803411968546,
"grad_norm": 1.3780273856595209,
"learning_rate": 3.9987375192699603e-05,
"loss": 1.7266,
"step": 120
},
{
"epoch": 0.021502510107068284,
"grad_norm": 1.1269352423107948,
"learning_rate": 3.998697126817841e-05,
"loss": 1.7176,
"step": 121
},
{
"epoch": 0.02168021680216802,
"grad_norm": 1.31322843982768,
"learning_rate": 3.998656098542578e-05,
"loss": 1.7587,
"step": 122
},
{
"epoch": 0.02185792349726776,
"grad_norm": 1.0200873692563535,
"learning_rate": 3.9986144344572244e-05,
"loss": 1.6356,
"step": 123
},
{
"epoch": 0.022035630192367497,
"grad_norm": 1.291844214487771,
"learning_rate": 3.998572134575033e-05,
"loss": 1.6403,
"step": 124
},
{
"epoch": 0.022213336887467235,
"grad_norm": 1.100492761429373,
"learning_rate": 3.998529198909461e-05,
"loss": 1.6716,
"step": 125
},
{
"epoch": 0.022391043582566973,
"grad_norm": 1.1796528285011687,
"learning_rate": 3.9984856274741666e-05,
"loss": 1.6616,
"step": 126
},
{
"epoch": 0.02256875027766671,
"grad_norm": 1.0047217735713243,
"learning_rate": 3.998441420283011e-05,
"loss": 1.6513,
"step": 127
},
{
"epoch": 0.02274645697276645,
"grad_norm": 1.093238980413397,
"learning_rate": 3.998396577350057e-05,
"loss": 1.6824,
"step": 128
},
{
"epoch": 0.022924163667866187,
"grad_norm": 1.1099834399188475,
"learning_rate": 3.9983510986895714e-05,
"loss": 1.6936,
"step": 129
},
{
"epoch": 0.023101870362965925,
"grad_norm": 0.9769860410091555,
"learning_rate": 3.998304984316019e-05,
"loss": 1.7095,
"step": 130
},
{
"epoch": 0.023279577058065663,
"grad_norm": 1.0110411319087698,
"learning_rate": 3.9982582342440726e-05,
"loss": 1.6822,
"step": 131
},
{
"epoch": 0.0234572837531654,
"grad_norm": 0.9871700264225557,
"learning_rate": 3.9982108484886016e-05,
"loss": 1.6502,
"step": 132
},
{
"epoch": 0.02363499044826514,
"grad_norm": 1.15902232276902,
"learning_rate": 3.998162827064683e-05,
"loss": 1.6866,
"step": 133
},
{
"epoch": 0.023812697143364876,
"grad_norm": 0.975759185331363,
"learning_rate": 3.998114169987591e-05,
"loss": 1.6819,
"step": 134
},
{
"epoch": 0.023990403838464614,
"grad_norm": 0.9778109277514414,
"learning_rate": 3.998064877272806e-05,
"loss": 1.7422,
"step": 135
},
{
"epoch": 0.024168110533564352,
"grad_norm": 1.1690411452558918,
"learning_rate": 3.998014948936008e-05,
"loss": 1.6836,
"step": 136
},
{
"epoch": 0.02434581722866409,
"grad_norm": 0.9791255523820125,
"learning_rate": 3.99796438499308e-05,
"loss": 1.6646,
"step": 137
},
{
"epoch": 0.024523523923763828,
"grad_norm": 0.9436587948404581,
"learning_rate": 3.997913185460108e-05,
"loss": 1.665,
"step": 138
},
{
"epoch": 0.024701230618863566,
"grad_norm": 0.9334401594093369,
"learning_rate": 3.997861350353379e-05,
"loss": 1.6624,
"step": 139
},
{
"epoch": 0.024878937313963304,
"grad_norm": 1.2435220088991672,
"learning_rate": 3.997808879689384e-05,
"loss": 1.7147,
"step": 140
},
{
"epoch": 0.02505664400906304,
"grad_norm": 0.9371603860001542,
"learning_rate": 3.9977557734848127e-05,
"loss": 1.6909,
"step": 141
},
{
"epoch": 0.02523435070416278,
"grad_norm": 1.1940950095473595,
"learning_rate": 3.997702031756561e-05,
"loss": 1.6584,
"step": 142
},
{
"epoch": 0.025412057399262517,
"grad_norm": 0.8316847990556294,
"learning_rate": 3.997647654521724e-05,
"loss": 1.6917,
"step": 143
},
{
"epoch": 0.025589764094362255,
"grad_norm": 0.9161593823641543,
"learning_rate": 3.997592641797601e-05,
"loss": 1.6634,
"step": 144
},
{
"epoch": 0.025767470789461993,
"grad_norm": 0.9786836086938804,
"learning_rate": 3.997536993601692e-05,
"loss": 1.6671,
"step": 145
},
{
"epoch": 0.02594517748456173,
"grad_norm": 0.855676840967063,
"learning_rate": 3.997480709951701e-05,
"loss": 1.675,
"step": 146
},
{
"epoch": 0.02612288417966147,
"grad_norm": 0.8363786507232833,
"learning_rate": 3.997423790865531e-05,
"loss": 1.6687,
"step": 147
},
{
"epoch": 0.026300590874761207,
"grad_norm": 0.8803951501876376,
"learning_rate": 3.99736623636129e-05,
"loss": 1.6241,
"step": 148
},
{
"epoch": 0.026478297569860944,
"grad_norm": 0.8129597973951936,
"learning_rate": 3.997308046457287e-05,
"loss": 1.6455,
"step": 149
},
{
"epoch": 0.026656004264960682,
"grad_norm": 0.8974539756506671,
"learning_rate": 3.997249221172033e-05,
"loss": 1.7277,
"step": 150
},
{
"epoch": 0.02683371096006042,
"grad_norm": 0.8505404038913564,
"learning_rate": 3.997189760524242e-05,
"loss": 1.637,
"step": 151
},
{
"epoch": 0.027011417655160158,
"grad_norm": 0.8826018344845956,
"learning_rate": 3.997129664532829e-05,
"loss": 1.6634,
"step": 152
},
{
"epoch": 0.027189124350259896,
"grad_norm": 0.8835092378704205,
"learning_rate": 3.9970689332169124e-05,
"loss": 1.6723,
"step": 153
},
{
"epoch": 0.027366831045359634,
"grad_norm": 0.9596883536883095,
"learning_rate": 3.9970075665958124e-05,
"loss": 1.6404,
"step": 154
},
{
"epoch": 0.027544537740459372,
"grad_norm": 0.9427189754608358,
"learning_rate": 3.996945564689049e-05,
"loss": 1.6465,
"step": 155
},
{
"epoch": 0.02772224443555911,
"grad_norm": 0.8397429952454217,
"learning_rate": 3.996882927516347e-05,
"loss": 1.6791,
"step": 156
},
{
"epoch": 0.027899951130658848,
"grad_norm": 0.8241856238462872,
"learning_rate": 3.9968196550976335e-05,
"loss": 1.6456,
"step": 157
},
{
"epoch": 0.028077657825758585,
"grad_norm": 0.7975147423320084,
"learning_rate": 3.996755747453036e-05,
"loss": 1.6253,
"step": 158
},
{
"epoch": 0.028255364520858323,
"grad_norm": 0.8292748177610278,
"learning_rate": 3.996691204602884e-05,
"loss": 1.6612,
"step": 159
},
{
"epoch": 0.02843307121595806,
"grad_norm": 1.058701245624631,
"learning_rate": 3.99662602656771e-05,
"loss": 1.696,
"step": 160
},
{
"epoch": 0.0286107779110578,
"grad_norm": 1.1500166857862704,
"learning_rate": 3.9965602133682495e-05,
"loss": 1.6818,
"step": 161
},
{
"epoch": 0.028788484606157537,
"grad_norm": 0.8648827311482044,
"learning_rate": 3.9964937650254375e-05,
"loss": 1.6113,
"step": 162
},
{
"epoch": 0.028966191301257275,
"grad_norm": 0.8689604770405944,
"learning_rate": 3.9964266815604135e-05,
"loss": 1.634,
"step": 163
},
{
"epoch": 0.029143897996357013,
"grad_norm": 1.0026218523152455,
"learning_rate": 3.9963589629945174e-05,
"loss": 1.6214,
"step": 164
},
{
"epoch": 0.02932160469145675,
"grad_norm": 0.9167927455780134,
"learning_rate": 3.996290609349292e-05,
"loss": 1.6764,
"step": 165
},
{
"epoch": 0.02949931138655649,
"grad_norm": 0.7271140307386422,
"learning_rate": 3.996221620646482e-05,
"loss": 1.6439,
"step": 166
},
{
"epoch": 0.029677018081656226,
"grad_norm": 0.8392855712331214,
"learning_rate": 3.996151996908034e-05,
"loss": 1.6424,
"step": 167
},
{
"epoch": 0.029854724776755964,
"grad_norm": 0.813753405491957,
"learning_rate": 3.996081738156096e-05,
"loss": 1.6484,
"step": 168
},
{
"epoch": 0.030032431471855702,
"grad_norm": 0.7444444633949494,
"learning_rate": 3.996010844413019e-05,
"loss": 1.6558,
"step": 169
},
{
"epoch": 0.03021013816695544,
"grad_norm": 0.7790910607327334,
"learning_rate": 3.995939315701356e-05,
"loss": 1.6892,
"step": 170
},
{
"epoch": 0.030387844862055178,
"grad_norm": 0.7440081448607498,
"learning_rate": 3.995867152043861e-05,
"loss": 1.6204,
"step": 171
},
{
"epoch": 0.030565551557154916,
"grad_norm": 0.7797579094093324,
"learning_rate": 3.9957943534634914e-05,
"loss": 1.6518,
"step": 172
},
{
"epoch": 0.030743258252254654,
"grad_norm": 0.7858342392616521,
"learning_rate": 3.9957209199834055e-05,
"loss": 1.6389,
"step": 173
},
{
"epoch": 0.03092096494735439,
"grad_norm": 0.7358598428408287,
"learning_rate": 3.995646851626964e-05,
"loss": 1.6426,
"step": 174
},
{
"epoch": 0.03109867164245413,
"grad_norm": 0.7073073818589722,
"learning_rate": 3.9955721484177285e-05,
"loss": 1.6373,
"step": 175
},
{
"epoch": 0.03127637833755387,
"grad_norm": 0.721663299651343,
"learning_rate": 3.9954968103794643e-05,
"loss": 1.659,
"step": 176
},
{
"epoch": 0.03145408503265361,
"grad_norm": 0.6874679004006741,
"learning_rate": 3.9954208375361376e-05,
"loss": 1.6385,
"step": 177
},
{
"epoch": 0.03163179172775334,
"grad_norm": 0.783565368224001,
"learning_rate": 3.9953442299119166e-05,
"loss": 1.6435,
"step": 178
},
{
"epoch": 0.031809498422853084,
"grad_norm": 0.7615375817234503,
"learning_rate": 3.995266987531173e-05,
"loss": 1.6432,
"step": 179
},
{
"epoch": 0.03198720511795282,
"grad_norm": 1.1280090472867612,
"learning_rate": 3.995189110418477e-05,
"loss": 1.6434,
"step": 180
},
{
"epoch": 0.03216491181305256,
"grad_norm": 0.7176808912027098,
"learning_rate": 3.9951105985986044e-05,
"loss": 1.603,
"step": 181
},
{
"epoch": 0.032342618508152295,
"grad_norm": 0.7821539572615381,
"learning_rate": 3.9950314520965304e-05,
"loss": 1.6698,
"step": 182
},
{
"epoch": 0.032520325203252036,
"grad_norm": 0.7551471893694591,
"learning_rate": 3.9949516709374337e-05,
"loss": 1.6363,
"step": 183
},
{
"epoch": 0.03269803189835177,
"grad_norm": 0.7708841023413825,
"learning_rate": 3.9948712551466925e-05,
"loss": 1.6761,
"step": 184
},
{
"epoch": 0.03287573859345151,
"grad_norm": 0.7834224848126599,
"learning_rate": 3.994790204749891e-05,
"loss": 1.6228,
"step": 185
},
{
"epoch": 0.033053445288551246,
"grad_norm": 0.7406639672958636,
"learning_rate": 3.994708519772811e-05,
"loss": 1.6578,
"step": 186
},
{
"epoch": 0.03323115198365099,
"grad_norm": 0.6607770280317816,
"learning_rate": 3.994626200241439e-05,
"loss": 1.6126,
"step": 187
},
{
"epoch": 0.03340885867875072,
"grad_norm": 0.839001914524302,
"learning_rate": 3.9945432461819615e-05,
"loss": 1.6306,
"step": 188
},
{
"epoch": 0.03358656537385046,
"grad_norm": 0.7500546857316235,
"learning_rate": 3.994459657620769e-05,
"loss": 1.6146,
"step": 189
},
{
"epoch": 0.0337642720689502,
"grad_norm": 0.7355766253764826,
"learning_rate": 3.994375434584452e-05,
"loss": 1.6378,
"step": 190
},
{
"epoch": 0.03394197876404994,
"grad_norm": 0.7295400356796966,
"learning_rate": 3.9942905770998025e-05,
"loss": 1.6357,
"step": 191
},
{
"epoch": 0.03411968545914967,
"grad_norm": 0.77507010333163,
"learning_rate": 3.994205085193817e-05,
"loss": 1.6567,
"step": 192
},
{
"epoch": 0.034297392154249415,
"grad_norm": 0.8103614938851825,
"learning_rate": 3.9941189588936905e-05,
"loss": 1.6044,
"step": 193
},
{
"epoch": 0.03447509884934915,
"grad_norm": 0.7425207358790483,
"learning_rate": 3.994032198226823e-05,
"loss": 1.626,
"step": 194
},
{
"epoch": 0.03465280554444889,
"grad_norm": 0.8263935010817769,
"learning_rate": 3.993944803220813e-05,
"loss": 1.6657,
"step": 195
},
{
"epoch": 0.034830512239548625,
"grad_norm": 0.8536080395881326,
"learning_rate": 3.9938567739034634e-05,
"loss": 1.6047,
"step": 196
},
{
"epoch": 0.035008218934648366,
"grad_norm": 0.974441967840995,
"learning_rate": 3.993768110302778e-05,
"loss": 1.5915,
"step": 197
},
{
"epoch": 0.0351859256297481,
"grad_norm": 0.7988837761727187,
"learning_rate": 3.9936788124469615e-05,
"loss": 1.6338,
"step": 198
},
{
"epoch": 0.03536363232484784,
"grad_norm": 0.7676134680550952,
"learning_rate": 3.993588880364423e-05,
"loss": 1.6309,
"step": 199
},
{
"epoch": 0.035541339019947576,
"grad_norm": 0.7562579386144687,
"learning_rate": 3.99349831408377e-05,
"loss": 1.6072,
"step": 200
},
{
"epoch": 0.03571904571504732,
"grad_norm": 0.8196744298270455,
"learning_rate": 3.993407113633814e-05,
"loss": 1.5952,
"step": 201
},
{
"epoch": 0.03589675241014705,
"grad_norm": 0.7442489022304094,
"learning_rate": 3.993315279043568e-05,
"loss": 1.6452,
"step": 202
},
{
"epoch": 0.036074459105246794,
"grad_norm": 0.7673867435262333,
"learning_rate": 3.9932228103422445e-05,
"loss": 1.5831,
"step": 203
},
{
"epoch": 0.03625216580034653,
"grad_norm": 0.6954150427802075,
"learning_rate": 3.993129707559262e-05,
"loss": 1.5641,
"step": 204
},
{
"epoch": 0.03642987249544627,
"grad_norm": 0.7241362287979707,
"learning_rate": 3.9930359707242364e-05,
"loss": 1.5688,
"step": 205
},
{
"epoch": 0.036607579190546004,
"grad_norm": 0.7239506361936963,
"learning_rate": 3.9929415998669875e-05,
"loss": 1.5621,
"step": 206
},
{
"epoch": 0.036785285885645745,
"grad_norm": 0.7276074174145452,
"learning_rate": 3.992846595017538e-05,
"loss": 1.6295,
"step": 207
},
{
"epoch": 0.03696299258074548,
"grad_norm": 0.7351058281999437,
"learning_rate": 3.9927509562061084e-05,
"loss": 1.5886,
"step": 208
},
{
"epoch": 0.03714069927584522,
"grad_norm": 0.7649108180401459,
"learning_rate": 3.9926546834631244e-05,
"loss": 1.598,
"step": 209
},
{
"epoch": 0.037318405970944955,
"grad_norm": 0.6922782792107084,
"learning_rate": 3.9925577768192116e-05,
"loss": 1.5967,
"step": 210
},
{
"epoch": 0.0374961126660447,
"grad_norm": 0.7227123961773478,
"learning_rate": 3.9924602363051995e-05,
"loss": 1.6137,
"step": 211
},
{
"epoch": 0.03767381936114443,
"grad_norm": 0.8658388985175065,
"learning_rate": 3.992362061952115e-05,
"loss": 1.6328,
"step": 212
},
{
"epoch": 0.03785152605624417,
"grad_norm": 0.7012845164672776,
"learning_rate": 3.9922632537911916e-05,
"loss": 1.5706,
"step": 213
},
{
"epoch": 0.03802923275134391,
"grad_norm": 0.7073533397111548,
"learning_rate": 3.9921638118538607e-05,
"loss": 1.5984,
"step": 214
},
{
"epoch": 0.03820693944644365,
"grad_norm": 0.749298903495991,
"learning_rate": 3.9920637361717566e-05,
"loss": 1.629,
"step": 215
},
{
"epoch": 0.03838464614154338,
"grad_norm": 0.7132627401864864,
"learning_rate": 3.991963026776716e-05,
"loss": 1.5745,
"step": 216
},
{
"epoch": 0.038562352836643124,
"grad_norm": 0.6582435691410105,
"learning_rate": 3.9918616837007755e-05,
"loss": 1.5943,
"step": 217
},
{
"epoch": 0.03874005953174286,
"grad_norm": 0.735051752704862,
"learning_rate": 3.9917597069761746e-05,
"loss": 1.5936,
"step": 218
},
{
"epoch": 0.0389177662268426,
"grad_norm": 0.7225736618203965,
"learning_rate": 3.991657096635355e-05,
"loss": 1.5849,
"step": 219
},
{
"epoch": 0.039095472921942334,
"grad_norm": 0.6932373304706703,
"learning_rate": 3.991553852710958e-05,
"loss": 1.6322,
"step": 220
},
{
"epoch": 0.039273179617042075,
"grad_norm": 0.7244114772444988,
"learning_rate": 3.991449975235827e-05,
"loss": 1.6001,
"step": 221
},
{
"epoch": 0.03945088631214181,
"grad_norm": 0.7001906274642198,
"learning_rate": 3.991345464243009e-05,
"loss": 1.5732,
"step": 222
},
{
"epoch": 0.03962859300724155,
"grad_norm": 0.7044927180232868,
"learning_rate": 3.9912403197657485e-05,
"loss": 1.5818,
"step": 223
},
{
"epoch": 0.039806299702341286,
"grad_norm": 0.7074092249786992,
"learning_rate": 3.9911345418374965e-05,
"loss": 1.5495,
"step": 224
},
{
"epoch": 0.03998400639744103,
"grad_norm": 0.7500180775295753,
"learning_rate": 3.991028130491901e-05,
"loss": 1.5785,
"step": 225
},
{
"epoch": 0.04016171309254076,
"grad_norm": 1.1740152594917923,
"learning_rate": 3.990921085762815e-05,
"loss": 1.5928,
"step": 226
},
{
"epoch": 0.0403394197876405,
"grad_norm": 0.7400121835824248,
"learning_rate": 3.99081340768429e-05,
"loss": 1.5949,
"step": 227
},
{
"epoch": 0.04051712648274024,
"grad_norm": 0.7334083256554693,
"learning_rate": 3.9907050962905814e-05,
"loss": 1.599,
"step": 228
},
{
"epoch": 0.04069483317783998,
"grad_norm": 1.0434597608192497,
"learning_rate": 3.990596151616145e-05,
"loss": 1.5809,
"step": 229
},
{
"epoch": 0.04087253987293971,
"grad_norm": 0.738087413250202,
"learning_rate": 3.9904865736956376e-05,
"loss": 1.6317,
"step": 230
},
{
"epoch": 0.041050246568039454,
"grad_norm": 0.8488465040384472,
"learning_rate": 3.990376362563918e-05,
"loss": 1.6248,
"step": 231
},
{
"epoch": 0.04122795326313919,
"grad_norm": 0.6756982156611561,
"learning_rate": 3.990265518256047e-05,
"loss": 1.628,
"step": 232
},
{
"epoch": 0.04140565995823893,
"grad_norm": 0.7288919279596535,
"learning_rate": 3.990154040807287e-05,
"loss": 1.5614,
"step": 233
},
{
"epoch": 0.041583366653338664,
"grad_norm": 0.6566145492118339,
"learning_rate": 3.9900419302530984e-05,
"loss": 1.5737,
"step": 234
},
{
"epoch": 0.041761073348438406,
"grad_norm": 0.6655058813405247,
"learning_rate": 3.989929186629149e-05,
"loss": 1.59,
"step": 235
},
{
"epoch": 0.04193878004353814,
"grad_norm": 0.6443441162576786,
"learning_rate": 3.989815809971302e-05,
"loss": 1.6081,
"step": 236
},
{
"epoch": 0.04211648673863788,
"grad_norm": 0.6332335311364995,
"learning_rate": 3.989701800315626e-05,
"loss": 1.6122,
"step": 237
},
{
"epoch": 0.042294193433737616,
"grad_norm": 0.6875098300505887,
"learning_rate": 3.989587157698389e-05,
"loss": 1.5899,
"step": 238
},
{
"epoch": 0.04247190012883736,
"grad_norm": 0.6329371889047603,
"learning_rate": 3.989471882156061e-05,
"loss": 1.5385,
"step": 239
},
{
"epoch": 0.04264960682393709,
"grad_norm": 0.6890438173900657,
"learning_rate": 3.989355973725315e-05,
"loss": 1.589,
"step": 240
},
{
"epoch": 0.04282731351903683,
"grad_norm": 0.649969034770369,
"learning_rate": 3.9892394324430215e-05,
"loss": 1.6116,
"step": 241
},
{
"epoch": 0.04300502021413657,
"grad_norm": 0.6539742573060966,
"learning_rate": 3.989122258346255e-05,
"loss": 1.561,
"step": 242
},
{
"epoch": 0.04318272690923631,
"grad_norm": 0.7151440071094135,
"learning_rate": 3.989004451472291e-05,
"loss": 1.6205,
"step": 243
},
{
"epoch": 0.04336043360433604,
"grad_norm": 0.7315343886210296,
"learning_rate": 3.988886011858606e-05,
"loss": 1.6104,
"step": 244
},
{
"epoch": 0.043538140299435785,
"grad_norm": 0.6554664982633776,
"learning_rate": 3.9887669395428776e-05,
"loss": 1.5922,
"step": 245
},
{
"epoch": 0.04371584699453552,
"grad_norm": 0.6237231229773131,
"learning_rate": 3.988647234562986e-05,
"loss": 1.5628,
"step": 246
},
{
"epoch": 0.04389355368963526,
"grad_norm": 0.6720642999697922,
"learning_rate": 3.988526896957011e-05,
"loss": 1.609,
"step": 247
},
{
"epoch": 0.044071260384734995,
"grad_norm": 0.6713016601670303,
"learning_rate": 3.988405926763234e-05,
"loss": 1.5913,
"step": 248
},
{
"epoch": 0.044248967079834736,
"grad_norm": 0.6084809353563728,
"learning_rate": 3.9882843240201374e-05,
"loss": 1.5972,
"step": 249
},
{
"epoch": 0.04442667377493447,
"grad_norm": 0.664863528640758,
"learning_rate": 3.988162088766406e-05,
"loss": 1.5783,
"step": 250
},
{
"epoch": 0.04460438047003421,
"grad_norm": 0.7712350649059323,
"learning_rate": 3.988039221040926e-05,
"loss": 1.587,
"step": 251
},
{
"epoch": 0.044782087165133946,
"grad_norm": 0.6283095113043053,
"learning_rate": 3.9879157208827826e-05,
"loss": 1.5918,
"step": 252
},
{
"epoch": 0.04495979386023369,
"grad_norm": 0.6285104812310706,
"learning_rate": 3.9877915883312636e-05,
"loss": 1.5574,
"step": 253
},
{
"epoch": 0.04513750055533342,
"grad_norm": 0.7408286163139809,
"learning_rate": 3.9876668234258586e-05,
"loss": 1.5939,
"step": 254
},
{
"epoch": 0.045315207250433163,
"grad_norm": 0.72365491278897,
"learning_rate": 3.9875414262062574e-05,
"loss": 1.61,
"step": 255
},
{
"epoch": 0.0454929139455329,
"grad_norm": 0.5987209939564704,
"learning_rate": 3.9874153967123506e-05,
"loss": 1.5715,
"step": 256
},
{
"epoch": 0.04567062064063264,
"grad_norm": 0.6491660256638561,
"learning_rate": 3.9872887349842314e-05,
"loss": 1.5788,
"step": 257
},
{
"epoch": 0.045848327335732374,
"grad_norm": 0.6870830106812563,
"learning_rate": 3.987161441062194e-05,
"loss": 1.6321,
"step": 258
},
{
"epoch": 0.046026034030832115,
"grad_norm": 0.6303705808197163,
"learning_rate": 3.98703351498673e-05,
"loss": 1.5522,
"step": 259
},
{
"epoch": 0.04620374072593185,
"grad_norm": 0.6046693837080016,
"learning_rate": 3.9869049567985384e-05,
"loss": 1.5573,
"step": 260
},
{
"epoch": 0.04638144742103159,
"grad_norm": 0.6200554784234972,
"learning_rate": 3.9867757665385146e-05,
"loss": 1.625,
"step": 261
},
{
"epoch": 0.046559154116131325,
"grad_norm": 0.6155575553865976,
"learning_rate": 3.986645944247756e-05,
"loss": 1.5578,
"step": 262
},
{
"epoch": 0.04673686081123107,
"grad_norm": 0.5926760796496237,
"learning_rate": 3.986515489967562e-05,
"loss": 1.5679,
"step": 263
},
{
"epoch": 0.0469145675063308,
"grad_norm": 0.6276850578652482,
"learning_rate": 3.9863844037394326e-05,
"loss": 1.5496,
"step": 264
},
{
"epoch": 0.04709227420143054,
"grad_norm": 0.5743247021427743,
"learning_rate": 3.986252685605069e-05,
"loss": 1.5568,
"step": 265
},
{
"epoch": 0.04726998089653028,
"grad_norm": 0.6188170008117156,
"learning_rate": 3.986120335606372e-05,
"loss": 1.5567,
"step": 266
},
{
"epoch": 0.04744768759163002,
"grad_norm": 0.6262590459143326,
"learning_rate": 3.985987353785446e-05,
"loss": 1.5729,
"step": 267
},
{
"epoch": 0.04762539428672975,
"grad_norm": 0.5882137236893985,
"learning_rate": 3.9858537401845955e-05,
"loss": 1.5716,
"step": 268
},
{
"epoch": 0.047803100981829494,
"grad_norm": 0.6429338106636636,
"learning_rate": 3.985719494846324e-05,
"loss": 1.5793,
"step": 269
},
{
"epoch": 0.04798080767692923,
"grad_norm": 0.6136292642722356,
"learning_rate": 3.985584617813338e-05,
"loss": 1.5812,
"step": 270
},
{
"epoch": 0.04815851437202897,
"grad_norm": 0.5921186172783467,
"learning_rate": 3.985449109128545e-05,
"loss": 1.5836,
"step": 271
},
{
"epoch": 0.048336221067128704,
"grad_norm": 0.592930912737563,
"learning_rate": 3.985312968835051e-05,
"loss": 1.5948,
"step": 272
},
{
"epoch": 0.048513927762228445,
"grad_norm": 0.5969070770962142,
"learning_rate": 3.9851761969761676e-05,
"loss": 1.5933,
"step": 273
},
{
"epoch": 0.04869163445732818,
"grad_norm": 0.5986540926826113,
"learning_rate": 3.985038793595402e-05,
"loss": 1.5873,
"step": 274
},
{
"epoch": 0.04886934115242792,
"grad_norm": 0.5653727257023415,
"learning_rate": 3.984900758736467e-05,
"loss": 1.5742,
"step": 275
},
{
"epoch": 0.049047047847527656,
"grad_norm": 0.5785321664888484,
"learning_rate": 3.984762092443271e-05,
"loss": 1.5751,
"step": 276
},
{
"epoch": 0.0492247545426274,
"grad_norm": 0.5752695057905393,
"learning_rate": 3.98462279475993e-05,
"loss": 1.5223,
"step": 277
},
{
"epoch": 0.04940246123772713,
"grad_norm": 0.5737855977009398,
"learning_rate": 3.984482865730755e-05,
"loss": 1.5701,
"step": 278
},
{
"epoch": 0.04958016793282687,
"grad_norm": 0.8859101697698951,
"learning_rate": 3.98434230540026e-05,
"loss": 1.5868,
"step": 279
},
{
"epoch": 0.04975787462792661,
"grad_norm": 0.6604103282294089,
"learning_rate": 3.9842011138131605e-05,
"loss": 1.5451,
"step": 280
},
{
"epoch": 0.04993558132302635,
"grad_norm": 0.6014445268968067,
"learning_rate": 3.984059291014373e-05,
"loss": 1.5473,
"step": 281
},
{
"epoch": 0.05011328801812608,
"grad_norm": 0.66274358840456,
"learning_rate": 3.9839168370490126e-05,
"loss": 1.5398,
"step": 282
},
{
"epoch": 0.050290994713225824,
"grad_norm": 0.6890497987915554,
"learning_rate": 3.983773751962397e-05,
"loss": 1.6196,
"step": 283
},
{
"epoch": 0.05046870140832556,
"grad_norm": 0.6367478417588196,
"learning_rate": 3.983630035800044e-05,
"loss": 1.5496,
"step": 284
},
{
"epoch": 0.0506464081034253,
"grad_norm": 0.6087646589129531,
"learning_rate": 3.9834856886076734e-05,
"loss": 1.5621,
"step": 285
},
{
"epoch": 0.050824114798525034,
"grad_norm": 0.7374819234836557,
"learning_rate": 3.983340710431204e-05,
"loss": 1.6027,
"step": 286
},
{
"epoch": 0.051001821493624776,
"grad_norm": 0.617901097330726,
"learning_rate": 3.983195101316756e-05,
"loss": 1.5942,
"step": 287
},
{
"epoch": 0.05117952818872451,
"grad_norm": 0.6505466150395768,
"learning_rate": 3.983048861310651e-05,
"loss": 1.594,
"step": 288
},
{
"epoch": 0.05135723488382425,
"grad_norm": 0.6068555972114835,
"learning_rate": 3.98290199045941e-05,
"loss": 1.5616,
"step": 289
},
{
"epoch": 0.051534941578923986,
"grad_norm": 0.6185627096486049,
"learning_rate": 3.982754488809756e-05,
"loss": 1.5636,
"step": 290
},
{
"epoch": 0.05171264827402373,
"grad_norm": 0.6746390299421239,
"learning_rate": 3.982606356408611e-05,
"loss": 1.5487,
"step": 291
},
{
"epoch": 0.05189035496912346,
"grad_norm": 0.6120936702178421,
"learning_rate": 3.9824575933031e-05,
"loss": 1.5767,
"step": 292
},
{
"epoch": 0.0520680616642232,
"grad_norm": 0.6225461853656687,
"learning_rate": 3.982308199540547e-05,
"loss": 1.586,
"step": 293
},
{
"epoch": 0.05224576835932294,
"grad_norm": 0.6625303950781601,
"learning_rate": 3.982158175168476e-05,
"loss": 1.5928,
"step": 294
},
{
"epoch": 0.05242347505442268,
"grad_norm": 0.6575997142271126,
"learning_rate": 3.982007520234614e-05,
"loss": 1.5465,
"step": 295
},
{
"epoch": 0.05260118174952241,
"grad_norm": 0.6486386858486308,
"learning_rate": 3.9818562347868864e-05,
"loss": 1.5574,
"step": 296
},
{
"epoch": 0.052778888444622155,
"grad_norm": 0.8903452381185941,
"learning_rate": 3.98170431887342e-05,
"loss": 1.5309,
"step": 297
},
{
"epoch": 0.05295659513972189,
"grad_norm": 0.6367931947373039,
"learning_rate": 3.981551772542542e-05,
"loss": 1.5313,
"step": 298
},
{
"epoch": 0.05313430183482163,
"grad_norm": 1.899467996416295,
"learning_rate": 3.98139859584278e-05,
"loss": 1.5646,
"step": 299
},
{
"epoch": 0.053312008529921365,
"grad_norm": 0.8418594054245243,
"learning_rate": 3.981244788822864e-05,
"loss": 1.6002,
"step": 300
},
{
"epoch": 0.053489715225021106,
"grad_norm": 0.9178875168224906,
"learning_rate": 3.98109035153172e-05,
"loss": 1.5156,
"step": 301
},
{
"epoch": 0.05366742192012084,
"grad_norm": 0.7450123163923421,
"learning_rate": 3.980935284018481e-05,
"loss": 1.608,
"step": 302
},
{
"epoch": 0.05384512861522058,
"grad_norm": 0.7163385028894111,
"learning_rate": 3.980779586332473e-05,
"loss": 1.5257,
"step": 303
},
{
"epoch": 0.054022835310320316,
"grad_norm": 0.7829324723135711,
"learning_rate": 3.98062325852323e-05,
"loss": 1.5491,
"step": 304
},
{
"epoch": 0.05420054200542006,
"grad_norm": 0.6092940042108528,
"learning_rate": 3.98046630064048e-05,
"loss": 1.5504,
"step": 305
},
{
"epoch": 0.05437824870051979,
"grad_norm": 0.7161334444570752,
"learning_rate": 3.980308712734157e-05,
"loss": 1.557,
"step": 306
},
{
"epoch": 0.05455595539561953,
"grad_norm": 0.7503673088239973,
"learning_rate": 3.9801504948543896e-05,
"loss": 1.5484,
"step": 307
},
{
"epoch": 0.05473366209071927,
"grad_norm": 0.65617471638868,
"learning_rate": 3.9799916470515115e-05,
"loss": 1.5941,
"step": 308
},
{
"epoch": 0.05491136878581901,
"grad_norm": 0.643618126346545,
"learning_rate": 3.979832169376056e-05,
"loss": 1.5573,
"step": 309
},
{
"epoch": 0.055089075480918744,
"grad_norm": 0.6864168160554526,
"learning_rate": 3.979672061878754e-05,
"loss": 1.5945,
"step": 310
},
{
"epoch": 0.055266782176018485,
"grad_norm": 0.648797058731368,
"learning_rate": 3.97951132461054e-05,
"loss": 1.5804,
"step": 311
},
{
"epoch": 0.05544448887111822,
"grad_norm": 0.6899666518096214,
"learning_rate": 3.979349957622548e-05,
"loss": 1.5959,
"step": 312
},
{
"epoch": 0.05562219556621796,
"grad_norm": 0.6261019598416873,
"learning_rate": 3.97918796096611e-05,
"loss": 1.5378,
"step": 313
},
{
"epoch": 0.055799902261317695,
"grad_norm": 0.6494902628089826,
"learning_rate": 3.979025334692762e-05,
"loss": 1.5597,
"step": 314
},
{
"epoch": 0.055977608956417436,
"grad_norm": 0.6412982229641065,
"learning_rate": 3.9788620788542376e-05,
"loss": 1.5489,
"step": 315
},
{
"epoch": 0.05615531565151717,
"grad_norm": 0.6709862507032615,
"learning_rate": 3.978698193502472e-05,
"loss": 1.5704,
"step": 316
},
{
"epoch": 0.05633302234661691,
"grad_norm": 0.6337245614502641,
"learning_rate": 3.9785336786896e-05,
"loss": 1.5349,
"step": 317
},
{
"epoch": 0.05651072904171665,
"grad_norm": 0.6012724710524419,
"learning_rate": 3.978368534467956e-05,
"loss": 1.591,
"step": 318
},
{
"epoch": 0.05668843573681639,
"grad_norm": 0.6078341476640726,
"learning_rate": 3.978202760890077e-05,
"loss": 1.5256,
"step": 319
},
{
"epoch": 0.05686614243191612,
"grad_norm": 0.6429040661947356,
"learning_rate": 3.978036358008697e-05,
"loss": 1.5387,
"step": 320
},
{
"epoch": 0.057043849127015864,
"grad_norm": 0.5848821870661242,
"learning_rate": 3.977869325876754e-05,
"loss": 1.5183,
"step": 321
},
{
"epoch": 0.0572215558221156,
"grad_norm": 0.6391402357974219,
"learning_rate": 3.977701664547383e-05,
"loss": 1.5027,
"step": 322
},
{
"epoch": 0.05739926251721534,
"grad_norm": 0.6300434804032938,
"learning_rate": 3.97753337407392e-05,
"loss": 1.5771,
"step": 323
},
{
"epoch": 0.057576969212315074,
"grad_norm": 0.6182421844326157,
"learning_rate": 3.977364454509901e-05,
"loss": 1.5468,
"step": 324
},
{
"epoch": 0.057754675907414815,
"grad_norm": 0.5781831951575187,
"learning_rate": 3.977194905909063e-05,
"loss": 1.5383,
"step": 325
},
{
"epoch": 0.05793238260251455,
"grad_norm": 0.610257465717385,
"learning_rate": 3.977024728325343e-05,
"loss": 1.57,
"step": 326
},
{
"epoch": 0.05811008929761429,
"grad_norm": 0.6223214249859234,
"learning_rate": 3.9768539218128776e-05,
"loss": 1.5211,
"step": 327
},
{
"epoch": 0.058287795992714025,
"grad_norm": 0.5359093559619611,
"learning_rate": 3.9766824864260024e-05,
"loss": 1.558,
"step": 328
},
{
"epoch": 0.05846550268781377,
"grad_norm": 0.6093365286737032,
"learning_rate": 3.976510422219256e-05,
"loss": 1.5931,
"step": 329
},
{
"epoch": 0.0586432093829135,
"grad_norm": 0.6121280670377821,
"learning_rate": 3.976337729247374e-05,
"loss": 1.5101,
"step": 330
},
{
"epoch": 0.05882091607801324,
"grad_norm": 0.5667408636367063,
"learning_rate": 3.976164407565293e-05,
"loss": 1.5339,
"step": 331
},
{
"epoch": 0.05899862277311298,
"grad_norm": 0.7843380739382102,
"learning_rate": 3.975990457228151e-05,
"loss": 1.5818,
"step": 332
},
{
"epoch": 0.05917632946821272,
"grad_norm": 0.5782320207600083,
"learning_rate": 3.9758158782912845e-05,
"loss": 1.5353,
"step": 333
},
{
"epoch": 0.05935403616331245,
"grad_norm": 0.5855260646088456,
"learning_rate": 3.97564067081023e-05,
"loss": 1.5514,
"step": 334
},
{
"epoch": 0.059531742858412194,
"grad_norm": 0.5986702962097536,
"learning_rate": 3.9754648348407255e-05,
"loss": 1.5587,
"step": 335
},
{
"epoch": 0.05970944955351193,
"grad_norm": 0.5514999586840627,
"learning_rate": 3.975288370438706e-05,
"loss": 1.528,
"step": 336
},
{
"epoch": 0.05988715624861167,
"grad_norm": 0.5814995743879351,
"learning_rate": 3.9751112776603085e-05,
"loss": 1.5572,
"step": 337
},
{
"epoch": 0.060064862943711404,
"grad_norm": 0.6009713143863742,
"learning_rate": 3.9749335565618703e-05,
"loss": 1.5531,
"step": 338
},
{
"epoch": 0.060242569638811146,
"grad_norm": 0.5677799903364621,
"learning_rate": 3.974755207199927e-05,
"loss": 1.5036,
"step": 339
},
{
"epoch": 0.06042027633391088,
"grad_norm": 0.545160032494763,
"learning_rate": 3.974576229631217e-05,
"loss": 1.5212,
"step": 340
},
{
"epoch": 0.06059798302901062,
"grad_norm": 0.6209116751414889,
"learning_rate": 3.974396623912672e-05,
"loss": 1.5936,
"step": 341
},
{
"epoch": 0.060775689724110356,
"grad_norm": 0.5589438447005669,
"learning_rate": 3.974216390101433e-05,
"loss": 1.5383,
"step": 342
},
{
"epoch": 0.0609533964192101,
"grad_norm": 0.597872435119531,
"learning_rate": 3.974035528254833e-05,
"loss": 1.5525,
"step": 343
},
{
"epoch": 0.06113110311430983,
"grad_norm": 0.5975059631565749,
"learning_rate": 3.973854038430408e-05,
"loss": 1.5615,
"step": 344
},
{
"epoch": 0.06130880980940957,
"grad_norm": 0.550517616249525,
"learning_rate": 3.973671920685893e-05,
"loss": 1.5911,
"step": 345
},
{
"epoch": 0.06148651650450931,
"grad_norm": 0.5946411782961919,
"learning_rate": 3.973489175079224e-05,
"loss": 1.5219,
"step": 346
},
{
"epoch": 0.06166422319960905,
"grad_norm": 0.5350874802875358,
"learning_rate": 3.973305801668535e-05,
"loss": 1.5655,
"step": 347
},
{
"epoch": 0.06184192989470878,
"grad_norm": 0.553879618406574,
"learning_rate": 3.973121800512161e-05,
"loss": 1.4782,
"step": 348
},
{
"epoch": 0.062019636589808524,
"grad_norm": 0.575623212764177,
"learning_rate": 3.9729371716686354e-05,
"loss": 1.5489,
"step": 349
},
{
"epoch": 0.06219734328490826,
"grad_norm": 0.604107113340426,
"learning_rate": 3.9727519151966934e-05,
"loss": 1.4991,
"step": 350
},
{
"epoch": 0.062375049980008,
"grad_norm": 0.576776013725218,
"learning_rate": 3.972566031155268e-05,
"loss": 1.5286,
"step": 351
},
{
"epoch": 0.06255275667510773,
"grad_norm": 0.5490720413377165,
"learning_rate": 3.9723795196034914e-05,
"loss": 1.5223,
"step": 352
},
{
"epoch": 0.06273046337020748,
"grad_norm": 0.5853459828358363,
"learning_rate": 3.972192380600698e-05,
"loss": 1.5403,
"step": 353
},
{
"epoch": 0.06290817006530722,
"grad_norm": 0.5777157905607645,
"learning_rate": 3.9720046142064195e-05,
"loss": 1.541,
"step": 354
},
{
"epoch": 0.06308587676040694,
"grad_norm": 0.6154830843864926,
"learning_rate": 3.9718162204803884e-05,
"loss": 1.5687,
"step": 355
},
{
"epoch": 0.06326358345550669,
"grad_norm": 0.5604282889512652,
"learning_rate": 3.9716271994825355e-05,
"loss": 1.526,
"step": 356
},
{
"epoch": 0.06344129015060643,
"grad_norm": 0.6074609283390823,
"learning_rate": 3.971437551272992e-05,
"loss": 1.536,
"step": 357
},
{
"epoch": 0.06361899684570617,
"grad_norm": 0.5486754003626192,
"learning_rate": 3.9712472759120895e-05,
"loss": 1.533,
"step": 358
},
{
"epoch": 0.0637967035408059,
"grad_norm": 0.5970220361137938,
"learning_rate": 3.971056373460357e-05,
"loss": 1.526,
"step": 359
},
{
"epoch": 0.06397441023590564,
"grad_norm": 0.5610880551757065,
"learning_rate": 3.970864843978525e-05,
"loss": 1.4867,
"step": 360
},
{
"epoch": 0.06415211693100538,
"grad_norm": 0.5771941844024673,
"learning_rate": 3.970672687527523e-05,
"loss": 1.5622,
"step": 361
},
{
"epoch": 0.06432982362610512,
"grad_norm": 0.5661346392767126,
"learning_rate": 3.9704799041684785e-05,
"loss": 1.522,
"step": 362
},
{
"epoch": 0.06450753032120485,
"grad_norm": 0.5399590569971943,
"learning_rate": 3.97028649396272e-05,
"loss": 1.4895,
"step": 363
},
{
"epoch": 0.06468523701630459,
"grad_norm": 0.5834059940781823,
"learning_rate": 3.9700924569717745e-05,
"loss": 1.5579,
"step": 364
},
{
"epoch": 0.06486294371140433,
"grad_norm": 0.537304474670225,
"learning_rate": 3.969897793257369e-05,
"loss": 1.5034,
"step": 365
},
{
"epoch": 0.06504065040650407,
"grad_norm": 0.5556055932875937,
"learning_rate": 3.96970250288143e-05,
"loss": 1.5278,
"step": 366
},
{
"epoch": 0.0652183571016038,
"grad_norm": 0.5328232585418551,
"learning_rate": 3.969506585906083e-05,
"loss": 1.4571,
"step": 367
},
{
"epoch": 0.06539606379670354,
"grad_norm": 0.5521064401355986,
"learning_rate": 3.9693100423936535e-05,
"loss": 1.5244,
"step": 368
},
{
"epoch": 0.06557377049180328,
"grad_norm": 0.5719961773544361,
"learning_rate": 3.969112872406664e-05,
"loss": 1.5019,
"step": 369
},
{
"epoch": 0.06575147718690302,
"grad_norm": 0.5193330550824518,
"learning_rate": 3.96891507600784e-05,
"loss": 1.4871,
"step": 370
},
{
"epoch": 0.06592918388200275,
"grad_norm": 0.5876066247823369,
"learning_rate": 3.968716653260102e-05,
"loss": 1.4944,
"step": 371
},
{
"epoch": 0.06610689057710249,
"grad_norm": 0.549611814994684,
"learning_rate": 3.9685176042265736e-05,
"loss": 1.5258,
"step": 372
},
{
"epoch": 0.06628459727220223,
"grad_norm": 1.050133268154343,
"learning_rate": 3.968317928970576e-05,
"loss": 1.5625,
"step": 373
},
{
"epoch": 0.06646230396730197,
"grad_norm": 0.5809041889878118,
"learning_rate": 3.9681176275556294e-05,
"loss": 1.5442,
"step": 374
},
{
"epoch": 0.0666400106624017,
"grad_norm": 0.5933713026641647,
"learning_rate": 3.9679167000454526e-05,
"loss": 1.5668,
"step": 375
},
{
"epoch": 0.06681771735750144,
"grad_norm": 0.5416918911061573,
"learning_rate": 3.967715146503966e-05,
"loss": 1.4702,
"step": 376
},
{
"epoch": 0.06699542405260119,
"grad_norm": 0.5597248575733753,
"learning_rate": 3.9675129669952864e-05,
"loss": 1.5214,
"step": 377
},
{
"epoch": 0.06717313074770093,
"grad_norm": 0.5639694457838478,
"learning_rate": 3.967310161583732e-05,
"loss": 1.4872,
"step": 378
},
{
"epoch": 0.06735083744280065,
"grad_norm": 0.569431900815657,
"learning_rate": 3.967106730333817e-05,
"loss": 1.596,
"step": 379
},
{
"epoch": 0.0675285441379004,
"grad_norm": 0.5609756371364222,
"learning_rate": 3.9669026733102584e-05,
"loss": 1.5466,
"step": 380
},
{
"epoch": 0.06770625083300014,
"grad_norm": 0.537543135843494,
"learning_rate": 3.9666979905779704e-05,
"loss": 1.5579,
"step": 381
},
{
"epoch": 0.06788395752809988,
"grad_norm": 0.5472966185809368,
"learning_rate": 3.9664926822020665e-05,
"loss": 1.5622,
"step": 382
},
{
"epoch": 0.0680616642231996,
"grad_norm": 0.5233765589078109,
"learning_rate": 3.966286748247858e-05,
"loss": 1.5221,
"step": 383
},
{
"epoch": 0.06823937091829935,
"grad_norm": 0.5722426540105747,
"learning_rate": 3.966080188780858e-05,
"loss": 1.5329,
"step": 384
},
{
"epoch": 0.06841707761339909,
"grad_norm": 0.5550094302510139,
"learning_rate": 3.965873003866776e-05,
"loss": 1.5542,
"step": 385
},
{
"epoch": 0.06859478430849883,
"grad_norm": 0.7448403111465802,
"learning_rate": 3.965665193571521e-05,
"loss": 1.5076,
"step": 386
},
{
"epoch": 0.06877249100359856,
"grad_norm": 0.6450334662296353,
"learning_rate": 3.965456757961202e-05,
"loss": 1.5126,
"step": 387
},
{
"epoch": 0.0689501976986983,
"grad_norm": 0.5803174181980905,
"learning_rate": 3.9652476971021265e-05,
"loss": 1.5552,
"step": 388
},
{
"epoch": 0.06912790439379804,
"grad_norm": 0.5438936744422824,
"learning_rate": 3.9650380110608e-05,
"loss": 1.5294,
"step": 389
},
{
"epoch": 0.06930561108889778,
"grad_norm": 0.552550406162032,
"learning_rate": 3.964827699903929e-05,
"loss": 1.5204,
"step": 390
},
{
"epoch": 0.06948331778399751,
"grad_norm": 0.5591498763993794,
"learning_rate": 3.964616763698416e-05,
"loss": 1.4811,
"step": 391
},
{
"epoch": 0.06966102447909725,
"grad_norm": 0.5461390461112752,
"learning_rate": 3.964405202511364e-05,
"loss": 1.5076,
"step": 392
},
{
"epoch": 0.06983873117419699,
"grad_norm": 0.5667345671447359,
"learning_rate": 3.964193016410074e-05,
"loss": 1.5143,
"step": 393
},
{
"epoch": 0.07001643786929673,
"grad_norm": 0.5501326063884165,
"learning_rate": 3.9639802054620484e-05,
"loss": 1.5531,
"step": 394
},
{
"epoch": 0.07019414456439646,
"grad_norm": 0.5491927478478683,
"learning_rate": 3.963766769734985e-05,
"loss": 1.5204,
"step": 395
},
{
"epoch": 0.0703718512594962,
"grad_norm": 0.5190437476306103,
"learning_rate": 3.963552709296781e-05,
"loss": 1.4993,
"step": 396
},
{
"epoch": 0.07054955795459594,
"grad_norm": 0.6244025903051255,
"learning_rate": 3.9633380242155353e-05,
"loss": 1.569,
"step": 397
},
{
"epoch": 0.07072726464969568,
"grad_norm": 0.5642428641363313,
"learning_rate": 3.9631227145595404e-05,
"loss": 1.537,
"step": 398
},
{
"epoch": 0.07090497134479541,
"grad_norm": 0.5469343616394088,
"learning_rate": 3.962906780397292e-05,
"loss": 1.5402,
"step": 399
},
{
"epoch": 0.07108267803989515,
"grad_norm": 0.5675167460055183,
"learning_rate": 3.962690221797484e-05,
"loss": 1.5655,
"step": 400
},
{
"epoch": 0.0712603847349949,
"grad_norm": 0.5236097648651717,
"learning_rate": 3.962473038829005e-05,
"loss": 1.4918,
"step": 401
},
{
"epoch": 0.07143809143009464,
"grad_norm": 0.5429230593918086,
"learning_rate": 3.962255231560947e-05,
"loss": 1.5348,
"step": 402
},
{
"epoch": 0.07161579812519436,
"grad_norm": 0.5394369725281741,
"learning_rate": 3.9620368000625974e-05,
"loss": 1.5993,
"step": 403
},
{
"epoch": 0.0717935048202941,
"grad_norm": 0.5619786247699265,
"learning_rate": 3.961817744403445e-05,
"loss": 1.5322,
"step": 404
},
{
"epoch": 0.07197121151539385,
"grad_norm": 0.5146776217568697,
"learning_rate": 3.961598064653173e-05,
"loss": 1.5153,
"step": 405
},
{
"epoch": 0.07214891821049359,
"grad_norm": 0.5593946922948098,
"learning_rate": 3.961377760881668e-05,
"loss": 1.5721,
"step": 406
},
{
"epoch": 0.07232662490559331,
"grad_norm": 0.5485812664257627,
"learning_rate": 3.961156833159012e-05,
"loss": 1.5679,
"step": 407
},
{
"epoch": 0.07250433160069306,
"grad_norm": 0.5771076252211235,
"learning_rate": 3.960935281555486e-05,
"loss": 1.5461,
"step": 408
},
{
"epoch": 0.0726820382957928,
"grad_norm": 0.5562843859673611,
"learning_rate": 3.96071310614157e-05,
"loss": 1.5101,
"step": 409
},
{
"epoch": 0.07285974499089254,
"grad_norm": 0.544398941101236,
"learning_rate": 3.9604903069879424e-05,
"loss": 1.5331,
"step": 410
},
{
"epoch": 0.07303745168599227,
"grad_norm": 0.5585943225343063,
"learning_rate": 3.960266884165479e-05,
"loss": 1.5576,
"step": 411
},
{
"epoch": 0.07321515838109201,
"grad_norm": 0.5086018234683867,
"learning_rate": 3.9600428377452556e-05,
"loss": 1.4657,
"step": 412
},
{
"epoch": 0.07339286507619175,
"grad_norm": 0.5714853267507225,
"learning_rate": 3.959818167798544e-05,
"loss": 1.5368,
"step": 413
},
{
"epoch": 0.07357057177129149,
"grad_norm": 0.5639335128661495,
"learning_rate": 3.959592874396819e-05,
"loss": 1.5121,
"step": 414
},
{
"epoch": 0.07374827846639122,
"grad_norm": 0.5180777572125466,
"learning_rate": 3.959366957611748e-05,
"loss": 1.4882,
"step": 415
},
{
"epoch": 0.07392598516149096,
"grad_norm": 0.5774338370157905,
"learning_rate": 3.9591404175152e-05,
"loss": 1.5091,
"step": 416
},
{
"epoch": 0.0741036918565907,
"grad_norm": 0.5371926480874282,
"learning_rate": 3.9589132541792425e-05,
"loss": 1.5108,
"step": 417
},
{
"epoch": 0.07428139855169044,
"grad_norm": 0.5283312128668677,
"learning_rate": 3.958685467676139e-05,
"loss": 1.4952,
"step": 418
},
{
"epoch": 0.07445910524679017,
"grad_norm": 0.5808755762362829,
"learning_rate": 3.958457058078354e-05,
"loss": 1.5015,
"step": 419
},
{
"epoch": 0.07463681194188991,
"grad_norm": 0.6151888447876378,
"learning_rate": 3.9582280254585484e-05,
"loss": 1.5109,
"step": 420
},
{
"epoch": 0.07481451863698965,
"grad_norm": 0.5559515324723363,
"learning_rate": 3.957998369889581e-05,
"loss": 1.5615,
"step": 421
},
{
"epoch": 0.0749922253320894,
"grad_norm": 0.5295433964963849,
"learning_rate": 3.957768091444512e-05,
"loss": 1.5335,
"step": 422
},
{
"epoch": 0.07516993202718912,
"grad_norm": 0.5475121543721021,
"learning_rate": 3.957537190196594e-05,
"loss": 1.492,
"step": 423
},
{
"epoch": 0.07534763872228886,
"grad_norm": 0.5198070204155584,
"learning_rate": 3.957305666219284e-05,
"loss": 1.5004,
"step": 424
},
{
"epoch": 0.0755253454173886,
"grad_norm": 0.525015142240077,
"learning_rate": 3.957073519586232e-05,
"loss": 1.4531,
"step": 425
},
{
"epoch": 0.07570305211248834,
"grad_norm": 0.5617177065366951,
"learning_rate": 3.9568407503712894e-05,
"loss": 1.5263,
"step": 426
},
{
"epoch": 0.07588075880758807,
"grad_norm": 0.540508238702485,
"learning_rate": 3.956607358648503e-05,
"loss": 1.5326,
"step": 427
},
{
"epoch": 0.07605846550268781,
"grad_norm": 0.5339185546015459,
"learning_rate": 3.956373344492121e-05,
"loss": 1.5037,
"step": 428
},
{
"epoch": 0.07623617219778756,
"grad_norm": 0.5629342594969254,
"learning_rate": 3.9561387079765873e-05,
"loss": 1.5425,
"step": 429
},
{
"epoch": 0.0764138788928873,
"grad_norm": 0.5474816387202716,
"learning_rate": 3.955903449176543e-05,
"loss": 1.5133,
"step": 430
},
{
"epoch": 0.07659158558798702,
"grad_norm": 0.5288482474897431,
"learning_rate": 3.9556675681668296e-05,
"loss": 1.5018,
"step": 431
},
{
"epoch": 0.07676929228308677,
"grad_norm": 0.5672426383218543,
"learning_rate": 3.9554310650224844e-05,
"loss": 1.4795,
"step": 432
},
{
"epoch": 0.0769469989781865,
"grad_norm": 0.5390429117186044,
"learning_rate": 3.955193939818745e-05,
"loss": 1.4719,
"step": 433
},
{
"epoch": 0.07712470567328625,
"grad_norm": 0.534963183346552,
"learning_rate": 3.9549561926310425e-05,
"loss": 1.4709,
"step": 434
},
{
"epoch": 0.07730241236838598,
"grad_norm": 0.546237129032285,
"learning_rate": 3.954717823535011e-05,
"loss": 1.5207,
"step": 435
},
{
"epoch": 0.07748011906348572,
"grad_norm": 0.5459557346009855,
"learning_rate": 3.95447883260648e-05,
"loss": 1.5445,
"step": 436
},
{
"epoch": 0.07765782575858546,
"grad_norm": 0.5295511199366542,
"learning_rate": 3.954239219921477e-05,
"loss": 1.4685,
"step": 437
},
{
"epoch": 0.0778355324536852,
"grad_norm": 0.5344744171190403,
"learning_rate": 3.9539989855562265e-05,
"loss": 1.4955,
"step": 438
},
{
"epoch": 0.07801323914878493,
"grad_norm": 0.5283678522250121,
"learning_rate": 3.953758129587152e-05,
"loss": 1.5542,
"step": 439
},
{
"epoch": 0.07819094584388467,
"grad_norm": 0.5517866909219805,
"learning_rate": 3.9535166520908756e-05,
"loss": 1.5318,
"step": 440
},
{
"epoch": 0.07836865253898441,
"grad_norm": 0.5330574312516522,
"learning_rate": 3.953274553144214e-05,
"loss": 1.5503,
"step": 441
},
{
"epoch": 0.07854635923408415,
"grad_norm": 0.5212943006691471,
"learning_rate": 3.9530318328241834e-05,
"loss": 1.5068,
"step": 442
},
{
"epoch": 0.07872406592918388,
"grad_norm": 0.5452270698740264,
"learning_rate": 3.952788491207999e-05,
"loss": 1.4995,
"step": 443
},
{
"epoch": 0.07890177262428362,
"grad_norm": 0.5506224655347005,
"learning_rate": 3.952544528373072e-05,
"loss": 1.5202,
"step": 444
},
{
"epoch": 0.07907947931938336,
"grad_norm": 0.5272476139419594,
"learning_rate": 3.952299944397011e-05,
"loss": 1.4879,
"step": 445
},
{
"epoch": 0.0792571860144831,
"grad_norm": 0.5378678859124546,
"learning_rate": 3.952054739357623e-05,
"loss": 1.5277,
"step": 446
},
{
"epoch": 0.07943489270958283,
"grad_norm": 0.5895663110882516,
"learning_rate": 3.951808913332912e-05,
"loss": 1.5235,
"step": 447
},
{
"epoch": 0.07961259940468257,
"grad_norm": 0.5024954346475309,
"learning_rate": 3.95156246640108e-05,
"loss": 1.5182,
"step": 448
},
{
"epoch": 0.07979030609978231,
"grad_norm": 0.5365400833163875,
"learning_rate": 3.951315398640527e-05,
"loss": 1.4943,
"step": 449
},
{
"epoch": 0.07996801279488205,
"grad_norm": 0.5231779880791568,
"learning_rate": 3.9510677101298505e-05,
"loss": 1.49,
"step": 450
},
{
"epoch": 0.08014571948998178,
"grad_norm": 0.5087260537087984,
"learning_rate": 3.950819400947843e-05,
"loss": 1.446,
"step": 451
},
{
"epoch": 0.08032342618508152,
"grad_norm": 0.5451053846808916,
"learning_rate": 3.950570471173497e-05,
"loss": 1.4651,
"step": 452
},
{
"epoch": 0.08050113288018126,
"grad_norm": 0.5047995394352821,
"learning_rate": 3.950320920886003e-05,
"loss": 1.5517,
"step": 453
},
{
"epoch": 0.080678839575281,
"grad_norm": 0.5551312788848669,
"learning_rate": 3.950070750164746e-05,
"loss": 1.5134,
"step": 454
},
{
"epoch": 0.08085654627038073,
"grad_norm": 0.5372815830192459,
"learning_rate": 3.94981995908931e-05,
"loss": 1.5114,
"step": 455
},
{
"epoch": 0.08103425296548047,
"grad_norm": 0.513330422605639,
"learning_rate": 3.9495685477394783e-05,
"loss": 1.4906,
"step": 456
},
{
"epoch": 0.08121195966058022,
"grad_norm": 0.5218392988566734,
"learning_rate": 3.9493165161952273e-05,
"loss": 1.4856,
"step": 457
},
{
"epoch": 0.08138966635567996,
"grad_norm": 0.5051300509153095,
"learning_rate": 3.949063864536735e-05,
"loss": 1.4854,
"step": 458
},
{
"epoch": 0.08156737305077968,
"grad_norm": 0.5275878363530845,
"learning_rate": 3.948810592844372e-05,
"loss": 1.5578,
"step": 459
},
{
"epoch": 0.08174507974587943,
"grad_norm": 0.5415658109074087,
"learning_rate": 3.948556701198712e-05,
"loss": 1.4905,
"step": 460
},
{
"epoch": 0.08192278644097917,
"grad_norm": 0.5253580913020117,
"learning_rate": 3.94830218968052e-05,
"loss": 1.517,
"step": 461
},
{
"epoch": 0.08210049313607891,
"grad_norm": 0.5072646123894281,
"learning_rate": 3.948047058370763e-05,
"loss": 1.4518,
"step": 462
},
{
"epoch": 0.08227819983117864,
"grad_norm": 0.5245261322602132,
"learning_rate": 3.9477913073506006e-05,
"loss": 1.5175,
"step": 463
},
{
"epoch": 0.08245590652627838,
"grad_norm": 0.5258975109830185,
"learning_rate": 3.947534936701395e-05,
"loss": 1.4888,
"step": 464
},
{
"epoch": 0.08263361322137812,
"grad_norm": 0.5103617872442405,
"learning_rate": 3.9472779465047e-05,
"loss": 1.467,
"step": 465
},
{
"epoch": 0.08281131991647786,
"grad_norm": 0.5857461455854438,
"learning_rate": 3.94702033684227e-05,
"loss": 1.5378,
"step": 466
},
{
"epoch": 0.08298902661157759,
"grad_norm": 0.5191561861580356,
"learning_rate": 3.9467621077960566e-05,
"loss": 1.4943,
"step": 467
},
{
"epoch": 0.08316673330667733,
"grad_norm": 0.5152465240046262,
"learning_rate": 3.9465032594482054e-05,
"loss": 1.529,
"step": 468
},
{
"epoch": 0.08334444000177707,
"grad_norm": 0.532427811165418,
"learning_rate": 3.946243791881061e-05,
"loss": 1.491,
"step": 469
},
{
"epoch": 0.08352214669687681,
"grad_norm": 0.48752277098536784,
"learning_rate": 3.945983705177167e-05,
"loss": 1.4816,
"step": 470
},
{
"epoch": 0.08369985339197654,
"grad_norm": 0.5475712849218061,
"learning_rate": 3.9457229994192594e-05,
"loss": 1.5232,
"step": 471
},
{
"epoch": 0.08387756008707628,
"grad_norm": 0.5714032527801315,
"learning_rate": 3.9454616746902754e-05,
"loss": 1.5211,
"step": 472
},
{
"epoch": 0.08405526678217602,
"grad_norm": 0.5044601490229016,
"learning_rate": 3.945199731073347e-05,
"loss": 1.4933,
"step": 473
},
{
"epoch": 0.08423297347727576,
"grad_norm": 0.5165372241090943,
"learning_rate": 3.944937168651802e-05,
"loss": 1.4357,
"step": 474
},
{
"epoch": 0.08441068017237549,
"grad_norm": 0.509781696381027,
"learning_rate": 3.944673987509168e-05,
"loss": 1.5037,
"step": 475
},
{
"epoch": 0.08458838686747523,
"grad_norm": 0.5352348055403786,
"learning_rate": 3.9444101877291675e-05,
"loss": 1.4597,
"step": 476
},
{
"epoch": 0.08476609356257497,
"grad_norm": 0.5100838219913741,
"learning_rate": 3.9441457693957194e-05,
"loss": 1.4961,
"step": 477
},
{
"epoch": 0.08494380025767471,
"grad_norm": 0.5478583859772086,
"learning_rate": 3.943880732592941e-05,
"loss": 1.4798,
"step": 478
},
{
"epoch": 0.08512150695277444,
"grad_norm": 0.5342721233935873,
"learning_rate": 3.943615077405146e-05,
"loss": 1.5268,
"step": 479
},
{
"epoch": 0.08529921364787418,
"grad_norm": 0.5236771939282089,
"learning_rate": 3.943348803916843e-05,
"loss": 1.5245,
"step": 480
},
{
"epoch": 0.08547692034297392,
"grad_norm": 0.5327592444459749,
"learning_rate": 3.9430819122127386e-05,
"loss": 1.466,
"step": 481
},
{
"epoch": 0.08565462703807367,
"grad_norm": 0.5439228747143829,
"learning_rate": 3.942814402377738e-05,
"loss": 1.4835,
"step": 482
},
{
"epoch": 0.0858323337331734,
"grad_norm": 1.0172684340436942,
"learning_rate": 3.942546274496939e-05,
"loss": 1.5713,
"step": 483
},
{
"epoch": 0.08601004042827314,
"grad_norm": 0.6564948646304566,
"learning_rate": 3.942277528655638e-05,
"loss": 1.4765,
"step": 484
},
{
"epoch": 0.08618774712337288,
"grad_norm": 0.5516641613359271,
"learning_rate": 3.942008164939329e-05,
"loss": 1.5352,
"step": 485
},
{
"epoch": 0.08636545381847262,
"grad_norm": 0.6017409340451898,
"learning_rate": 3.941738183433703e-05,
"loss": 1.5116,
"step": 486
},
{
"epoch": 0.08654316051357235,
"grad_norm": 0.6098364933062008,
"learning_rate": 3.9414675842246444e-05,
"loss": 1.4949,
"step": 487
},
{
"epoch": 0.08672086720867209,
"grad_norm": 0.5294695163749313,
"learning_rate": 3.941196367398236e-05,
"loss": 1.5262,
"step": 488
},
{
"epoch": 0.08689857390377183,
"grad_norm": 0.6725732027381797,
"learning_rate": 3.9409245330407575e-05,
"loss": 1.4644,
"step": 489
},
{
"epoch": 0.08707628059887157,
"grad_norm": 0.5274761904259598,
"learning_rate": 3.9406520812386844e-05,
"loss": 1.5538,
"step": 490
},
{
"epoch": 0.0872539872939713,
"grad_norm": 0.5858748977761022,
"learning_rate": 3.940379012078688e-05,
"loss": 1.5144,
"step": 491
},
{
"epoch": 0.08743169398907104,
"grad_norm": 1.0779873696474975,
"learning_rate": 3.940105325647638e-05,
"loss": 1.459,
"step": 492
},
{
"epoch": 0.08760940068417078,
"grad_norm": 0.5545632357106861,
"learning_rate": 3.939831022032598e-05,
"loss": 1.4787,
"step": 493
},
{
"epoch": 0.08778710737927052,
"grad_norm": 0.62234717349128,
"learning_rate": 3.9395561013208306e-05,
"loss": 1.4958,
"step": 494
},
{
"epoch": 0.08796481407437025,
"grad_norm": 0.5056102080897161,
"learning_rate": 3.939280563599792e-05,
"loss": 1.5026,
"step": 495
},
{
"epoch": 0.08814252076946999,
"grad_norm": 0.6201921335807054,
"learning_rate": 3.9390044089571363e-05,
"loss": 1.5225,
"step": 496
},
{
"epoch": 0.08832022746456973,
"grad_norm": 0.5529675953467025,
"learning_rate": 3.938727637480713e-05,
"loss": 1.5512,
"step": 497
},
{
"epoch": 0.08849793415966947,
"grad_norm": 0.6042259813216383,
"learning_rate": 3.938450249258569e-05,
"loss": 1.507,
"step": 498
},
{
"epoch": 0.0886756408547692,
"grad_norm": 1.142432004786422,
"learning_rate": 3.938172244378947e-05,
"loss": 1.5717,
"step": 499
},
{
"epoch": 0.08885334754986894,
"grad_norm": 0.5480087872975086,
"learning_rate": 3.937893622930285e-05,
"loss": 1.4975,
"step": 500
},
{
"epoch": 0.08903105424496868,
"grad_norm": 0.613822816385908,
"learning_rate": 3.937614385001218e-05,
"loss": 1.5139,
"step": 501
},
{
"epoch": 0.08920876094006842,
"grad_norm": 0.5510372647165712,
"learning_rate": 3.937334530680576e-05,
"loss": 1.5124,
"step": 502
},
{
"epoch": 0.08938646763516815,
"grad_norm": 0.5979595047504339,
"learning_rate": 3.9370540600573866e-05,
"loss": 1.5971,
"step": 503
},
{
"epoch": 0.08956417433026789,
"grad_norm": 0.5717822921297301,
"learning_rate": 3.936772973220873e-05,
"loss": 1.4976,
"step": 504
},
{
"epoch": 0.08974188102536763,
"grad_norm": 0.559609202180981,
"learning_rate": 3.9364912702604546e-05,
"loss": 1.484,
"step": 505
},
{
"epoch": 0.08991958772046738,
"grad_norm": 0.5745166317597064,
"learning_rate": 3.936208951265745e-05,
"loss": 1.5011,
"step": 506
},
{
"epoch": 0.0900972944155671,
"grad_norm": 0.5201571707053739,
"learning_rate": 3.9359260163265565e-05,
"loss": 1.5203,
"step": 507
},
{
"epoch": 0.09027500111066684,
"grad_norm": 0.5536454796094519,
"learning_rate": 3.935642465532895e-05,
"loss": 1.4759,
"step": 508
},
{
"epoch": 0.09045270780576659,
"grad_norm": 0.5647602108555851,
"learning_rate": 3.935358298974964e-05,
"loss": 1.4926,
"step": 509
},
{
"epoch": 0.09063041450086633,
"grad_norm": 0.5025494360655633,
"learning_rate": 3.9350735167431625e-05,
"loss": 1.4797,
"step": 510
},
{
"epoch": 0.09080812119596605,
"grad_norm": 0.5682293349767763,
"learning_rate": 3.934788118928084e-05,
"loss": 1.4884,
"step": 511
},
{
"epoch": 0.0909858278910658,
"grad_norm": 0.5007448700097056,
"learning_rate": 3.93450210562052e-05,
"loss": 1.5034,
"step": 512
},
{
"epoch": 0.09116353458616554,
"grad_norm": 0.5466078236904641,
"learning_rate": 3.9342154769114554e-05,
"loss": 1.5469,
"step": 513
},
{
"epoch": 0.09134124128126528,
"grad_norm": 0.5031228905491764,
"learning_rate": 3.933928232892074e-05,
"loss": 1.454,
"step": 514
},
{
"epoch": 0.091518947976365,
"grad_norm": 0.6438023444166119,
"learning_rate": 3.933640373653752e-05,
"loss": 1.5253,
"step": 515
},
{
"epoch": 0.09169665467146475,
"grad_norm": 0.520705544718714,
"learning_rate": 3.933351899288064e-05,
"loss": 1.4795,
"step": 516
},
{
"epoch": 0.09187436136656449,
"grad_norm": 0.5258831815198243,
"learning_rate": 3.9330628098867775e-05,
"loss": 1.5148,
"step": 517
},
{
"epoch": 0.09205206806166423,
"grad_norm": 0.558897413073457,
"learning_rate": 3.932773105541859e-05,
"loss": 1.5162,
"step": 518
},
{
"epoch": 0.09222977475676396,
"grad_norm": 0.5414042892993217,
"learning_rate": 3.932482786345468e-05,
"loss": 1.4595,
"step": 519
},
{
"epoch": 0.0924074814518637,
"grad_norm": 0.5723469085591627,
"learning_rate": 3.9321918523899605e-05,
"loss": 1.5141,
"step": 520
},
{
"epoch": 0.09258518814696344,
"grad_norm": 0.6136392461879155,
"learning_rate": 3.931900303767889e-05,
"loss": 1.485,
"step": 521
},
{
"epoch": 0.09276289484206318,
"grad_norm": 0.5666687211228527,
"learning_rate": 3.9316081405719996e-05,
"loss": 1.4766,
"step": 522
},
{
"epoch": 0.09294060153716291,
"grad_norm": 0.5329475558287496,
"learning_rate": 3.931315362895235e-05,
"loss": 1.4927,
"step": 523
},
{
"epoch": 0.09311830823226265,
"grad_norm": 0.531309077248914,
"learning_rate": 3.931021970830733e-05,
"loss": 1.4581,
"step": 524
},
{
"epoch": 0.09329601492736239,
"grad_norm": 0.5352765018248674,
"learning_rate": 3.930727964471828e-05,
"loss": 1.5379,
"step": 525
},
{
"epoch": 0.09347372162246213,
"grad_norm": 0.5596795970775909,
"learning_rate": 3.930433343912048e-05,
"loss": 1.5122,
"step": 526
},
{
"epoch": 0.09365142831756186,
"grad_norm": 0.5198775267084861,
"learning_rate": 3.9301381092451184e-05,
"loss": 1.4587,
"step": 527
},
{
"epoch": 0.0938291350126616,
"grad_norm": 0.5517072411756261,
"learning_rate": 3.929842260564959e-05,
"loss": 1.4421,
"step": 528
},
{
"epoch": 0.09400684170776134,
"grad_norm": 0.5326209672863561,
"learning_rate": 3.929545797965683e-05,
"loss": 1.4964,
"step": 529
},
{
"epoch": 0.09418454840286108,
"grad_norm": 0.5389455780527872,
"learning_rate": 3.9292487215416025e-05,
"loss": 1.5112,
"step": 530
},
{
"epoch": 0.09436225509796081,
"grad_norm": 0.5268530661465088,
"learning_rate": 3.9289510313872224e-05,
"loss": 1.4764,
"step": 531
},
{
"epoch": 0.09453996179306055,
"grad_norm": 0.5093430385008781,
"learning_rate": 3.928652727597244e-05,
"loss": 1.5006,
"step": 532
},
{
"epoch": 0.0947176684881603,
"grad_norm": 0.5442649770044492,
"learning_rate": 3.928353810266563e-05,
"loss": 1.4676,
"step": 533
},
{
"epoch": 0.09489537518326004,
"grad_norm": 0.4866742770369063,
"learning_rate": 3.9280542794902704e-05,
"loss": 1.458,
"step": 534
},
{
"epoch": 0.09507308187835976,
"grad_norm": 0.5377800117025099,
"learning_rate": 3.927754135363652e-05,
"loss": 1.5127,
"step": 535
},
{
"epoch": 0.0952507885734595,
"grad_norm": 0.49828153701162353,
"learning_rate": 3.9274533779821915e-05,
"loss": 1.4745,
"step": 536
},
{
"epoch": 0.09542849526855925,
"grad_norm": 0.4949654206040928,
"learning_rate": 3.927152007441564e-05,
"loss": 1.4567,
"step": 537
},
{
"epoch": 0.09560620196365899,
"grad_norm": 0.4980224254890624,
"learning_rate": 3.926850023837641e-05,
"loss": 1.4979,
"step": 538
},
{
"epoch": 0.09578390865875872,
"grad_norm": 0.512278422615779,
"learning_rate": 3.926547427266489e-05,
"loss": 1.4887,
"step": 539
},
{
"epoch": 0.09596161535385846,
"grad_norm": 0.5308497180957137,
"learning_rate": 3.926244217824369e-05,
"loss": 1.5188,
"step": 540
},
{
"epoch": 0.0961393220489582,
"grad_norm": 0.5038192859879042,
"learning_rate": 3.92594039560774e-05,
"loss": 1.5279,
"step": 541
},
{
"epoch": 0.09631702874405794,
"grad_norm": 0.5117116646802263,
"learning_rate": 3.925635960713252e-05,
"loss": 1.4839,
"step": 542
},
{
"epoch": 0.09649473543915767,
"grad_norm": 0.5264997966628753,
"learning_rate": 3.9253309132377525e-05,
"loss": 1.4827,
"step": 543
},
{
"epoch": 0.09667244213425741,
"grad_norm": 0.5194266517864689,
"learning_rate": 3.9250252532782804e-05,
"loss": 1.4858,
"step": 544
},
{
"epoch": 0.09685014882935715,
"grad_norm": 0.5241339936438614,
"learning_rate": 3.9247189809320746e-05,
"loss": 1.4837,
"step": 545
},
{
"epoch": 0.09702785552445689,
"grad_norm": 0.5005230639346845,
"learning_rate": 3.924412096296565e-05,
"loss": 1.4712,
"step": 546
},
{
"epoch": 0.09720556221955662,
"grad_norm": 0.5221898629652085,
"learning_rate": 3.9241045994693764e-05,
"loss": 1.4553,
"step": 547
},
{
"epoch": 0.09738326891465636,
"grad_norm": 0.5169252455550911,
"learning_rate": 3.923796490548332e-05,
"loss": 1.4949,
"step": 548
},
{
"epoch": 0.0975609756097561,
"grad_norm": 0.5036403314700781,
"learning_rate": 3.9234877696314435e-05,
"loss": 1.5009,
"step": 549
},
{
"epoch": 0.09773868230485584,
"grad_norm": 0.5197610404921802,
"learning_rate": 3.9231784368169236e-05,
"loss": 1.4957,
"step": 550
},
{
"epoch": 0.09791638899995557,
"grad_norm": 0.48634498646574387,
"learning_rate": 3.9228684922031754e-05,
"loss": 1.4561,
"step": 551
},
{
"epoch": 0.09809409569505531,
"grad_norm": 0.5039715387382648,
"learning_rate": 3.9225579358888e-05,
"loss": 1.5036,
"step": 552
},
{
"epoch": 0.09827180239015505,
"grad_norm": 0.5281110815424767,
"learning_rate": 3.9222467679725884e-05,
"loss": 1.5436,
"step": 553
},
{
"epoch": 0.0984495090852548,
"grad_norm": 0.4802840061816844,
"learning_rate": 3.921934988553531e-05,
"loss": 1.4642,
"step": 554
},
{
"epoch": 0.09862721578035452,
"grad_norm": 0.6499862274367488,
"learning_rate": 3.92162259773081e-05,
"loss": 1.4708,
"step": 555
},
{
"epoch": 0.09880492247545426,
"grad_norm": 0.4951732552179106,
"learning_rate": 3.921309595603803e-05,
"loss": 1.4926,
"step": 556
},
{
"epoch": 0.098982629170554,
"grad_norm": 0.49247627255163745,
"learning_rate": 3.9209959822720825e-05,
"loss": 1.479,
"step": 557
},
{
"epoch": 0.09916033586565375,
"grad_norm": 0.5214804486972271,
"learning_rate": 3.920681757835413e-05,
"loss": 1.4862,
"step": 558
},
{
"epoch": 0.09933804256075347,
"grad_norm": 0.5268517377951137,
"learning_rate": 3.920366922393757e-05,
"loss": 1.5212,
"step": 559
},
{
"epoch": 0.09951574925585321,
"grad_norm": 0.5086369091612315,
"learning_rate": 3.920051476047269e-05,
"loss": 1.4441,
"step": 560
},
{
"epoch": 0.09969345595095296,
"grad_norm": 0.5668506227578013,
"learning_rate": 3.9197354188962974e-05,
"loss": 1.492,
"step": 561
},
{
"epoch": 0.0998711626460527,
"grad_norm": 0.5214837748772414,
"learning_rate": 3.919418751041387e-05,
"loss": 1.4975,
"step": 562
},
{
"epoch": 0.10004886934115242,
"grad_norm": 0.5121146073485657,
"learning_rate": 3.919101472583276e-05,
"loss": 1.4779,
"step": 563
},
{
"epoch": 0.10022657603625217,
"grad_norm": 0.5309494197368134,
"learning_rate": 3.918783583622896e-05,
"loss": 1.4749,
"step": 564
},
{
"epoch": 0.10040428273135191,
"grad_norm": 0.5328054151881617,
"learning_rate": 3.9184650842613733e-05,
"loss": 1.4864,
"step": 565
},
{
"epoch": 0.10058198942645165,
"grad_norm": 0.5271985653200746,
"learning_rate": 3.9181459746000306e-05,
"loss": 1.4575,
"step": 566
},
{
"epoch": 0.10075969612155138,
"grad_norm": 0.5227990970834103,
"learning_rate": 3.917826254740379e-05,
"loss": 1.5159,
"step": 567
},
{
"epoch": 0.10093740281665112,
"grad_norm": 0.5274994534195018,
"learning_rate": 3.917505924784131e-05,
"loss": 1.5027,
"step": 568
},
{
"epoch": 0.10111510951175086,
"grad_norm": 0.5577044665041152,
"learning_rate": 3.9171849848331866e-05,
"loss": 1.5073,
"step": 569
},
{
"epoch": 0.1012928162068506,
"grad_norm": 0.5652239348110253,
"learning_rate": 3.916863434989645e-05,
"loss": 1.4524,
"step": 570
},
{
"epoch": 0.10147052290195033,
"grad_norm": 0.517176536839651,
"learning_rate": 3.9165412753557965e-05,
"loss": 1.4854,
"step": 571
},
{
"epoch": 0.10164822959705007,
"grad_norm": 0.541735408962882,
"learning_rate": 3.916218506034127e-05,
"loss": 1.463,
"step": 572
},
{
"epoch": 0.10182593629214981,
"grad_norm": 0.5367423622916893,
"learning_rate": 3.915895127127313e-05,
"loss": 1.5302,
"step": 573
},
{
"epoch": 0.10200364298724955,
"grad_norm": 0.5420246645626278,
"learning_rate": 3.91557113873823e-05,
"loss": 1.4344,
"step": 574
},
{
"epoch": 0.10218134968234928,
"grad_norm": 0.5033312743174577,
"learning_rate": 3.9152465409699434e-05,
"loss": 1.4605,
"step": 575
},
{
"epoch": 0.10235905637744902,
"grad_norm": 0.5069274245026906,
"learning_rate": 3.914921333925714e-05,
"loss": 1.5087,
"step": 576
},
{
"epoch": 0.10253676307254876,
"grad_norm": 0.4982242990006382,
"learning_rate": 3.9145955177089976e-05,
"loss": 1.5041,
"step": 577
},
{
"epoch": 0.1027144697676485,
"grad_norm": 0.5476001683545442,
"learning_rate": 3.914269092423441e-05,
"loss": 1.48,
"step": 578
},
{
"epoch": 0.10289217646274823,
"grad_norm": 0.5136845078510653,
"learning_rate": 3.913942058172886e-05,
"loss": 1.4722,
"step": 579
},
{
"epoch": 0.10306988315784797,
"grad_norm": 0.508162472694997,
"learning_rate": 3.913614415061369e-05,
"loss": 1.4923,
"step": 580
},
{
"epoch": 0.10324758985294771,
"grad_norm": 0.5460644620291856,
"learning_rate": 3.9132861631931206e-05,
"loss": 1.4864,
"step": 581
},
{
"epoch": 0.10342529654804745,
"grad_norm": 0.5029387412741884,
"learning_rate": 3.912957302672562e-05,
"loss": 1.4438,
"step": 582
},
{
"epoch": 0.10360300324314718,
"grad_norm": 0.5515764139248127,
"learning_rate": 3.912627833604311e-05,
"loss": 1.4744,
"step": 583
},
{
"epoch": 0.10378070993824692,
"grad_norm": 0.9102325494571837,
"learning_rate": 3.9122977560931776e-05,
"loss": 1.4742,
"step": 584
},
{
"epoch": 0.10395841663334666,
"grad_norm": 0.5123150193400218,
"learning_rate": 3.9119670702441654e-05,
"loss": 1.4463,
"step": 585
},
{
"epoch": 0.1041361233284464,
"grad_norm": 0.5585050665448704,
"learning_rate": 3.911635776162472e-05,
"loss": 1.5245,
"step": 586
},
{
"epoch": 0.10431383002354613,
"grad_norm": 0.5525106766014709,
"learning_rate": 3.911303873953488e-05,
"loss": 1.4958,
"step": 587
},
{
"epoch": 0.10449153671864587,
"grad_norm": 0.5097684621300219,
"learning_rate": 3.910971363722798e-05,
"loss": 1.484,
"step": 588
},
{
"epoch": 0.10466924341374562,
"grad_norm": 0.5354300351242434,
"learning_rate": 3.91063824557618e-05,
"loss": 1.4575,
"step": 589
},
{
"epoch": 0.10484695010884536,
"grad_norm": 0.4938255086931856,
"learning_rate": 3.9103045196196044e-05,
"loss": 1.4208,
"step": 590
},
{
"epoch": 0.10502465680394509,
"grad_norm": 0.5627478062284161,
"learning_rate": 3.909970185959237e-05,
"loss": 1.492,
"step": 591
},
{
"epoch": 0.10520236349904483,
"grad_norm": 0.4905293847643088,
"learning_rate": 3.909635244701434e-05,
"loss": 1.5343,
"step": 592
},
{
"epoch": 0.10538007019414457,
"grad_norm": 0.5583618838981772,
"learning_rate": 3.9092996959527483e-05,
"loss": 1.4763,
"step": 593
},
{
"epoch": 0.10555777688924431,
"grad_norm": 0.5023479365682482,
"learning_rate": 3.908963539819923e-05,
"loss": 1.4858,
"step": 594
},
{
"epoch": 0.10573548358434404,
"grad_norm": 0.5452681366578828,
"learning_rate": 3.908626776409896e-05,
"loss": 1.5071,
"step": 595
},
{
"epoch": 0.10591319027944378,
"grad_norm": 0.49825368012799215,
"learning_rate": 3.908289405829797e-05,
"loss": 1.5111,
"step": 596
},
{
"epoch": 0.10609089697454352,
"grad_norm": 0.5041330157752002,
"learning_rate": 3.907951428186953e-05,
"loss": 1.4981,
"step": 597
},
{
"epoch": 0.10626860366964326,
"grad_norm": 0.5736577756068604,
"learning_rate": 3.907612843588878e-05,
"loss": 1.4482,
"step": 598
},
{
"epoch": 0.10644631036474299,
"grad_norm": 0.5264917355211317,
"learning_rate": 3.9072736521432826e-05,
"loss": 1.5242,
"step": 599
},
{
"epoch": 0.10662401705984273,
"grad_norm": 0.5659047757584583,
"learning_rate": 3.9069338539580715e-05,
"loss": 1.5262,
"step": 600
},
{
"epoch": 0.10680172375494247,
"grad_norm": 0.6473632807332391,
"learning_rate": 3.90659344914134e-05,
"loss": 1.5001,
"step": 601
},
{
"epoch": 0.10697943045004221,
"grad_norm": 0.4782362275636608,
"learning_rate": 3.906252437801377e-05,
"loss": 1.4752,
"step": 602
},
{
"epoch": 0.10715713714514194,
"grad_norm": 0.4860938153673032,
"learning_rate": 3.905910820046664e-05,
"loss": 1.4896,
"step": 603
},
{
"epoch": 0.10733484384024168,
"grad_norm": 0.46276368033569376,
"learning_rate": 3.9055685959858785e-05,
"loss": 1.4665,
"step": 604
},
{
"epoch": 0.10751255053534142,
"grad_norm": 0.4901955345036761,
"learning_rate": 3.905225765727886e-05,
"loss": 1.4143,
"step": 605
},
{
"epoch": 0.10769025723044116,
"grad_norm": 0.4908376040367661,
"learning_rate": 3.9048823293817475e-05,
"loss": 1.4591,
"step": 606
},
{
"epoch": 0.10786796392554089,
"grad_norm": 0.48726853696554795,
"learning_rate": 3.9045382870567176e-05,
"loss": 1.4857,
"step": 607
},
{
"epoch": 0.10804567062064063,
"grad_norm": 0.5230141735996594,
"learning_rate": 3.904193638862242e-05,
"loss": 1.5043,
"step": 608
},
{
"epoch": 0.10822337731574037,
"grad_norm": 0.5089586948916339,
"learning_rate": 3.90384838490796e-05,
"loss": 1.4739,
"step": 609
},
{
"epoch": 0.10840108401084012,
"grad_norm": 0.4958062557037028,
"learning_rate": 3.9035025253037035e-05,
"loss": 1.4763,
"step": 610
},
{
"epoch": 0.10857879070593984,
"grad_norm": 0.518336195386355,
"learning_rate": 3.9031560601594964e-05,
"loss": 1.5382,
"step": 611
},
{
"epoch": 0.10875649740103958,
"grad_norm": 0.49075499981701515,
"learning_rate": 3.9028089895855564e-05,
"loss": 1.4608,
"step": 612
},
{
"epoch": 0.10893420409613933,
"grad_norm": 0.5046447950113302,
"learning_rate": 3.9024613136922925e-05,
"loss": 1.4677,
"step": 613
},
{
"epoch": 0.10911191079123907,
"grad_norm": 0.4856083949336938,
"learning_rate": 3.9021130325903076e-05,
"loss": 1.4524,
"step": 614
},
{
"epoch": 0.1092896174863388,
"grad_norm": 0.4860570733631214,
"learning_rate": 3.901764146390396e-05,
"loss": 1.4562,
"step": 615
},
{
"epoch": 0.10946732418143854,
"grad_norm": 0.48476547981185947,
"learning_rate": 3.901414655203545e-05,
"loss": 1.4723,
"step": 616
},
{
"epoch": 0.10964503087653828,
"grad_norm": 0.4891494607879471,
"learning_rate": 3.901064559140935e-05,
"loss": 1.4773,
"step": 617
},
{
"epoch": 0.10982273757163802,
"grad_norm": 0.4812843053929585,
"learning_rate": 3.900713858313937e-05,
"loss": 1.4884,
"step": 618
},
{
"epoch": 0.11000044426673775,
"grad_norm": 0.48930824519155997,
"learning_rate": 3.900362552834117e-05,
"loss": 1.4456,
"step": 619
},
{
"epoch": 0.11017815096183749,
"grad_norm": 0.4676911145411502,
"learning_rate": 3.9000106428132304e-05,
"loss": 1.4631,
"step": 620
},
{
"epoch": 0.11035585765693723,
"grad_norm": 0.48125906587739264,
"learning_rate": 3.899658128363227e-05,
"loss": 1.4886,
"step": 621
},
{
"epoch": 0.11053356435203697,
"grad_norm": 0.49243107792650775,
"learning_rate": 3.8993050095962485e-05,
"loss": 1.4774,
"step": 622
},
{
"epoch": 0.1107112710471367,
"grad_norm": 0.4947129617234653,
"learning_rate": 3.8989512866246287e-05,
"loss": 1.4329,
"step": 623
},
{
"epoch": 0.11088897774223644,
"grad_norm": 0.4608759461683559,
"learning_rate": 3.898596959560893e-05,
"loss": 1.4492,
"step": 624
},
{
"epoch": 0.11106668443733618,
"grad_norm": 0.48616444719902874,
"learning_rate": 3.898242028517759e-05,
"loss": 1.4352,
"step": 625
},
{
"epoch": 0.11124439113243592,
"grad_norm": 0.5032073396014028,
"learning_rate": 3.897886493608139e-05,
"loss": 1.4466,
"step": 626
},
{
"epoch": 0.11142209782753565,
"grad_norm": 0.4864657727782443,
"learning_rate": 3.897530354945133e-05,
"loss": 1.5155,
"step": 627
},
{
"epoch": 0.11159980452263539,
"grad_norm": 0.49051682871149144,
"learning_rate": 3.897173612642036e-05,
"loss": 1.4409,
"step": 628
},
{
"epoch": 0.11177751121773513,
"grad_norm": 0.5127109813382803,
"learning_rate": 3.8968162668123367e-05,
"loss": 1.4775,
"step": 629
},
{
"epoch": 0.11195521791283487,
"grad_norm": 0.4760182786422667,
"learning_rate": 3.8964583175697107e-05,
"loss": 1.483,
"step": 630
},
{
"epoch": 0.1121329246079346,
"grad_norm": 0.46780942067574105,
"learning_rate": 3.8960997650280286e-05,
"loss": 1.4843,
"step": 631
},
{
"epoch": 0.11231063130303434,
"grad_norm": 0.4635158944278821,
"learning_rate": 3.8957406093013546e-05,
"loss": 1.4835,
"step": 632
},
{
"epoch": 0.11248833799813408,
"grad_norm": 0.4810605937932081,
"learning_rate": 3.8953808505039405e-05,
"loss": 1.4823,
"step": 633
},
{
"epoch": 0.11266604469323382,
"grad_norm": 0.4654089208927498,
"learning_rate": 3.895020488750235e-05,
"loss": 1.4789,
"step": 634
},
{
"epoch": 0.11284375138833355,
"grad_norm": 0.4617790707505816,
"learning_rate": 3.894659524154874e-05,
"loss": 1.4685,
"step": 635
},
{
"epoch": 0.1130214580834333,
"grad_norm": 0.47207831609776124,
"learning_rate": 3.894297956832688e-05,
"loss": 1.5068,
"step": 636
},
{
"epoch": 0.11319916477853303,
"grad_norm": 0.5021079363313464,
"learning_rate": 3.8939357868986975e-05,
"loss": 1.4925,
"step": 637
},
{
"epoch": 0.11337687147363278,
"grad_norm": 0.596480787672527,
"learning_rate": 3.8935730144681165e-05,
"loss": 1.5098,
"step": 638
},
{
"epoch": 0.1135545781687325,
"grad_norm": 0.4773589471632914,
"learning_rate": 3.8932096396563494e-05,
"loss": 1.4503,
"step": 639
},
{
"epoch": 0.11373228486383224,
"grad_norm": 0.49324858223107665,
"learning_rate": 3.8928456625789925e-05,
"loss": 1.4881,
"step": 640
},
{
"epoch": 0.11390999155893199,
"grad_norm": 0.48414500363029345,
"learning_rate": 3.892481083351833e-05,
"loss": 1.4751,
"step": 641
},
{
"epoch": 0.11408769825403173,
"grad_norm": 0.4981660322552098,
"learning_rate": 3.8921159020908524e-05,
"loss": 1.486,
"step": 642
},
{
"epoch": 0.11426540494913145,
"grad_norm": 0.4930120438440902,
"learning_rate": 3.89175011891222e-05,
"loss": 1.4597,
"step": 643
},
{
"epoch": 0.1144431116442312,
"grad_norm": 0.5036548254612124,
"learning_rate": 3.8913837339322986e-05,
"loss": 1.4613,
"step": 644
},
{
"epoch": 0.11462081833933094,
"grad_norm": 0.516111651924831,
"learning_rate": 3.8910167472676425e-05,
"loss": 1.4994,
"step": 645
},
{
"epoch": 0.11479852503443068,
"grad_norm": 0.48258343799305153,
"learning_rate": 3.890649159034997e-05,
"loss": 1.4712,
"step": 646
},
{
"epoch": 0.1149762317295304,
"grad_norm": 0.5306152457963272,
"learning_rate": 3.890280969351299e-05,
"loss": 1.5128,
"step": 647
},
{
"epoch": 0.11515393842463015,
"grad_norm": 0.4925000700925326,
"learning_rate": 3.889912178333676e-05,
"loss": 1.5119,
"step": 648
},
{
"epoch": 0.11533164511972989,
"grad_norm": 0.4980533760256093,
"learning_rate": 3.889542786099448e-05,
"loss": 1.4637,
"step": 649
},
{
"epoch": 0.11550935181482963,
"grad_norm": 0.5097124031719039,
"learning_rate": 3.889172792766125e-05,
"loss": 1.4967,
"step": 650
},
{
"epoch": 0.11568705850992936,
"grad_norm": 0.4869220749741633,
"learning_rate": 3.888802198451409e-05,
"loss": 1.4711,
"step": 651
},
{
"epoch": 0.1158647652050291,
"grad_norm": 0.5146028919363663,
"learning_rate": 3.888431003273193e-05,
"loss": 1.51,
"step": 652
},
{
"epoch": 0.11604247190012884,
"grad_norm": 0.4909919591307829,
"learning_rate": 3.888059207349562e-05,
"loss": 1.4735,
"step": 653
},
{
"epoch": 0.11622017859522858,
"grad_norm": 0.48748291351166484,
"learning_rate": 3.8876868107987905e-05,
"loss": 1.4578,
"step": 654
},
{
"epoch": 0.11639788529032831,
"grad_norm": 0.5190716263723744,
"learning_rate": 3.887313813739344e-05,
"loss": 1.4759,
"step": 655
},
{
"epoch": 0.11657559198542805,
"grad_norm": 0.4664148989865334,
"learning_rate": 3.886940216289882e-05,
"loss": 1.4471,
"step": 656
},
{
"epoch": 0.11675329868052779,
"grad_norm": 0.49847228933626675,
"learning_rate": 3.8865660185692506e-05,
"loss": 1.4735,
"step": 657
},
{
"epoch": 0.11693100537562753,
"grad_norm": 0.4681379549536594,
"learning_rate": 3.886191220696491e-05,
"loss": 1.4388,
"step": 658
},
{
"epoch": 0.11710871207072726,
"grad_norm": 0.49445940242462844,
"learning_rate": 3.885815822790833e-05,
"loss": 1.4803,
"step": 659
},
{
"epoch": 0.117286418765827,
"grad_norm": 0.5065811112795368,
"learning_rate": 3.885439824971697e-05,
"loss": 1.5397,
"step": 660
},
{
"epoch": 0.11746412546092674,
"grad_norm": 0.4784167370137502,
"learning_rate": 3.8850632273586944e-05,
"loss": 1.4616,
"step": 661
},
{
"epoch": 0.11764183215602649,
"grad_norm": 0.4787213183958859,
"learning_rate": 3.88468603007163e-05,
"loss": 1.4876,
"step": 662
},
{
"epoch": 0.11781953885112621,
"grad_norm": 0.4929963460524961,
"learning_rate": 3.884308233230496e-05,
"loss": 1.4667,
"step": 663
},
{
"epoch": 0.11799724554622595,
"grad_norm": 0.48673239626795956,
"learning_rate": 3.8839298369554777e-05,
"loss": 1.4663,
"step": 664
},
{
"epoch": 0.1181749522413257,
"grad_norm": 0.4774600572146014,
"learning_rate": 3.8835508413669485e-05,
"loss": 1.4666,
"step": 665
},
{
"epoch": 0.11835265893642544,
"grad_norm": 0.5336252086405456,
"learning_rate": 3.8831712465854754e-05,
"loss": 1.4537,
"step": 666
},
{
"epoch": 0.11853036563152516,
"grad_norm": 0.4727770652478497,
"learning_rate": 3.882791052731814e-05,
"loss": 1.4251,
"step": 667
},
{
"epoch": 0.1187080723266249,
"grad_norm": 0.48461117388773034,
"learning_rate": 3.8824102599269114e-05,
"loss": 1.4464,
"step": 668
},
{
"epoch": 0.11888577902172465,
"grad_norm": 0.47725727885791813,
"learning_rate": 3.8820288682919045e-05,
"loss": 1.4244,
"step": 669
},
{
"epoch": 0.11906348571682439,
"grad_norm": 0.48075091646627244,
"learning_rate": 3.881646877948122e-05,
"loss": 1.4171,
"step": 670
},
{
"epoch": 0.11924119241192412,
"grad_norm": 0.4749308149608574,
"learning_rate": 3.881264289017081e-05,
"loss": 1.457,
"step": 671
},
{
"epoch": 0.11941889910702386,
"grad_norm": 0.48032949429178895,
"learning_rate": 3.880881101620491e-05,
"loss": 1.4765,
"step": 672
},
{
"epoch": 0.1195966058021236,
"grad_norm": 0.4816146004845993,
"learning_rate": 3.8804973158802514e-05,
"loss": 1.4581,
"step": 673
},
{
"epoch": 0.11977431249722334,
"grad_norm": 0.4512891810145045,
"learning_rate": 3.880112931918451e-05,
"loss": 1.4652,
"step": 674
},
{
"epoch": 0.11995201919232307,
"grad_norm": 0.481979871710216,
"learning_rate": 3.87972794985737e-05,
"loss": 1.4824,
"step": 675
},
{
"epoch": 0.12012972588742281,
"grad_norm": 0.5382655661222737,
"learning_rate": 3.879342369819478e-05,
"loss": 1.5031,
"step": 676
},
{
"epoch": 0.12030743258252255,
"grad_norm": 0.4809045140163986,
"learning_rate": 3.878956191927436e-05,
"loss": 1.4648,
"step": 677
},
{
"epoch": 0.12048513927762229,
"grad_norm": 0.47168636122777535,
"learning_rate": 3.8785694163040934e-05,
"loss": 1.4673,
"step": 678
},
{
"epoch": 0.12066284597272202,
"grad_norm": 0.4670740591655531,
"learning_rate": 3.878182043072492e-05,
"loss": 1.491,
"step": 679
},
{
"epoch": 0.12084055266782176,
"grad_norm": 0.47995169421722605,
"learning_rate": 3.8777940723558606e-05,
"loss": 1.4869,
"step": 680
},
{
"epoch": 0.1210182593629215,
"grad_norm": 0.45979816135535656,
"learning_rate": 3.877405504277623e-05,
"loss": 1.4756,
"step": 681
},
{
"epoch": 0.12119596605802124,
"grad_norm": 0.4778885764775004,
"learning_rate": 3.8770163389613874e-05,
"loss": 1.4499,
"step": 682
},
{
"epoch": 0.12137367275312097,
"grad_norm": 0.46620669651475805,
"learning_rate": 3.8766265765309554e-05,
"loss": 1.4654,
"step": 683
},
{
"epoch": 0.12155137944822071,
"grad_norm": 0.47872776011103996,
"learning_rate": 3.876236217110318e-05,
"loss": 1.4975,
"step": 684
},
{
"epoch": 0.12172908614332045,
"grad_norm": 0.48214716191790824,
"learning_rate": 3.8758452608236565e-05,
"loss": 1.4751,
"step": 685
},
{
"epoch": 0.1219067928384202,
"grad_norm": 0.4742697201635599,
"learning_rate": 3.8754537077953395e-05,
"loss": 1.4567,
"step": 686
},
{
"epoch": 0.12208449953351992,
"grad_norm": 0.485460739872924,
"learning_rate": 3.8750615581499295e-05,
"loss": 1.4762,
"step": 687
},
{
"epoch": 0.12226220622861966,
"grad_norm": 0.4754027299875628,
"learning_rate": 3.874668812012175e-05,
"loss": 1.4595,
"step": 688
},
{
"epoch": 0.1224399129237194,
"grad_norm": 0.4520916680098931,
"learning_rate": 3.874275469507017e-05,
"loss": 1.4273,
"step": 689
},
{
"epoch": 0.12261761961881915,
"grad_norm": 0.4894174200677997,
"learning_rate": 3.873881530759585e-05,
"loss": 1.4762,
"step": 690
},
{
"epoch": 0.12279532631391887,
"grad_norm": 0.4421531852162167,
"learning_rate": 3.873486995895198e-05,
"loss": 1.4186,
"step": 691
},
{
"epoch": 0.12297303300901861,
"grad_norm": 0.4885666756036006,
"learning_rate": 3.873091865039365e-05,
"loss": 1.4875,
"step": 692
},
{
"epoch": 0.12315073970411836,
"grad_norm": 0.4598511480454751,
"learning_rate": 3.872696138317785e-05,
"loss": 1.4254,
"step": 693
},
{
"epoch": 0.1233284463992181,
"grad_norm": 0.484962786655213,
"learning_rate": 3.872299815856345e-05,
"loss": 1.4692,
"step": 694
},
{
"epoch": 0.12350615309431782,
"grad_norm": 0.47332036847914627,
"learning_rate": 3.871902897781124e-05,
"loss": 1.4536,
"step": 695
},
{
"epoch": 0.12368385978941757,
"grad_norm": 0.47727685335759695,
"learning_rate": 3.871505384218388e-05,
"loss": 1.4611,
"step": 696
},
{
"epoch": 0.12386156648451731,
"grad_norm": 0.4923640698229617,
"learning_rate": 3.871107275294595e-05,
"loss": 1.4527,
"step": 697
},
{
"epoch": 0.12403927317961705,
"grad_norm": 0.47127038970270313,
"learning_rate": 3.870708571136389e-05,
"loss": 1.4599,
"step": 698
},
{
"epoch": 0.12421697987471678,
"grad_norm": 0.4706198476806867,
"learning_rate": 3.870309271870607e-05,
"loss": 1.463,
"step": 699
},
{
"epoch": 0.12439468656981652,
"grad_norm": 0.463254800304209,
"learning_rate": 3.869909377624272e-05,
"loss": 1.4657,
"step": 700
},
{
"epoch": 0.12457239326491626,
"grad_norm": 0.4729136270227897,
"learning_rate": 3.8695088885246e-05,
"loss": 1.4317,
"step": 701
},
{
"epoch": 0.124750099960016,
"grad_norm": 0.486149741199063,
"learning_rate": 3.869107804698992e-05,
"loss": 1.4827,
"step": 702
},
{
"epoch": 0.12492780665511573,
"grad_norm": 0.48334721027187433,
"learning_rate": 3.868706126275041e-05,
"loss": 1.4612,
"step": 703
},
{
"epoch": 0.12510551335021547,
"grad_norm": 0.5047390411054135,
"learning_rate": 3.868303853380529e-05,
"loss": 1.4719,
"step": 704
},
{
"epoch": 0.1252832200453152,
"grad_norm": 0.4839027419547605,
"learning_rate": 3.867900986143427e-05,
"loss": 1.4563,
"step": 705
},
{
"epoch": 0.12546092674041495,
"grad_norm": 0.4966959343595079,
"learning_rate": 3.867497524691892e-05,
"loss": 1.5031,
"step": 706
},
{
"epoch": 0.12563863343551468,
"grad_norm": 0.5088174250986103,
"learning_rate": 3.867093469154275e-05,
"loss": 1.4565,
"step": 707
},
{
"epoch": 0.12581634013061443,
"grad_norm": 0.4654788932468796,
"learning_rate": 3.8666888196591144e-05,
"loss": 1.4477,
"step": 708
},
{
"epoch": 0.12599404682571416,
"grad_norm": 0.47580786595841335,
"learning_rate": 3.8662835763351345e-05,
"loss": 1.4067,
"step": 709
},
{
"epoch": 0.1261717535208139,
"grad_norm": 0.4701649671303287,
"learning_rate": 3.8658777393112524e-05,
"loss": 1.4196,
"step": 710
},
{
"epoch": 0.12634946021591364,
"grad_norm": 0.4785588984231743,
"learning_rate": 3.8654713087165725e-05,
"loss": 1.4896,
"step": 711
},
{
"epoch": 0.12652716691101337,
"grad_norm": 0.47466721773844667,
"learning_rate": 3.865064284680387e-05,
"loss": 1.5,
"step": 712
},
{
"epoch": 0.1267048736061131,
"grad_norm": 0.4746208560306717,
"learning_rate": 3.864656667332178e-05,
"loss": 1.4528,
"step": 713
},
{
"epoch": 0.12688258030121286,
"grad_norm": 0.4804157062493404,
"learning_rate": 3.864248456801618e-05,
"loss": 1.5158,
"step": 714
},
{
"epoch": 0.12706028699631258,
"grad_norm": 0.46831560596679445,
"learning_rate": 3.863839653218564e-05,
"loss": 1.4706,
"step": 715
},
{
"epoch": 0.12723799369141234,
"grad_norm": 0.4911192733065015,
"learning_rate": 3.8634302567130655e-05,
"loss": 1.4721,
"step": 716
},
{
"epoch": 0.12741570038651207,
"grad_norm": 0.46098084262758277,
"learning_rate": 3.8630202674153584e-05,
"loss": 1.417,
"step": 717
},
{
"epoch": 0.1275934070816118,
"grad_norm": 0.46675085827441604,
"learning_rate": 3.8626096854558694e-05,
"loss": 1.4037,
"step": 718
},
{
"epoch": 0.12777111377671155,
"grad_norm": 0.4848626701524921,
"learning_rate": 3.862198510965211e-05,
"loss": 1.4833,
"step": 719
},
{
"epoch": 0.12794882047181128,
"grad_norm": 0.4756416118624315,
"learning_rate": 3.861786744074186e-05,
"loss": 1.4421,
"step": 720
},
{
"epoch": 0.128126527166911,
"grad_norm": 0.5235270979092916,
"learning_rate": 3.861374384913786e-05,
"loss": 1.4352,
"step": 721
},
{
"epoch": 0.12830423386201076,
"grad_norm": 0.464690184599356,
"learning_rate": 3.860961433615189e-05,
"loss": 1.4953,
"step": 722
},
{
"epoch": 0.12848194055711049,
"grad_norm": 0.4940248580046805,
"learning_rate": 3.860547890309763e-05,
"loss": 1.4438,
"step": 723
},
{
"epoch": 0.12865964725221024,
"grad_norm": 0.5785734547707853,
"learning_rate": 3.8601337551290635e-05,
"loss": 1.5034,
"step": 724
},
{
"epoch": 0.12883735394730997,
"grad_norm": 0.4948015364285141,
"learning_rate": 3.859719028204836e-05,
"loss": 1.4879,
"step": 725
},
{
"epoch": 0.1290150606424097,
"grad_norm": 0.44873843870187885,
"learning_rate": 3.8593037096690115e-05,
"loss": 1.4079,
"step": 726
},
{
"epoch": 0.12919276733750945,
"grad_norm": 0.5484139875457255,
"learning_rate": 3.858887799653711e-05,
"loss": 1.4573,
"step": 727
},
{
"epoch": 0.12937047403260918,
"grad_norm": 0.46830512450085493,
"learning_rate": 3.858471298291244e-05,
"loss": 1.4681,
"step": 728
},
{
"epoch": 0.1295481807277089,
"grad_norm": 0.5108503374647384,
"learning_rate": 3.858054205714107e-05,
"loss": 1.4661,
"step": 729
},
{
"epoch": 0.12972588742280866,
"grad_norm": 0.4986482480983646,
"learning_rate": 3.857636522054984e-05,
"loss": 1.472,
"step": 730
},
{
"epoch": 0.1299035941179084,
"grad_norm": 0.48272812761667444,
"learning_rate": 3.857218247446749e-05,
"loss": 1.4635,
"step": 731
},
{
"epoch": 0.13008130081300814,
"grad_norm": 0.4828369681005407,
"learning_rate": 3.8567993820224634e-05,
"loss": 1.4283,
"step": 732
},
{
"epoch": 0.13025900750810787,
"grad_norm": 0.4760464397291005,
"learning_rate": 3.856379925915376e-05,
"loss": 1.4781,
"step": 733
},
{
"epoch": 0.1304367142032076,
"grad_norm": 0.5227822167836678,
"learning_rate": 3.855959879258923e-05,
"loss": 1.491,
"step": 734
},
{
"epoch": 0.13061442089830735,
"grad_norm": 0.4649924108993964,
"learning_rate": 3.855539242186729e-05,
"loss": 1.4787,
"step": 735
},
{
"epoch": 0.13079212759340708,
"grad_norm": 0.5020549475999919,
"learning_rate": 3.855118014832608e-05,
"loss": 1.4343,
"step": 736
},
{
"epoch": 0.1309698342885068,
"grad_norm": 0.4806439308832536,
"learning_rate": 3.854696197330559e-05,
"loss": 1.4459,
"step": 737
},
{
"epoch": 0.13114754098360656,
"grad_norm": 0.5081708563410391,
"learning_rate": 3.854273789814771e-05,
"loss": 1.4544,
"step": 738
},
{
"epoch": 0.1313252476787063,
"grad_norm": 0.49861624121844145,
"learning_rate": 3.853850792419618e-05,
"loss": 1.437,
"step": 739
},
{
"epoch": 0.13150295437380605,
"grad_norm": 0.5361281845642616,
"learning_rate": 3.853427205279665e-05,
"loss": 1.4544,
"step": 740
},
{
"epoch": 0.13168066106890577,
"grad_norm": 0.48222596386794553,
"learning_rate": 3.8530030285296635e-05,
"loss": 1.4566,
"step": 741
},
{
"epoch": 0.1318583677640055,
"grad_norm": 0.6825270708467718,
"learning_rate": 3.8525782623045513e-05,
"loss": 1.4758,
"step": 742
},
{
"epoch": 0.13203607445910526,
"grad_norm": 0.4787808650505015,
"learning_rate": 3.852152906739454e-05,
"loss": 1.4733,
"step": 743
},
{
"epoch": 0.13221378115420498,
"grad_norm": 0.494498927353553,
"learning_rate": 3.851726961969686e-05,
"loss": 1.4866,
"step": 744
},
{
"epoch": 0.1323914878493047,
"grad_norm": 0.46892323723667,
"learning_rate": 3.851300428130748e-05,
"loss": 1.4464,
"step": 745
},
{
"epoch": 0.13256919454440447,
"grad_norm": 0.47837887665046896,
"learning_rate": 3.8508733053583294e-05,
"loss": 1.4544,
"step": 746
},
{
"epoch": 0.1327469012395042,
"grad_norm": 0.47982373569009784,
"learning_rate": 3.8504455937883046e-05,
"loss": 1.4409,
"step": 747
},
{
"epoch": 0.13292460793460395,
"grad_norm": 0.46281753991981844,
"learning_rate": 3.850017293556737e-05,
"loss": 1.451,
"step": 748
},
{
"epoch": 0.13310231462970368,
"grad_norm": 0.48052054953115775,
"learning_rate": 3.849588404799877e-05,
"loss": 1.4923,
"step": 749
},
{
"epoch": 0.1332800213248034,
"grad_norm": 0.48298793919639504,
"learning_rate": 3.8491589276541626e-05,
"loss": 1.4932,
"step": 750
},
{
"epoch": 0.13345772801990316,
"grad_norm": 0.8342297631644411,
"learning_rate": 3.848728862256218e-05,
"loss": 1.438,
"step": 751
},
{
"epoch": 0.1336354347150029,
"grad_norm": 0.4926107491214109,
"learning_rate": 3.848298208742856e-05,
"loss": 1.5045,
"step": 752
},
{
"epoch": 0.13381314141010262,
"grad_norm": 0.5592357343802832,
"learning_rate": 3.847866967251075e-05,
"loss": 1.4806,
"step": 753
},
{
"epoch": 0.13399084810520237,
"grad_norm": 0.5373051737485756,
"learning_rate": 3.8474351379180606e-05,
"loss": 1.4745,
"step": 754
},
{
"epoch": 0.1341685548003021,
"grad_norm": 0.5019587654926602,
"learning_rate": 3.8470027208811866e-05,
"loss": 1.4693,
"step": 755
},
{
"epoch": 0.13434626149540185,
"grad_norm": 0.4731530404646903,
"learning_rate": 3.846569716278012e-05,
"loss": 1.4645,
"step": 756
},
{
"epoch": 0.13452396819050158,
"grad_norm": 0.5064035322039119,
"learning_rate": 3.846136124246285e-05,
"loss": 1.4771,
"step": 757
},
{
"epoch": 0.1347016748856013,
"grad_norm": 0.46921666896806546,
"learning_rate": 3.845701944923939e-05,
"loss": 1.3972,
"step": 758
},
{
"epoch": 0.13487938158070106,
"grad_norm": 0.4759417960453021,
"learning_rate": 3.8452671784490934e-05,
"loss": 1.4469,
"step": 759
},
{
"epoch": 0.1350570882758008,
"grad_norm": 0.46684492853786036,
"learning_rate": 3.844831824960057e-05,
"loss": 1.4982,
"step": 760
},
{
"epoch": 0.13523479497090052,
"grad_norm": 0.4730046597112984,
"learning_rate": 3.844395884595323e-05,
"loss": 1.4264,
"step": 761
},
{
"epoch": 0.13541250166600027,
"grad_norm": 0.4641460404586111,
"learning_rate": 3.8439593574935734e-05,
"loss": 1.4319,
"step": 762
},
{
"epoch": 0.1355902083611,
"grad_norm": 0.45861793155267555,
"learning_rate": 3.843522243793674e-05,
"loss": 1.4747,
"step": 763
},
{
"epoch": 0.13576791505619976,
"grad_norm": 0.4603304649890127,
"learning_rate": 3.8430845436346815e-05,
"loss": 1.4167,
"step": 764
},
{
"epoch": 0.13594562175129948,
"grad_norm": 0.4667481987097305,
"learning_rate": 3.842646257155834e-05,
"loss": 1.4764,
"step": 765
},
{
"epoch": 0.1361233284463992,
"grad_norm": 0.46307910234365796,
"learning_rate": 3.8422073844965596e-05,
"loss": 1.4584,
"step": 766
},
{
"epoch": 0.13630103514149897,
"grad_norm": 0.47099038972989193,
"learning_rate": 3.8417679257964717e-05,
"loss": 1.4516,
"step": 767
},
{
"epoch": 0.1364787418365987,
"grad_norm": 0.44767605123520193,
"learning_rate": 3.841327881195371e-05,
"loss": 1.4934,
"step": 768
},
{
"epoch": 0.13665644853169842,
"grad_norm": 0.4748517908296488,
"learning_rate": 3.840887250833243e-05,
"loss": 1.4289,
"step": 769
},
{
"epoch": 0.13683415522679818,
"grad_norm": 0.5646427886996436,
"learning_rate": 3.840446034850262e-05,
"loss": 1.5259,
"step": 770
},
{
"epoch": 0.1370118619218979,
"grad_norm": 0.48240875458142385,
"learning_rate": 3.8400042333867855e-05,
"loss": 1.4582,
"step": 771
},
{
"epoch": 0.13718956861699766,
"grad_norm": 0.4576174371237635,
"learning_rate": 3.8395618465833594e-05,
"loss": 1.4371,
"step": 772
},
{
"epoch": 0.1373672753120974,
"grad_norm": 0.4613616620324441,
"learning_rate": 3.839118874580715e-05,
"loss": 1.4183,
"step": 773
},
{
"epoch": 0.13754498200719711,
"grad_norm": 0.46537858429498186,
"learning_rate": 3.838675317519771e-05,
"loss": 1.4711,
"step": 774
},
{
"epoch": 0.13772268870229687,
"grad_norm": 0.5041265832128041,
"learning_rate": 3.838231175541631e-05,
"loss": 1.4433,
"step": 775
},
{
"epoch": 0.1379003953973966,
"grad_norm": 0.46750320445645044,
"learning_rate": 3.8377864487875845e-05,
"loss": 1.4925,
"step": 776
},
{
"epoch": 0.13807810209249632,
"grad_norm": 0.4663420347719685,
"learning_rate": 3.837341137399107e-05,
"loss": 1.5048,
"step": 777
},
{
"epoch": 0.13825580878759608,
"grad_norm": 0.47174908380768454,
"learning_rate": 3.8368952415178613e-05,
"loss": 1.4668,
"step": 778
},
{
"epoch": 0.1384335154826958,
"grad_norm": 0.4635859740523269,
"learning_rate": 3.8364487612856946e-05,
"loss": 1.4585,
"step": 779
},
{
"epoch": 0.13861122217779556,
"grad_norm": 0.4698214916407571,
"learning_rate": 3.8360016968446415e-05,
"loss": 1.4965,
"step": 780
},
{
"epoch": 0.1387889288728953,
"grad_norm": 0.4767542746971557,
"learning_rate": 3.835554048336921e-05,
"loss": 1.4142,
"step": 781
},
{
"epoch": 0.13896663556799502,
"grad_norm": 0.44238855541558036,
"learning_rate": 3.835105815904938e-05,
"loss": 1.422,
"step": 782
},
{
"epoch": 0.13914434226309477,
"grad_norm": 0.8546282371706397,
"learning_rate": 3.8346569996912844e-05,
"loss": 1.4023,
"step": 783
},
{
"epoch": 0.1393220489581945,
"grad_norm": 0.4797489188095591,
"learning_rate": 3.834207599838737e-05,
"loss": 1.4715,
"step": 784
},
{
"epoch": 0.13949975565329423,
"grad_norm": 0.4732427504828133,
"learning_rate": 3.833757616490259e-05,
"loss": 1.454,
"step": 785
},
{
"epoch": 0.13967746234839398,
"grad_norm": 0.4821878860160217,
"learning_rate": 3.833307049788996e-05,
"loss": 1.483,
"step": 786
},
{
"epoch": 0.1398551690434937,
"grad_norm": 0.46851750784285123,
"learning_rate": 3.832855899878285e-05,
"loss": 1.4391,
"step": 787
},
{
"epoch": 0.14003287573859347,
"grad_norm": 0.49217111284837145,
"learning_rate": 3.832404166901644e-05,
"loss": 1.4721,
"step": 788
},
{
"epoch": 0.1402105824336932,
"grad_norm": 0.4795026852934791,
"learning_rate": 3.831951851002777e-05,
"loss": 1.4447,
"step": 789
},
{
"epoch": 0.14038828912879292,
"grad_norm": 0.4881146432201372,
"learning_rate": 3.831498952325575e-05,
"loss": 1.4448,
"step": 790
},
{
"epoch": 0.14056599582389268,
"grad_norm": 0.4752831154747966,
"learning_rate": 3.831045471014113e-05,
"loss": 1.444,
"step": 791
},
{
"epoch": 0.1407437025189924,
"grad_norm": 0.46518916612474115,
"learning_rate": 3.8305914072126536e-05,
"loss": 1.4239,
"step": 792
},
{
"epoch": 0.14092140921409213,
"grad_norm": 0.49907781359026854,
"learning_rate": 3.8301367610656405e-05,
"loss": 1.4383,
"step": 793
},
{
"epoch": 0.14109911590919189,
"grad_norm": 0.4610687490576072,
"learning_rate": 3.8296815327177064e-05,
"loss": 1.453,
"step": 794
},
{
"epoch": 0.1412768226042916,
"grad_norm": 0.4920091684315072,
"learning_rate": 3.829225722313669e-05,
"loss": 1.4525,
"step": 795
},
{
"epoch": 0.14145452929939137,
"grad_norm": 0.5015566664857107,
"learning_rate": 3.828769329998528e-05,
"loss": 1.4898,
"step": 796
},
{
"epoch": 0.1416322359944911,
"grad_norm": 0.44330588750190814,
"learning_rate": 3.8283123559174725e-05,
"loss": 1.4197,
"step": 797
},
{
"epoch": 0.14180994268959082,
"grad_norm": 0.4855755250548181,
"learning_rate": 3.8278548002158735e-05,
"loss": 1.4697,
"step": 798
},
{
"epoch": 0.14198764938469058,
"grad_norm": 0.44944607158425637,
"learning_rate": 3.827396663039288e-05,
"loss": 1.4922,
"step": 799
},
{
"epoch": 0.1421653560797903,
"grad_norm": 0.47896084368785063,
"learning_rate": 3.826937944533458e-05,
"loss": 1.4076,
"step": 800
},
{
"epoch": 0.14234306277489003,
"grad_norm": 0.4513565896420756,
"learning_rate": 3.826478644844311e-05,
"loss": 1.5012,
"step": 801
},
{
"epoch": 0.1425207694699898,
"grad_norm": 0.5022425674772737,
"learning_rate": 3.826018764117958e-05,
"loss": 1.4363,
"step": 802
},
{
"epoch": 0.14269847616508952,
"grad_norm": 0.4733673466302544,
"learning_rate": 3.8255583025006974e-05,
"loss": 1.4915,
"step": 803
},
{
"epoch": 0.14287618286018927,
"grad_norm": 0.5033670062385162,
"learning_rate": 3.825097260139009e-05,
"loss": 1.4454,
"step": 804
},
{
"epoch": 0.143053889555289,
"grad_norm": 0.4476473659393791,
"learning_rate": 3.82463563717956e-05,
"loss": 1.4116,
"step": 805
},
{
"epoch": 0.14323159625038873,
"grad_norm": 0.48689233292334644,
"learning_rate": 3.8241734337692e-05,
"loss": 1.3993,
"step": 806
},
{
"epoch": 0.14340930294548848,
"grad_norm": 0.4621394785576905,
"learning_rate": 3.8237106500549665e-05,
"loss": 1.4406,
"step": 807
},
{
"epoch": 0.1435870096405882,
"grad_norm": 0.4634360239576694,
"learning_rate": 3.823247286184079e-05,
"loss": 1.457,
"step": 808
},
{
"epoch": 0.14376471633568794,
"grad_norm": 0.4862803784024682,
"learning_rate": 3.822783342303942e-05,
"loss": 1.5153,
"step": 809
},
{
"epoch": 0.1439424230307877,
"grad_norm": 0.46200466884813207,
"learning_rate": 3.822318818562145e-05,
"loss": 1.4679,
"step": 810
},
{
"epoch": 0.14412012972588742,
"grad_norm": 0.4656559467798048,
"learning_rate": 3.821853715106461e-05,
"loss": 1.4144,
"step": 811
},
{
"epoch": 0.14429783642098717,
"grad_norm": 0.48480255856074,
"learning_rate": 3.8213880320848486e-05,
"loss": 1.4704,
"step": 812
},
{
"epoch": 0.1444755431160869,
"grad_norm": 0.45722251270153513,
"learning_rate": 3.8209217696454504e-05,
"loss": 1.464,
"step": 813
},
{
"epoch": 0.14465324981118663,
"grad_norm": 0.4457067628912401,
"learning_rate": 3.820454927936594e-05,
"loss": 1.396,
"step": 814
},
{
"epoch": 0.14483095650628638,
"grad_norm": 0.4522619259939123,
"learning_rate": 3.819987507106789e-05,
"loss": 1.4419,
"step": 815
},
{
"epoch": 0.1450086632013861,
"grad_norm": 0.4397569062561208,
"learning_rate": 3.8195195073047325e-05,
"loss": 1.3881,
"step": 816
},
{
"epoch": 0.14518636989648584,
"grad_norm": 0.46122250708272317,
"learning_rate": 3.819050928679303e-05,
"loss": 1.4499,
"step": 817
},
{
"epoch": 0.1453640765915856,
"grad_norm": 0.46068741531869645,
"learning_rate": 3.818581771379563e-05,
"loss": 1.4477,
"step": 818
},
{
"epoch": 0.14554178328668532,
"grad_norm": 0.4480079204762609,
"learning_rate": 3.818112035554763e-05,
"loss": 1.4869,
"step": 819
},
{
"epoch": 0.14571948998178508,
"grad_norm": 0.4682670983238948,
"learning_rate": 3.8176417213543324e-05,
"loss": 1.4965,
"step": 820
},
{
"epoch": 0.1458971966768848,
"grad_norm": 0.46488736553438076,
"learning_rate": 3.817170828927889e-05,
"loss": 1.4622,
"step": 821
},
{
"epoch": 0.14607490337198453,
"grad_norm": 0.46512381516219214,
"learning_rate": 3.816699358425231e-05,
"loss": 1.4083,
"step": 822
},
{
"epoch": 0.1462526100670843,
"grad_norm": 0.4644074091508032,
"learning_rate": 3.8162273099963425e-05,
"loss": 1.4152,
"step": 823
},
{
"epoch": 0.14643031676218402,
"grad_norm": 0.4484534198149306,
"learning_rate": 3.81575468379139e-05,
"loss": 1.47,
"step": 824
},
{
"epoch": 0.14660802345728374,
"grad_norm": 0.456874479702208,
"learning_rate": 3.815281479960727e-05,
"loss": 1.4627,
"step": 825
},
{
"epoch": 0.1467857301523835,
"grad_norm": 0.45485530826228243,
"learning_rate": 3.814807698654887e-05,
"loss": 1.4599,
"step": 826
},
{
"epoch": 0.14696343684748323,
"grad_norm": 0.46145854600600467,
"learning_rate": 3.814333340024589e-05,
"loss": 1.4061,
"step": 827
},
{
"epoch": 0.14714114354258298,
"grad_norm": 0.5392392781062655,
"learning_rate": 3.813858404220736e-05,
"loss": 1.4476,
"step": 828
},
{
"epoch": 0.1473188502376827,
"grad_norm": 0.48171751933860474,
"learning_rate": 3.8133828913944126e-05,
"loss": 1.4055,
"step": 829
},
{
"epoch": 0.14749655693278244,
"grad_norm": 0.45027030983766647,
"learning_rate": 3.81290680169689e-05,
"loss": 1.4779,
"step": 830
},
{
"epoch": 0.1476742636278822,
"grad_norm": 0.4821003139127576,
"learning_rate": 3.81243013527962e-05,
"loss": 1.4798,
"step": 831
},
{
"epoch": 0.14785197032298192,
"grad_norm": 0.4647031207552602,
"learning_rate": 3.81195289229424e-05,
"loss": 1.487,
"step": 832
},
{
"epoch": 0.14802967701808165,
"grad_norm": 0.48333009636842367,
"learning_rate": 3.8114750728925695e-05,
"loss": 1.4231,
"step": 833
},
{
"epoch": 0.1482073837131814,
"grad_norm": 0.45138963902111956,
"learning_rate": 3.810996677226612e-05,
"loss": 1.4739,
"step": 834
},
{
"epoch": 0.14838509040828113,
"grad_norm": 0.5056074750741885,
"learning_rate": 3.810517705448554e-05,
"loss": 1.4835,
"step": 835
},
{
"epoch": 0.14856279710338088,
"grad_norm": 0.45052344387048043,
"learning_rate": 3.8100381577107664e-05,
"loss": 1.4572,
"step": 836
},
{
"epoch": 0.1487405037984806,
"grad_norm": 0.5069433730246271,
"learning_rate": 3.809558034165801e-05,
"loss": 1.4479,
"step": 837
},
{
"epoch": 0.14891821049358034,
"grad_norm": 0.47697564989465263,
"learning_rate": 3.8090773349663946e-05,
"loss": 1.5111,
"step": 838
},
{
"epoch": 0.1490959171886801,
"grad_norm": 0.4601949439717773,
"learning_rate": 3.8085960602654663e-05,
"loss": 1.4214,
"step": 839
},
{
"epoch": 0.14927362388377982,
"grad_norm": 0.49498081221877455,
"learning_rate": 3.80811421021612e-05,
"loss": 1.4762,
"step": 840
},
{
"epoch": 0.14945133057887955,
"grad_norm": 0.4742796273607144,
"learning_rate": 3.8076317849716395e-05,
"loss": 1.4348,
"step": 841
},
{
"epoch": 0.1496290372739793,
"grad_norm": 0.46431709448542696,
"learning_rate": 3.807148784685494e-05,
"loss": 1.4968,
"step": 842
},
{
"epoch": 0.14980674396907903,
"grad_norm": 0.4711103707349645,
"learning_rate": 3.8066652095113365e-05,
"loss": 1.523,
"step": 843
},
{
"epoch": 0.1499844506641788,
"grad_norm": 0.5277202854997737,
"learning_rate": 3.806181059602999e-05,
"loss": 1.4628,
"step": 844
},
{
"epoch": 0.15016215735927851,
"grad_norm": 0.4649839036640465,
"learning_rate": 3.805696335114499e-05,
"loss": 1.4572,
"step": 845
},
{
"epoch": 0.15033986405437824,
"grad_norm": 0.5041524279251177,
"learning_rate": 3.805211036200038e-05,
"loss": 1.414,
"step": 846
},
{
"epoch": 0.150517570749478,
"grad_norm": 0.468876695230687,
"learning_rate": 3.804725163013998e-05,
"loss": 1.4349,
"step": 847
},
{
"epoch": 0.15069527744457772,
"grad_norm": 0.4564204598836714,
"learning_rate": 3.804238715710944e-05,
"loss": 1.4192,
"step": 848
},
{
"epoch": 0.15087298413967745,
"grad_norm": 0.4778879819877622,
"learning_rate": 3.8037516944456244e-05,
"loss": 1.4314,
"step": 849
},
{
"epoch": 0.1510506908347772,
"grad_norm": 0.45253071404383505,
"learning_rate": 3.80326409937297e-05,
"loss": 1.4853,
"step": 850
},
{
"epoch": 0.15122839752987693,
"grad_norm": 0.5029937246677569,
"learning_rate": 3.8027759306480925e-05,
"loss": 1.3804,
"step": 851
},
{
"epoch": 0.1514061042249767,
"grad_norm": 0.4716452078420153,
"learning_rate": 3.80228718842629e-05,
"loss": 1.4146,
"step": 852
},
{
"epoch": 0.15158381092007642,
"grad_norm": 0.4780526006310195,
"learning_rate": 3.8017978728630386e-05,
"loss": 1.4566,
"step": 853
},
{
"epoch": 0.15176151761517614,
"grad_norm": 0.4974188750193863,
"learning_rate": 3.8013079841139996e-05,
"loss": 1.4883,
"step": 854
},
{
"epoch": 0.1519392243102759,
"grad_norm": 0.5080336045263969,
"learning_rate": 3.8008175223350165e-05,
"loss": 1.4643,
"step": 855
},
{
"epoch": 0.15211693100537563,
"grad_norm": 0.48418462740980084,
"learning_rate": 3.800326487682113e-05,
"loss": 1.4401,
"step": 856
},
{
"epoch": 0.15229463770047535,
"grad_norm": 0.42506743418835535,
"learning_rate": 3.7998348803114976e-05,
"loss": 1.4679,
"step": 857
},
{
"epoch": 0.1524723443955751,
"grad_norm": 0.4621213968115969,
"learning_rate": 3.7993427003795583e-05,
"loss": 1.4753,
"step": 858
},
{
"epoch": 0.15265005109067484,
"grad_norm": 0.44699278495039435,
"learning_rate": 3.798849948042869e-05,
"loss": 1.4163,
"step": 859
},
{
"epoch": 0.1528277577857746,
"grad_norm": 0.47779446315566365,
"learning_rate": 3.798356623458182e-05,
"loss": 1.4594,
"step": 860
},
{
"epoch": 0.15300546448087432,
"grad_norm": 0.44716553214735444,
"learning_rate": 3.797862726782433e-05,
"loss": 1.4529,
"step": 861
},
{
"epoch": 0.15318317117597405,
"grad_norm": 0.5395637078935029,
"learning_rate": 3.797368258172741e-05,
"loss": 1.4794,
"step": 862
},
{
"epoch": 0.1533608778710738,
"grad_norm": 0.4650695455494772,
"learning_rate": 3.7968732177864046e-05,
"loss": 1.4138,
"step": 863
},
{
"epoch": 0.15353858456617353,
"grad_norm": 0.45800662640311324,
"learning_rate": 3.796377605780906e-05,
"loss": 1.4687,
"step": 864
},
{
"epoch": 0.15371629126127326,
"grad_norm": 0.5029807596406858,
"learning_rate": 3.7958814223139085e-05,
"loss": 1.4731,
"step": 865
},
{
"epoch": 0.153893997956373,
"grad_norm": 0.4377617596055028,
"learning_rate": 3.795384667543257e-05,
"loss": 1.4219,
"step": 866
},
{
"epoch": 0.15407170465147274,
"grad_norm": 0.48072694757059053,
"learning_rate": 3.79488734162698e-05,
"loss": 1.4221,
"step": 867
},
{
"epoch": 0.1542494113465725,
"grad_norm": 0.44471872525401335,
"learning_rate": 3.794389444723285e-05,
"loss": 1.4605,
"step": 868
},
{
"epoch": 0.15442711804167222,
"grad_norm": 0.49190879287502176,
"learning_rate": 3.7938909769905625e-05,
"loss": 1.473,
"step": 869
},
{
"epoch": 0.15460482473677195,
"grad_norm": 0.5500326732539106,
"learning_rate": 3.7933919385873846e-05,
"loss": 1.4798,
"step": 870
},
{
"epoch": 0.1547825314318717,
"grad_norm": 0.4791971289087994,
"learning_rate": 3.7928923296725045e-05,
"loss": 1.4586,
"step": 871
},
{
"epoch": 0.15496023812697143,
"grad_norm": 0.4805601836856273,
"learning_rate": 3.792392150404858e-05,
"loss": 1.433,
"step": 872
},
{
"epoch": 0.15513794482207116,
"grad_norm": 0.4594602519219576,
"learning_rate": 3.79189140094356e-05,
"loss": 1.4764,
"step": 873
},
{
"epoch": 0.15531565151717092,
"grad_norm": 0.47751622766956187,
"learning_rate": 3.791390081447911e-05,
"loss": 1.4702,
"step": 874
},
{
"epoch": 0.15549335821227064,
"grad_norm": 0.49781862893380596,
"learning_rate": 3.790888192077387e-05,
"loss": 1.4473,
"step": 875
},
{
"epoch": 0.1556710649073704,
"grad_norm": 0.5351244148304558,
"learning_rate": 3.7903857329916504e-05,
"loss": 1.4574,
"step": 876
},
{
"epoch": 0.15584877160247013,
"grad_norm": 0.49666160157996647,
"learning_rate": 3.7898827043505426e-05,
"loss": 1.4754,
"step": 877
},
{
"epoch": 0.15602647829756985,
"grad_norm": 0.489587695101352,
"learning_rate": 3.789379106314086e-05,
"loss": 1.4436,
"step": 878
},
{
"epoch": 0.1562041849926696,
"grad_norm": 0.4534051429886331,
"learning_rate": 3.788874939042485e-05,
"loss": 1.4101,
"step": 879
},
{
"epoch": 0.15638189168776934,
"grad_norm": 0.45698994509829893,
"learning_rate": 3.7883702026961245e-05,
"loss": 1.4248,
"step": 880
},
{
"epoch": 0.15655959838286906,
"grad_norm": 0.461169810998765,
"learning_rate": 3.787864897435571e-05,
"loss": 1.4569,
"step": 881
},
{
"epoch": 0.15673730507796882,
"grad_norm": 0.448056811391795,
"learning_rate": 3.787359023421571e-05,
"loss": 1.4106,
"step": 882
},
{
"epoch": 0.15691501177306855,
"grad_norm": 0.4756438256157798,
"learning_rate": 3.786852580815054e-05,
"loss": 1.4463,
"step": 883
},
{
"epoch": 0.1570927184681683,
"grad_norm": 0.44061679566657647,
"learning_rate": 3.786345569777126e-05,
"loss": 1.4134,
"step": 884
},
{
"epoch": 0.15727042516326803,
"grad_norm": 0.45289003864472993,
"learning_rate": 3.785837990469079e-05,
"loss": 1.4622,
"step": 885
},
{
"epoch": 0.15744813185836776,
"grad_norm": 0.4516911841449598,
"learning_rate": 3.7853298430523835e-05,
"loss": 1.4394,
"step": 886
},
{
"epoch": 0.1576258385534675,
"grad_norm": 0.4330056439097637,
"learning_rate": 3.78482112768869e-05,
"loss": 1.445,
"step": 887
},
{
"epoch": 0.15780354524856724,
"grad_norm": 0.46495280554456486,
"learning_rate": 3.7843118445398316e-05,
"loss": 1.4395,
"step": 888
},
{
"epoch": 0.15798125194366697,
"grad_norm": 0.47061814298731763,
"learning_rate": 3.783801993767819e-05,
"loss": 1.4203,
"step": 889
},
{
"epoch": 0.15815895863876672,
"grad_norm": 0.4597537304502784,
"learning_rate": 3.783291575534847e-05,
"loss": 1.465,
"step": 890
},
{
"epoch": 0.15833666533386645,
"grad_norm": 0.43418252891773623,
"learning_rate": 3.7827805900032874e-05,
"loss": 1.4429,
"step": 891
},
{
"epoch": 0.1585143720289662,
"grad_norm": 0.4450135274809499,
"learning_rate": 3.7822690373356964e-05,
"loss": 1.4793,
"step": 892
},
{
"epoch": 0.15869207872406593,
"grad_norm": 0.44535180112607203,
"learning_rate": 3.781756917694807e-05,
"loss": 1.442,
"step": 893
},
{
"epoch": 0.15886978541916566,
"grad_norm": 0.43179849795393616,
"learning_rate": 3.781244231243535e-05,
"loss": 1.4338,
"step": 894
},
{
"epoch": 0.15904749211426542,
"grad_norm": 0.4341998841076116,
"learning_rate": 3.780730978144975e-05,
"loss": 1.4262,
"step": 895
},
{
"epoch": 0.15922519880936514,
"grad_norm": 0.4473673502459626,
"learning_rate": 3.780217158562403e-05,
"loss": 1.4608,
"step": 896
},
{
"epoch": 0.15940290550446487,
"grad_norm": 0.4372258657908458,
"learning_rate": 3.779702772659274e-05,
"loss": 1.4911,
"step": 897
},
{
"epoch": 0.15958061219956463,
"grad_norm": 0.4364679346292376,
"learning_rate": 3.7791878205992246e-05,
"loss": 1.4199,
"step": 898
},
{
"epoch": 0.15975831889466435,
"grad_norm": 0.43223688529570037,
"learning_rate": 3.77867230254607e-05,
"loss": 1.421,
"step": 899
},
{
"epoch": 0.1599360255897641,
"grad_norm": 0.4325282271197267,
"learning_rate": 3.7781562186638066e-05,
"loss": 1.4587,
"step": 900
},
{
"epoch": 0.16011373228486384,
"grad_norm": 0.45249174587018237,
"learning_rate": 3.7776395691166104e-05,
"loss": 1.4315,
"step": 901
},
{
"epoch": 0.16029143897996356,
"grad_norm": 0.4334439906574456,
"learning_rate": 3.777122354068837e-05,
"loss": 1.4377,
"step": 902
},
{
"epoch": 0.16046914567506332,
"grad_norm": 0.45028642375923883,
"learning_rate": 3.7766045736850224e-05,
"loss": 1.44,
"step": 903
},
{
"epoch": 0.16064685237016305,
"grad_norm": 0.441543215144827,
"learning_rate": 3.7760862281298824e-05,
"loss": 1.4084,
"step": 904
},
{
"epoch": 0.16082455906526277,
"grad_norm": 0.4376553952680813,
"learning_rate": 3.775567317568313e-05,
"loss": 1.4699,
"step": 905
},
{
"epoch": 0.16100226576036253,
"grad_norm": 0.4578896627589206,
"learning_rate": 3.7750478421653886e-05,
"loss": 1.4969,
"step": 906
},
{
"epoch": 0.16117997245546226,
"grad_norm": 0.44576370443197105,
"learning_rate": 3.774527802086364e-05,
"loss": 1.451,
"step": 907
},
{
"epoch": 0.161357679150562,
"grad_norm": 0.4724398404997984,
"learning_rate": 3.7740071974966746e-05,
"loss": 1.4196,
"step": 908
},
{
"epoch": 0.16153538584566174,
"grad_norm": 0.45698680754269466,
"learning_rate": 3.7734860285619334e-05,
"loss": 1.4462,
"step": 909
},
{
"epoch": 0.16171309254076147,
"grad_norm": 0.46401310207237945,
"learning_rate": 3.7729642954479355e-05,
"loss": 1.4606,
"step": 910
},
{
"epoch": 0.16189079923586122,
"grad_norm": 0.44426741105272494,
"learning_rate": 3.772441998320652e-05,
"loss": 1.4302,
"step": 911
},
{
"epoch": 0.16206850593096095,
"grad_norm": 0.47049303554504585,
"learning_rate": 3.7719191373462375e-05,
"loss": 1.4641,
"step": 912
},
{
"epoch": 0.16224621262606068,
"grad_norm": 0.4431301545601661,
"learning_rate": 3.771395712691022e-05,
"loss": 1.4069,
"step": 913
},
{
"epoch": 0.16242391932116043,
"grad_norm": 0.444746728235462,
"learning_rate": 3.7708717245215185e-05,
"loss": 1.4228,
"step": 914
},
{
"epoch": 0.16260162601626016,
"grad_norm": 0.4699488604702284,
"learning_rate": 3.770347173004417e-05,
"loss": 1.4431,
"step": 915
},
{
"epoch": 0.16277933271135991,
"grad_norm": 0.44792816984133355,
"learning_rate": 3.769822058306586e-05,
"loss": 1.4583,
"step": 916
},
{
"epoch": 0.16295703940645964,
"grad_norm": 0.6107426975985023,
"learning_rate": 3.769296380595076e-05,
"loss": 1.4542,
"step": 917
},
{
"epoch": 0.16313474610155937,
"grad_norm": 0.4444313078418792,
"learning_rate": 3.7687701400371133e-05,
"loss": 1.4678,
"step": 918
},
{
"epoch": 0.16331245279665912,
"grad_norm": 0.4753070079809744,
"learning_rate": 3.768243336800106e-05,
"loss": 1.4192,
"step": 919
},
{
"epoch": 0.16349015949175885,
"grad_norm": 0.4654275681939166,
"learning_rate": 3.7677159710516403e-05,
"loss": 1.4621,
"step": 920
},
{
"epoch": 0.16366786618685858,
"grad_norm": 0.4426709979809801,
"learning_rate": 3.767188042959481e-05,
"loss": 1.4553,
"step": 921
},
{
"epoch": 0.16384557288195833,
"grad_norm": 0.4513614826590076,
"learning_rate": 3.766659552691572e-05,
"loss": 1.4666,
"step": 922
},
{
"epoch": 0.16402327957705806,
"grad_norm": 0.45907242588791797,
"learning_rate": 3.766130500416035e-05,
"loss": 1.5026,
"step": 923
},
{
"epoch": 0.16420098627215782,
"grad_norm": 0.44527853762004493,
"learning_rate": 3.765600886301173e-05,
"loss": 1.4369,
"step": 924
},
{
"epoch": 0.16437869296725754,
"grad_norm": 0.4393309013728393,
"learning_rate": 3.765070710515465e-05,
"loss": 1.4056,
"step": 925
},
{
"epoch": 0.16455639966235727,
"grad_norm": 0.4414334268482012,
"learning_rate": 3.76453997322757e-05,
"loss": 1.4311,
"step": 926
},
{
"epoch": 0.16473410635745703,
"grad_norm": 0.464621520080213,
"learning_rate": 3.764008674606326e-05,
"loss": 1.4279,
"step": 927
},
{
"epoch": 0.16491181305255675,
"grad_norm": 0.43379407526783853,
"learning_rate": 3.7634768148207496e-05,
"loss": 1.4558,
"step": 928
},
{
"epoch": 0.16508951974765648,
"grad_norm": 0.47217359400765396,
"learning_rate": 3.7629443940400346e-05,
"loss": 1.4639,
"step": 929
},
{
"epoch": 0.16526722644275624,
"grad_norm": 0.4275303365467779,
"learning_rate": 3.7624114124335535e-05,
"loss": 1.4242,
"step": 930
},
{
"epoch": 0.16544493313785597,
"grad_norm": 0.4757647231499221,
"learning_rate": 3.761877870170859e-05,
"loss": 1.4421,
"step": 931
},
{
"epoch": 0.16562263983295572,
"grad_norm": 0.4461871194650327,
"learning_rate": 3.76134376742168e-05,
"loss": 1.423,
"step": 932
},
{
"epoch": 0.16580034652805545,
"grad_norm": 0.4489193412881358,
"learning_rate": 3.760809104355926e-05,
"loss": 1.4265,
"step": 933
},
{
"epoch": 0.16597805322315518,
"grad_norm": 0.4534302413804035,
"learning_rate": 3.760273881143681e-05,
"loss": 1.4689,
"step": 934
},
{
"epoch": 0.16615575991825493,
"grad_norm": 0.4839990895196798,
"learning_rate": 3.759738097955212e-05,
"loss": 1.4845,
"step": 935
},
{
"epoch": 0.16633346661335466,
"grad_norm": 0.431423542466281,
"learning_rate": 3.75920175496096e-05,
"loss": 1.3853,
"step": 936
},
{
"epoch": 0.16651117330845439,
"grad_norm": 0.44585304641125434,
"learning_rate": 3.7586648523315476e-05,
"loss": 1.4457,
"step": 937
},
{
"epoch": 0.16668888000355414,
"grad_norm": 0.4400262243540558,
"learning_rate": 3.758127390237771e-05,
"loss": 1.4501,
"step": 938
},
{
"epoch": 0.16686658669865387,
"grad_norm": 0.4734813254407016,
"learning_rate": 3.757589368850609e-05,
"loss": 1.4234,
"step": 939
},
{
"epoch": 0.16704429339375362,
"grad_norm": 0.4217820656233788,
"learning_rate": 3.757050788341216e-05,
"loss": 1.4375,
"step": 940
},
{
"epoch": 0.16722200008885335,
"grad_norm": 0.46133486720822525,
"learning_rate": 3.756511648880925e-05,
"loss": 1.4453,
"step": 941
},
{
"epoch": 0.16739970678395308,
"grad_norm": 0.4360634640669039,
"learning_rate": 3.755971950641245e-05,
"loss": 1.4314,
"step": 942
},
{
"epoch": 0.16757741347905283,
"grad_norm": 0.46601965201822415,
"learning_rate": 3.755431693793865e-05,
"loss": 1.4956,
"step": 943
},
{
"epoch": 0.16775512017415256,
"grad_norm": 0.4365788362677747,
"learning_rate": 3.7548908785106515e-05,
"loss": 1.39,
"step": 944
},
{
"epoch": 0.1679328268692523,
"grad_norm": 0.4821065775333637,
"learning_rate": 3.7543495049636466e-05,
"loss": 1.4779,
"step": 945
},
{
"epoch": 0.16811053356435204,
"grad_norm": 0.5082679333449232,
"learning_rate": 3.7538075733250724e-05,
"loss": 1.4606,
"step": 946
},
{
"epoch": 0.16828824025945177,
"grad_norm": 0.4578632229130094,
"learning_rate": 3.753265083767328e-05,
"loss": 1.4688,
"step": 947
},
{
"epoch": 0.16846594695455153,
"grad_norm": 0.45972888541142687,
"learning_rate": 3.752722036462988e-05,
"loss": 1.4417,
"step": 948
},
{
"epoch": 0.16864365364965125,
"grad_norm": 0.46551544825460855,
"learning_rate": 3.752178431584806e-05,
"loss": 1.4293,
"step": 949
},
{
"epoch": 0.16882136034475098,
"grad_norm": 0.472046643783578,
"learning_rate": 3.751634269305715e-05,
"loss": 1.4177,
"step": 950
},
{
"epoch": 0.16899906703985074,
"grad_norm": 0.4709713268723435,
"learning_rate": 3.751089549798822e-05,
"loss": 1.4401,
"step": 951
},
{
"epoch": 0.16917677373495046,
"grad_norm": 0.4392993516067808,
"learning_rate": 3.750544273237412e-05,
"loss": 1.4344,
"step": 952
},
{
"epoch": 0.1693544804300502,
"grad_norm": 0.4546057693802621,
"learning_rate": 3.749998439794948e-05,
"loss": 1.3917,
"step": 953
},
{
"epoch": 0.16953218712514995,
"grad_norm": 0.45635953241267835,
"learning_rate": 3.7494520496450706e-05,
"loss": 1.4762,
"step": 954
},
{
"epoch": 0.16970989382024967,
"grad_norm": 0.49199295438824736,
"learning_rate": 3.7489051029615964e-05,
"loss": 1.4518,
"step": 955
},
{
"epoch": 0.16988760051534943,
"grad_norm": 0.4426917759639636,
"learning_rate": 3.7483575999185184e-05,
"loss": 1.4382,
"step": 956
},
{
"epoch": 0.17006530721044916,
"grad_norm": 0.42887914121305687,
"learning_rate": 3.7478095406900095e-05,
"loss": 1.4529,
"step": 957
},
{
"epoch": 0.17024301390554888,
"grad_norm": 0.4551079790360319,
"learning_rate": 3.7472609254504163e-05,
"loss": 1.4466,
"step": 958
},
{
"epoch": 0.17042072060064864,
"grad_norm": 0.4345721494831705,
"learning_rate": 3.746711754374264e-05,
"loss": 1.4758,
"step": 959
},
{
"epoch": 0.17059842729574837,
"grad_norm": 0.5020756502414094,
"learning_rate": 3.7461620276362546e-05,
"loss": 1.4548,
"step": 960
},
{
"epoch": 0.1707761339908481,
"grad_norm": 0.4523902983706086,
"learning_rate": 3.7456117454112654e-05,
"loss": 1.4399,
"step": 961
},
{
"epoch": 0.17095384068594785,
"grad_norm": 0.45274481583164505,
"learning_rate": 3.7450609078743525e-05,
"loss": 1.4534,
"step": 962
},
{
"epoch": 0.17113154738104758,
"grad_norm": 0.45457871051034965,
"learning_rate": 3.744509515200748e-05,
"loss": 1.4139,
"step": 963
},
{
"epoch": 0.17130925407614733,
"grad_norm": 0.45400588013878695,
"learning_rate": 3.74395756756586e-05,
"loss": 1.3987,
"step": 964
},
{
"epoch": 0.17148696077124706,
"grad_norm": 0.4554342391133547,
"learning_rate": 3.743405065145272e-05,
"loss": 1.4263,
"step": 965
},
{
"epoch": 0.1716646674663468,
"grad_norm": 0.45838743426062856,
"learning_rate": 3.742852008114747e-05,
"loss": 1.4265,
"step": 966
},
{
"epoch": 0.17184237416144654,
"grad_norm": 0.44097581986033074,
"learning_rate": 3.7422983966502226e-05,
"loss": 1.435,
"step": 967
},
{
"epoch": 0.17202008085654627,
"grad_norm": 0.43286286359567755,
"learning_rate": 3.741744230927813e-05,
"loss": 1.4267,
"step": 968
},
{
"epoch": 0.172197787551646,
"grad_norm": 0.4629962018060795,
"learning_rate": 3.741189511123808e-05,
"loss": 1.4123,
"step": 969
},
{
"epoch": 0.17237549424674575,
"grad_norm": 0.4179209440155217,
"learning_rate": 3.7406342374146755e-05,
"loss": 1.3918,
"step": 970
},
{
"epoch": 0.17255320094184548,
"grad_norm": 0.7169335291129493,
"learning_rate": 3.740078409977057e-05,
"loss": 1.3663,
"step": 971
},
{
"epoch": 0.17273090763694524,
"grad_norm": 1.0784358303062258,
"learning_rate": 3.739522028987774e-05,
"loss": 1.4095,
"step": 972
},
{
"epoch": 0.17290861433204496,
"grad_norm": 0.484475456652512,
"learning_rate": 3.73896509462382e-05,
"loss": 1.4687,
"step": 973
},
{
"epoch": 0.1730863210271447,
"grad_norm": 0.43464828497745317,
"learning_rate": 3.7384076070623663e-05,
"loss": 1.4301,
"step": 974
},
{
"epoch": 0.17326402772224445,
"grad_norm": 0.44960888976398267,
"learning_rate": 3.737849566480761e-05,
"loss": 1.4159,
"step": 975
},
{
"epoch": 0.17344173441734417,
"grad_norm": 0.45629591652914586,
"learning_rate": 3.737290973056527e-05,
"loss": 1.4284,
"step": 976
},
{
"epoch": 0.1736194411124439,
"grad_norm": 0.47311161543293684,
"learning_rate": 3.7367318269673626e-05,
"loss": 1.4012,
"step": 977
},
{
"epoch": 0.17379714780754366,
"grad_norm": 0.46815200049267125,
"learning_rate": 3.736172128391144e-05,
"loss": 1.414,
"step": 978
},
{
"epoch": 0.17397485450264338,
"grad_norm": 0.4911497626391726,
"learning_rate": 3.7356118775059205e-05,
"loss": 1.4885,
"step": 979
},
{
"epoch": 0.17415256119774314,
"grad_norm": 0.7719250056861239,
"learning_rate": 3.73505107448992e-05,
"loss": 1.4819,
"step": 980
},
{
"epoch": 0.17433026789284287,
"grad_norm": 0.46233587038957197,
"learning_rate": 3.7344897195215427e-05,
"loss": 1.4333,
"step": 981
},
{
"epoch": 0.1745079745879426,
"grad_norm": 0.4455737298756048,
"learning_rate": 3.733927812779367e-05,
"loss": 1.4143,
"step": 982
},
{
"epoch": 0.17468568128304235,
"grad_norm": 2.542396348686165,
"learning_rate": 3.733365354442147e-05,
"loss": 1.4314,
"step": 983
},
{
"epoch": 0.17486338797814208,
"grad_norm": 0.7260753362621891,
"learning_rate": 3.73280234468881e-05,
"loss": 1.4562,
"step": 984
},
{
"epoch": 0.1750410946732418,
"grad_norm": 0.6656077695759083,
"learning_rate": 3.73223878369846e-05,
"loss": 1.4026,
"step": 985
},
{
"epoch": 0.17521880136834156,
"grad_norm": 0.5228997068419738,
"learning_rate": 3.731674671650377e-05,
"loss": 1.4557,
"step": 986
},
{
"epoch": 0.1753965080634413,
"grad_norm": 0.4854694705680882,
"learning_rate": 3.731110008724015e-05,
"loss": 1.425,
"step": 987
},
{
"epoch": 0.17557421475854104,
"grad_norm": 0.5374457610601567,
"learning_rate": 3.7305447950990045e-05,
"loss": 1.4341,
"step": 988
},
{
"epoch": 0.17575192145364077,
"grad_norm": 1.1321481862855063,
"learning_rate": 3.72997903095515e-05,
"loss": 1.4304,
"step": 989
},
{
"epoch": 0.1759296281487405,
"grad_norm": 0.5703633779559925,
"learning_rate": 3.729412716472433e-05,
"loss": 1.476,
"step": 990
},
{
"epoch": 0.17610733484384025,
"grad_norm": 0.47565588635255274,
"learning_rate": 3.7288458518310064e-05,
"loss": 1.4548,
"step": 991
},
{
"epoch": 0.17628504153893998,
"grad_norm": 0.5299818633177216,
"learning_rate": 3.7282784372112034e-05,
"loss": 1.4276,
"step": 992
},
{
"epoch": 0.1764627482340397,
"grad_norm": 0.5028913779384946,
"learning_rate": 3.727710472793527e-05,
"loss": 1.4573,
"step": 993
},
{
"epoch": 0.17664045492913946,
"grad_norm": 0.48914408894264777,
"learning_rate": 3.727141958758658e-05,
"loss": 1.4266,
"step": 994
},
{
"epoch": 0.1768181616242392,
"grad_norm": 0.47144386059460486,
"learning_rate": 3.726572895287451e-05,
"loss": 1.4303,
"step": 995
},
{
"epoch": 0.17699586831933894,
"grad_norm": 0.46618408848488435,
"learning_rate": 3.726003282560938e-05,
"loss": 1.4311,
"step": 996
},
{
"epoch": 0.17717357501443867,
"grad_norm": 0.45907263594171704,
"learning_rate": 3.7254331207603206e-05,
"loss": 1.4178,
"step": 997
},
{
"epoch": 0.1773512817095384,
"grad_norm": 0.47958400626869224,
"learning_rate": 3.72486241006698e-05,
"loss": 1.4508,
"step": 998
},
{
"epoch": 0.17752898840463815,
"grad_norm": 0.45760467485561257,
"learning_rate": 3.724291150662469e-05,
"loss": 1.4025,
"step": 999
},
{
"epoch": 0.17770669509973788,
"grad_norm": 0.47270245571488745,
"learning_rate": 3.723719342728516e-05,
"loss": 1.3934,
"step": 1000
},
{
"epoch": 0.1778844017948376,
"grad_norm": 0.4417648790178083,
"learning_rate": 3.7231469864470245e-05,
"loss": 1.4058,
"step": 1001
},
{
"epoch": 0.17806210848993737,
"grad_norm": 1.151544841848841,
"learning_rate": 3.722574082000071e-05,
"loss": 1.4123,
"step": 1002
},
{
"epoch": 0.1782398151850371,
"grad_norm": 0.44901614823004227,
"learning_rate": 3.7220006295699076e-05,
"loss": 1.4182,
"step": 1003
},
{
"epoch": 0.17841752188013685,
"grad_norm": 0.4990677281235097,
"learning_rate": 3.72142662933896e-05,
"loss": 1.4117,
"step": 1004
},
{
"epoch": 0.17859522857523658,
"grad_norm": 0.4408293777835546,
"learning_rate": 3.7208520814898295e-05,
"loss": 1.4299,
"step": 1005
},
{
"epoch": 0.1787729352703363,
"grad_norm": 0.5163333407518188,
"learning_rate": 3.720276986205289e-05,
"loss": 1.4871,
"step": 1006
},
{
"epoch": 0.17895064196543606,
"grad_norm": 0.43061042506049724,
"learning_rate": 3.719701343668289e-05,
"loss": 1.4397,
"step": 1007
},
{
"epoch": 0.17912834866053579,
"grad_norm": 0.7057354172024568,
"learning_rate": 3.71912515406195e-05,
"loss": 1.4145,
"step": 1008
},
{
"epoch": 0.1793060553556355,
"grad_norm": 0.4592206378284871,
"learning_rate": 3.71854841756957e-05,
"loss": 1.4283,
"step": 1009
},
{
"epoch": 0.17948376205073527,
"grad_norm": 0.45253562053240415,
"learning_rate": 3.71797113437462e-05,
"loss": 1.4437,
"step": 1010
},
{
"epoch": 0.179661468745835,
"grad_norm": 0.4689265090922591,
"learning_rate": 3.717393304660744e-05,
"loss": 1.4706,
"step": 1011
},
{
"epoch": 0.17983917544093475,
"grad_norm": 0.46323731180223027,
"learning_rate": 3.7168149286117614e-05,
"loss": 1.4088,
"step": 1012
},
{
"epoch": 0.18001688213603448,
"grad_norm": 0.45273185542420297,
"learning_rate": 3.716236006411663e-05,
"loss": 1.4849,
"step": 1013
},
{
"epoch": 0.1801945888311342,
"grad_norm": 0.49096191295327735,
"learning_rate": 3.7156565382446164e-05,
"loss": 1.4479,
"step": 1014
},
{
"epoch": 0.18037229552623396,
"grad_norm": 0.4247216427520264,
"learning_rate": 3.71507652429496e-05,
"loss": 1.464,
"step": 1015
},
{
"epoch": 0.1805500022213337,
"grad_norm": 0.42960100271307966,
"learning_rate": 3.714495964747208e-05,
"loss": 1.4486,
"step": 1016
},
{
"epoch": 0.18072770891643342,
"grad_norm": 0.44001336493839505,
"learning_rate": 3.7139148597860475e-05,
"loss": 1.4764,
"step": 1017
},
{
"epoch": 0.18090541561153317,
"grad_norm": 0.5609801230610598,
"learning_rate": 3.713333209596338e-05,
"loss": 1.4832,
"step": 1018
},
{
"epoch": 0.1810831223066329,
"grad_norm": 0.441685930233975,
"learning_rate": 3.712751014363114e-05,
"loss": 1.4644,
"step": 1019
},
{
"epoch": 0.18126082900173265,
"grad_norm": 0.4321524688727574,
"learning_rate": 3.712168274271583e-05,
"loss": 1.459,
"step": 1020
},
{
"epoch": 0.18143853569683238,
"grad_norm": 0.4348386579564695,
"learning_rate": 3.7115849895071244e-05,
"loss": 1.4277,
"step": 1021
},
{
"epoch": 0.1816162423919321,
"grad_norm": 0.44433582537351246,
"learning_rate": 3.711001160255293e-05,
"loss": 1.4354,
"step": 1022
},
{
"epoch": 0.18179394908703186,
"grad_norm": 0.5164899950523163,
"learning_rate": 3.710416786701816e-05,
"loss": 1.4657,
"step": 1023
},
{
"epoch": 0.1819716557821316,
"grad_norm": 0.4235832067805997,
"learning_rate": 3.709831869032593e-05,
"loss": 1.3902,
"step": 1024
},
{
"epoch": 0.18214936247723132,
"grad_norm": 0.44639408704260364,
"learning_rate": 3.709246407433698e-05,
"loss": 1.4434,
"step": 1025
},
{
"epoch": 0.18232706917233107,
"grad_norm": 0.4515274121894754,
"learning_rate": 3.708660402091377e-05,
"loss": 1.4889,
"step": 1026
},
{
"epoch": 0.1825047758674308,
"grad_norm": 0.4368982467523272,
"learning_rate": 3.7080738531920485e-05,
"loss": 1.4204,
"step": 1027
},
{
"epoch": 0.18268248256253056,
"grad_norm": 0.45549065615175693,
"learning_rate": 3.707486760922306e-05,
"loss": 1.4517,
"step": 1028
},
{
"epoch": 0.18286018925763028,
"grad_norm": 0.44706570716978783,
"learning_rate": 3.706899125468915e-05,
"loss": 1.4319,
"step": 1029
},
{
"epoch": 0.18303789595273,
"grad_norm": 0.4352661297114883,
"learning_rate": 3.706310947018812e-05,
"loss": 1.4457,
"step": 1030
},
{
"epoch": 0.18321560264782977,
"grad_norm": 0.4470875921005806,
"learning_rate": 3.7057222257591076e-05,
"loss": 1.4284,
"step": 1031
},
{
"epoch": 0.1833933093429295,
"grad_norm": 0.41471703311188945,
"learning_rate": 3.705132961877086e-05,
"loss": 1.4464,
"step": 1032
},
{
"epoch": 0.18357101603802922,
"grad_norm": 0.45787313617931447,
"learning_rate": 3.7045431555602027e-05,
"loss": 1.4108,
"step": 1033
},
{
"epoch": 0.18374872273312898,
"grad_norm": 0.4196656407976471,
"learning_rate": 3.703952806996086e-05,
"loss": 1.4043,
"step": 1034
},
{
"epoch": 0.1839264294282287,
"grad_norm": 0.4443610528078267,
"learning_rate": 3.703361916372537e-05,
"loss": 1.4388,
"step": 1035
},
{
"epoch": 0.18410413612332846,
"grad_norm": 0.406979449894464,
"learning_rate": 3.702770483877529e-05,
"loss": 1.3696,
"step": 1036
},
{
"epoch": 0.1842818428184282,
"grad_norm": 0.4446302305308523,
"learning_rate": 3.7021785096992094e-05,
"loss": 1.4496,
"step": 1037
},
{
"epoch": 0.18445954951352792,
"grad_norm": 0.4148766509424914,
"learning_rate": 3.7015859940258945e-05,
"loss": 1.4199,
"step": 1038
},
{
"epoch": 0.18463725620862767,
"grad_norm": 0.42128718072797816,
"learning_rate": 3.700992937046074e-05,
"loss": 1.4086,
"step": 1039
},
{
"epoch": 0.1848149629037274,
"grad_norm": 0.42288488632167615,
"learning_rate": 3.700399338948413e-05,
"loss": 1.447,
"step": 1040
},
{
"epoch": 0.18499266959882713,
"grad_norm": 0.4425728394764142,
"learning_rate": 3.6998051999217446e-05,
"loss": 1.4242,
"step": 1041
},
{
"epoch": 0.18517037629392688,
"grad_norm": 0.48060748531657493,
"learning_rate": 3.699210520155075e-05,
"loss": 1.4233,
"step": 1042
},
{
"epoch": 0.1853480829890266,
"grad_norm": 0.4409773177975972,
"learning_rate": 3.6986152998375845e-05,
"loss": 1.4743,
"step": 1043
},
{
"epoch": 0.18552578968412636,
"grad_norm": 0.4181857266567988,
"learning_rate": 3.6980195391586234e-05,
"loss": 1.3724,
"step": 1044
},
{
"epoch": 0.1857034963792261,
"grad_norm": 0.4443847116503312,
"learning_rate": 3.697423238307714e-05,
"loss": 1.4644,
"step": 1045
},
{
"epoch": 0.18588120307432582,
"grad_norm": 0.42505659970898246,
"learning_rate": 3.69682639747455e-05,
"loss": 1.3964,
"step": 1046
},
{
"epoch": 0.18605890976942557,
"grad_norm": 0.4556176956547751,
"learning_rate": 3.696229016849001e-05,
"loss": 1.449,
"step": 1047
},
{
"epoch": 0.1862366164645253,
"grad_norm": 0.44753158083748945,
"learning_rate": 3.6956310966211e-05,
"loss": 1.4513,
"step": 1048
},
{
"epoch": 0.18641432315962503,
"grad_norm": 0.5316954434592434,
"learning_rate": 3.6950326369810616e-05,
"loss": 1.4245,
"step": 1049
},
{
"epoch": 0.18659202985472478,
"grad_norm": 0.4454172320927584,
"learning_rate": 3.694433638119264e-05,
"loss": 1.4325,
"step": 1050
},
{
"epoch": 0.1867697365498245,
"grad_norm": 0.5447391361346448,
"learning_rate": 3.6938341002262605e-05,
"loss": 1.4624,
"step": 1051
},
{
"epoch": 0.18694744324492427,
"grad_norm": 0.4407169048948348,
"learning_rate": 3.693234023492776e-05,
"loss": 1.4552,
"step": 1052
},
{
"epoch": 0.187125149940024,
"grad_norm": 0.43493992423985844,
"learning_rate": 3.692633408109706e-05,
"loss": 1.4227,
"step": 1053
},
{
"epoch": 0.18730285663512372,
"grad_norm": 0.44083912401234027,
"learning_rate": 3.6920322542681175e-05,
"loss": 1.4198,
"step": 1054
},
{
"epoch": 0.18748056333022348,
"grad_norm": 0.4621359478684846,
"learning_rate": 3.6914305621592486e-05,
"loss": 1.4267,
"step": 1055
},
{
"epoch": 0.1876582700253232,
"grad_norm": 0.4483266453827309,
"learning_rate": 3.690828331974509e-05,
"loss": 1.4468,
"step": 1056
},
{
"epoch": 0.18783597672042293,
"grad_norm": 0.6272970554658519,
"learning_rate": 3.6902255639054806e-05,
"loss": 1.4216,
"step": 1057
},
{
"epoch": 0.1880136834155227,
"grad_norm": 0.4565286470605675,
"learning_rate": 3.6896222581439134e-05,
"loss": 1.4127,
"step": 1058
},
{
"epoch": 0.18819139011062241,
"grad_norm": 0.45484780288626986,
"learning_rate": 3.689018414881731e-05,
"loss": 1.4292,
"step": 1059
},
{
"epoch": 0.18836909680572217,
"grad_norm": 0.44153967035667124,
"learning_rate": 3.688414034311028e-05,
"loss": 1.45,
"step": 1060
},
{
"epoch": 0.1885468035008219,
"grad_norm": 0.46125845944019317,
"learning_rate": 3.6878091166240676e-05,
"loss": 1.4517,
"step": 1061
},
{
"epoch": 0.18872451019592162,
"grad_norm": 0.43109007115006603,
"learning_rate": 3.687203662013287e-05,
"loss": 1.4698,
"step": 1062
},
{
"epoch": 0.18890221689102138,
"grad_norm": 0.43886539035427746,
"learning_rate": 3.686597670671293e-05,
"loss": 1.4651,
"step": 1063
},
{
"epoch": 0.1890799235861211,
"grad_norm": 0.45882472879108416,
"learning_rate": 3.685991142790861e-05,
"loss": 1.4395,
"step": 1064
},
{
"epoch": 0.18925763028122083,
"grad_norm": 0.4182727754238493,
"learning_rate": 3.6853840785649404e-05,
"loss": 1.415,
"step": 1065
},
{
"epoch": 0.1894353369763206,
"grad_norm": 0.44034128627992397,
"learning_rate": 3.684776478186649e-05,
"loss": 1.3752,
"step": 1066
},
{
"epoch": 0.18961304367142032,
"grad_norm": 0.4260836583689225,
"learning_rate": 3.6841683418492765e-05,
"loss": 1.4299,
"step": 1067
},
{
"epoch": 0.18979075036652007,
"grad_norm": 0.4299751591386746,
"learning_rate": 3.683559669746283e-05,
"loss": 1.356,
"step": 1068
},
{
"epoch": 0.1899684570616198,
"grad_norm": 0.42372324142294765,
"learning_rate": 3.682950462071297e-05,
"loss": 1.4422,
"step": 1069
},
{
"epoch": 0.19014616375671953,
"grad_norm": 0.46155167017662735,
"learning_rate": 3.68234071901812e-05,
"loss": 1.4045,
"step": 1070
},
{
"epoch": 0.19032387045181928,
"grad_norm": 0.4124338783463005,
"learning_rate": 3.6817304407807226e-05,
"loss": 1.3599,
"step": 1071
},
{
"epoch": 0.190501577146919,
"grad_norm": 0.4415821121035263,
"learning_rate": 3.681119627553245e-05,
"loss": 1.4003,
"step": 1072
},
{
"epoch": 0.19067928384201874,
"grad_norm": 0.554555589370907,
"learning_rate": 3.680508279529999e-05,
"loss": 1.4006,
"step": 1073
},
{
"epoch": 0.1908569905371185,
"grad_norm": 0.42224485149835916,
"learning_rate": 3.679896396905467e-05,
"loss": 1.4271,
"step": 1074
},
{
"epoch": 0.19103469723221822,
"grad_norm": 0.4453601401162064,
"learning_rate": 3.679283979874298e-05,
"loss": 1.4272,
"step": 1075
},
{
"epoch": 0.19121240392731798,
"grad_norm": 0.4241383460960415,
"learning_rate": 3.6786710286313154e-05,
"loss": 1.4757,
"step": 1076
},
{
"epoch": 0.1913901106224177,
"grad_norm": 0.545592522263147,
"learning_rate": 3.678057543371509e-05,
"loss": 1.4158,
"step": 1077
},
{
"epoch": 0.19156781731751743,
"grad_norm": 0.41305698421585296,
"learning_rate": 3.677443524290042e-05,
"loss": 1.406,
"step": 1078
},
{
"epoch": 0.19174552401261719,
"grad_norm": 0.44851387739473364,
"learning_rate": 3.676828971582243e-05,
"loss": 1.4287,
"step": 1079
},
{
"epoch": 0.1919232307077169,
"grad_norm": 0.443779564254347,
"learning_rate": 3.6762138854436146e-05,
"loss": 1.4437,
"step": 1080
},
{
"epoch": 0.19210093740281664,
"grad_norm": 0.4238218613442605,
"learning_rate": 3.6755982660698265e-05,
"loss": 1.379,
"step": 1081
},
{
"epoch": 0.1922786440979164,
"grad_norm": 0.4333094397051119,
"learning_rate": 3.674982113656719e-05,
"loss": 1.4358,
"step": 1082
},
{
"epoch": 0.19245635079301612,
"grad_norm": 0.44169505670078996,
"learning_rate": 3.674365428400301e-05,
"loss": 1.4723,
"step": 1083
},
{
"epoch": 0.19263405748811588,
"grad_norm": 0.44891172143636165,
"learning_rate": 3.673748210496754e-05,
"loss": 1.4563,
"step": 1084
},
{
"epoch": 0.1928117641832156,
"grad_norm": 0.41001700514615447,
"learning_rate": 3.6731304601424234e-05,
"loss": 1.3653,
"step": 1085
},
{
"epoch": 0.19298947087831533,
"grad_norm": 0.4632061676291482,
"learning_rate": 3.672512177533829e-05,
"loss": 1.4368,
"step": 1086
},
{
"epoch": 0.1931671775734151,
"grad_norm": 0.44792990007076405,
"learning_rate": 3.671893362867658e-05,
"loss": 1.4542,
"step": 1087
},
{
"epoch": 0.19334488426851482,
"grad_norm": 0.4401351542901058,
"learning_rate": 3.671274016340768e-05,
"loss": 1.4564,
"step": 1088
},
{
"epoch": 0.19352259096361454,
"grad_norm": 0.43635851816781446,
"learning_rate": 3.670654138150182e-05,
"loss": 1.4095,
"step": 1089
},
{
"epoch": 0.1937002976587143,
"grad_norm": 0.4218573775655246,
"learning_rate": 3.670033728493097e-05,
"loss": 1.4146,
"step": 1090
},
{
"epoch": 0.19387800435381403,
"grad_norm": 0.4192302297576943,
"learning_rate": 3.669412787566878e-05,
"loss": 1.3842,
"step": 1091
},
{
"epoch": 0.19405571104891378,
"grad_norm": 0.4193916891563698,
"learning_rate": 3.668791315569055e-05,
"loss": 1.4237,
"step": 1092
},
{
"epoch": 0.1942334177440135,
"grad_norm": 0.41391009661385514,
"learning_rate": 3.668169312697332e-05,
"loss": 1.4268,
"step": 1093
},
{
"epoch": 0.19441112443911324,
"grad_norm": 0.4237113832337278,
"learning_rate": 3.66754677914958e-05,
"loss": 1.4235,
"step": 1094
},
{
"epoch": 0.194588831134213,
"grad_norm": 0.41043837532114125,
"learning_rate": 3.666923715123837e-05,
"loss": 1.3755,
"step": 1095
},
{
"epoch": 0.19476653782931272,
"grad_norm": 0.4171480825962362,
"learning_rate": 3.666300120818313e-05,
"loss": 1.4178,
"step": 1096
},
{
"epoch": 0.19494424452441245,
"grad_norm": 0.42595846763815437,
"learning_rate": 3.665675996431383e-05,
"loss": 1.4342,
"step": 1097
},
{
"epoch": 0.1951219512195122,
"grad_norm": 0.41521033840204497,
"learning_rate": 3.6650513421615955e-05,
"loss": 1.3695,
"step": 1098
},
{
"epoch": 0.19529965791461193,
"grad_norm": 0.45086260254516475,
"learning_rate": 3.664426158207663e-05,
"loss": 1.4286,
"step": 1099
},
{
"epoch": 0.19547736460971168,
"grad_norm": 0.623194223881299,
"learning_rate": 3.663800444768468e-05,
"loss": 1.3698,
"step": 1100
},
{
"epoch": 0.1956550713048114,
"grad_norm": 0.4317215418384982,
"learning_rate": 3.663174202043063e-05,
"loss": 1.4154,
"step": 1101
},
{
"epoch": 0.19583277799991114,
"grad_norm": 0.4373409314762208,
"learning_rate": 3.662547430230667e-05,
"loss": 1.4345,
"step": 1102
},
{
"epoch": 0.1960104846950109,
"grad_norm": 0.4499722016240007,
"learning_rate": 3.661920129530668e-05,
"loss": 1.4452,
"step": 1103
},
{
"epoch": 0.19618819139011062,
"grad_norm": 0.4047165375054159,
"learning_rate": 3.661292300142622e-05,
"loss": 1.3921,
"step": 1104
},
{
"epoch": 0.19636589808521035,
"grad_norm": 0.4334611124662071,
"learning_rate": 3.6606639422662525e-05,
"loss": 1.4038,
"step": 1105
},
{
"epoch": 0.1965436047803101,
"grad_norm": 0.4306778453261507,
"learning_rate": 3.660035056101453e-05,
"loss": 1.421,
"step": 1106
},
{
"epoch": 0.19672131147540983,
"grad_norm": 0.45990842150645517,
"learning_rate": 3.6594056418482844e-05,
"loss": 1.4462,
"step": 1107
},
{
"epoch": 0.1968990181705096,
"grad_norm": 0.41533954165630327,
"learning_rate": 3.658775699706974e-05,
"loss": 1.4093,
"step": 1108
},
{
"epoch": 0.19707672486560932,
"grad_norm": 0.46169125055390864,
"learning_rate": 3.658145229877919e-05,
"loss": 1.3761,
"step": 1109
},
{
"epoch": 0.19725443156070904,
"grad_norm": 0.4346877161885415,
"learning_rate": 3.657514232561684e-05,
"loss": 1.4235,
"step": 1110
},
{
"epoch": 0.1974321382558088,
"grad_norm": 0.46430207637930576,
"learning_rate": 3.656882707959e-05,
"loss": 1.4529,
"step": 1111
},
{
"epoch": 0.19760984495090853,
"grad_norm": 0.4682054383112406,
"learning_rate": 3.656250656270768e-05,
"loss": 1.4558,
"step": 1112
},
{
"epoch": 0.19778755164600825,
"grad_norm": 1.2745175303188616,
"learning_rate": 3.655618077698055e-05,
"loss": 1.3949,
"step": 1113
},
{
"epoch": 0.197965258341108,
"grad_norm": 0.4387928375863726,
"learning_rate": 3.654984972442096e-05,
"loss": 1.4583,
"step": 1114
},
{
"epoch": 0.19814296503620774,
"grad_norm": 0.46491515552883766,
"learning_rate": 3.654351340704294e-05,
"loss": 1.5043,
"step": 1115
},
{
"epoch": 0.1983206717313075,
"grad_norm": 0.4261919011505481,
"learning_rate": 3.6537171826862186e-05,
"loss": 1.4217,
"step": 1116
},
{
"epoch": 0.19849837842640722,
"grad_norm": 0.5065489892660902,
"learning_rate": 3.653082498589608e-05,
"loss": 1.4183,
"step": 1117
},
{
"epoch": 0.19867608512150695,
"grad_norm": 0.43715677068644615,
"learning_rate": 3.6524472886163676e-05,
"loss": 1.4176,
"step": 1118
},
{
"epoch": 0.1988537918166067,
"grad_norm": 0.433475214258197,
"learning_rate": 3.6518115529685683e-05,
"loss": 1.4529,
"step": 1119
},
{
"epoch": 0.19903149851170643,
"grad_norm": 0.43976021381503744,
"learning_rate": 3.651175291848451e-05,
"loss": 1.3847,
"step": 1120
},
{
"epoch": 0.19920920520680616,
"grad_norm": 0.4450224998408098,
"learning_rate": 3.650538505458421e-05,
"loss": 1.4167,
"step": 1121
},
{
"epoch": 0.1993869119019059,
"grad_norm": 0.4417400921323867,
"learning_rate": 3.649901194001053e-05,
"loss": 1.4472,
"step": 1122
},
{
"epoch": 0.19956461859700564,
"grad_norm": 0.4219750538139898,
"learning_rate": 3.649263357679087e-05,
"loss": 1.386,
"step": 1123
},
{
"epoch": 0.1997423252921054,
"grad_norm": 0.44177517737095756,
"learning_rate": 3.648624996695432e-05,
"loss": 1.4516,
"step": 1124
},
{
"epoch": 0.19992003198720512,
"grad_norm": 0.4438188691207675,
"learning_rate": 3.647986111253161e-05,
"loss": 1.3987,
"step": 1125
},
{
"epoch": 0.20009773868230485,
"grad_norm": 0.4267600776622297,
"learning_rate": 3.647346701555516e-05,
"loss": 1.4433,
"step": 1126
},
{
"epoch": 0.2002754453774046,
"grad_norm": 0.4490087825461045,
"learning_rate": 3.646706767805906e-05,
"loss": 1.4625,
"step": 1127
},
{
"epoch": 0.20045315207250433,
"grad_norm": 0.638674334865335,
"learning_rate": 3.646066310207905e-05,
"loss": 1.4103,
"step": 1128
},
{
"epoch": 0.20063085876760406,
"grad_norm": 0.44414979738016686,
"learning_rate": 3.645425328965256e-05,
"loss": 1.4233,
"step": 1129
},
{
"epoch": 0.20080856546270381,
"grad_norm": 0.43270033001876534,
"learning_rate": 3.6447838242818655e-05,
"loss": 1.4465,
"step": 1130
},
{
"epoch": 0.20098627215780354,
"grad_norm": 0.4403635134676768,
"learning_rate": 3.6441417963618094e-05,
"loss": 1.4402,
"step": 1131
},
{
"epoch": 0.2011639788529033,
"grad_norm": 0.442380623807038,
"learning_rate": 3.643499245409328e-05,
"loss": 1.4108,
"step": 1132
},
{
"epoch": 0.20134168554800302,
"grad_norm": 0.42645280130540014,
"learning_rate": 3.6428561716288295e-05,
"loss": 1.3844,
"step": 1133
},
{
"epoch": 0.20151939224310275,
"grad_norm": 0.44573622749209074,
"learning_rate": 3.6422125752248876e-05,
"loss": 1.4135,
"step": 1134
},
{
"epoch": 0.2016970989382025,
"grad_norm": 0.43900025310384394,
"learning_rate": 3.6415684564022415e-05,
"loss": 1.4449,
"step": 1135
},
{
"epoch": 0.20187480563330223,
"grad_norm": 0.4300234348045378,
"learning_rate": 3.640923815365799e-05,
"loss": 1.437,
"step": 1136
},
{
"epoch": 0.20205251232840196,
"grad_norm": 0.4169040830591406,
"learning_rate": 3.640278652320632e-05,
"loss": 1.4137,
"step": 1137
},
{
"epoch": 0.20223021902350172,
"grad_norm": 0.44149967944120305,
"learning_rate": 3.639632967471978e-05,
"loss": 1.4307,
"step": 1138
},
{
"epoch": 0.20240792571860144,
"grad_norm": 0.41927755392906363,
"learning_rate": 3.6389867610252434e-05,
"loss": 1.3972,
"step": 1139
},
{
"epoch": 0.2025856324137012,
"grad_norm": 0.4526716906912448,
"learning_rate": 3.6383400331859975e-05,
"loss": 1.4245,
"step": 1140
},
{
"epoch": 0.20276333910880093,
"grad_norm": 0.4334120620381292,
"learning_rate": 3.637692784159976e-05,
"loss": 1.3987,
"step": 1141
},
{
"epoch": 0.20294104580390065,
"grad_norm": 0.49542410551432847,
"learning_rate": 3.637045014153082e-05,
"loss": 1.4574,
"step": 1142
},
{
"epoch": 0.2031187524990004,
"grad_norm": 0.49603332865584326,
"learning_rate": 3.636396723371383e-05,
"loss": 1.4945,
"step": 1143
},
{
"epoch": 0.20329645919410014,
"grad_norm": 0.4336106881461919,
"learning_rate": 3.635747912021113e-05,
"loss": 1.423,
"step": 1144
},
{
"epoch": 0.20347416588919987,
"grad_norm": 0.7202149885328238,
"learning_rate": 3.63509858030867e-05,
"loss": 1.407,
"step": 1145
},
{
"epoch": 0.20365187258429962,
"grad_norm": 0.4311900756451114,
"learning_rate": 3.6344487284406195e-05,
"loss": 1.4103,
"step": 1146
},
{
"epoch": 0.20382957927939935,
"grad_norm": 0.4724036413229147,
"learning_rate": 3.633798356623691e-05,
"loss": 1.4598,
"step": 1147
},
{
"epoch": 0.2040072859744991,
"grad_norm": 0.4332198873160031,
"learning_rate": 3.63314746506478e-05,
"loss": 1.4025,
"step": 1148
},
{
"epoch": 0.20418499266959883,
"grad_norm": 0.46435489157239224,
"learning_rate": 3.6324960539709485e-05,
"loss": 1.4097,
"step": 1149
},
{
"epoch": 0.20436269936469856,
"grad_norm": 0.4217359909178058,
"learning_rate": 3.631844123549421e-05,
"loss": 1.3719,
"step": 1150
},
{
"epoch": 0.2045404060597983,
"grad_norm": 0.4485266334648088,
"learning_rate": 3.63119167400759e-05,
"loss": 1.3873,
"step": 1151
},
{
"epoch": 0.20471811275489804,
"grad_norm": 0.4341148107217547,
"learning_rate": 3.6305387055530115e-05,
"loss": 1.4329,
"step": 1152
},
{
"epoch": 0.20489581944999777,
"grad_norm": 0.4375687293042603,
"learning_rate": 3.6298852183934066e-05,
"loss": 1.4342,
"step": 1153
},
{
"epoch": 0.20507352614509752,
"grad_norm": 0.4721103690811158,
"learning_rate": 3.6292312127366634e-05,
"loss": 1.4441,
"step": 1154
},
{
"epoch": 0.20525123284019725,
"grad_norm": 0.42813838393743325,
"learning_rate": 3.6285766887908316e-05,
"loss": 1.3781,
"step": 1155
},
{
"epoch": 0.205428939535297,
"grad_norm": 0.42369288670760147,
"learning_rate": 3.6279216467641287e-05,
"loss": 1.4048,
"step": 1156
},
{
"epoch": 0.20560664623039673,
"grad_norm": 0.5012652205970332,
"learning_rate": 3.627266086864935e-05,
"loss": 1.4361,
"step": 1157
},
{
"epoch": 0.20578435292549646,
"grad_norm": 0.43357430908266625,
"learning_rate": 3.6266100093017975e-05,
"loss": 1.4369,
"step": 1158
},
{
"epoch": 0.20596205962059622,
"grad_norm": 0.4186399140051493,
"learning_rate": 3.625953414283426e-05,
"loss": 1.4115,
"step": 1159
},
{
"epoch": 0.20613976631569594,
"grad_norm": 0.4363301160509203,
"learning_rate": 3.6252963020186956e-05,
"loss": 1.3844,
"step": 1160
},
{
"epoch": 0.20631747301079567,
"grad_norm": 0.42287668020667984,
"learning_rate": 3.624638672716647e-05,
"loss": 1.3773,
"step": 1161
},
{
"epoch": 0.20649517970589543,
"grad_norm": 0.4187817093800454,
"learning_rate": 3.6239805265864837e-05,
"loss": 1.3842,
"step": 1162
},
{
"epoch": 0.20667288640099515,
"grad_norm": 0.4060300383333441,
"learning_rate": 3.623321863837575e-05,
"loss": 1.3682,
"step": 1163
},
{
"epoch": 0.2068505930960949,
"grad_norm": 0.4230565431334674,
"learning_rate": 3.622662684679453e-05,
"loss": 1.4128,
"step": 1164
},
{
"epoch": 0.20702829979119464,
"grad_norm": 0.4535601565465648,
"learning_rate": 3.622002989321815e-05,
"loss": 1.4617,
"step": 1165
},
{
"epoch": 0.20720600648629436,
"grad_norm": 0.4422346390344988,
"learning_rate": 3.621342777974524e-05,
"loss": 1.4209,
"step": 1166
},
{
"epoch": 0.20738371318139412,
"grad_norm": 0.4984460570898308,
"learning_rate": 3.620682050847604e-05,
"loss": 1.4148,
"step": 1167
},
{
"epoch": 0.20756141987649385,
"grad_norm": 0.4339110959998564,
"learning_rate": 3.620020808151246e-05,
"loss": 1.4385,
"step": 1168
},
{
"epoch": 0.20773912657159357,
"grad_norm": 0.4485639468512291,
"learning_rate": 3.6193590500958024e-05,
"loss": 1.4321,
"step": 1169
},
{
"epoch": 0.20791683326669333,
"grad_norm": 0.44181447909950217,
"learning_rate": 3.6186967768917916e-05,
"loss": 1.4402,
"step": 1170
},
{
"epoch": 0.20809453996179306,
"grad_norm": 0.48016138168226113,
"learning_rate": 3.6180339887498953e-05,
"loss": 1.4596,
"step": 1171
},
{
"epoch": 0.2082722466568928,
"grad_norm": 0.4299062776323167,
"learning_rate": 3.617370685880959e-05,
"loss": 1.4248,
"step": 1172
},
{
"epoch": 0.20844995335199254,
"grad_norm": 0.4693714347876249,
"learning_rate": 3.616706868495991e-05,
"loss": 1.4026,
"step": 1173
},
{
"epoch": 0.20862766004709227,
"grad_norm": 0.4250949965681123,
"learning_rate": 3.616042536806164e-05,
"loss": 1.4551,
"step": 1174
},
{
"epoch": 0.20880536674219202,
"grad_norm": 0.5009015072792162,
"learning_rate": 3.615377691022816e-05,
"loss": 1.4352,
"step": 1175
},
{
"epoch": 0.20898307343729175,
"grad_norm": 0.4189644426644193,
"learning_rate": 3.614712331357446e-05,
"loss": 1.3868,
"step": 1176
},
{
"epoch": 0.20916078013239148,
"grad_norm": 0.44665661832487225,
"learning_rate": 3.614046458021717e-05,
"loss": 1.4475,
"step": 1177
},
{
"epoch": 0.20933848682749123,
"grad_norm": 0.4437631757440217,
"learning_rate": 3.6133800712274555e-05,
"loss": 1.4317,
"step": 1178
},
{
"epoch": 0.20951619352259096,
"grad_norm": 0.5287024185398435,
"learning_rate": 3.612713171186653e-05,
"loss": 1.4472,
"step": 1179
},
{
"epoch": 0.20969390021769072,
"grad_norm": 0.4285863865384925,
"learning_rate": 3.612045758111463e-05,
"loss": 1.4011,
"step": 1180
},
{
"epoch": 0.20987160691279044,
"grad_norm": 0.4288702138602202,
"learning_rate": 3.6113778322142005e-05,
"loss": 1.4121,
"step": 1181
},
{
"epoch": 0.21004931360789017,
"grad_norm": 0.45272467255722865,
"learning_rate": 3.610709393707347e-05,
"loss": 1.4066,
"step": 1182
},
{
"epoch": 0.21022702030298993,
"grad_norm": 0.43679236478664973,
"learning_rate": 3.6100404428035444e-05,
"loss": 1.4025,
"step": 1183
},
{
"epoch": 0.21040472699808965,
"grad_norm": 0.4388088804168485,
"learning_rate": 3.609370979715598e-05,
"loss": 1.4255,
"step": 1184
},
{
"epoch": 0.21058243369318938,
"grad_norm": 0.4783645189240892,
"learning_rate": 3.608701004656478e-05,
"loss": 1.4376,
"step": 1185
},
{
"epoch": 0.21076014038828914,
"grad_norm": 0.43981620785117215,
"learning_rate": 3.608030517839315e-05,
"loss": 1.4228,
"step": 1186
},
{
"epoch": 0.21093784708338886,
"grad_norm": 0.4210881179095043,
"learning_rate": 3.6073595194774046e-05,
"loss": 1.4011,
"step": 1187
},
{
"epoch": 0.21111555377848862,
"grad_norm": 0.4267775930760924,
"learning_rate": 3.606688009784203e-05,
"loss": 1.4527,
"step": 1188
},
{
"epoch": 0.21129326047358835,
"grad_norm": 0.42631874365630096,
"learning_rate": 3.6060159889733307e-05,
"loss": 1.4357,
"step": 1189
},
{
"epoch": 0.21147096716868807,
"grad_norm": 0.4199547392318526,
"learning_rate": 3.6053434572585696e-05,
"loss": 1.4481,
"step": 1190
},
{
"epoch": 0.21164867386378783,
"grad_norm": 0.41649718059632407,
"learning_rate": 3.6046704148538645e-05,
"loss": 1.3996,
"step": 1191
},
{
"epoch": 0.21182638055888756,
"grad_norm": 0.41336486583078785,
"learning_rate": 3.6039968619733234e-05,
"loss": 1.4377,
"step": 1192
},
{
"epoch": 0.21200408725398728,
"grad_norm": 0.43479183622855827,
"learning_rate": 3.603322798831216e-05,
"loss": 1.4572,
"step": 1193
},
{
"epoch": 0.21218179394908704,
"grad_norm": 0.41163007628889786,
"learning_rate": 3.602648225641975e-05,
"loss": 1.3785,
"step": 1194
},
{
"epoch": 0.21235950064418677,
"grad_norm": 0.4108953956663432,
"learning_rate": 3.6019731426201936e-05,
"loss": 1.4305,
"step": 1195
},
{
"epoch": 0.21253720733928652,
"grad_norm": 0.41439602303893464,
"learning_rate": 3.601297549980629e-05,
"loss": 1.4385,
"step": 1196
},
{
"epoch": 0.21271491403438625,
"grad_norm": 0.45014221047477554,
"learning_rate": 3.600621447938201e-05,
"loss": 1.4039,
"step": 1197
},
{
"epoch": 0.21289262072948598,
"grad_norm": 0.41297775375136125,
"learning_rate": 3.5999448367079886e-05,
"loss": 1.4212,
"step": 1198
},
{
"epoch": 0.21307032742458573,
"grad_norm": 0.7210832585394966,
"learning_rate": 3.5992677165052354e-05,
"loss": 1.4186,
"step": 1199
},
{
"epoch": 0.21324803411968546,
"grad_norm": 0.44656746977476314,
"learning_rate": 3.598590087545346e-05,
"loss": 1.415,
"step": 1200
},
{
"epoch": 0.2134257408147852,
"grad_norm": 0.4125182829943099,
"learning_rate": 3.597911950043887e-05,
"loss": 1.4368,
"step": 1201
},
{
"epoch": 0.21360344750988494,
"grad_norm": 0.42688332753485914,
"learning_rate": 3.597233304216587e-05,
"loss": 1.4253,
"step": 1202
},
{
"epoch": 0.21378115420498467,
"grad_norm": 0.41505573657851175,
"learning_rate": 3.596554150279334e-05,
"loss": 1.3809,
"step": 1203
},
{
"epoch": 0.21395886090008442,
"grad_norm": 0.6587188819392091,
"learning_rate": 3.595874488448183e-05,
"loss": 1.3921,
"step": 1204
},
{
"epoch": 0.21413656759518415,
"grad_norm": 0.411137584196251,
"learning_rate": 3.5951943189393445e-05,
"loss": 1.4007,
"step": 1205
},
{
"epoch": 0.21431427429028388,
"grad_norm": 0.4164190875087277,
"learning_rate": 3.5945136419691945e-05,
"loss": 1.4163,
"step": 1206
},
{
"epoch": 0.21449198098538363,
"grad_norm": 0.42725129515696025,
"learning_rate": 3.593832457754269e-05,
"loss": 1.4171,
"step": 1207
},
{
"epoch": 0.21466968768048336,
"grad_norm": 0.4115782209635488,
"learning_rate": 3.593150766511265e-05,
"loss": 1.4043,
"step": 1208
},
{
"epoch": 0.2148473943755831,
"grad_norm": 0.42680655179517346,
"learning_rate": 3.592468568457042e-05,
"loss": 1.4625,
"step": 1209
},
{
"epoch": 0.21502510107068284,
"grad_norm": 0.4168510203765361,
"learning_rate": 3.591785863808619e-05,
"loss": 1.4299,
"step": 1210
},
{
"epoch": 0.21520280776578257,
"grad_norm": 0.40979466771838496,
"learning_rate": 3.5911026527831786e-05,
"loss": 1.4232,
"step": 1211
},
{
"epoch": 0.21538051446088233,
"grad_norm": 0.5333306968959777,
"learning_rate": 3.590418935598062e-05,
"loss": 1.4425,
"step": 1212
},
{
"epoch": 0.21555822115598205,
"grad_norm": 0.4558854633088275,
"learning_rate": 3.5897347124707734e-05,
"loss": 1.3903,
"step": 1213
},
{
"epoch": 0.21573592785108178,
"grad_norm": 0.4295791985194421,
"learning_rate": 3.5890499836189755e-05,
"loss": 1.4018,
"step": 1214
},
{
"epoch": 0.21591363454618154,
"grad_norm": 0.4333912871309332,
"learning_rate": 3.588364749260495e-05,
"loss": 1.4391,
"step": 1215
},
{
"epoch": 0.21609134124128127,
"grad_norm": 0.43213220791343854,
"learning_rate": 3.587679009613317e-05,
"loss": 1.422,
"step": 1216
},
{
"epoch": 0.216269047936381,
"grad_norm": 0.4033611631052586,
"learning_rate": 3.5869927648955886e-05,
"loss": 1.3919,
"step": 1217
},
{
"epoch": 0.21644675463148075,
"grad_norm": 0.420725939576407,
"learning_rate": 3.586306015325616e-05,
"loss": 1.4101,
"step": 1218
},
{
"epoch": 0.21662446132658048,
"grad_norm": 0.4141696059837607,
"learning_rate": 3.585618761121869e-05,
"loss": 1.4145,
"step": 1219
},
{
"epoch": 0.21680216802168023,
"grad_norm": 1.2074754450956195,
"learning_rate": 3.584931002502975e-05,
"loss": 1.4257,
"step": 1220
},
{
"epoch": 0.21697987471677996,
"grad_norm": 0.4472549672251518,
"learning_rate": 3.5842427396877235e-05,
"loss": 1.4179,
"step": 1221
},
{
"epoch": 0.21715758141187969,
"grad_norm": 0.45828518736790463,
"learning_rate": 3.583553972895063e-05,
"loss": 1.4768,
"step": 1222
},
{
"epoch": 0.21733528810697944,
"grad_norm": 0.4326027341024743,
"learning_rate": 3.582864702344104e-05,
"loss": 1.3948,
"step": 1223
},
{
"epoch": 0.21751299480207917,
"grad_norm": 0.4388141321535259,
"learning_rate": 3.582174928254116e-05,
"loss": 1.4487,
"step": 1224
},
{
"epoch": 0.2176907014971789,
"grad_norm": 0.4403750923551395,
"learning_rate": 3.581484650844528e-05,
"loss": 1.4529,
"step": 1225
},
{
"epoch": 0.21786840819227865,
"grad_norm": 0.420972575084007,
"learning_rate": 3.580793870334933e-05,
"loss": 1.4374,
"step": 1226
},
{
"epoch": 0.21804611488737838,
"grad_norm": 0.4214436601582466,
"learning_rate": 3.58010258694508e-05,
"loss": 1.3888,
"step": 1227
},
{
"epoch": 0.21822382158247813,
"grad_norm": 0.42420946955250544,
"learning_rate": 3.579410800894877e-05,
"loss": 1.409,
"step": 1228
},
{
"epoch": 0.21840152827757786,
"grad_norm": 0.43561251042108373,
"learning_rate": 3.578718512404398e-05,
"loss": 1.3912,
"step": 1229
},
{
"epoch": 0.2185792349726776,
"grad_norm": 0.41325786536811815,
"learning_rate": 3.578025721693869e-05,
"loss": 1.4108,
"step": 1230
},
{
"epoch": 0.21875694166777734,
"grad_norm": 0.4309168564160067,
"learning_rate": 3.577332428983684e-05,
"loss": 1.4577,
"step": 1231
},
{
"epoch": 0.21893464836287707,
"grad_norm": 0.4283338502460869,
"learning_rate": 3.576638634494389e-05,
"loss": 1.4343,
"step": 1232
},
{
"epoch": 0.2191123550579768,
"grad_norm": 0.45146052681812293,
"learning_rate": 3.5759443384466946e-05,
"loss": 1.4759,
"step": 1233
},
{
"epoch": 0.21929006175307655,
"grad_norm": 0.41256816081537157,
"learning_rate": 3.575249541061469e-05,
"loss": 1.4402,
"step": 1234
},
{
"epoch": 0.21946776844817628,
"grad_norm": 0.42181710496620006,
"learning_rate": 3.574554242559742e-05,
"loss": 1.4304,
"step": 1235
},
{
"epoch": 0.21964547514327604,
"grad_norm": 0.4214228474552675,
"learning_rate": 3.573858443162698e-05,
"loss": 1.3918,
"step": 1236
},
{
"epoch": 0.21982318183837576,
"grad_norm": 0.45323747725052366,
"learning_rate": 3.573162143091685e-05,
"loss": 1.4033,
"step": 1237
},
{
"epoch": 0.2200008885334755,
"grad_norm": 0.41567206049943956,
"learning_rate": 3.5724653425682105e-05,
"loss": 1.3661,
"step": 1238
},
{
"epoch": 0.22017859522857525,
"grad_norm": 0.41548506106839833,
"learning_rate": 3.57176804181394e-05,
"loss": 1.3778,
"step": 1239
},
{
"epoch": 0.22035630192367497,
"grad_norm": 0.45256017524097486,
"learning_rate": 3.571070241050695e-05,
"loss": 1.4387,
"step": 1240
},
{
"epoch": 0.2205340086187747,
"grad_norm": 0.4535535765828799,
"learning_rate": 3.570371940500462e-05,
"loss": 1.3897,
"step": 1241
},
{
"epoch": 0.22071171531387446,
"grad_norm": 0.4237809032901047,
"learning_rate": 3.569673140385383e-05,
"loss": 1.4222,
"step": 1242
},
{
"epoch": 0.22088942200897418,
"grad_norm": 0.409349994952873,
"learning_rate": 3.568973840927759e-05,
"loss": 1.384,
"step": 1243
},
{
"epoch": 0.22106712870407394,
"grad_norm": 0.4346066422118836,
"learning_rate": 3.5682740423500494e-05,
"loss": 1.4186,
"step": 1244
},
{
"epoch": 0.22124483539917367,
"grad_norm": 0.41077879103023524,
"learning_rate": 3.567573744874874e-05,
"loss": 1.4068,
"step": 1245
},
{
"epoch": 0.2214225420942734,
"grad_norm": 0.4678068813251344,
"learning_rate": 3.5668729487250125e-05,
"loss": 1.4012,
"step": 1246
},
{
"epoch": 0.22160024878937315,
"grad_norm": 0.4071283710576322,
"learning_rate": 3.5661716541233984e-05,
"loss": 1.4224,
"step": 1247
},
{
"epoch": 0.22177795548447288,
"grad_norm": 0.4232367119547039,
"learning_rate": 3.565469861293128e-05,
"loss": 1.4008,
"step": 1248
},
{
"epoch": 0.2219556621795726,
"grad_norm": 0.41469081860433954,
"learning_rate": 3.564767570457455e-05,
"loss": 1.4528,
"step": 1249
},
{
"epoch": 0.22213336887467236,
"grad_norm": 0.4419502966591746,
"learning_rate": 3.564064781839791e-05,
"loss": 1.4051,
"step": 1250
},
{
"epoch": 0.2223110755697721,
"grad_norm": 0.41406326812956407,
"learning_rate": 3.563361495663706e-05,
"loss": 1.4256,
"step": 1251
},
{
"epoch": 0.22248878226487184,
"grad_norm": 0.4644774558570271,
"learning_rate": 3.56265771215293e-05,
"loss": 1.3785,
"step": 1252
},
{
"epoch": 0.22266648895997157,
"grad_norm": 0.4041011589241617,
"learning_rate": 3.5619534315313476e-05,
"loss": 1.3784,
"step": 1253
},
{
"epoch": 0.2228441956550713,
"grad_norm": 0.4345844961982715,
"learning_rate": 3.561248654023005e-05,
"loss": 1.3985,
"step": 1254
},
{
"epoch": 0.22302190235017105,
"grad_norm": 0.4008466107529262,
"learning_rate": 3.5605433798521046e-05,
"loss": 1.3514,
"step": 1255
},
{
"epoch": 0.22319960904527078,
"grad_norm": 0.43531834074557524,
"learning_rate": 3.559837609243008e-05,
"loss": 1.4021,
"step": 1256
},
{
"epoch": 0.2233773157403705,
"grad_norm": 0.44784478385241305,
"learning_rate": 3.559131342420235e-05,
"loss": 1.3722,
"step": 1257
},
{
"epoch": 0.22355502243547026,
"grad_norm": 0.5563381171526215,
"learning_rate": 3.5584245796084593e-05,
"loss": 1.4375,
"step": 1258
},
{
"epoch": 0.22373272913057,
"grad_norm": 0.4247835822460256,
"learning_rate": 3.557717321032519e-05,
"loss": 1.4526,
"step": 1259
},
{
"epoch": 0.22391043582566975,
"grad_norm": 0.9606577871605956,
"learning_rate": 3.557009566917403e-05,
"loss": 1.4316,
"step": 1260
},
{
"epoch": 0.22408814252076947,
"grad_norm": 0.45767138716777445,
"learning_rate": 3.556301317488264e-05,
"loss": 1.3979,
"step": 1261
},
{
"epoch": 0.2242658492158692,
"grad_norm": 0.4320209516650993,
"learning_rate": 3.555592572970408e-05,
"loss": 1.3959,
"step": 1262
},
{
"epoch": 0.22444355591096896,
"grad_norm": 0.4518357313945011,
"learning_rate": 3.554883333589301e-05,
"loss": 1.4311,
"step": 1263
},
{
"epoch": 0.22462126260606868,
"grad_norm": 0.5663265558376351,
"learning_rate": 3.5541735995705635e-05,
"loss": 1.3484,
"step": 1264
},
{
"epoch": 0.2247989693011684,
"grad_norm": 0.4592704509814157,
"learning_rate": 3.553463371139978e-05,
"loss": 1.4326,
"step": 1265
},
{
"epoch": 0.22497667599626817,
"grad_norm": 0.4443304508745431,
"learning_rate": 3.552752648523478e-05,
"loss": 1.4187,
"step": 1266
},
{
"epoch": 0.2251543826913679,
"grad_norm": 0.43380266644289006,
"learning_rate": 3.552041431947161e-05,
"loss": 1.4241,
"step": 1267
},
{
"epoch": 0.22533208938646765,
"grad_norm": 0.41739874487269,
"learning_rate": 3.551329721637277e-05,
"loss": 1.3832,
"step": 1268
},
{
"epoch": 0.22550979608156738,
"grad_norm": 0.4353137981126273,
"learning_rate": 3.550617517820234e-05,
"loss": 1.3836,
"step": 1269
},
{
"epoch": 0.2256875027766671,
"grad_norm": 0.4229798435659934,
"learning_rate": 3.549904820722598e-05,
"loss": 1.3979,
"step": 1270
},
{
"epoch": 0.22586520947176686,
"grad_norm": 0.42429173306338686,
"learning_rate": 3.549191630571091e-05,
"loss": 1.4455,
"step": 1271
},
{
"epoch": 0.2260429161668666,
"grad_norm": 0.4123675389665796,
"learning_rate": 3.548477947592593e-05,
"loss": 1.3818,
"step": 1272
},
{
"epoch": 0.22622062286196631,
"grad_norm": 0.42763291967742956,
"learning_rate": 3.5477637720141396e-05,
"loss": 1.4733,
"step": 1273
},
{
"epoch": 0.22639832955706607,
"grad_norm": 0.40756892020456703,
"learning_rate": 3.547049104062923e-05,
"loss": 1.4362,
"step": 1274
},
{
"epoch": 0.2265760362521658,
"grad_norm": 0.4327492402187981,
"learning_rate": 3.5463339439662924e-05,
"loss": 1.391,
"step": 1275
},
{
"epoch": 0.22675374294726555,
"grad_norm": 0.3986156265509585,
"learning_rate": 3.5456182919517546e-05,
"loss": 1.385,
"step": 1276
},
{
"epoch": 0.22693144964236528,
"grad_norm": 0.4058757820583695,
"learning_rate": 3.544902148246972e-05,
"loss": 1.4115,
"step": 1277
},
{
"epoch": 0.227109156337465,
"grad_norm": 0.41268112385834427,
"learning_rate": 3.5441855130797615e-05,
"loss": 1.3958,
"step": 1278
},
{
"epoch": 0.22728686303256476,
"grad_norm": 0.4280779514909465,
"learning_rate": 3.5434683866781e-05,
"loss": 1.3982,
"step": 1279
},
{
"epoch": 0.2274645697276645,
"grad_norm": 0.4213240669350423,
"learning_rate": 3.54275076927012e-05,
"loss": 1.4551,
"step": 1280
},
{
"epoch": 0.22764227642276422,
"grad_norm": 0.409113058809916,
"learning_rate": 3.542032661084106e-05,
"loss": 1.4236,
"step": 1281
},
{
"epoch": 0.22781998311786397,
"grad_norm": 0.40707874005765143,
"learning_rate": 3.541314062348503e-05,
"loss": 1.4073,
"step": 1282
},
{
"epoch": 0.2279976898129637,
"grad_norm": 0.4131893671064229,
"learning_rate": 3.5405949732919124e-05,
"loss": 1.409,
"step": 1283
},
{
"epoch": 0.22817539650806345,
"grad_norm": 0.4051551673167467,
"learning_rate": 3.5398753941430875e-05,
"loss": 1.4336,
"step": 1284
},
{
"epoch": 0.22835310320316318,
"grad_norm": 0.42517677014915223,
"learning_rate": 3.539155325130942e-05,
"loss": 1.409,
"step": 1285
},
{
"epoch": 0.2285308098982629,
"grad_norm": 0.4221567198367165,
"learning_rate": 3.538434766484542e-05,
"loss": 1.4492,
"step": 1286
},
{
"epoch": 0.22870851659336267,
"grad_norm": 0.4231102722126546,
"learning_rate": 3.5377137184331105e-05,
"loss": 1.4553,
"step": 1287
},
{
"epoch": 0.2288862232884624,
"grad_norm": 0.4413978107571921,
"learning_rate": 3.536992181206028e-05,
"loss": 1.4943,
"step": 1288
},
{
"epoch": 0.22906392998356212,
"grad_norm": 0.41329693933303224,
"learning_rate": 3.536270155032828e-05,
"loss": 1.3834,
"step": 1289
},
{
"epoch": 0.22924163667866188,
"grad_norm": 0.8239021892480644,
"learning_rate": 3.535547640143201e-05,
"loss": 1.4005,
"step": 1290
},
{
"epoch": 0.2294193433737616,
"grad_norm": 0.4092518709236838,
"learning_rate": 3.5348246367669925e-05,
"loss": 1.3854,
"step": 1291
},
{
"epoch": 0.22959705006886136,
"grad_norm": 0.40808675000652667,
"learning_rate": 3.534101145134203e-05,
"loss": 1.3739,
"step": 1292
},
{
"epoch": 0.22977475676396109,
"grad_norm": 0.4191932832386246,
"learning_rate": 3.533377165474989e-05,
"loss": 1.3788,
"step": 1293
},
{
"epoch": 0.2299524634590608,
"grad_norm": 0.4108451758888158,
"learning_rate": 3.532652698019662e-05,
"loss": 1.3912,
"step": 1294
},
{
"epoch": 0.23013017015416057,
"grad_norm": 0.4270583586338924,
"learning_rate": 3.53192774299869e-05,
"loss": 1.4522,
"step": 1295
},
{
"epoch": 0.2303078768492603,
"grad_norm": 0.41820657318974835,
"learning_rate": 3.531202300642693e-05,
"loss": 1.4141,
"step": 1296
},
{
"epoch": 0.23048558354436002,
"grad_norm": 0.43880577188165293,
"learning_rate": 3.530476371182449e-05,
"loss": 1.3964,
"step": 1297
},
{
"epoch": 0.23066329023945978,
"grad_norm": 0.40605237872879885,
"learning_rate": 3.5297499548488896e-05,
"loss": 1.4231,
"step": 1298
},
{
"epoch": 0.2308409969345595,
"grad_norm": 0.41971274546058,
"learning_rate": 3.5290230518731005e-05,
"loss": 1.4172,
"step": 1299
},
{
"epoch": 0.23101870362965926,
"grad_norm": 0.4042860861024759,
"learning_rate": 3.5282956624863246e-05,
"loss": 1.3853,
"step": 1300
},
{
"epoch": 0.231196410324759,
"grad_norm": 0.39292435787956914,
"learning_rate": 3.527567786919957e-05,
"loss": 1.3735,
"step": 1301
},
{
"epoch": 0.23137411701985872,
"grad_norm": 0.41058229559063725,
"learning_rate": 3.52683942540555e-05,
"loss": 1.3993,
"step": 1302
},
{
"epoch": 0.23155182371495847,
"grad_norm": 0.4166459797153265,
"learning_rate": 3.526110578174808e-05,
"loss": 1.3832,
"step": 1303
},
{
"epoch": 0.2317295304100582,
"grad_norm": 0.3993184425974253,
"learning_rate": 3.525381245459591e-05,
"loss": 1.405,
"step": 1304
},
{
"epoch": 0.23190723710515793,
"grad_norm": 0.4258275457274125,
"learning_rate": 3.524651427491914e-05,
"loss": 1.3947,
"step": 1305
},
{
"epoch": 0.23208494380025768,
"grad_norm": 0.398953553045706,
"learning_rate": 3.523921124503946e-05,
"loss": 1.3855,
"step": 1306
},
{
"epoch": 0.2322626504953574,
"grad_norm": 0.42332360690948206,
"learning_rate": 3.523190336728009e-05,
"loss": 1.4149,
"step": 1307
},
{
"epoch": 0.23244035719045716,
"grad_norm": 0.4198901694210052,
"learning_rate": 3.522459064396581e-05,
"loss": 1.3872,
"step": 1308
},
{
"epoch": 0.2326180638855569,
"grad_norm": 0.5502553505608405,
"learning_rate": 3.521727307742294e-05,
"loss": 1.4515,
"step": 1309
},
{
"epoch": 0.23279577058065662,
"grad_norm": 0.427356814468813,
"learning_rate": 3.520995066997932e-05,
"loss": 1.4451,
"step": 1310
},
{
"epoch": 0.23297347727575637,
"grad_norm": 0.40852838188384516,
"learning_rate": 3.520262342396437e-05,
"loss": 1.3789,
"step": 1311
},
{
"epoch": 0.2331511839708561,
"grad_norm": 0.42910805234691424,
"learning_rate": 3.5195291341709e-05,
"loss": 1.3888,
"step": 1312
},
{
"epoch": 0.23332889066595583,
"grad_norm": 0.4811463994160335,
"learning_rate": 3.51879544255457e-05,
"loss": 1.4176,
"step": 1313
},
{
"epoch": 0.23350659736105558,
"grad_norm": 0.4198015585565707,
"learning_rate": 3.518061267780847e-05,
"loss": 1.4227,
"step": 1314
},
{
"epoch": 0.2336843040561553,
"grad_norm": 0.4126591267226317,
"learning_rate": 3.517326610083286e-05,
"loss": 1.4247,
"step": 1315
},
{
"epoch": 0.23386201075125507,
"grad_norm": 0.43890661601442504,
"learning_rate": 3.516591469695597e-05,
"loss": 1.3904,
"step": 1316
},
{
"epoch": 0.2340397174463548,
"grad_norm": 0.4133112377442189,
"learning_rate": 3.51585584685164e-05,
"loss": 1.3953,
"step": 1317
},
{
"epoch": 0.23421742414145452,
"grad_norm": 0.45327758686735065,
"learning_rate": 3.515119741785431e-05,
"loss": 1.4733,
"step": 1318
},
{
"epoch": 0.23439513083655428,
"grad_norm": 0.4032816451984864,
"learning_rate": 3.514383154731139e-05,
"loss": 1.3941,
"step": 1319
},
{
"epoch": 0.234572837531654,
"grad_norm": 0.4172081816955108,
"learning_rate": 3.513646085923086e-05,
"loss": 1.392,
"step": 1320
},
{
"epoch": 0.23475054422675373,
"grad_norm": 0.43340167729422235,
"learning_rate": 3.5129085355957486e-05,
"loss": 1.4473,
"step": 1321
},
{
"epoch": 0.2349282509218535,
"grad_norm": 0.4061183634901522,
"learning_rate": 3.512170503983754e-05,
"loss": 1.3956,
"step": 1322
},
{
"epoch": 0.23510595761695322,
"grad_norm": 0.43256394358459643,
"learning_rate": 3.5114319913218844e-05,
"loss": 1.4421,
"step": 1323
},
{
"epoch": 0.23528366431205297,
"grad_norm": 0.4796382751089028,
"learning_rate": 3.510692997845074e-05,
"loss": 1.3881,
"step": 1324
},
{
"epoch": 0.2354613710071527,
"grad_norm": 0.5551151272008323,
"learning_rate": 3.509953523788412e-05,
"loss": 1.4287,
"step": 1325
},
{
"epoch": 0.23563907770225243,
"grad_norm": 0.4217063547117016,
"learning_rate": 3.5092135693871384e-05,
"loss": 1.3775,
"step": 1326
},
{
"epoch": 0.23581678439735218,
"grad_norm": 0.4228264880871538,
"learning_rate": 3.508473134876646e-05,
"loss": 1.4384,
"step": 1327
},
{
"epoch": 0.2359944910924519,
"grad_norm": 0.42411027917039257,
"learning_rate": 3.5077322204924806e-05,
"loss": 1.3874,
"step": 1328
},
{
"epoch": 0.23617219778755164,
"grad_norm": 0.4328744760158063,
"learning_rate": 3.506990826470342e-05,
"loss": 1.442,
"step": 1329
},
{
"epoch": 0.2363499044826514,
"grad_norm": 0.41280661822172604,
"learning_rate": 3.5062489530460816e-05,
"loss": 1.4138,
"step": 1330
},
{
"epoch": 0.23652761117775112,
"grad_norm": 0.417238301514077,
"learning_rate": 3.5055066004557027e-05,
"loss": 1.4185,
"step": 1331
},
{
"epoch": 0.23670531787285087,
"grad_norm": 0.42137641201577586,
"learning_rate": 3.504763768935362e-05,
"loss": 1.3977,
"step": 1332
},
{
"epoch": 0.2368830245679506,
"grad_norm": 0.4071250955211016,
"learning_rate": 3.504020458721368e-05,
"loss": 1.3751,
"step": 1333
},
{
"epoch": 0.23706073126305033,
"grad_norm": 0.4112842447077109,
"learning_rate": 3.503276670050181e-05,
"loss": 1.3761,
"step": 1334
},
{
"epoch": 0.23723843795815008,
"grad_norm": 0.4082649528123595,
"learning_rate": 3.502532403158416e-05,
"loss": 1.3828,
"step": 1335
},
{
"epoch": 0.2374161446532498,
"grad_norm": 0.42810614913592565,
"learning_rate": 3.501787658282837e-05,
"loss": 1.4219,
"step": 1336
},
{
"epoch": 0.23759385134834954,
"grad_norm": 0.4196521470044464,
"learning_rate": 3.5010424356603614e-05,
"loss": 1.4429,
"step": 1337
},
{
"epoch": 0.2377715580434493,
"grad_norm": 0.4183773511423429,
"learning_rate": 3.5002967355280583e-05,
"loss": 1.4462,
"step": 1338
},
{
"epoch": 0.23794926473854902,
"grad_norm": 0.46808565764186855,
"learning_rate": 3.49955055812315e-05,
"loss": 1.39,
"step": 1339
},
{
"epoch": 0.23812697143364878,
"grad_norm": 0.4066648234130081,
"learning_rate": 3.498803903683008e-05,
"loss": 1.4593,
"step": 1340
},
{
"epoch": 0.2383046781287485,
"grad_norm": 0.3992629372153268,
"learning_rate": 3.4980567724451584e-05,
"loss": 1.413,
"step": 1341
},
{
"epoch": 0.23848238482384823,
"grad_norm": 0.3953104231647446,
"learning_rate": 3.497309164647277e-05,
"loss": 1.3961,
"step": 1342
},
{
"epoch": 0.238660091518948,
"grad_norm": 0.41028849354628794,
"learning_rate": 3.496561080527192e-05,
"loss": 1.4091,
"step": 1343
},
{
"epoch": 0.23883779821404771,
"grad_norm": 0.39844337650190265,
"learning_rate": 3.4958125203228834e-05,
"loss": 1.4242,
"step": 1344
},
{
"epoch": 0.23901550490914744,
"grad_norm": 0.4060888353517679,
"learning_rate": 3.4950634842724826e-05,
"loss": 1.3922,
"step": 1345
},
{
"epoch": 0.2391932116042472,
"grad_norm": 0.42281588157007005,
"learning_rate": 3.494313972614271e-05,
"loss": 1.4183,
"step": 1346
},
{
"epoch": 0.23937091829934692,
"grad_norm": 0.3914267580507275,
"learning_rate": 3.4935639855866835e-05,
"loss": 1.3444,
"step": 1347
},
{
"epoch": 0.23954862499444668,
"grad_norm": 0.40809915444593825,
"learning_rate": 3.4928135234283036e-05,
"loss": 1.4369,
"step": 1348
},
{
"epoch": 0.2397263316895464,
"grad_norm": 0.41275650832900634,
"learning_rate": 3.492062586377869e-05,
"loss": 1.4289,
"step": 1349
},
{
"epoch": 0.23990403838464613,
"grad_norm": 0.40668594982713024,
"learning_rate": 3.4913111746742653e-05,
"loss": 1.3863,
"step": 1350
},
{
"epoch": 0.2400817450797459,
"grad_norm": 0.41309812113853794,
"learning_rate": 3.490559288556532e-05,
"loss": 1.3775,
"step": 1351
},
{
"epoch": 0.24025945177484562,
"grad_norm": 0.4246994748953506,
"learning_rate": 3.4898069282638576e-05,
"loss": 1.4249,
"step": 1352
},
{
"epoch": 0.24043715846994534,
"grad_norm": 0.4307764105028778,
"learning_rate": 3.489054094035583e-05,
"loss": 1.49,
"step": 1353
},
{
"epoch": 0.2406148651650451,
"grad_norm": 0.8923821436103766,
"learning_rate": 3.4883007861111974e-05,
"loss": 1.4068,
"step": 1354
},
{
"epoch": 0.24079257186014483,
"grad_norm": 0.4546266349687631,
"learning_rate": 3.4875470047303436e-05,
"loss": 1.4135,
"step": 1355
},
{
"epoch": 0.24097027855524458,
"grad_norm": 0.4284288452730429,
"learning_rate": 3.4867927501328125e-05,
"loss": 1.4203,
"step": 1356
},
{
"epoch": 0.2411479852503443,
"grad_norm": 0.42672539977832674,
"learning_rate": 3.4860380225585475e-05,
"loss": 1.4335,
"step": 1357
},
{
"epoch": 0.24132569194544404,
"grad_norm": 0.4772138886314652,
"learning_rate": 3.4852828222476405e-05,
"loss": 1.3969,
"step": 1358
},
{
"epoch": 0.2415033986405438,
"grad_norm": 0.4083247611995249,
"learning_rate": 3.484527149440337e-05,
"loss": 1.4082,
"step": 1359
},
{
"epoch": 0.24168110533564352,
"grad_norm": 0.4112576880766219,
"learning_rate": 3.4837710043770286e-05,
"loss": 1.3966,
"step": 1360
},
{
"epoch": 0.24185881203074325,
"grad_norm": 0.4049685198565756,
"learning_rate": 3.483014387298261e-05,
"loss": 1.434,
"step": 1361
},
{
"epoch": 0.242036518725843,
"grad_norm": 0.4066657350843835,
"learning_rate": 3.482257298444727e-05,
"loss": 1.4325,
"step": 1362
},
{
"epoch": 0.24221422542094273,
"grad_norm": 0.3950827266173155,
"learning_rate": 3.481499738057271e-05,
"loss": 1.4091,
"step": 1363
},
{
"epoch": 0.24239193211604249,
"grad_norm": 0.4231016438260705,
"learning_rate": 3.480741706376887e-05,
"loss": 1.4067,
"step": 1364
},
{
"epoch": 0.2425696388111422,
"grad_norm": 0.40667835827041027,
"learning_rate": 3.479983203644721e-05,
"loss": 1.407,
"step": 1365
},
{
"epoch": 0.24274734550624194,
"grad_norm": 0.3826059969815751,
"learning_rate": 3.479224230102064e-05,
"loss": 1.3729,
"step": 1366
},
{
"epoch": 0.2429250522013417,
"grad_norm": 0.4303744066854862,
"learning_rate": 3.478464785990363e-05,
"loss": 1.4904,
"step": 1367
},
{
"epoch": 0.24310275889644142,
"grad_norm": 0.4125269060186455,
"learning_rate": 3.477704871551208e-05,
"loss": 1.4173,
"step": 1368
},
{
"epoch": 0.24328046559154115,
"grad_norm": 0.4630165580588375,
"learning_rate": 3.4769444870263456e-05,
"loss": 1.4038,
"step": 1369
},
{
"epoch": 0.2434581722866409,
"grad_norm": 0.40247418448391115,
"learning_rate": 3.476183632657666e-05,
"loss": 1.3843,
"step": 1370
},
{
"epoch": 0.24363587898174063,
"grad_norm": 0.41324466881790134,
"learning_rate": 3.4754223086872115e-05,
"loss": 1.4272,
"step": 1371
},
{
"epoch": 0.2438135856768404,
"grad_norm": 0.3942930584941148,
"learning_rate": 3.4746605153571746e-05,
"loss": 1.3692,
"step": 1372
},
{
"epoch": 0.24399129237194012,
"grad_norm": 0.4235182199460832,
"learning_rate": 3.473898252909895e-05,
"loss": 1.4708,
"step": 1373
},
{
"epoch": 0.24416899906703984,
"grad_norm": 0.39242009274884626,
"learning_rate": 3.473135521587864e-05,
"loss": 1.3431,
"step": 1374
},
{
"epoch": 0.2443467057621396,
"grad_norm": 0.41286319477494693,
"learning_rate": 3.472372321633719e-05,
"loss": 1.4658,
"step": 1375
},
{
"epoch": 0.24452441245723933,
"grad_norm": 0.4007752899907668,
"learning_rate": 3.4716086532902505e-05,
"loss": 1.4073,
"step": 1376
},
{
"epoch": 0.24470211915233905,
"grad_norm": 0.4079407990157625,
"learning_rate": 3.470844516800394e-05,
"loss": 1.4131,
"step": 1377
},
{
"epoch": 0.2448798258474388,
"grad_norm": 0.40215981930493233,
"learning_rate": 3.4700799124072365e-05,
"loss": 1.4172,
"step": 1378
},
{
"epoch": 0.24505753254253854,
"grad_norm": 0.4020622968662787,
"learning_rate": 3.469314840354012e-05,
"loss": 1.3953,
"step": 1379
},
{
"epoch": 0.2452352392376383,
"grad_norm": 0.40720156057449114,
"learning_rate": 3.468549300884106e-05,
"loss": 1.4054,
"step": 1380
},
{
"epoch": 0.24541294593273802,
"grad_norm": 0.40217170380296496,
"learning_rate": 3.467783294241049e-05,
"loss": 1.3895,
"step": 1381
},
{
"epoch": 0.24559065262783775,
"grad_norm": 0.38997974439453087,
"learning_rate": 3.467016820668524e-05,
"loss": 1.3984,
"step": 1382
},
{
"epoch": 0.2457683593229375,
"grad_norm": 0.4201054404870361,
"learning_rate": 3.46624988041036e-05,
"loss": 1.4437,
"step": 1383
},
{
"epoch": 0.24594606601803723,
"grad_norm": 0.4121553377990555,
"learning_rate": 3.465482473710534e-05,
"loss": 1.433,
"step": 1384
},
{
"epoch": 0.24612377271313696,
"grad_norm": 0.40251823071085335,
"learning_rate": 3.464714600813174e-05,
"loss": 1.4056,
"step": 1385
},
{
"epoch": 0.2463014794082367,
"grad_norm": 0.4174810636428583,
"learning_rate": 3.463946261962555e-05,
"loss": 1.4029,
"step": 1386
},
{
"epoch": 0.24647918610333644,
"grad_norm": 0.39852212960883077,
"learning_rate": 3.463177457403099e-05,
"loss": 1.3733,
"step": 1387
},
{
"epoch": 0.2466568927984362,
"grad_norm": 0.4039618677196372,
"learning_rate": 3.462408187379377e-05,
"loss": 1.4041,
"step": 1388
},
{
"epoch": 0.24683459949353592,
"grad_norm": 0.39107187607578137,
"learning_rate": 3.461638452136109e-05,
"loss": 1.3969,
"step": 1389
},
{
"epoch": 0.24701230618863565,
"grad_norm": 0.391869741711078,
"learning_rate": 3.460868251918162e-05,
"loss": 1.3565,
"step": 1390
},
{
"epoch": 0.2471900128837354,
"grad_norm": 0.402016897183356,
"learning_rate": 3.460097586970551e-05,
"loss": 1.4079,
"step": 1391
},
{
"epoch": 0.24736771957883513,
"grad_norm": 0.3984762953900053,
"learning_rate": 3.45932645753844e-05,
"loss": 1.4277,
"step": 1392
},
{
"epoch": 0.24754542627393486,
"grad_norm": 0.6125189325509154,
"learning_rate": 3.458554863867139e-05,
"loss": 1.4053,
"step": 1393
},
{
"epoch": 0.24772313296903462,
"grad_norm": 0.404234847544222,
"learning_rate": 3.457782806202105e-05,
"loss": 1.4102,
"step": 1394
},
{
"epoch": 0.24790083966413434,
"grad_norm": 0.40700028030996593,
"learning_rate": 3.457010284788947e-05,
"loss": 1.4087,
"step": 1395
},
{
"epoch": 0.2480785463592341,
"grad_norm": 0.3969254779534257,
"learning_rate": 3.456237299873416e-05,
"loss": 1.3924,
"step": 1396
},
{
"epoch": 0.24825625305433383,
"grad_norm": 0.4125955257872514,
"learning_rate": 3.4554638517014146e-05,
"loss": 1.3965,
"step": 1397
},
{
"epoch": 0.24843395974943355,
"grad_norm": 0.4011467457330639,
"learning_rate": 3.454689940518991e-05,
"loss": 1.4343,
"step": 1398
},
{
"epoch": 0.2486116664445333,
"grad_norm": 0.3953963427371629,
"learning_rate": 3.453915566572341e-05,
"loss": 1.3814,
"step": 1399
},
{
"epoch": 0.24878937313963304,
"grad_norm": 0.41029454043957303,
"learning_rate": 3.4531407301078056e-05,
"loss": 1.412,
"step": 1400
},
{
"epoch": 0.24896707983473276,
"grad_norm": 0.4119264000562141,
"learning_rate": 3.452365431371878e-05,
"loss": 1.458,
"step": 1401
},
{
"epoch": 0.24914478652983252,
"grad_norm": 0.40954146702714184,
"learning_rate": 3.451589670611193e-05,
"loss": 1.453,
"step": 1402
},
{
"epoch": 0.24932249322493225,
"grad_norm": 0.4103030083489122,
"learning_rate": 3.450813448072536e-05,
"loss": 1.4067,
"step": 1403
},
{
"epoch": 0.249500199920032,
"grad_norm": 0.4143225260793054,
"learning_rate": 3.450036764002837e-05,
"loss": 1.4084,
"step": 1404
},
{
"epoch": 0.24967790661513173,
"grad_norm": 0.40986839325941216,
"learning_rate": 3.449259618649174e-05,
"loss": 1.416,
"step": 1405
},
{
"epoch": 0.24985561331023146,
"grad_norm": 0.40010907679641156,
"learning_rate": 3.448482012258772e-05,
"loss": 1.4321,
"step": 1406
},
{
"epoch": 0.2500333200053312,
"grad_norm": 0.4209601630068639,
"learning_rate": 3.4477039450790015e-05,
"loss": 1.3935,
"step": 1407
},
{
"epoch": 0.25021102670043094,
"grad_norm": 0.4097846355318961,
"learning_rate": 3.4469254173573815e-05,
"loss": 1.4304,
"step": 1408
},
{
"epoch": 0.2503887333955307,
"grad_norm": 0.4273244974183935,
"learning_rate": 3.446146429341575e-05,
"loss": 1.446,
"step": 1409
},
{
"epoch": 0.2505664400906304,
"grad_norm": 0.40009954889644894,
"learning_rate": 3.445366981279394e-05,
"loss": 1.4104,
"step": 1410
},
{
"epoch": 0.25074414678573015,
"grad_norm": 0.42036735604879694,
"learning_rate": 3.4445870734187945e-05,
"loss": 1.4342,
"step": 1411
},
{
"epoch": 0.2509218534808299,
"grad_norm": 0.39697668437600453,
"learning_rate": 3.4438067060078795e-05,
"loss": 1.3581,
"step": 1412
},
{
"epoch": 0.2510995601759296,
"grad_norm": 0.46576563504113877,
"learning_rate": 3.4430258792949006e-05,
"loss": 1.3605,
"step": 1413
},
{
"epoch": 0.25127726687102936,
"grad_norm": 0.4004807035457663,
"learning_rate": 3.442244593528251e-05,
"loss": 1.3587,
"step": 1414
},
{
"epoch": 0.2514549735661291,
"grad_norm": 0.41489939685491206,
"learning_rate": 3.4414628489564746e-05,
"loss": 1.3855,
"step": 1415
},
{
"epoch": 0.25163268026122887,
"grad_norm": 0.4026257164141241,
"learning_rate": 3.4406806458282575e-05,
"loss": 1.3826,
"step": 1416
},
{
"epoch": 0.25181038695632857,
"grad_norm": 0.4103673574102238,
"learning_rate": 3.439897984392434e-05,
"loss": 1.3547,
"step": 1417
},
{
"epoch": 0.2519880936514283,
"grad_norm": 0.3936898455284017,
"learning_rate": 3.439114864897983e-05,
"loss": 1.3653,
"step": 1418
},
{
"epoch": 0.2521658003465281,
"grad_norm": 0.4844357072407005,
"learning_rate": 3.43833128759403e-05,
"loss": 1.4061,
"step": 1419
},
{
"epoch": 0.2523435070416278,
"grad_norm": 0.4088858027976452,
"learning_rate": 3.437547252729845e-05,
"loss": 1.3859,
"step": 1420
},
{
"epoch": 0.25252121373672753,
"grad_norm": 0.40428153326350913,
"learning_rate": 3.436762760554845e-05,
"loss": 1.4089,
"step": 1421
},
{
"epoch": 0.2526989204318273,
"grad_norm": 0.4142292441840726,
"learning_rate": 3.4359778113185905e-05,
"loss": 1.4552,
"step": 1422
},
{
"epoch": 0.252876627126927,
"grad_norm": 0.3997829535396069,
"learning_rate": 3.4351924052707904e-05,
"loss": 1.3953,
"step": 1423
},
{
"epoch": 0.25305433382202674,
"grad_norm": 0.4220205605578233,
"learning_rate": 3.434406542661296e-05,
"loss": 1.4492,
"step": 1424
},
{
"epoch": 0.2532320405171265,
"grad_norm": 0.3987455392173384,
"learning_rate": 3.4336202237401045e-05,
"loss": 1.4147,
"step": 1425
},
{
"epoch": 0.2534097472122262,
"grad_norm": 0.4114073862225607,
"learning_rate": 3.43283344875736e-05,
"loss": 1.3981,
"step": 1426
},
{
"epoch": 0.25358745390732595,
"grad_norm": 0.4044618734280721,
"learning_rate": 3.4320462179633496e-05,
"loss": 1.3855,
"step": 1427
},
{
"epoch": 0.2537651606024257,
"grad_norm": 0.39923119602369345,
"learning_rate": 3.431258531608506e-05,
"loss": 1.3592,
"step": 1428
},
{
"epoch": 0.2539428672975254,
"grad_norm": 0.4072338424150161,
"learning_rate": 3.4304703899434083e-05,
"loss": 1.4164,
"step": 1429
},
{
"epoch": 0.25412057399262516,
"grad_norm": 0.4198432353559508,
"learning_rate": 3.4296817932187785e-05,
"loss": 1.4236,
"step": 1430
},
{
"epoch": 0.2542982806877249,
"grad_norm": 0.3925188456141566,
"learning_rate": 3.428892741685483e-05,
"loss": 1.3975,
"step": 1431
},
{
"epoch": 0.2544759873828247,
"grad_norm": 0.4195768275679181,
"learning_rate": 3.4281032355945356e-05,
"loss": 1.4129,
"step": 1432
},
{
"epoch": 0.2546536940779244,
"grad_norm": 0.4023575760851173,
"learning_rate": 3.427313275197092e-05,
"loss": 1.3902,
"step": 1433
},
{
"epoch": 0.25483140077302413,
"grad_norm": 0.4154001919824938,
"learning_rate": 3.426522860744453e-05,
"loss": 1.4271,
"step": 1434
},
{
"epoch": 0.2550091074681239,
"grad_norm": 0.393198028213413,
"learning_rate": 3.4257319924880656e-05,
"loss": 1.4063,
"step": 1435
},
{
"epoch": 0.2551868141632236,
"grad_norm": 0.4139692307441467,
"learning_rate": 3.42494067067952e-05,
"loss": 1.3465,
"step": 1436
},
{
"epoch": 0.25536452085832334,
"grad_norm": 0.39896302423713187,
"learning_rate": 3.424148895570549e-05,
"loss": 1.3482,
"step": 1437
},
{
"epoch": 0.2555422275534231,
"grad_norm": 0.4028895467392937,
"learning_rate": 3.423356667413032e-05,
"loss": 1.363,
"step": 1438
},
{
"epoch": 0.2557199342485228,
"grad_norm": 0.4605777965399149,
"learning_rate": 3.422563986458992e-05,
"loss": 1.3814,
"step": 1439
},
{
"epoch": 0.25589764094362255,
"grad_norm": 0.4156298870975026,
"learning_rate": 3.421770852960595e-05,
"loss": 1.4496,
"step": 1440
},
{
"epoch": 0.2560753476387223,
"grad_norm": 0.4045057042454162,
"learning_rate": 3.420977267170153e-05,
"loss": 1.3948,
"step": 1441
},
{
"epoch": 0.256253054333822,
"grad_norm": 0.4112185248066729,
"learning_rate": 3.4201832293401184e-05,
"loss": 1.4057,
"step": 1442
},
{
"epoch": 0.25643076102892176,
"grad_norm": 0.39170921902921363,
"learning_rate": 3.4193887397230916e-05,
"loss": 1.41,
"step": 1443
},
{
"epoch": 0.2566084677240215,
"grad_norm": 0.4321471711199227,
"learning_rate": 3.418593798571814e-05,
"loss": 1.3999,
"step": 1444
},
{
"epoch": 0.2567861744191212,
"grad_norm": 0.40429252232263074,
"learning_rate": 3.417798406139171e-05,
"loss": 1.4485,
"step": 1445
},
{
"epoch": 0.25696388111422097,
"grad_norm": 0.46638759675517955,
"learning_rate": 3.417002562678191e-05,
"loss": 1.4535,
"step": 1446
},
{
"epoch": 0.2571415878093207,
"grad_norm": 0.39494496668873735,
"learning_rate": 3.416206268442049e-05,
"loss": 1.3663,
"step": 1447
},
{
"epoch": 0.2573192945044205,
"grad_norm": 0.4031389055220716,
"learning_rate": 3.41540952368406e-05,
"loss": 1.4056,
"step": 1448
},
{
"epoch": 0.2574970011995202,
"grad_norm": 0.42295562058627356,
"learning_rate": 3.414612328657684e-05,
"loss": 1.4182,
"step": 1449
},
{
"epoch": 0.25767470789461994,
"grad_norm": 0.42225115541414654,
"learning_rate": 3.413814683616522e-05,
"loss": 1.4626,
"step": 1450
},
{
"epoch": 0.2578524145897197,
"grad_norm": 0.42325789847461065,
"learning_rate": 3.413016588814322e-05,
"loss": 1.4396,
"step": 1451
},
{
"epoch": 0.2580301212848194,
"grad_norm": 0.41003262891449194,
"learning_rate": 3.412218044504973e-05,
"loss": 1.3706,
"step": 1452
},
{
"epoch": 0.25820782797991915,
"grad_norm": 0.42615291211319023,
"learning_rate": 3.411419050942505e-05,
"loss": 1.4408,
"step": 1453
},
{
"epoch": 0.2583855346750189,
"grad_norm": 0.411848840267133,
"learning_rate": 3.410619608381093e-05,
"loss": 1.4226,
"step": 1454
},
{
"epoch": 0.2585632413701186,
"grad_norm": 0.478057068501874,
"learning_rate": 3.409819717075058e-05,
"loss": 1.3946,
"step": 1455
},
{
"epoch": 0.25874094806521836,
"grad_norm": 0.433569524595364,
"learning_rate": 3.409019377278856e-05,
"loss": 1.412,
"step": 1456
},
{
"epoch": 0.2589186547603181,
"grad_norm": 0.418064039634807,
"learning_rate": 3.408218589247094e-05,
"loss": 1.3944,
"step": 1457
},
{
"epoch": 0.2590963614554178,
"grad_norm": 0.40826125932848734,
"learning_rate": 3.407417353234514e-05,
"loss": 1.4023,
"step": 1458
},
{
"epoch": 0.25927406815051757,
"grad_norm": 0.40065403419220375,
"learning_rate": 3.406615669496008e-05,
"loss": 1.3373,
"step": 1459
},
{
"epoch": 0.2594517748456173,
"grad_norm": 0.4350504805880367,
"learning_rate": 3.405813538286605e-05,
"loss": 1.4381,
"step": 1460
},
{
"epoch": 0.259629481540717,
"grad_norm": 0.489549801161693,
"learning_rate": 3.405010959861477e-05,
"loss": 1.3767,
"step": 1461
},
{
"epoch": 0.2598071882358168,
"grad_norm": 0.4210833659134268,
"learning_rate": 3.404207934475941e-05,
"loss": 1.4309,
"step": 1462
},
{
"epoch": 0.25998489493091653,
"grad_norm": 0.4284789231932698,
"learning_rate": 3.403404462385453e-05,
"loss": 1.409,
"step": 1463
},
{
"epoch": 0.2601626016260163,
"grad_norm": 0.419449962123283,
"learning_rate": 3.402600543845614e-05,
"loss": 1.4069,
"step": 1464
},
{
"epoch": 0.260340308321116,
"grad_norm": 0.41738399174149476,
"learning_rate": 3.401796179112164e-05,
"loss": 1.39,
"step": 1465
},
{
"epoch": 0.26051801501621574,
"grad_norm": 0.43465908256250635,
"learning_rate": 3.400991368440988e-05,
"loss": 1.3791,
"step": 1466
},
{
"epoch": 0.2606957217113155,
"grad_norm": 0.4041223638462964,
"learning_rate": 3.400186112088111e-05,
"loss": 1.4069,
"step": 1467
},
{
"epoch": 0.2608734284064152,
"grad_norm": 0.4214174654062767,
"learning_rate": 3.3993804103097e-05,
"loss": 1.3916,
"step": 1468
},
{
"epoch": 0.26105113510151495,
"grad_norm": 0.40833699777055193,
"learning_rate": 3.398574263362064e-05,
"loss": 1.4379,
"step": 1469
},
{
"epoch": 0.2612288417966147,
"grad_norm": 0.414159998298676,
"learning_rate": 3.397767671501654e-05,
"loss": 1.3912,
"step": 1470
},
{
"epoch": 0.2614065484917144,
"grad_norm": 0.407778026980415,
"learning_rate": 3.3969606349850605e-05,
"loss": 1.389,
"step": 1471
},
{
"epoch": 0.26158425518681416,
"grad_norm": 0.40490522615633934,
"learning_rate": 3.396153154069018e-05,
"loss": 1.4075,
"step": 1472
},
{
"epoch": 0.2617619618819139,
"grad_norm": 0.4287023990781641,
"learning_rate": 3.3953452290104015e-05,
"loss": 1.4055,
"step": 1473
},
{
"epoch": 0.2619396685770136,
"grad_norm": 0.39999545599550895,
"learning_rate": 3.3945368600662275e-05,
"loss": 1.4003,
"step": 1474
},
{
"epoch": 0.2621173752721134,
"grad_norm": 0.422853566992036,
"learning_rate": 3.3937280474936533e-05,
"loss": 1.3392,
"step": 1475
},
{
"epoch": 0.26229508196721313,
"grad_norm": 0.3971862648489382,
"learning_rate": 3.392918791549976e-05,
"loss": 1.3775,
"step": 1476
},
{
"epoch": 0.26247278866231283,
"grad_norm": 0.43871125352733753,
"learning_rate": 3.3921090924926364e-05,
"loss": 1.3957,
"step": 1477
},
{
"epoch": 0.2626504953574126,
"grad_norm": 0.4055888945084895,
"learning_rate": 3.391298950579215e-05,
"loss": 1.3775,
"step": 1478
},
{
"epoch": 0.26282820205251234,
"grad_norm": 0.43260590595920856,
"learning_rate": 3.390488366067432e-05,
"loss": 1.404,
"step": 1479
},
{
"epoch": 0.2630059087476121,
"grad_norm": 0.40893492441252577,
"learning_rate": 3.389677339215151e-05,
"loss": 1.3985,
"step": 1480
},
{
"epoch": 0.2631836154427118,
"grad_norm": 0.4189102682759715,
"learning_rate": 3.3888658702803746e-05,
"loss": 1.3905,
"step": 1481
},
{
"epoch": 0.26336132213781155,
"grad_norm": 0.4121318979230604,
"learning_rate": 3.388053959521245e-05,
"loss": 1.3657,
"step": 1482
},
{
"epoch": 0.2635390288329113,
"grad_norm": 0.4106250783713469,
"learning_rate": 3.3872416071960485e-05,
"loss": 1.4152,
"step": 1483
},
{
"epoch": 0.263716735528011,
"grad_norm": 0.41297143329763053,
"learning_rate": 3.386428813563208e-05,
"loss": 1.3939,
"step": 1484
},
{
"epoch": 0.26389444222311076,
"grad_norm": 0.4038393328003671,
"learning_rate": 3.3856155788812896e-05,
"loss": 1.4133,
"step": 1485
},
{
"epoch": 0.2640721489182105,
"grad_norm": 0.40929286407655574,
"learning_rate": 3.384801903408997e-05,
"loss": 1.3957,
"step": 1486
},
{
"epoch": 0.2642498556133102,
"grad_norm": 0.40358864124781363,
"learning_rate": 3.383987787405177e-05,
"loss": 1.3833,
"step": 1487
},
{
"epoch": 0.26442756230840997,
"grad_norm": 0.4117619060719583,
"learning_rate": 3.383173231128815e-05,
"loss": 1.4129,
"step": 1488
},
{
"epoch": 0.2646052690035097,
"grad_norm": 0.4108170684660174,
"learning_rate": 3.3823582348390353e-05,
"loss": 1.4237,
"step": 1489
},
{
"epoch": 0.2647829756986094,
"grad_norm": 0.3983691747376825,
"learning_rate": 3.381542798795106e-05,
"loss": 1.3762,
"step": 1490
},
{
"epoch": 0.2649606823937092,
"grad_norm": 0.4011355193821098,
"learning_rate": 3.3807269232564306e-05,
"loss": 1.4207,
"step": 1491
},
{
"epoch": 0.26513838908880893,
"grad_norm": 0.4033565373244467,
"learning_rate": 3.379910608482556e-05,
"loss": 1.4176,
"step": 1492
},
{
"epoch": 0.26531609578390863,
"grad_norm": 0.39975323881407954,
"learning_rate": 3.3790938547331656e-05,
"loss": 1.3909,
"step": 1493
},
{
"epoch": 0.2654938024790084,
"grad_norm": 0.40331452813084595,
"learning_rate": 3.378276662268085e-05,
"loss": 1.4103,
"step": 1494
},
{
"epoch": 0.26567150917410814,
"grad_norm": 0.4048844555897097,
"learning_rate": 3.3774590313472785e-05,
"loss": 1.4009,
"step": 1495
},
{
"epoch": 0.2658492158692079,
"grad_norm": 0.40891071466537193,
"learning_rate": 3.376640962230851e-05,
"loss": 1.3697,
"step": 1496
},
{
"epoch": 0.2660269225643076,
"grad_norm": 0.40945914085911916,
"learning_rate": 3.375822455179043e-05,
"loss": 1.3977,
"step": 1497
},
{
"epoch": 0.26620462925940735,
"grad_norm": 0.41230474478689183,
"learning_rate": 3.375003510452239e-05,
"loss": 1.4257,
"step": 1498
},
{
"epoch": 0.2663823359545071,
"grad_norm": 0.39679055509217737,
"learning_rate": 3.3741841283109604e-05,
"loss": 1.3837,
"step": 1499
},
{
"epoch": 0.2665600426496068,
"grad_norm": 0.40229044646430634,
"learning_rate": 3.373364309015868e-05,
"loss": 1.4043,
"step": 1500
},
{
"epoch": 0.26673774934470657,
"grad_norm": 0.4521506835547598,
"learning_rate": 3.3725440528277614e-05,
"loss": 1.4034,
"step": 1501
},
{
"epoch": 0.2669154560398063,
"grad_norm": 0.4087028391598625,
"learning_rate": 3.3717233600075795e-05,
"loss": 1.3606,
"step": 1502
},
{
"epoch": 0.267093162734906,
"grad_norm": 0.4270441272389639,
"learning_rate": 3.370902230816401e-05,
"loss": 1.4142,
"step": 1503
},
{
"epoch": 0.2672708694300058,
"grad_norm": 0.42020944289053036,
"learning_rate": 3.3700806655154415e-05,
"loss": 1.4203,
"step": 1504
},
{
"epoch": 0.26744857612510553,
"grad_norm": 0.4157531809184675,
"learning_rate": 3.3692586643660565e-05,
"loss": 1.4433,
"step": 1505
},
{
"epoch": 0.26762628282020523,
"grad_norm": 0.4123571538806253,
"learning_rate": 3.3684362276297406e-05,
"loss": 1.4147,
"step": 1506
},
{
"epoch": 0.267803989515305,
"grad_norm": 0.40688487432237336,
"learning_rate": 3.367613355568126e-05,
"loss": 1.3858,
"step": 1507
},
{
"epoch": 0.26798169621040474,
"grad_norm": 0.4074554826846542,
"learning_rate": 3.366790048442984e-05,
"loss": 1.3965,
"step": 1508
},
{
"epoch": 0.26815940290550444,
"grad_norm": 0.41091211552050755,
"learning_rate": 3.365966306516224e-05,
"loss": 1.3912,
"step": 1509
},
{
"epoch": 0.2683371096006042,
"grad_norm": 0.40578913266787037,
"learning_rate": 3.3651421300498936e-05,
"loss": 1.3832,
"step": 1510
},
{
"epoch": 0.26851481629570395,
"grad_norm": 0.42596587125472835,
"learning_rate": 3.3643175193061795e-05,
"loss": 1.4135,
"step": 1511
},
{
"epoch": 0.2686925229908037,
"grad_norm": 0.38887241444059667,
"learning_rate": 3.363492474547404e-05,
"loss": 1.3956,
"step": 1512
},
{
"epoch": 0.2688702296859034,
"grad_norm": 0.4193797422834709,
"learning_rate": 3.362666996036033e-05,
"loss": 1.3979,
"step": 1513
},
{
"epoch": 0.26904793638100316,
"grad_norm": 0.39973207430837404,
"learning_rate": 3.361841084034662e-05,
"loss": 1.3805,
"step": 1514
},
{
"epoch": 0.2692256430761029,
"grad_norm": 0.40169465829512757,
"learning_rate": 3.361014738806033e-05,
"loss": 1.3787,
"step": 1515
},
{
"epoch": 0.2694033497712026,
"grad_norm": 0.4230149703749249,
"learning_rate": 3.36018796061302e-05,
"loss": 1.4047,
"step": 1516
},
{
"epoch": 0.26958105646630237,
"grad_norm": 0.40458523046321393,
"learning_rate": 3.359360749718638e-05,
"loss": 1.3902,
"step": 1517
},
{
"epoch": 0.2697587631614021,
"grad_norm": 0.4159800568664621,
"learning_rate": 3.3585331063860365e-05,
"loss": 1.3982,
"step": 1518
},
{
"epoch": 0.2699364698565018,
"grad_norm": 0.404911813344907,
"learning_rate": 3.3577050308785065e-05,
"loss": 1.4325,
"step": 1519
},
{
"epoch": 0.2701141765516016,
"grad_norm": 0.39565956158233667,
"learning_rate": 3.3568765234594733e-05,
"loss": 1.3841,
"step": 1520
},
{
"epoch": 0.27029188324670134,
"grad_norm": 0.395257268373819,
"learning_rate": 3.3560475843925004e-05,
"loss": 1.3741,
"step": 1521
},
{
"epoch": 0.27046958994180104,
"grad_norm": 0.4009478958769014,
"learning_rate": 3.3552182139412886e-05,
"loss": 1.3811,
"step": 1522
},
{
"epoch": 0.2706472966369008,
"grad_norm": 0.40332948406440877,
"learning_rate": 3.354388412369679e-05,
"loss": 1.4244,
"step": 1523
},
{
"epoch": 0.27082500333200055,
"grad_norm": 0.40445847779240834,
"learning_rate": 3.353558179941643e-05,
"loss": 1.3899,
"step": 1524
},
{
"epoch": 0.27100271002710025,
"grad_norm": 0.41659706143120684,
"learning_rate": 3.3527275169212956e-05,
"loss": 1.4559,
"step": 1525
},
{
"epoch": 0.2711804167222,
"grad_norm": 0.3797275166135395,
"learning_rate": 3.351896423572886e-05,
"loss": 1.3734,
"step": 1526
},
{
"epoch": 0.27135812341729976,
"grad_norm": 0.42653579773894135,
"learning_rate": 3.3510649001608005e-05,
"loss": 1.4177,
"step": 1527
},
{
"epoch": 0.2715358301123995,
"grad_norm": 0.3824161841000674,
"learning_rate": 3.350232946949563e-05,
"loss": 1.3569,
"step": 1528
},
{
"epoch": 0.2717135368074992,
"grad_norm": 0.4225419002287009,
"learning_rate": 3.349400564203832e-05,
"loss": 1.4024,
"step": 1529
},
{
"epoch": 0.27189124350259897,
"grad_norm": 0.39466940551621815,
"learning_rate": 3.348567752188405e-05,
"loss": 1.377,
"step": 1530
},
{
"epoch": 0.2720689501976987,
"grad_norm": 0.4010951030346351,
"learning_rate": 3.347734511168215e-05,
"loss": 1.3676,
"step": 1531
},
{
"epoch": 0.2722466568927984,
"grad_norm": 0.5063570615741713,
"learning_rate": 3.346900841408332e-05,
"loss": 1.4435,
"step": 1532
},
{
"epoch": 0.2724243635878982,
"grad_norm": 0.39219731529474244,
"learning_rate": 3.346066743173962e-05,
"loss": 1.3872,
"step": 1533
},
{
"epoch": 0.27260207028299793,
"grad_norm": 0.42300756758462943,
"learning_rate": 3.345232216730446e-05,
"loss": 1.3827,
"step": 1534
},
{
"epoch": 0.27277977697809763,
"grad_norm": 0.41876110802193633,
"learning_rate": 3.3443972623432645e-05,
"loss": 1.3877,
"step": 1535
},
{
"epoch": 0.2729574836731974,
"grad_norm": 0.4126442224578322,
"learning_rate": 3.343561880278031e-05,
"loss": 1.3663,
"step": 1536
},
{
"epoch": 0.27313519036829714,
"grad_norm": 0.3919799564393195,
"learning_rate": 3.342726070800497e-05,
"loss": 1.3871,
"step": 1537
},
{
"epoch": 0.27331289706339684,
"grad_norm": 0.43304735075880263,
"learning_rate": 3.341889834176549e-05,
"loss": 1.4009,
"step": 1538
},
{
"epoch": 0.2734906037584966,
"grad_norm": 0.3841164432242966,
"learning_rate": 3.341053170672209e-05,
"loss": 1.3765,
"step": 1539
},
{
"epoch": 0.27366831045359635,
"grad_norm": 0.4063604678068933,
"learning_rate": 3.340216080553636e-05,
"loss": 1.3665,
"step": 1540
},
{
"epoch": 0.27384601714869605,
"grad_norm": 0.42964707793794155,
"learning_rate": 3.339378564087123e-05,
"loss": 1.4599,
"step": 1541
},
{
"epoch": 0.2740237238437958,
"grad_norm": 0.409699265448033,
"learning_rate": 3.3385406215391016e-05,
"loss": 1.3976,
"step": 1542
},
{
"epoch": 0.27420143053889556,
"grad_norm": 0.3959143469445942,
"learning_rate": 3.337702253176136e-05,
"loss": 1.4101,
"step": 1543
},
{
"epoch": 0.2743791372339953,
"grad_norm": 0.397496601630699,
"learning_rate": 3.336863459264926e-05,
"loss": 1.4259,
"step": 1544
},
{
"epoch": 0.274556843929095,
"grad_norm": 0.39618005789658367,
"learning_rate": 3.33602424007231e-05,
"loss": 1.392,
"step": 1545
},
{
"epoch": 0.2747345506241948,
"grad_norm": 0.38142163345536595,
"learning_rate": 3.3351845958652575e-05,
"loss": 1.3402,
"step": 1546
},
{
"epoch": 0.27491225731929453,
"grad_norm": 0.4046215353009362,
"learning_rate": 3.334344526910876e-05,
"loss": 1.4524,
"step": 1547
},
{
"epoch": 0.27508996401439423,
"grad_norm": 0.39183355159053,
"learning_rate": 3.333504033476407e-05,
"loss": 1.4105,
"step": 1548
},
{
"epoch": 0.275267670709494,
"grad_norm": 0.3994596205670117,
"learning_rate": 3.332663115829227e-05,
"loss": 1.4115,
"step": 1549
},
{
"epoch": 0.27544537740459374,
"grad_norm": 0.38873509328398664,
"learning_rate": 3.331821774236849e-05,
"loss": 1.4007,
"step": 1550
},
{
"epoch": 0.27562308409969344,
"grad_norm": 0.42474404287915785,
"learning_rate": 3.3309800089669175e-05,
"loss": 1.4299,
"step": 1551
},
{
"epoch": 0.2758007907947932,
"grad_norm": 0.3964283899871285,
"learning_rate": 3.330137820287215e-05,
"loss": 1.3932,
"step": 1552
},
{
"epoch": 0.27597849748989295,
"grad_norm": 0.40334326624088107,
"learning_rate": 3.329295208465658e-05,
"loss": 1.3891,
"step": 1553
},
{
"epoch": 0.27615620418499265,
"grad_norm": 0.39566899117970694,
"learning_rate": 3.328452173770296e-05,
"loss": 1.3352,
"step": 1554
},
{
"epoch": 0.2763339108800924,
"grad_norm": 0.4014164153309157,
"learning_rate": 3.327608716469316e-05,
"loss": 1.3649,
"step": 1555
},
{
"epoch": 0.27651161757519216,
"grad_norm": 0.39022684326859636,
"learning_rate": 3.3267648368310354e-05,
"loss": 1.3826,
"step": 1556
},
{
"epoch": 0.27668932427029186,
"grad_norm": 0.5063434996784149,
"learning_rate": 3.32592053512391e-05,
"loss": 1.4134,
"step": 1557
},
{
"epoch": 0.2768670309653916,
"grad_norm": 0.3818301683041679,
"learning_rate": 3.325075811616527e-05,
"loss": 1.3741,
"step": 1558
},
{
"epoch": 0.27704473766049137,
"grad_norm": 0.42666065554043625,
"learning_rate": 3.3242306665776084e-05,
"loss": 1.4169,
"step": 1559
},
{
"epoch": 0.2772224443555911,
"grad_norm": 0.39478348878271535,
"learning_rate": 3.323385100276013e-05,
"loss": 1.3758,
"step": 1560
},
{
"epoch": 0.2774001510506908,
"grad_norm": 0.3935711385208444,
"learning_rate": 3.322539112980729e-05,
"loss": 1.3992,
"step": 1561
},
{
"epoch": 0.2775778577457906,
"grad_norm": 0.4039836404994471,
"learning_rate": 3.321692704960881e-05,
"loss": 1.3665,
"step": 1562
},
{
"epoch": 0.27775556444089033,
"grad_norm": 0.39737908823200796,
"learning_rate": 3.320845876485729e-05,
"loss": 1.4043,
"step": 1563
},
{
"epoch": 0.27793327113599003,
"grad_norm": 0.4115849266885589,
"learning_rate": 3.319998627824664e-05,
"loss": 1.4154,
"step": 1564
},
{
"epoch": 0.2781109778310898,
"grad_norm": 0.4544614768122568,
"learning_rate": 3.3191509592472117e-05,
"loss": 1.3852,
"step": 1565
},
{
"epoch": 0.27828868452618954,
"grad_norm": 0.39960824755776425,
"learning_rate": 3.318302871023032e-05,
"loss": 1.4232,
"step": 1566
},
{
"epoch": 0.27846639122128924,
"grad_norm": 0.3978897552883364,
"learning_rate": 3.317454363421916e-05,
"loss": 1.3998,
"step": 1567
},
{
"epoch": 0.278644097916389,
"grad_norm": 0.39399979959090914,
"learning_rate": 3.3166054367137915e-05,
"loss": 1.4317,
"step": 1568
},
{
"epoch": 0.27882180461148875,
"grad_norm": 0.39333702130758064,
"learning_rate": 3.315756091168719e-05,
"loss": 1.3844,
"step": 1569
},
{
"epoch": 0.27899951130658845,
"grad_norm": 0.3820040968197348,
"learning_rate": 3.314906327056888e-05,
"loss": 1.3985,
"step": 1570
},
{
"epoch": 0.2791772180016882,
"grad_norm": 0.4044341419222239,
"learning_rate": 3.314056144648628e-05,
"loss": 1.4083,
"step": 1571
},
{
"epoch": 0.27935492469678797,
"grad_norm": 0.38804010008057804,
"learning_rate": 3.313205544214396e-05,
"loss": 1.3858,
"step": 1572
},
{
"epoch": 0.27953263139188766,
"grad_norm": 0.4175669028600208,
"learning_rate": 3.312354526024784e-05,
"loss": 1.3977,
"step": 1573
},
{
"epoch": 0.2797103380869874,
"grad_norm": 0.3968599897090016,
"learning_rate": 3.311503090350518e-05,
"loss": 1.4252,
"step": 1574
},
{
"epoch": 0.2798880447820872,
"grad_norm": 0.40607472583007825,
"learning_rate": 3.3106512374624544e-05,
"loss": 1.3792,
"step": 1575
},
{
"epoch": 0.28006575147718693,
"grad_norm": 0.4063505282680437,
"learning_rate": 3.3097989676315846e-05,
"loss": 1.4165,
"step": 1576
},
{
"epoch": 0.28024345817228663,
"grad_norm": 0.39467961244767036,
"learning_rate": 3.308946281129031e-05,
"loss": 1.4086,
"step": 1577
},
{
"epoch": 0.2804211648673864,
"grad_norm": 0.7452462331713218,
"learning_rate": 3.308093178226051e-05,
"loss": 1.3971,
"step": 1578
},
{
"epoch": 0.28059887156248614,
"grad_norm": 0.4103747449194099,
"learning_rate": 3.3072396591940296e-05,
"loss": 1.4438,
"step": 1579
},
{
"epoch": 0.28077657825758584,
"grad_norm": 0.3945192355187906,
"learning_rate": 3.306385724304489e-05,
"loss": 1.3916,
"step": 1580
},
{
"epoch": 0.2809542849526856,
"grad_norm": 0.4186833421546783,
"learning_rate": 3.305531373829082e-05,
"loss": 1.3975,
"step": 1581
},
{
"epoch": 0.28113199164778535,
"grad_norm": 0.43661631843903514,
"learning_rate": 3.304676608039594e-05,
"loss": 1.393,
"step": 1582
},
{
"epoch": 0.28130969834288505,
"grad_norm": 0.39942312024139526,
"learning_rate": 3.303821427207941e-05,
"loss": 1.3938,
"step": 1583
},
{
"epoch": 0.2814874050379848,
"grad_norm": 0.42533593961945426,
"learning_rate": 3.302965831606172e-05,
"loss": 1.395,
"step": 1584
},
{
"epoch": 0.28166511173308456,
"grad_norm": 0.42766323190515876,
"learning_rate": 3.302109821506469e-05,
"loss": 1.4738,
"step": 1585
},
{
"epoch": 0.28184281842818426,
"grad_norm": 0.40009620942154245,
"learning_rate": 3.301253397181145e-05,
"loss": 1.4015,
"step": 1586
},
{
"epoch": 0.282020525123284,
"grad_norm": 0.43498293133784205,
"learning_rate": 3.3003965589026436e-05,
"loss": 1.4381,
"step": 1587
},
{
"epoch": 0.28219823181838377,
"grad_norm": 0.40201798053248416,
"learning_rate": 3.2995393069435424e-05,
"loss": 1.4101,
"step": 1588
},
{
"epoch": 0.28237593851348347,
"grad_norm": 0.44420450279225676,
"learning_rate": 3.2986816415765476e-05,
"loss": 1.4404,
"step": 1589
},
{
"epoch": 0.2825536452085832,
"grad_norm": 0.3902539364169751,
"learning_rate": 3.2978235630745006e-05,
"loss": 1.3884,
"step": 1590
},
{
"epoch": 0.282731351903683,
"grad_norm": 0.4264358967910341,
"learning_rate": 3.296965071710371e-05,
"loss": 1.4122,
"step": 1591
},
{
"epoch": 0.28290905859878274,
"grad_norm": 0.39791990145190537,
"learning_rate": 3.296106167757263e-05,
"loss": 1.3751,
"step": 1592
},
{
"epoch": 0.28308676529388244,
"grad_norm": 0.4076885805287037,
"learning_rate": 3.295246851488407e-05,
"loss": 1.3684,
"step": 1593
},
{
"epoch": 0.2832644719889822,
"grad_norm": 0.38624796100879505,
"learning_rate": 3.2943871231771696e-05,
"loss": 1.4014,
"step": 1594
},
{
"epoch": 0.28344217868408195,
"grad_norm": 0.4036826725814489,
"learning_rate": 3.293526983097047e-05,
"loss": 1.4107,
"step": 1595
},
{
"epoch": 0.28361988537918165,
"grad_norm": 0.39644941593489663,
"learning_rate": 3.292666431521664e-05,
"loss": 1.3825,
"step": 1596
},
{
"epoch": 0.2837975920742814,
"grad_norm": 0.4131582452874447,
"learning_rate": 3.291805468724781e-05,
"loss": 1.4083,
"step": 1597
},
{
"epoch": 0.28397529876938116,
"grad_norm": 0.40876336853298856,
"learning_rate": 3.290944094980284e-05,
"loss": 1.3698,
"step": 1598
},
{
"epoch": 0.28415300546448086,
"grad_norm": 0.4163366672426127,
"learning_rate": 3.290082310562194e-05,
"loss": 1.4206,
"step": 1599
},
{
"epoch": 0.2843307121595806,
"grad_norm": 0.4202313101356313,
"learning_rate": 3.2892201157446585e-05,
"loss": 1.4011,
"step": 1600
},
{
"epoch": 0.28450841885468037,
"grad_norm": 0.4159053690757171,
"learning_rate": 3.28835751080196e-05,
"loss": 1.3986,
"step": 1601
},
{
"epoch": 0.28468612554978007,
"grad_norm": 0.42181651947626014,
"learning_rate": 3.2874944960085086e-05,
"loss": 1.386,
"step": 1602
},
{
"epoch": 0.2848638322448798,
"grad_norm": 0.4333788126001272,
"learning_rate": 3.2866310716388464e-05,
"loss": 1.4092,
"step": 1603
},
{
"epoch": 0.2850415389399796,
"grad_norm": 0.39991112724954747,
"learning_rate": 3.285767237967643e-05,
"loss": 1.3554,
"step": 1604
},
{
"epoch": 0.2852192456350793,
"grad_norm": 0.4239677203185931,
"learning_rate": 3.284902995269701e-05,
"loss": 1.3638,
"step": 1605
},
{
"epoch": 0.28539695233017903,
"grad_norm": 0.39721886482847046,
"learning_rate": 3.284038343819954e-05,
"loss": 1.4141,
"step": 1606
},
{
"epoch": 0.2855746590252788,
"grad_norm": 0.42208714969133787,
"learning_rate": 3.2831732838934615e-05,
"loss": 1.4219,
"step": 1607
},
{
"epoch": 0.28575236572037854,
"grad_norm": 0.3991316682175627,
"learning_rate": 3.282307815765416e-05,
"loss": 1.3903,
"step": 1608
},
{
"epoch": 0.28593007241547824,
"grad_norm": 0.4391594378440536,
"learning_rate": 3.28144193971114e-05,
"loss": 1.4051,
"step": 1609
},
{
"epoch": 0.286107779110578,
"grad_norm": 0.40552356304580467,
"learning_rate": 3.2805756560060844e-05,
"loss": 1.3791,
"step": 1610
},
{
"epoch": 0.28628548580567775,
"grad_norm": 0.4188134425150147,
"learning_rate": 3.27970896492583e-05,
"loss": 1.3997,
"step": 1611
},
{
"epoch": 0.28646319250077745,
"grad_norm": 0.40441593193212794,
"learning_rate": 3.2788418667460873e-05,
"loss": 1.3519,
"step": 1612
},
{
"epoch": 0.2866408991958772,
"grad_norm": 0.4220351718939253,
"learning_rate": 3.277974361742698e-05,
"loss": 1.3809,
"step": 1613
},
{
"epoch": 0.28681860589097696,
"grad_norm": 0.4079900863817792,
"learning_rate": 3.27710645019163e-05,
"loss": 1.4453,
"step": 1614
},
{
"epoch": 0.28699631258607666,
"grad_norm": 0.4076445057269382,
"learning_rate": 3.276238132368984e-05,
"loss": 1.3809,
"step": 1615
},
{
"epoch": 0.2871740192811764,
"grad_norm": 0.4110468169633426,
"learning_rate": 3.275369408550987e-05,
"loss": 1.3778,
"step": 1616
},
{
"epoch": 0.2873517259762762,
"grad_norm": 0.4093288346776009,
"learning_rate": 3.274500279013997e-05,
"loss": 1.3721,
"step": 1617
},
{
"epoch": 0.2875294326713759,
"grad_norm": 0.4232851884958083,
"learning_rate": 3.2736307440345e-05,
"loss": 1.3889,
"step": 1618
},
{
"epoch": 0.28770713936647563,
"grad_norm": 0.3946482958150229,
"learning_rate": 3.272760803889111e-05,
"loss": 1.3815,
"step": 1619
},
{
"epoch": 0.2878848460615754,
"grad_norm": 0.42197964306152186,
"learning_rate": 3.271890458854576e-05,
"loss": 1.42,
"step": 1620
},
{
"epoch": 0.2880625527566751,
"grad_norm": 0.38702598400063803,
"learning_rate": 3.271019709207767e-05,
"loss": 1.3679,
"step": 1621
},
{
"epoch": 0.28824025945177484,
"grad_norm": 0.3982874007838377,
"learning_rate": 3.2701485552256846e-05,
"loss": 1.3791,
"step": 1622
},
{
"epoch": 0.2884179661468746,
"grad_norm": 0.398321736745097,
"learning_rate": 3.269276997185461e-05,
"loss": 1.4031,
"step": 1623
},
{
"epoch": 0.28859567284197435,
"grad_norm": 0.40388997131194854,
"learning_rate": 3.268405035364356e-05,
"loss": 1.3914,
"step": 1624
},
{
"epoch": 0.28877337953707405,
"grad_norm": 0.3809944371654938,
"learning_rate": 3.2675326700397544e-05,
"loss": 1.3756,
"step": 1625
},
{
"epoch": 0.2889510862321738,
"grad_norm": 0.42009222674340824,
"learning_rate": 3.266659901489174e-05,
"loss": 1.4025,
"step": 1626
},
{
"epoch": 0.28912879292727356,
"grad_norm": 0.4024351126442283,
"learning_rate": 3.2657867299902594e-05,
"loss": 1.3953,
"step": 1627
},
{
"epoch": 0.28930649962237326,
"grad_norm": 0.3935931687789224,
"learning_rate": 3.264913155820781e-05,
"loss": 1.3694,
"step": 1628
},
{
"epoch": 0.289484206317473,
"grad_norm": 0.39922607990594977,
"learning_rate": 3.264039179258639e-05,
"loss": 1.3993,
"step": 1629
},
{
"epoch": 0.28966191301257277,
"grad_norm": 0.38992975570713506,
"learning_rate": 3.2631648005818634e-05,
"loss": 1.3945,
"step": 1630
},
{
"epoch": 0.28983961970767247,
"grad_norm": 0.4024419864708078,
"learning_rate": 3.2622900200686096e-05,
"loss": 1.3804,
"step": 1631
},
{
"epoch": 0.2900173264027722,
"grad_norm": 0.3863561890919406,
"learning_rate": 3.261414837997163e-05,
"loss": 1.3857,
"step": 1632
},
{
"epoch": 0.290195033097872,
"grad_norm": 0.3981306969291142,
"learning_rate": 3.260539254645934e-05,
"loss": 1.3797,
"step": 1633
},
{
"epoch": 0.2903727397929717,
"grad_norm": 0.4269522317555827,
"learning_rate": 3.259663270293462e-05,
"loss": 1.4093,
"step": 1634
},
{
"epoch": 0.29055044648807143,
"grad_norm": 0.40318290709870525,
"learning_rate": 3.258786885218415e-05,
"loss": 1.4133,
"step": 1635
},
{
"epoch": 0.2907281531831712,
"grad_norm": 0.41707796643328604,
"learning_rate": 3.2579100996995876e-05,
"loss": 1.399,
"step": 1636
},
{
"epoch": 0.2909058598782709,
"grad_norm": 0.4049682690781572,
"learning_rate": 3.257032914015901e-05,
"loss": 1.4059,
"step": 1637
},
{
"epoch": 0.29108356657337064,
"grad_norm": 0.3933614482345727,
"learning_rate": 3.256155328446405e-05,
"loss": 1.4036,
"step": 1638
},
{
"epoch": 0.2912612732684704,
"grad_norm": 0.3974220852598693,
"learning_rate": 3.255277343270276e-05,
"loss": 1.3656,
"step": 1639
},
{
"epoch": 0.29143897996357016,
"grad_norm": 0.4460896207080004,
"learning_rate": 3.2543989587668174e-05,
"loss": 1.3788,
"step": 1640
},
{
"epoch": 0.29161668665866985,
"grad_norm": 0.3974283209575224,
"learning_rate": 3.25352017521546e-05,
"loss": 1.3927,
"step": 1641
},
{
"epoch": 0.2917943933537696,
"grad_norm": 0.40482572552380786,
"learning_rate": 3.252640992895762e-05,
"loss": 1.4,
"step": 1642
},
{
"epoch": 0.29197210004886937,
"grad_norm": 0.4084934809027206,
"learning_rate": 3.251761412087406e-05,
"loss": 1.3975,
"step": 1643
},
{
"epoch": 0.29214980674396906,
"grad_norm": 0.39785446841501804,
"learning_rate": 3.250881433070206e-05,
"loss": 1.3667,
"step": 1644
},
{
"epoch": 0.2923275134390688,
"grad_norm": 0.464693741380573,
"learning_rate": 3.2500010561240966e-05,
"loss": 1.4156,
"step": 1645
},
{
"epoch": 0.2925052201341686,
"grad_norm": 0.39891875789491243,
"learning_rate": 3.249120281529145e-05,
"loss": 1.3772,
"step": 1646
},
{
"epoch": 0.2926829268292683,
"grad_norm": 0.3855430733334076,
"learning_rate": 3.2482391095655405e-05,
"loss": 1.3503,
"step": 1647
},
{
"epoch": 0.29286063352436803,
"grad_norm": 0.4102377735238566,
"learning_rate": 3.247357540513602e-05,
"loss": 1.352,
"step": 1648
},
{
"epoch": 0.2930383402194678,
"grad_norm": 0.38176412011283334,
"learning_rate": 3.246475574653771e-05,
"loss": 1.3353,
"step": 1649
},
{
"epoch": 0.2932160469145675,
"grad_norm": 0.39481755222220544,
"learning_rate": 3.245593212266619e-05,
"loss": 1.431,
"step": 1650
},
{
"epoch": 0.29339375360966724,
"grad_norm": 0.3964506398001558,
"learning_rate": 3.244710453632842e-05,
"loss": 1.4137,
"step": 1651
},
{
"epoch": 0.293571460304767,
"grad_norm": 0.38102838872946093,
"learning_rate": 3.243827299033262e-05,
"loss": 1.402,
"step": 1652
},
{
"epoch": 0.2937491669998667,
"grad_norm": 0.4039223675880007,
"learning_rate": 3.242943748748827e-05,
"loss": 1.3883,
"step": 1653
},
{
"epoch": 0.29392687369496645,
"grad_norm": 0.3849542305136165,
"learning_rate": 3.2420598030606106e-05,
"loss": 1.3681,
"step": 1654
},
{
"epoch": 0.2941045803900662,
"grad_norm": 0.396638556987177,
"learning_rate": 3.241175462249813e-05,
"loss": 1.4236,
"step": 1655
},
{
"epoch": 0.29428228708516596,
"grad_norm": 0.3974496423181337,
"learning_rate": 3.24029072659776e-05,
"loss": 1.4038,
"step": 1656
},
{
"epoch": 0.29445999378026566,
"grad_norm": 0.38954024997027087,
"learning_rate": 3.239405596385902e-05,
"loss": 1.3501,
"step": 1657
},
{
"epoch": 0.2946377004753654,
"grad_norm": 0.39576624884291123,
"learning_rate": 3.2385200718958147e-05,
"loss": 1.3419,
"step": 1658
},
{
"epoch": 0.29481540717046517,
"grad_norm": 0.3915805649468901,
"learning_rate": 3.237634153409202e-05,
"loss": 1.3903,
"step": 1659
},
{
"epoch": 0.29499311386556487,
"grad_norm": 0.38825577392928856,
"learning_rate": 3.23674784120789e-05,
"loss": 1.375,
"step": 1660
},
{
"epoch": 0.2951708205606646,
"grad_norm": 0.39928106209699527,
"learning_rate": 3.2358611355738316e-05,
"loss": 1.3835,
"step": 1661
},
{
"epoch": 0.2953485272557644,
"grad_norm": 0.3792899764377308,
"learning_rate": 3.234974036789105e-05,
"loss": 1.3488,
"step": 1662
},
{
"epoch": 0.2955262339508641,
"grad_norm": 0.40778368437430934,
"learning_rate": 3.234086545135912e-05,
"loss": 1.4245,
"step": 1663
},
{
"epoch": 0.29570394064596384,
"grad_norm": 0.38469336346166605,
"learning_rate": 3.233198660896581e-05,
"loss": 1.4012,
"step": 1664
},
{
"epoch": 0.2958816473410636,
"grad_norm": 0.5805469638483082,
"learning_rate": 3.2323103843535654e-05,
"loss": 1.3923,
"step": 1665
},
{
"epoch": 0.2960593540361633,
"grad_norm": 0.3920958287034638,
"learning_rate": 3.231421715789441e-05,
"loss": 1.4146,
"step": 1666
},
{
"epoch": 0.29623706073126305,
"grad_norm": 0.42798898780017963,
"learning_rate": 3.2305326554869113e-05,
"loss": 1.4368,
"step": 1667
},
{
"epoch": 0.2964147674263628,
"grad_norm": 0.39135204789914735,
"learning_rate": 3.229643203728802e-05,
"loss": 1.3817,
"step": 1668
},
{
"epoch": 0.2965924741214625,
"grad_norm": 0.40778556857486137,
"learning_rate": 3.228753360798067e-05,
"loss": 1.4456,
"step": 1669
},
{
"epoch": 0.29677018081656226,
"grad_norm": 0.3731132574979778,
"learning_rate": 3.227863126977778e-05,
"loss": 1.3684,
"step": 1670
},
{
"epoch": 0.296947887511662,
"grad_norm": 0.3991898902802162,
"learning_rate": 3.226972502551139e-05,
"loss": 1.4111,
"step": 1671
},
{
"epoch": 0.29712559420676177,
"grad_norm": 0.3962811405636046,
"learning_rate": 3.2260814878014715e-05,
"loss": 1.4037,
"step": 1672
},
{
"epoch": 0.29730330090186147,
"grad_norm": 0.3935519612258737,
"learning_rate": 3.2251900830122255e-05,
"loss": 1.4237,
"step": 1673
},
{
"epoch": 0.2974810075969612,
"grad_norm": 0.421779100118435,
"learning_rate": 3.224298288466974e-05,
"loss": 1.4265,
"step": 1674
},
{
"epoch": 0.297658714292061,
"grad_norm": 0.39123499790967425,
"learning_rate": 3.2234061044494116e-05,
"loss": 1.398,
"step": 1675
},
{
"epoch": 0.2978364209871607,
"grad_norm": 0.3956770363895871,
"learning_rate": 3.222513531243362e-05,
"loss": 1.3328,
"step": 1676
},
{
"epoch": 0.29801412768226043,
"grad_norm": 0.3937175967810584,
"learning_rate": 3.221620569132767e-05,
"loss": 1.4267,
"step": 1677
},
{
"epoch": 0.2981918343773602,
"grad_norm": 0.3980663252074694,
"learning_rate": 3.220727218401694e-05,
"loss": 1.4245,
"step": 1678
},
{
"epoch": 0.2983695410724599,
"grad_norm": 0.38560516504414605,
"learning_rate": 3.219833479334337e-05,
"loss": 1.3825,
"step": 1679
},
{
"epoch": 0.29854724776755964,
"grad_norm": 0.39827183035455255,
"learning_rate": 3.218939352215011e-05,
"loss": 1.3628,
"step": 1680
},
{
"epoch": 0.2987249544626594,
"grad_norm": 0.40282044667502276,
"learning_rate": 3.218044837328153e-05,
"loss": 1.383,
"step": 1681
},
{
"epoch": 0.2989026611577591,
"grad_norm": 0.38288832599294975,
"learning_rate": 3.217149934958326e-05,
"loss": 1.3477,
"step": 1682
},
{
"epoch": 0.29908036785285885,
"grad_norm": 0.42839146266150513,
"learning_rate": 3.2162546453902156e-05,
"loss": 1.411,
"step": 1683
},
{
"epoch": 0.2992580745479586,
"grad_norm": 0.3728771393975353,
"learning_rate": 3.21535896890863e-05,
"loss": 1.3626,
"step": 1684
},
{
"epoch": 0.2994357812430583,
"grad_norm": 0.4095472698617107,
"learning_rate": 3.2144629057985e-05,
"loss": 1.3767,
"step": 1685
},
{
"epoch": 0.29961348793815806,
"grad_norm": 0.3890754130529206,
"learning_rate": 3.21356645634488e-05,
"loss": 1.3957,
"step": 1686
},
{
"epoch": 0.2997911946332578,
"grad_norm": 0.394776988426533,
"learning_rate": 3.212669620832949e-05,
"loss": 1.3477,
"step": 1687
},
{
"epoch": 0.2999689013283576,
"grad_norm": 0.3880027151202906,
"learning_rate": 3.211772399548006e-05,
"loss": 1.3746,
"step": 1688
},
{
"epoch": 0.3001466080234573,
"grad_norm": 0.40469740899803336,
"learning_rate": 3.210874792775474e-05,
"loss": 1.3922,
"step": 1689
},
{
"epoch": 0.30032431471855703,
"grad_norm": 0.40797766255136025,
"learning_rate": 3.2099768008009e-05,
"loss": 1.424,
"step": 1690
},
{
"epoch": 0.3005020214136568,
"grad_norm": 0.39587049118402245,
"learning_rate": 3.209078423909951e-05,
"loss": 1.3758,
"step": 1691
},
{
"epoch": 0.3006797281087565,
"grad_norm": 0.41031824801943484,
"learning_rate": 3.208179662388416e-05,
"loss": 1.3939,
"step": 1692
},
{
"epoch": 0.30085743480385624,
"grad_norm": 0.41488130859786965,
"learning_rate": 3.207280516522211e-05,
"loss": 1.4049,
"step": 1693
},
{
"epoch": 0.301035141498956,
"grad_norm": 0.39793865481142593,
"learning_rate": 3.206380986597369e-05,
"loss": 1.3761,
"step": 1694
},
{
"epoch": 0.3012128481940557,
"grad_norm": 0.37855960317497056,
"learning_rate": 3.205481072900049e-05,
"loss": 1.3634,
"step": 1695
},
{
"epoch": 0.30139055488915545,
"grad_norm": 0.407598246908826,
"learning_rate": 3.204580775716529e-05,
"loss": 1.4006,
"step": 1696
},
{
"epoch": 0.3015682615842552,
"grad_norm": 0.37823132794884584,
"learning_rate": 3.203680095333211e-05,
"loss": 1.3704,
"step": 1697
},
{
"epoch": 0.3017459682793549,
"grad_norm": 0.39322719495335956,
"learning_rate": 3.202779032036619e-05,
"loss": 1.4464,
"step": 1698
},
{
"epoch": 0.30192367497445466,
"grad_norm": 0.3812291633383872,
"learning_rate": 3.201877586113397e-05,
"loss": 1.3624,
"step": 1699
},
{
"epoch": 0.3021013816695544,
"grad_norm": 0.4062537352361028,
"learning_rate": 3.200975757850312e-05,
"loss": 1.4064,
"step": 1700
},
{
"epoch": 0.3022790883646541,
"grad_norm": 0.3859127178085196,
"learning_rate": 3.2000735475342546e-05,
"loss": 1.3808,
"step": 1701
},
{
"epoch": 0.30245679505975387,
"grad_norm": 0.3956382866962086,
"learning_rate": 3.199170955452232e-05,
"loss": 1.3807,
"step": 1702
},
{
"epoch": 0.3026345017548536,
"grad_norm": 0.3960554073361072,
"learning_rate": 3.1982679818913775e-05,
"loss": 1.4077,
"step": 1703
},
{
"epoch": 0.3028122084499534,
"grad_norm": 0.3907105830360487,
"learning_rate": 3.197364627138944e-05,
"loss": 1.384,
"step": 1704
},
{
"epoch": 0.3029899151450531,
"grad_norm": 0.3874503958625912,
"learning_rate": 3.196460891482305e-05,
"loss": 1.3683,
"step": 1705
},
{
"epoch": 0.30316762184015283,
"grad_norm": 0.39236377976142384,
"learning_rate": 3.195556775208956e-05,
"loss": 1.4252,
"step": 1706
},
{
"epoch": 0.3033453285352526,
"grad_norm": 0.3907350244014331,
"learning_rate": 3.1946522786065125e-05,
"loss": 1.377,
"step": 1707
},
{
"epoch": 0.3035230352303523,
"grad_norm": 0.40711868622924857,
"learning_rate": 3.1937474019627135e-05,
"loss": 1.3677,
"step": 1708
},
{
"epoch": 0.30370074192545204,
"grad_norm": 0.39183232074521307,
"learning_rate": 3.1928421455654166e-05,
"loss": 1.3755,
"step": 1709
},
{
"epoch": 0.3038784486205518,
"grad_norm": 0.39363626362771287,
"learning_rate": 3.191936509702601e-05,
"loss": 1.3845,
"step": 1710
},
{
"epoch": 0.3040561553156515,
"grad_norm": 0.3928302898937467,
"learning_rate": 3.191030494662365e-05,
"loss": 1.3704,
"step": 1711
},
{
"epoch": 0.30423386201075125,
"grad_norm": 0.3949231106889793,
"learning_rate": 3.190124100732931e-05,
"loss": 1.3936,
"step": 1712
},
{
"epoch": 0.304411568705851,
"grad_norm": 0.3896429521006358,
"learning_rate": 3.1892173282026395e-05,
"loss": 1.402,
"step": 1713
},
{
"epoch": 0.3045892754009507,
"grad_norm": 0.39358950608098114,
"learning_rate": 3.1883101773599516e-05,
"loss": 1.3837,
"step": 1714
},
{
"epoch": 0.30476698209605046,
"grad_norm": 0.4550369289759249,
"learning_rate": 3.187402648493449e-05,
"loss": 1.3822,
"step": 1715
},
{
"epoch": 0.3049446887911502,
"grad_norm": 0.39311726201340064,
"learning_rate": 3.186494741891834e-05,
"loss": 1.4064,
"step": 1716
},
{
"epoch": 0.3051223954862499,
"grad_norm": 0.3855903980160069,
"learning_rate": 3.1855864578439283e-05,
"loss": 1.4198,
"step": 1717
},
{
"epoch": 0.3053001021813497,
"grad_norm": 0.4036364187035828,
"learning_rate": 3.184677796638675e-05,
"loss": 1.4056,
"step": 1718
},
{
"epoch": 0.30547780887644943,
"grad_norm": 0.40182253767855214,
"learning_rate": 3.183768758565135e-05,
"loss": 1.4188,
"step": 1719
},
{
"epoch": 0.3056555155715492,
"grad_norm": 0.4031283687604397,
"learning_rate": 3.1828593439124915e-05,
"loss": 1.3847,
"step": 1720
},
{
"epoch": 0.3058332222666489,
"grad_norm": 0.40834336351674094,
"learning_rate": 3.1819495529700465e-05,
"loss": 1.4272,
"step": 1721
},
{
"epoch": 0.30601092896174864,
"grad_norm": 0.40645031754955535,
"learning_rate": 3.18103938602722e-05,
"loss": 1.365,
"step": 1722
},
{
"epoch": 0.3061886356568484,
"grad_norm": 0.40896747773286224,
"learning_rate": 3.180128843373555e-05,
"loss": 1.4315,
"step": 1723
},
{
"epoch": 0.3063663423519481,
"grad_norm": 0.4007914207865842,
"learning_rate": 3.179217925298712e-05,
"loss": 1.3713,
"step": 1724
},
{
"epoch": 0.30654404904704785,
"grad_norm": 0.3844939369315128,
"learning_rate": 3.17830663209247e-05,
"loss": 1.4105,
"step": 1725
},
{
"epoch": 0.3067217557421476,
"grad_norm": 0.40306042343676945,
"learning_rate": 3.1773949640447295e-05,
"loss": 1.3393,
"step": 1726
},
{
"epoch": 0.3068994624372473,
"grad_norm": 0.39028605768072533,
"learning_rate": 3.176482921445509e-05,
"loss": 1.3942,
"step": 1727
},
{
"epoch": 0.30707716913234706,
"grad_norm": 0.4007505580822187,
"learning_rate": 3.1755705045849465e-05,
"loss": 1.387,
"step": 1728
},
{
"epoch": 0.3072548758274468,
"grad_norm": 0.38390136498298344,
"learning_rate": 3.174657713753299e-05,
"loss": 1.3881,
"step": 1729
},
{
"epoch": 0.3074325825225465,
"grad_norm": 0.407235129101364,
"learning_rate": 3.173744549240942e-05,
"loss": 1.3934,
"step": 1730
},
{
"epoch": 0.30761028921764627,
"grad_norm": 0.3755877486744155,
"learning_rate": 3.1728310113383715e-05,
"loss": 1.3701,
"step": 1731
},
{
"epoch": 0.307787995912746,
"grad_norm": 0.389801398537014,
"learning_rate": 3.1719171003361996e-05,
"loss": 1.4031,
"step": 1732
},
{
"epoch": 0.3079657026078457,
"grad_norm": 0.39386792292517,
"learning_rate": 3.171002816525159e-05,
"loss": 1.3989,
"step": 1733
},
{
"epoch": 0.3081434093029455,
"grad_norm": 0.3949183258614981,
"learning_rate": 3.170088160196101e-05,
"loss": 1.4067,
"step": 1734
},
{
"epoch": 0.30832111599804524,
"grad_norm": 0.38621958864303724,
"learning_rate": 3.169173131639995e-05,
"loss": 1.3904,
"step": 1735
},
{
"epoch": 0.308498822693145,
"grad_norm": 0.4016218794880723,
"learning_rate": 3.168257731147928e-05,
"loss": 1.4124,
"step": 1736
},
{
"epoch": 0.3086765293882447,
"grad_norm": 0.38077920029390383,
"learning_rate": 3.167341959011107e-05,
"loss": 1.3434,
"step": 1737
},
{
"epoch": 0.30885423608334445,
"grad_norm": 0.37712607417130817,
"learning_rate": 3.1664258155208555e-05,
"loss": 1.3732,
"step": 1738
},
{
"epoch": 0.3090319427784442,
"grad_norm": 0.38127618336972374,
"learning_rate": 3.165509300968617e-05,
"loss": 1.3804,
"step": 1739
},
{
"epoch": 0.3092096494735439,
"grad_norm": 0.38212737779449363,
"learning_rate": 3.1645924156459515e-05,
"loss": 1.3983,
"step": 1740
},
{
"epoch": 0.30938735616864366,
"grad_norm": 0.3791030728111567,
"learning_rate": 3.1636751598445367e-05,
"loss": 1.3769,
"step": 1741
},
{
"epoch": 0.3095650628637434,
"grad_norm": 0.3683547743981304,
"learning_rate": 3.162757533856171e-05,
"loss": 1.3462,
"step": 1742
},
{
"epoch": 0.3097427695588431,
"grad_norm": 0.3710925539466437,
"learning_rate": 3.1618395379727664e-05,
"loss": 1.3869,
"step": 1743
},
{
"epoch": 0.30992047625394287,
"grad_norm": 0.3733581259666212,
"learning_rate": 3.1609211724863555e-05,
"loss": 1.3653,
"step": 1744
},
{
"epoch": 0.3100981829490426,
"grad_norm": 0.37721119755183447,
"learning_rate": 3.1600024376890876e-05,
"loss": 1.3535,
"step": 1745
},
{
"epoch": 0.3102758896441423,
"grad_norm": 0.38250155430906946,
"learning_rate": 3.159083333873229e-05,
"loss": 1.3835,
"step": 1746
},
{
"epoch": 0.3104535963392421,
"grad_norm": 0.39623835986112244,
"learning_rate": 3.158163861331164e-05,
"loss": 1.418,
"step": 1747
},
{
"epoch": 0.31063130303434183,
"grad_norm": 0.37350308493972084,
"learning_rate": 3.1572440203553956e-05,
"loss": 1.3875,
"step": 1748
},
{
"epoch": 0.31080900972944153,
"grad_norm": 0.39653181346304683,
"learning_rate": 3.156323811238541e-05,
"loss": 1.4208,
"step": 1749
},
{
"epoch": 0.3109867164245413,
"grad_norm": 0.3882518444960288,
"learning_rate": 3.155403234273336e-05,
"loss": 1.3517,
"step": 1750
},
{
"epoch": 0.31116442311964104,
"grad_norm": 0.393033720262576,
"learning_rate": 3.154482289752634e-05,
"loss": 1.3906,
"step": 1751
},
{
"epoch": 0.3113421298147408,
"grad_norm": 0.38532747933351663,
"learning_rate": 3.153560977969405e-05,
"loss": 1.3349,
"step": 1752
},
{
"epoch": 0.3115198365098405,
"grad_norm": 0.3894550574668196,
"learning_rate": 3.152639299216734e-05,
"loss": 1.3585,
"step": 1753
},
{
"epoch": 0.31169754320494025,
"grad_norm": 0.403069140343191,
"learning_rate": 3.151717253787827e-05,
"loss": 1.4384,
"step": 1754
},
{
"epoch": 0.31187524990004,
"grad_norm": 0.3962395569014596,
"learning_rate": 3.150794841976002e-05,
"loss": 1.4045,
"step": 1755
},
{
"epoch": 0.3120529565951397,
"grad_norm": 0.39105302509666257,
"learning_rate": 3.149872064074696e-05,
"loss": 1.3931,
"step": 1756
},
{
"epoch": 0.31223066329023946,
"grad_norm": 0.39402156442460873,
"learning_rate": 3.1489489203774627e-05,
"loss": 1.4271,
"step": 1757
},
{
"epoch": 0.3124083699853392,
"grad_norm": 0.39548972614655586,
"learning_rate": 3.14802541117797e-05,
"loss": 1.3845,
"step": 1758
},
{
"epoch": 0.3125860766804389,
"grad_norm": 0.4088832651384783,
"learning_rate": 3.147101536770005e-05,
"loss": 1.3865,
"step": 1759
},
{
"epoch": 0.3127637833755387,
"grad_norm": 0.4050016567505205,
"learning_rate": 3.1461772974474686e-05,
"loss": 1.38,
"step": 1760
},
{
"epoch": 0.31294149007063843,
"grad_norm": 0.3850984271596602,
"learning_rate": 3.14525269350438e-05,
"loss": 1.3569,
"step": 1761
},
{
"epoch": 0.31311919676573813,
"grad_norm": 0.42396749266499106,
"learning_rate": 3.144327725234871e-05,
"loss": 1.4248,
"step": 1762
},
{
"epoch": 0.3132969034608379,
"grad_norm": 0.38649448237711687,
"learning_rate": 3.143402392933193e-05,
"loss": 1.3393,
"step": 1763
},
{
"epoch": 0.31347461015593764,
"grad_norm": 0.3911409421992,
"learning_rate": 3.142476696893711e-05,
"loss": 1.3562,
"step": 1764
},
{
"epoch": 0.31365231685103734,
"grad_norm": 0.3867918736327642,
"learning_rate": 3.141550637410906e-05,
"loss": 1.4108,
"step": 1765
},
{
"epoch": 0.3138300235461371,
"grad_norm": 0.3896252656015036,
"learning_rate": 3.1406242147793764e-05,
"loss": 1.3841,
"step": 1766
},
{
"epoch": 0.31400773024123685,
"grad_norm": 0.3944956213686029,
"learning_rate": 3.139697429293833e-05,
"loss": 1.3588,
"step": 1767
},
{
"epoch": 0.3141854369363366,
"grad_norm": 0.4079286098638695,
"learning_rate": 3.138770281249105e-05,
"loss": 1.4005,
"step": 1768
},
{
"epoch": 0.3143631436314363,
"grad_norm": 0.39309645416300937,
"learning_rate": 3.137842770940134e-05,
"loss": 1.3802,
"step": 1769
},
{
"epoch": 0.31454085032653606,
"grad_norm": 0.4091024994526025,
"learning_rate": 3.1369148986619805e-05,
"loss": 1.4164,
"step": 1770
},
{
"epoch": 0.3147185570216358,
"grad_norm": 0.38786909038899126,
"learning_rate": 3.1359866647098164e-05,
"loss": 1.3841,
"step": 1771
},
{
"epoch": 0.3148962637167355,
"grad_norm": 0.39562547098074097,
"learning_rate": 3.1350580693789315e-05,
"loss": 1.4103,
"step": 1772
},
{
"epoch": 0.31507397041183527,
"grad_norm": 0.3745077874003781,
"learning_rate": 3.134129112964729e-05,
"loss": 1.3807,
"step": 1773
},
{
"epoch": 0.315251677106935,
"grad_norm": 0.39175799452235666,
"learning_rate": 3.133199795762727e-05,
"loss": 1.3627,
"step": 1774
},
{
"epoch": 0.3154293838020347,
"grad_norm": 0.39908939284623834,
"learning_rate": 3.13227011806856e-05,
"loss": 1.3358,
"step": 1775
},
{
"epoch": 0.3156070904971345,
"grad_norm": 0.39521937785899547,
"learning_rate": 3.131340080177974e-05,
"loss": 1.3962,
"step": 1776
},
{
"epoch": 0.31578479719223423,
"grad_norm": 0.38997888630052696,
"learning_rate": 3.130409682386834e-05,
"loss": 1.3919,
"step": 1777
},
{
"epoch": 0.31596250388733393,
"grad_norm": 0.38023845033215115,
"learning_rate": 3.129478924991114e-05,
"loss": 1.3694,
"step": 1778
},
{
"epoch": 0.3161402105824337,
"grad_norm": 0.38997148692561384,
"learning_rate": 3.128547808286909e-05,
"loss": 1.3719,
"step": 1779
},
{
"epoch": 0.31631791727753344,
"grad_norm": 0.37419039731758624,
"learning_rate": 3.127616332570422e-05,
"loss": 1.3633,
"step": 1780
},
{
"epoch": 0.31649562397263314,
"grad_norm": 0.38967807646949976,
"learning_rate": 3.1266844981379736e-05,
"loss": 1.4169,
"step": 1781
},
{
"epoch": 0.3166733306677329,
"grad_norm": 0.38823451964989025,
"learning_rate": 3.1257523052859985e-05,
"loss": 1.4106,
"step": 1782
},
{
"epoch": 0.31685103736283265,
"grad_norm": 0.4058016638885152,
"learning_rate": 3.124819754311044e-05,
"loss": 1.4122,
"step": 1783
},
{
"epoch": 0.3170287440579324,
"grad_norm": 0.3781632866921193,
"learning_rate": 3.123886845509773e-05,
"loss": 1.3788,
"step": 1784
},
{
"epoch": 0.3172064507530321,
"grad_norm": 0.3894033146872224,
"learning_rate": 3.12295357917896e-05,
"loss": 1.39,
"step": 1785
},
{
"epoch": 0.31738415744813187,
"grad_norm": 0.39116647932889187,
"learning_rate": 3.122019955615496e-05,
"loss": 1.3638,
"step": 1786
},
{
"epoch": 0.3175618641432316,
"grad_norm": 0.38186242416560334,
"learning_rate": 3.121085975116384e-05,
"loss": 1.3783,
"step": 1787
},
{
"epoch": 0.3177395708383313,
"grad_norm": 0.39461468874914757,
"learning_rate": 3.1201516379787395e-05,
"loss": 1.4036,
"step": 1788
},
{
"epoch": 0.3179172775334311,
"grad_norm": 0.39576125879021357,
"learning_rate": 3.119216944499794e-05,
"loss": 1.4504,
"step": 1789
},
{
"epoch": 0.31809498422853083,
"grad_norm": 0.39512815692494174,
"learning_rate": 3.118281894976891e-05,
"loss": 1.3847,
"step": 1790
},
{
"epoch": 0.31827269092363053,
"grad_norm": 0.38497200040027063,
"learning_rate": 3.117346489707486e-05,
"loss": 1.3929,
"step": 1791
},
{
"epoch": 0.3184503976187303,
"grad_norm": 0.3994560964591088,
"learning_rate": 3.1164107289891505e-05,
"loss": 1.3996,
"step": 1792
},
{
"epoch": 0.31862810431383004,
"grad_norm": 0.38412965195317883,
"learning_rate": 3.115474613119567e-05,
"loss": 1.3506,
"step": 1793
},
{
"epoch": 0.31880581100892974,
"grad_norm": 0.4175920718904223,
"learning_rate": 3.1145381423965316e-05,
"loss": 1.3702,
"step": 1794
},
{
"epoch": 0.3189835177040295,
"grad_norm": 0.38055015830058964,
"learning_rate": 3.113601317117953e-05,
"loss": 1.3666,
"step": 1795
},
{
"epoch": 0.31916122439912925,
"grad_norm": 0.395568172410078,
"learning_rate": 3.1126641375818544e-05,
"loss": 1.3312,
"step": 1796
},
{
"epoch": 0.31933893109422895,
"grad_norm": 0.45522421478868835,
"learning_rate": 3.1117266040863676e-05,
"loss": 1.3463,
"step": 1797
},
{
"epoch": 0.3195166377893287,
"grad_norm": 0.39511902851133723,
"learning_rate": 3.110788716929742e-05,
"loss": 1.3467,
"step": 1798
},
{
"epoch": 0.31969434448442846,
"grad_norm": 0.3902721964386902,
"learning_rate": 3.109850476410335e-05,
"loss": 1.3976,
"step": 1799
},
{
"epoch": 0.3198720511795282,
"grad_norm": 0.387942316670381,
"learning_rate": 3.108911882826621e-05,
"loss": 1.3689,
"step": 1800
},
{
"epoch": 0.3200497578746279,
"grad_norm": 0.37620718149954774,
"learning_rate": 3.107972936477183e-05,
"loss": 1.3546,
"step": 1801
},
{
"epoch": 0.32022746456972767,
"grad_norm": 0.3911530654061591,
"learning_rate": 3.107033637660717e-05,
"loss": 1.3628,
"step": 1802
},
{
"epoch": 0.3204051712648274,
"grad_norm": 0.3884359042829105,
"learning_rate": 3.1060939866760324e-05,
"loss": 1.3455,
"step": 1803
},
{
"epoch": 0.3205828779599271,
"grad_norm": 0.38787882764424353,
"learning_rate": 3.10515398382205e-05,
"loss": 1.3618,
"step": 1804
},
{
"epoch": 0.3207605846550269,
"grad_norm": 0.400394865121721,
"learning_rate": 3.1042136293978015e-05,
"loss": 1.3984,
"step": 1805
},
{
"epoch": 0.32093829135012664,
"grad_norm": 0.3811460603755443,
"learning_rate": 3.103272923702432e-05,
"loss": 1.3869,
"step": 1806
},
{
"epoch": 0.32111599804522634,
"grad_norm": 0.3901063944836667,
"learning_rate": 3.102331867035197e-05,
"loss": 1.4101,
"step": 1807
},
{
"epoch": 0.3212937047403261,
"grad_norm": 0.3865329240358804,
"learning_rate": 3.101390459695465e-05,
"loss": 1.3406,
"step": 1808
},
{
"epoch": 0.32147141143542585,
"grad_norm": 0.37690283812206615,
"learning_rate": 3.100448701982716e-05,
"loss": 1.3722,
"step": 1809
},
{
"epoch": 0.32164911813052555,
"grad_norm": 0.535754563145781,
"learning_rate": 3.099506594196539e-05,
"loss": 1.4249,
"step": 1810
},
{
"epoch": 0.3218268248256253,
"grad_norm": 0.3915409306391556,
"learning_rate": 3.098564136636638e-05,
"loss": 1.4342,
"step": 1811
},
{
"epoch": 0.32200453152072506,
"grad_norm": 0.38385104803322634,
"learning_rate": 3.0976213296028256e-05,
"loss": 1.4316,
"step": 1812
},
{
"epoch": 0.32218223821582476,
"grad_norm": 0.39016669651862046,
"learning_rate": 3.0966781733950265e-05,
"loss": 1.4049,
"step": 1813
},
{
"epoch": 0.3223599449109245,
"grad_norm": 0.3965624734000128,
"learning_rate": 3.0957346683132765e-05,
"loss": 1.456,
"step": 1814
},
{
"epoch": 0.32253765160602427,
"grad_norm": 0.4046102928431785,
"learning_rate": 3.094790814657722e-05,
"loss": 1.4117,
"step": 1815
},
{
"epoch": 0.322715358301124,
"grad_norm": 0.38448696757874945,
"learning_rate": 3.0938466127286205e-05,
"loss": 1.356,
"step": 1816
},
{
"epoch": 0.3228930649962237,
"grad_norm": 0.39453271283639346,
"learning_rate": 3.0929020628263415e-05,
"loss": 1.3935,
"step": 1817
},
{
"epoch": 0.3230707716913235,
"grad_norm": 0.39238367855417733,
"learning_rate": 3.091957165251363e-05,
"loss": 1.3826,
"step": 1818
},
{
"epoch": 0.32324847838642323,
"grad_norm": 0.39163811237135726,
"learning_rate": 3.0910119203042755e-05,
"loss": 1.337,
"step": 1819
},
{
"epoch": 0.32342618508152293,
"grad_norm": 0.4056124126345571,
"learning_rate": 3.090066328285779e-05,
"loss": 1.4315,
"step": 1820
},
{
"epoch": 0.3236038917766227,
"grad_norm": 0.39448427777806605,
"learning_rate": 3.089120389496683e-05,
"loss": 1.3801,
"step": 1821
},
{
"epoch": 0.32378159847172244,
"grad_norm": 0.38855699607832905,
"learning_rate": 3.0881741042379096e-05,
"loss": 1.3648,
"step": 1822
},
{
"epoch": 0.32395930516682214,
"grad_norm": 0.37276922751948827,
"learning_rate": 3.0872274728104895e-05,
"loss": 1.3506,
"step": 1823
},
{
"epoch": 0.3241370118619219,
"grad_norm": 0.38987456226339795,
"learning_rate": 3.086280495515564e-05,
"loss": 1.3584,
"step": 1824
},
{
"epoch": 0.32431471855702165,
"grad_norm": 0.38526011600193577,
"learning_rate": 3.085333172654385e-05,
"loss": 1.3709,
"step": 1825
},
{
"epoch": 0.32449242525212135,
"grad_norm": 0.45335425369041166,
"learning_rate": 3.084385504528313e-05,
"loss": 1.3841,
"step": 1826
},
{
"epoch": 0.3246701319472211,
"grad_norm": 0.387922570189183,
"learning_rate": 3.083437491438819e-05,
"loss": 1.4075,
"step": 1827
},
{
"epoch": 0.32484783864232086,
"grad_norm": 0.3734787403936944,
"learning_rate": 3.082489133687484e-05,
"loss": 1.3518,
"step": 1828
},
{
"epoch": 0.32502554533742056,
"grad_norm": 0.383743854418042,
"learning_rate": 3.0815404315759985e-05,
"loss": 1.3829,
"step": 1829
},
{
"epoch": 0.3252032520325203,
"grad_norm": 0.3889626385449473,
"learning_rate": 3.080591385406162e-05,
"loss": 1.4266,
"step": 1830
},
{
"epoch": 0.3253809587276201,
"grad_norm": 0.39903475109986664,
"learning_rate": 3.079641995479885e-05,
"loss": 1.3871,
"step": 1831
},
{
"epoch": 0.32555866542271983,
"grad_norm": 0.37241242229289784,
"learning_rate": 3.078692262099185e-05,
"loss": 1.3643,
"step": 1832
},
{
"epoch": 0.32573637211781953,
"grad_norm": 0.4004546603347606,
"learning_rate": 3.0777421855661914e-05,
"loss": 1.4036,
"step": 1833
},
{
"epoch": 0.3259140788129193,
"grad_norm": 0.37718501955241385,
"learning_rate": 3.076791766183141e-05,
"loss": 1.3653,
"step": 1834
},
{
"epoch": 0.32609178550801904,
"grad_norm": 0.3801857474043455,
"learning_rate": 3.075841004252379e-05,
"loss": 1.3816,
"step": 1835
},
{
"epoch": 0.32626949220311874,
"grad_norm": 0.37739169189773436,
"learning_rate": 3.074889900076362e-05,
"loss": 1.3778,
"step": 1836
},
{
"epoch": 0.3264471988982185,
"grad_norm": 0.3882920803440394,
"learning_rate": 3.073938453957653e-05,
"loss": 1.3665,
"step": 1837
},
{
"epoch": 0.32662490559331825,
"grad_norm": 0.38773649864919885,
"learning_rate": 3.072986666198926e-05,
"loss": 1.3628,
"step": 1838
},
{
"epoch": 0.32680261228841795,
"grad_norm": 0.3843587978553194,
"learning_rate": 3.072034537102962e-05,
"loss": 1.4058,
"step": 1839
},
{
"epoch": 0.3269803189835177,
"grad_norm": 0.37421893457961775,
"learning_rate": 3.071082066972651e-05,
"loss": 1.3552,
"step": 1840
},
{
"epoch": 0.32715802567861746,
"grad_norm": 0.372967803012153,
"learning_rate": 3.070129256110992e-05,
"loss": 1.3665,
"step": 1841
},
{
"epoch": 0.32733573237371716,
"grad_norm": 0.37151984631524004,
"learning_rate": 3.069176104821092e-05,
"loss": 1.3871,
"step": 1842
},
{
"epoch": 0.3275134390688169,
"grad_norm": 0.3789196207398101,
"learning_rate": 3.0682226134061667e-05,
"loss": 1.377,
"step": 1843
},
{
"epoch": 0.32769114576391667,
"grad_norm": 0.37938167458298927,
"learning_rate": 3.067268782169538e-05,
"loss": 1.3915,
"step": 1844
},
{
"epoch": 0.32786885245901637,
"grad_norm": 0.38403679799423507,
"learning_rate": 3.066314611414639e-05,
"loss": 1.3601,
"step": 1845
},
{
"epoch": 0.3280465591541161,
"grad_norm": 0.3899144176602821,
"learning_rate": 3.065360101445011e-05,
"loss": 1.4021,
"step": 1846
},
{
"epoch": 0.3282242658492159,
"grad_norm": 0.3867651643913956,
"learning_rate": 3.064405252564298e-05,
"loss": 1.3955,
"step": 1847
},
{
"epoch": 0.32840197254431563,
"grad_norm": 0.3922295973025351,
"learning_rate": 3.063450065076257e-05,
"loss": 1.3983,
"step": 1848
},
{
"epoch": 0.32857967923941533,
"grad_norm": 0.3796860790098293,
"learning_rate": 3.062494539284752e-05,
"loss": 1.3919,
"step": 1849
},
{
"epoch": 0.3287573859345151,
"grad_norm": 0.37714265773407546,
"learning_rate": 3.0615386754937535e-05,
"loss": 1.3553,
"step": 1850
},
{
"epoch": 0.32893509262961484,
"grad_norm": 0.3833067508454648,
"learning_rate": 3.060582474007338e-05,
"loss": 1.3528,
"step": 1851
},
{
"epoch": 0.32911279932471454,
"grad_norm": 0.4038329668125259,
"learning_rate": 3.0596259351296925e-05,
"loss": 1.3974,
"step": 1852
},
{
"epoch": 0.3292905060198143,
"grad_norm": 0.4234736647581316,
"learning_rate": 3.058669059165111e-05,
"loss": 1.3762,
"step": 1853
},
{
"epoch": 0.32946821271491405,
"grad_norm": 0.3772030108499336,
"learning_rate": 3.0577118464179915e-05,
"loss": 1.3912,
"step": 1854
},
{
"epoch": 0.32964591941001375,
"grad_norm": 0.3937113218893644,
"learning_rate": 3.0567542971928426e-05,
"loss": 1.4124,
"step": 1855
},
{
"epoch": 0.3298236261051135,
"grad_norm": 0.3728529787925495,
"learning_rate": 3.055796411794279e-05,
"loss": 1.3481,
"step": 1856
},
{
"epoch": 0.33000133280021327,
"grad_norm": 0.3854430443777905,
"learning_rate": 3.054838190527021e-05,
"loss": 1.3935,
"step": 1857
},
{
"epoch": 0.33017903949531296,
"grad_norm": 0.38862383822723556,
"learning_rate": 3.053879633695898e-05,
"loss": 1.365,
"step": 1858
},
{
"epoch": 0.3303567461904127,
"grad_norm": 0.39362543083356244,
"learning_rate": 3.052920741605845e-05,
"loss": 1.3589,
"step": 1859
},
{
"epoch": 0.3305344528855125,
"grad_norm": 0.3965340724162032,
"learning_rate": 3.051961514561902e-05,
"loss": 1.4306,
"step": 1860
},
{
"epoch": 0.3307121595806122,
"grad_norm": 0.3865733891090966,
"learning_rate": 3.051001952869219e-05,
"loss": 1.4036,
"step": 1861
},
{
"epoch": 0.33088986627571193,
"grad_norm": 0.3949717795113669,
"learning_rate": 3.050042056833049e-05,
"loss": 1.3993,
"step": 1862
},
{
"epoch": 0.3310675729708117,
"grad_norm": 0.38595486821388914,
"learning_rate": 3.0490818267587543e-05,
"loss": 1.4456,
"step": 1863
},
{
"epoch": 0.33124527966591144,
"grad_norm": 0.38360853627267566,
"learning_rate": 3.0481212629518015e-05,
"loss": 1.3711,
"step": 1864
},
{
"epoch": 0.33142298636101114,
"grad_norm": 0.392302215551794,
"learning_rate": 3.047160365717764e-05,
"loss": 1.4177,
"step": 1865
},
{
"epoch": 0.3316006930561109,
"grad_norm": 0.3634513789296566,
"learning_rate": 3.046199135362322e-05,
"loss": 1.3653,
"step": 1866
},
{
"epoch": 0.33177839975121065,
"grad_norm": 0.39597990726862486,
"learning_rate": 3.0452375721912593e-05,
"loss": 1.4257,
"step": 1867
},
{
"epoch": 0.33195610644631035,
"grad_norm": 0.37900022924289106,
"learning_rate": 3.044275676510469e-05,
"loss": 1.3733,
"step": 1868
},
{
"epoch": 0.3321338131414101,
"grad_norm": 0.3899025110248986,
"learning_rate": 3.043313448625948e-05,
"loss": 1.4035,
"step": 1869
},
{
"epoch": 0.33231151983650986,
"grad_norm": 0.3908524853367731,
"learning_rate": 3.0423508888437977e-05,
"loss": 1.3984,
"step": 1870
},
{
"epoch": 0.33248922653160956,
"grad_norm": 0.38571811035955983,
"learning_rate": 3.0413879974702283e-05,
"loss": 1.3949,
"step": 1871
},
{
"epoch": 0.3326669332267093,
"grad_norm": 0.40108424735035103,
"learning_rate": 3.0404247748115523e-05,
"loss": 1.4128,
"step": 1872
},
{
"epoch": 0.33284463992180907,
"grad_norm": 0.39620608190377626,
"learning_rate": 3.0394612211741897e-05,
"loss": 1.3904,
"step": 1873
},
{
"epoch": 0.33302234661690877,
"grad_norm": 0.3820450499155962,
"learning_rate": 3.0384973368646636e-05,
"loss": 1.3677,
"step": 1874
},
{
"epoch": 0.3332000533120085,
"grad_norm": 0.3961286799861153,
"learning_rate": 3.0375331221896058e-05,
"loss": 1.4322,
"step": 1875
},
{
"epoch": 0.3333777600071083,
"grad_norm": 0.39887544912544043,
"learning_rate": 3.0365685774557497e-05,
"loss": 1.3955,
"step": 1876
},
{
"epoch": 0.333555466702208,
"grad_norm": 0.38383035732080034,
"learning_rate": 3.035603702969936e-05,
"loss": 1.4242,
"step": 1877
},
{
"epoch": 0.33373317339730774,
"grad_norm": 0.38398135607668943,
"learning_rate": 3.0346384990391082e-05,
"loss": 1.3457,
"step": 1878
},
{
"epoch": 0.3339108800924075,
"grad_norm": 0.37226405672558444,
"learning_rate": 3.033672965970317e-05,
"loss": 1.3507,
"step": 1879
},
{
"epoch": 0.33408858678750725,
"grad_norm": 0.38951460100340457,
"learning_rate": 3.032707104070717e-05,
"loss": 1.3936,
"step": 1880
},
{
"epoch": 0.33426629348260695,
"grad_norm": 0.3891401229424884,
"learning_rate": 3.031740913647565e-05,
"loss": 1.3992,
"step": 1881
},
{
"epoch": 0.3344440001777067,
"grad_norm": 0.38051407311218477,
"learning_rate": 3.0307743950082263e-05,
"loss": 1.3608,
"step": 1882
},
{
"epoch": 0.33462170687280646,
"grad_norm": 0.3948498895115259,
"learning_rate": 3.029807548460168e-05,
"loss": 1.4108,
"step": 1883
},
{
"epoch": 0.33479941356790616,
"grad_norm": 0.374286337209788,
"learning_rate": 3.0288403743109622e-05,
"loss": 1.3555,
"step": 1884
},
{
"epoch": 0.3349771202630059,
"grad_norm": 0.3935545848108091,
"learning_rate": 3.027872872868285e-05,
"loss": 1.3866,
"step": 1885
},
{
"epoch": 0.33515482695810567,
"grad_norm": 0.39153795710842026,
"learning_rate": 3.0269050444399167e-05,
"loss": 1.3792,
"step": 1886
},
{
"epoch": 0.33533253365320537,
"grad_norm": 0.3941631798044439,
"learning_rate": 3.0259368893337426e-05,
"loss": 1.3972,
"step": 1887
},
{
"epoch": 0.3355102403483051,
"grad_norm": 0.37568110424062484,
"learning_rate": 3.024968407857749e-05,
"loss": 1.3985,
"step": 1888
},
{
"epoch": 0.3356879470434049,
"grad_norm": 0.40183265235860816,
"learning_rate": 3.0239996003200308e-05,
"loss": 1.3711,
"step": 1889
},
{
"epoch": 0.3358656537385046,
"grad_norm": 0.387374069403729,
"learning_rate": 3.0230304670287815e-05,
"loss": 1.3646,
"step": 1890
},
{
"epoch": 0.33604336043360433,
"grad_norm": 0.3742482275920783,
"learning_rate": 3.022061008292303e-05,
"loss": 1.3503,
"step": 1891
},
{
"epoch": 0.3362210671287041,
"grad_norm": 0.38828119066908645,
"learning_rate": 3.0210912244189968e-05,
"loss": 1.3366,
"step": 1892
},
{
"epoch": 0.3363987738238038,
"grad_norm": 0.3870104285995852,
"learning_rate": 3.0201211157173684e-05,
"loss": 1.3752,
"step": 1893
},
{
"epoch": 0.33657648051890354,
"grad_norm": 0.3688781683040064,
"learning_rate": 3.0191506824960296e-05,
"loss": 1.3752,
"step": 1894
},
{
"epoch": 0.3367541872140033,
"grad_norm": 0.3857567828477677,
"learning_rate": 3.018179925063693e-05,
"loss": 1.3568,
"step": 1895
},
{
"epoch": 0.33693189390910305,
"grad_norm": 0.40288149024256487,
"learning_rate": 3.017208843729174e-05,
"loss": 1.4,
"step": 1896
},
{
"epoch": 0.33710960060420275,
"grad_norm": 0.36727376208495327,
"learning_rate": 3.016237438801392e-05,
"loss": 1.3334,
"step": 1897
},
{
"epoch": 0.3372873072993025,
"grad_norm": 0.40278316010113185,
"learning_rate": 3.015265710589371e-05,
"loss": 1.371,
"step": 1898
},
{
"epoch": 0.33746501399440226,
"grad_norm": 0.37398059995907673,
"learning_rate": 3.014293659402234e-05,
"loss": 1.4052,
"step": 1899
},
{
"epoch": 0.33764272068950196,
"grad_norm": 0.4105041104440446,
"learning_rate": 3.0133212855492083e-05,
"loss": 1.4098,
"step": 1900
},
{
"epoch": 0.3378204273846017,
"grad_norm": 0.3651968031441386,
"learning_rate": 3.012348589339626e-05,
"loss": 1.3564,
"step": 1901
},
{
"epoch": 0.3379981340797015,
"grad_norm": 0.38808186301835124,
"learning_rate": 3.0113755710829192e-05,
"loss": 1.3815,
"step": 1902
},
{
"epoch": 0.3381758407748012,
"grad_norm": 0.38261444070690964,
"learning_rate": 3.010402231088624e-05,
"loss": 1.3336,
"step": 1903
},
{
"epoch": 0.33835354746990093,
"grad_norm": 0.3754014568108369,
"learning_rate": 3.009428569666377e-05,
"loss": 1.3585,
"step": 1904
},
{
"epoch": 0.3385312541650007,
"grad_norm": 0.39535983893845417,
"learning_rate": 3.0084545871259187e-05,
"loss": 1.3906,
"step": 1905
},
{
"epoch": 0.3387089608601004,
"grad_norm": 0.38993577189402234,
"learning_rate": 3.007480283777092e-05,
"loss": 1.3723,
"step": 1906
},
{
"epoch": 0.33888666755520014,
"grad_norm": 0.3880794439415563,
"learning_rate": 3.00650565992984e-05,
"loss": 1.3775,
"step": 1907
},
{
"epoch": 0.3390643742502999,
"grad_norm": 0.3749440109607459,
"learning_rate": 3.0055307158942096e-05,
"loss": 1.3621,
"step": 1908
},
{
"epoch": 0.3392420809453996,
"grad_norm": 0.37475587172963887,
"learning_rate": 3.0045554519803483e-05,
"loss": 1.3542,
"step": 1909
},
{
"epoch": 0.33941978764049935,
"grad_norm": 0.3798318232025045,
"learning_rate": 3.0035798684985074e-05,
"loss": 1.4002,
"step": 1910
},
{
"epoch": 0.3395974943355991,
"grad_norm": 0.39700324927199315,
"learning_rate": 3.002603965759036e-05,
"loss": 1.3449,
"step": 1911
},
{
"epoch": 0.33977520103069886,
"grad_norm": 0.36502690026577905,
"learning_rate": 3.0016277440723883e-05,
"loss": 1.3643,
"step": 1912
},
{
"epoch": 0.33995290772579856,
"grad_norm": 0.3774449465772432,
"learning_rate": 3.000651203749119e-05,
"loss": 1.3897,
"step": 1913
},
{
"epoch": 0.3401306144208983,
"grad_norm": 0.39435452885313843,
"learning_rate": 2.999674345099884e-05,
"loss": 1.4203,
"step": 1914
},
{
"epoch": 0.34030832111599807,
"grad_norm": 0.3683068925678174,
"learning_rate": 2.99869716843544e-05,
"loss": 1.3556,
"step": 1915
},
{
"epoch": 0.34048602781109777,
"grad_norm": 0.375367523209389,
"learning_rate": 2.9977196740666447e-05,
"loss": 1.3545,
"step": 1916
},
{
"epoch": 0.3406637345061975,
"grad_norm": 0.39426383316586766,
"learning_rate": 2.9967418623044594e-05,
"loss": 1.4145,
"step": 1917
},
{
"epoch": 0.3408414412012973,
"grad_norm": 0.402180090509128,
"learning_rate": 2.9957637334599417e-05,
"loss": 1.4081,
"step": 1918
},
{
"epoch": 0.341019147896397,
"grad_norm": 0.3920066787215482,
"learning_rate": 2.9947852878442545e-05,
"loss": 1.4138,
"step": 1919
},
{
"epoch": 0.34119685459149673,
"grad_norm": 0.38020355256010196,
"learning_rate": 2.99380652576866e-05,
"loss": 1.3876,
"step": 1920
},
{
"epoch": 0.3413745612865965,
"grad_norm": 0.3859004929349077,
"learning_rate": 2.9928274475445206e-05,
"loss": 1.3871,
"step": 1921
},
{
"epoch": 0.3415522679816962,
"grad_norm": 0.37758309523822325,
"learning_rate": 2.9918480534832985e-05,
"loss": 1.3677,
"step": 1922
},
{
"epoch": 0.34172997467679594,
"grad_norm": 0.37932571384976704,
"learning_rate": 2.990868343896558e-05,
"loss": 1.3612,
"step": 1923
},
{
"epoch": 0.3419076813718957,
"grad_norm": 0.3684495759062843,
"learning_rate": 2.9898883190959637e-05,
"loss": 1.3911,
"step": 1924
},
{
"epoch": 0.3420853880669954,
"grad_norm": 0.37667640298155297,
"learning_rate": 2.9889079793932788e-05,
"loss": 1.3677,
"step": 1925
},
{
"epoch": 0.34226309476209515,
"grad_norm": 0.3762706845981104,
"learning_rate": 2.9879273251003692e-05,
"loss": 1.3988,
"step": 1926
},
{
"epoch": 0.3424408014571949,
"grad_norm": 0.37300303341742524,
"learning_rate": 2.9869463565291982e-05,
"loss": 1.3369,
"step": 1927
},
{
"epoch": 0.34261850815229467,
"grad_norm": 0.37888849359110005,
"learning_rate": 2.9859650739918307e-05,
"loss": 1.3807,
"step": 1928
},
{
"epoch": 0.34279621484739436,
"grad_norm": 0.375897871554685,
"learning_rate": 2.9849834778004315e-05,
"loss": 1.3409,
"step": 1929
},
{
"epoch": 0.3429739215424941,
"grad_norm": 0.3881702629242823,
"learning_rate": 2.984001568267264e-05,
"loss": 1.3731,
"step": 1930
},
{
"epoch": 0.3431516282375939,
"grad_norm": 0.3828745704791317,
"learning_rate": 2.9830193457046932e-05,
"loss": 1.4209,
"step": 1931
},
{
"epoch": 0.3433293349326936,
"grad_norm": 0.38727275103290604,
"learning_rate": 2.982036810425182e-05,
"loss": 1.3731,
"step": 1932
},
{
"epoch": 0.34350704162779333,
"grad_norm": 0.37796242512505795,
"learning_rate": 2.9810539627412932e-05,
"loss": 1.3555,
"step": 1933
},
{
"epoch": 0.3436847483228931,
"grad_norm": 0.37151394273213734,
"learning_rate": 2.9800708029656896e-05,
"loss": 1.3798,
"step": 1934
},
{
"epoch": 0.3438624550179928,
"grad_norm": 0.3893574703625444,
"learning_rate": 2.9790873314111322e-05,
"loss": 1.3823,
"step": 1935
},
{
"epoch": 0.34404016171309254,
"grad_norm": 0.37049082787542664,
"learning_rate": 2.9781035483904824e-05,
"loss": 1.365,
"step": 1936
},
{
"epoch": 0.3442178684081923,
"grad_norm": 0.3722409855075822,
"learning_rate": 2.9771194542166996e-05,
"loss": 1.3785,
"step": 1937
},
{
"epoch": 0.344395575103292,
"grad_norm": 0.37863651900710327,
"learning_rate": 2.976135049202843e-05,
"loss": 1.3631,
"step": 1938
},
{
"epoch": 0.34457328179839175,
"grad_norm": 0.3918775544801439,
"learning_rate": 2.9751503336620698e-05,
"loss": 1.3494,
"step": 1939
},
{
"epoch": 0.3447509884934915,
"grad_norm": 0.37833150020135425,
"learning_rate": 2.9741653079076375e-05,
"loss": 1.3833,
"step": 1940
},
{
"epoch": 0.3449286951885912,
"grad_norm": 0.36852505433159194,
"learning_rate": 2.9731799722529007e-05,
"loss": 1.3638,
"step": 1941
},
{
"epoch": 0.34510640188369096,
"grad_norm": 0.39585830201783523,
"learning_rate": 2.9721943270113123e-05,
"loss": 1.4272,
"step": 1942
},
{
"epoch": 0.3452841085787907,
"grad_norm": 0.38146744683522815,
"learning_rate": 2.9712083724964268e-05,
"loss": 1.391,
"step": 1943
},
{
"epoch": 0.34546181527389047,
"grad_norm": 0.3704157349795045,
"learning_rate": 2.9702221090218932e-05,
"loss": 1.3541,
"step": 1944
},
{
"epoch": 0.34563952196899017,
"grad_norm": 0.3893696245867969,
"learning_rate": 2.9692355369014607e-05,
"loss": 1.3924,
"step": 1945
},
{
"epoch": 0.3458172286640899,
"grad_norm": 0.3780282244944519,
"learning_rate": 2.9682486564489764e-05,
"loss": 1.3797,
"step": 1946
},
{
"epoch": 0.3459949353591897,
"grad_norm": 0.37974018855854225,
"learning_rate": 2.9672614679783865e-05,
"loss": 1.4155,
"step": 1947
},
{
"epoch": 0.3461726420542894,
"grad_norm": 0.3842283437624976,
"learning_rate": 2.966273971803733e-05,
"loss": 1.3824,
"step": 1948
},
{
"epoch": 0.34635034874938914,
"grad_norm": 0.38565083096901065,
"learning_rate": 2.9652861682391574e-05,
"loss": 1.3701,
"step": 1949
},
{
"epoch": 0.3465280554444889,
"grad_norm": 0.373062882547069,
"learning_rate": 2.9642980575988986e-05,
"loss": 1.3938,
"step": 1950
},
{
"epoch": 0.3467057621395886,
"grad_norm": 0.38257966767692486,
"learning_rate": 2.9633096401972934e-05,
"loss": 1.4018,
"step": 1951
},
{
"epoch": 0.34688346883468835,
"grad_norm": 0.38371141784994217,
"learning_rate": 2.962320916348776e-05,
"loss": 1.3939,
"step": 1952
},
{
"epoch": 0.3470611755297881,
"grad_norm": 0.3848642393824481,
"learning_rate": 2.9613318863678773e-05,
"loss": 1.3623,
"step": 1953
},
{
"epoch": 0.3472388822248878,
"grad_norm": 0.3721844602724466,
"learning_rate": 2.9603425505692273e-05,
"loss": 1.4262,
"step": 1954
},
{
"epoch": 0.34741658891998756,
"grad_norm": 0.3882685648677181,
"learning_rate": 2.959352909267552e-05,
"loss": 1.382,
"step": 1955
},
{
"epoch": 0.3475942956150873,
"grad_norm": 0.3781500127791112,
"learning_rate": 2.958362962777675e-05,
"loss": 1.3915,
"step": 1956
},
{
"epoch": 0.347772002310187,
"grad_norm": 0.39529503006282524,
"learning_rate": 2.9573727114145162e-05,
"loss": 1.3783,
"step": 1957
},
{
"epoch": 0.34794970900528677,
"grad_norm": 0.3787054447113773,
"learning_rate": 2.956382155493094e-05,
"loss": 1.3632,
"step": 1958
},
{
"epoch": 0.3481274157003865,
"grad_norm": 0.397627679260632,
"learning_rate": 2.9553912953285226e-05,
"loss": 1.3589,
"step": 1959
},
{
"epoch": 0.3483051223954863,
"grad_norm": 0.37004364746831403,
"learning_rate": 2.9544001312360126e-05,
"loss": 1.3428,
"step": 1960
},
{
"epoch": 0.348482829090586,
"grad_norm": 0.3970292251817273,
"learning_rate": 2.9534086635308728e-05,
"loss": 1.3607,
"step": 1961
},
{
"epoch": 0.34866053578568573,
"grad_norm": 0.4064331271231644,
"learning_rate": 2.9524168925285077e-05,
"loss": 1.3848,
"step": 1962
},
{
"epoch": 0.3488382424807855,
"grad_norm": 0.3946971025669671,
"learning_rate": 2.951424818544418e-05,
"loss": 1.4168,
"step": 1963
},
{
"epoch": 0.3490159491758852,
"grad_norm": 0.39269868011457854,
"learning_rate": 2.9504324418942015e-05,
"loss": 1.3604,
"step": 1964
},
{
"epoch": 0.34919365587098494,
"grad_norm": 0.39745485270026465,
"learning_rate": 2.949439762893551e-05,
"loss": 1.431,
"step": 1965
},
{
"epoch": 0.3493713625660847,
"grad_norm": 0.40799641605318854,
"learning_rate": 2.9484467818582576e-05,
"loss": 1.3869,
"step": 1966
},
{
"epoch": 0.3495490692611844,
"grad_norm": 0.37996918608627567,
"learning_rate": 2.947453499104206e-05,
"loss": 1.368,
"step": 1967
},
{
"epoch": 0.34972677595628415,
"grad_norm": 0.38628262289733706,
"learning_rate": 2.9464599149473786e-05,
"loss": 1.3765,
"step": 1968
},
{
"epoch": 0.3499044826513839,
"grad_norm": 0.37526432911010715,
"learning_rate": 2.9454660297038535e-05,
"loss": 1.3631,
"step": 1969
},
{
"epoch": 0.3500821893464836,
"grad_norm": 0.38540261846975876,
"learning_rate": 2.9444718436898045e-05,
"loss": 1.3509,
"step": 1970
},
{
"epoch": 0.35025989604158336,
"grad_norm": 0.3685583964816267,
"learning_rate": 2.9434773572215e-05,
"loss": 1.3866,
"step": 1971
},
{
"epoch": 0.3504376027366831,
"grad_norm": 0.3768596267780326,
"learning_rate": 2.9424825706153047e-05,
"loss": 1.4072,
"step": 1972
},
{
"epoch": 0.3506153094317828,
"grad_norm": 0.3670700781753171,
"learning_rate": 2.94148748418768e-05,
"loss": 1.3617,
"step": 1973
},
{
"epoch": 0.3507930161268826,
"grad_norm": 0.3775949675562243,
"learning_rate": 2.940492098255182e-05,
"loss": 1.3893,
"step": 1974
},
{
"epoch": 0.35097072282198233,
"grad_norm": 0.38501476018501546,
"learning_rate": 2.9394964131344595e-05,
"loss": 1.3853,
"step": 1975
},
{
"epoch": 0.3511484295170821,
"grad_norm": 0.3873266690455701,
"learning_rate": 2.9385004291422605e-05,
"loss": 1.3677,
"step": 1976
},
{
"epoch": 0.3513261362121818,
"grad_norm": 0.39068746653904796,
"learning_rate": 2.9375041465954255e-05,
"loss": 1.3609,
"step": 1977
},
{
"epoch": 0.35150384290728154,
"grad_norm": 0.39181641678116447,
"learning_rate": 2.9365075658108916e-05,
"loss": 1.3928,
"step": 1978
},
{
"epoch": 0.3516815496023813,
"grad_norm": 0.37704436442893074,
"learning_rate": 2.935510687105688e-05,
"loss": 1.3607,
"step": 1979
},
{
"epoch": 0.351859256297481,
"grad_norm": 0.3819133370272372,
"learning_rate": 2.9345135107969427e-05,
"loss": 1.3829,
"step": 1980
},
{
"epoch": 0.35203696299258075,
"grad_norm": 0.37519110865522765,
"learning_rate": 2.933516037201875e-05,
"loss": 1.3501,
"step": 1981
},
{
"epoch": 0.3522146696876805,
"grad_norm": 0.3725068224905714,
"learning_rate": 2.9325182666378e-05,
"loss": 1.3645,
"step": 1982
},
{
"epoch": 0.3523923763827802,
"grad_norm": 0.3760451462851054,
"learning_rate": 2.9315201994221283e-05,
"loss": 1.3921,
"step": 1983
},
{
"epoch": 0.35257008307787996,
"grad_norm": 0.3764395018134201,
"learning_rate": 2.9305218358723625e-05,
"loss": 1.3691,
"step": 1984
},
{
"epoch": 0.3527477897729797,
"grad_norm": 0.38721279760015465,
"learning_rate": 2.929523176306102e-05,
"loss": 1.438,
"step": 1985
},
{
"epoch": 0.3529254964680794,
"grad_norm": 0.37709983837598404,
"learning_rate": 2.928524221041038e-05,
"loss": 1.3587,
"step": 1986
},
{
"epoch": 0.35310320316317917,
"grad_norm": 0.3660185788798451,
"learning_rate": 2.9275249703949578e-05,
"loss": 1.3764,
"step": 1987
},
{
"epoch": 0.3532809098582789,
"grad_norm": 0.3941656042794553,
"learning_rate": 2.9265254246857422e-05,
"loss": 1.4131,
"step": 1988
},
{
"epoch": 0.3534586165533786,
"grad_norm": 0.38806306223734477,
"learning_rate": 2.9255255842313643e-05,
"loss": 1.4477,
"step": 1989
},
{
"epoch": 0.3536363232484784,
"grad_norm": 0.38378703449397306,
"learning_rate": 2.9245254493498932e-05,
"loss": 1.3864,
"step": 1990
},
{
"epoch": 0.35381402994357813,
"grad_norm": 0.37403048717502896,
"learning_rate": 2.9235250203594897e-05,
"loss": 1.4094,
"step": 1991
},
{
"epoch": 0.3539917366386779,
"grad_norm": 0.3772346704712321,
"learning_rate": 2.9225242975784097e-05,
"loss": 1.36,
"step": 1992
},
{
"epoch": 0.3541694433337776,
"grad_norm": 0.3902810922419991,
"learning_rate": 2.921523281325002e-05,
"loss": 1.3977,
"step": 1993
},
{
"epoch": 0.35434715002887734,
"grad_norm": 0.3650888292826055,
"learning_rate": 2.9205219719177083e-05,
"loss": 1.3853,
"step": 1994
},
{
"epoch": 0.3545248567239771,
"grad_norm": 0.3774722850386825,
"learning_rate": 2.9195203696750643e-05,
"loss": 1.3401,
"step": 1995
},
{
"epoch": 0.3547025634190768,
"grad_norm": 0.3626149600075632,
"learning_rate": 2.918518474915698e-05,
"loss": 1.3031,
"step": 1996
},
{
"epoch": 0.35488027011417655,
"grad_norm": 0.3752606635367311,
"learning_rate": 2.917516287958332e-05,
"loss": 1.3951,
"step": 1997
},
{
"epoch": 0.3550579768092763,
"grad_norm": 0.37019867220992364,
"learning_rate": 2.9165138091217798e-05,
"loss": 1.323,
"step": 1998
},
{
"epoch": 0.355235683504376,
"grad_norm": 0.3685861147042038,
"learning_rate": 2.9155110387249486e-05,
"loss": 1.3425,
"step": 1999
},
{
"epoch": 0.35541339019947576,
"grad_norm": 0.3848290403978416,
"learning_rate": 2.9145079770868398e-05,
"loss": 1.3716,
"step": 2000
},
{
"epoch": 0.3555910968945755,
"grad_norm": 0.3693261760822017,
"learning_rate": 2.913504624526545e-05,
"loss": 1.3785,
"step": 2001
},
{
"epoch": 0.3557688035896752,
"grad_norm": 0.3707254957868877,
"learning_rate": 2.91250098136325e-05,
"loss": 1.3403,
"step": 2002
},
{
"epoch": 0.355946510284775,
"grad_norm": 0.3843575197454317,
"learning_rate": 2.911497047916232e-05,
"loss": 1.3743,
"step": 2003
},
{
"epoch": 0.35612421697987473,
"grad_norm": 0.37178865194268135,
"learning_rate": 2.9104928245048624e-05,
"loss": 1.3898,
"step": 2004
},
{
"epoch": 0.35630192367497443,
"grad_norm": 0.37513690080854045,
"learning_rate": 2.909488311448602e-05,
"loss": 1.3447,
"step": 2005
},
{
"epoch": 0.3564796303700742,
"grad_norm": 0.3867753904105246,
"learning_rate": 2.9084835090670065e-05,
"loss": 1.4281,
"step": 2006
},
{
"epoch": 0.35665733706517394,
"grad_norm": 0.36652520664868254,
"learning_rate": 2.907478417679722e-05,
"loss": 1.3405,
"step": 2007
},
{
"epoch": 0.3568350437602737,
"grad_norm": 0.3704892000065961,
"learning_rate": 2.9064730376064866e-05,
"loss": 1.3707,
"step": 2008
},
{
"epoch": 0.3570127504553734,
"grad_norm": 0.375891552531725,
"learning_rate": 2.9054673691671317e-05,
"loss": 1.3933,
"step": 2009
},
{
"epoch": 0.35719045715047315,
"grad_norm": 0.363650574782824,
"learning_rate": 2.9044614126815775e-05,
"loss": 1.3729,
"step": 2010
},
{
"epoch": 0.3573681638455729,
"grad_norm": 0.376110213674763,
"learning_rate": 2.90345516846984e-05,
"loss": 1.3852,
"step": 2011
},
{
"epoch": 0.3575458705406726,
"grad_norm": 0.39309439348070885,
"learning_rate": 2.9024486368520218e-05,
"loss": 1.386,
"step": 2012
},
{
"epoch": 0.35772357723577236,
"grad_norm": 0.3760635796988099,
"learning_rate": 2.9014418181483216e-05,
"loss": 1.3927,
"step": 2013
},
{
"epoch": 0.3579012839308721,
"grad_norm": 0.37910246522314794,
"learning_rate": 2.9004347126790266e-05,
"loss": 1.4208,
"step": 2014
},
{
"epoch": 0.3580789906259718,
"grad_norm": 0.3965835030124175,
"learning_rate": 2.8994273207645164e-05,
"loss": 1.4005,
"step": 2015
},
{
"epoch": 0.35825669732107157,
"grad_norm": 0.38406955431589307,
"learning_rate": 2.8984196427252606e-05,
"loss": 1.3768,
"step": 2016
},
{
"epoch": 0.3584344040161713,
"grad_norm": 0.3753176602541864,
"learning_rate": 2.8974116788818207e-05,
"loss": 1.3687,
"step": 2017
},
{
"epoch": 0.358612110711271,
"grad_norm": 0.4004400912156542,
"learning_rate": 2.8964034295548497e-05,
"loss": 1.3593,
"step": 2018
},
{
"epoch": 0.3587898174063708,
"grad_norm": 0.3754131843013086,
"learning_rate": 2.8953948950650893e-05,
"loss": 1.3423,
"step": 2019
},
{
"epoch": 0.35896752410147054,
"grad_norm": 0.38474819716759195,
"learning_rate": 2.8943860757333754e-05,
"loss": 1.349,
"step": 2020
},
{
"epoch": 0.35914523079657024,
"grad_norm": 0.3710225265582783,
"learning_rate": 2.89337697188063e-05,
"loss": 1.2953,
"step": 2021
},
{
"epoch": 0.35932293749167,
"grad_norm": 0.39226363833021793,
"learning_rate": 2.89236758382787e-05,
"loss": 1.4019,
"step": 2022
},
{
"epoch": 0.35950064418676975,
"grad_norm": 0.38259423648887203,
"learning_rate": 2.8913579118961993e-05,
"loss": 1.3873,
"step": 2023
},
{
"epoch": 0.3596783508818695,
"grad_norm": 0.3618017870914052,
"learning_rate": 2.8903479564068138e-05,
"loss": 1.3305,
"step": 2024
},
{
"epoch": 0.3598560575769692,
"grad_norm": 0.3837616956001988,
"learning_rate": 2.8893377176810004e-05,
"loss": 1.4158,
"step": 2025
},
{
"epoch": 0.36003376427206896,
"grad_norm": 0.37604389160424323,
"learning_rate": 2.888327196040134e-05,
"loss": 1.3914,
"step": 2026
},
{
"epoch": 0.3602114709671687,
"grad_norm": 0.3580125510889005,
"learning_rate": 2.8873163918056808e-05,
"loss": 1.3464,
"step": 2027
},
{
"epoch": 0.3603891776622684,
"grad_norm": 0.3680715507490336,
"learning_rate": 2.886305305299196e-05,
"loss": 1.3674,
"step": 2028
},
{
"epoch": 0.36056688435736817,
"grad_norm": 0.36927625380941664,
"learning_rate": 2.8852939368423265e-05,
"loss": 1.3692,
"step": 2029
},
{
"epoch": 0.3607445910524679,
"grad_norm": 0.3812559716218701,
"learning_rate": 2.8842822867568073e-05,
"loss": 1.3818,
"step": 2030
},
{
"epoch": 0.3609222977475676,
"grad_norm": 0.3742526726867375,
"learning_rate": 2.883270355364462e-05,
"loss": 1.3663,
"step": 2031
},
{
"epoch": 0.3611000044426674,
"grad_norm": 0.367329865958006,
"learning_rate": 2.8822581429872066e-05,
"loss": 1.3642,
"step": 2032
},
{
"epoch": 0.36127771113776713,
"grad_norm": 0.3827810563556126,
"learning_rate": 2.881245649947045e-05,
"loss": 1.3638,
"step": 2033
},
{
"epoch": 0.36145541783286683,
"grad_norm": 0.37552937422063326,
"learning_rate": 2.8802328765660684e-05,
"loss": 1.3825,
"step": 2034
},
{
"epoch": 0.3616331245279666,
"grad_norm": 0.38342713216141494,
"learning_rate": 2.8792198231664605e-05,
"loss": 1.4095,
"step": 2035
},
{
"epoch": 0.36181083122306634,
"grad_norm": 0.36787329704293664,
"learning_rate": 2.8782064900704924e-05,
"loss": 1.3833,
"step": 2036
},
{
"epoch": 0.36198853791816604,
"grad_norm": 0.3779302658172981,
"learning_rate": 2.8771928776005248e-05,
"loss": 1.4015,
"step": 2037
},
{
"epoch": 0.3621662446132658,
"grad_norm": 0.3669498632513325,
"learning_rate": 2.8761789860790066e-05,
"loss": 1.3261,
"step": 2038
},
{
"epoch": 0.36234395130836555,
"grad_norm": 0.39890854905924855,
"learning_rate": 2.875164815828475e-05,
"loss": 1.3957,
"step": 2039
},
{
"epoch": 0.3625216580034653,
"grad_norm": 0.37498597579474835,
"learning_rate": 2.8741503671715576e-05,
"loss": 1.3897,
"step": 2040
},
{
"epoch": 0.362699364698565,
"grad_norm": 0.3719172947862393,
"learning_rate": 2.8731356404309694e-05,
"loss": 1.352,
"step": 2041
},
{
"epoch": 0.36287707139366476,
"grad_norm": 0.38606186525255665,
"learning_rate": 2.8721206359295135e-05,
"loss": 1.3088,
"step": 2042
},
{
"epoch": 0.3630547780887645,
"grad_norm": 0.3669411686310488,
"learning_rate": 2.871105353990083e-05,
"loss": 1.3525,
"step": 2043
},
{
"epoch": 0.3632324847838642,
"grad_norm": 0.3899749312976543,
"learning_rate": 2.870089794935657e-05,
"loss": 1.3345,
"step": 2044
},
{
"epoch": 0.363410191478964,
"grad_norm": 0.3794010443770452,
"learning_rate": 2.869073959089305e-05,
"loss": 1.3423,
"step": 2045
},
{
"epoch": 0.36358789817406373,
"grad_norm": 0.38784599558430893,
"learning_rate": 2.8680578467741823e-05,
"loss": 1.3707,
"step": 2046
},
{
"epoch": 0.36376560486916343,
"grad_norm": 0.38445552559097823,
"learning_rate": 2.867041458313534e-05,
"loss": 1.3375,
"step": 2047
},
{
"epoch": 0.3639433115642632,
"grad_norm": 0.5466198478554991,
"learning_rate": 2.8660247940306924e-05,
"loss": 1.4015,
"step": 2048
},
{
"epoch": 0.36412101825936294,
"grad_norm": 0.38232037613731357,
"learning_rate": 2.865007854249078e-05,
"loss": 1.3855,
"step": 2049
},
{
"epoch": 0.36429872495446264,
"grad_norm": 0.3919740951998763,
"learning_rate": 2.863990639292198e-05,
"loss": 1.3376,
"step": 2050
},
{
"epoch": 0.3644764316495624,
"grad_norm": 0.3833790783921882,
"learning_rate": 2.862973149483647e-05,
"loss": 1.3714,
"step": 2051
},
{
"epoch": 0.36465413834466215,
"grad_norm": 0.3855022831483702,
"learning_rate": 2.8619553851471082e-05,
"loss": 1.3569,
"step": 2052
},
{
"epoch": 0.36483184503976185,
"grad_norm": 0.3874899753273773,
"learning_rate": 2.860937346606352e-05,
"loss": 1.4209,
"step": 2053
},
{
"epoch": 0.3650095517348616,
"grad_norm": 0.36852020629023935,
"learning_rate": 2.8599190341852348e-05,
"loss": 1.3137,
"step": 2054
},
{
"epoch": 0.36518725842996136,
"grad_norm": 0.3938847141677878,
"learning_rate": 2.8589004482077016e-05,
"loss": 1.3759,
"step": 2055
},
{
"epoch": 0.3653649651250611,
"grad_norm": 0.3930925887041744,
"learning_rate": 2.8578815889977835e-05,
"loss": 1.3783,
"step": 2056
},
{
"epoch": 0.3655426718201608,
"grad_norm": 0.3711568117892979,
"learning_rate": 2.856862456879599e-05,
"loss": 1.3452,
"step": 2057
},
{
"epoch": 0.36572037851526057,
"grad_norm": 0.3748795637081748,
"learning_rate": 2.8558430521773525e-05,
"loss": 1.4045,
"step": 2058
},
{
"epoch": 0.3658980852103603,
"grad_norm": 0.5502497826727013,
"learning_rate": 2.854823375215336e-05,
"loss": 1.3678,
"step": 2059
},
{
"epoch": 0.36607579190546,
"grad_norm": 0.37290642062329465,
"learning_rate": 2.853803426317929e-05,
"loss": 1.3442,
"step": 2060
},
{
"epoch": 0.3662534986005598,
"grad_norm": 0.3932488144259585,
"learning_rate": 2.8527832058095946e-05,
"loss": 1.3578,
"step": 2061
},
{
"epoch": 0.36643120529565953,
"grad_norm": 0.3816904829602918,
"learning_rate": 2.8517627140148855e-05,
"loss": 1.4011,
"step": 2062
},
{
"epoch": 0.36660891199075923,
"grad_norm": 0.3831389681932397,
"learning_rate": 2.8507419512584396e-05,
"loss": 1.3462,
"step": 2063
},
{
"epoch": 0.366786618685859,
"grad_norm": 0.38654684755899327,
"learning_rate": 2.8497209178649793e-05,
"loss": 1.3793,
"step": 2064
},
{
"epoch": 0.36696432538095874,
"grad_norm": 0.38258351987208283,
"learning_rate": 2.848699614159316e-05,
"loss": 1.3531,
"step": 2065
},
{
"epoch": 0.36714203207605844,
"grad_norm": 0.3808042585224028,
"learning_rate": 2.847678040466344e-05,
"loss": 1.3764,
"step": 2066
},
{
"epoch": 0.3673197387711582,
"grad_norm": 0.3791640165606861,
"learning_rate": 2.8466561971110466e-05,
"loss": 1.3326,
"step": 2067
},
{
"epoch": 0.36749744546625795,
"grad_norm": 0.39194450685469906,
"learning_rate": 2.8456340844184907e-05,
"loss": 1.3865,
"step": 2068
},
{
"epoch": 0.36767515216135765,
"grad_norm": 0.3929205452910296,
"learning_rate": 2.84461170271383e-05,
"loss": 1.3735,
"step": 2069
},
{
"epoch": 0.3678528588564574,
"grad_norm": 0.4544662135840749,
"learning_rate": 2.8435890523223025e-05,
"loss": 1.3802,
"step": 2070
},
{
"epoch": 0.36803056555155717,
"grad_norm": 0.38546107926871354,
"learning_rate": 2.8425661335692338e-05,
"loss": 1.3678,
"step": 2071
},
{
"epoch": 0.3682082722466569,
"grad_norm": 0.3845536892076783,
"learning_rate": 2.8415429467800323e-05,
"loss": 1.3879,
"step": 2072
},
{
"epoch": 0.3683859789417566,
"grad_norm": 0.3927559840082437,
"learning_rate": 2.8405194922801932e-05,
"loss": 1.3699,
"step": 2073
},
{
"epoch": 0.3685636856368564,
"grad_norm": 0.38542537329624194,
"learning_rate": 2.8394957703952976e-05,
"loss": 1.3555,
"step": 2074
},
{
"epoch": 0.36874139233195613,
"grad_norm": 0.37487125313974307,
"learning_rate": 2.8384717814510097e-05,
"loss": 1.3893,
"step": 2075
},
{
"epoch": 0.36891909902705583,
"grad_norm": 0.38311285287605673,
"learning_rate": 2.8374475257730797e-05,
"loss": 1.3309,
"step": 2076
},
{
"epoch": 0.3690968057221556,
"grad_norm": 0.36868604644207165,
"learning_rate": 2.836423003687343e-05,
"loss": 1.3821,
"step": 2077
},
{
"epoch": 0.36927451241725534,
"grad_norm": 0.39039358376794725,
"learning_rate": 2.8353982155197192e-05,
"loss": 1.3489,
"step": 2078
},
{
"epoch": 0.36945221911235504,
"grad_norm": 0.39247553911691213,
"learning_rate": 2.8343731615962135e-05,
"loss": 1.373,
"step": 2079
},
{
"epoch": 0.3696299258074548,
"grad_norm": 0.38837524518800964,
"learning_rate": 2.833347842242913e-05,
"loss": 1.3115,
"step": 2080
},
{
"epoch": 0.36980763250255455,
"grad_norm": 0.3826286359238273,
"learning_rate": 2.8323222577859917e-05,
"loss": 1.3603,
"step": 2081
},
{
"epoch": 0.36998533919765425,
"grad_norm": 0.37901948418877646,
"learning_rate": 2.8312964085517086e-05,
"loss": 1.3201,
"step": 2082
},
{
"epoch": 0.370163045892754,
"grad_norm": 0.3850322629101244,
"learning_rate": 2.8302702948664044e-05,
"loss": 1.3853,
"step": 2083
},
{
"epoch": 0.37034075258785376,
"grad_norm": 0.38097775693701824,
"learning_rate": 2.8292439170565056e-05,
"loss": 1.3559,
"step": 2084
},
{
"epoch": 0.37051845928295346,
"grad_norm": 0.38744862059479873,
"learning_rate": 2.8282172754485214e-05,
"loss": 1.3561,
"step": 2085
},
{
"epoch": 0.3706961659780532,
"grad_norm": 0.3768085023138019,
"learning_rate": 2.8271903703690474e-05,
"loss": 1.3449,
"step": 2086
},
{
"epoch": 0.37087387267315297,
"grad_norm": 0.39315306800923067,
"learning_rate": 2.8261632021447608e-05,
"loss": 1.3624,
"step": 2087
},
{
"epoch": 0.3710515793682527,
"grad_norm": 0.37548743908320453,
"learning_rate": 2.8251357711024224e-05,
"loss": 1.3654,
"step": 2088
},
{
"epoch": 0.3712292860633524,
"grad_norm": 0.40984879200269053,
"learning_rate": 2.8241080775688777e-05,
"loss": 1.3675,
"step": 2089
},
{
"epoch": 0.3714069927584522,
"grad_norm": 0.39191387423006097,
"learning_rate": 2.823080121871056e-05,
"loss": 1.4043,
"step": 2090
},
{
"epoch": 0.37158469945355194,
"grad_norm": 0.3919655435883125,
"learning_rate": 2.822051904335968e-05,
"loss": 1.3365,
"step": 2091
},
{
"epoch": 0.37176240614865164,
"grad_norm": 0.40582687473012025,
"learning_rate": 2.8210234252907107e-05,
"loss": 1.3837,
"step": 2092
},
{
"epoch": 0.3719401128437514,
"grad_norm": 0.39641148135149445,
"learning_rate": 2.8199946850624614e-05,
"loss": 1.4086,
"step": 2093
},
{
"epoch": 0.37211781953885115,
"grad_norm": 0.39158857609962955,
"learning_rate": 2.818965683978482e-05,
"loss": 1.4071,
"step": 2094
},
{
"epoch": 0.37229552623395085,
"grad_norm": 0.3772136442166356,
"learning_rate": 2.8179364223661176e-05,
"loss": 1.3553,
"step": 2095
},
{
"epoch": 0.3724732329290506,
"grad_norm": 0.3911934644076744,
"learning_rate": 2.8169069005527947e-05,
"loss": 1.396,
"step": 2096
},
{
"epoch": 0.37265093962415036,
"grad_norm": 0.3786146356176828,
"learning_rate": 2.8158771188660244e-05,
"loss": 1.383,
"step": 2097
},
{
"epoch": 0.37282864631925006,
"grad_norm": 0.3917714068276506,
"learning_rate": 2.8148470776333988e-05,
"loss": 1.4306,
"step": 2098
},
{
"epoch": 0.3730063530143498,
"grad_norm": 0.3826229071713996,
"learning_rate": 2.813816777182595e-05,
"loss": 1.3832,
"step": 2099
},
{
"epoch": 0.37318405970944957,
"grad_norm": 0.4134575225994191,
"learning_rate": 2.8127862178413692e-05,
"loss": 1.3783,
"step": 2100
},
{
"epoch": 0.37336176640454927,
"grad_norm": 0.3878088293432596,
"learning_rate": 2.8117553999375626e-05,
"loss": 1.4201,
"step": 2101
},
{
"epoch": 0.373539473099649,
"grad_norm": 0.38431931106807216,
"learning_rate": 2.8107243237990974e-05,
"loss": 1.3546,
"step": 2102
},
{
"epoch": 0.3737171797947488,
"grad_norm": 0.3738982476201066,
"learning_rate": 2.8096929897539782e-05,
"loss": 1.3723,
"step": 2103
},
{
"epoch": 0.37389488648984853,
"grad_norm": 0.36994547589519333,
"learning_rate": 2.8086613981302925e-05,
"loss": 1.3694,
"step": 2104
},
{
"epoch": 0.37407259318494823,
"grad_norm": 0.39674224510999045,
"learning_rate": 2.8076295492562077e-05,
"loss": 1.361,
"step": 2105
},
{
"epoch": 0.374250299880048,
"grad_norm": 0.4102815026265374,
"learning_rate": 2.806597443459976e-05,
"loss": 1.3816,
"step": 2106
},
{
"epoch": 0.37442800657514774,
"grad_norm": 0.38227597468206215,
"learning_rate": 2.8055650810699286e-05,
"loss": 1.3376,
"step": 2107
},
{
"epoch": 0.37460571327024744,
"grad_norm": 0.37850444060685934,
"learning_rate": 2.804532462414479e-05,
"loss": 1.387,
"step": 2108
},
{
"epoch": 0.3747834199653472,
"grad_norm": 0.38451005594438303,
"learning_rate": 2.803499587822124e-05,
"loss": 1.3401,
"step": 2109
},
{
"epoch": 0.37496112666044695,
"grad_norm": 0.38499984993292624,
"learning_rate": 2.8024664576214387e-05,
"loss": 1.4156,
"step": 2110
},
{
"epoch": 0.37513883335554665,
"grad_norm": 0.36495403082361727,
"learning_rate": 2.801433072141083e-05,
"loss": 1.3434,
"step": 2111
},
{
"epoch": 0.3753165400506464,
"grad_norm": 0.38120210101103197,
"learning_rate": 2.800399431709795e-05,
"loss": 1.356,
"step": 2112
},
{
"epoch": 0.37549424674574616,
"grad_norm": 0.367466282318448,
"learning_rate": 2.799365536656396e-05,
"loss": 1.3656,
"step": 2113
},
{
"epoch": 0.37567195344084586,
"grad_norm": 0.37642793409710207,
"learning_rate": 2.7983313873097868e-05,
"loss": 1.3932,
"step": 2114
},
{
"epoch": 0.3758496601359456,
"grad_norm": 0.3853526175367025,
"learning_rate": 2.7972969839989495e-05,
"loss": 1.3992,
"step": 2115
},
{
"epoch": 0.3760273668310454,
"grad_norm": 0.3652098512506825,
"learning_rate": 2.796262327052949e-05,
"loss": 1.3428,
"step": 2116
},
{
"epoch": 0.3762050735261451,
"grad_norm": 0.3687840191593852,
"learning_rate": 2.7952274168009265e-05,
"loss": 1.358,
"step": 2117
},
{
"epoch": 0.37638278022124483,
"grad_norm": 0.3792236055814281,
"learning_rate": 2.7941922535721083e-05,
"loss": 1.3585,
"step": 2118
},
{
"epoch": 0.3765604869163446,
"grad_norm": 0.38475883667540234,
"learning_rate": 2.793156837695799e-05,
"loss": 1.39,
"step": 2119
},
{
"epoch": 0.37673819361144434,
"grad_norm": 0.37132752897505217,
"learning_rate": 2.7921211695013836e-05,
"loss": 1.3289,
"step": 2120
},
{
"epoch": 0.37691590030654404,
"grad_norm": 0.3693443745785962,
"learning_rate": 2.791085249318328e-05,
"loss": 1.357,
"step": 2121
},
{
"epoch": 0.3770936070016438,
"grad_norm": 0.3817370385206183,
"learning_rate": 2.7900490774761766e-05,
"loss": 1.3774,
"step": 2122
},
{
"epoch": 0.37727131369674355,
"grad_norm": 0.3847374719754216,
"learning_rate": 2.7890126543045566e-05,
"loss": 1.3909,
"step": 2123
},
{
"epoch": 0.37744902039184325,
"grad_norm": 0.37846996732480104,
"learning_rate": 2.7879759801331733e-05,
"loss": 1.3517,
"step": 2124
},
{
"epoch": 0.377626727086943,
"grad_norm": 0.3860969671107221,
"learning_rate": 2.7869390552918124e-05,
"loss": 1.389,
"step": 2125
},
{
"epoch": 0.37780443378204276,
"grad_norm": 0.3701093687869099,
"learning_rate": 2.785901880110338e-05,
"loss": 1.3201,
"step": 2126
},
{
"epoch": 0.37798214047714246,
"grad_norm": 0.37287429988285337,
"learning_rate": 2.7848644549186974e-05,
"loss": 1.335,
"step": 2127
},
{
"epoch": 0.3781598471722422,
"grad_norm": 0.40818306621556444,
"learning_rate": 2.783826780046913e-05,
"loss": 1.3744,
"step": 2128
},
{
"epoch": 0.37833755386734197,
"grad_norm": 0.36984971271511946,
"learning_rate": 2.782788855825089e-05,
"loss": 1.3407,
"step": 2129
},
{
"epoch": 0.37851526056244167,
"grad_norm": 0.3806308835124078,
"learning_rate": 2.7817506825834093e-05,
"loss": 1.3961,
"step": 2130
},
{
"epoch": 0.3786929672575414,
"grad_norm": 0.3741530163556233,
"learning_rate": 2.780712260652136e-05,
"loss": 1.3789,
"step": 2131
},
{
"epoch": 0.3788706739526412,
"grad_norm": 0.372668547949379,
"learning_rate": 2.7796735903616107e-05,
"loss": 1.3963,
"step": 2132
},
{
"epoch": 0.3790483806477409,
"grad_norm": 0.3979414169386232,
"learning_rate": 2.7786346720422536e-05,
"loss": 1.3872,
"step": 2133
},
{
"epoch": 0.37922608734284063,
"grad_norm": 0.365369011654538,
"learning_rate": 2.7775955060245645e-05,
"loss": 1.3727,
"step": 2134
},
{
"epoch": 0.3794037940379404,
"grad_norm": 0.3878602028797145,
"learning_rate": 2.776556092639122e-05,
"loss": 1.3938,
"step": 2135
},
{
"epoch": 0.37958150073304014,
"grad_norm": 0.3764632803844917,
"learning_rate": 2.775516432216582e-05,
"loss": 1.3549,
"step": 2136
},
{
"epoch": 0.37975920742813984,
"grad_norm": 0.4800489372765138,
"learning_rate": 2.774476525087681e-05,
"loss": 1.403,
"step": 2137
},
{
"epoch": 0.3799369141232396,
"grad_norm": 0.3840008330003893,
"learning_rate": 2.773436371583233e-05,
"loss": 1.3898,
"step": 2138
},
{
"epoch": 0.38011462081833935,
"grad_norm": 0.37055255539553345,
"learning_rate": 2.77239597203413e-05,
"loss": 1.417,
"step": 2139
},
{
"epoch": 0.38029232751343905,
"grad_norm": 0.3862633055022592,
"learning_rate": 2.7713553267713416e-05,
"loss": 1.3895,
"step": 2140
},
{
"epoch": 0.3804700342085388,
"grad_norm": 0.3717795090544308,
"learning_rate": 2.7703144361259186e-05,
"loss": 1.3688,
"step": 2141
},
{
"epoch": 0.38064774090363857,
"grad_norm": 0.383532129090514,
"learning_rate": 2.7692733004289873e-05,
"loss": 1.3714,
"step": 2142
},
{
"epoch": 0.38082544759873826,
"grad_norm": 0.3786001147192247,
"learning_rate": 2.7682319200117524e-05,
"loss": 1.4002,
"step": 2143
},
{
"epoch": 0.381003154293838,
"grad_norm": 0.38526844543360167,
"learning_rate": 2.767190295205496e-05,
"loss": 1.3497,
"step": 2144
},
{
"epoch": 0.3811808609889378,
"grad_norm": 0.36265953743190005,
"learning_rate": 2.766148426341579e-05,
"loss": 1.3516,
"step": 2145
},
{
"epoch": 0.3813585676840375,
"grad_norm": 0.40339135715887287,
"learning_rate": 2.7651063137514405e-05,
"loss": 1.3993,
"step": 2146
},
{
"epoch": 0.38153627437913723,
"grad_norm": 0.36499021850069835,
"learning_rate": 2.764063957766594e-05,
"loss": 1.3683,
"step": 2147
},
{
"epoch": 0.381713981074237,
"grad_norm": 0.3703688735124043,
"learning_rate": 2.763021358718634e-05,
"loss": 1.3928,
"step": 2148
},
{
"epoch": 0.3818916877693367,
"grad_norm": 0.37723125234606564,
"learning_rate": 2.7619785169392303e-05,
"loss": 1.3559,
"step": 2149
},
{
"epoch": 0.38206939446443644,
"grad_norm": 0.36324639264459596,
"learning_rate": 2.7609354327601313e-05,
"loss": 1.3561,
"step": 2150
},
{
"epoch": 0.3822471011595362,
"grad_norm": 0.37050264099248426,
"learning_rate": 2.759892106513161e-05,
"loss": 1.3524,
"step": 2151
},
{
"epoch": 0.38242480785463595,
"grad_norm": 0.3827956353434813,
"learning_rate": 2.7588485385302207e-05,
"loss": 1.3582,
"step": 2152
},
{
"epoch": 0.38260251454973565,
"grad_norm": 0.3678763455040647,
"learning_rate": 2.7578047291432898e-05,
"loss": 1.3493,
"step": 2153
},
{
"epoch": 0.3827802212448354,
"grad_norm": 0.3809404233444153,
"learning_rate": 2.7567606786844233e-05,
"loss": 1.3801,
"step": 2154
},
{
"epoch": 0.38295792793993516,
"grad_norm": 0.3691894269375115,
"learning_rate": 2.7557163874857536e-05,
"loss": 1.3431,
"step": 2155
},
{
"epoch": 0.38313563463503486,
"grad_norm": 0.35855099993474865,
"learning_rate": 2.7546718558794894e-05,
"loss": 1.3429,
"step": 2156
},
{
"epoch": 0.3833133413301346,
"grad_norm": 0.376539636653492,
"learning_rate": 2.7536270841979153e-05,
"loss": 1.4068,
"step": 2157
},
{
"epoch": 0.38349104802523437,
"grad_norm": 0.3729960486441172,
"learning_rate": 2.752582072773393e-05,
"loss": 1.3737,
"step": 2158
},
{
"epoch": 0.38366875472033407,
"grad_norm": 0.37335478540456274,
"learning_rate": 2.75153682193836e-05,
"loss": 1.3663,
"step": 2159
},
{
"epoch": 0.3838464614154338,
"grad_norm": 0.3583265867146266,
"learning_rate": 2.7504913320253312e-05,
"loss": 1.3565,
"step": 2160
},
{
"epoch": 0.3840241681105336,
"grad_norm": 0.37623669090396245,
"learning_rate": 2.749445603366896e-05,
"loss": 1.3576,
"step": 2161
},
{
"epoch": 0.3842018748056333,
"grad_norm": 0.3714993560258395,
"learning_rate": 2.7483996362957205e-05,
"loss": 1.3898,
"step": 2162
},
{
"epoch": 0.38437958150073304,
"grad_norm": 0.3677395205785789,
"learning_rate": 2.7473534311445463e-05,
"loss": 1.3661,
"step": 2163
},
{
"epoch": 0.3845572881958328,
"grad_norm": 0.3836427679343959,
"learning_rate": 2.746306988246191e-05,
"loss": 1.3618,
"step": 2164
},
{
"epoch": 0.3847349948909325,
"grad_norm": 0.3673439931273699,
"learning_rate": 2.745260307933548e-05,
"loss": 1.3587,
"step": 2165
},
{
"epoch": 0.38491270158603225,
"grad_norm": 0.3795729301405178,
"learning_rate": 2.7442133905395855e-05,
"loss": 1.3667,
"step": 2166
},
{
"epoch": 0.385090408281132,
"grad_norm": 0.3714090719765484,
"learning_rate": 2.7431662363973483e-05,
"loss": 1.384,
"step": 2167
},
{
"epoch": 0.38526811497623176,
"grad_norm": 0.3786886257590711,
"learning_rate": 2.7421188458399552e-05,
"loss": 1.347,
"step": 2168
},
{
"epoch": 0.38544582167133146,
"grad_norm": 0.4398314007332105,
"learning_rate": 2.741071219200601e-05,
"loss": 1.3534,
"step": 2169
},
{
"epoch": 0.3856235283664312,
"grad_norm": 0.3860393738500145,
"learning_rate": 2.7400233568125556e-05,
"loss": 1.378,
"step": 2170
},
{
"epoch": 0.38580123506153097,
"grad_norm": 0.37804928421098777,
"learning_rate": 2.7389752590091637e-05,
"loss": 1.3393,
"step": 2171
},
{
"epoch": 0.38597894175663067,
"grad_norm": 0.37630821570814543,
"learning_rate": 2.7379269261238445e-05,
"loss": 1.4025,
"step": 2172
},
{
"epoch": 0.3861566484517304,
"grad_norm": 0.7800582199137145,
"learning_rate": 2.7368783584900927e-05,
"loss": 1.4008,
"step": 2173
},
{
"epoch": 0.3863343551468302,
"grad_norm": 0.36366135941047745,
"learning_rate": 2.7358295564414782e-05,
"loss": 1.3613,
"step": 2174
},
{
"epoch": 0.3865120618419299,
"grad_norm": 0.3968335555548615,
"learning_rate": 2.734780520311643e-05,
"loss": 1.3789,
"step": 2175
},
{
"epoch": 0.38668976853702963,
"grad_norm": 0.37158786173068864,
"learning_rate": 2.733731250434307e-05,
"loss": 1.3614,
"step": 2176
},
{
"epoch": 0.3868674752321294,
"grad_norm": 0.3797194551343313,
"learning_rate": 2.7326817471432616e-05,
"loss": 1.3453,
"step": 2177
},
{
"epoch": 0.3870451819272291,
"grad_norm": 0.36906908806732547,
"learning_rate": 2.7316320107723732e-05,
"loss": 1.3588,
"step": 2178
},
{
"epoch": 0.38722288862232884,
"grad_norm": 0.38074304938224646,
"learning_rate": 2.7305820416555838e-05,
"loss": 1.387,
"step": 2179
},
{
"epoch": 0.3874005953174286,
"grad_norm": 0.3816018368878402,
"learning_rate": 2.7295318401269074e-05,
"loss": 1.3828,
"step": 2180
},
{
"epoch": 0.3875783020125283,
"grad_norm": 0.38851697097867094,
"learning_rate": 2.728481406520433e-05,
"loss": 1.4035,
"step": 2181
},
{
"epoch": 0.38775600870762805,
"grad_norm": 0.3874736122300011,
"learning_rate": 2.7274307411703237e-05,
"loss": 1.4391,
"step": 2182
},
{
"epoch": 0.3879337154027278,
"grad_norm": 0.3820879709926199,
"learning_rate": 2.726379844410816e-05,
"loss": 1.4032,
"step": 2183
},
{
"epoch": 0.38811142209782756,
"grad_norm": 0.3700821195528636,
"learning_rate": 2.7253287165762196e-05,
"loss": 1.3525,
"step": 2184
},
{
"epoch": 0.38828912879292726,
"grad_norm": 0.37592397826141766,
"learning_rate": 2.7242773580009174e-05,
"loss": 1.3384,
"step": 2185
},
{
"epoch": 0.388466835488027,
"grad_norm": 0.37863600814037157,
"learning_rate": 2.7232257690193673e-05,
"loss": 1.3926,
"step": 2186
},
{
"epoch": 0.3886445421831268,
"grad_norm": 0.38313579484972543,
"learning_rate": 2.7221739499660996e-05,
"loss": 1.371,
"step": 2187
},
{
"epoch": 0.3888222488782265,
"grad_norm": 0.3805435332526512,
"learning_rate": 2.7211219011757166e-05,
"loss": 1.3944,
"step": 2188
},
{
"epoch": 0.38899995557332623,
"grad_norm": 0.3828223497241651,
"learning_rate": 2.7200696229828957e-05,
"loss": 1.3444,
"step": 2189
},
{
"epoch": 0.389177662268426,
"grad_norm": 0.3700508917864514,
"learning_rate": 2.7190171157223867e-05,
"loss": 1.3446,
"step": 2190
},
{
"epoch": 0.3893553689635257,
"grad_norm": 0.3998394047129767,
"learning_rate": 2.7179643797290108e-05,
"loss": 1.3789,
"step": 2191
},
{
"epoch": 0.38953307565862544,
"grad_norm": 0.3745299908901653,
"learning_rate": 2.7169114153376646e-05,
"loss": 1.3249,
"step": 2192
},
{
"epoch": 0.3897107823537252,
"grad_norm": 0.38215659614142616,
"learning_rate": 2.7158582228833146e-05,
"loss": 1.3229,
"step": 2193
},
{
"epoch": 0.3898884890488249,
"grad_norm": 0.3705002991656851,
"learning_rate": 2.714804802701001e-05,
"loss": 1.3424,
"step": 2194
},
{
"epoch": 0.39006619574392465,
"grad_norm": 0.37710872512299437,
"learning_rate": 2.7137511551258386e-05,
"loss": 1.3927,
"step": 2195
},
{
"epoch": 0.3902439024390244,
"grad_norm": 0.3799472244887311,
"learning_rate": 2.7126972804930097e-05,
"loss": 1.342,
"step": 2196
},
{
"epoch": 0.3904216091341241,
"grad_norm": 0.38048939391655195,
"learning_rate": 2.7116431791377738e-05,
"loss": 1.3825,
"step": 2197
},
{
"epoch": 0.39059931582922386,
"grad_norm": 0.39204920142181177,
"learning_rate": 2.7105888513954593e-05,
"loss": 1.3924,
"step": 2198
},
{
"epoch": 0.3907770225243236,
"grad_norm": 0.3769480345378491,
"learning_rate": 2.709534297601468e-05,
"loss": 1.387,
"step": 2199
},
{
"epoch": 0.39095472921942337,
"grad_norm": 0.38770629389115324,
"learning_rate": 2.7084795180912727e-05,
"loss": 1.386,
"step": 2200
},
{
"epoch": 0.39113243591452307,
"grad_norm": 0.3786570159735101,
"learning_rate": 2.70742451320042e-05,
"loss": 1.3825,
"step": 2201
},
{
"epoch": 0.3913101426096228,
"grad_norm": 0.37293528997933595,
"learning_rate": 2.7063692832645254e-05,
"loss": 1.3949,
"step": 2202
},
{
"epoch": 0.3914878493047226,
"grad_norm": 0.3796336842472203,
"learning_rate": 2.7053138286192783e-05,
"loss": 1.3289,
"step": 2203
},
{
"epoch": 0.3916655559998223,
"grad_norm": 0.3909015339746708,
"learning_rate": 2.704258149600438e-05,
"loss": 1.3629,
"step": 2204
},
{
"epoch": 0.39184326269492203,
"grad_norm": 0.3817333771324511,
"learning_rate": 2.7032022465438362e-05,
"loss": 1.3794,
"step": 2205
},
{
"epoch": 0.3920209693900218,
"grad_norm": 0.3789955260995565,
"learning_rate": 2.7021461197853756e-05,
"loss": 1.3664,
"step": 2206
},
{
"epoch": 0.3921986760851215,
"grad_norm": 0.38544064428960867,
"learning_rate": 2.70108976966103e-05,
"loss": 1.393,
"step": 2207
},
{
"epoch": 0.39237638278022124,
"grad_norm": 0.38442331663479956,
"learning_rate": 2.700033196506843e-05,
"loss": 1.3566,
"step": 2208
},
{
"epoch": 0.392554089475321,
"grad_norm": 0.3736648780597236,
"learning_rate": 2.6989764006589325e-05,
"loss": 1.3616,
"step": 2209
},
{
"epoch": 0.3927317961704207,
"grad_norm": 0.36854101001463496,
"learning_rate": 2.6979193824534842e-05,
"loss": 1.3929,
"step": 2210
},
{
"epoch": 0.39290950286552045,
"grad_norm": 0.36291581597881556,
"learning_rate": 2.696862142226755e-05,
"loss": 1.3443,
"step": 2211
},
{
"epoch": 0.3930872095606202,
"grad_norm": 0.3867018963457568,
"learning_rate": 2.6958046803150733e-05,
"loss": 1.4153,
"step": 2212
},
{
"epoch": 0.3932649162557199,
"grad_norm": 0.3746429445925624,
"learning_rate": 2.694746997054837e-05,
"loss": 1.4251,
"step": 2213
},
{
"epoch": 0.39344262295081966,
"grad_norm": 0.38054859950973724,
"learning_rate": 2.693689092782517e-05,
"loss": 1.4121,
"step": 2214
},
{
"epoch": 0.3936203296459194,
"grad_norm": 0.3751670054719357,
"learning_rate": 2.69263096783465e-05,
"loss": 1.3861,
"step": 2215
},
{
"epoch": 0.3937980363410192,
"grad_norm": 0.36661633681258476,
"learning_rate": 2.691572622547847e-05,
"loss": 1.3723,
"step": 2216
},
{
"epoch": 0.3939757430361189,
"grad_norm": 0.3680499021848459,
"learning_rate": 2.6905140572587876e-05,
"loss": 1.3533,
"step": 2217
},
{
"epoch": 0.39415344973121863,
"grad_norm": 0.37276198879725636,
"learning_rate": 2.6894552723042205e-05,
"loss": 1.3708,
"step": 2218
},
{
"epoch": 0.3943311564263184,
"grad_norm": 0.36392983860194444,
"learning_rate": 2.6883962680209657e-05,
"loss": 1.3698,
"step": 2219
},
{
"epoch": 0.3945088631214181,
"grad_norm": 0.37148741457688733,
"learning_rate": 2.6873370447459114e-05,
"loss": 1.3438,
"step": 2220
},
{
"epoch": 0.39468656981651784,
"grad_norm": 0.37658159554045656,
"learning_rate": 2.6862776028160184e-05,
"loss": 1.379,
"step": 2221
},
{
"epoch": 0.3948642765116176,
"grad_norm": 0.36918130766323926,
"learning_rate": 2.6852179425683126e-05,
"loss": 1.331,
"step": 2222
},
{
"epoch": 0.3950419832067173,
"grad_norm": 0.3813060841403589,
"learning_rate": 2.684158064339894e-05,
"loss": 1.3329,
"step": 2223
},
{
"epoch": 0.39521968990181705,
"grad_norm": 0.3794639187450798,
"learning_rate": 2.6830979684679293e-05,
"loss": 1.3637,
"step": 2224
},
{
"epoch": 0.3953973965969168,
"grad_norm": 0.3638426474434246,
"learning_rate": 2.682037655289654e-05,
"loss": 1.3328,
"step": 2225
},
{
"epoch": 0.3955751032920165,
"grad_norm": 0.37816316424284646,
"learning_rate": 2.6809771251423746e-05,
"loss": 1.3944,
"step": 2226
},
{
"epoch": 0.39575280998711626,
"grad_norm": 0.37918587612906796,
"learning_rate": 2.6799163783634647e-05,
"loss": 1.3804,
"step": 2227
},
{
"epoch": 0.395930516682216,
"grad_norm": 0.3801526192472718,
"learning_rate": 2.678855415290369e-05,
"loss": 1.3622,
"step": 2228
},
{
"epoch": 0.3961082233773157,
"grad_norm": 0.37094280910114624,
"learning_rate": 2.677794236260599e-05,
"loss": 1.3557,
"step": 2229
},
{
"epoch": 0.39628593007241547,
"grad_norm": 0.46923870200435464,
"learning_rate": 2.676732841611736e-05,
"loss": 1.3149,
"step": 2230
},
{
"epoch": 0.3964636367675152,
"grad_norm": 0.3754507532286325,
"learning_rate": 2.6756712316814297e-05,
"loss": 1.3967,
"step": 2231
},
{
"epoch": 0.396641343462615,
"grad_norm": 0.3827458436658888,
"learning_rate": 2.6746094068073976e-05,
"loss": 1.3712,
"step": 2232
},
{
"epoch": 0.3968190501577147,
"grad_norm": 0.36858605464363037,
"learning_rate": 2.6735473673274273e-05,
"loss": 1.3512,
"step": 2233
},
{
"epoch": 0.39699675685281444,
"grad_norm": 0.40764798123038853,
"learning_rate": 2.6724851135793725e-05,
"loss": 1.3756,
"step": 2234
},
{
"epoch": 0.3971744635479142,
"grad_norm": 0.375877895416509,
"learning_rate": 2.6714226459011562e-05,
"loss": 1.3813,
"step": 2235
},
{
"epoch": 0.3973521702430139,
"grad_norm": 0.3701898935275692,
"learning_rate": 2.6703599646307698e-05,
"loss": 1.3405,
"step": 2236
},
{
"epoch": 0.39752987693811365,
"grad_norm": 0.3689702170603083,
"learning_rate": 2.669297070106272e-05,
"loss": 1.3361,
"step": 2237
},
{
"epoch": 0.3977075836332134,
"grad_norm": 0.38309159812747956,
"learning_rate": 2.6682339626657895e-05,
"loss": 1.3572,
"step": 2238
},
{
"epoch": 0.3978852903283131,
"grad_norm": 0.36679319871184973,
"learning_rate": 2.6671706426475164e-05,
"loss": 1.3495,
"step": 2239
},
{
"epoch": 0.39806299702341286,
"grad_norm": 0.3783807086598179,
"learning_rate": 2.666107110389716e-05,
"loss": 1.4046,
"step": 2240
},
{
"epoch": 0.3982407037185126,
"grad_norm": 0.37627245208013843,
"learning_rate": 2.665043366230716e-05,
"loss": 1.3553,
"step": 2241
},
{
"epoch": 0.3984184104136123,
"grad_norm": 0.3804367258759181,
"learning_rate": 2.6639794105089154e-05,
"loss": 1.3418,
"step": 2242
},
{
"epoch": 0.39859611710871207,
"grad_norm": 0.3837942238523859,
"learning_rate": 2.662915243562777e-05,
"loss": 1.367,
"step": 2243
},
{
"epoch": 0.3987738238038118,
"grad_norm": 0.3759728525380926,
"learning_rate": 2.6618508657308332e-05,
"loss": 1.3539,
"step": 2244
},
{
"epoch": 0.3989515304989115,
"grad_norm": 0.38170541137113095,
"learning_rate": 2.6607862773516826e-05,
"loss": 1.4033,
"step": 2245
},
{
"epoch": 0.3991292371940113,
"grad_norm": 0.37109731804300555,
"learning_rate": 2.6597214787639897e-05,
"loss": 1.3431,
"step": 2246
},
{
"epoch": 0.39930694388911103,
"grad_norm": 0.6960921903837297,
"learning_rate": 2.6586564703064887e-05,
"loss": 1.3625,
"step": 2247
},
{
"epoch": 0.3994846505842108,
"grad_norm": 0.3766372857217262,
"learning_rate": 2.6575912523179773e-05,
"loss": 1.4163,
"step": 2248
},
{
"epoch": 0.3996623572793105,
"grad_norm": 0.39369145095745955,
"learning_rate": 2.6565258251373225e-05,
"loss": 1.3445,
"step": 2249
},
{
"epoch": 0.39984006397441024,
"grad_norm": 0.41418421465744304,
"learning_rate": 2.6554601891034555e-05,
"loss": 1.3596,
"step": 2250
},
{
"epoch": 0.40001777066951,
"grad_norm": 0.3838837583781622,
"learning_rate": 2.6543943445553773e-05,
"loss": 1.3553,
"step": 2251
},
{
"epoch": 0.4001954773646097,
"grad_norm": 0.3917608265208986,
"learning_rate": 2.6533282918321503e-05,
"loss": 1.4201,
"step": 2252
},
{
"epoch": 0.40037318405970945,
"grad_norm": 0.3686529047421648,
"learning_rate": 2.6522620312729074e-05,
"loss": 1.3553,
"step": 2253
},
{
"epoch": 0.4005508907548092,
"grad_norm": 0.3745070055917828,
"learning_rate": 2.651195563216846e-05,
"loss": 1.3665,
"step": 2254
},
{
"epoch": 0.4007285974499089,
"grad_norm": 0.3653778510228011,
"learning_rate": 2.6501288880032304e-05,
"loss": 1.4024,
"step": 2255
},
{
"epoch": 0.40090630414500866,
"grad_norm": 0.3742831408164309,
"learning_rate": 2.6490620059713886e-05,
"loss": 1.3102,
"step": 2256
},
{
"epoch": 0.4010840108401084,
"grad_norm": 0.37701316599087775,
"learning_rate": 2.6479949174607166e-05,
"loss": 1.3795,
"step": 2257
},
{
"epoch": 0.4012617175352081,
"grad_norm": 0.3771767986404963,
"learning_rate": 2.6469276228106754e-05,
"loss": 1.4101,
"step": 2258
},
{
"epoch": 0.4014394242303079,
"grad_norm": 0.3800665891949158,
"learning_rate": 2.6458601223607923e-05,
"loss": 1.3621,
"step": 2259
},
{
"epoch": 0.40161713092540763,
"grad_norm": 0.3679116596213652,
"learning_rate": 2.6447924164506572e-05,
"loss": 1.3561,
"step": 2260
},
{
"epoch": 0.40179483762050733,
"grad_norm": 0.3878758500551551,
"learning_rate": 2.6437245054199285e-05,
"loss": 1.4075,
"step": 2261
},
{
"epoch": 0.4019725443156071,
"grad_norm": 0.3656877718520377,
"learning_rate": 2.6426563896083295e-05,
"loss": 1.3828,
"step": 2262
},
{
"epoch": 0.40215025101070684,
"grad_norm": 0.3976637029921766,
"learning_rate": 2.6415880693556467e-05,
"loss": 1.3946,
"step": 2263
},
{
"epoch": 0.4023279577058066,
"grad_norm": 0.36851713765657096,
"learning_rate": 2.640519545001733e-05,
"loss": 1.3432,
"step": 2264
},
{
"epoch": 0.4025056644009063,
"grad_norm": 0.49124439575842643,
"learning_rate": 2.6394508168865076e-05,
"loss": 1.3606,
"step": 2265
},
{
"epoch": 0.40268337109600605,
"grad_norm": 0.3914941274208213,
"learning_rate": 2.6383818853499518e-05,
"loss": 1.4301,
"step": 2266
},
{
"epoch": 0.4028610777911058,
"grad_norm": 0.3886976465176677,
"learning_rate": 2.6373127507321124e-05,
"loss": 1.3724,
"step": 2267
},
{
"epoch": 0.4030387844862055,
"grad_norm": 0.36324650779651735,
"learning_rate": 2.6362434133731022e-05,
"loss": 1.3785,
"step": 2268
},
{
"epoch": 0.40321649118130526,
"grad_norm": 0.38071888700082046,
"learning_rate": 2.635173873613097e-05,
"loss": 1.3754,
"step": 2269
},
{
"epoch": 0.403394197876405,
"grad_norm": 0.38689658753726724,
"learning_rate": 2.6341041317923374e-05,
"loss": 1.4078,
"step": 2270
},
{
"epoch": 0.4035719045715047,
"grad_norm": 0.36428077174364204,
"learning_rate": 2.6330341882511285e-05,
"loss": 1.3329,
"step": 2271
},
{
"epoch": 0.40374961126660447,
"grad_norm": 0.37973950431865044,
"learning_rate": 2.63196404332984e-05,
"loss": 1.3631,
"step": 2272
},
{
"epoch": 0.4039273179617042,
"grad_norm": 0.3840587315053361,
"learning_rate": 2.6308936973689045e-05,
"loss": 1.3526,
"step": 2273
},
{
"epoch": 0.4041050246568039,
"grad_norm": 0.38137734355205466,
"learning_rate": 2.62982315070882e-05,
"loss": 1.3625,
"step": 2274
},
{
"epoch": 0.4042827313519037,
"grad_norm": 0.37572945605051283,
"learning_rate": 2.628752403690146e-05,
"loss": 1.3959,
"step": 2275
},
{
"epoch": 0.40446043804700343,
"grad_norm": 0.36967537753531604,
"learning_rate": 2.627681456653508e-05,
"loss": 1.3942,
"step": 2276
},
{
"epoch": 0.40463814474210313,
"grad_norm": 0.3756502029239785,
"learning_rate": 2.6266103099395953e-05,
"loss": 1.4146,
"step": 2277
},
{
"epoch": 0.4048158514372029,
"grad_norm": 0.36433147221752377,
"learning_rate": 2.625538963889159e-05,
"loss": 1.3342,
"step": 2278
},
{
"epoch": 0.40499355813230264,
"grad_norm": 0.3647409157994278,
"learning_rate": 2.6244674188430145e-05,
"loss": 1.3433,
"step": 2279
},
{
"epoch": 0.4051712648274024,
"grad_norm": 0.39477654688647756,
"learning_rate": 2.6233956751420403e-05,
"loss": 1.3613,
"step": 2280
},
{
"epoch": 0.4053489715225021,
"grad_norm": 0.363324301498071,
"learning_rate": 2.6223237331271785e-05,
"loss": 1.394,
"step": 2281
},
{
"epoch": 0.40552667821760185,
"grad_norm": 0.37196892174562574,
"learning_rate": 2.6212515931394337e-05,
"loss": 1.371,
"step": 2282
},
{
"epoch": 0.4057043849127016,
"grad_norm": 0.3689203250859051,
"learning_rate": 2.620179255519873e-05,
"loss": 1.3743,
"step": 2283
},
{
"epoch": 0.4058820916078013,
"grad_norm": 0.36484216317901697,
"learning_rate": 2.6191067206096293e-05,
"loss": 1.3537,
"step": 2284
},
{
"epoch": 0.40605979830290106,
"grad_norm": 0.37548204718921047,
"learning_rate": 2.618033988749895e-05,
"loss": 1.384,
"step": 2285
},
{
"epoch": 0.4062375049980008,
"grad_norm": 0.381238958957241,
"learning_rate": 2.6169610602819262e-05,
"loss": 1.3652,
"step": 2286
},
{
"epoch": 0.4064152116931005,
"grad_norm": 0.36683229717233773,
"learning_rate": 2.6158879355470418e-05,
"loss": 1.3659,
"step": 2287
},
{
"epoch": 0.4065929183882003,
"grad_norm": 0.3766564752197739,
"learning_rate": 2.6148146148866217e-05,
"loss": 1.3516,
"step": 2288
},
{
"epoch": 0.40677062508330003,
"grad_norm": 0.38272038200092195,
"learning_rate": 2.6137410986421118e-05,
"loss": 1.367,
"step": 2289
},
{
"epoch": 0.40694833177839973,
"grad_norm": 0.37396853168675803,
"learning_rate": 2.6126673871550155e-05,
"loss": 1.3748,
"step": 2290
},
{
"epoch": 0.4071260384734995,
"grad_norm": 0.37419046601363093,
"learning_rate": 2.611593480766902e-05,
"loss": 1.3722,
"step": 2291
},
{
"epoch": 0.40730374516859924,
"grad_norm": 0.36879969445797145,
"learning_rate": 2.610519379819401e-05,
"loss": 1.3674,
"step": 2292
},
{
"epoch": 0.40748145186369894,
"grad_norm": 0.4014595143794716,
"learning_rate": 2.609445084654204e-05,
"loss": 1.3681,
"step": 2293
},
{
"epoch": 0.4076591585587987,
"grad_norm": 0.3674780088056944,
"learning_rate": 2.6083705956130638e-05,
"loss": 1.373,
"step": 2294
},
{
"epoch": 0.40783686525389845,
"grad_norm": 0.37434556908215866,
"learning_rate": 2.6072959130377965e-05,
"loss": 1.4071,
"step": 2295
},
{
"epoch": 0.4080145719489982,
"grad_norm": 0.36978595938703845,
"learning_rate": 2.6062210372702784e-05,
"loss": 1.3701,
"step": 2296
},
{
"epoch": 0.4081922786440979,
"grad_norm": 0.3728928263297197,
"learning_rate": 2.6051459686524484e-05,
"loss": 1.3916,
"step": 2297
},
{
"epoch": 0.40836998533919766,
"grad_norm": 0.3817637096482412,
"learning_rate": 2.604070707526305e-05,
"loss": 1.3927,
"step": 2298
},
{
"epoch": 0.4085476920342974,
"grad_norm": 0.3701625269030061,
"learning_rate": 2.602995254233909e-05,
"loss": 1.3547,
"step": 2299
},
{
"epoch": 0.4087253987293971,
"grad_norm": 0.3692082834679575,
"learning_rate": 2.6019196091173843e-05,
"loss": 1.4041,
"step": 2300
},
{
"epoch": 0.40890310542449687,
"grad_norm": 0.36521702830936914,
"learning_rate": 2.6008437725189116e-05,
"loss": 1.3272,
"step": 2301
},
{
"epoch": 0.4090808121195966,
"grad_norm": 0.3745097034754463,
"learning_rate": 2.599767744780735e-05,
"loss": 1.3625,
"step": 2302
},
{
"epoch": 0.4092585188146963,
"grad_norm": 0.3631276386939826,
"learning_rate": 2.59869152624516e-05,
"loss": 1.3173,
"step": 2303
},
{
"epoch": 0.4094362255097961,
"grad_norm": 0.3653002182323476,
"learning_rate": 2.5976151172545514e-05,
"loss": 1.3131,
"step": 2304
},
{
"epoch": 0.40961393220489584,
"grad_norm": 0.35938410847728314,
"learning_rate": 2.596538518151336e-05,
"loss": 1.329,
"step": 2305
},
{
"epoch": 0.40979163889999554,
"grad_norm": 0.3756698359485644,
"learning_rate": 2.5954617292779984e-05,
"loss": 1.3751,
"step": 2306
},
{
"epoch": 0.4099693455950953,
"grad_norm": 0.3591273757317011,
"learning_rate": 2.5943847509770878e-05,
"loss": 1.3345,
"step": 2307
},
{
"epoch": 0.41014705229019505,
"grad_norm": 0.3704273395113719,
"learning_rate": 2.5933075835912095e-05,
"loss": 1.4132,
"step": 2308
},
{
"epoch": 0.41032475898529475,
"grad_norm": 0.38010583196825565,
"learning_rate": 2.592230227463031e-05,
"loss": 1.376,
"step": 2309
},
{
"epoch": 0.4105024656803945,
"grad_norm": 0.6448396847403634,
"learning_rate": 2.5911526829352803e-05,
"loss": 1.3473,
"step": 2310
},
{
"epoch": 0.41068017237549426,
"grad_norm": 0.38676244842113655,
"learning_rate": 2.5900749503507438e-05,
"loss": 1.3377,
"step": 2311
},
{
"epoch": 0.410857879070594,
"grad_norm": 0.3769930458940102,
"learning_rate": 2.5889970300522684e-05,
"loss": 1.4034,
"step": 2312
},
{
"epoch": 0.4110355857656937,
"grad_norm": 0.39016758752896535,
"learning_rate": 2.5879189223827607e-05,
"loss": 1.3499,
"step": 2313
},
{
"epoch": 0.41121329246079347,
"grad_norm": 0.4885064844128314,
"learning_rate": 2.5868406276851886e-05,
"loss": 1.3552,
"step": 2314
},
{
"epoch": 0.4113909991558932,
"grad_norm": 0.39079990161170225,
"learning_rate": 2.5857621463025765e-05,
"loss": 1.3965,
"step": 2315
},
{
"epoch": 0.4115687058509929,
"grad_norm": 0.4454390221647807,
"learning_rate": 2.5846834785780096e-05,
"loss": 1.412,
"step": 2316
},
{
"epoch": 0.4117464125460927,
"grad_norm": 0.3769735278708659,
"learning_rate": 2.583604624854633e-05,
"loss": 1.3294,
"step": 2317
},
{
"epoch": 0.41192411924119243,
"grad_norm": 0.3715972275644503,
"learning_rate": 2.5825255854756494e-05,
"loss": 1.3355,
"step": 2318
},
{
"epoch": 0.41210182593629213,
"grad_norm": 0.3648385476503677,
"learning_rate": 2.581446360784323e-05,
"loss": 1.309,
"step": 2319
},
{
"epoch": 0.4122795326313919,
"grad_norm": 0.38492071118601323,
"learning_rate": 2.5803669511239743e-05,
"loss": 1.3424,
"step": 2320
},
{
"epoch": 0.41245723932649164,
"grad_norm": 0.37192028840848074,
"learning_rate": 2.579287356837984e-05,
"loss": 1.3546,
"step": 2321
},
{
"epoch": 0.41263494602159134,
"grad_norm": 0.37093144651678667,
"learning_rate": 2.578207578269792e-05,
"loss": 1.3658,
"step": 2322
},
{
"epoch": 0.4128126527166911,
"grad_norm": 0.37634505200600166,
"learning_rate": 2.577127615762895e-05,
"loss": 1.3365,
"step": 2323
},
{
"epoch": 0.41299035941179085,
"grad_norm": 0.37111922601044206,
"learning_rate": 2.576047469660851e-05,
"loss": 1.3304,
"step": 2324
},
{
"epoch": 0.41316806610689055,
"grad_norm": 0.3614174021506481,
"learning_rate": 2.5749671403072726e-05,
"loss": 1.3204,
"step": 2325
},
{
"epoch": 0.4133457728019903,
"grad_norm": 0.36951618389514373,
"learning_rate": 2.5738866280458347e-05,
"loss": 1.3387,
"step": 2326
},
{
"epoch": 0.41352347949709006,
"grad_norm": 0.3782220016632725,
"learning_rate": 2.5728059332202683e-05,
"loss": 1.3518,
"step": 2327
},
{
"epoch": 0.4137011861921898,
"grad_norm": 0.3761130947236687,
"learning_rate": 2.571725056174362e-05,
"loss": 1.3613,
"step": 2328
},
{
"epoch": 0.4138788928872895,
"grad_norm": 0.3772009802920589,
"learning_rate": 2.570643997251964e-05,
"loss": 1.3411,
"step": 2329
},
{
"epoch": 0.4140565995823893,
"grad_norm": 0.3782566947270513,
"learning_rate": 2.5695627567969786e-05,
"loss": 1.3497,
"step": 2330
},
{
"epoch": 0.41423430627748903,
"grad_norm": 0.3708416081240294,
"learning_rate": 2.5684813351533693e-05,
"loss": 1.3771,
"step": 2331
},
{
"epoch": 0.41441201297258873,
"grad_norm": 0.3660558662558747,
"learning_rate": 2.567399732665156e-05,
"loss": 1.3451,
"step": 2332
},
{
"epoch": 0.4145897196676885,
"grad_norm": 0.3703186872009514,
"learning_rate": 2.5663179496764184e-05,
"loss": 1.3057,
"step": 2333
},
{
"epoch": 0.41476742636278824,
"grad_norm": 0.38118204841398584,
"learning_rate": 2.5652359865312907e-05,
"loss": 1.3896,
"step": 2334
},
{
"epoch": 0.41494513305788794,
"grad_norm": 0.38405359765919656,
"learning_rate": 2.5641538435739656e-05,
"loss": 1.3446,
"step": 2335
},
{
"epoch": 0.4151228397529877,
"grad_norm": 0.3569880594121409,
"learning_rate": 2.5630715211486935e-05,
"loss": 1.3069,
"step": 2336
},
{
"epoch": 0.41530054644808745,
"grad_norm": 0.3770268321988843,
"learning_rate": 2.561989019599781e-05,
"loss": 1.3869,
"step": 2337
},
{
"epoch": 0.41547825314318715,
"grad_norm": 0.37831543388037175,
"learning_rate": 2.5609063392715937e-05,
"loss": 1.3679,
"step": 2338
},
{
"epoch": 0.4156559598382869,
"grad_norm": 0.36324953941678656,
"learning_rate": 2.5598234805085505e-05,
"loss": 1.344,
"step": 2339
},
{
"epoch": 0.41583366653338666,
"grad_norm": 0.37776845678876897,
"learning_rate": 2.5587404436551307e-05,
"loss": 1.3036,
"step": 2340
},
{
"epoch": 0.41601137322848636,
"grad_norm": 0.37493529109966117,
"learning_rate": 2.5576572290558686e-05,
"loss": 1.3757,
"step": 2341
},
{
"epoch": 0.4161890799235861,
"grad_norm": 0.35571555778923974,
"learning_rate": 2.5565738370553542e-05,
"loss": 1.3198,
"step": 2342
},
{
"epoch": 0.41636678661868587,
"grad_norm": 0.38268191988498174,
"learning_rate": 2.555490267998236e-05,
"loss": 1.3833,
"step": 2343
},
{
"epoch": 0.4165444933137856,
"grad_norm": 0.36631398389539993,
"learning_rate": 2.554406522229216e-05,
"loss": 1.3987,
"step": 2344
},
{
"epoch": 0.4167222000088853,
"grad_norm": 0.36447645789443867,
"learning_rate": 2.5533226000930563e-05,
"loss": 1.3619,
"step": 2345
},
{
"epoch": 0.4168999067039851,
"grad_norm": 0.369724239727089,
"learning_rate": 2.552238501934571e-05,
"loss": 1.3656,
"step": 2346
},
{
"epoch": 0.41707761339908483,
"grad_norm": 0.3622063506807623,
"learning_rate": 2.5511542280986334e-05,
"loss": 1.3606,
"step": 2347
},
{
"epoch": 0.41725532009418453,
"grad_norm": 0.3523011985067753,
"learning_rate": 2.5500697789301705e-05,
"loss": 1.3257,
"step": 2348
},
{
"epoch": 0.4174330267892843,
"grad_norm": 0.42100246905369515,
"learning_rate": 2.5489851547741672e-05,
"loss": 1.3246,
"step": 2349
},
{
"epoch": 0.41761073348438404,
"grad_norm": 0.37106627714345597,
"learning_rate": 2.5479003559756613e-05,
"loss": 1.4184,
"step": 2350
},
{
"epoch": 0.41778844017948374,
"grad_norm": 0.3655606003182911,
"learning_rate": 2.5468153828797486e-05,
"loss": 1.3506,
"step": 2351
},
{
"epoch": 0.4179661468745835,
"grad_norm": 0.3581987603850164,
"learning_rate": 2.545730235831579e-05,
"loss": 1.3407,
"step": 2352
},
{
"epoch": 0.41814385356968325,
"grad_norm": 0.37165528632356354,
"learning_rate": 2.5446449151763593e-05,
"loss": 1.3497,
"step": 2353
},
{
"epoch": 0.41832156026478295,
"grad_norm": 0.3694306120930717,
"learning_rate": 2.543559421259349e-05,
"loss": 1.3801,
"step": 2354
},
{
"epoch": 0.4184992669598827,
"grad_norm": 0.36890138999389693,
"learning_rate": 2.5424737544258644e-05,
"loss": 1.3637,
"step": 2355
},
{
"epoch": 0.41867697365498246,
"grad_norm": 0.37576834042952373,
"learning_rate": 2.541387915021278e-05,
"loss": 1.3789,
"step": 2356
},
{
"epoch": 0.41885468035008216,
"grad_norm": 0.38155947019695174,
"learning_rate": 2.5403019033910137e-05,
"loss": 1.4147,
"step": 2357
},
{
"epoch": 0.4190323870451819,
"grad_norm": 0.37030929568445786,
"learning_rate": 2.5392157198805527e-05,
"loss": 1.3544,
"step": 2358
},
{
"epoch": 0.4192100937402817,
"grad_norm": 0.3776214213750152,
"learning_rate": 2.538129364835431e-05,
"loss": 1.3865,
"step": 2359
},
{
"epoch": 0.41938780043538143,
"grad_norm": 0.38428987431906775,
"learning_rate": 2.537042838601239e-05,
"loss": 1.3937,
"step": 2360
},
{
"epoch": 0.41956550713048113,
"grad_norm": 0.3764978408999048,
"learning_rate": 2.53595614152362e-05,
"loss": 1.3975,
"step": 2361
},
{
"epoch": 0.4197432138255809,
"grad_norm": 0.4022274083838417,
"learning_rate": 2.5348692739482733e-05,
"loss": 1.3668,
"step": 2362
},
{
"epoch": 0.41992092052068064,
"grad_norm": 0.36611524877086477,
"learning_rate": 2.533782236220952e-05,
"loss": 1.3393,
"step": 2363
},
{
"epoch": 0.42009862721578034,
"grad_norm": 0.3831727674422021,
"learning_rate": 2.5326950286874636e-05,
"loss": 1.3516,
"step": 2364
},
{
"epoch": 0.4202763339108801,
"grad_norm": 0.3796039763154656,
"learning_rate": 2.5316076516936683e-05,
"loss": 1.3767,
"step": 2365
},
{
"epoch": 0.42045404060597985,
"grad_norm": 0.3613483537317841,
"learning_rate": 2.5305201055854815e-05,
"loss": 1.3421,
"step": 2366
},
{
"epoch": 0.42063174730107955,
"grad_norm": 0.38039216702237644,
"learning_rate": 2.5294323907088724e-05,
"loss": 1.3887,
"step": 2367
},
{
"epoch": 0.4208094539961793,
"grad_norm": 0.37332063147033506,
"learning_rate": 2.5283445074098634e-05,
"loss": 1.3564,
"step": 2368
},
{
"epoch": 0.42098716069127906,
"grad_norm": 0.3863617987784046,
"learning_rate": 2.5272564560345306e-05,
"loss": 1.4083,
"step": 2369
},
{
"epoch": 0.42116486738637876,
"grad_norm": 0.3784515461580594,
"learning_rate": 2.526168236929004e-05,
"loss": 1.343,
"step": 2370
},
{
"epoch": 0.4213425740814785,
"grad_norm": 0.3504632545263714,
"learning_rate": 2.5250798504394656e-05,
"loss": 1.3738,
"step": 2371
},
{
"epoch": 0.42152028077657827,
"grad_norm": 0.3657041776916383,
"learning_rate": 2.5239912969121527e-05,
"loss": 1.3127,
"step": 2372
},
{
"epoch": 0.42169798747167797,
"grad_norm": 0.39021725773059884,
"learning_rate": 2.5229025766933538e-05,
"loss": 1.3685,
"step": 2373
},
{
"epoch": 0.4218756941667777,
"grad_norm": 0.3776908094233572,
"learning_rate": 2.5218136901294115e-05,
"loss": 1.3659,
"step": 2374
},
{
"epoch": 0.4220534008618775,
"grad_norm": 0.38070218376462694,
"learning_rate": 2.5207246375667217e-05,
"loss": 1.3792,
"step": 2375
},
{
"epoch": 0.42223110755697724,
"grad_norm": 0.36666936901038155,
"learning_rate": 2.5196354193517317e-05,
"loss": 1.3082,
"step": 2376
},
{
"epoch": 0.42240881425207694,
"grad_norm": 0.3861134954156074,
"learning_rate": 2.5185460358309426e-05,
"loss": 1.3869,
"step": 2377
},
{
"epoch": 0.4225865209471767,
"grad_norm": 0.3688997106821003,
"learning_rate": 2.5174564873509086e-05,
"loss": 1.35,
"step": 2378
},
{
"epoch": 0.42276422764227645,
"grad_norm": 0.37208934956180695,
"learning_rate": 2.5163667742582337e-05,
"loss": 1.3823,
"step": 2379
},
{
"epoch": 0.42294193433737615,
"grad_norm": 0.37186059402784377,
"learning_rate": 2.515276896899578e-05,
"loss": 1.3506,
"step": 2380
},
{
"epoch": 0.4231196410324759,
"grad_norm": 0.37897797736318933,
"learning_rate": 2.5141868556216504e-05,
"loss": 1.3775,
"step": 2381
},
{
"epoch": 0.42329734772757566,
"grad_norm": 0.3580118662022412,
"learning_rate": 2.5130966507712146e-05,
"loss": 1.2847,
"step": 2382
},
{
"epoch": 0.42347505442267536,
"grad_norm": 0.38649161741796545,
"learning_rate": 2.5120062826950853e-05,
"loss": 1.3802,
"step": 2383
},
{
"epoch": 0.4236527611177751,
"grad_norm": 0.39547579457425697,
"learning_rate": 2.5109157517401283e-05,
"loss": 1.3831,
"step": 2384
},
{
"epoch": 0.42383046781287487,
"grad_norm": 0.40301291170276105,
"learning_rate": 2.509825058253263e-05,
"loss": 1.3891,
"step": 2385
},
{
"epoch": 0.42400817450797457,
"grad_norm": 0.37214182887704234,
"learning_rate": 2.5087342025814584e-05,
"loss": 1.3449,
"step": 2386
},
{
"epoch": 0.4241858812030743,
"grad_norm": 0.3780278261711136,
"learning_rate": 2.5076431850717375e-05,
"loss": 1.4096,
"step": 2387
},
{
"epoch": 0.4243635878981741,
"grad_norm": 0.38598308732512915,
"learning_rate": 2.5065520060711717e-05,
"loss": 1.3598,
"step": 2388
},
{
"epoch": 0.4245412945932738,
"grad_norm": 0.36593642280975036,
"learning_rate": 2.5054606659268876e-05,
"loss": 1.3642,
"step": 2389
},
{
"epoch": 0.42471900128837353,
"grad_norm": 0.3938478706883254,
"learning_rate": 2.50436916498606e-05,
"loss": 1.3681,
"step": 2390
},
{
"epoch": 0.4248967079834733,
"grad_norm": 0.3753203870640419,
"learning_rate": 2.503277503595915e-05,
"loss": 1.3988,
"step": 2391
},
{
"epoch": 0.42507441467857304,
"grad_norm": 0.37133668408237525,
"learning_rate": 2.5021856821037328e-05,
"loss": 1.3972,
"step": 2392
},
{
"epoch": 0.42525212137367274,
"grad_norm": 0.38652078312362,
"learning_rate": 2.5010937008568398e-05,
"loss": 1.3559,
"step": 2393
},
{
"epoch": 0.4254298280687725,
"grad_norm": 0.35873140963275923,
"learning_rate": 2.5000015602026183e-05,
"loss": 1.3492,
"step": 2394
},
{
"epoch": 0.42560753476387225,
"grad_norm": 0.37542617610833,
"learning_rate": 2.4989092604884966e-05,
"loss": 1.3856,
"step": 2395
},
{
"epoch": 0.42578524145897195,
"grad_norm": 0.3755752117648042,
"learning_rate": 2.4978168020619574e-05,
"loss": 1.3987,
"step": 2396
},
{
"epoch": 0.4259629481540717,
"grad_norm": 0.3730599691370295,
"learning_rate": 2.4967241852705316e-05,
"loss": 1.3601,
"step": 2397
},
{
"epoch": 0.42614065484917146,
"grad_norm": 0.3651611075277475,
"learning_rate": 2.4956314104618007e-05,
"loss": 1.3509,
"step": 2398
},
{
"epoch": 0.42631836154427116,
"grad_norm": 0.35847872066029823,
"learning_rate": 2.4945384779833974e-05,
"loss": 1.341,
"step": 2399
},
{
"epoch": 0.4264960682393709,
"grad_norm": 0.3582103695198002,
"learning_rate": 2.493445388183004e-05,
"loss": 1.3001,
"step": 2400
},
{
"epoch": 0.4266737749344707,
"grad_norm": 0.3563974600563056,
"learning_rate": 2.4923521414083532e-05,
"loss": 1.3216,
"step": 2401
},
{
"epoch": 0.4268514816295704,
"grad_norm": 0.3799486762593359,
"learning_rate": 2.4912587380072273e-05,
"loss": 1.3716,
"step": 2402
},
{
"epoch": 0.42702918832467013,
"grad_norm": 0.3735513184866531,
"learning_rate": 2.490165178327458e-05,
"loss": 1.3508,
"step": 2403
},
{
"epoch": 0.4272068950197699,
"grad_norm": 0.3808941436091705,
"learning_rate": 2.4890714627169273e-05,
"loss": 1.3434,
"step": 2404
},
{
"epoch": 0.4273846017148696,
"grad_norm": 0.38474303631771356,
"learning_rate": 2.4879775915235674e-05,
"loss": 1.378,
"step": 2405
},
{
"epoch": 0.42756230840996934,
"grad_norm": 0.3668329112834644,
"learning_rate": 2.486883565095359e-05,
"loss": 1.369,
"step": 2406
},
{
"epoch": 0.4277400151050691,
"grad_norm": 0.38296709397377376,
"learning_rate": 2.4857893837803313e-05,
"loss": 1.3593,
"step": 2407
},
{
"epoch": 0.42791772180016885,
"grad_norm": 0.37541262776896134,
"learning_rate": 2.4846950479265656e-05,
"loss": 1.3643,
"step": 2408
},
{
"epoch": 0.42809542849526855,
"grad_norm": 0.44179600608271563,
"learning_rate": 2.48360055788219e-05,
"loss": 1.3698,
"step": 2409
},
{
"epoch": 0.4282731351903683,
"grad_norm": 0.3704271827379285,
"learning_rate": 2.4825059139953826e-05,
"loss": 1.3464,
"step": 2410
},
{
"epoch": 0.42845084188546806,
"grad_norm": 0.37195575205082904,
"learning_rate": 2.4814111166143693e-05,
"loss": 1.3656,
"step": 2411
},
{
"epoch": 0.42862854858056776,
"grad_norm": 0.3802181821378268,
"learning_rate": 2.4803161660874272e-05,
"loss": 1.3767,
"step": 2412
},
{
"epoch": 0.4288062552756675,
"grad_norm": 0.36943973496177035,
"learning_rate": 2.4792210627628802e-05,
"loss": 1.3502,
"step": 2413
},
{
"epoch": 0.42898396197076727,
"grad_norm": 0.3685747575826343,
"learning_rate": 2.4781258069891e-05,
"loss": 1.3598,
"step": 2414
},
{
"epoch": 0.42916166866586697,
"grad_norm": 0.3741228383415003,
"learning_rate": 2.4770303991145097e-05,
"loss": 1.3202,
"step": 2415
},
{
"epoch": 0.4293393753609667,
"grad_norm": 0.3682257708532437,
"learning_rate": 2.4759348394875782e-05,
"loss": 1.3055,
"step": 2416
},
{
"epoch": 0.4295170820560665,
"grad_norm": 0.3763061976956641,
"learning_rate": 2.4748391284568244e-05,
"loss": 1.3605,
"step": 2417
},
{
"epoch": 0.4296947887511662,
"grad_norm": 0.6037316819956852,
"learning_rate": 2.4737432663708128e-05,
"loss": 1.3597,
"step": 2418
},
{
"epoch": 0.42987249544626593,
"grad_norm": 0.3663663781550869,
"learning_rate": 2.47264725357816e-05,
"loss": 1.3228,
"step": 2419
},
{
"epoch": 0.4300502021413657,
"grad_norm": 0.3820202368263099,
"learning_rate": 2.4715510904275276e-05,
"loss": 1.353,
"step": 2420
},
{
"epoch": 0.4302279088364654,
"grad_norm": 0.42682356634839697,
"learning_rate": 2.4704547772676247e-05,
"loss": 1.384,
"step": 2421
},
{
"epoch": 0.43040561553156514,
"grad_norm": 0.3695863652788531,
"learning_rate": 2.4693583144472105e-05,
"loss": 1.3833,
"step": 2422
},
{
"epoch": 0.4305833222266649,
"grad_norm": 0.3709696612822972,
"learning_rate": 2.468261702315089e-05,
"loss": 1.3656,
"step": 2423
},
{
"epoch": 0.43076102892176465,
"grad_norm": 0.38311420504737276,
"learning_rate": 2.4671649412201154e-05,
"loss": 1.3614,
"step": 2424
},
{
"epoch": 0.43093873561686435,
"grad_norm": 0.3703093793779499,
"learning_rate": 2.466068031511187e-05,
"loss": 1.3518,
"step": 2425
},
{
"epoch": 0.4311164423119641,
"grad_norm": 0.3736138756673212,
"learning_rate": 2.4649709735372538e-05,
"loss": 1.352,
"step": 2426
},
{
"epoch": 0.43129414900706387,
"grad_norm": 0.38183067854973496,
"learning_rate": 2.4638737676473095e-05,
"loss": 1.3366,
"step": 2427
},
{
"epoch": 0.43147185570216356,
"grad_norm": 0.3715035752996116,
"learning_rate": 2.462776414190396e-05,
"loss": 1.327,
"step": 2428
},
{
"epoch": 0.4316495623972633,
"grad_norm": 0.36720829775952324,
"learning_rate": 2.4616789135156024e-05,
"loss": 1.3551,
"step": 2429
},
{
"epoch": 0.4318272690923631,
"grad_norm": 0.36951530160770313,
"learning_rate": 2.460581265972064e-05,
"loss": 1.3807,
"step": 2430
},
{
"epoch": 0.4320049757874628,
"grad_norm": 0.3777968754516929,
"learning_rate": 2.4594834719089634e-05,
"loss": 1.4109,
"step": 2431
},
{
"epoch": 0.43218268248256253,
"grad_norm": 0.36895648932383046,
"learning_rate": 2.4583855316755293e-05,
"loss": 1.3552,
"step": 2432
},
{
"epoch": 0.4323603891776623,
"grad_norm": 0.3765602204984262,
"learning_rate": 2.4572874456210375e-05,
"loss": 1.3633,
"step": 2433
},
{
"epoch": 0.432538095872762,
"grad_norm": 0.3626466960393105,
"learning_rate": 2.45618921409481e-05,
"loss": 1.3664,
"step": 2434
},
{
"epoch": 0.43271580256786174,
"grad_norm": 0.36704397118823406,
"learning_rate": 2.4550908374462137e-05,
"loss": 1.3701,
"step": 2435
},
{
"epoch": 0.4328935092629615,
"grad_norm": 0.37572094628419545,
"learning_rate": 2.4539923160246638e-05,
"loss": 1.3907,
"step": 2436
},
{
"epoch": 0.4330712159580612,
"grad_norm": 0.36723032950078355,
"learning_rate": 2.4528936501796206e-05,
"loss": 1.2855,
"step": 2437
},
{
"epoch": 0.43324892265316095,
"grad_norm": 0.36733445142713567,
"learning_rate": 2.4517948402605903e-05,
"loss": 1.3332,
"step": 2438
},
{
"epoch": 0.4334266293482607,
"grad_norm": 0.3828742020744556,
"learning_rate": 2.450695886617125e-05,
"loss": 1.4047,
"step": 2439
},
{
"epoch": 0.43360433604336046,
"grad_norm": 0.37430284097762684,
"learning_rate": 2.4495967895988223e-05,
"loss": 1.3814,
"step": 2440
},
{
"epoch": 0.43378204273846016,
"grad_norm": 0.3770474026701448,
"learning_rate": 2.4484975495553256e-05,
"loss": 1.3727,
"step": 2441
},
{
"epoch": 0.4339597494335599,
"grad_norm": 0.3704061189946518,
"learning_rate": 2.4473981668363237e-05,
"loss": 1.3652,
"step": 2442
},
{
"epoch": 0.43413745612865967,
"grad_norm": 0.3784017602011827,
"learning_rate": 2.446298641791552e-05,
"loss": 1.3729,
"step": 2443
},
{
"epoch": 0.43431516282375937,
"grad_norm": 0.3758920961055329,
"learning_rate": 2.4451989747707887e-05,
"loss": 1.3654,
"step": 2444
},
{
"epoch": 0.4344928695188591,
"grad_norm": 0.37155786583024464,
"learning_rate": 2.44409916612386e-05,
"loss": 1.3302,
"step": 2445
},
{
"epoch": 0.4346705762139589,
"grad_norm": 0.3707735305871686,
"learning_rate": 2.442999216200634e-05,
"loss": 1.3399,
"step": 2446
},
{
"epoch": 0.4348482829090586,
"grad_norm": 0.36516776215342295,
"learning_rate": 2.441899125351027e-05,
"loss": 1.339,
"step": 2447
},
{
"epoch": 0.43502598960415834,
"grad_norm": 0.3837988129754436,
"learning_rate": 2.4407988939249978e-05,
"loss": 1.3806,
"step": 2448
},
{
"epoch": 0.4352036962992581,
"grad_norm": 0.3656160623536884,
"learning_rate": 2.4396985222725504e-05,
"loss": 1.3588,
"step": 2449
},
{
"epoch": 0.4353814029943578,
"grad_norm": 0.4206463711173713,
"learning_rate": 2.438598010743735e-05,
"loss": 1.3407,
"step": 2450
},
{
"epoch": 0.43555910968945755,
"grad_norm": 0.3648900508379003,
"learning_rate": 2.437497359688643e-05,
"loss": 1.3026,
"step": 2451
},
{
"epoch": 0.4357368163845573,
"grad_norm": 0.3722804443242332,
"learning_rate": 2.4363965694574142e-05,
"loss": 1.3709,
"step": 2452
},
{
"epoch": 0.435914523079657,
"grad_norm": 0.38199095847070424,
"learning_rate": 2.4352956404002293e-05,
"loss": 1.3658,
"step": 2453
},
{
"epoch": 0.43609222977475676,
"grad_norm": 0.35291410669252193,
"learning_rate": 2.4341945728673162e-05,
"loss": 1.3223,
"step": 2454
},
{
"epoch": 0.4362699364698565,
"grad_norm": 0.3696159173567162,
"learning_rate": 2.4330933672089434e-05,
"loss": 1.3679,
"step": 2455
},
{
"epoch": 0.43644764316495627,
"grad_norm": 0.36186971866916035,
"learning_rate": 2.431992023775425e-05,
"loss": 1.3439,
"step": 2456
},
{
"epoch": 0.43662534986005597,
"grad_norm": 0.3600051392442093,
"learning_rate": 2.430890542917121e-05,
"loss": 1.3457,
"step": 2457
},
{
"epoch": 0.4368030565551557,
"grad_norm": 0.37030565483931305,
"learning_rate": 2.4297889249844318e-05,
"loss": 1.3674,
"step": 2458
},
{
"epoch": 0.4369807632502555,
"grad_norm": 0.36449548453285147,
"learning_rate": 2.428687170327803e-05,
"loss": 1.3201,
"step": 2459
},
{
"epoch": 0.4371584699453552,
"grad_norm": 0.3706868786031279,
"learning_rate": 2.4275852792977227e-05,
"loss": 1.3788,
"step": 2460
},
{
"epoch": 0.43733617664045493,
"grad_norm": 0.3797452262332168,
"learning_rate": 2.426483252244725e-05,
"loss": 1.3675,
"step": 2461
},
{
"epoch": 0.4375138833355547,
"grad_norm": 0.37209389567459616,
"learning_rate": 2.4253810895193844e-05,
"loss": 1.3587,
"step": 2462
},
{
"epoch": 0.4376915900306544,
"grad_norm": 0.35551082416615015,
"learning_rate": 2.4242787914723188e-05,
"loss": 1.3394,
"step": 2463
},
{
"epoch": 0.43786929672575414,
"grad_norm": 0.3689469560495925,
"learning_rate": 2.4231763584541912e-05,
"loss": 1.3169,
"step": 2464
},
{
"epoch": 0.4380470034208539,
"grad_norm": 0.3606483136603109,
"learning_rate": 2.422073790815706e-05,
"loss": 1.3466,
"step": 2465
},
{
"epoch": 0.4382247101159536,
"grad_norm": 0.37891049161067963,
"learning_rate": 2.4209710889076095e-05,
"loss": 1.3563,
"step": 2466
},
{
"epoch": 0.43840241681105335,
"grad_norm": 0.36275891408148236,
"learning_rate": 2.4198682530806942e-05,
"loss": 1.2869,
"step": 2467
},
{
"epoch": 0.4385801235061531,
"grad_norm": 0.3776473569026916,
"learning_rate": 2.4187652836857904e-05,
"loss": 1.3842,
"step": 2468
},
{
"epoch": 0.4387578302012528,
"grad_norm": 0.36634649039683875,
"learning_rate": 2.4176621810737757e-05,
"loss": 1.3564,
"step": 2469
},
{
"epoch": 0.43893553689635256,
"grad_norm": 0.363406339305815,
"learning_rate": 2.4165589455955658e-05,
"loss": 1.3647,
"step": 2470
},
{
"epoch": 0.4391132435914523,
"grad_norm": 0.35808092460195007,
"learning_rate": 2.415455577602122e-05,
"loss": 1.3126,
"step": 2471
},
{
"epoch": 0.4392909502865521,
"grad_norm": 0.36498855003634634,
"learning_rate": 2.4143520774444458e-05,
"loss": 1.3629,
"step": 2472
},
{
"epoch": 0.4394686569816518,
"grad_norm": 0.3630769779739509,
"learning_rate": 2.413248445473582e-05,
"loss": 1.3748,
"step": 2473
},
{
"epoch": 0.43964636367675153,
"grad_norm": 0.37667976271819054,
"learning_rate": 2.4121446820406157e-05,
"loss": 1.3969,
"step": 2474
},
{
"epoch": 0.4398240703718513,
"grad_norm": 0.489233668993759,
"learning_rate": 2.4110407874966753e-05,
"loss": 1.3222,
"step": 2475
},
{
"epoch": 0.440001777066951,
"grad_norm": 0.36855495051436266,
"learning_rate": 2.4099367621929308e-05,
"loss": 1.3578,
"step": 2476
},
{
"epoch": 0.44017948376205074,
"grad_norm": 0.36651610330892465,
"learning_rate": 2.408832606480593e-05,
"loss": 1.3762,
"step": 2477
},
{
"epoch": 0.4403571904571505,
"grad_norm": 0.36678670340726954,
"learning_rate": 2.4077283207109145e-05,
"loss": 1.3426,
"step": 2478
},
{
"epoch": 0.4405348971522502,
"grad_norm": 0.36310801947240706,
"learning_rate": 2.406623905235189e-05,
"loss": 1.3816,
"step": 2479
},
{
"epoch": 0.44071260384734995,
"grad_norm": 0.36274985866414183,
"learning_rate": 2.4055193604047534e-05,
"loss": 1.3592,
"step": 2480
},
{
"epoch": 0.4408903105424497,
"grad_norm": 0.3724773838239447,
"learning_rate": 2.4044146865709825e-05,
"loss": 1.3535,
"step": 2481
},
{
"epoch": 0.4410680172375494,
"grad_norm": 0.35793349328478513,
"learning_rate": 2.403309884085294e-05,
"loss": 1.3368,
"step": 2482
},
{
"epoch": 0.44124572393264916,
"grad_norm": 0.36343033817960085,
"learning_rate": 2.4022049532991476e-05,
"loss": 1.3778,
"step": 2483
},
{
"epoch": 0.4414234306277489,
"grad_norm": 0.34936756584407097,
"learning_rate": 2.4010998945640415e-05,
"loss": 1.2654,
"step": 2484
},
{
"epoch": 0.4416011373228486,
"grad_norm": 0.3721265121923162,
"learning_rate": 2.3999947082315162e-05,
"loss": 1.3746,
"step": 2485
},
{
"epoch": 0.44177884401794837,
"grad_norm": 0.3568307889846751,
"learning_rate": 2.3988893946531513e-05,
"loss": 1.293,
"step": 2486
},
{
"epoch": 0.4419565507130481,
"grad_norm": 0.3662659139613045,
"learning_rate": 2.397783954180569e-05,
"loss": 1.342,
"step": 2487
},
{
"epoch": 0.4421342574081479,
"grad_norm": 0.3665303587462419,
"learning_rate": 2.3966783871654304e-05,
"loss": 1.3358,
"step": 2488
},
{
"epoch": 0.4423119641032476,
"grad_norm": 0.35931217344788907,
"learning_rate": 2.3955726939594363e-05,
"loss": 1.343,
"step": 2489
},
{
"epoch": 0.44248967079834733,
"grad_norm": 0.35781525770695205,
"learning_rate": 2.3944668749143295e-05,
"loss": 1.3545,
"step": 2490
},
{
"epoch": 0.4426673774934471,
"grad_norm": 0.36893829329356326,
"learning_rate": 2.3933609303818916e-05,
"loss": 1.3162,
"step": 2491
},
{
"epoch": 0.4428450841885468,
"grad_norm": 0.37703908118369167,
"learning_rate": 2.3922548607139442e-05,
"loss": 1.3928,
"step": 2492
},
{
"epoch": 0.44302279088364654,
"grad_norm": 0.36956659483986803,
"learning_rate": 2.3911486662623485e-05,
"loss": 1.3653,
"step": 2493
},
{
"epoch": 0.4432004975787463,
"grad_norm": 0.3657052195379849,
"learning_rate": 2.390042347379007e-05,
"loss": 1.3913,
"step": 2494
},
{
"epoch": 0.443378204273846,
"grad_norm": 0.3616298772997449,
"learning_rate": 2.388935904415859e-05,
"loss": 1.3818,
"step": 2495
},
{
"epoch": 0.44355591096894575,
"grad_norm": 0.3599845260099251,
"learning_rate": 2.387829337724886e-05,
"loss": 1.3857,
"step": 2496
},
{
"epoch": 0.4437336176640455,
"grad_norm": 0.36825199567307,
"learning_rate": 2.386722647658107e-05,
"loss": 1.3341,
"step": 2497
},
{
"epoch": 0.4439113243591452,
"grad_norm": 0.37579768747093106,
"learning_rate": 2.385615834567581e-05,
"loss": 1.3611,
"step": 2498
},
{
"epoch": 0.44408903105424496,
"grad_norm": 0.47175414022288126,
"learning_rate": 2.3845088988054067e-05,
"loss": 1.343,
"step": 2499
},
{
"epoch": 0.4442667377493447,
"grad_norm": 0.3753793352760324,
"learning_rate": 2.3834018407237203e-05,
"loss": 1.3661,
"step": 2500
},
{
"epoch": 0.4444444444444444,
"grad_norm": 0.35868214905063567,
"learning_rate": 2.382294660674698e-05,
"loss": 1.3289,
"step": 2501
},
{
"epoch": 0.4446221511395442,
"grad_norm": 0.3760300117959523,
"learning_rate": 2.381187359010555e-05,
"loss": 1.3636,
"step": 2502
},
{
"epoch": 0.44479985783464393,
"grad_norm": 0.3692546989749041,
"learning_rate": 2.3800799360835444e-05,
"loss": 1.3879,
"step": 2503
},
{
"epoch": 0.4449775645297437,
"grad_norm": 0.38139816336101673,
"learning_rate": 2.378972392245959e-05,
"loss": 1.368,
"step": 2504
},
{
"epoch": 0.4451552712248434,
"grad_norm": 0.371630132004165,
"learning_rate": 2.3778647278501277e-05,
"loss": 1.3387,
"step": 2505
},
{
"epoch": 0.44533297791994314,
"grad_norm": 0.363557543612448,
"learning_rate": 2.3767569432484212e-05,
"loss": 1.365,
"step": 2506
},
{
"epoch": 0.4455106846150429,
"grad_norm": 0.3772943098942852,
"learning_rate": 2.3756490387932458e-05,
"loss": 1.3652,
"step": 2507
},
{
"epoch": 0.4456883913101426,
"grad_norm": 0.3583627029781593,
"learning_rate": 2.3745410148370464e-05,
"loss": 1.3453,
"step": 2508
},
{
"epoch": 0.44586609800524235,
"grad_norm": 0.3604169394693747,
"learning_rate": 2.3734328717323073e-05,
"loss": 1.3663,
"step": 2509
},
{
"epoch": 0.4460438047003421,
"grad_norm": 0.3818164893681944,
"learning_rate": 2.372324609831548e-05,
"loss": 1.3719,
"step": 2510
},
{
"epoch": 0.4462215113954418,
"grad_norm": 0.36951250899291455,
"learning_rate": 2.3712162294873293e-05,
"loss": 1.3882,
"step": 2511
},
{
"epoch": 0.44639921809054156,
"grad_norm": 0.37082678119087104,
"learning_rate": 2.370107731052246e-05,
"loss": 1.3734,
"step": 2512
},
{
"epoch": 0.4465769247856413,
"grad_norm": 0.36561232543984495,
"learning_rate": 2.3689991148789337e-05,
"loss": 1.3458,
"step": 2513
},
{
"epoch": 0.446754631480741,
"grad_norm": 0.3624345049988776,
"learning_rate": 2.367890381320064e-05,
"loss": 1.3544,
"step": 2514
},
{
"epoch": 0.44693233817584077,
"grad_norm": 0.37237445408576886,
"learning_rate": 2.3667815307283457e-05,
"loss": 1.3595,
"step": 2515
},
{
"epoch": 0.4471100448709405,
"grad_norm": 0.3603569698073013,
"learning_rate": 2.3656725634565244e-05,
"loss": 1.3824,
"step": 2516
},
{
"epoch": 0.4472877515660402,
"grad_norm": 0.3494519226194827,
"learning_rate": 2.3645634798573832e-05,
"loss": 1.3386,
"step": 2517
},
{
"epoch": 0.44746545826114,
"grad_norm": 0.37890218620856275,
"learning_rate": 2.3634542802837445e-05,
"loss": 1.3852,
"step": 2518
},
{
"epoch": 0.44764316495623974,
"grad_norm": 0.36837881028718755,
"learning_rate": 2.362344965088463e-05,
"loss": 1.3575,
"step": 2519
},
{
"epoch": 0.4478208716513395,
"grad_norm": 0.363458165980323,
"learning_rate": 2.3612355346244346e-05,
"loss": 1.3481,
"step": 2520
},
{
"epoch": 0.4479985783464392,
"grad_norm": 0.34991766612535047,
"learning_rate": 2.3601259892445892e-05,
"loss": 1.3081,
"step": 2521
},
{
"epoch": 0.44817628504153895,
"grad_norm": 0.3757617737950117,
"learning_rate": 2.359016329301894e-05,
"loss": 1.3939,
"step": 2522
},
{
"epoch": 0.4483539917366387,
"grad_norm": 0.3706052849750369,
"learning_rate": 2.3579065551493526e-05,
"loss": 1.407,
"step": 2523
},
{
"epoch": 0.4485316984317384,
"grad_norm": 0.365245108358911,
"learning_rate": 2.3567966671400055e-05,
"loss": 1.3691,
"step": 2524
},
{
"epoch": 0.44870940512683816,
"grad_norm": 0.36340219928223094,
"learning_rate": 2.3556866656269288e-05,
"loss": 1.381,
"step": 2525
},
{
"epoch": 0.4488871118219379,
"grad_norm": 0.3689119172290008,
"learning_rate": 2.354576550963235e-05,
"loss": 1.3798,
"step": 2526
},
{
"epoch": 0.4490648185170376,
"grad_norm": 0.35986130666235555,
"learning_rate": 2.3534663235020715e-05,
"loss": 1.3362,
"step": 2527
},
{
"epoch": 0.44924252521213737,
"grad_norm": 0.3655341946386817,
"learning_rate": 2.3523559835966236e-05,
"loss": 1.3716,
"step": 2528
},
{
"epoch": 0.4494202319072371,
"grad_norm": 0.3539503151036961,
"learning_rate": 2.3512455316001117e-05,
"loss": 1.3221,
"step": 2529
},
{
"epoch": 0.4495979386023368,
"grad_norm": 0.35970275276049474,
"learning_rate": 2.35013496786579e-05,
"loss": 1.3243,
"step": 2530
},
{
"epoch": 0.4497756452974366,
"grad_norm": 0.36583732953029,
"learning_rate": 2.3490242927469506e-05,
"loss": 1.3166,
"step": 2531
},
{
"epoch": 0.44995335199253633,
"grad_norm": 0.36641328917657445,
"learning_rate": 2.34791350659692e-05,
"loss": 1.352,
"step": 2532
},
{
"epoch": 0.45013105868763603,
"grad_norm": 0.35967754356135523,
"learning_rate": 2.34680260976906e-05,
"loss": 1.3376,
"step": 2533
},
{
"epoch": 0.4503087653827358,
"grad_norm": 0.3743407483287956,
"learning_rate": 2.3456916026167683e-05,
"loss": 1.3934,
"step": 2534
},
{
"epoch": 0.45048647207783554,
"grad_norm": 0.35741092896705196,
"learning_rate": 2.344580485493476e-05,
"loss": 1.3505,
"step": 2535
},
{
"epoch": 0.4506641787729353,
"grad_norm": 0.4007646141340526,
"learning_rate": 2.3434692587526517e-05,
"loss": 1.3473,
"step": 2536
},
{
"epoch": 0.450841885468035,
"grad_norm": 0.3679769858150435,
"learning_rate": 2.3423579227477972e-05,
"loss": 1.3502,
"step": 2537
},
{
"epoch": 0.45101959216313475,
"grad_norm": 0.37470004140158025,
"learning_rate": 2.3412464778324485e-05,
"loss": 1.3742,
"step": 2538
},
{
"epoch": 0.4511972988582345,
"grad_norm": 0.3534727787328941,
"learning_rate": 2.3401349243601783e-05,
"loss": 1.3124,
"step": 2539
},
{
"epoch": 0.4513750055533342,
"grad_norm": 0.3663653540908419,
"learning_rate": 2.3390232626845922e-05,
"loss": 1.3769,
"step": 2540
},
{
"epoch": 0.45155271224843396,
"grad_norm": 0.36359691579944314,
"learning_rate": 2.33791149315933e-05,
"loss": 1.3257,
"step": 2541
},
{
"epoch": 0.4517304189435337,
"grad_norm": 0.3692494633829404,
"learning_rate": 2.336799616138067e-05,
"loss": 1.3418,
"step": 2542
},
{
"epoch": 0.4519081256386334,
"grad_norm": 0.3644286598902068,
"learning_rate": 2.335687631974513e-05,
"loss": 1.3349,
"step": 2543
},
{
"epoch": 0.4520858323337332,
"grad_norm": 0.3837693502229869,
"learning_rate": 2.3345755410224107e-05,
"loss": 1.3742,
"step": 2544
},
{
"epoch": 0.45226353902883293,
"grad_norm": 0.36632355918280635,
"learning_rate": 2.3334633436355364e-05,
"loss": 1.3874,
"step": 2545
},
{
"epoch": 0.45244124572393263,
"grad_norm": 0.35741126687898817,
"learning_rate": 2.332351040167701e-05,
"loss": 1.3283,
"step": 2546
},
{
"epoch": 0.4526189524190324,
"grad_norm": 0.36466220212848427,
"learning_rate": 2.3312386309727496e-05,
"loss": 1.3366,
"step": 2547
},
{
"epoch": 0.45279665911413214,
"grad_norm": 0.36817498895381734,
"learning_rate": 2.3301261164045613e-05,
"loss": 1.3765,
"step": 2548
},
{
"epoch": 0.45297436580923184,
"grad_norm": 0.3609299776017188,
"learning_rate": 2.3290134968170462e-05,
"loss": 1.3212,
"step": 2549
},
{
"epoch": 0.4531520725043316,
"grad_norm": 0.36300358661192106,
"learning_rate": 2.3279007725641506e-05,
"loss": 1.3779,
"step": 2550
},
{
"epoch": 0.45332977919943135,
"grad_norm": 0.3658623872078622,
"learning_rate": 2.3267879439998533e-05,
"loss": 1.3731,
"step": 2551
},
{
"epoch": 0.4535074858945311,
"grad_norm": 0.3593989384887376,
"learning_rate": 2.325675011478165e-05,
"loss": 1.3418,
"step": 2552
},
{
"epoch": 0.4536851925896308,
"grad_norm": 0.3819162508246344,
"learning_rate": 2.324561975353131e-05,
"loss": 1.3494,
"step": 2553
},
{
"epoch": 0.45386289928473056,
"grad_norm": 0.3686363732200014,
"learning_rate": 2.323448835978829e-05,
"loss": 1.3703,
"step": 2554
},
{
"epoch": 0.4540406059798303,
"grad_norm": 0.3577800070810984,
"learning_rate": 2.3223355937093697e-05,
"loss": 1.3165,
"step": 2555
},
{
"epoch": 0.45421831267493,
"grad_norm": 0.36624689225122303,
"learning_rate": 2.321222248898896e-05,
"loss": 1.3712,
"step": 2556
},
{
"epoch": 0.45439601937002977,
"grad_norm": 0.37828754195986,
"learning_rate": 2.3201088019015843e-05,
"loss": 1.391,
"step": 2557
},
{
"epoch": 0.4545737260651295,
"grad_norm": 0.37287571108834033,
"learning_rate": 2.3189952530716427e-05,
"loss": 1.3768,
"step": 2558
},
{
"epoch": 0.4547514327602292,
"grad_norm": 0.38078944198365794,
"learning_rate": 2.317881602763312e-05,
"loss": 1.3347,
"step": 2559
},
{
"epoch": 0.454929139455329,
"grad_norm": 0.37191650591950726,
"learning_rate": 2.316767851330866e-05,
"loss": 1.3635,
"step": 2560
},
{
"epoch": 0.45510684615042873,
"grad_norm": 0.3831073633926542,
"learning_rate": 2.3156539991286088e-05,
"loss": 1.3294,
"step": 2561
},
{
"epoch": 0.45528455284552843,
"grad_norm": 0.38610381854553955,
"learning_rate": 2.314540046510878e-05,
"loss": 1.4203,
"step": 2562
},
{
"epoch": 0.4554622595406282,
"grad_norm": 0.37090625201277466,
"learning_rate": 2.313425993832044e-05,
"loss": 1.3746,
"step": 2563
},
{
"epoch": 0.45563996623572794,
"grad_norm": 0.3728914256499427,
"learning_rate": 2.312311841446507e-05,
"loss": 1.3717,
"step": 2564
},
{
"epoch": 0.45581767293082764,
"grad_norm": 0.40310803526355243,
"learning_rate": 2.3111975897086997e-05,
"loss": 1.3845,
"step": 2565
},
{
"epoch": 0.4559953796259274,
"grad_norm": 0.3670825673314574,
"learning_rate": 2.3100832389730865e-05,
"loss": 1.3288,
"step": 2566
},
{
"epoch": 0.45617308632102715,
"grad_norm": 0.3829298949695095,
"learning_rate": 2.3089687895941647e-05,
"loss": 1.3454,
"step": 2567
},
{
"epoch": 0.4563507930161269,
"grad_norm": 0.3689696029602471,
"learning_rate": 2.3078542419264593e-05,
"loss": 1.3588,
"step": 2568
},
{
"epoch": 0.4565284997112266,
"grad_norm": 0.36880397809146925,
"learning_rate": 2.3067395963245307e-05,
"loss": 1.375,
"step": 2569
},
{
"epoch": 0.45670620640632636,
"grad_norm": 0.3602156816712697,
"learning_rate": 2.305624853142968e-05,
"loss": 1.3302,
"step": 2570
},
{
"epoch": 0.4568839131014261,
"grad_norm": 0.38075767144484307,
"learning_rate": 2.3045100127363917e-05,
"loss": 1.3854,
"step": 2571
},
{
"epoch": 0.4570616197965258,
"grad_norm": 0.35805346107680147,
"learning_rate": 2.303395075459454e-05,
"loss": 1.3464,
"step": 2572
},
{
"epoch": 0.4572393264916256,
"grad_norm": 0.3641153373207373,
"learning_rate": 2.302280041666837e-05,
"loss": 1.3212,
"step": 2573
},
{
"epoch": 0.45741703318672533,
"grad_norm": 0.36097113768018957,
"learning_rate": 2.3011649117132543e-05,
"loss": 1.3637,
"step": 2574
},
{
"epoch": 0.45759473988182503,
"grad_norm": 0.35570931844201364,
"learning_rate": 2.3000496859534493e-05,
"loss": 1.3128,
"step": 2575
},
{
"epoch": 0.4577724465769248,
"grad_norm": 0.3684573369916239,
"learning_rate": 2.2989343647421967e-05,
"loss": 1.3439,
"step": 2576
},
{
"epoch": 0.45795015327202454,
"grad_norm": 0.3657872718028968,
"learning_rate": 2.2978189484343007e-05,
"loss": 1.3335,
"step": 2577
},
{
"epoch": 0.45812785996712424,
"grad_norm": 0.36068035739037163,
"learning_rate": 2.2967034373845972e-05,
"loss": 1.3537,
"step": 2578
},
{
"epoch": 0.458305566662224,
"grad_norm": 0.3676315043246795,
"learning_rate": 2.29558783194795e-05,
"loss": 1.3723,
"step": 2579
},
{
"epoch": 0.45848327335732375,
"grad_norm": 0.3664525750444746,
"learning_rate": 2.2944721324792542e-05,
"loss": 1.3184,
"step": 2580
},
{
"epoch": 0.45866098005242345,
"grad_norm": 0.3662849413848241,
"learning_rate": 2.2933563393334364e-05,
"loss": 1.3446,
"step": 2581
},
{
"epoch": 0.4588386867475232,
"grad_norm": 0.36516980559353845,
"learning_rate": 2.2922404528654493e-05,
"loss": 1.3636,
"step": 2582
},
{
"epoch": 0.45901639344262296,
"grad_norm": 0.3729721472111294,
"learning_rate": 2.2911244734302788e-05,
"loss": 1.3615,
"step": 2583
},
{
"epoch": 0.4591941001377227,
"grad_norm": 0.36755445826632194,
"learning_rate": 2.290008401382938e-05,
"loss": 1.3702,
"step": 2584
},
{
"epoch": 0.4593718068328224,
"grad_norm": 0.3715568477576965,
"learning_rate": 2.2888922370784712e-05,
"loss": 1.3611,
"step": 2585
},
{
"epoch": 0.45954951352792217,
"grad_norm": 0.36389415361946975,
"learning_rate": 2.2877759808719513e-05,
"loss": 1.3273,
"step": 2586
},
{
"epoch": 0.4597272202230219,
"grad_norm": 0.37000295805490635,
"learning_rate": 2.2866596331184795e-05,
"loss": 1.3442,
"step": 2587
},
{
"epoch": 0.4599049269181216,
"grad_norm": 0.35401390066807403,
"learning_rate": 2.2855431941731877e-05,
"loss": 1.3463,
"step": 2588
},
{
"epoch": 0.4600826336132214,
"grad_norm": 0.3669906243253639,
"learning_rate": 2.2844266643912357e-05,
"loss": 1.3797,
"step": 2589
},
{
"epoch": 0.46026034030832114,
"grad_norm": 0.3857061046289869,
"learning_rate": 2.2833100441278128e-05,
"loss": 1.3875,
"step": 2590
},
{
"epoch": 0.46043804700342084,
"grad_norm": 0.36666155822743435,
"learning_rate": 2.282193333738137e-05,
"loss": 1.3602,
"step": 2591
},
{
"epoch": 0.4606157536985206,
"grad_norm": 0.37940327200794394,
"learning_rate": 2.2810765335774553e-05,
"loss": 1.3774,
"step": 2592
},
{
"epoch": 0.46079346039362035,
"grad_norm": 0.37090926469094876,
"learning_rate": 2.2799596440010428e-05,
"loss": 1.3044,
"step": 2593
},
{
"epoch": 0.46097116708872005,
"grad_norm": 0.36699177238256386,
"learning_rate": 2.278842665364201e-05,
"loss": 1.3762,
"step": 2594
},
{
"epoch": 0.4611488737838198,
"grad_norm": 0.36350404337437,
"learning_rate": 2.277725598022265e-05,
"loss": 1.3159,
"step": 2595
},
{
"epoch": 0.46132658047891956,
"grad_norm": 0.36041818514525864,
"learning_rate": 2.2766084423305933e-05,
"loss": 1.3185,
"step": 2596
},
{
"epoch": 0.46150428717401926,
"grad_norm": 0.3737702301543396,
"learning_rate": 2.2754911986445736e-05,
"loss": 1.3434,
"step": 2597
},
{
"epoch": 0.461681993869119,
"grad_norm": 0.35758725753853576,
"learning_rate": 2.2743738673196227e-05,
"loss": 1.3412,
"step": 2598
},
{
"epoch": 0.46185970056421877,
"grad_norm": 0.35756142321775597,
"learning_rate": 2.273256448711185e-05,
"loss": 1.3121,
"step": 2599
},
{
"epoch": 0.4620374072593185,
"grad_norm": 0.3617446513906599,
"learning_rate": 2.2721389431747322e-05,
"loss": 1.3357,
"step": 2600
},
{
"epoch": 0.4622151139544182,
"grad_norm": 0.36067688987890945,
"learning_rate": 2.2710213510657638e-05,
"loss": 1.3575,
"step": 2601
},
{
"epoch": 0.462392820649518,
"grad_norm": 0.357950134504415,
"learning_rate": 2.2699036727398074e-05,
"loss": 1.301,
"step": 2602
},
{
"epoch": 0.46257052734461773,
"grad_norm": 0.36142836111594073,
"learning_rate": 2.268785908552416e-05,
"loss": 1.339,
"step": 2603
},
{
"epoch": 0.46274823403971743,
"grad_norm": 0.3571226546807297,
"learning_rate": 2.2676680588591734e-05,
"loss": 1.366,
"step": 2604
},
{
"epoch": 0.4629259407348172,
"grad_norm": 0.37356430511904215,
"learning_rate": 2.2665501240156864e-05,
"loss": 1.3494,
"step": 2605
},
{
"epoch": 0.46310364742991694,
"grad_norm": 0.36242743321505566,
"learning_rate": 2.2654321043775925e-05,
"loss": 1.387,
"step": 2606
},
{
"epoch": 0.46328135412501664,
"grad_norm": 0.3644387718365672,
"learning_rate": 2.264314000300555e-05,
"loss": 1.3831,
"step": 2607
},
{
"epoch": 0.4634590608201164,
"grad_norm": 0.37045697467880256,
"learning_rate": 2.263195812140263e-05,
"loss": 1.3266,
"step": 2608
},
{
"epoch": 0.46363676751521615,
"grad_norm": 0.3620704399442718,
"learning_rate": 2.2620775402524338e-05,
"loss": 1.3301,
"step": 2609
},
{
"epoch": 0.46381447421031585,
"grad_norm": 0.3635890568365683,
"learning_rate": 2.26095918499281e-05,
"loss": 1.3422,
"step": 2610
},
{
"epoch": 0.4639921809054156,
"grad_norm": 0.363623122482695,
"learning_rate": 2.2598407467171623e-05,
"loss": 1.3348,
"step": 2611
},
{
"epoch": 0.46416988760051536,
"grad_norm": 0.368340606927619,
"learning_rate": 2.2587222257812865e-05,
"loss": 1.3692,
"step": 2612
},
{
"epoch": 0.46434759429561506,
"grad_norm": 0.37078675177928555,
"learning_rate": 2.257603622541005e-05,
"loss": 1.3854,
"step": 2613
},
{
"epoch": 0.4645253009907148,
"grad_norm": 0.37197816305365483,
"learning_rate": 2.256484937352167e-05,
"loss": 1.3492,
"step": 2614
},
{
"epoch": 0.4647030076858146,
"grad_norm": 0.3553643616299594,
"learning_rate": 2.255366170570647e-05,
"loss": 1.3136,
"step": 2615
},
{
"epoch": 0.46488071438091433,
"grad_norm": 0.3680394994439378,
"learning_rate": 2.2542473225523457e-05,
"loss": 1.3376,
"step": 2616
},
{
"epoch": 0.46505842107601403,
"grad_norm": 0.3685097246807061,
"learning_rate": 2.253128393653189e-05,
"loss": 1.3443,
"step": 2617
},
{
"epoch": 0.4652361277711138,
"grad_norm": 0.3637030697027318,
"learning_rate": 2.252009384229131e-05,
"loss": 1.3274,
"step": 2618
},
{
"epoch": 0.46541383446621354,
"grad_norm": 0.36965766529856053,
"learning_rate": 2.2508902946361485e-05,
"loss": 1.3126,
"step": 2619
},
{
"epoch": 0.46559154116131324,
"grad_norm": 0.36352250798321967,
"learning_rate": 2.249771125230245e-05,
"loss": 1.3367,
"step": 2620
},
{
"epoch": 0.465769247856413,
"grad_norm": 0.3637091803630393,
"learning_rate": 2.248651876367449e-05,
"loss": 1.3694,
"step": 2621
},
{
"epoch": 0.46594695455151275,
"grad_norm": 0.37484745308843825,
"learning_rate": 2.247532548403815e-05,
"loss": 1.3929,
"step": 2622
},
{
"epoch": 0.46612466124661245,
"grad_norm": 0.3559823929201626,
"learning_rate": 2.246413141695423e-05,
"loss": 1.3096,
"step": 2623
},
{
"epoch": 0.4663023679417122,
"grad_norm": 0.3697707871027922,
"learning_rate": 2.245293656598376e-05,
"loss": 1.3511,
"step": 2624
},
{
"epoch": 0.46648007463681196,
"grad_norm": 0.36824586687461225,
"learning_rate": 2.2441740934688042e-05,
"loss": 1.3295,
"step": 2625
},
{
"epoch": 0.46665778133191166,
"grad_norm": 0.38213332337637496,
"learning_rate": 2.2430544526628615e-05,
"loss": 1.3414,
"step": 2626
},
{
"epoch": 0.4668354880270114,
"grad_norm": 0.38043225021308125,
"learning_rate": 2.2419347345367265e-05,
"loss": 1.4011,
"step": 2627
},
{
"epoch": 0.46701319472211117,
"grad_norm": 0.3702055043780779,
"learning_rate": 2.2408149394466022e-05,
"loss": 1.3779,
"step": 2628
},
{
"epoch": 0.46719090141721087,
"grad_norm": 0.34739519353554854,
"learning_rate": 2.239695067748717e-05,
"loss": 1.3022,
"step": 2629
},
{
"epoch": 0.4673686081123106,
"grad_norm": 0.37708817243013343,
"learning_rate": 2.2385751197993234e-05,
"loss": 1.3793,
"step": 2630
},
{
"epoch": 0.4675463148074104,
"grad_norm": 0.3520454573366823,
"learning_rate": 2.2374550959546974e-05,
"loss": 1.3356,
"step": 2631
},
{
"epoch": 0.46772402150251013,
"grad_norm": 0.3652903482248608,
"learning_rate": 2.2363349965711398e-05,
"loss": 1.3619,
"step": 2632
},
{
"epoch": 0.46790172819760983,
"grad_norm": 0.37025141788854643,
"learning_rate": 2.2352148220049755e-05,
"loss": 1.3633,
"step": 2633
},
{
"epoch": 0.4680794348927096,
"grad_norm": 0.3652129204530205,
"learning_rate": 2.2340945726125528e-05,
"loss": 1.3837,
"step": 2634
},
{
"epoch": 0.46825714158780934,
"grad_norm": 0.3655416847940541,
"learning_rate": 2.2329742487502446e-05,
"loss": 1.3429,
"step": 2635
},
{
"epoch": 0.46843484828290904,
"grad_norm": 0.3862218694119032,
"learning_rate": 2.2318538507744458e-05,
"loss": 1.4004,
"step": 2636
},
{
"epoch": 0.4686125549780088,
"grad_norm": 0.3742234567796507,
"learning_rate": 2.2307333790415774e-05,
"loss": 1.3486,
"step": 2637
},
{
"epoch": 0.46879026167310855,
"grad_norm": 0.36834148195899913,
"learning_rate": 2.229612833908082e-05,
"loss": 1.3806,
"step": 2638
},
{
"epoch": 0.46896796836820825,
"grad_norm": 0.38227974876188614,
"learning_rate": 2.2284922157304258e-05,
"loss": 1.3769,
"step": 2639
},
{
"epoch": 0.469145675063308,
"grad_norm": 0.3741378516464579,
"learning_rate": 2.2273715248650988e-05,
"loss": 1.3946,
"step": 2640
},
{
"epoch": 0.46932338175840776,
"grad_norm": 0.3557786674604447,
"learning_rate": 2.226250761668614e-05,
"loss": 1.3231,
"step": 2641
},
{
"epoch": 0.46950108845350746,
"grad_norm": 0.37569734665344795,
"learning_rate": 2.2251299264975076e-05,
"loss": 1.3454,
"step": 2642
},
{
"epoch": 0.4696787951486072,
"grad_norm": 0.3572781404162457,
"learning_rate": 2.224009019708337e-05,
"loss": 1.3094,
"step": 2643
},
{
"epoch": 0.469856501843707,
"grad_norm": 0.3625741688764768,
"learning_rate": 2.2228880416576846e-05,
"loss": 1.3438,
"step": 2644
},
{
"epoch": 0.4700342085388067,
"grad_norm": 0.37014806397227973,
"learning_rate": 2.221766992702155e-05,
"loss": 1.3514,
"step": 2645
},
{
"epoch": 0.47021191523390643,
"grad_norm": 0.35545148601056614,
"learning_rate": 2.2206458731983745e-05,
"loss": 1.2937,
"step": 2646
},
{
"epoch": 0.4703896219290062,
"grad_norm": 0.37174921903910463,
"learning_rate": 2.2195246835029914e-05,
"loss": 1.3755,
"step": 2647
},
{
"epoch": 0.47056732862410594,
"grad_norm": 0.37356047165924144,
"learning_rate": 2.218403423972679e-05,
"loss": 1.337,
"step": 2648
},
{
"epoch": 0.47074503531920564,
"grad_norm": 0.3649296914270244,
"learning_rate": 2.21728209496413e-05,
"loss": 1.3284,
"step": 2649
},
{
"epoch": 0.4709227420143054,
"grad_norm": 0.36161419352277974,
"learning_rate": 2.216160696834061e-05,
"loss": 1.3547,
"step": 2650
},
{
"epoch": 0.47110044870940515,
"grad_norm": 0.3576902152703704,
"learning_rate": 2.215039229939208e-05,
"loss": 1.3188,
"step": 2651
},
{
"epoch": 0.47127815540450485,
"grad_norm": 0.35801418279744485,
"learning_rate": 2.2139176946363326e-05,
"loss": 1.3072,
"step": 2652
},
{
"epoch": 0.4714558620996046,
"grad_norm": 0.3669673003794392,
"learning_rate": 2.2127960912822162e-05,
"loss": 1.3543,
"step": 2653
},
{
"epoch": 0.47163356879470436,
"grad_norm": 0.3705971690686729,
"learning_rate": 2.2116744202336603e-05,
"loss": 1.3736,
"step": 2654
},
{
"epoch": 0.47181127548980406,
"grad_norm": 0.3703908239707703,
"learning_rate": 2.2105526818474914e-05,
"loss": 1.3782,
"step": 2655
},
{
"epoch": 0.4719889821849038,
"grad_norm": 0.3550612617891946,
"learning_rate": 2.2094308764805545e-05,
"loss": 1.3104,
"step": 2656
},
{
"epoch": 0.47216668888000357,
"grad_norm": 0.3617614143760198,
"learning_rate": 2.2083090044897172e-05,
"loss": 1.326,
"step": 2657
},
{
"epoch": 0.47234439557510327,
"grad_norm": 0.3655246733114863,
"learning_rate": 2.2071870662318683e-05,
"loss": 1.3432,
"step": 2658
},
{
"epoch": 0.472522102270203,
"grad_norm": 0.3589101375744327,
"learning_rate": 2.206065062063917e-05,
"loss": 1.3487,
"step": 2659
},
{
"epoch": 0.4726998089653028,
"grad_norm": 0.34965802393464246,
"learning_rate": 2.2049429923427942e-05,
"loss": 1.2698,
"step": 2660
},
{
"epoch": 0.4728775156604025,
"grad_norm": 0.35397520858965603,
"learning_rate": 2.2038208574254522e-05,
"loss": 1.3101,
"step": 2661
},
{
"epoch": 0.47305522235550224,
"grad_norm": 0.3650289333083801,
"learning_rate": 2.2026986576688616e-05,
"loss": 1.3116,
"step": 2662
},
{
"epoch": 0.473232929050602,
"grad_norm": 0.36040172487658534,
"learning_rate": 2.2015763934300166e-05,
"loss": 1.3311,
"step": 2663
},
{
"epoch": 0.47341063574570175,
"grad_norm": 0.3520903335633863,
"learning_rate": 2.2004540650659297e-05,
"loss": 1.3022,
"step": 2664
},
{
"epoch": 0.47358834244080145,
"grad_norm": 0.3702044559030688,
"learning_rate": 2.1993316729336353e-05,
"loss": 1.3517,
"step": 2665
},
{
"epoch": 0.4737660491359012,
"grad_norm": 0.3593013954260921,
"learning_rate": 2.1982092173901863e-05,
"loss": 1.3735,
"step": 2666
},
{
"epoch": 0.47394375583100096,
"grad_norm": 0.37294453014284673,
"learning_rate": 2.1970866987926585e-05,
"loss": 1.3836,
"step": 2667
},
{
"epoch": 0.47412146252610066,
"grad_norm": 0.3663044094099886,
"learning_rate": 2.1959641174981457e-05,
"loss": 1.3723,
"step": 2668
},
{
"epoch": 0.4742991692212004,
"grad_norm": 0.3521659829623544,
"learning_rate": 2.1948414738637612e-05,
"loss": 1.3527,
"step": 2669
},
{
"epoch": 0.47447687591630017,
"grad_norm": 0.36541943119922443,
"learning_rate": 2.1937187682466404e-05,
"loss": 1.3125,
"step": 2670
},
{
"epoch": 0.47465458261139987,
"grad_norm": 0.366226610526978,
"learning_rate": 2.1925960010039353e-05,
"loss": 1.3626,
"step": 2671
},
{
"epoch": 0.4748322893064996,
"grad_norm": 0.35125651457041435,
"learning_rate": 2.191473172492821e-05,
"loss": 1.3521,
"step": 2672
},
{
"epoch": 0.4750099960015994,
"grad_norm": 0.35768712169065753,
"learning_rate": 2.19035028307049e-05,
"loss": 1.3538,
"step": 2673
},
{
"epoch": 0.4751877026966991,
"grad_norm": 0.36221204633366816,
"learning_rate": 2.189227333094154e-05,
"loss": 1.359,
"step": 2674
},
{
"epoch": 0.47536540939179883,
"grad_norm": 0.37103939557277726,
"learning_rate": 2.1881043229210446e-05,
"loss": 1.3863,
"step": 2675
},
{
"epoch": 0.4755431160868986,
"grad_norm": 0.3585714609277535,
"learning_rate": 2.186981252908413e-05,
"loss": 1.3356,
"step": 2676
},
{
"epoch": 0.4757208227819983,
"grad_norm": 0.3678116096344312,
"learning_rate": 2.185858123413528e-05,
"loss": 1.3692,
"step": 2677
},
{
"epoch": 0.47589852947709804,
"grad_norm": 0.3589095463864255,
"learning_rate": 2.184734934793679e-05,
"loss": 1.3236,
"step": 2678
},
{
"epoch": 0.4760762361721978,
"grad_norm": 0.3569743736787915,
"learning_rate": 2.1836116874061734e-05,
"loss": 1.3481,
"step": 2679
},
{
"epoch": 0.47625394286729755,
"grad_norm": 0.3559962451453174,
"learning_rate": 2.1824883816083365e-05,
"loss": 1.3372,
"step": 2680
},
{
"epoch": 0.47643164956239725,
"grad_norm": 0.35287193640732023,
"learning_rate": 2.181365017757514e-05,
"loss": 1.3269,
"step": 2681
},
{
"epoch": 0.476609356257497,
"grad_norm": 0.3646787851581461,
"learning_rate": 2.180241596211068e-05,
"loss": 1.3903,
"step": 2682
},
{
"epoch": 0.47678706295259676,
"grad_norm": 0.36791090128333925,
"learning_rate": 2.1791181173263815e-05,
"loss": 1.3577,
"step": 2683
},
{
"epoch": 0.47696476964769646,
"grad_norm": 0.35492266554328017,
"learning_rate": 2.1779945814608534e-05,
"loss": 1.3462,
"step": 2684
},
{
"epoch": 0.4771424763427962,
"grad_norm": 0.359579655810323,
"learning_rate": 2.1768709889719005e-05,
"loss": 1.3266,
"step": 2685
},
{
"epoch": 0.477320183037896,
"grad_norm": 0.36395792917588615,
"learning_rate": 2.175747340216961e-05,
"loss": 1.3831,
"step": 2686
},
{
"epoch": 0.4774978897329957,
"grad_norm": 0.3515510667764322,
"learning_rate": 2.174623635553487e-05,
"loss": 1.3079,
"step": 2687
},
{
"epoch": 0.47767559642809543,
"grad_norm": 0.3617117769734793,
"learning_rate": 2.173499875338951e-05,
"loss": 1.3482,
"step": 2688
},
{
"epoch": 0.4778533031231952,
"grad_norm": 0.3674443491628358,
"learning_rate": 2.1723760599308408e-05,
"loss": 1.3644,
"step": 2689
},
{
"epoch": 0.4780310098182949,
"grad_norm": 0.37369527839372874,
"learning_rate": 2.1712521896866657e-05,
"loss": 1.373,
"step": 2690
},
{
"epoch": 0.47820871651339464,
"grad_norm": 0.3714660267826562,
"learning_rate": 2.1701282649639474e-05,
"loss": 1.3542,
"step": 2691
},
{
"epoch": 0.4783864232084944,
"grad_norm": 0.3776047275372483,
"learning_rate": 2.1690042861202286e-05,
"loss": 1.3352,
"step": 2692
},
{
"epoch": 0.4785641299035941,
"grad_norm": 0.36266241816809414,
"learning_rate": 2.1678802535130688e-05,
"loss": 1.3096,
"step": 2693
},
{
"epoch": 0.47874183659869385,
"grad_norm": 0.3826471236490594,
"learning_rate": 2.166756167500043e-05,
"loss": 1.3433,
"step": 2694
},
{
"epoch": 0.4789195432937936,
"grad_norm": 0.36385025977849184,
"learning_rate": 2.1656320284387446e-05,
"loss": 1.3251,
"step": 2695
},
{
"epoch": 0.47909724998889336,
"grad_norm": 0.37886629897590895,
"learning_rate": 2.164507836686782e-05,
"loss": 1.3844,
"step": 2696
},
{
"epoch": 0.47927495668399306,
"grad_norm": 0.3726511491684368,
"learning_rate": 2.1633835926017833e-05,
"loss": 1.3405,
"step": 2697
},
{
"epoch": 0.4794526633790928,
"grad_norm": 0.36445320693615957,
"learning_rate": 2.1622592965413923e-05,
"loss": 1.3551,
"step": 2698
},
{
"epoch": 0.47963037007419257,
"grad_norm": 0.368657259760148,
"learning_rate": 2.1611349488632665e-05,
"loss": 1.3353,
"step": 2699
},
{
"epoch": 0.47980807676929227,
"grad_norm": 0.3795934543677101,
"learning_rate": 2.1600105499250835e-05,
"loss": 1.3358,
"step": 2700
},
{
"epoch": 0.479985783464392,
"grad_norm": 0.3668045790843027,
"learning_rate": 2.158886100084536e-05,
"loss": 1.3423,
"step": 2701
},
{
"epoch": 0.4801634901594918,
"grad_norm": 0.35714332031502666,
"learning_rate": 2.157761599699331e-05,
"loss": 1.3348,
"step": 2702
},
{
"epoch": 0.4803411968545915,
"grad_norm": 0.3720617154996512,
"learning_rate": 2.156637049127195e-05,
"loss": 1.3005,
"step": 2703
},
{
"epoch": 0.48051890354969123,
"grad_norm": 0.36338863318301323,
"learning_rate": 2.1555124487258676e-05,
"loss": 1.3491,
"step": 2704
},
{
"epoch": 0.480696610244791,
"grad_norm": 0.359122654509386,
"learning_rate": 2.1543877988531057e-05,
"loss": 1.3901,
"step": 2705
},
{
"epoch": 0.4808743169398907,
"grad_norm": 0.3719580006892696,
"learning_rate": 2.153263099866682e-05,
"loss": 1.3334,
"step": 2706
},
{
"epoch": 0.48105202363499044,
"grad_norm": 0.3735747087215803,
"learning_rate": 2.1521383521243842e-05,
"loss": 1.3278,
"step": 2707
},
{
"epoch": 0.4812297303300902,
"grad_norm": 0.3645900067525138,
"learning_rate": 2.151013555984015e-05,
"loss": 1.3126,
"step": 2708
},
{
"epoch": 0.4814074370251899,
"grad_norm": 0.3972215853580638,
"learning_rate": 2.149888711803394e-05,
"loss": 1.3763,
"step": 2709
},
{
"epoch": 0.48158514372028965,
"grad_norm": 0.3775115965857101,
"learning_rate": 2.1487638199403548e-05,
"loss": 1.3715,
"step": 2710
},
{
"epoch": 0.4817628504153894,
"grad_norm": 0.3770204555496595,
"learning_rate": 2.1476388807527467e-05,
"loss": 1.3726,
"step": 2711
},
{
"epoch": 0.48194055711048917,
"grad_norm": 0.3881218467055191,
"learning_rate": 2.1465138945984342e-05,
"loss": 1.3861,
"step": 2712
},
{
"epoch": 0.48211826380558886,
"grad_norm": 0.3771831467372762,
"learning_rate": 2.1453888618352966e-05,
"loss": 1.345,
"step": 2713
},
{
"epoch": 0.4822959705006886,
"grad_norm": 0.36773897288536794,
"learning_rate": 2.144263782821228e-05,
"loss": 1.3067,
"step": 2714
},
{
"epoch": 0.4824736771957884,
"grad_norm": 0.3523413183718623,
"learning_rate": 2.143138657914137e-05,
"loss": 1.3253,
"step": 2715
},
{
"epoch": 0.4826513838908881,
"grad_norm": 0.3780601974012108,
"learning_rate": 2.142013487471947e-05,
"loss": 1.3902,
"step": 2716
},
{
"epoch": 0.48282909058598783,
"grad_norm": 0.3669348658595694,
"learning_rate": 2.1408882718525965e-05,
"loss": 1.33,
"step": 2717
},
{
"epoch": 0.4830067972810876,
"grad_norm": 0.3709859678216607,
"learning_rate": 2.1397630114140365e-05,
"loss": 1.3194,
"step": 2718
},
{
"epoch": 0.4831845039761873,
"grad_norm": 0.36583669979596484,
"learning_rate": 2.1386377065142346e-05,
"loss": 1.3379,
"step": 2719
},
{
"epoch": 0.48336221067128704,
"grad_norm": 0.37085221230036125,
"learning_rate": 2.1375123575111706e-05,
"loss": 1.3601,
"step": 2720
},
{
"epoch": 0.4835399173663868,
"grad_norm": 0.35904215723155725,
"learning_rate": 2.1363869647628404e-05,
"loss": 1.3419,
"step": 2721
},
{
"epoch": 0.4837176240614865,
"grad_norm": 0.36553572829177144,
"learning_rate": 2.135261528627251e-05,
"loss": 1.3347,
"step": 2722
},
{
"epoch": 0.48389533075658625,
"grad_norm": 0.3609072616544804,
"learning_rate": 2.1341360494624262e-05,
"loss": 1.3417,
"step": 2723
},
{
"epoch": 0.484073037451686,
"grad_norm": 0.3567302211282411,
"learning_rate": 2.1330105276264012e-05,
"loss": 1.3533,
"step": 2724
},
{
"epoch": 0.4842507441467857,
"grad_norm": 0.3725952669861981,
"learning_rate": 2.131884963477226e-05,
"loss": 1.3003,
"step": 2725
},
{
"epoch": 0.48442845084188546,
"grad_norm": 0.3679438937601488,
"learning_rate": 2.1307593573729642e-05,
"loss": 1.3284,
"step": 2726
},
{
"epoch": 0.4846061575369852,
"grad_norm": 0.3539359392461221,
"learning_rate": 2.129633709671691e-05,
"loss": 1.3499,
"step": 2727
},
{
"epoch": 0.48478386423208497,
"grad_norm": 0.385903867573792,
"learning_rate": 2.1285080207314976e-05,
"loss": 1.3349,
"step": 2728
},
{
"epoch": 0.48496157092718467,
"grad_norm": 0.36320761525786033,
"learning_rate": 2.1273822909104856e-05,
"loss": 1.3458,
"step": 2729
},
{
"epoch": 0.4851392776222844,
"grad_norm": 0.37410873096574154,
"learning_rate": 2.1262565205667714e-05,
"loss": 1.3505,
"step": 2730
},
{
"epoch": 0.4853169843173842,
"grad_norm": 0.3506174502038526,
"learning_rate": 2.125130710058484e-05,
"loss": 1.3172,
"step": 2731
},
{
"epoch": 0.4854946910124839,
"grad_norm": 0.36114231316280015,
"learning_rate": 2.1240048597437645e-05,
"loss": 1.3434,
"step": 2732
},
{
"epoch": 0.48567239770758364,
"grad_norm": 0.367592480295572,
"learning_rate": 2.122878969980767e-05,
"loss": 1.3529,
"step": 2733
},
{
"epoch": 0.4858501044026834,
"grad_norm": 0.3521422737899018,
"learning_rate": 2.121753041127658e-05,
"loss": 1.3303,
"step": 2734
},
{
"epoch": 0.4860278110977831,
"grad_norm": 0.36588192034123895,
"learning_rate": 2.120627073542617e-05,
"loss": 1.3446,
"step": 2735
},
{
"epoch": 0.48620551779288285,
"grad_norm": 0.36081444934875834,
"learning_rate": 2.1195010675838356e-05,
"loss": 1.3245,
"step": 2736
},
{
"epoch": 0.4863832244879826,
"grad_norm": 0.35910345118532927,
"learning_rate": 2.1183750236095176e-05,
"loss": 1.3502,
"step": 2737
},
{
"epoch": 0.4865609311830823,
"grad_norm": 0.36208003395460875,
"learning_rate": 2.1172489419778782e-05,
"loss": 1.3501,
"step": 2738
},
{
"epoch": 0.48673863787818206,
"grad_norm": 0.36121485258395153,
"learning_rate": 2.116122823047145e-05,
"loss": 1.3533,
"step": 2739
},
{
"epoch": 0.4869163445732818,
"grad_norm": 0.36534059322884777,
"learning_rate": 2.1149966671755585e-05,
"loss": 1.3489,
"step": 2740
},
{
"epoch": 0.4870940512683815,
"grad_norm": 0.35257381022660517,
"learning_rate": 2.113870474721369e-05,
"loss": 1.3395,
"step": 2741
},
{
"epoch": 0.48727175796348127,
"grad_norm": 0.3641210546790651,
"learning_rate": 2.1127442460428406e-05,
"loss": 1.3589,
"step": 2742
},
{
"epoch": 0.487449464658581,
"grad_norm": 0.3569506325466698,
"learning_rate": 2.1116179814982473e-05,
"loss": 1.3191,
"step": 2743
},
{
"epoch": 0.4876271713536808,
"grad_norm": 0.4564426150795033,
"learning_rate": 2.1104916814458746e-05,
"loss": 1.3378,
"step": 2744
},
{
"epoch": 0.4878048780487805,
"grad_norm": 0.35223748078857897,
"learning_rate": 2.1093653462440208e-05,
"loss": 1.3203,
"step": 2745
},
{
"epoch": 0.48798258474388023,
"grad_norm": 0.36425817623068074,
"learning_rate": 2.1082389762509928e-05,
"loss": 1.3639,
"step": 2746
},
{
"epoch": 0.48816029143898,
"grad_norm": 0.3531712139147734,
"learning_rate": 2.107112571825112e-05,
"loss": 1.3041,
"step": 2747
},
{
"epoch": 0.4883379981340797,
"grad_norm": 0.3766408899431714,
"learning_rate": 2.1059861333247063e-05,
"loss": 1.3726,
"step": 2748
},
{
"epoch": 0.48851570482917944,
"grad_norm": 0.35928138909638146,
"learning_rate": 2.1048596611081192e-05,
"loss": 1.3273,
"step": 2749
},
{
"epoch": 0.4886934115242792,
"grad_norm": 0.37305158280413625,
"learning_rate": 2.103733155533702e-05,
"loss": 1.3744,
"step": 2750
},
{
"epoch": 0.4888711182193789,
"grad_norm": 0.36493846042791933,
"learning_rate": 2.1026066169598174e-05,
"loss": 1.3787,
"step": 2751
},
{
"epoch": 0.48904882491447865,
"grad_norm": 0.3613167406333977,
"learning_rate": 2.1014800457448384e-05,
"loss": 1.3499,
"step": 2752
},
{
"epoch": 0.4892265316095784,
"grad_norm": 0.3609273185205677,
"learning_rate": 2.1003534422471475e-05,
"loss": 1.3479,
"step": 2753
},
{
"epoch": 0.4894042383046781,
"grad_norm": 0.35642644725642936,
"learning_rate": 2.0992268068251406e-05,
"loss": 1.3244,
"step": 2754
},
{
"epoch": 0.48958194499977786,
"grad_norm": 0.3644070632950628,
"learning_rate": 2.09810013983722e-05,
"loss": 1.3729,
"step": 2755
},
{
"epoch": 0.4897596516948776,
"grad_norm": 0.3640488442246515,
"learning_rate": 2.0969734416417995e-05,
"loss": 1.4008,
"step": 2756
},
{
"epoch": 0.4899373583899773,
"grad_norm": 0.37332911126965146,
"learning_rate": 2.095846712597304e-05,
"loss": 1.3813,
"step": 2757
},
{
"epoch": 0.4901150650850771,
"grad_norm": 0.3571142567393805,
"learning_rate": 2.094719953062167e-05,
"loss": 1.3482,
"step": 2758
},
{
"epoch": 0.49029277178017683,
"grad_norm": 0.37103299954625935,
"learning_rate": 2.0935931633948313e-05,
"loss": 1.3116,
"step": 2759
},
{
"epoch": 0.4904704784752766,
"grad_norm": 0.37733149055740745,
"learning_rate": 2.09246634395375e-05,
"loss": 1.3502,
"step": 2760
},
{
"epoch": 0.4906481851703763,
"grad_norm": 0.36185900854796244,
"learning_rate": 2.0913394950973863e-05,
"loss": 1.3448,
"step": 2761
},
{
"epoch": 0.49082589186547604,
"grad_norm": 0.37692248104914244,
"learning_rate": 2.0902126171842113e-05,
"loss": 1.3729,
"step": 2762
},
{
"epoch": 0.4910035985605758,
"grad_norm": 0.35609722922186443,
"learning_rate": 2.0890857105727065e-05,
"loss": 1.34,
"step": 2763
},
{
"epoch": 0.4911813052556755,
"grad_norm": 0.36660400286419315,
"learning_rate": 2.087958775621361e-05,
"loss": 1.3585,
"step": 2764
},
{
"epoch": 0.49135901195077525,
"grad_norm": 0.3621427752437197,
"learning_rate": 2.0868318126886753e-05,
"loss": 1.3159,
"step": 2765
},
{
"epoch": 0.491536718645875,
"grad_norm": 0.3626978069664154,
"learning_rate": 2.085704822133158e-05,
"loss": 1.3437,
"step": 2766
},
{
"epoch": 0.4917144253409747,
"grad_norm": 0.36165041592634645,
"learning_rate": 2.0845778043133235e-05,
"loss": 1.3183,
"step": 2767
},
{
"epoch": 0.49189213203607446,
"grad_norm": 0.3667628473974172,
"learning_rate": 2.0834507595876997e-05,
"loss": 1.3616,
"step": 2768
},
{
"epoch": 0.4920698387311742,
"grad_norm": 0.35344555001890005,
"learning_rate": 2.0823236883148195e-05,
"loss": 1.3396,
"step": 2769
},
{
"epoch": 0.4922475454262739,
"grad_norm": 0.3510174598606899,
"learning_rate": 2.0811965908532263e-05,
"loss": 1.3418,
"step": 2770
},
{
"epoch": 0.49242525212137367,
"grad_norm": 0.3579370895602885,
"learning_rate": 2.08006946756147e-05,
"loss": 1.3374,
"step": 2771
},
{
"epoch": 0.4926029588164734,
"grad_norm": 0.35098037760562983,
"learning_rate": 2.0789423187981108e-05,
"loss": 1.3563,
"step": 2772
},
{
"epoch": 0.4927806655115731,
"grad_norm": 0.35759418255366543,
"learning_rate": 2.0778151449217157e-05,
"loss": 1.3528,
"step": 2773
},
{
"epoch": 0.4929583722066729,
"grad_norm": 0.3531344126411576,
"learning_rate": 2.0766879462908593e-05,
"loss": 1.3288,
"step": 2774
},
{
"epoch": 0.49313607890177263,
"grad_norm": 0.35177715300961,
"learning_rate": 2.0755607232641252e-05,
"loss": 1.3148,
"step": 2775
},
{
"epoch": 0.4933137855968724,
"grad_norm": 0.3567003459409521,
"learning_rate": 2.074433476200103e-05,
"loss": 1.3265,
"step": 2776
},
{
"epoch": 0.4934914922919721,
"grad_norm": 0.3629866588815171,
"learning_rate": 2.0733062054573936e-05,
"loss": 1.3679,
"step": 2777
},
{
"epoch": 0.49366919898707184,
"grad_norm": 0.34455253753497683,
"learning_rate": 2.0721789113946008e-05,
"loss": 1.3264,
"step": 2778
},
{
"epoch": 0.4938469056821716,
"grad_norm": 0.3555532899783648,
"learning_rate": 2.071051594370339e-05,
"loss": 1.3138,
"step": 2779
},
{
"epoch": 0.4940246123772713,
"grad_norm": 0.3682937557228829,
"learning_rate": 2.0699242547432293e-05,
"loss": 1.3705,
"step": 2780
},
{
"epoch": 0.49420231907237105,
"grad_norm": 0.3619964546653071,
"learning_rate": 2.068796892871899e-05,
"loss": 1.3431,
"step": 2781
},
{
"epoch": 0.4943800257674708,
"grad_norm": 0.36541699859649135,
"learning_rate": 2.0676695091149833e-05,
"loss": 1.3497,
"step": 2782
},
{
"epoch": 0.4945577324625705,
"grad_norm": 0.37068345212365994,
"learning_rate": 2.0665421038311234e-05,
"loss": 1.3456,
"step": 2783
},
{
"epoch": 0.49473543915767026,
"grad_norm": 0.38745504013089405,
"learning_rate": 2.0654146773789705e-05,
"loss": 1.3511,
"step": 2784
},
{
"epoch": 0.49491314585277,
"grad_norm": 0.3597505131998641,
"learning_rate": 2.0642872301171772e-05,
"loss": 1.3433,
"step": 2785
},
{
"epoch": 0.4950908525478697,
"grad_norm": 0.3701598765174156,
"learning_rate": 2.0631597624044076e-05,
"loss": 1.36,
"step": 2786
},
{
"epoch": 0.4952685592429695,
"grad_norm": 0.3750446062922082,
"learning_rate": 2.0620322745993294e-05,
"loss": 1.3494,
"step": 2787
},
{
"epoch": 0.49544626593806923,
"grad_norm": 0.36681489210873613,
"learning_rate": 2.0609047670606187e-05,
"loss": 1.3534,
"step": 2788
},
{
"epoch": 0.49562397263316893,
"grad_norm": 0.3572674533847875,
"learning_rate": 2.059777240146956e-05,
"loss": 1.3516,
"step": 2789
},
{
"epoch": 0.4958016793282687,
"grad_norm": 0.36186684305571715,
"learning_rate": 2.0586496942170284e-05,
"loss": 1.3675,
"step": 2790
},
{
"epoch": 0.49597938602336844,
"grad_norm": 0.3470851149557133,
"learning_rate": 2.0575221296295306e-05,
"loss": 1.333,
"step": 2791
},
{
"epoch": 0.4961570927184682,
"grad_norm": 0.3591370385584767,
"learning_rate": 2.0563945467431616e-05,
"loss": 1.3444,
"step": 2792
},
{
"epoch": 0.4963347994135679,
"grad_norm": 0.35625601878515634,
"learning_rate": 2.055266945916627e-05,
"loss": 1.3425,
"step": 2793
},
{
"epoch": 0.49651250610866765,
"grad_norm": 0.36206029792873917,
"learning_rate": 2.0541393275086374e-05,
"loss": 1.3968,
"step": 2794
},
{
"epoch": 0.4966902128037674,
"grad_norm": 0.3574031620725376,
"learning_rate": 2.0530116918779097e-05,
"loss": 1.354,
"step": 2795
},
{
"epoch": 0.4968679194988671,
"grad_norm": 0.3662154484296931,
"learning_rate": 2.0518840393831655e-05,
"loss": 1.3532,
"step": 2796
},
{
"epoch": 0.49704562619396686,
"grad_norm": 0.3594205479851579,
"learning_rate": 2.0507563703831327e-05,
"loss": 1.3056,
"step": 2797
},
{
"epoch": 0.4972233328890666,
"grad_norm": 0.3735308931598829,
"learning_rate": 2.049628685236544e-05,
"loss": 1.3423,
"step": 2798
},
{
"epoch": 0.4974010395841663,
"grad_norm": 0.3649770648919041,
"learning_rate": 2.0485009843021375e-05,
"loss": 1.3577,
"step": 2799
},
{
"epoch": 0.49757874627926607,
"grad_norm": 0.3770256530833716,
"learning_rate": 2.0473732679386558e-05,
"loss": 1.3773,
"step": 2800
},
{
"epoch": 0.4977564529743658,
"grad_norm": 0.4041104609142492,
"learning_rate": 2.0462455365048462e-05,
"loss": 1.3743,
"step": 2801
},
{
"epoch": 0.4979341596694655,
"grad_norm": 0.36105967451707277,
"learning_rate": 2.0451177903594618e-05,
"loss": 1.3868,
"step": 2802
},
{
"epoch": 0.4981118663645653,
"grad_norm": 0.37786484125317515,
"learning_rate": 2.0439900298612606e-05,
"loss": 1.3572,
"step": 2803
},
{
"epoch": 0.49828957305966504,
"grad_norm": 0.37014516386977825,
"learning_rate": 2.0428622553690028e-05,
"loss": 1.3694,
"step": 2804
},
{
"epoch": 0.49846727975476474,
"grad_norm": 0.34916691013213275,
"learning_rate": 2.041734467241456e-05,
"loss": 1.3376,
"step": 2805
},
{
"epoch": 0.4986449864498645,
"grad_norm": 0.35857544028852656,
"learning_rate": 2.0406066658373897e-05,
"loss": 1.3514,
"step": 2806
},
{
"epoch": 0.49882269314496425,
"grad_norm": 0.3527759854097344,
"learning_rate": 2.0394788515155803e-05,
"loss": 1.2705,
"step": 2807
},
{
"epoch": 0.499000399840064,
"grad_norm": 0.3598420608671938,
"learning_rate": 2.038351024634805e-05,
"loss": 1.3819,
"step": 2808
},
{
"epoch": 0.4991781065351637,
"grad_norm": 0.3578966259376419,
"learning_rate": 2.0372231855538475e-05,
"loss": 1.3358,
"step": 2809
},
{
"epoch": 0.49935581323026346,
"grad_norm": 0.35172270780014814,
"learning_rate": 2.0360953346314952e-05,
"loss": 1.3472,
"step": 2810
},
{
"epoch": 0.4995335199253632,
"grad_norm": 0.3626406061330495,
"learning_rate": 2.034967472226538e-05,
"loss": 1.3434,
"step": 2811
},
{
"epoch": 0.4997112266204629,
"grad_norm": 0.3515804697540496,
"learning_rate": 2.0338395986977703e-05,
"loss": 1.3185,
"step": 2812
},
{
"epoch": 0.49988893331556267,
"grad_norm": 0.35056760602075804,
"learning_rate": 2.0327117144039895e-05,
"loss": 1.3161,
"step": 2813
},
{
"epoch": 0.5000666400106624,
"grad_norm": 0.3634975431865219,
"learning_rate": 2.0315838197039976e-05,
"loss": 1.3616,
"step": 2814
},
{
"epoch": 0.5002443467057621,
"grad_norm": 0.3558160832827394,
"learning_rate": 2.030455914956599e-05,
"loss": 1.3533,
"step": 2815
},
{
"epoch": 0.5004220534008619,
"grad_norm": 0.35020432648918964,
"learning_rate": 2.0293280005206003e-05,
"loss": 1.3227,
"step": 2816
},
{
"epoch": 0.5005997600959616,
"grad_norm": 0.3520780012350196,
"learning_rate": 2.0282000767548134e-05,
"loss": 1.3349,
"step": 2817
},
{
"epoch": 0.5007774667910614,
"grad_norm": 0.3619806435860585,
"learning_rate": 2.027072144018052e-05,
"loss": 1.369,
"step": 2818
},
{
"epoch": 0.5009551734861611,
"grad_norm": 0.35577411944977455,
"learning_rate": 2.0259442026691322e-05,
"loss": 1.3242,
"step": 2819
},
{
"epoch": 0.5011328801812608,
"grad_norm": 0.35190298295845057,
"learning_rate": 2.0248162530668733e-05,
"loss": 1.3358,
"step": 2820
},
{
"epoch": 0.5013105868763605,
"grad_norm": 0.3537088318944128,
"learning_rate": 2.0236882955700983e-05,
"loss": 1.3318,
"step": 2821
},
{
"epoch": 0.5014882935714603,
"grad_norm": 0.36756360111221986,
"learning_rate": 2.0225603305376313e-05,
"loss": 1.3882,
"step": 2822
},
{
"epoch": 0.50166600026656,
"grad_norm": 0.4093022353186475,
"learning_rate": 2.0214323583282978e-05,
"loss": 1.3498,
"step": 2823
},
{
"epoch": 0.5018437069616598,
"grad_norm": 0.35888442097835466,
"learning_rate": 2.0203043793009285e-05,
"loss": 1.3359,
"step": 2824
},
{
"epoch": 0.5020214136567596,
"grad_norm": 0.37545772976363817,
"learning_rate": 2.0191763938143546e-05,
"loss": 1.3404,
"step": 2825
},
{
"epoch": 0.5021991203518592,
"grad_norm": 0.3638957753136997,
"learning_rate": 2.0180484022274087e-05,
"loss": 1.3739,
"step": 2826
},
{
"epoch": 0.502376827046959,
"grad_norm": 0.36076021335259456,
"learning_rate": 2.016920404898927e-05,
"loss": 1.3262,
"step": 2827
},
{
"epoch": 0.5025545337420587,
"grad_norm": 0.3568024412890916,
"learning_rate": 2.0157924021877463e-05,
"loss": 1.3742,
"step": 2828
},
{
"epoch": 0.5027322404371585,
"grad_norm": 0.35488574786615973,
"learning_rate": 2.014664394452705e-05,
"loss": 1.3369,
"step": 2829
},
{
"epoch": 0.5029099471322582,
"grad_norm": 0.3622310025113257,
"learning_rate": 2.0135363820526446e-05,
"loss": 1.3012,
"step": 2830
},
{
"epoch": 0.503087653827358,
"grad_norm": 0.3576470828133915,
"learning_rate": 2.0124083653464065e-05,
"loss": 1.3158,
"step": 2831
},
{
"epoch": 0.5032653605224577,
"grad_norm": 0.36230481963891736,
"learning_rate": 2.0112803446928332e-05,
"loss": 1.372,
"step": 2832
},
{
"epoch": 0.5034430672175574,
"grad_norm": 0.36382786915407267,
"learning_rate": 2.0101523204507716e-05,
"loss": 1.3547,
"step": 2833
},
{
"epoch": 0.5036207739126571,
"grad_norm": 0.3651286695585329,
"learning_rate": 2.009024292979065e-05,
"loss": 1.3457,
"step": 2834
},
{
"epoch": 0.5037984806077569,
"grad_norm": 0.363431104675626,
"learning_rate": 2.0078962626365613e-05,
"loss": 1.3507,
"step": 2835
},
{
"epoch": 0.5039761873028566,
"grad_norm": 0.36847720638816134,
"learning_rate": 2.006768229782108e-05,
"loss": 1.3518,
"step": 2836
},
{
"epoch": 0.5041538939979564,
"grad_norm": 0.3698164924914968,
"learning_rate": 2.0056401947745533e-05,
"loss": 1.3464,
"step": 2837
},
{
"epoch": 0.5043316006930562,
"grad_norm": 0.3607331053456653,
"learning_rate": 2.0045121579727465e-05,
"loss": 1.3425,
"step": 2838
},
{
"epoch": 0.5045093073881558,
"grad_norm": 0.35519255563921104,
"learning_rate": 2.0033841197355373e-05,
"loss": 1.2976,
"step": 2839
},
{
"epoch": 0.5046870140832556,
"grad_norm": 0.3557740565237606,
"learning_rate": 2.0022560804217767e-05,
"loss": 1.3281,
"step": 2840
},
{
"epoch": 0.5048647207783553,
"grad_norm": 0.3545651346890717,
"learning_rate": 2.001128040390314e-05,
"loss": 1.3452,
"step": 2841
},
{
"epoch": 0.5050424274734551,
"grad_norm": 0.3605508469169887,
"learning_rate": 2e-05,
"loss": 1.337,
"step": 2842
},
{
"epoch": 0.5052201341685548,
"grad_norm": 0.3703084340193821,
"learning_rate": 1.9988719596096868e-05,
"loss": 1.3815,
"step": 2843
},
{
"epoch": 0.5053978408636546,
"grad_norm": 0.3787621593152529,
"learning_rate": 1.9977439195782243e-05,
"loss": 1.3561,
"step": 2844
},
{
"epoch": 0.5055755475587542,
"grad_norm": 0.36818002630657176,
"learning_rate": 1.996615880264463e-05,
"loss": 1.3482,
"step": 2845
},
{
"epoch": 0.505753254253854,
"grad_norm": 0.36777348455104447,
"learning_rate": 1.9954878420272538e-05,
"loss": 1.3525,
"step": 2846
},
{
"epoch": 0.5059309609489537,
"grad_norm": 0.3541087304225588,
"learning_rate": 1.9943598052254473e-05,
"loss": 1.3419,
"step": 2847
},
{
"epoch": 0.5061086676440535,
"grad_norm": 0.3651887508956978,
"learning_rate": 1.9932317702178928e-05,
"loss": 1.3587,
"step": 2848
},
{
"epoch": 0.5062863743391532,
"grad_norm": 0.36451032784277587,
"learning_rate": 1.99210373736344e-05,
"loss": 1.3756,
"step": 2849
},
{
"epoch": 0.506464081034253,
"grad_norm": 0.3603103333777577,
"learning_rate": 1.9909757070209354e-05,
"loss": 1.3289,
"step": 2850
},
{
"epoch": 0.5066417877293528,
"grad_norm": 0.3599355488064004,
"learning_rate": 1.989847679549229e-05,
"loss": 1.3623,
"step": 2851
},
{
"epoch": 0.5068194944244524,
"grad_norm": 0.35666043643053913,
"learning_rate": 1.988719655307167e-05,
"loss": 1.3554,
"step": 2852
},
{
"epoch": 0.5069972011195522,
"grad_norm": 0.3675415272107359,
"learning_rate": 1.9875916346535945e-05,
"loss": 1.338,
"step": 2853
},
{
"epoch": 0.5071749078146519,
"grad_norm": 0.3859448516119809,
"learning_rate": 1.9864636179473557e-05,
"loss": 1.4065,
"step": 2854
},
{
"epoch": 0.5073526145097517,
"grad_norm": 0.39254084361545905,
"learning_rate": 1.9853356055472955e-05,
"loss": 1.3582,
"step": 2855
},
{
"epoch": 0.5075303212048514,
"grad_norm": 0.3673711773545292,
"learning_rate": 1.9842075978122547e-05,
"loss": 1.3636,
"step": 2856
},
{
"epoch": 0.5077080278999512,
"grad_norm": 0.36321533194023914,
"learning_rate": 1.9830795951010737e-05,
"loss": 1.3197,
"step": 2857
},
{
"epoch": 0.5078857345950508,
"grad_norm": 0.3632833787762882,
"learning_rate": 1.981951597772592e-05,
"loss": 1.3472,
"step": 2858
},
{
"epoch": 0.5080634412901506,
"grad_norm": 0.366536545958968,
"learning_rate": 1.980823606185646e-05,
"loss": 1.3505,
"step": 2859
},
{
"epoch": 0.5082411479852503,
"grad_norm": 0.36865122145628043,
"learning_rate": 1.9796956206990722e-05,
"loss": 1.3527,
"step": 2860
},
{
"epoch": 0.5084188546803501,
"grad_norm": 0.3640445429233462,
"learning_rate": 1.978567641671703e-05,
"loss": 1.328,
"step": 2861
},
{
"epoch": 0.5085965613754498,
"grad_norm": 0.35951971391060233,
"learning_rate": 1.9774396694623697e-05,
"loss": 1.3545,
"step": 2862
},
{
"epoch": 0.5087742680705496,
"grad_norm": 0.36094109268922947,
"learning_rate": 1.9763117044299024e-05,
"loss": 1.3431,
"step": 2863
},
{
"epoch": 0.5089519747656494,
"grad_norm": 0.36443151906200194,
"learning_rate": 1.9751837469331267e-05,
"loss": 1.3518,
"step": 2864
},
{
"epoch": 0.509129681460749,
"grad_norm": 0.3582642190341808,
"learning_rate": 1.9740557973308684e-05,
"loss": 1.3419,
"step": 2865
},
{
"epoch": 0.5093073881558488,
"grad_norm": 0.3672586338565434,
"learning_rate": 1.9729278559819488e-05,
"loss": 1.3395,
"step": 2866
},
{
"epoch": 0.5094850948509485,
"grad_norm": 0.4192387449140547,
"learning_rate": 1.9717999232451876e-05,
"loss": 1.3423,
"step": 2867
},
{
"epoch": 0.5096628015460483,
"grad_norm": 0.35685950488953844,
"learning_rate": 1.9706719994794e-05,
"loss": 1.2923,
"step": 2868
},
{
"epoch": 0.509840508241148,
"grad_norm": 0.36047929504288423,
"learning_rate": 1.969544085043402e-05,
"loss": 1.3105,
"step": 2869
},
{
"epoch": 0.5100182149362478,
"grad_norm": 0.36211312092600784,
"learning_rate": 1.9684161802960028e-05,
"loss": 1.3302,
"step": 2870
},
{
"epoch": 0.5101959216313474,
"grad_norm": 0.3642067069684092,
"learning_rate": 1.9672882855960112e-05,
"loss": 1.3518,
"step": 2871
},
{
"epoch": 0.5103736283264472,
"grad_norm": 0.3790365226323441,
"learning_rate": 1.9661604013022307e-05,
"loss": 1.3755,
"step": 2872
},
{
"epoch": 0.5105513350215469,
"grad_norm": 0.3460647589020893,
"learning_rate": 1.965032527773462e-05,
"loss": 1.3331,
"step": 2873
},
{
"epoch": 0.5107290417166467,
"grad_norm": 0.3653377849193898,
"learning_rate": 1.9639046653685055e-05,
"loss": 1.387,
"step": 2874
},
{
"epoch": 0.5109067484117464,
"grad_norm": 0.3578003372830014,
"learning_rate": 1.962776814446153e-05,
"loss": 1.3379,
"step": 2875
},
{
"epoch": 0.5110844551068462,
"grad_norm": 0.36781398039717217,
"learning_rate": 1.9616489753651957e-05,
"loss": 1.3708,
"step": 2876
},
{
"epoch": 0.5112621618019458,
"grad_norm": 0.36564762968960335,
"learning_rate": 1.960521148484421e-05,
"loss": 1.3743,
"step": 2877
},
{
"epoch": 0.5114398684970456,
"grad_norm": 0.35499060556149525,
"learning_rate": 1.9593933341626107e-05,
"loss": 1.3451,
"step": 2878
},
{
"epoch": 0.5116175751921453,
"grad_norm": 0.35279036845476486,
"learning_rate": 1.9582655327585447e-05,
"loss": 1.3438,
"step": 2879
},
{
"epoch": 0.5117952818872451,
"grad_norm": 0.3756676667430106,
"learning_rate": 1.957137744630998e-05,
"loss": 1.3799,
"step": 2880
},
{
"epoch": 0.5119729885823449,
"grad_norm": 0.36455971074787796,
"learning_rate": 1.9560099701387404e-05,
"loss": 1.3553,
"step": 2881
},
{
"epoch": 0.5121506952774446,
"grad_norm": 0.36587302263997973,
"learning_rate": 1.9548822096405382e-05,
"loss": 1.3125,
"step": 2882
},
{
"epoch": 0.5123284019725444,
"grad_norm": 0.3666454463033673,
"learning_rate": 1.953754463495154e-05,
"loss": 1.3952,
"step": 2883
},
{
"epoch": 0.512506108667644,
"grad_norm": 0.3653675310479251,
"learning_rate": 1.952626732061345e-05,
"loss": 1.299,
"step": 2884
},
{
"epoch": 0.5126838153627438,
"grad_norm": 0.36346652636051296,
"learning_rate": 1.9514990156978632e-05,
"loss": 1.3658,
"step": 2885
},
{
"epoch": 0.5128615220578435,
"grad_norm": 0.35787348819454073,
"learning_rate": 1.950371314763457e-05,
"loss": 1.344,
"step": 2886
},
{
"epoch": 0.5130392287529433,
"grad_norm": 0.38885333863926036,
"learning_rate": 1.9492436296168677e-05,
"loss": 1.3421,
"step": 2887
},
{
"epoch": 0.513216935448043,
"grad_norm": 0.3548343070030667,
"learning_rate": 1.9481159606168348e-05,
"loss": 1.3087,
"step": 2888
},
{
"epoch": 0.5133946421431428,
"grad_norm": 0.3725963754691008,
"learning_rate": 1.946988308122091e-05,
"loss": 1.4117,
"step": 2889
},
{
"epoch": 0.5135723488382424,
"grad_norm": 0.4177376856614558,
"learning_rate": 1.9458606724913636e-05,
"loss": 1.3302,
"step": 2890
},
{
"epoch": 0.5137500555333422,
"grad_norm": 0.3553195555421267,
"learning_rate": 1.944733054083374e-05,
"loss": 1.3419,
"step": 2891
},
{
"epoch": 0.5139277622284419,
"grad_norm": 0.3612262945013363,
"learning_rate": 1.9436054532568384e-05,
"loss": 1.363,
"step": 2892
},
{
"epoch": 0.5141054689235417,
"grad_norm": 0.34986337370641657,
"learning_rate": 1.9424778703704697e-05,
"loss": 1.3198,
"step": 2893
},
{
"epoch": 0.5142831756186415,
"grad_norm": 0.362574151798206,
"learning_rate": 1.9413503057829722e-05,
"loss": 1.31,
"step": 2894
},
{
"epoch": 0.5144608823137412,
"grad_norm": 0.35634533355215053,
"learning_rate": 1.940222759853045e-05,
"loss": 1.3109,
"step": 2895
},
{
"epoch": 0.514638589008841,
"grad_norm": 0.3520413960052993,
"learning_rate": 1.939095232939382e-05,
"loss": 1.3206,
"step": 2896
},
{
"epoch": 0.5148162957039406,
"grad_norm": 0.3747616295514053,
"learning_rate": 1.937967725400671e-05,
"loss": 1.3638,
"step": 2897
},
{
"epoch": 0.5149940023990404,
"grad_norm": 0.5688692722354263,
"learning_rate": 1.936840237595593e-05,
"loss": 1.3189,
"step": 2898
},
{
"epoch": 0.5151717090941401,
"grad_norm": 0.3624855151520999,
"learning_rate": 1.935712769882823e-05,
"loss": 1.3674,
"step": 2899
},
{
"epoch": 0.5153494157892399,
"grad_norm": 0.35890446040133,
"learning_rate": 1.9345853226210308e-05,
"loss": 1.3235,
"step": 2900
},
{
"epoch": 0.5155271224843396,
"grad_norm": 0.3607004012800823,
"learning_rate": 1.9334578961688763e-05,
"loss": 1.3417,
"step": 2901
},
{
"epoch": 0.5157048291794394,
"grad_norm": 0.4616624764461598,
"learning_rate": 1.9323304908850173e-05,
"loss": 1.3129,
"step": 2902
},
{
"epoch": 0.515882535874539,
"grad_norm": 0.3559897931715917,
"learning_rate": 1.9312031071281013e-05,
"loss": 1.3508,
"step": 2903
},
{
"epoch": 0.5160602425696388,
"grad_norm": 0.36548416170396786,
"learning_rate": 1.930075745256771e-05,
"loss": 1.3001,
"step": 2904
},
{
"epoch": 0.5162379492647385,
"grad_norm": 0.3618579337285716,
"learning_rate": 1.9289484056296617e-05,
"loss": 1.3528,
"step": 2905
},
{
"epoch": 0.5164156559598383,
"grad_norm": 0.36604656326028606,
"learning_rate": 1.9278210886053995e-05,
"loss": 1.3851,
"step": 2906
},
{
"epoch": 0.516593362654938,
"grad_norm": 0.3650119506862482,
"learning_rate": 1.926693794542607e-05,
"loss": 1.3677,
"step": 2907
},
{
"epoch": 0.5167710693500378,
"grad_norm": 0.35582472216870936,
"learning_rate": 1.9255665237998976e-05,
"loss": 1.3382,
"step": 2908
},
{
"epoch": 0.5169487760451374,
"grad_norm": 0.3555804929988836,
"learning_rate": 1.924439276735876e-05,
"loss": 1.3274,
"step": 2909
},
{
"epoch": 0.5171264827402372,
"grad_norm": 0.36581018768779455,
"learning_rate": 1.923312053709141e-05,
"loss": 1.3466,
"step": 2910
},
{
"epoch": 0.517304189435337,
"grad_norm": 0.3559123553300019,
"learning_rate": 1.9221848550782846e-05,
"loss": 1.3672,
"step": 2911
},
{
"epoch": 0.5174818961304367,
"grad_norm": 0.3561087221511621,
"learning_rate": 1.9210576812018895e-05,
"loss": 1.3572,
"step": 2912
},
{
"epoch": 0.5176596028255365,
"grad_norm": 0.3676517881213993,
"learning_rate": 1.9199305324385306e-05,
"loss": 1.3117,
"step": 2913
},
{
"epoch": 0.5178373095206362,
"grad_norm": 0.3512025912786583,
"learning_rate": 1.9188034091467747e-05,
"loss": 1.297,
"step": 2914
},
{
"epoch": 0.518015016215736,
"grad_norm": 0.3623519044155793,
"learning_rate": 1.9176763116851808e-05,
"loss": 1.3625,
"step": 2915
},
{
"epoch": 0.5181927229108356,
"grad_norm": 0.35783389036603636,
"learning_rate": 1.916549240412301e-05,
"loss": 1.3612,
"step": 2916
},
{
"epoch": 0.5183704296059354,
"grad_norm": 0.349359408341441,
"learning_rate": 1.915422195686677e-05,
"loss": 1.3154,
"step": 2917
},
{
"epoch": 0.5185481363010351,
"grad_norm": 0.35334523802932266,
"learning_rate": 1.9142951778668432e-05,
"loss": 1.3548,
"step": 2918
},
{
"epoch": 0.5187258429961349,
"grad_norm": 0.37640890385226705,
"learning_rate": 1.9131681873113254e-05,
"loss": 1.3688,
"step": 2919
},
{
"epoch": 0.5189035496912346,
"grad_norm": 0.35562745840968546,
"learning_rate": 1.9120412243786393e-05,
"loss": 1.331,
"step": 2920
},
{
"epoch": 0.5190812563863344,
"grad_norm": 0.3623982590068297,
"learning_rate": 1.910914289427294e-05,
"loss": 1.3651,
"step": 2921
},
{
"epoch": 0.519258963081434,
"grad_norm": 0.35850837670935226,
"learning_rate": 1.9097873828157894e-05,
"loss": 1.3425,
"step": 2922
},
{
"epoch": 0.5194366697765338,
"grad_norm": 0.3586252815162115,
"learning_rate": 1.9086605049026143e-05,
"loss": 1.3507,
"step": 2923
},
{
"epoch": 0.5196143764716336,
"grad_norm": 0.36792366348624045,
"learning_rate": 1.90753365604625e-05,
"loss": 1.3476,
"step": 2924
},
{
"epoch": 0.5197920831667333,
"grad_norm": 0.35857927759557046,
"learning_rate": 1.906406836605169e-05,
"loss": 1.3332,
"step": 2925
},
{
"epoch": 0.5199697898618331,
"grad_norm": 0.3534072877854019,
"learning_rate": 1.9052800469378336e-05,
"loss": 1.3224,
"step": 2926
},
{
"epoch": 0.5201474965569328,
"grad_norm": 0.36190487471757754,
"learning_rate": 1.9041532874026967e-05,
"loss": 1.361,
"step": 2927
},
{
"epoch": 0.5203252032520326,
"grad_norm": 0.3522938000906696,
"learning_rate": 1.903026558358201e-05,
"loss": 1.3066,
"step": 2928
},
{
"epoch": 0.5205029099471322,
"grad_norm": 0.35275002695605007,
"learning_rate": 1.9018998601627804e-05,
"loss": 1.3293,
"step": 2929
},
{
"epoch": 0.520680616642232,
"grad_norm": 0.35716746535788046,
"learning_rate": 1.9007731931748604e-05,
"loss": 1.3631,
"step": 2930
},
{
"epoch": 0.5208583233373317,
"grad_norm": 0.35733149699769795,
"learning_rate": 1.899646557752853e-05,
"loss": 1.3546,
"step": 2931
},
{
"epoch": 0.5210360300324315,
"grad_norm": 0.3499106490479668,
"learning_rate": 1.8985199542551626e-05,
"loss": 1.3219,
"step": 2932
},
{
"epoch": 0.5212137367275312,
"grad_norm": 0.36322913962107684,
"learning_rate": 1.8973933830401836e-05,
"loss": 1.3351,
"step": 2933
},
{
"epoch": 0.521391443422631,
"grad_norm": 0.3579017155971588,
"learning_rate": 1.8962668444662983e-05,
"loss": 1.3266,
"step": 2934
},
{
"epoch": 0.5215691501177306,
"grad_norm": 0.3509816375159486,
"learning_rate": 1.895140338891881e-05,
"loss": 1.3491,
"step": 2935
},
{
"epoch": 0.5217468568128304,
"grad_norm": 0.368376079576896,
"learning_rate": 1.8940138666752944e-05,
"loss": 1.3406,
"step": 2936
},
{
"epoch": 0.5219245635079302,
"grad_norm": 0.3536562933615042,
"learning_rate": 1.8928874281748894e-05,
"loss": 1.3117,
"step": 2937
},
{
"epoch": 0.5221022702030299,
"grad_norm": 0.3512987235282378,
"learning_rate": 1.8917610237490075e-05,
"loss": 1.3516,
"step": 2938
},
{
"epoch": 0.5222799768981297,
"grad_norm": 0.3646995595294746,
"learning_rate": 1.8906346537559802e-05,
"loss": 1.3257,
"step": 2939
},
{
"epoch": 0.5224576835932294,
"grad_norm": 0.3649987919801856,
"learning_rate": 1.8895083185541257e-05,
"loss": 1.3279,
"step": 2940
},
{
"epoch": 0.5226353902883291,
"grad_norm": 0.35300145612225364,
"learning_rate": 1.8883820185017537e-05,
"loss": 1.3241,
"step": 2941
},
{
"epoch": 0.5228130969834288,
"grad_norm": 0.3536557449738324,
"learning_rate": 1.88725575395716e-05,
"loss": 1.3468,
"step": 2942
},
{
"epoch": 0.5229908036785286,
"grad_norm": 0.3562664734891774,
"learning_rate": 1.8861295252786312e-05,
"loss": 1.3433,
"step": 2943
},
{
"epoch": 0.5231685103736283,
"grad_norm": 0.3603408016433828,
"learning_rate": 1.885003332824442e-05,
"loss": 1.3566,
"step": 2944
},
{
"epoch": 0.5233462170687281,
"grad_norm": 0.363925978391125,
"learning_rate": 1.8838771769528556e-05,
"loss": 1.3343,
"step": 2945
},
{
"epoch": 0.5235239237638278,
"grad_norm": 0.3548548147302873,
"learning_rate": 1.882751058022123e-05,
"loss": 1.3484,
"step": 2946
},
{
"epoch": 0.5237016304589276,
"grad_norm": 0.3568949336250508,
"learning_rate": 1.8816249763904838e-05,
"loss": 1.328,
"step": 2947
},
{
"epoch": 0.5238793371540272,
"grad_norm": 0.37373807167335776,
"learning_rate": 1.8804989324161644e-05,
"loss": 1.3295,
"step": 2948
},
{
"epoch": 0.524057043849127,
"grad_norm": 0.35985600632311593,
"learning_rate": 1.8793729264573836e-05,
"loss": 1.3555,
"step": 2949
},
{
"epoch": 0.5242347505442267,
"grad_norm": 0.3661676464318552,
"learning_rate": 1.878246958872343e-05,
"loss": 1.355,
"step": 2950
},
{
"epoch": 0.5244124572393265,
"grad_norm": 0.37189172105045504,
"learning_rate": 1.877121030019234e-05,
"loss": 1.3224,
"step": 2951
},
{
"epoch": 0.5245901639344263,
"grad_norm": 0.36064330501919944,
"learning_rate": 1.8759951402562362e-05,
"loss": 1.3695,
"step": 2952
},
{
"epoch": 0.524767870629526,
"grad_norm": 0.36192366386463803,
"learning_rate": 1.8748692899415166e-05,
"loss": 1.3302,
"step": 2953
},
{
"epoch": 0.5249455773246257,
"grad_norm": 0.36384411610698797,
"learning_rate": 1.873743479433229e-05,
"loss": 1.319,
"step": 2954
},
{
"epoch": 0.5251232840197254,
"grad_norm": 0.35551202124849285,
"learning_rate": 1.872617709089515e-05,
"loss": 1.3265,
"step": 2955
},
{
"epoch": 0.5253009907148252,
"grad_norm": 0.3629321629368281,
"learning_rate": 1.8714919792685034e-05,
"loss": 1.332,
"step": 2956
},
{
"epoch": 0.5254786974099249,
"grad_norm": 0.3671764460411999,
"learning_rate": 1.8703662903283092e-05,
"loss": 1.3571,
"step": 2957
},
{
"epoch": 0.5256564041050247,
"grad_norm": 0.36952090517941927,
"learning_rate": 1.8692406426270368e-05,
"loss": 1.3034,
"step": 2958
},
{
"epoch": 0.5258341108001244,
"grad_norm": 0.3554421862204248,
"learning_rate": 1.8681150365227745e-05,
"loss": 1.3492,
"step": 2959
},
{
"epoch": 0.5260118174952242,
"grad_norm": 0.36510056019301185,
"learning_rate": 1.8669894723735995e-05,
"loss": 1.3149,
"step": 2960
},
{
"epoch": 0.5261895241903238,
"grad_norm": 0.3701366813399121,
"learning_rate": 1.865863950537575e-05,
"loss": 1.3172,
"step": 2961
},
{
"epoch": 0.5263672308854236,
"grad_norm": 0.36550565409129565,
"learning_rate": 1.864738471372749e-05,
"loss": 1.3222,
"step": 2962
},
{
"epoch": 0.5265449375805233,
"grad_norm": 0.37001664437091586,
"learning_rate": 1.8636130352371603e-05,
"loss": 1.3508,
"step": 2963
},
{
"epoch": 0.5267226442756231,
"grad_norm": 0.3475367050337639,
"learning_rate": 1.8624876424888297e-05,
"loss": 1.3087,
"step": 2964
},
{
"epoch": 0.5269003509707229,
"grad_norm": 0.3551245262782015,
"learning_rate": 1.8613622934857664e-05,
"loss": 1.3154,
"step": 2965
},
{
"epoch": 0.5270780576658226,
"grad_norm": 0.3638534759328045,
"learning_rate": 1.860236988585964e-05,
"loss": 1.2953,
"step": 2966
},
{
"epoch": 0.5272557643609223,
"grad_norm": 0.36085447979622576,
"learning_rate": 1.859111728147404e-05,
"loss": 1.3573,
"step": 2967
},
{
"epoch": 0.527433471056022,
"grad_norm": 0.36185216872707787,
"learning_rate": 1.8579865125280536e-05,
"loss": 1.3452,
"step": 2968
},
{
"epoch": 0.5276111777511218,
"grad_norm": 0.36347667732096267,
"learning_rate": 1.8568613420858636e-05,
"loss": 1.3107,
"step": 2969
},
{
"epoch": 0.5277888844462215,
"grad_norm": 0.35183967590753756,
"learning_rate": 1.8557362171787727e-05,
"loss": 1.3254,
"step": 2970
},
{
"epoch": 0.5279665911413213,
"grad_norm": 0.3601618391966797,
"learning_rate": 1.8546111381647037e-05,
"loss": 1.3462,
"step": 2971
},
{
"epoch": 0.528144297836421,
"grad_norm": 0.35570667546527074,
"learning_rate": 1.853486105401566e-05,
"loss": 1.3281,
"step": 2972
},
{
"epoch": 0.5283220045315207,
"grad_norm": 0.3562226277853729,
"learning_rate": 1.852361119247254e-05,
"loss": 1.3407,
"step": 2973
},
{
"epoch": 0.5284997112266204,
"grad_norm": 0.3629270053341451,
"learning_rate": 1.8512361800596462e-05,
"loss": 1.333,
"step": 2974
},
{
"epoch": 0.5286774179217202,
"grad_norm": 0.3630284908991094,
"learning_rate": 1.850111288196607e-05,
"loss": 1.3862,
"step": 2975
},
{
"epoch": 0.5288551246168199,
"grad_norm": 0.3726164620110594,
"learning_rate": 1.8489864440159853e-05,
"loss": 1.3511,
"step": 2976
},
{
"epoch": 0.5290328313119197,
"grad_norm": 0.3594614880609556,
"learning_rate": 1.8478616478756164e-05,
"loss": 1.3605,
"step": 2977
},
{
"epoch": 0.5292105380070194,
"grad_norm": 0.3581251956783206,
"learning_rate": 1.8467369001333183e-05,
"loss": 1.3706,
"step": 2978
},
{
"epoch": 0.5293882447021192,
"grad_norm": 0.35339409747433875,
"learning_rate": 1.8456122011468946e-05,
"loss": 1.3567,
"step": 2979
},
{
"epoch": 0.5295659513972188,
"grad_norm": 0.3698712300793728,
"learning_rate": 1.8444875512741324e-05,
"loss": 1.3271,
"step": 2980
},
{
"epoch": 0.5297436580923186,
"grad_norm": 0.35387871348504213,
"learning_rate": 1.8433629508728054e-05,
"loss": 1.3303,
"step": 2981
},
{
"epoch": 0.5299213647874184,
"grad_norm": 0.35157733760375803,
"learning_rate": 1.8422384003006694e-05,
"loss": 1.3498,
"step": 2982
},
{
"epoch": 0.5300990714825181,
"grad_norm": 0.36807704652491297,
"learning_rate": 1.8411138999154648e-05,
"loss": 1.3656,
"step": 2983
},
{
"epoch": 0.5302767781776179,
"grad_norm": 0.3644340020278979,
"learning_rate": 1.8399894500749175e-05,
"loss": 1.3629,
"step": 2984
},
{
"epoch": 0.5304544848727176,
"grad_norm": 0.3616278541905301,
"learning_rate": 1.8388650511367335e-05,
"loss": 1.3613,
"step": 2985
},
{
"epoch": 0.5306321915678173,
"grad_norm": 0.35452826409508276,
"learning_rate": 1.837740703458608e-05,
"loss": 1.3317,
"step": 2986
},
{
"epoch": 0.530809898262917,
"grad_norm": 0.3623991604301386,
"learning_rate": 1.836616407398217e-05,
"loss": 1.3692,
"step": 2987
},
{
"epoch": 0.5309876049580168,
"grad_norm": 0.359206355859447,
"learning_rate": 1.8354921633132185e-05,
"loss": 1.3531,
"step": 2988
},
{
"epoch": 0.5311653116531165,
"grad_norm": 0.3483298630300049,
"learning_rate": 1.8343679715612568e-05,
"loss": 1.3365,
"step": 2989
},
{
"epoch": 0.5313430183482163,
"grad_norm": 0.3756244406004804,
"learning_rate": 1.8332438324999577e-05,
"loss": 1.3006,
"step": 2990
},
{
"epoch": 0.531520725043316,
"grad_norm": 0.3539411212473202,
"learning_rate": 1.832119746486932e-05,
"loss": 1.319,
"step": 2991
},
{
"epoch": 0.5316984317384158,
"grad_norm": 0.359992781470055,
"learning_rate": 1.8309957138797717e-05,
"loss": 1.3053,
"step": 2992
},
{
"epoch": 0.5318761384335154,
"grad_norm": 0.35629066682146243,
"learning_rate": 1.8298717350360533e-05,
"loss": 1.3451,
"step": 2993
},
{
"epoch": 0.5320538451286152,
"grad_norm": 0.35342551711102815,
"learning_rate": 1.8287478103133353e-05,
"loss": 1.3605,
"step": 2994
},
{
"epoch": 0.532231551823715,
"grad_norm": 0.3572835124162402,
"learning_rate": 1.827623940069159e-05,
"loss": 1.3485,
"step": 2995
},
{
"epoch": 0.5324092585188147,
"grad_norm": 0.36098039965347856,
"learning_rate": 1.82650012466105e-05,
"loss": 1.3619,
"step": 2996
},
{
"epoch": 0.5325869652139145,
"grad_norm": 0.5014364575450393,
"learning_rate": 1.8253763644465133e-05,
"loss": 1.3418,
"step": 2997
},
{
"epoch": 0.5327646719090142,
"grad_norm": 0.35213821430867126,
"learning_rate": 1.8242526597830397e-05,
"loss": 1.3402,
"step": 2998
},
{
"epoch": 0.5329423786041139,
"grad_norm": 0.3816278526641126,
"learning_rate": 1.823129011028099e-05,
"loss": 1.3457,
"step": 2999
},
{
"epoch": 0.5331200852992136,
"grad_norm": 0.3637638394424721,
"learning_rate": 1.8220054185391473e-05,
"loss": 1.3655,
"step": 3000
},
{
"epoch": 0.5332977919943134,
"grad_norm": 0.3496823303625333,
"learning_rate": 1.8208818826736188e-05,
"loss": 1.2962,
"step": 3001
},
{
"epoch": 0.5334754986894131,
"grad_norm": 0.3465990610637408,
"learning_rate": 1.8197584037889325e-05,
"loss": 1.2788,
"step": 3002
},
{
"epoch": 0.5336532053845129,
"grad_norm": 0.3547409009221713,
"learning_rate": 1.818634982242487e-05,
"loss": 1.3624,
"step": 3003
},
{
"epoch": 0.5338309120796126,
"grad_norm": 0.35995599573033155,
"learning_rate": 1.8175116183916635e-05,
"loss": 1.3464,
"step": 3004
},
{
"epoch": 0.5340086187747123,
"grad_norm": 0.3534648527493373,
"learning_rate": 1.8163883125938272e-05,
"loss": 1.2971,
"step": 3005
},
{
"epoch": 0.534186325469812,
"grad_norm": 0.36308980615771086,
"learning_rate": 1.8152650652063218e-05,
"loss": 1.3468,
"step": 3006
},
{
"epoch": 0.5343640321649118,
"grad_norm": 0.3573183032571732,
"learning_rate": 1.8141418765864726e-05,
"loss": 1.3253,
"step": 3007
},
{
"epoch": 0.5345417388600116,
"grad_norm": 0.35109052070713,
"learning_rate": 1.813018747091587e-05,
"loss": 1.3018,
"step": 3008
},
{
"epoch": 0.5347194455551113,
"grad_norm": 0.34597883541704333,
"learning_rate": 1.811895677078956e-05,
"loss": 1.3173,
"step": 3009
},
{
"epoch": 0.5348971522502111,
"grad_norm": 0.35239071661063703,
"learning_rate": 1.8107726669058468e-05,
"loss": 1.3113,
"step": 3010
},
{
"epoch": 0.5350748589453108,
"grad_norm": 0.37348919474797276,
"learning_rate": 1.8096497169295107e-05,
"loss": 1.3523,
"step": 3011
},
{
"epoch": 0.5352525656404105,
"grad_norm": 0.35398862186577446,
"learning_rate": 1.8085268275071795e-05,
"loss": 1.3494,
"step": 3012
},
{
"epoch": 0.5354302723355102,
"grad_norm": 0.36198861676237026,
"learning_rate": 1.8074039989960647e-05,
"loss": 1.3531,
"step": 3013
},
{
"epoch": 0.53560797903061,
"grad_norm": 0.3543355509784464,
"learning_rate": 1.8062812317533606e-05,
"loss": 1.2781,
"step": 3014
},
{
"epoch": 0.5357856857257097,
"grad_norm": 0.34477783817711466,
"learning_rate": 1.805158526136239e-05,
"loss": 1.2989,
"step": 3015
},
{
"epoch": 0.5359633924208095,
"grad_norm": 0.34663354588691064,
"learning_rate": 1.804035882501855e-05,
"loss": 1.2895,
"step": 3016
},
{
"epoch": 0.5361410991159092,
"grad_norm": 0.3657167383906152,
"learning_rate": 1.802913301207342e-05,
"loss": 1.346,
"step": 3017
},
{
"epoch": 0.5363188058110089,
"grad_norm": 0.3495902173220004,
"learning_rate": 1.8017907826098137e-05,
"loss": 1.3309,
"step": 3018
},
{
"epoch": 0.5364965125061086,
"grad_norm": 0.3549079415601038,
"learning_rate": 1.8006683270663654e-05,
"loss": 1.3569,
"step": 3019
},
{
"epoch": 0.5366742192012084,
"grad_norm": 0.3512322008361958,
"learning_rate": 1.799545934934071e-05,
"loss": 1.3442,
"step": 3020
},
{
"epoch": 0.5368519258963081,
"grad_norm": 0.35991781262001044,
"learning_rate": 1.7984236065699844e-05,
"loss": 1.3447,
"step": 3021
},
{
"epoch": 0.5370296325914079,
"grad_norm": 0.3598805530485889,
"learning_rate": 1.7973013423311384e-05,
"loss": 1.365,
"step": 3022
},
{
"epoch": 0.5372073392865077,
"grad_norm": 0.35015929218715575,
"learning_rate": 1.796179142574548e-05,
"loss": 1.2975,
"step": 3023
},
{
"epoch": 0.5373850459816074,
"grad_norm": 0.3508625891411885,
"learning_rate": 1.795057007657206e-05,
"loss": 1.2904,
"step": 3024
},
{
"epoch": 0.5375627526767071,
"grad_norm": 0.3494995761432889,
"learning_rate": 1.7939349379360836e-05,
"loss": 1.2959,
"step": 3025
},
{
"epoch": 0.5377404593718068,
"grad_norm": 0.35341436658858655,
"learning_rate": 1.7928129337681327e-05,
"loss": 1.322,
"step": 3026
},
{
"epoch": 0.5379181660669066,
"grad_norm": 0.34971611308889927,
"learning_rate": 1.7916909955102827e-05,
"loss": 1.3161,
"step": 3027
},
{
"epoch": 0.5380958727620063,
"grad_norm": 0.38961704621842136,
"learning_rate": 1.7905691235194462e-05,
"loss": 1.3028,
"step": 3028
},
{
"epoch": 0.5382735794571061,
"grad_norm": 0.35677472810992455,
"learning_rate": 1.7894473181525092e-05,
"loss": 1.3397,
"step": 3029
},
{
"epoch": 0.5384512861522058,
"grad_norm": 0.3649535441281923,
"learning_rate": 1.78832557976634e-05,
"loss": 1.378,
"step": 3030
},
{
"epoch": 0.5386289928473055,
"grad_norm": 0.35723591638127455,
"learning_rate": 1.7872039087177848e-05,
"loss": 1.341,
"step": 3031
},
{
"epoch": 0.5388066995424052,
"grad_norm": 0.35625122565538725,
"learning_rate": 1.7860823053636677e-05,
"loss": 1.3043,
"step": 3032
},
{
"epoch": 0.538984406237505,
"grad_norm": 0.35708876308003723,
"learning_rate": 1.7849607700607922e-05,
"loss": 1.34,
"step": 3033
},
{
"epoch": 0.5391621129326047,
"grad_norm": 0.3543149424488148,
"learning_rate": 1.78383930316594e-05,
"loss": 1.3582,
"step": 3034
},
{
"epoch": 0.5393398196277045,
"grad_norm": 0.355221743684264,
"learning_rate": 1.7827179050358704e-05,
"loss": 1.3097,
"step": 3035
},
{
"epoch": 0.5395175263228043,
"grad_norm": 0.35526011374492955,
"learning_rate": 1.781596576027321e-05,
"loss": 1.3328,
"step": 3036
},
{
"epoch": 0.5396952330179039,
"grad_norm": 0.386357198429028,
"learning_rate": 1.7804753164970086e-05,
"loss": 1.3241,
"step": 3037
},
{
"epoch": 0.5398729397130037,
"grad_norm": 0.35714896040963595,
"learning_rate": 1.779354126801626e-05,
"loss": 1.3609,
"step": 3038
},
{
"epoch": 0.5400506464081034,
"grad_norm": 0.36567638648350054,
"learning_rate": 1.7782330072978454e-05,
"loss": 1.3417,
"step": 3039
},
{
"epoch": 0.5402283531032032,
"grad_norm": 0.35993581359518056,
"learning_rate": 1.7771119583423164e-05,
"loss": 1.3242,
"step": 3040
},
{
"epoch": 0.5404060597983029,
"grad_norm": 0.3555776449931194,
"learning_rate": 1.7759909802916633e-05,
"loss": 1.3071,
"step": 3041
},
{
"epoch": 0.5405837664934027,
"grad_norm": 0.36503317848006256,
"learning_rate": 1.774870073502493e-05,
"loss": 1.3673,
"step": 3042
},
{
"epoch": 0.5407614731885024,
"grad_norm": 0.3597704600444554,
"learning_rate": 1.7737492383313866e-05,
"loss": 1.3685,
"step": 3043
},
{
"epoch": 0.5409391798836021,
"grad_norm": 0.3573516271753079,
"learning_rate": 1.772628475134902e-05,
"loss": 1.3141,
"step": 3044
},
{
"epoch": 0.5411168865787018,
"grad_norm": 0.3677540315029491,
"learning_rate": 1.771507784269575e-05,
"loss": 1.3534,
"step": 3045
},
{
"epoch": 0.5412945932738016,
"grad_norm": 0.3552128078720035,
"learning_rate": 1.770387166091918e-05,
"loss": 1.3129,
"step": 3046
},
{
"epoch": 0.5414722999689013,
"grad_norm": 0.3819426965377775,
"learning_rate": 1.769266620958423e-05,
"loss": 1.3113,
"step": 3047
},
{
"epoch": 0.5416500066640011,
"grad_norm": 0.35618132966541255,
"learning_rate": 1.768146149225555e-05,
"loss": 1.2995,
"step": 3048
},
{
"epoch": 0.5418277133591008,
"grad_norm": 0.35724589797614714,
"learning_rate": 1.7670257512497564e-05,
"loss": 1.3358,
"step": 3049
},
{
"epoch": 0.5420054200542005,
"grad_norm": 0.3588161918993778,
"learning_rate": 1.7659054273874476e-05,
"loss": 1.323,
"step": 3050
},
{
"epoch": 0.5421831267493002,
"grad_norm": 0.3575850726531929,
"learning_rate": 1.764785177995025e-05,
"loss": 1.3407,
"step": 3051
},
{
"epoch": 0.5423608334444,
"grad_norm": 0.35243129018811564,
"learning_rate": 1.7636650034288605e-05,
"loss": 1.3162,
"step": 3052
},
{
"epoch": 0.5425385401394998,
"grad_norm": 0.35327567253855763,
"learning_rate": 1.762544904045303e-05,
"loss": 1.3537,
"step": 3053
},
{
"epoch": 0.5427162468345995,
"grad_norm": 0.35277026179544885,
"learning_rate": 1.7614248802006773e-05,
"loss": 1.3059,
"step": 3054
},
{
"epoch": 0.5428939535296993,
"grad_norm": 0.3496131098162887,
"learning_rate": 1.7603049322512834e-05,
"loss": 1.3531,
"step": 3055
},
{
"epoch": 0.543071660224799,
"grad_norm": 0.4074062166992718,
"learning_rate": 1.759185060553398e-05,
"loss": 1.322,
"step": 3056
},
{
"epoch": 0.5432493669198987,
"grad_norm": 0.3562788240116262,
"learning_rate": 1.7580652654632745e-05,
"loss": 1.3445,
"step": 3057
},
{
"epoch": 0.5434270736149984,
"grad_norm": 0.35871445418872067,
"learning_rate": 1.756945547337139e-05,
"loss": 1.3234,
"step": 3058
},
{
"epoch": 0.5436047803100982,
"grad_norm": 0.35669561372998515,
"learning_rate": 1.755825906531197e-05,
"loss": 1.3563,
"step": 3059
},
{
"epoch": 0.5437824870051979,
"grad_norm": 0.3567260191406541,
"learning_rate": 1.7547063434016242e-05,
"loss": 1.3664,
"step": 3060
},
{
"epoch": 0.5439601937002977,
"grad_norm": 0.35571425325610206,
"learning_rate": 1.7535868583045773e-05,
"loss": 1.3245,
"step": 3061
},
{
"epoch": 0.5441379003953974,
"grad_norm": 0.3590728562137645,
"learning_rate": 1.7524674515961853e-05,
"loss": 1.3882,
"step": 3062
},
{
"epoch": 0.5443156070904971,
"grad_norm": 0.3619417521619826,
"learning_rate": 1.751348123632552e-05,
"loss": 1.328,
"step": 3063
},
{
"epoch": 0.5444933137855968,
"grad_norm": 0.34701791246927793,
"learning_rate": 1.7502288747697552e-05,
"loss": 1.2806,
"step": 3064
},
{
"epoch": 0.5446710204806966,
"grad_norm": 0.3575668437803739,
"learning_rate": 1.7491097053638522e-05,
"loss": 1.3586,
"step": 3065
},
{
"epoch": 0.5448487271757964,
"grad_norm": 0.35812062014771007,
"learning_rate": 1.7479906157708693e-05,
"loss": 1.3392,
"step": 3066
},
{
"epoch": 0.5450264338708961,
"grad_norm": 0.3590297708568551,
"learning_rate": 1.7468716063468112e-05,
"loss": 1.3529,
"step": 3067
},
{
"epoch": 0.5452041405659959,
"grad_norm": 0.35219205364100037,
"learning_rate": 1.7457526774476554e-05,
"loss": 1.3414,
"step": 3068
},
{
"epoch": 0.5453818472610955,
"grad_norm": 0.3544861097378537,
"learning_rate": 1.7446338294293537e-05,
"loss": 1.3128,
"step": 3069
},
{
"epoch": 0.5455595539561953,
"grad_norm": 0.35002751355318656,
"learning_rate": 1.7435150626478336e-05,
"loss": 1.3295,
"step": 3070
},
{
"epoch": 0.545737260651295,
"grad_norm": 0.36288101499062586,
"learning_rate": 1.7423963774589953e-05,
"loss": 1.3299,
"step": 3071
},
{
"epoch": 0.5459149673463948,
"grad_norm": 0.35572605627120046,
"learning_rate": 1.7412777742187142e-05,
"loss": 1.345,
"step": 3072
},
{
"epoch": 0.5460926740414945,
"grad_norm": 0.36077172396005225,
"learning_rate": 1.7401592532828384e-05,
"loss": 1.338,
"step": 3073
},
{
"epoch": 0.5462703807365943,
"grad_norm": 0.3585725726952749,
"learning_rate": 1.73904081500719e-05,
"loss": 1.3553,
"step": 3074
},
{
"epoch": 0.546448087431694,
"grad_norm": 0.3725317248026827,
"learning_rate": 1.737922459747567e-05,
"loss": 1.2935,
"step": 3075
},
{
"epoch": 0.5466257941267937,
"grad_norm": 0.35902413703841757,
"learning_rate": 1.7368041878597375e-05,
"loss": 1.3572,
"step": 3076
},
{
"epoch": 0.5468035008218934,
"grad_norm": 0.35003186660553554,
"learning_rate": 1.7356859996994456e-05,
"loss": 1.3736,
"step": 3077
},
{
"epoch": 0.5469812075169932,
"grad_norm": 0.3517635948492715,
"learning_rate": 1.7345678956224075e-05,
"loss": 1.346,
"step": 3078
},
{
"epoch": 0.547158914212093,
"grad_norm": 0.3570032145902501,
"learning_rate": 1.733449875984314e-05,
"loss": 1.3641,
"step": 3079
},
{
"epoch": 0.5473366209071927,
"grad_norm": 0.3479475227847456,
"learning_rate": 1.7323319411408276e-05,
"loss": 1.2896,
"step": 3080
},
{
"epoch": 0.5475143276022925,
"grad_norm": 0.35604039988300484,
"learning_rate": 1.7312140914475848e-05,
"loss": 1.3612,
"step": 3081
},
{
"epoch": 0.5476920342973921,
"grad_norm": 0.35887943462098254,
"learning_rate": 1.730096327260194e-05,
"loss": 1.3721,
"step": 3082
},
{
"epoch": 0.5478697409924919,
"grad_norm": 0.3566269346555792,
"learning_rate": 1.728978648934236e-05,
"loss": 1.3481,
"step": 3083
},
{
"epoch": 0.5480474476875916,
"grad_norm": 0.3469964850556751,
"learning_rate": 1.727861056825268e-05,
"loss": 1.356,
"step": 3084
},
{
"epoch": 0.5482251543826914,
"grad_norm": 0.35206929136748466,
"learning_rate": 1.7267435512888156e-05,
"loss": 1.3571,
"step": 3085
},
{
"epoch": 0.5484028610777911,
"grad_norm": 0.33846926140930417,
"learning_rate": 1.725626132680378e-05,
"loss": 1.3069,
"step": 3086
},
{
"epoch": 0.5485805677728909,
"grad_norm": 0.35085091367425897,
"learning_rate": 1.7245088013554275e-05,
"loss": 1.3614,
"step": 3087
},
{
"epoch": 0.5487582744679906,
"grad_norm": 0.34600661934185983,
"learning_rate": 1.7233915576694077e-05,
"loss": 1.2973,
"step": 3088
},
{
"epoch": 0.5489359811630903,
"grad_norm": 0.35748934480367617,
"learning_rate": 1.7222744019777356e-05,
"loss": 1.334,
"step": 3089
},
{
"epoch": 0.54911368785819,
"grad_norm": 0.3529501407107381,
"learning_rate": 1.7211573346357992e-05,
"loss": 1.3026,
"step": 3090
},
{
"epoch": 0.5492913945532898,
"grad_norm": 0.362601025336206,
"learning_rate": 1.7200403559989586e-05,
"loss": 1.359,
"step": 3091
},
{
"epoch": 0.5494691012483895,
"grad_norm": 0.3606285981160576,
"learning_rate": 1.718923466422545e-05,
"loss": 1.331,
"step": 3092
},
{
"epoch": 0.5496468079434893,
"grad_norm": 0.353932627507291,
"learning_rate": 1.7178066662618633e-05,
"loss": 1.352,
"step": 3093
},
{
"epoch": 0.5498245146385891,
"grad_norm": 0.3671176423998276,
"learning_rate": 1.7166899558721876e-05,
"loss": 1.3619,
"step": 3094
},
{
"epoch": 0.5500022213336887,
"grad_norm": 0.37166334479742885,
"learning_rate": 1.715573335608765e-05,
"loss": 1.3534,
"step": 3095
},
{
"epoch": 0.5501799280287885,
"grad_norm": 0.3585155480725968,
"learning_rate": 1.7144568058268136e-05,
"loss": 1.2939,
"step": 3096
},
{
"epoch": 0.5503576347238882,
"grad_norm": 0.3476489841910763,
"learning_rate": 1.713340366881521e-05,
"loss": 1.3171,
"step": 3097
},
{
"epoch": 0.550535341418988,
"grad_norm": 0.36089357995071264,
"learning_rate": 1.7122240191280493e-05,
"loss": 1.3312,
"step": 3098
},
{
"epoch": 0.5507130481140877,
"grad_norm": 0.3606355818482498,
"learning_rate": 1.711107762921529e-05,
"loss": 1.3612,
"step": 3099
},
{
"epoch": 0.5508907548091875,
"grad_norm": 0.35963905731919565,
"learning_rate": 1.7099915986170628e-05,
"loss": 1.2836,
"step": 3100
},
{
"epoch": 0.5510684615042871,
"grad_norm": 0.3619954231171682,
"learning_rate": 1.7088755265697222e-05,
"loss": 1.3464,
"step": 3101
},
{
"epoch": 0.5512461681993869,
"grad_norm": 0.3531074933627891,
"learning_rate": 1.7077595471345507e-05,
"loss": 1.3183,
"step": 3102
},
{
"epoch": 0.5514238748944866,
"grad_norm": 0.4372574883289275,
"learning_rate": 1.7066436606665642e-05,
"loss": 1.3441,
"step": 3103
},
{
"epoch": 0.5516015815895864,
"grad_norm": 0.3537939767190582,
"learning_rate": 1.705527867520746e-05,
"loss": 1.3214,
"step": 3104
},
{
"epoch": 0.5517792882846861,
"grad_norm": 0.365278924288947,
"learning_rate": 1.704412168052051e-05,
"loss": 1.3528,
"step": 3105
},
{
"epoch": 0.5519569949797859,
"grad_norm": 0.3763367785721599,
"learning_rate": 1.7032965626154038e-05,
"loss": 1.3752,
"step": 3106
},
{
"epoch": 0.5521347016748857,
"grad_norm": 0.34705508717764383,
"learning_rate": 1.7021810515656993e-05,
"loss": 1.3156,
"step": 3107
},
{
"epoch": 0.5523124083699853,
"grad_norm": 0.34835072222060387,
"learning_rate": 1.7010656352578036e-05,
"loss": 1.3291,
"step": 3108
},
{
"epoch": 0.552490115065085,
"grad_norm": 0.4580993098436763,
"learning_rate": 1.6999503140465514e-05,
"loss": 1.3661,
"step": 3109
},
{
"epoch": 0.5526678217601848,
"grad_norm": 0.35900709238175893,
"learning_rate": 1.6988350882867464e-05,
"loss": 1.3398,
"step": 3110
},
{
"epoch": 0.5528455284552846,
"grad_norm": 0.35284705143877854,
"learning_rate": 1.6977199583331633e-05,
"loss": 1.3351,
"step": 3111
},
{
"epoch": 0.5530232351503843,
"grad_norm": 0.34305857552882685,
"learning_rate": 1.6966049245405466e-05,
"loss": 1.2939,
"step": 3112
},
{
"epoch": 0.5532009418454841,
"grad_norm": 0.35847511603474036,
"learning_rate": 1.6954899872636087e-05,
"loss": 1.349,
"step": 3113
},
{
"epoch": 0.5533786485405837,
"grad_norm": 0.3462736865695819,
"learning_rate": 1.6943751468570327e-05,
"loss": 1.2828,
"step": 3114
},
{
"epoch": 0.5535563552356835,
"grad_norm": 0.3568513652018462,
"learning_rate": 1.6932604036754706e-05,
"loss": 1.3258,
"step": 3115
},
{
"epoch": 0.5537340619307832,
"grad_norm": 0.35968340847110947,
"learning_rate": 1.692145758073541e-05,
"loss": 1.3438,
"step": 3116
},
{
"epoch": 0.553911768625883,
"grad_norm": 0.35653677735737366,
"learning_rate": 1.691031210405836e-05,
"loss": 1.3113,
"step": 3117
},
{
"epoch": 0.5540894753209827,
"grad_norm": 0.5240154486520902,
"learning_rate": 1.689916761026914e-05,
"loss": 1.3239,
"step": 3118
},
{
"epoch": 0.5542671820160825,
"grad_norm": 0.35957037671848324,
"learning_rate": 1.6888024102913013e-05,
"loss": 1.3685,
"step": 3119
},
{
"epoch": 0.5544448887111822,
"grad_norm": 0.34923791840275853,
"learning_rate": 1.6876881585534943e-05,
"loss": 1.2904,
"step": 3120
},
{
"epoch": 0.5546225954062819,
"grad_norm": 0.3547767396722677,
"learning_rate": 1.686574006167956e-05,
"loss": 1.3379,
"step": 3121
},
{
"epoch": 0.5548003021013816,
"grad_norm": 0.3634242515485292,
"learning_rate": 1.6854599534891223e-05,
"loss": 1.3454,
"step": 3122
},
{
"epoch": 0.5549780087964814,
"grad_norm": 0.36149375504685927,
"learning_rate": 1.6843460008713922e-05,
"loss": 1.3201,
"step": 3123
},
{
"epoch": 0.5551557154915812,
"grad_norm": 0.35819806088013895,
"learning_rate": 1.683232148669135e-05,
"loss": 1.338,
"step": 3124
},
{
"epoch": 0.5553334221866809,
"grad_norm": 0.3565175050855168,
"learning_rate": 1.6821183972366882e-05,
"loss": 1.315,
"step": 3125
},
{
"epoch": 0.5555111288817807,
"grad_norm": 0.34945516594482867,
"learning_rate": 1.6810047469283577e-05,
"loss": 1.3153,
"step": 3126
},
{
"epoch": 0.5556888355768803,
"grad_norm": 0.37481237517516475,
"learning_rate": 1.6798911980984163e-05,
"loss": 1.3832,
"step": 3127
},
{
"epoch": 0.5558665422719801,
"grad_norm": 0.3595863188078745,
"learning_rate": 1.6787777511011046e-05,
"loss": 1.3414,
"step": 3128
},
{
"epoch": 0.5560442489670798,
"grad_norm": 0.3624400748523654,
"learning_rate": 1.677664406290631e-05,
"loss": 1.3761,
"step": 3129
},
{
"epoch": 0.5562219556621796,
"grad_norm": 0.3649961984077324,
"learning_rate": 1.6765511640211714e-05,
"loss": 1.3637,
"step": 3130
},
{
"epoch": 0.5563996623572793,
"grad_norm": 0.3550793370272792,
"learning_rate": 1.6754380246468694e-05,
"loss": 1.3631,
"step": 3131
},
{
"epoch": 0.5565773690523791,
"grad_norm": 0.3548229769389945,
"learning_rate": 1.6743249885218355e-05,
"loss": 1.3215,
"step": 3132
},
{
"epoch": 0.5567550757474787,
"grad_norm": 0.36251495804038464,
"learning_rate": 1.6732120560001474e-05,
"loss": 1.319,
"step": 3133
},
{
"epoch": 0.5569327824425785,
"grad_norm": 0.35566342417657576,
"learning_rate": 1.6720992274358504e-05,
"loss": 1.3486,
"step": 3134
},
{
"epoch": 0.5571104891376782,
"grad_norm": 0.35326391704113047,
"learning_rate": 1.6709865031829538e-05,
"loss": 1.3834,
"step": 3135
},
{
"epoch": 0.557288195832778,
"grad_norm": 0.35615412607248265,
"learning_rate": 1.6698738835954394e-05,
"loss": 1.3267,
"step": 3136
},
{
"epoch": 0.5574659025278778,
"grad_norm": 0.34959900784242404,
"learning_rate": 1.668761369027251e-05,
"loss": 1.3258,
"step": 3137
},
{
"epoch": 0.5576436092229775,
"grad_norm": 0.36640842013392394,
"learning_rate": 1.6676489598323002e-05,
"loss": 1.3253,
"step": 3138
},
{
"epoch": 0.5578213159180773,
"grad_norm": 0.39939342676383416,
"learning_rate": 1.666536656364464e-05,
"loss": 1.3022,
"step": 3139
},
{
"epoch": 0.5579990226131769,
"grad_norm": 0.34989206821726687,
"learning_rate": 1.6654244589775896e-05,
"loss": 1.3085,
"step": 3140
},
{
"epoch": 0.5581767293082767,
"grad_norm": 0.34874797446072786,
"learning_rate": 1.6643123680254873e-05,
"loss": 1.3027,
"step": 3141
},
{
"epoch": 0.5583544360033764,
"grad_norm": 0.36344274162640366,
"learning_rate": 1.6632003838619333e-05,
"loss": 1.3442,
"step": 3142
},
{
"epoch": 0.5585321426984762,
"grad_norm": 0.3418024375247051,
"learning_rate": 1.6620885068406707e-05,
"loss": 1.2915,
"step": 3143
},
{
"epoch": 0.5587098493935759,
"grad_norm": 0.35103880021691486,
"learning_rate": 1.6609767373154088e-05,
"loss": 1.296,
"step": 3144
},
{
"epoch": 0.5588875560886757,
"grad_norm": 0.3664900538075101,
"learning_rate": 1.6598650756398224e-05,
"loss": 1.3564,
"step": 3145
},
{
"epoch": 0.5590652627837753,
"grad_norm": 0.36895711224948674,
"learning_rate": 1.6587535221675518e-05,
"loss": 1.3625,
"step": 3146
},
{
"epoch": 0.5592429694788751,
"grad_norm": 0.3529813030125228,
"learning_rate": 1.6576420772522038e-05,
"loss": 1.356,
"step": 3147
},
{
"epoch": 0.5594206761739748,
"grad_norm": 0.3666503520441967,
"learning_rate": 1.656530741247349e-05,
"loss": 1.3251,
"step": 3148
},
{
"epoch": 0.5595983828690746,
"grad_norm": 0.3611460278271957,
"learning_rate": 1.6554195145065242e-05,
"loss": 1.3243,
"step": 3149
},
{
"epoch": 0.5597760895641744,
"grad_norm": 0.35058843427442477,
"learning_rate": 1.6543083973832327e-05,
"loss": 1.3021,
"step": 3150
},
{
"epoch": 0.5599537962592741,
"grad_norm": 0.35316736846492747,
"learning_rate": 1.6531973902309406e-05,
"loss": 1.3385,
"step": 3151
},
{
"epoch": 0.5601315029543739,
"grad_norm": 0.35910993188327567,
"learning_rate": 1.6520864934030808e-05,
"loss": 1.3289,
"step": 3152
},
{
"epoch": 0.5603092096494735,
"grad_norm": 0.35387489800799665,
"learning_rate": 1.6509757072530498e-05,
"loss": 1.3571,
"step": 3153
},
{
"epoch": 0.5604869163445733,
"grad_norm": 0.34603162757437156,
"learning_rate": 1.6498650321342106e-05,
"loss": 1.3123,
"step": 3154
},
{
"epoch": 0.560664623039673,
"grad_norm": 0.36117466593088743,
"learning_rate": 1.648754468399889e-05,
"loss": 1.3664,
"step": 3155
},
{
"epoch": 0.5608423297347728,
"grad_norm": 0.35083761152022125,
"learning_rate": 1.6476440164033768e-05,
"loss": 1.3148,
"step": 3156
},
{
"epoch": 0.5610200364298725,
"grad_norm": 0.3527868710534661,
"learning_rate": 1.6465336764979292e-05,
"loss": 1.3455,
"step": 3157
},
{
"epoch": 0.5611977431249723,
"grad_norm": 0.36724013950610684,
"learning_rate": 1.6454234490367653e-05,
"loss": 1.3264,
"step": 3158
},
{
"epoch": 0.5613754498200719,
"grad_norm": 0.38691132838249376,
"learning_rate": 1.644313334373072e-05,
"loss": 1.3596,
"step": 3159
},
{
"epoch": 0.5615531565151717,
"grad_norm": 0.3671559049545463,
"learning_rate": 1.6432033328599952e-05,
"loss": 1.3239,
"step": 3160
},
{
"epoch": 0.5617308632102714,
"grad_norm": 0.4085327367617323,
"learning_rate": 1.642093444850648e-05,
"loss": 1.3388,
"step": 3161
},
{
"epoch": 0.5619085699053712,
"grad_norm": 0.366021203537141,
"learning_rate": 1.640983670698107e-05,
"loss": 1.3164,
"step": 3162
},
{
"epoch": 0.562086276600471,
"grad_norm": 0.3638684117269417,
"learning_rate": 1.6398740107554118e-05,
"loss": 1.3138,
"step": 3163
},
{
"epoch": 0.5622639832955707,
"grad_norm": 0.3582766330608195,
"learning_rate": 1.638764465375566e-05,
"loss": 1.3376,
"step": 3164
},
{
"epoch": 0.5624416899906703,
"grad_norm": 0.36132658155922703,
"learning_rate": 1.6376550349115378e-05,
"loss": 1.3365,
"step": 3165
},
{
"epoch": 0.5626193966857701,
"grad_norm": 0.35596273175243176,
"learning_rate": 1.6365457197162565e-05,
"loss": 1.319,
"step": 3166
},
{
"epoch": 0.5627971033808699,
"grad_norm": 0.36480806830707185,
"learning_rate": 1.635436520142617e-05,
"loss": 1.3252,
"step": 3167
},
{
"epoch": 0.5629748100759696,
"grad_norm": 0.3498872083721939,
"learning_rate": 1.6343274365434766e-05,
"loss": 1.303,
"step": 3168
},
{
"epoch": 0.5631525167710694,
"grad_norm": 0.36113176584927664,
"learning_rate": 1.6332184692716553e-05,
"loss": 1.3016,
"step": 3169
},
{
"epoch": 0.5633302234661691,
"grad_norm": 0.36790245884700934,
"learning_rate": 1.6321096186799365e-05,
"loss": 1.3583,
"step": 3170
},
{
"epoch": 0.5635079301612689,
"grad_norm": 0.3498005018215792,
"learning_rate": 1.6310008851210666e-05,
"loss": 1.3292,
"step": 3171
},
{
"epoch": 0.5636856368563685,
"grad_norm": 0.3706804926717746,
"learning_rate": 1.6298922689477542e-05,
"loss": 1.3673,
"step": 3172
},
{
"epoch": 0.5638633435514683,
"grad_norm": 0.3533417787854198,
"learning_rate": 1.6287837705126714e-05,
"loss": 1.3052,
"step": 3173
},
{
"epoch": 0.564041050246568,
"grad_norm": 0.37064753003851436,
"learning_rate": 1.6276753901684524e-05,
"loss": 1.3368,
"step": 3174
},
{
"epoch": 0.5642187569416678,
"grad_norm": 0.3647135958943936,
"learning_rate": 1.626567128267694e-05,
"loss": 1.3428,
"step": 3175
},
{
"epoch": 0.5643964636367675,
"grad_norm": 0.3579611374920639,
"learning_rate": 1.6254589851629546e-05,
"loss": 1.335,
"step": 3176
},
{
"epoch": 0.5645741703318673,
"grad_norm": 0.370601135161346,
"learning_rate": 1.6243509612067545e-05,
"loss": 1.3797,
"step": 3177
},
{
"epoch": 0.5647518770269669,
"grad_norm": 0.3629090828705271,
"learning_rate": 1.6232430567515794e-05,
"loss": 1.3551,
"step": 3178
},
{
"epoch": 0.5649295837220667,
"grad_norm": 0.3622790767015303,
"learning_rate": 1.6221352721498726e-05,
"loss": 1.3681,
"step": 3179
},
{
"epoch": 0.5651072904171665,
"grad_norm": 0.3471163223224285,
"learning_rate": 1.6210276077540422e-05,
"loss": 1.2967,
"step": 3180
},
{
"epoch": 0.5652849971122662,
"grad_norm": 0.36694944616670594,
"learning_rate": 1.619920063916456e-05,
"loss": 1.3344,
"step": 3181
},
{
"epoch": 0.565462703807366,
"grad_norm": 0.3698756203737894,
"learning_rate": 1.6188126409894452e-05,
"loss": 1.4027,
"step": 3182
},
{
"epoch": 0.5656404105024657,
"grad_norm": 0.3516163913323672,
"learning_rate": 1.6177053393253026e-05,
"loss": 1.2938,
"step": 3183
},
{
"epoch": 0.5658181171975655,
"grad_norm": 0.3706599579228091,
"learning_rate": 1.6165981592762807e-05,
"loss": 1.3467,
"step": 3184
},
{
"epoch": 0.5659958238926651,
"grad_norm": 0.3690083798372657,
"learning_rate": 1.6154911011945943e-05,
"loss": 1.3351,
"step": 3185
},
{
"epoch": 0.5661735305877649,
"grad_norm": 0.3701106811890601,
"learning_rate": 1.6143841654324192e-05,
"loss": 1.4176,
"step": 3186
},
{
"epoch": 0.5663512372828646,
"grad_norm": 0.3513633874600803,
"learning_rate": 1.6132773523418933e-05,
"loss": 1.3176,
"step": 3187
},
{
"epoch": 0.5665289439779644,
"grad_norm": 0.3607221197394773,
"learning_rate": 1.6121706622751147e-05,
"loss": 1.311,
"step": 3188
},
{
"epoch": 0.5667066506730641,
"grad_norm": 0.36019188579581696,
"learning_rate": 1.6110640955841415e-05,
"loss": 1.363,
"step": 3189
},
{
"epoch": 0.5668843573681639,
"grad_norm": 0.35264667170816555,
"learning_rate": 1.6099576526209944e-05,
"loss": 1.3077,
"step": 3190
},
{
"epoch": 0.5670620640632635,
"grad_norm": 0.3638200593889725,
"learning_rate": 1.6088513337376515e-05,
"loss": 1.3285,
"step": 3191
},
{
"epoch": 0.5672397707583633,
"grad_norm": 0.381826937536756,
"learning_rate": 1.6077451392860565e-05,
"loss": 1.3294,
"step": 3192
},
{
"epoch": 0.567417477453463,
"grad_norm": 0.3563212636669049,
"learning_rate": 1.606639069618109e-05,
"loss": 1.3207,
"step": 3193
},
{
"epoch": 0.5675951841485628,
"grad_norm": 0.3652014433716863,
"learning_rate": 1.6055331250856715e-05,
"loss": 1.3755,
"step": 3194
},
{
"epoch": 0.5677728908436626,
"grad_norm": 0.35088420798678743,
"learning_rate": 1.604427306040564e-05,
"loss": 1.3357,
"step": 3195
},
{
"epoch": 0.5679505975387623,
"grad_norm": 0.36113584904182816,
"learning_rate": 1.60332161283457e-05,
"loss": 1.36,
"step": 3196
},
{
"epoch": 0.568128304233862,
"grad_norm": 0.36781248043336534,
"learning_rate": 1.602216045819432e-05,
"loss": 1.3769,
"step": 3197
},
{
"epoch": 0.5683060109289617,
"grad_norm": 0.3494576248740234,
"learning_rate": 1.6011106053468494e-05,
"loss": 1.2954,
"step": 3198
},
{
"epoch": 0.5684837176240615,
"grad_norm": 0.3582667203690463,
"learning_rate": 1.600005291768485e-05,
"loss": 1.3736,
"step": 3199
},
{
"epoch": 0.5686614243191612,
"grad_norm": 0.35909854837601946,
"learning_rate": 1.598900105435959e-05,
"loss": 1.316,
"step": 3200
},
{
"epoch": 0.568839131014261,
"grad_norm": 0.3463470363162854,
"learning_rate": 1.5977950467008527e-05,
"loss": 1.3439,
"step": 3201
},
{
"epoch": 0.5690168377093607,
"grad_norm": 0.35926301577301817,
"learning_rate": 1.5966901159147063e-05,
"loss": 1.3545,
"step": 3202
},
{
"epoch": 0.5691945444044605,
"grad_norm": 0.3671054438992789,
"learning_rate": 1.595585313429018e-05,
"loss": 1.3776,
"step": 3203
},
{
"epoch": 0.5693722510995601,
"grad_norm": 0.3472875993083112,
"learning_rate": 1.5944806395952473e-05,
"loss": 1.3275,
"step": 3204
},
{
"epoch": 0.5695499577946599,
"grad_norm": 0.3522607929545398,
"learning_rate": 1.593376094764811e-05,
"loss": 1.3226,
"step": 3205
},
{
"epoch": 0.5697276644897596,
"grad_norm": 0.368788247840791,
"learning_rate": 1.592271679289086e-05,
"loss": 1.3897,
"step": 3206
},
{
"epoch": 0.5699053711848594,
"grad_norm": 0.35026937797876284,
"learning_rate": 1.5911673935194076e-05,
"loss": 1.3049,
"step": 3207
},
{
"epoch": 0.5700830778799592,
"grad_norm": 0.3990470937262973,
"learning_rate": 1.59006323780707e-05,
"loss": 1.3198,
"step": 3208
},
{
"epoch": 0.5702607845750589,
"grad_norm": 0.361318233042466,
"learning_rate": 1.588959212503325e-05,
"loss": 1.3816,
"step": 3209
},
{
"epoch": 0.5704384912701586,
"grad_norm": 0.34903541347334777,
"learning_rate": 1.5878553179593847e-05,
"loss": 1.3261,
"step": 3210
},
{
"epoch": 0.5706161979652583,
"grad_norm": 0.35116018630985574,
"learning_rate": 1.5867515545264186e-05,
"loss": 1.2892,
"step": 3211
},
{
"epoch": 0.5707939046603581,
"grad_norm": 0.36176007554064704,
"learning_rate": 1.585647922555555e-05,
"loss": 1.3762,
"step": 3212
},
{
"epoch": 0.5709716113554578,
"grad_norm": 0.35153628095762707,
"learning_rate": 1.584544422397879e-05,
"loss": 1.323,
"step": 3213
},
{
"epoch": 0.5711493180505576,
"grad_norm": 0.3632437642154713,
"learning_rate": 1.5834410544044342e-05,
"loss": 1.3439,
"step": 3214
},
{
"epoch": 0.5713270247456573,
"grad_norm": 0.34574596818089126,
"learning_rate": 1.582337818926225e-05,
"loss": 1.2832,
"step": 3215
},
{
"epoch": 0.5715047314407571,
"grad_norm": 0.3578436850359886,
"learning_rate": 1.58123471631421e-05,
"loss": 1.3258,
"step": 3216
},
{
"epoch": 0.5716824381358567,
"grad_norm": 0.36087813575138256,
"learning_rate": 1.580131746919307e-05,
"loss": 1.3453,
"step": 3217
},
{
"epoch": 0.5718601448309565,
"grad_norm": 0.35312898658473296,
"learning_rate": 1.579028911092391e-05,
"loss": 1.343,
"step": 3218
},
{
"epoch": 0.5720378515260562,
"grad_norm": 0.3533368916444126,
"learning_rate": 1.577926209184295e-05,
"loss": 1.3113,
"step": 3219
},
{
"epoch": 0.572215558221156,
"grad_norm": 0.3545981966774702,
"learning_rate": 1.5768236415458095e-05,
"loss": 1.3334,
"step": 3220
},
{
"epoch": 0.5723932649162558,
"grad_norm": 0.34431380887665286,
"learning_rate": 1.575721208527682e-05,
"loss": 1.2672,
"step": 3221
},
{
"epoch": 0.5725709716113555,
"grad_norm": 0.35849091926681326,
"learning_rate": 1.5746189104806167e-05,
"loss": 1.3472,
"step": 3222
},
{
"epoch": 0.5727486783064552,
"grad_norm": 0.3510242678554384,
"learning_rate": 1.5735167477552752e-05,
"loss": 1.3011,
"step": 3223
},
{
"epoch": 0.5729263850015549,
"grad_norm": 0.3598878342789821,
"learning_rate": 1.5724147207022773e-05,
"loss": 1.3753,
"step": 3224
},
{
"epoch": 0.5731040916966547,
"grad_norm": 0.34479298590246177,
"learning_rate": 1.5713128296721978e-05,
"loss": 1.3028,
"step": 3225
},
{
"epoch": 0.5732817983917544,
"grad_norm": 0.3514177894533139,
"learning_rate": 1.570211075015569e-05,
"loss": 1.2756,
"step": 3226
},
{
"epoch": 0.5734595050868542,
"grad_norm": 0.35047472978574734,
"learning_rate": 1.5691094570828798e-05,
"loss": 1.3366,
"step": 3227
},
{
"epoch": 0.5736372117819539,
"grad_norm": 0.35472909146708137,
"learning_rate": 1.5680079762245747e-05,
"loss": 1.331,
"step": 3228
},
{
"epoch": 0.5738149184770536,
"grad_norm": 0.35207134089893716,
"learning_rate": 1.5669066327910573e-05,
"loss": 1.3477,
"step": 3229
},
{
"epoch": 0.5739926251721533,
"grad_norm": 0.3510568184797083,
"learning_rate": 1.5658054271326844e-05,
"loss": 1.3353,
"step": 3230
},
{
"epoch": 0.5741703318672531,
"grad_norm": 0.35600104034540386,
"learning_rate": 1.5647043595997713e-05,
"loss": 1.3858,
"step": 3231
},
{
"epoch": 0.5743480385623528,
"grad_norm": 0.38237579124959836,
"learning_rate": 1.5636034305425868e-05,
"loss": 1.3474,
"step": 3232
},
{
"epoch": 0.5745257452574526,
"grad_norm": 0.3542852400985203,
"learning_rate": 1.562502640311357e-05,
"loss": 1.3127,
"step": 3233
},
{
"epoch": 0.5747034519525523,
"grad_norm": 0.3597333691166982,
"learning_rate": 1.5614019892562657e-05,
"loss": 1.3416,
"step": 3234
},
{
"epoch": 0.5748811586476521,
"grad_norm": 0.3485553389939112,
"learning_rate": 1.5603014777274503e-05,
"loss": 1.3566,
"step": 3235
},
{
"epoch": 0.5750588653427517,
"grad_norm": 0.3529553014024367,
"learning_rate": 1.5592011060750036e-05,
"loss": 1.3245,
"step": 3236
},
{
"epoch": 0.5752365720378515,
"grad_norm": 0.39995773840563853,
"learning_rate": 1.558100874648973e-05,
"loss": 1.3415,
"step": 3237
},
{
"epoch": 0.5754142787329513,
"grad_norm": 0.35051088254721285,
"learning_rate": 1.5570007837993663e-05,
"loss": 1.3369,
"step": 3238
},
{
"epoch": 0.575591985428051,
"grad_norm": 0.3533619630171019,
"learning_rate": 1.555900833876141e-05,
"loss": 1.3313,
"step": 3239
},
{
"epoch": 0.5757696921231508,
"grad_norm": 0.3798945598733388,
"learning_rate": 1.5548010252292116e-05,
"loss": 1.3033,
"step": 3240
},
{
"epoch": 0.5759473988182505,
"grad_norm": 0.34458667222670897,
"learning_rate": 1.5537013582084486e-05,
"loss": 1.3309,
"step": 3241
},
{
"epoch": 0.5761251055133502,
"grad_norm": 0.3696610785830197,
"learning_rate": 1.5526018331636766e-05,
"loss": 1.3465,
"step": 3242
},
{
"epoch": 0.5763028122084499,
"grad_norm": 0.35269801117298816,
"learning_rate": 1.551502450444675e-05,
"loss": 1.3682,
"step": 3243
},
{
"epoch": 0.5764805189035497,
"grad_norm": 0.4238727420219645,
"learning_rate": 1.5504032104011787e-05,
"loss": 1.2999,
"step": 3244
},
{
"epoch": 0.5766582255986494,
"grad_norm": 0.3516608659324545,
"learning_rate": 1.549304113382876e-05,
"loss": 1.3089,
"step": 3245
},
{
"epoch": 0.5768359322937492,
"grad_norm": 0.3493096051586714,
"learning_rate": 1.5482051597394104e-05,
"loss": 1.3087,
"step": 3246
},
{
"epoch": 0.5770136389888489,
"grad_norm": 0.35735127922305004,
"learning_rate": 1.5471063498203797e-05,
"loss": 1.3312,
"step": 3247
},
{
"epoch": 0.5771913456839487,
"grad_norm": 0.343972749306401,
"learning_rate": 1.5460076839753365e-05,
"loss": 1.3259,
"step": 3248
},
{
"epoch": 0.5773690523790483,
"grad_norm": 0.35654822325568947,
"learning_rate": 1.5449091625537866e-05,
"loss": 1.3841,
"step": 3249
},
{
"epoch": 0.5775467590741481,
"grad_norm": 0.3559314308280083,
"learning_rate": 1.543810785905191e-05,
"loss": 1.3428,
"step": 3250
},
{
"epoch": 0.5777244657692479,
"grad_norm": 0.35398032520178174,
"learning_rate": 1.542712554378962e-05,
"loss": 1.3584,
"step": 3251
},
{
"epoch": 0.5779021724643476,
"grad_norm": 0.4609732313628664,
"learning_rate": 1.5416144683244704e-05,
"loss": 1.3502,
"step": 3252
},
{
"epoch": 0.5780798791594474,
"grad_norm": 0.3672728559384176,
"learning_rate": 1.540516528091037e-05,
"loss": 1.3374,
"step": 3253
},
{
"epoch": 0.5782575858545471,
"grad_norm": 0.358254423588708,
"learning_rate": 1.5394187340279366e-05,
"loss": 1.3331,
"step": 3254
},
{
"epoch": 0.5784352925496468,
"grad_norm": 0.3550856723536352,
"learning_rate": 1.5383210864843986e-05,
"loss": 1.3424,
"step": 3255
},
{
"epoch": 0.5786129992447465,
"grad_norm": 0.3579560610196115,
"learning_rate": 1.5372235858096042e-05,
"loss": 1.3836,
"step": 3256
},
{
"epoch": 0.5787907059398463,
"grad_norm": 0.3582174116761431,
"learning_rate": 1.536126232352691e-05,
"loss": 1.3441,
"step": 3257
},
{
"epoch": 0.578968412634946,
"grad_norm": 0.3522822363052389,
"learning_rate": 1.535029026462747e-05,
"loss": 1.3286,
"step": 3258
},
{
"epoch": 0.5791461193300458,
"grad_norm": 0.3752313541973127,
"learning_rate": 1.5339319684888137e-05,
"loss": 1.3497,
"step": 3259
},
{
"epoch": 0.5793238260251455,
"grad_norm": 0.3497695925402451,
"learning_rate": 1.532835058779886e-05,
"loss": 1.2876,
"step": 3260
},
{
"epoch": 0.5795015327202452,
"grad_norm": 0.3518677734619256,
"learning_rate": 1.531738297684911e-05,
"loss": 1.3378,
"step": 3261
},
{
"epoch": 0.5796792394153449,
"grad_norm": 0.35533776301135683,
"learning_rate": 1.53064168555279e-05,
"loss": 1.3096,
"step": 3262
},
{
"epoch": 0.5798569461104447,
"grad_norm": 0.36041267961354073,
"learning_rate": 1.5295452227323756e-05,
"loss": 1.3346,
"step": 3263
},
{
"epoch": 0.5800346528055444,
"grad_norm": 0.35034600086244017,
"learning_rate": 1.528448909572473e-05,
"loss": 1.2637,
"step": 3264
},
{
"epoch": 0.5802123595006442,
"grad_norm": 0.36925125051668767,
"learning_rate": 1.5273527464218398e-05,
"loss": 1.3285,
"step": 3265
},
{
"epoch": 0.580390066195744,
"grad_norm": 0.35347552230985674,
"learning_rate": 1.526256733629187e-05,
"loss": 1.3322,
"step": 3266
},
{
"epoch": 0.5805677728908437,
"grad_norm": 0.3510632967953715,
"learning_rate": 1.5251608715431764e-05,
"loss": 1.2948,
"step": 3267
},
{
"epoch": 0.5807454795859434,
"grad_norm": 0.35127152990541166,
"learning_rate": 1.5240651605124224e-05,
"loss": 1.3177,
"step": 3268
},
{
"epoch": 0.5809231862810431,
"grad_norm": 0.3575381676930619,
"learning_rate": 1.5229696008854913e-05,
"loss": 1.324,
"step": 3269
},
{
"epoch": 0.5811008929761429,
"grad_norm": 0.3498894547189853,
"learning_rate": 1.5218741930109e-05,
"loss": 1.3233,
"step": 3270
},
{
"epoch": 0.5812785996712426,
"grad_norm": 0.46392480788700874,
"learning_rate": 1.5207789372371205e-05,
"loss": 1.334,
"step": 3271
},
{
"epoch": 0.5814563063663424,
"grad_norm": 0.3549345803719622,
"learning_rate": 1.5196838339125735e-05,
"loss": 1.358,
"step": 3272
},
{
"epoch": 0.5816340130614421,
"grad_norm": 0.429205976671245,
"learning_rate": 1.5185888833856313e-05,
"loss": 1.3235,
"step": 3273
},
{
"epoch": 0.5818117197565418,
"grad_norm": 0.35294133493769286,
"learning_rate": 1.5174940860046184e-05,
"loss": 1.3229,
"step": 3274
},
{
"epoch": 0.5819894264516415,
"grad_norm": 0.355114990183835,
"learning_rate": 1.5163994421178105e-05,
"loss": 1.3376,
"step": 3275
},
{
"epoch": 0.5821671331467413,
"grad_norm": 0.3515651695075507,
"learning_rate": 1.515304952073435e-05,
"loss": 1.3347,
"step": 3276
},
{
"epoch": 0.582344839841841,
"grad_norm": 0.3586565469716434,
"learning_rate": 1.5142106162196692e-05,
"loss": 1.3322,
"step": 3277
},
{
"epoch": 0.5825225465369408,
"grad_norm": 0.36229082757738135,
"learning_rate": 1.5131164349046421e-05,
"loss": 1.3611,
"step": 3278
},
{
"epoch": 0.5827002532320406,
"grad_norm": 0.3673505302785486,
"learning_rate": 1.512022408476433e-05,
"loss": 1.333,
"step": 3279
},
{
"epoch": 0.5828779599271403,
"grad_norm": 0.3460388101266975,
"learning_rate": 1.5109285372830729e-05,
"loss": 1.3122,
"step": 3280
},
{
"epoch": 0.58305566662224,
"grad_norm": 0.34947351322121956,
"learning_rate": 1.5098348216725425e-05,
"loss": 1.305,
"step": 3281
},
{
"epoch": 0.5832333733173397,
"grad_norm": 0.35712074878631417,
"learning_rate": 1.5087412619927736e-05,
"loss": 1.3321,
"step": 3282
},
{
"epoch": 0.5834110800124395,
"grad_norm": 0.35978153789840633,
"learning_rate": 1.5076478585916471e-05,
"loss": 1.3693,
"step": 3283
},
{
"epoch": 0.5835887867075392,
"grad_norm": 0.35067212862536573,
"learning_rate": 1.506554611816996e-05,
"loss": 1.3075,
"step": 3284
},
{
"epoch": 0.583766493402639,
"grad_norm": 0.34625143323763347,
"learning_rate": 1.5054615220166029e-05,
"loss": 1.3426,
"step": 3285
},
{
"epoch": 0.5839442000977387,
"grad_norm": 0.426437127319821,
"learning_rate": 1.5043685895381998e-05,
"loss": 1.3228,
"step": 3286
},
{
"epoch": 0.5841219067928384,
"grad_norm": 0.35546433936846183,
"learning_rate": 1.5032758147294692e-05,
"loss": 1.3629,
"step": 3287
},
{
"epoch": 0.5842996134879381,
"grad_norm": 0.3515534171845095,
"learning_rate": 1.5021831979380436e-05,
"loss": 1.3141,
"step": 3288
},
{
"epoch": 0.5844773201830379,
"grad_norm": 0.35971508100065025,
"learning_rate": 1.5010907395115033e-05,
"loss": 1.3121,
"step": 3289
},
{
"epoch": 0.5846550268781376,
"grad_norm": 0.3538811239344762,
"learning_rate": 1.499998439797382e-05,
"loss": 1.3892,
"step": 3290
},
{
"epoch": 0.5848327335732374,
"grad_norm": 0.3580210692093392,
"learning_rate": 1.4989062991431607e-05,
"loss": 1.3555,
"step": 3291
},
{
"epoch": 0.5850104402683372,
"grad_norm": 0.3563351170750152,
"learning_rate": 1.4978143178962685e-05,
"loss": 1.3431,
"step": 3292
},
{
"epoch": 0.5851881469634368,
"grad_norm": 0.3570521446937506,
"learning_rate": 1.4967224964040847e-05,
"loss": 1.3233,
"step": 3293
},
{
"epoch": 0.5853658536585366,
"grad_norm": 0.3652116318099997,
"learning_rate": 1.495630835013941e-05,
"loss": 1.3099,
"step": 3294
},
{
"epoch": 0.5855435603536363,
"grad_norm": 0.3538480448102535,
"learning_rate": 1.4945393340731131e-05,
"loss": 1.3232,
"step": 3295
},
{
"epoch": 0.5857212670487361,
"grad_norm": 0.36727097077346565,
"learning_rate": 1.493447993928829e-05,
"loss": 1.3271,
"step": 3296
},
{
"epoch": 0.5858989737438358,
"grad_norm": 0.3624193203079002,
"learning_rate": 1.4923568149282636e-05,
"loss": 1.345,
"step": 3297
},
{
"epoch": 0.5860766804389356,
"grad_norm": 0.3574498417530265,
"learning_rate": 1.4912657974185418e-05,
"loss": 1.3363,
"step": 3298
},
{
"epoch": 0.5862543871340353,
"grad_norm": 0.36234777284338704,
"learning_rate": 1.4901749417467377e-05,
"loss": 1.3135,
"step": 3299
},
{
"epoch": 0.586432093829135,
"grad_norm": 0.3592970980549119,
"learning_rate": 1.4890842482598722e-05,
"loss": 1.3528,
"step": 3300
},
{
"epoch": 0.5866098005242347,
"grad_norm": 0.3665654810053385,
"learning_rate": 1.4879937173049156e-05,
"loss": 1.3636,
"step": 3301
},
{
"epoch": 0.5867875072193345,
"grad_norm": 0.36883431302419983,
"learning_rate": 1.486903349228786e-05,
"loss": 1.3721,
"step": 3302
},
{
"epoch": 0.5869652139144342,
"grad_norm": 0.3871603709956387,
"learning_rate": 1.48581314437835e-05,
"loss": 1.3361,
"step": 3303
},
{
"epoch": 0.587142920609534,
"grad_norm": 0.3678710581702569,
"learning_rate": 1.4847231031004227e-05,
"loss": 1.3245,
"step": 3304
},
{
"epoch": 0.5873206273046337,
"grad_norm": 0.35091399498488307,
"learning_rate": 1.4836332257417668e-05,
"loss": 1.3128,
"step": 3305
},
{
"epoch": 0.5874983339997334,
"grad_norm": 0.3619278392353973,
"learning_rate": 1.4825435126490924e-05,
"loss": 1.3621,
"step": 3306
},
{
"epoch": 0.5876760406948331,
"grad_norm": 0.3584833297560503,
"learning_rate": 1.4814539641690574e-05,
"loss": 1.3588,
"step": 3307
},
{
"epoch": 0.5878537473899329,
"grad_norm": 0.35742314642348666,
"learning_rate": 1.4803645806482685e-05,
"loss": 1.3692,
"step": 3308
},
{
"epoch": 0.5880314540850327,
"grad_norm": 0.3593254294467593,
"learning_rate": 1.4792753624332784e-05,
"loss": 1.3255,
"step": 3309
},
{
"epoch": 0.5882091607801324,
"grad_norm": 0.3584890643372212,
"learning_rate": 1.4781863098705891e-05,
"loss": 1.3538,
"step": 3310
},
{
"epoch": 0.5883868674752322,
"grad_norm": 0.3472485407229358,
"learning_rate": 1.477097423306647e-05,
"loss": 1.2882,
"step": 3311
},
{
"epoch": 0.5885645741703319,
"grad_norm": 0.35457528429614343,
"learning_rate": 1.4760087030878473e-05,
"loss": 1.3496,
"step": 3312
},
{
"epoch": 0.5887422808654316,
"grad_norm": 0.3642206320106112,
"learning_rate": 1.474920149560535e-05,
"loss": 1.3428,
"step": 3313
},
{
"epoch": 0.5889199875605313,
"grad_norm": 0.4674885525444342,
"learning_rate": 1.473831763070997e-05,
"loss": 1.3665,
"step": 3314
},
{
"epoch": 0.5890976942556311,
"grad_norm": 0.355436914305763,
"learning_rate": 1.47274354396547e-05,
"loss": 1.337,
"step": 3315
},
{
"epoch": 0.5892754009507308,
"grad_norm": 0.3569204713244432,
"learning_rate": 1.4716554925901374e-05,
"loss": 1.3743,
"step": 3316
},
{
"epoch": 0.5894531076458306,
"grad_norm": 0.3428640341241889,
"learning_rate": 1.470567609291128e-05,
"loss": 1.3007,
"step": 3317
},
{
"epoch": 0.5896308143409303,
"grad_norm": 0.3532062422664561,
"learning_rate": 1.469479894414519e-05,
"loss": 1.3392,
"step": 3318
},
{
"epoch": 0.58980852103603,
"grad_norm": 0.35519775472288906,
"learning_rate": 1.4683923483063325e-05,
"loss": 1.3331,
"step": 3319
},
{
"epoch": 0.5899862277311297,
"grad_norm": 0.3654162550108507,
"learning_rate": 1.4673049713125372e-05,
"loss": 1.3142,
"step": 3320
},
{
"epoch": 0.5901639344262295,
"grad_norm": 0.3480085556587412,
"learning_rate": 1.466217763779048e-05,
"loss": 1.2939,
"step": 3321
},
{
"epoch": 0.5903416411213293,
"grad_norm": 0.35086270221502536,
"learning_rate": 1.4651307260517267e-05,
"loss": 1.3607,
"step": 3322
},
{
"epoch": 0.590519347816429,
"grad_norm": 0.3533034091437995,
"learning_rate": 1.4640438584763803e-05,
"loss": 1.3514,
"step": 3323
},
{
"epoch": 0.5906970545115288,
"grad_norm": 0.34592595049040153,
"learning_rate": 1.4629571613987614e-05,
"loss": 1.3014,
"step": 3324
},
{
"epoch": 0.5908747612066284,
"grad_norm": 0.356558266252796,
"learning_rate": 1.4618706351645697e-05,
"loss": 1.3493,
"step": 3325
},
{
"epoch": 0.5910524679017282,
"grad_norm": 0.3520732851507656,
"learning_rate": 1.4607842801194476e-05,
"loss": 1.3468,
"step": 3326
},
{
"epoch": 0.5912301745968279,
"grad_norm": 0.3578469769464203,
"learning_rate": 1.459698096608987e-05,
"loss": 1.372,
"step": 3327
},
{
"epoch": 0.5914078812919277,
"grad_norm": 0.3480134635162249,
"learning_rate": 1.4586120849787228e-05,
"loss": 1.3155,
"step": 3328
},
{
"epoch": 0.5915855879870274,
"grad_norm": 0.3565644832240233,
"learning_rate": 1.4575262455741361e-05,
"loss": 1.3317,
"step": 3329
},
{
"epoch": 0.5917632946821272,
"grad_norm": 0.34852062720230415,
"learning_rate": 1.4564405787406521e-05,
"loss": 1.3298,
"step": 3330
},
{
"epoch": 0.5919410013772269,
"grad_norm": 0.3495525376955254,
"learning_rate": 1.455355084823641e-05,
"loss": 1.3344,
"step": 3331
},
{
"epoch": 0.5921187080723266,
"grad_norm": 0.34770076716242626,
"learning_rate": 1.4542697641684211e-05,
"loss": 1.2946,
"step": 3332
},
{
"epoch": 0.5922964147674263,
"grad_norm": 0.3526824993769257,
"learning_rate": 1.4531846171202522e-05,
"loss": 1.3146,
"step": 3333
},
{
"epoch": 0.5924741214625261,
"grad_norm": 0.3516366662201093,
"learning_rate": 1.4520996440243393e-05,
"loss": 1.3159,
"step": 3334
},
{
"epoch": 0.5926518281576258,
"grad_norm": 0.3519054630565006,
"learning_rate": 1.4510148452258333e-05,
"loss": 1.3282,
"step": 3335
},
{
"epoch": 0.5928295348527256,
"grad_norm": 0.3624284254546288,
"learning_rate": 1.4499302210698296e-05,
"loss": 1.3184,
"step": 3336
},
{
"epoch": 0.5930072415478254,
"grad_norm": 0.3722079674471134,
"learning_rate": 1.4488457719013671e-05,
"loss": 1.3714,
"step": 3337
},
{
"epoch": 0.593184948242925,
"grad_norm": 0.35940070403711927,
"learning_rate": 1.4477614980654294e-05,
"loss": 1.3422,
"step": 3338
},
{
"epoch": 0.5933626549380248,
"grad_norm": 0.3524020929618335,
"learning_rate": 1.4466773999069445e-05,
"loss": 1.3306,
"step": 3339
},
{
"epoch": 0.5935403616331245,
"grad_norm": 0.358864523450581,
"learning_rate": 1.445593477770784e-05,
"loss": 1.3771,
"step": 3340
},
{
"epoch": 0.5937180683282243,
"grad_norm": 0.3654822573234683,
"learning_rate": 1.4445097320017647e-05,
"loss": 1.3406,
"step": 3341
},
{
"epoch": 0.593895775023324,
"grad_norm": 0.3583539868428544,
"learning_rate": 1.443426162944646e-05,
"loss": 1.3534,
"step": 3342
},
{
"epoch": 0.5940734817184238,
"grad_norm": 0.35949593161233495,
"learning_rate": 1.4423427709441317e-05,
"loss": 1.3837,
"step": 3343
},
{
"epoch": 0.5942511884135235,
"grad_norm": 0.35513562443514374,
"learning_rate": 1.44125955634487e-05,
"loss": 1.3357,
"step": 3344
},
{
"epoch": 0.5944288951086232,
"grad_norm": 0.3602125413723429,
"learning_rate": 1.4401765194914493e-05,
"loss": 1.3927,
"step": 3345
},
{
"epoch": 0.5946066018037229,
"grad_norm": 0.3515154294057817,
"learning_rate": 1.4390936607284068e-05,
"loss": 1.3277,
"step": 3346
},
{
"epoch": 0.5947843084988227,
"grad_norm": 0.35421470608637734,
"learning_rate": 1.4380109804002196e-05,
"loss": 1.3344,
"step": 3347
},
{
"epoch": 0.5949620151939224,
"grad_norm": 0.345666755185299,
"learning_rate": 1.4369284788513077e-05,
"loss": 1.2739,
"step": 3348
},
{
"epoch": 0.5951397218890222,
"grad_norm": 0.35150015801975987,
"learning_rate": 1.4358461564260356e-05,
"loss": 1.2932,
"step": 3349
},
{
"epoch": 0.595317428584122,
"grad_norm": 0.3532111985045056,
"learning_rate": 1.4347640134687098e-05,
"loss": 1.3167,
"step": 3350
},
{
"epoch": 0.5954951352792216,
"grad_norm": 0.3492944212338751,
"learning_rate": 1.4336820503235819e-05,
"loss": 1.3109,
"step": 3351
},
{
"epoch": 0.5956728419743214,
"grad_norm": 0.3533004028402618,
"learning_rate": 1.432600267334844e-05,
"loss": 1.3523,
"step": 3352
},
{
"epoch": 0.5958505486694211,
"grad_norm": 0.34423302974491937,
"learning_rate": 1.4315186648466313e-05,
"loss": 1.323,
"step": 3353
},
{
"epoch": 0.5960282553645209,
"grad_norm": 0.3496763866858666,
"learning_rate": 1.4304372432030218e-05,
"loss": 1.3093,
"step": 3354
},
{
"epoch": 0.5962059620596206,
"grad_norm": 0.3462153687253022,
"learning_rate": 1.4293560027480367e-05,
"loss": 1.2954,
"step": 3355
},
{
"epoch": 0.5963836687547204,
"grad_norm": 0.3469812289706861,
"learning_rate": 1.4282749438256385e-05,
"loss": 1.3201,
"step": 3356
},
{
"epoch": 0.59656137544982,
"grad_norm": 0.3627836093883163,
"learning_rate": 1.4271940667797324e-05,
"loss": 1.3677,
"step": 3357
},
{
"epoch": 0.5967390821449198,
"grad_norm": 0.33862354072520096,
"learning_rate": 1.4261133719541658e-05,
"loss": 1.2909,
"step": 3358
},
{
"epoch": 0.5969167888400195,
"grad_norm": 0.34504198392449603,
"learning_rate": 1.4250328596927277e-05,
"loss": 1.3321,
"step": 3359
},
{
"epoch": 0.5970944955351193,
"grad_norm": 0.3475951302460396,
"learning_rate": 1.42395253033915e-05,
"loss": 1.3272,
"step": 3360
},
{
"epoch": 0.597272202230219,
"grad_norm": 0.34195846819097087,
"learning_rate": 1.4228723842371053e-05,
"loss": 1.312,
"step": 3361
},
{
"epoch": 0.5974499089253188,
"grad_norm": 0.34197884962262975,
"learning_rate": 1.4217924217302088e-05,
"loss": 1.3084,
"step": 3362
},
{
"epoch": 0.5976276156204186,
"grad_norm": 0.3362986933034551,
"learning_rate": 1.4207126431620171e-05,
"loss": 1.2749,
"step": 3363
},
{
"epoch": 0.5978053223155182,
"grad_norm": 0.36148542505697445,
"learning_rate": 1.419633048876026e-05,
"loss": 1.307,
"step": 3364
},
{
"epoch": 0.597983029010618,
"grad_norm": 0.36026797879051387,
"learning_rate": 1.4185536392156776e-05,
"loss": 1.3541,
"step": 3365
},
{
"epoch": 0.5981607357057177,
"grad_norm": 0.3487388711668628,
"learning_rate": 1.4174744145243513e-05,
"loss": 1.3241,
"step": 3366
},
{
"epoch": 0.5983384424008175,
"grad_norm": 0.3535850386555065,
"learning_rate": 1.4163953751453683e-05,
"loss": 1.3493,
"step": 3367
},
{
"epoch": 0.5985161490959172,
"grad_norm": 0.35088206270137384,
"learning_rate": 1.4153165214219906e-05,
"loss": 1.336,
"step": 3368
},
{
"epoch": 0.598693855791017,
"grad_norm": 0.3634355096218182,
"learning_rate": 1.4142378536974243e-05,
"loss": 1.2718,
"step": 3369
},
{
"epoch": 0.5988715624861166,
"grad_norm": 0.3513092447595911,
"learning_rate": 1.4131593723148122e-05,
"loss": 1.3591,
"step": 3370
},
{
"epoch": 0.5990492691812164,
"grad_norm": 0.3535645382717685,
"learning_rate": 1.4120810776172396e-05,
"loss": 1.3722,
"step": 3371
},
{
"epoch": 0.5992269758763161,
"grad_norm": 0.35515319156027025,
"learning_rate": 1.4110029699477327e-05,
"loss": 1.3422,
"step": 3372
},
{
"epoch": 0.5994046825714159,
"grad_norm": 0.3554626993213754,
"learning_rate": 1.409925049649257e-05,
"loss": 1.3505,
"step": 3373
},
{
"epoch": 0.5995823892665156,
"grad_norm": 0.35638173726697353,
"learning_rate": 1.4088473170647205e-05,
"loss": 1.2791,
"step": 3374
},
{
"epoch": 0.5997600959616154,
"grad_norm": 0.3426686423345436,
"learning_rate": 1.4077697725369696e-05,
"loss": 1.3008,
"step": 3375
},
{
"epoch": 0.5999378026567151,
"grad_norm": 0.3497819274354154,
"learning_rate": 1.4066924164087912e-05,
"loss": 1.3138,
"step": 3376
},
{
"epoch": 0.6001155093518148,
"grad_norm": 0.35418800693554636,
"learning_rate": 1.405615249022913e-05,
"loss": 1.3157,
"step": 3377
},
{
"epoch": 0.6002932160469145,
"grad_norm": 0.3457024172191111,
"learning_rate": 1.4045382707220014e-05,
"loss": 1.2996,
"step": 3378
},
{
"epoch": 0.6004709227420143,
"grad_norm": 0.3767809033791895,
"learning_rate": 1.4034614818486647e-05,
"loss": 1.3244,
"step": 3379
},
{
"epoch": 0.6006486294371141,
"grad_norm": 0.35324349974171776,
"learning_rate": 1.402384882745449e-05,
"loss": 1.2991,
"step": 3380
},
{
"epoch": 0.6008263361322138,
"grad_norm": 0.3948559139054159,
"learning_rate": 1.4013084737548405e-05,
"loss": 1.3447,
"step": 3381
},
{
"epoch": 0.6010040428273136,
"grad_norm": 0.35257161812593213,
"learning_rate": 1.4002322552192654e-05,
"loss": 1.3078,
"step": 3382
},
{
"epoch": 0.6011817495224132,
"grad_norm": 0.7775892006647773,
"learning_rate": 1.3991562274810891e-05,
"loss": 1.2957,
"step": 3383
},
{
"epoch": 0.601359456217513,
"grad_norm": 0.35392610879244596,
"learning_rate": 1.3980803908826164e-05,
"loss": 1.3302,
"step": 3384
},
{
"epoch": 0.6015371629126127,
"grad_norm": 0.35909837502269015,
"learning_rate": 1.397004745766091e-05,
"loss": 1.326,
"step": 3385
},
{
"epoch": 0.6017148696077125,
"grad_norm": 0.3541038114278233,
"learning_rate": 1.3959292924736958e-05,
"loss": 1.3262,
"step": 3386
},
{
"epoch": 0.6018925763028122,
"grad_norm": 0.35304389372419415,
"learning_rate": 1.3948540313475518e-05,
"loss": 1.3163,
"step": 3387
},
{
"epoch": 0.602070282997912,
"grad_norm": 0.37929502564754813,
"learning_rate": 1.393778962729722e-05,
"loss": 1.3252,
"step": 3388
},
{
"epoch": 0.6022479896930116,
"grad_norm": 0.36403888417517616,
"learning_rate": 1.3927040869622044e-05,
"loss": 1.3323,
"step": 3389
},
{
"epoch": 0.6024256963881114,
"grad_norm": 0.36093193049555194,
"learning_rate": 1.3916294043869369e-05,
"loss": 1.3195,
"step": 3390
},
{
"epoch": 0.6026034030832111,
"grad_norm": 0.3504058470447915,
"learning_rate": 1.3905549153457974e-05,
"loss": 1.338,
"step": 3391
},
{
"epoch": 0.6027811097783109,
"grad_norm": 0.35253976962035305,
"learning_rate": 1.3894806201805997e-05,
"loss": 1.2804,
"step": 3392
},
{
"epoch": 0.6029588164734107,
"grad_norm": 0.35663805495351575,
"learning_rate": 1.3884065192330985e-05,
"loss": 1.3009,
"step": 3393
},
{
"epoch": 0.6031365231685104,
"grad_norm": 0.3757383719991757,
"learning_rate": 1.387332612844985e-05,
"loss": 1.3674,
"step": 3394
},
{
"epoch": 0.6033142298636102,
"grad_norm": 0.356126654885609,
"learning_rate": 1.3862589013578894e-05,
"loss": 1.3305,
"step": 3395
},
{
"epoch": 0.6034919365587098,
"grad_norm": 0.3553041157592863,
"learning_rate": 1.3851853851133784e-05,
"loss": 1.2987,
"step": 3396
},
{
"epoch": 0.6036696432538096,
"grad_norm": 0.35669594780169706,
"learning_rate": 1.384112064452959e-05,
"loss": 1.3368,
"step": 3397
},
{
"epoch": 0.6038473499489093,
"grad_norm": 0.35954925427186796,
"learning_rate": 1.3830389397180744e-05,
"loss": 1.3469,
"step": 3398
},
{
"epoch": 0.6040250566440091,
"grad_norm": 0.3548128211337011,
"learning_rate": 1.3819660112501054e-05,
"loss": 1.2965,
"step": 3399
},
{
"epoch": 0.6042027633391088,
"grad_norm": 0.349665034973616,
"learning_rate": 1.3808932793903709e-05,
"loss": 1.2836,
"step": 3400
},
{
"epoch": 0.6043804700342086,
"grad_norm": 0.37771782808565246,
"learning_rate": 1.3798207444801267e-05,
"loss": 1.3403,
"step": 3401
},
{
"epoch": 0.6045581767293082,
"grad_norm": 0.35294652954771016,
"learning_rate": 1.378748406860567e-05,
"loss": 1.3223,
"step": 3402
},
{
"epoch": 0.604735883424408,
"grad_norm": 0.35713961194565763,
"learning_rate": 1.3776762668728224e-05,
"loss": 1.355,
"step": 3403
},
{
"epoch": 0.6049135901195077,
"grad_norm": 0.34021766228241174,
"learning_rate": 1.3766043248579605e-05,
"loss": 1.2794,
"step": 3404
},
{
"epoch": 0.6050912968146075,
"grad_norm": 0.3488045121154714,
"learning_rate": 1.3755325811569863e-05,
"loss": 1.313,
"step": 3405
},
{
"epoch": 0.6052690035097072,
"grad_norm": 0.3598559048139747,
"learning_rate": 1.3744610361108412e-05,
"loss": 1.3419,
"step": 3406
},
{
"epoch": 0.605446710204807,
"grad_norm": 0.34841228078930575,
"learning_rate": 1.373389690060405e-05,
"loss": 1.3071,
"step": 3407
},
{
"epoch": 0.6056244168999068,
"grad_norm": 0.3582960062140285,
"learning_rate": 1.3723185433464923e-05,
"loss": 1.334,
"step": 3408
},
{
"epoch": 0.6058021235950064,
"grad_norm": 0.3514805555144828,
"learning_rate": 1.3712475963098548e-05,
"loss": 1.2882,
"step": 3409
},
{
"epoch": 0.6059798302901062,
"grad_norm": 0.36118994996643794,
"learning_rate": 1.3701768492911808e-05,
"loss": 1.3654,
"step": 3410
},
{
"epoch": 0.6061575369852059,
"grad_norm": 0.36147154720464797,
"learning_rate": 1.3691063026310958e-05,
"loss": 1.3802,
"step": 3411
},
{
"epoch": 0.6063352436803057,
"grad_norm": 0.35207992759249124,
"learning_rate": 1.3680359566701605e-05,
"loss": 1.3022,
"step": 3412
},
{
"epoch": 0.6065129503754054,
"grad_norm": 0.3535034028295533,
"learning_rate": 1.3669658117488717e-05,
"loss": 1.3104,
"step": 3413
},
{
"epoch": 0.6066906570705052,
"grad_norm": 0.3457652226249156,
"learning_rate": 1.3658958682076633e-05,
"loss": 1.2883,
"step": 3414
},
{
"epoch": 0.6068683637656048,
"grad_norm": 0.3507966248708197,
"learning_rate": 1.3648261263869036e-05,
"loss": 1.3456,
"step": 3415
},
{
"epoch": 0.6070460704607046,
"grad_norm": 0.3621360022205465,
"learning_rate": 1.3637565866268985e-05,
"loss": 1.3546,
"step": 3416
},
{
"epoch": 0.6072237771558043,
"grad_norm": 0.36235380086983665,
"learning_rate": 1.362687249267888e-05,
"loss": 1.3777,
"step": 3417
},
{
"epoch": 0.6074014838509041,
"grad_norm": 0.35765822328946123,
"learning_rate": 1.361618114650049e-05,
"loss": 1.3719,
"step": 3418
},
{
"epoch": 0.6075791905460038,
"grad_norm": 0.3564972731008636,
"learning_rate": 1.3605491831134936e-05,
"loss": 1.3328,
"step": 3419
},
{
"epoch": 0.6077568972411036,
"grad_norm": 0.38162366136796994,
"learning_rate": 1.3594804549982667e-05,
"loss": 1.3079,
"step": 3420
},
{
"epoch": 0.6079346039362032,
"grad_norm": 0.3502623701677519,
"learning_rate": 1.358411930644354e-05,
"loss": 1.3463,
"step": 3421
},
{
"epoch": 0.608112310631303,
"grad_norm": 0.3478311662147964,
"learning_rate": 1.3573436103916712e-05,
"loss": 1.2918,
"step": 3422
},
{
"epoch": 0.6082900173264028,
"grad_norm": 0.34333915102124024,
"learning_rate": 1.3562754945800725e-05,
"loss": 1.3216,
"step": 3423
},
{
"epoch": 0.6084677240215025,
"grad_norm": 0.36154332608882356,
"learning_rate": 1.3552075835493433e-05,
"loss": 1.3975,
"step": 3424
},
{
"epoch": 0.6086454307166023,
"grad_norm": 0.35238874630176287,
"learning_rate": 1.3541398776392085e-05,
"loss": 1.3285,
"step": 3425
},
{
"epoch": 0.608823137411702,
"grad_norm": 0.34650775810937595,
"learning_rate": 1.3530723771893248e-05,
"loss": 1.3279,
"step": 3426
},
{
"epoch": 0.6090008441068018,
"grad_norm": 0.3458946393733133,
"learning_rate": 1.3520050825392839e-05,
"loss": 1.3295,
"step": 3427
},
{
"epoch": 0.6091785508019014,
"grad_norm": 0.3660695913501381,
"learning_rate": 1.350937994028612e-05,
"loss": 1.3774,
"step": 3428
},
{
"epoch": 0.6093562574970012,
"grad_norm": 0.3579558570309594,
"learning_rate": 1.34987111199677e-05,
"loss": 1.3388,
"step": 3429
},
{
"epoch": 0.6095339641921009,
"grad_norm": 0.3440428892053028,
"learning_rate": 1.348804436783154e-05,
"loss": 1.2819,
"step": 3430
},
{
"epoch": 0.6097116708872007,
"grad_norm": 0.35047490870339304,
"learning_rate": 1.347737968727093e-05,
"loss": 1.3184,
"step": 3431
},
{
"epoch": 0.6098893775823004,
"grad_norm": 0.36302077145323347,
"learning_rate": 1.3466717081678504e-05,
"loss": 1.3336,
"step": 3432
},
{
"epoch": 0.6100670842774002,
"grad_norm": 0.35930584300584023,
"learning_rate": 1.345605655444624e-05,
"loss": 1.3792,
"step": 3433
},
{
"epoch": 0.6102447909724998,
"grad_norm": 0.34850453761408146,
"learning_rate": 1.3445398108965443e-05,
"loss": 1.3076,
"step": 3434
},
{
"epoch": 0.6104224976675996,
"grad_norm": 0.35930769006601837,
"learning_rate": 1.3434741748626778e-05,
"loss": 1.3561,
"step": 3435
},
{
"epoch": 0.6106002043626994,
"grad_norm": 0.3598028686972987,
"learning_rate": 1.3424087476820228e-05,
"loss": 1.3436,
"step": 3436
},
{
"epoch": 0.6107779110577991,
"grad_norm": 0.3556321628638517,
"learning_rate": 1.341343529693512e-05,
"loss": 1.3442,
"step": 3437
},
{
"epoch": 0.6109556177528989,
"grad_norm": 0.358092107483723,
"learning_rate": 1.3402785212360102e-05,
"loss": 1.3531,
"step": 3438
},
{
"epoch": 0.6111333244479986,
"grad_norm": 0.3579478154837523,
"learning_rate": 1.3392137226483179e-05,
"loss": 1.3688,
"step": 3439
},
{
"epoch": 0.6113110311430984,
"grad_norm": 0.3595318314443169,
"learning_rate": 1.3381491342691671e-05,
"loss": 1.3323,
"step": 3440
},
{
"epoch": 0.611488737838198,
"grad_norm": 0.3434803347418904,
"learning_rate": 1.3370847564372238e-05,
"loss": 1.2688,
"step": 3441
},
{
"epoch": 0.6116664445332978,
"grad_norm": 0.34716539881076947,
"learning_rate": 1.3360205894910858e-05,
"loss": 1.3338,
"step": 3442
},
{
"epoch": 0.6118441512283975,
"grad_norm": 0.3558241649399973,
"learning_rate": 1.3349566337692841e-05,
"loss": 1.3463,
"step": 3443
},
{
"epoch": 0.6120218579234973,
"grad_norm": 0.35014843448718375,
"learning_rate": 1.3338928896102847e-05,
"loss": 1.3344,
"step": 3444
},
{
"epoch": 0.612199564618597,
"grad_norm": 0.36066684304118046,
"learning_rate": 1.3328293573524841e-05,
"loss": 1.3096,
"step": 3445
},
{
"epoch": 0.6123772713136968,
"grad_norm": 0.35088320614217394,
"learning_rate": 1.3317660373342115e-05,
"loss": 1.2987,
"step": 3446
},
{
"epoch": 0.6125549780087964,
"grad_norm": 0.3452658746609516,
"learning_rate": 1.3307029298937288e-05,
"loss": 1.2939,
"step": 3447
},
{
"epoch": 0.6127326847038962,
"grad_norm": 0.34319440469889845,
"learning_rate": 1.3296400353692307e-05,
"loss": 1.2773,
"step": 3448
},
{
"epoch": 0.612910391398996,
"grad_norm": 0.35817010320683185,
"learning_rate": 1.3285773540988443e-05,
"loss": 1.3159,
"step": 3449
},
{
"epoch": 0.6130880980940957,
"grad_norm": 0.35515782252851075,
"learning_rate": 1.3275148864206283e-05,
"loss": 1.29,
"step": 3450
},
{
"epoch": 0.6132658047891955,
"grad_norm": 0.35487950706352067,
"learning_rate": 1.3264526326725735e-05,
"loss": 1.3477,
"step": 3451
},
{
"epoch": 0.6134435114842952,
"grad_norm": 0.34588878820892416,
"learning_rate": 1.3253905931926025e-05,
"loss": 1.3075,
"step": 3452
},
{
"epoch": 0.6136212181793949,
"grad_norm": 0.34397424681541394,
"learning_rate": 1.3243287683185708e-05,
"loss": 1.3156,
"step": 3453
},
{
"epoch": 0.6137989248744946,
"grad_norm": 0.35366375932630956,
"learning_rate": 1.3232671583882645e-05,
"loss": 1.3046,
"step": 3454
},
{
"epoch": 0.6139766315695944,
"grad_norm": 0.3603361334571491,
"learning_rate": 1.3222057637394016e-05,
"loss": 1.3399,
"step": 3455
},
{
"epoch": 0.6141543382646941,
"grad_norm": 0.3648964889968698,
"learning_rate": 1.3211445847096319e-05,
"loss": 1.3578,
"step": 3456
},
{
"epoch": 0.6143320449597939,
"grad_norm": 0.3514922925320543,
"learning_rate": 1.3200836216365357e-05,
"loss": 1.3464,
"step": 3457
},
{
"epoch": 0.6145097516548936,
"grad_norm": 0.34969958902453024,
"learning_rate": 1.3190228748576265e-05,
"loss": 1.3056,
"step": 3458
},
{
"epoch": 0.6146874583499934,
"grad_norm": 0.35090713872527757,
"learning_rate": 1.3179623447103466e-05,
"loss": 1.3364,
"step": 3459
},
{
"epoch": 0.614865165045093,
"grad_norm": 0.3513232004757053,
"learning_rate": 1.316902031532072e-05,
"loss": 1.3175,
"step": 3460
},
{
"epoch": 0.6150428717401928,
"grad_norm": 0.3384327118243738,
"learning_rate": 1.3158419356601069e-05,
"loss": 1.3057,
"step": 3461
},
{
"epoch": 0.6152205784352925,
"grad_norm": 0.34630904505240434,
"learning_rate": 1.3147820574316874e-05,
"loss": 1.3079,
"step": 3462
},
{
"epoch": 0.6153982851303923,
"grad_norm": 0.3513758096633715,
"learning_rate": 1.3137223971839823e-05,
"loss": 1.3341,
"step": 3463
},
{
"epoch": 0.615575991825492,
"grad_norm": 0.3498204131593602,
"learning_rate": 1.3126629552540893e-05,
"loss": 1.3385,
"step": 3464
},
{
"epoch": 0.6157536985205918,
"grad_norm": 0.35092940903969533,
"learning_rate": 1.3116037319790356e-05,
"loss": 1.3055,
"step": 3465
},
{
"epoch": 0.6159314052156915,
"grad_norm": 0.35344248496148667,
"learning_rate": 1.3105447276957798e-05,
"loss": 1.325,
"step": 3466
},
{
"epoch": 0.6161091119107912,
"grad_norm": 0.35202575310436035,
"learning_rate": 1.3094859427412132e-05,
"loss": 1.3039,
"step": 3467
},
{
"epoch": 0.616286818605891,
"grad_norm": 0.3483695842530992,
"learning_rate": 1.3084273774521534e-05,
"loss": 1.3468,
"step": 3468
},
{
"epoch": 0.6164645253009907,
"grad_norm": 0.3959357905709929,
"learning_rate": 1.3073690321653505e-05,
"loss": 1.298,
"step": 3469
},
{
"epoch": 0.6166422319960905,
"grad_norm": 0.34248997144547194,
"learning_rate": 1.3063109072174842e-05,
"loss": 1.3061,
"step": 3470
},
{
"epoch": 0.6168199386911902,
"grad_norm": 0.35193713015110695,
"learning_rate": 1.3052530029451628e-05,
"loss": 1.3444,
"step": 3471
},
{
"epoch": 0.61699764538629,
"grad_norm": 0.3610908653827003,
"learning_rate": 1.3041953196849276e-05,
"loss": 1.34,
"step": 3472
},
{
"epoch": 0.6171753520813896,
"grad_norm": 0.36302864356259795,
"learning_rate": 1.3031378577732459e-05,
"loss": 1.3565,
"step": 3473
},
{
"epoch": 0.6173530587764894,
"grad_norm": 0.35684466712961455,
"learning_rate": 1.3020806175465168e-05,
"loss": 1.325,
"step": 3474
},
{
"epoch": 0.6175307654715891,
"grad_norm": 0.3476267205029857,
"learning_rate": 1.3010235993410683e-05,
"loss": 1.3226,
"step": 3475
},
{
"epoch": 0.6177084721666889,
"grad_norm": 0.3541495132971835,
"learning_rate": 1.299966803493157e-05,
"loss": 1.3071,
"step": 3476
},
{
"epoch": 0.6178861788617886,
"grad_norm": 0.3530855538034206,
"learning_rate": 1.2989102303389708e-05,
"loss": 1.3172,
"step": 3477
},
{
"epoch": 0.6180638855568884,
"grad_norm": 0.352741230389903,
"learning_rate": 1.2978538802146252e-05,
"loss": 1.3087,
"step": 3478
},
{
"epoch": 0.618241592251988,
"grad_norm": 0.35098827117813136,
"learning_rate": 1.2967977534561648e-05,
"loss": 1.3086,
"step": 3479
},
{
"epoch": 0.6184192989470878,
"grad_norm": 0.36053657354714624,
"learning_rate": 1.2957418503995625e-05,
"loss": 1.3706,
"step": 3480
},
{
"epoch": 0.6185970056421876,
"grad_norm": 0.3573251621263162,
"learning_rate": 1.2946861713807222e-05,
"loss": 1.357,
"step": 3481
},
{
"epoch": 0.6187747123372873,
"grad_norm": 0.3648854987431012,
"learning_rate": 1.2936307167354753e-05,
"loss": 1.3206,
"step": 3482
},
{
"epoch": 0.6189524190323871,
"grad_norm": 0.35075119725142023,
"learning_rate": 1.292575486799581e-05,
"loss": 1.3333,
"step": 3483
},
{
"epoch": 0.6191301257274868,
"grad_norm": 0.3576639539935488,
"learning_rate": 1.291520481908728e-05,
"loss": 1.3479,
"step": 3484
},
{
"epoch": 0.6193078324225865,
"grad_norm": 0.3408548933739274,
"learning_rate": 1.2904657023985323e-05,
"loss": 1.3139,
"step": 3485
},
{
"epoch": 0.6194855391176862,
"grad_norm": 0.34827712255318927,
"learning_rate": 1.2894111486045416e-05,
"loss": 1.3164,
"step": 3486
},
{
"epoch": 0.619663245812786,
"grad_norm": 0.3530284512740369,
"learning_rate": 1.288356820862227e-05,
"loss": 1.3021,
"step": 3487
},
{
"epoch": 0.6198409525078857,
"grad_norm": 0.3487448358288206,
"learning_rate": 1.287302719506991e-05,
"loss": 1.3278,
"step": 3488
},
{
"epoch": 0.6200186592029855,
"grad_norm": 0.3546498559937753,
"learning_rate": 1.2862488448741623e-05,
"loss": 1.352,
"step": 3489
},
{
"epoch": 0.6201963658980852,
"grad_norm": 0.3414282080083184,
"learning_rate": 1.2851951972989988e-05,
"loss": 1.2896,
"step": 3490
},
{
"epoch": 0.620374072593185,
"grad_norm": 0.3481440177399318,
"learning_rate": 1.284141777116686e-05,
"loss": 1.3355,
"step": 3491
},
{
"epoch": 0.6205517792882846,
"grad_norm": 0.3555978004518865,
"learning_rate": 1.283088584662336e-05,
"loss": 1.345,
"step": 3492
},
{
"epoch": 0.6207294859833844,
"grad_norm": 0.3487013872756217,
"learning_rate": 1.2820356202709893e-05,
"loss": 1.3406,
"step": 3493
},
{
"epoch": 0.6209071926784842,
"grad_norm": 0.36431135589082125,
"learning_rate": 1.2809828842776135e-05,
"loss": 1.3666,
"step": 3494
},
{
"epoch": 0.6210848993735839,
"grad_norm": 0.3415748872215295,
"learning_rate": 1.2799303770171043e-05,
"loss": 1.3024,
"step": 3495
},
{
"epoch": 0.6212626060686837,
"grad_norm": 0.346755002742322,
"learning_rate": 1.2788780988242837e-05,
"loss": 1.3282,
"step": 3496
},
{
"epoch": 0.6214403127637834,
"grad_norm": 0.34617069790531374,
"learning_rate": 1.2778260500339013e-05,
"loss": 1.2872,
"step": 3497
},
{
"epoch": 0.6216180194588831,
"grad_norm": 0.3511723114281262,
"learning_rate": 1.2767742309806335e-05,
"loss": 1.3256,
"step": 3498
},
{
"epoch": 0.6217957261539828,
"grad_norm": 0.3560121169532296,
"learning_rate": 1.275722641999083e-05,
"loss": 1.3038,
"step": 3499
},
{
"epoch": 0.6219734328490826,
"grad_norm": 0.3428252230527555,
"learning_rate": 1.274671283423781e-05,
"loss": 1.3362,
"step": 3500
},
{
"epoch": 0.6221511395441823,
"grad_norm": 0.35595904506239096,
"learning_rate": 1.273620155589185e-05,
"loss": 1.3577,
"step": 3501
},
{
"epoch": 0.6223288462392821,
"grad_norm": 0.35364506147165736,
"learning_rate": 1.2725692588296768e-05,
"loss": 1.319,
"step": 3502
},
{
"epoch": 0.6225065529343818,
"grad_norm": 0.36371012528566943,
"learning_rate": 1.2715185934795678e-05,
"loss": 1.3025,
"step": 3503
},
{
"epoch": 0.6226842596294816,
"grad_norm": 0.3525034334520896,
"learning_rate": 1.2704681598730933e-05,
"loss": 1.3418,
"step": 3504
},
{
"epoch": 0.6228619663245812,
"grad_norm": 0.36707531277622457,
"learning_rate": 1.269417958344417e-05,
"loss": 1.3492,
"step": 3505
},
{
"epoch": 0.623039673019681,
"grad_norm": 0.3527874038572763,
"learning_rate": 1.2683679892276275e-05,
"loss": 1.3248,
"step": 3506
},
{
"epoch": 0.6232173797147808,
"grad_norm": 0.3805042949176451,
"learning_rate": 1.2673182528567394e-05,
"loss": 1.2803,
"step": 3507
},
{
"epoch": 0.6233950864098805,
"grad_norm": 0.3553290720721577,
"learning_rate": 1.2662687495656934e-05,
"loss": 1.3828,
"step": 3508
},
{
"epoch": 0.6235727931049803,
"grad_norm": 0.3502024829437212,
"learning_rate": 1.265219479688357e-05,
"loss": 1.3314,
"step": 3509
},
{
"epoch": 0.62375049980008,
"grad_norm": 0.36167146615962387,
"learning_rate": 1.2641704435585225e-05,
"loss": 1.377,
"step": 3510
},
{
"epoch": 0.6239282064951797,
"grad_norm": 0.3541914340991388,
"learning_rate": 1.2631216415099074e-05,
"loss": 1.3184,
"step": 3511
},
{
"epoch": 0.6241059131902794,
"grad_norm": 0.3511230442086655,
"learning_rate": 1.262073073876156e-05,
"loss": 1.32,
"step": 3512
},
{
"epoch": 0.6242836198853792,
"grad_norm": 0.348281527612528,
"learning_rate": 1.2610247409908368e-05,
"loss": 1.3026,
"step": 3513
},
{
"epoch": 0.6244613265804789,
"grad_norm": 0.5010936057062909,
"learning_rate": 1.2599766431874447e-05,
"loss": 1.3319,
"step": 3514
},
{
"epoch": 0.6246390332755787,
"grad_norm": 0.34982481338698207,
"learning_rate": 1.2589287807993994e-05,
"loss": 1.3028,
"step": 3515
},
{
"epoch": 0.6248167399706784,
"grad_norm": 0.35271161337064644,
"learning_rate": 1.2578811541600455e-05,
"loss": 1.3313,
"step": 3516
},
{
"epoch": 0.6249944466657781,
"grad_norm": 0.3500532385358204,
"learning_rate": 1.2568337636026526e-05,
"loss": 1.3269,
"step": 3517
},
{
"epoch": 0.6251721533608778,
"grad_norm": 0.35177103555905587,
"learning_rate": 1.2557866094604147e-05,
"loss": 1.3519,
"step": 3518
},
{
"epoch": 0.6253498600559776,
"grad_norm": 0.3584582389565658,
"learning_rate": 1.2547396920664524e-05,
"loss": 1.3183,
"step": 3519
},
{
"epoch": 0.6255275667510773,
"grad_norm": 0.34713114070942014,
"learning_rate": 1.2536930117538097e-05,
"loss": 1.3282,
"step": 3520
},
{
"epoch": 0.6257052734461771,
"grad_norm": 0.3505607946810754,
"learning_rate": 1.2526465688554543e-05,
"loss": 1.2847,
"step": 3521
},
{
"epoch": 0.6258829801412769,
"grad_norm": 0.3512982570310662,
"learning_rate": 1.2516003637042795e-05,
"loss": 1.3364,
"step": 3522
},
{
"epoch": 0.6260606868363766,
"grad_norm": 0.34666749618680676,
"learning_rate": 1.2505543966331045e-05,
"loss": 1.3096,
"step": 3523
},
{
"epoch": 0.6262383935314763,
"grad_norm": 0.34523876294436856,
"learning_rate": 1.2495086679746693e-05,
"loss": 1.3247,
"step": 3524
},
{
"epoch": 0.626416100226576,
"grad_norm": 0.35131784447409253,
"learning_rate": 1.2484631780616405e-05,
"loss": 1.2967,
"step": 3525
},
{
"epoch": 0.6265938069216758,
"grad_norm": 0.34895807051574573,
"learning_rate": 1.247417927226608e-05,
"loss": 1.316,
"step": 3526
},
{
"epoch": 0.6267715136167755,
"grad_norm": 0.3435311453437375,
"learning_rate": 1.2463729158020854e-05,
"loss": 1.3212,
"step": 3527
},
{
"epoch": 0.6269492203118753,
"grad_norm": 0.3504293859491529,
"learning_rate": 1.2453281441205115e-05,
"loss": 1.2841,
"step": 3528
},
{
"epoch": 0.627126927006975,
"grad_norm": 0.3452047492994417,
"learning_rate": 1.2442836125142468e-05,
"loss": 1.3029,
"step": 3529
},
{
"epoch": 0.6273046337020747,
"grad_norm": 0.3538870466975276,
"learning_rate": 1.243239321315577e-05,
"loss": 1.3334,
"step": 3530
},
{
"epoch": 0.6274823403971744,
"grad_norm": 0.4329482976640203,
"learning_rate": 1.2421952708567107e-05,
"loss": 1.3481,
"step": 3531
},
{
"epoch": 0.6276600470922742,
"grad_norm": 0.36199803471473163,
"learning_rate": 1.2411514614697798e-05,
"loss": 1.3566,
"step": 3532
},
{
"epoch": 0.6278377537873739,
"grad_norm": 0.3556286416595564,
"learning_rate": 1.2401078934868397e-05,
"loss": 1.3486,
"step": 3533
},
{
"epoch": 0.6280154604824737,
"grad_norm": 0.35600530329186614,
"learning_rate": 1.2390645672398693e-05,
"loss": 1.3646,
"step": 3534
},
{
"epoch": 0.6281931671775735,
"grad_norm": 0.3479904072114199,
"learning_rate": 1.2380214830607705e-05,
"loss": 1.3184,
"step": 3535
},
{
"epoch": 0.6283708738726732,
"grad_norm": 0.3660021801651203,
"learning_rate": 1.236978641281366e-05,
"loss": 1.3155,
"step": 3536
},
{
"epoch": 0.6285485805677729,
"grad_norm": 0.34826092642233225,
"learning_rate": 1.2359360422334064e-05,
"loss": 1.3026,
"step": 3537
},
{
"epoch": 0.6287262872628726,
"grad_norm": 0.3589388236277339,
"learning_rate": 1.2348936862485603e-05,
"loss": 1.2885,
"step": 3538
},
{
"epoch": 0.6289039939579724,
"grad_norm": 0.34406238927711147,
"learning_rate": 1.2338515736584213e-05,
"loss": 1.3149,
"step": 3539
},
{
"epoch": 0.6290817006530721,
"grad_norm": 0.34534939258453673,
"learning_rate": 1.2328097047945046e-05,
"loss": 1.3082,
"step": 3540
},
{
"epoch": 0.6292594073481719,
"grad_norm": 0.3479698754552064,
"learning_rate": 1.2317680799882478e-05,
"loss": 1.3089,
"step": 3541
},
{
"epoch": 0.6294371140432716,
"grad_norm": 0.3522472221669943,
"learning_rate": 1.230726699571013e-05,
"loss": 1.3436,
"step": 3542
},
{
"epoch": 0.6296148207383713,
"grad_norm": 0.3421187408811723,
"learning_rate": 1.2296855638740816e-05,
"loss": 1.3339,
"step": 3543
},
{
"epoch": 0.629792527433471,
"grad_norm": 0.371775519640679,
"learning_rate": 1.2286446732286587e-05,
"loss": 1.334,
"step": 3544
},
{
"epoch": 0.6299702341285708,
"grad_norm": 0.3516808475528218,
"learning_rate": 1.2276040279658714e-05,
"loss": 1.3459,
"step": 3545
},
{
"epoch": 0.6301479408236705,
"grad_norm": 0.34433514983690117,
"learning_rate": 1.2265636284167677e-05,
"loss": 1.3039,
"step": 3546
},
{
"epoch": 0.6303256475187703,
"grad_norm": 0.3595187422589055,
"learning_rate": 1.2255234749123195e-05,
"loss": 1.3402,
"step": 3547
},
{
"epoch": 0.63050335421387,
"grad_norm": 0.35500381815338017,
"learning_rate": 1.2244835677834183e-05,
"loss": 1.3545,
"step": 3548
},
{
"epoch": 0.6306810609089697,
"grad_norm": 0.35227984564181875,
"learning_rate": 1.223443907360879e-05,
"loss": 1.3279,
"step": 3549
},
{
"epoch": 0.6308587676040694,
"grad_norm": 0.3539012241116081,
"learning_rate": 1.2224044939754358e-05,
"loss": 1.366,
"step": 3550
},
{
"epoch": 0.6310364742991692,
"grad_norm": 0.3535723377466579,
"learning_rate": 1.2213653279577469e-05,
"loss": 1.3329,
"step": 3551
},
{
"epoch": 0.631214180994269,
"grad_norm": 0.3412230079184341,
"learning_rate": 1.22032640963839e-05,
"loss": 1.3111,
"step": 3552
},
{
"epoch": 0.6313918876893687,
"grad_norm": 0.3489158241352377,
"learning_rate": 1.2192877393478646e-05,
"loss": 1.3258,
"step": 3553
},
{
"epoch": 0.6315695943844685,
"grad_norm": 0.3502172954682903,
"learning_rate": 1.2182493174165917e-05,
"loss": 1.263,
"step": 3554
},
{
"epoch": 0.6317473010795682,
"grad_norm": 0.35442301222365435,
"learning_rate": 1.217211144174911e-05,
"loss": 1.3474,
"step": 3555
},
{
"epoch": 0.6319250077746679,
"grad_norm": 0.3447772063140135,
"learning_rate": 1.2161732199530874e-05,
"loss": 1.2988,
"step": 3556
},
{
"epoch": 0.6321027144697676,
"grad_norm": 0.359170476528373,
"learning_rate": 1.2151355450813032e-05,
"loss": 1.298,
"step": 3557
},
{
"epoch": 0.6322804211648674,
"grad_norm": 0.3889303169393706,
"learning_rate": 1.2140981198896622e-05,
"loss": 1.3322,
"step": 3558
},
{
"epoch": 0.6324581278599671,
"grad_norm": 0.34100859278037526,
"learning_rate": 1.2130609447081887e-05,
"loss": 1.2805,
"step": 3559
},
{
"epoch": 0.6326358345550669,
"grad_norm": 0.3371161733492959,
"learning_rate": 1.2120240198668267e-05,
"loss": 1.2838,
"step": 3560
},
{
"epoch": 0.6328135412501666,
"grad_norm": 0.3544709930045916,
"learning_rate": 1.2109873456954438e-05,
"loss": 1.328,
"step": 3561
},
{
"epoch": 0.6329912479452663,
"grad_norm": 0.3499247543553148,
"learning_rate": 1.2099509225238242e-05,
"loss": 1.3194,
"step": 3562
},
{
"epoch": 0.633168954640366,
"grad_norm": 0.3500964827390187,
"learning_rate": 1.2089147506816732e-05,
"loss": 1.3296,
"step": 3563
},
{
"epoch": 0.6333466613354658,
"grad_norm": 0.3539910136551971,
"learning_rate": 1.2078788304986168e-05,
"loss": 1.3238,
"step": 3564
},
{
"epoch": 0.6335243680305656,
"grad_norm": 0.35891071751053333,
"learning_rate": 1.2068431623042014e-05,
"loss": 1.338,
"step": 3565
},
{
"epoch": 0.6337020747256653,
"grad_norm": 0.3638873067049612,
"learning_rate": 1.205807746427892e-05,
"loss": 1.3636,
"step": 3566
},
{
"epoch": 0.6338797814207651,
"grad_norm": 0.6027654256216256,
"learning_rate": 1.2047725831990741e-05,
"loss": 1.3387,
"step": 3567
},
{
"epoch": 0.6340574881158648,
"grad_norm": 0.33576689188154824,
"learning_rate": 1.2037376729470522e-05,
"loss": 1.2514,
"step": 3568
},
{
"epoch": 0.6342351948109645,
"grad_norm": 0.3472133122929411,
"learning_rate": 1.2027030160010504e-05,
"loss": 1.3005,
"step": 3569
},
{
"epoch": 0.6344129015060642,
"grad_norm": 0.3587314884749791,
"learning_rate": 1.2016686126902137e-05,
"loss": 1.3292,
"step": 3570
},
{
"epoch": 0.634590608201164,
"grad_norm": 0.3572555695033759,
"learning_rate": 1.2006344633436045e-05,
"loss": 1.3583,
"step": 3571
},
{
"epoch": 0.6347683148962637,
"grad_norm": 0.352422896827629,
"learning_rate": 1.1996005682902054e-05,
"loss": 1.3536,
"step": 3572
},
{
"epoch": 0.6349460215913635,
"grad_norm": 0.3451986400193679,
"learning_rate": 1.198566927858918e-05,
"loss": 1.3127,
"step": 3573
},
{
"epoch": 0.6351237282864632,
"grad_norm": 0.35802191205505696,
"learning_rate": 1.1975335423785613e-05,
"loss": 1.3504,
"step": 3574
},
{
"epoch": 0.6353014349815629,
"grad_norm": 0.35446832909500153,
"learning_rate": 1.1965004121778768e-05,
"loss": 1.2733,
"step": 3575
},
{
"epoch": 0.6354791416766626,
"grad_norm": 0.35319387761890314,
"learning_rate": 1.1954675375855216e-05,
"loss": 1.3071,
"step": 3576
},
{
"epoch": 0.6356568483717624,
"grad_norm": 0.35483079117715055,
"learning_rate": 1.1944349189300726e-05,
"loss": 1.329,
"step": 3577
},
{
"epoch": 0.6358345550668622,
"grad_norm": 0.35331841236658873,
"learning_rate": 1.1934025565400242e-05,
"loss": 1.3075,
"step": 3578
},
{
"epoch": 0.6360122617619619,
"grad_norm": 0.3522955931944837,
"learning_rate": 1.1923704507437921e-05,
"loss": 1.281,
"step": 3579
},
{
"epoch": 0.6361899684570617,
"grad_norm": 0.36626422839522904,
"learning_rate": 1.1913386018697084e-05,
"loss": 1.3183,
"step": 3580
},
{
"epoch": 0.6363676751521613,
"grad_norm": 0.3566237535774367,
"learning_rate": 1.1903070102460225e-05,
"loss": 1.3362,
"step": 3581
},
{
"epoch": 0.6365453818472611,
"grad_norm": 0.3606735180431276,
"learning_rate": 1.1892756762009037e-05,
"loss": 1.3697,
"step": 3582
},
{
"epoch": 0.6367230885423608,
"grad_norm": 0.3544471811114998,
"learning_rate": 1.1882446000624381e-05,
"loss": 1.3473,
"step": 3583
},
{
"epoch": 0.6369007952374606,
"grad_norm": 0.34911300255163463,
"learning_rate": 1.1872137821586313e-05,
"loss": 1.2832,
"step": 3584
},
{
"epoch": 0.6370785019325603,
"grad_norm": 0.35043064903490645,
"learning_rate": 1.1861832228174057e-05,
"loss": 1.3224,
"step": 3585
},
{
"epoch": 0.6372562086276601,
"grad_norm": 0.3579147549809315,
"learning_rate": 1.1851529223666013e-05,
"loss": 1.2834,
"step": 3586
},
{
"epoch": 0.6374339153227598,
"grad_norm": 0.35684170651571745,
"learning_rate": 1.1841228811339764e-05,
"loss": 1.36,
"step": 3587
},
{
"epoch": 0.6376116220178595,
"grad_norm": 0.36910686942563103,
"learning_rate": 1.1830930994472057e-05,
"loss": 1.3324,
"step": 3588
},
{
"epoch": 0.6377893287129592,
"grad_norm": 0.35431582769181197,
"learning_rate": 1.182063577633883e-05,
"loss": 1.3193,
"step": 3589
},
{
"epoch": 0.637967035408059,
"grad_norm": 0.3662434623903684,
"learning_rate": 1.1810343160215183e-05,
"loss": 1.3446,
"step": 3590
},
{
"epoch": 0.6381447421031587,
"grad_norm": 0.3533396012906131,
"learning_rate": 1.1800053149375392e-05,
"loss": 1.3519,
"step": 3591
},
{
"epoch": 0.6383224487982585,
"grad_norm": 0.3546451578524887,
"learning_rate": 1.1789765747092896e-05,
"loss": 1.305,
"step": 3592
},
{
"epoch": 0.6385001554933583,
"grad_norm": 0.35417807509773597,
"learning_rate": 1.1779480956640322e-05,
"loss": 1.3334,
"step": 3593
},
{
"epoch": 0.6386778621884579,
"grad_norm": 0.3475884479077818,
"learning_rate": 1.1769198781289445e-05,
"loss": 1.318,
"step": 3594
},
{
"epoch": 0.6388555688835577,
"grad_norm": 0.35800434905644596,
"learning_rate": 1.175891922431123e-05,
"loss": 1.3127,
"step": 3595
},
{
"epoch": 0.6390332755786574,
"grad_norm": 0.3522202519786556,
"learning_rate": 1.1748642288975786e-05,
"loss": 1.3135,
"step": 3596
},
{
"epoch": 0.6392109822737572,
"grad_norm": 0.36084947516536625,
"learning_rate": 1.1738367978552394e-05,
"loss": 1.3278,
"step": 3597
},
{
"epoch": 0.6393886889688569,
"grad_norm": 0.3512062462924096,
"learning_rate": 1.1728096296309528e-05,
"loss": 1.3195,
"step": 3598
},
{
"epoch": 0.6395663956639567,
"grad_norm": 0.35289477670882013,
"learning_rate": 1.1717827245514787e-05,
"loss": 1.346,
"step": 3599
},
{
"epoch": 0.6397441023590564,
"grad_norm": 0.3527175209273952,
"learning_rate": 1.1707560829434952e-05,
"loss": 1.3143,
"step": 3600
},
{
"epoch": 0.6399218090541561,
"grad_norm": 0.3640141946618361,
"learning_rate": 1.1697297051335962e-05,
"loss": 1.3037,
"step": 3601
},
{
"epoch": 0.6400995157492558,
"grad_norm": 0.34995605232997207,
"learning_rate": 1.1687035914482919e-05,
"loss": 1.3098,
"step": 3602
},
{
"epoch": 0.6402772224443556,
"grad_norm": 0.3482300519280768,
"learning_rate": 1.1676777422140079e-05,
"loss": 1.3009,
"step": 3603
},
{
"epoch": 0.6404549291394553,
"grad_norm": 0.35937732045522675,
"learning_rate": 1.1666521577570875e-05,
"loss": 1.3508,
"step": 3604
},
{
"epoch": 0.6406326358345551,
"grad_norm": 0.35927073486236644,
"learning_rate": 1.165626838403788e-05,
"loss": 1.3194,
"step": 3605
},
{
"epoch": 0.6408103425296549,
"grad_norm": 0.3504476687300422,
"learning_rate": 1.1646017844802818e-05,
"loss": 1.3343,
"step": 3606
},
{
"epoch": 0.6409880492247545,
"grad_norm": 0.3490767533627115,
"learning_rate": 1.1635769963126573e-05,
"loss": 1.3319,
"step": 3607
},
{
"epoch": 0.6411657559198543,
"grad_norm": 0.3458004044475459,
"learning_rate": 1.1625524742269207e-05,
"loss": 1.3104,
"step": 3608
},
{
"epoch": 0.641343462614954,
"grad_norm": 0.35242397859525815,
"learning_rate": 1.1615282185489912e-05,
"loss": 1.3265,
"step": 3609
},
{
"epoch": 0.6415211693100538,
"grad_norm": 0.34601854464718834,
"learning_rate": 1.1605042296047034e-05,
"loss": 1.3115,
"step": 3610
},
{
"epoch": 0.6416988760051535,
"grad_norm": 0.3661617334284252,
"learning_rate": 1.1594805077198074e-05,
"loss": 1.2955,
"step": 3611
},
{
"epoch": 0.6418765827002533,
"grad_norm": 0.3684702839274687,
"learning_rate": 1.1584570532199688e-05,
"loss": 1.3931,
"step": 3612
},
{
"epoch": 0.6420542893953529,
"grad_norm": 0.34788510855090826,
"learning_rate": 1.1574338664307674e-05,
"loss": 1.2869,
"step": 3613
},
{
"epoch": 0.6422319960904527,
"grad_norm": 0.34370742810559013,
"learning_rate": 1.1564109476776983e-05,
"loss": 1.3223,
"step": 3614
},
{
"epoch": 0.6424097027855524,
"grad_norm": 0.3568583405736582,
"learning_rate": 1.155388297286171e-05,
"loss": 1.3258,
"step": 3615
},
{
"epoch": 0.6425874094806522,
"grad_norm": 0.3591957849927233,
"learning_rate": 1.1543659155815092e-05,
"loss": 1.3455,
"step": 3616
},
{
"epoch": 0.6427651161757519,
"grad_norm": 0.3544383209421418,
"learning_rate": 1.1533438028889537e-05,
"loss": 1.3422,
"step": 3617
},
{
"epoch": 0.6429428228708517,
"grad_norm": 0.3514316832526391,
"learning_rate": 1.1523219595336562e-05,
"loss": 1.3298,
"step": 3618
},
{
"epoch": 0.6431205295659514,
"grad_norm": 0.38974183894498776,
"learning_rate": 1.1513003858406848e-05,
"loss": 1.2817,
"step": 3619
},
{
"epoch": 0.6432982362610511,
"grad_norm": 0.3422279426283475,
"learning_rate": 1.1502790821350217e-05,
"loss": 1.2749,
"step": 3620
},
{
"epoch": 0.6434759429561508,
"grad_norm": 0.37370222240068574,
"learning_rate": 1.1492580487415612e-05,
"loss": 1.3079,
"step": 3621
},
{
"epoch": 0.6436536496512506,
"grad_norm": 0.3451495761033728,
"learning_rate": 1.1482372859851148e-05,
"loss": 1.281,
"step": 3622
},
{
"epoch": 0.6438313563463504,
"grad_norm": 0.34729055134458614,
"learning_rate": 1.1472167941904057e-05,
"loss": 1.3025,
"step": 3623
},
{
"epoch": 0.6440090630414501,
"grad_norm": 0.3461291927553688,
"learning_rate": 1.1461965736820719e-05,
"loss": 1.3212,
"step": 3624
},
{
"epoch": 0.6441867697365499,
"grad_norm": 0.3452232727609853,
"learning_rate": 1.1451766247846638e-05,
"loss": 1.3369,
"step": 3625
},
{
"epoch": 0.6443644764316495,
"grad_norm": 0.34826671737925735,
"learning_rate": 1.1441569478226478e-05,
"loss": 1.3019,
"step": 3626
},
{
"epoch": 0.6445421831267493,
"grad_norm": 0.34297546744899143,
"learning_rate": 1.1431375431204021e-05,
"loss": 1.2989,
"step": 3627
},
{
"epoch": 0.644719889821849,
"grad_norm": 0.34798360002674356,
"learning_rate": 1.1421184110022175e-05,
"loss": 1.3184,
"step": 3628
},
{
"epoch": 0.6448975965169488,
"grad_norm": 0.3583250054761565,
"learning_rate": 1.1410995517922991e-05,
"loss": 1.3139,
"step": 3629
},
{
"epoch": 0.6450753032120485,
"grad_norm": 0.3563166474028965,
"learning_rate": 1.1400809658147653e-05,
"loss": 1.3136,
"step": 3630
},
{
"epoch": 0.6452530099071483,
"grad_norm": 0.33680110116586953,
"learning_rate": 1.1390626533936482e-05,
"loss": 1.2922,
"step": 3631
},
{
"epoch": 0.645430716602248,
"grad_norm": 0.34764640599981006,
"learning_rate": 1.1380446148528921e-05,
"loss": 1.3304,
"step": 3632
},
{
"epoch": 0.6456084232973477,
"grad_norm": 0.34153223840634284,
"learning_rate": 1.1370268505163536e-05,
"loss": 1.2813,
"step": 3633
},
{
"epoch": 0.6457861299924474,
"grad_norm": 0.35084946690801455,
"learning_rate": 1.136009360707803e-05,
"loss": 1.3149,
"step": 3634
},
{
"epoch": 0.6459638366875472,
"grad_norm": 0.3437263386686483,
"learning_rate": 1.1349921457509226e-05,
"loss": 1.2783,
"step": 3635
},
{
"epoch": 0.646141543382647,
"grad_norm": 0.33985062583568526,
"learning_rate": 1.1339752059693078e-05,
"loss": 1.2968,
"step": 3636
},
{
"epoch": 0.6463192500777467,
"grad_norm": 0.35462496454508874,
"learning_rate": 1.1329585416864666e-05,
"loss": 1.3676,
"step": 3637
},
{
"epoch": 0.6464969567728465,
"grad_norm": 0.3450593816439494,
"learning_rate": 1.1319421532258185e-05,
"loss": 1.3305,
"step": 3638
},
{
"epoch": 0.6466746634679461,
"grad_norm": 0.3416057081749839,
"learning_rate": 1.1309260409106955e-05,
"loss": 1.3,
"step": 3639
},
{
"epoch": 0.6468523701630459,
"grad_norm": 0.3588587950187614,
"learning_rate": 1.1299102050643431e-05,
"loss": 1.3229,
"step": 3640
},
{
"epoch": 0.6470300768581456,
"grad_norm": 0.4204773179685022,
"learning_rate": 1.1288946460099173e-05,
"loss": 1.3182,
"step": 3641
},
{
"epoch": 0.6472077835532454,
"grad_norm": 0.34243059101949913,
"learning_rate": 1.1278793640704873e-05,
"loss": 1.2925,
"step": 3642
},
{
"epoch": 0.6473854902483451,
"grad_norm": 0.34519423240233155,
"learning_rate": 1.1268643595690318e-05,
"loss": 1.2934,
"step": 3643
},
{
"epoch": 0.6475631969434449,
"grad_norm": 0.35333237328784,
"learning_rate": 1.1258496328284427e-05,
"loss": 1.3349,
"step": 3644
},
{
"epoch": 0.6477409036385445,
"grad_norm": 0.34555008399123893,
"learning_rate": 1.1248351841715252e-05,
"loss": 1.3285,
"step": 3645
},
{
"epoch": 0.6479186103336443,
"grad_norm": 0.3586080671378361,
"learning_rate": 1.1238210139209942e-05,
"loss": 1.3151,
"step": 3646
},
{
"epoch": 0.648096317028744,
"grad_norm": 0.3508471840852196,
"learning_rate": 1.1228071223994757e-05,
"loss": 1.3267,
"step": 3647
},
{
"epoch": 0.6482740237238438,
"grad_norm": 0.3510992429683992,
"learning_rate": 1.121793509929508e-05,
"loss": 1.3294,
"step": 3648
},
{
"epoch": 0.6484517304189436,
"grad_norm": 0.3542094273190764,
"learning_rate": 1.12078017683354e-05,
"loss": 1.332,
"step": 3649
},
{
"epoch": 0.6486294371140433,
"grad_norm": 0.3389532684343894,
"learning_rate": 1.1197671234339324e-05,
"loss": 1.3182,
"step": 3650
},
{
"epoch": 0.6488071438091431,
"grad_norm": 0.3528520413474479,
"learning_rate": 1.1187543500529565e-05,
"loss": 1.3573,
"step": 3651
},
{
"epoch": 0.6489848505042427,
"grad_norm": 0.35284675047864744,
"learning_rate": 1.1177418570127943e-05,
"loss": 1.3123,
"step": 3652
},
{
"epoch": 0.6491625571993425,
"grad_norm": 0.35100765292428593,
"learning_rate": 1.116729644635538e-05,
"loss": 1.3347,
"step": 3653
},
{
"epoch": 0.6493402638944422,
"grad_norm": 0.40385935103769904,
"learning_rate": 1.1157177132431934e-05,
"loss": 1.2831,
"step": 3654
},
{
"epoch": 0.649517970589542,
"grad_norm": 0.3511078875767884,
"learning_rate": 1.1147060631576738e-05,
"loss": 1.3123,
"step": 3655
},
{
"epoch": 0.6496956772846417,
"grad_norm": 0.3550586122022803,
"learning_rate": 1.1136946947008042e-05,
"loss": 1.3267,
"step": 3656
},
{
"epoch": 0.6498733839797415,
"grad_norm": 0.4601289072957297,
"learning_rate": 1.1126836081943199e-05,
"loss": 1.3618,
"step": 3657
},
{
"epoch": 0.6500510906748411,
"grad_norm": 0.3380549678726808,
"learning_rate": 1.1116728039598668e-05,
"loss": 1.2991,
"step": 3658
},
{
"epoch": 0.6502287973699409,
"grad_norm": 0.35533370997942487,
"learning_rate": 1.1106622823190003e-05,
"loss": 1.3344,
"step": 3659
},
{
"epoch": 0.6504065040650406,
"grad_norm": 0.351216726506674,
"learning_rate": 1.1096520435931865e-05,
"loss": 1.3384,
"step": 3660
},
{
"epoch": 0.6505842107601404,
"grad_norm": 0.35029403833027334,
"learning_rate": 1.1086420881038016e-05,
"loss": 1.3345,
"step": 3661
},
{
"epoch": 0.6507619174552401,
"grad_norm": 0.3568126350628331,
"learning_rate": 1.107632416172131e-05,
"loss": 1.2866,
"step": 3662
},
{
"epoch": 0.6509396241503399,
"grad_norm": 0.34958623854560367,
"learning_rate": 1.10662302811937e-05,
"loss": 1.3337,
"step": 3663
},
{
"epoch": 0.6511173308454397,
"grad_norm": 0.3483133182419292,
"learning_rate": 1.105613924266626e-05,
"loss": 1.3094,
"step": 3664
},
{
"epoch": 0.6512950375405393,
"grad_norm": 0.3634436091274065,
"learning_rate": 1.1046051049349114e-05,
"loss": 1.3543,
"step": 3665
},
{
"epoch": 0.6514727442356391,
"grad_norm": 0.3974618856367738,
"learning_rate": 1.1035965704451515e-05,
"loss": 1.3101,
"step": 3666
},
{
"epoch": 0.6516504509307388,
"grad_norm": 0.3500587822843521,
"learning_rate": 1.1025883211181796e-05,
"loss": 1.3168,
"step": 3667
},
{
"epoch": 0.6518281576258386,
"grad_norm": 0.36009561151895203,
"learning_rate": 1.10158035727474e-05,
"loss": 1.3466,
"step": 3668
},
{
"epoch": 0.6520058643209383,
"grad_norm": 0.3513127427137808,
"learning_rate": 1.1005726792354843e-05,
"loss": 1.344,
"step": 3669
},
{
"epoch": 0.6521835710160381,
"grad_norm": 0.34741838690748694,
"learning_rate": 1.0995652873209739e-05,
"loss": 1.2868,
"step": 3670
},
{
"epoch": 0.6523612777111377,
"grad_norm": 0.3445656726790001,
"learning_rate": 1.0985581818516789e-05,
"loss": 1.3175,
"step": 3671
},
{
"epoch": 0.6525389844062375,
"grad_norm": 0.3464368571062268,
"learning_rate": 1.0975513631479788e-05,
"loss": 1.3203,
"step": 3672
},
{
"epoch": 0.6527166911013372,
"grad_norm": 0.3663077271992055,
"learning_rate": 1.0965448315301614e-05,
"loss": 1.3322,
"step": 3673
},
{
"epoch": 0.652894397796437,
"grad_norm": 0.36870857953823827,
"learning_rate": 1.0955385873184231e-05,
"loss": 1.3384,
"step": 3674
},
{
"epoch": 0.6530721044915367,
"grad_norm": 0.3544591607081791,
"learning_rate": 1.0945326308328696e-05,
"loss": 1.3326,
"step": 3675
},
{
"epoch": 0.6532498111866365,
"grad_norm": 0.3582443098591744,
"learning_rate": 1.093526962393514e-05,
"loss": 1.3537,
"step": 3676
},
{
"epoch": 0.6534275178817361,
"grad_norm": 0.35547344852330076,
"learning_rate": 1.0925215823202781e-05,
"loss": 1.3269,
"step": 3677
},
{
"epoch": 0.6536052245768359,
"grad_norm": 0.3503800971138621,
"learning_rate": 1.0915164909329937e-05,
"loss": 1.3166,
"step": 3678
},
{
"epoch": 0.6537829312719357,
"grad_norm": 0.34468421015362666,
"learning_rate": 1.090511688551398e-05,
"loss": 1.3054,
"step": 3679
},
{
"epoch": 0.6539606379670354,
"grad_norm": 0.34591137839759856,
"learning_rate": 1.0895071754951388e-05,
"loss": 1.3239,
"step": 3680
},
{
"epoch": 0.6541383446621352,
"grad_norm": 0.3614123029712968,
"learning_rate": 1.088502952083768e-05,
"loss": 1.3462,
"step": 3681
},
{
"epoch": 0.6543160513572349,
"grad_norm": 0.35073936314270276,
"learning_rate": 1.0874990186367507e-05,
"loss": 1.3382,
"step": 3682
},
{
"epoch": 0.6544937580523347,
"grad_norm": 0.33873860967612485,
"learning_rate": 1.0864953754734557e-05,
"loss": 1.3139,
"step": 3683
},
{
"epoch": 0.6546714647474343,
"grad_norm": 0.3390867183464163,
"learning_rate": 1.0854920229131609e-05,
"loss": 1.2753,
"step": 3684
},
{
"epoch": 0.6548491714425341,
"grad_norm": 0.356218239452208,
"learning_rate": 1.0844889612750517e-05,
"loss": 1.3017,
"step": 3685
},
{
"epoch": 0.6550268781376338,
"grad_norm": 0.34356572156436105,
"learning_rate": 1.083486190878221e-05,
"loss": 1.2819,
"step": 3686
},
{
"epoch": 0.6552045848327336,
"grad_norm": 0.34591191830786133,
"learning_rate": 1.0824837120416687e-05,
"loss": 1.2969,
"step": 3687
},
{
"epoch": 0.6553822915278333,
"grad_norm": 0.3667712797516585,
"learning_rate": 1.0814815250843025e-05,
"loss": 1.3188,
"step": 3688
},
{
"epoch": 0.6555599982229331,
"grad_norm": 0.3473359033613071,
"learning_rate": 1.0804796303249365e-05,
"loss": 1.3458,
"step": 3689
},
{
"epoch": 0.6557377049180327,
"grad_norm": 0.34239430019584516,
"learning_rate": 1.0794780280822926e-05,
"loss": 1.3281,
"step": 3690
},
{
"epoch": 0.6559154116131325,
"grad_norm": 0.34549645109197147,
"learning_rate": 1.078476718674998e-05,
"loss": 1.292,
"step": 3691
},
{
"epoch": 0.6560931183082322,
"grad_norm": 0.3521789609320512,
"learning_rate": 1.0774757024215904e-05,
"loss": 1.2964,
"step": 3692
},
{
"epoch": 0.656270825003332,
"grad_norm": 0.3556081642393652,
"learning_rate": 1.0764749796405106e-05,
"loss": 1.353,
"step": 3693
},
{
"epoch": 0.6564485316984318,
"grad_norm": 0.35815888133742413,
"learning_rate": 1.0754745506501074e-05,
"loss": 1.3497,
"step": 3694
},
{
"epoch": 0.6566262383935315,
"grad_norm": 0.3477932143773736,
"learning_rate": 1.074474415768636e-05,
"loss": 1.3013,
"step": 3695
},
{
"epoch": 0.6568039450886313,
"grad_norm": 0.3698423677668519,
"learning_rate": 1.0734745753142586e-05,
"loss": 1.3024,
"step": 3696
},
{
"epoch": 0.6569816517837309,
"grad_norm": 0.35058727241275217,
"learning_rate": 1.0724750296050425e-05,
"loss": 1.3152,
"step": 3697
},
{
"epoch": 0.6571593584788307,
"grad_norm": 0.3537993203463864,
"learning_rate": 1.0714757789589628e-05,
"loss": 1.3301,
"step": 3698
},
{
"epoch": 0.6573370651739304,
"grad_norm": 0.3485996636857801,
"learning_rate": 1.070476823693899e-05,
"loss": 1.3467,
"step": 3699
},
{
"epoch": 0.6575147718690302,
"grad_norm": 0.3609665237480074,
"learning_rate": 1.0694781641276375e-05,
"loss": 1.3351,
"step": 3700
},
{
"epoch": 0.6576924785641299,
"grad_norm": 0.34865060020943806,
"learning_rate": 1.068479800577872e-05,
"loss": 1.3233,
"step": 3701
},
{
"epoch": 0.6578701852592297,
"grad_norm": 0.3428667524605816,
"learning_rate": 1.0674817333622007e-05,
"loss": 1.3135,
"step": 3702
},
{
"epoch": 0.6580478919543293,
"grad_norm": 0.3581101323332052,
"learning_rate": 1.066483962798126e-05,
"loss": 1.2984,
"step": 3703
},
{
"epoch": 0.6582255986494291,
"grad_norm": 0.35579423922310016,
"learning_rate": 1.0654864892030585e-05,
"loss": 1.3084,
"step": 3704
},
{
"epoch": 0.6584033053445288,
"grad_norm": 0.34008483451330707,
"learning_rate": 1.0644893128943122e-05,
"loss": 1.3186,
"step": 3705
},
{
"epoch": 0.6585810120396286,
"grad_norm": 0.3486531832758974,
"learning_rate": 1.0634924341891093e-05,
"loss": 1.334,
"step": 3706
},
{
"epoch": 0.6587587187347284,
"grad_norm": 0.35702910853334957,
"learning_rate": 1.0624958534045748e-05,
"loss": 1.3032,
"step": 3707
},
{
"epoch": 0.6589364254298281,
"grad_norm": 0.35397638354198,
"learning_rate": 1.06149957085774e-05,
"loss": 1.3399,
"step": 3708
},
{
"epoch": 0.6591141321249278,
"grad_norm": 0.3632629208119128,
"learning_rate": 1.0605035868655411e-05,
"loss": 1.3008,
"step": 3709
},
{
"epoch": 0.6592918388200275,
"grad_norm": 0.3570549594307501,
"learning_rate": 1.0595079017448191e-05,
"loss": 1.3534,
"step": 3710
},
{
"epoch": 0.6594695455151273,
"grad_norm": 0.3481870247694426,
"learning_rate": 1.0585125158123204e-05,
"loss": 1.3194,
"step": 3711
},
{
"epoch": 0.659647252210227,
"grad_norm": 0.3446325208652196,
"learning_rate": 1.0575174293846957e-05,
"loss": 1.3069,
"step": 3712
},
{
"epoch": 0.6598249589053268,
"grad_norm": 0.3434266805852896,
"learning_rate": 1.0565226427785011e-05,
"loss": 1.3063,
"step": 3713
},
{
"epoch": 0.6600026656004265,
"grad_norm": 0.3493183979966537,
"learning_rate": 1.0555281563101957e-05,
"loss": 1.315,
"step": 3714
},
{
"epoch": 0.6601803722955263,
"grad_norm": 0.35377016786009574,
"learning_rate": 1.0545339702961463e-05,
"loss": 1.3376,
"step": 3715
},
{
"epoch": 0.6603580789906259,
"grad_norm": 0.3510733126283065,
"learning_rate": 1.0535400850526214e-05,
"loss": 1.3009,
"step": 3716
},
{
"epoch": 0.6605357856857257,
"grad_norm": 0.34692448274099447,
"learning_rate": 1.0525465008957951e-05,
"loss": 1.3414,
"step": 3717
},
{
"epoch": 0.6607134923808254,
"grad_norm": 0.34943164747198535,
"learning_rate": 1.0515532181417436e-05,
"loss": 1.3215,
"step": 3718
},
{
"epoch": 0.6608911990759252,
"grad_norm": 0.36068294818835905,
"learning_rate": 1.0505602371064492e-05,
"loss": 1.3167,
"step": 3719
},
{
"epoch": 0.661068905771025,
"grad_norm": 0.3495084316909251,
"learning_rate": 1.0495675581057992e-05,
"loss": 1.3295,
"step": 3720
},
{
"epoch": 0.6612466124661247,
"grad_norm": 0.34352898374800256,
"learning_rate": 1.0485751814555822e-05,
"loss": 1.3275,
"step": 3721
},
{
"epoch": 0.6614243191612243,
"grad_norm": 0.36387810229574835,
"learning_rate": 1.0475831074714927e-05,
"loss": 1.3041,
"step": 3722
},
{
"epoch": 0.6616020258563241,
"grad_norm": 0.3527236688170157,
"learning_rate": 1.0465913364691268e-05,
"loss": 1.3354,
"step": 3723
},
{
"epoch": 0.6617797325514239,
"grad_norm": 0.34576831683320336,
"learning_rate": 1.045599868763988e-05,
"loss": 1.3203,
"step": 3724
},
{
"epoch": 0.6619574392465236,
"grad_norm": 0.34922117786100354,
"learning_rate": 1.0446087046714788e-05,
"loss": 1.338,
"step": 3725
},
{
"epoch": 0.6621351459416234,
"grad_norm": 0.3519063927728252,
"learning_rate": 1.0436178445069071e-05,
"loss": 1.3325,
"step": 3726
},
{
"epoch": 0.6623128526367231,
"grad_norm": 0.35565998364823265,
"learning_rate": 1.042627288585485e-05,
"loss": 1.3454,
"step": 3727
},
{
"epoch": 0.6624905593318229,
"grad_norm": 0.348560134851858,
"learning_rate": 1.0416370372223254e-05,
"loss": 1.3361,
"step": 3728
},
{
"epoch": 0.6626682660269225,
"grad_norm": 0.36575705151723664,
"learning_rate": 1.0406470907324482e-05,
"loss": 1.3255,
"step": 3729
},
{
"epoch": 0.6628459727220223,
"grad_norm": 0.34383301139071126,
"learning_rate": 1.0396574494307727e-05,
"loss": 1.3077,
"step": 3730
},
{
"epoch": 0.663023679417122,
"grad_norm": 0.34208135835628173,
"learning_rate": 1.0386681136321228e-05,
"loss": 1.3184,
"step": 3731
},
{
"epoch": 0.6632013861122218,
"grad_norm": 0.40370700872959003,
"learning_rate": 1.0376790836512245e-05,
"loss": 1.3,
"step": 3732
},
{
"epoch": 0.6633790928073215,
"grad_norm": 0.3434374649475327,
"learning_rate": 1.0366903598027069e-05,
"loss": 1.3009,
"step": 3733
},
{
"epoch": 0.6635567995024213,
"grad_norm": 0.37173012973967334,
"learning_rate": 1.0357019424011018e-05,
"loss": 1.288,
"step": 3734
},
{
"epoch": 0.663734506197521,
"grad_norm": 0.33663340096220656,
"learning_rate": 1.0347138317608434e-05,
"loss": 1.2853,
"step": 3735
},
{
"epoch": 0.6639122128926207,
"grad_norm": 0.3510964898624168,
"learning_rate": 1.0337260281962678e-05,
"loss": 1.348,
"step": 3736
},
{
"epoch": 0.6640899195877205,
"grad_norm": 0.3535250806302582,
"learning_rate": 1.0327385320216136e-05,
"loss": 1.3508,
"step": 3737
},
{
"epoch": 0.6642676262828202,
"grad_norm": 0.3506383033136998,
"learning_rate": 1.0317513435510233e-05,
"loss": 1.3456,
"step": 3738
},
{
"epoch": 0.66444533297792,
"grad_norm": 0.34485357454112564,
"learning_rate": 1.0307644630985401e-05,
"loss": 1.3178,
"step": 3739
},
{
"epoch": 0.6646230396730197,
"grad_norm": 0.3519673944887055,
"learning_rate": 1.0297778909781078e-05,
"loss": 1.3051,
"step": 3740
},
{
"epoch": 0.6648007463681194,
"grad_norm": 0.3382798743591559,
"learning_rate": 1.028791627503574e-05,
"loss": 1.261,
"step": 3741
},
{
"epoch": 0.6649784530632191,
"grad_norm": 0.3546852828415783,
"learning_rate": 1.0278056729886873e-05,
"loss": 1.3194,
"step": 3742
},
{
"epoch": 0.6651561597583189,
"grad_norm": 0.3550205225395426,
"learning_rate": 1.0268200277470998e-05,
"loss": 1.3563,
"step": 3743
},
{
"epoch": 0.6653338664534186,
"grad_norm": 0.3423325756112575,
"learning_rate": 1.0258346920923628e-05,
"loss": 1.3314,
"step": 3744
},
{
"epoch": 0.6655115731485184,
"grad_norm": 0.345406103789718,
"learning_rate": 1.0248496663379304e-05,
"loss": 1.311,
"step": 3745
},
{
"epoch": 0.6656892798436181,
"grad_norm": 0.3542280193127132,
"learning_rate": 1.0238649507971577e-05,
"loss": 1.3121,
"step": 3746
},
{
"epoch": 0.6658669865387179,
"grad_norm": 0.3485568610656086,
"learning_rate": 1.0228805457833009e-05,
"loss": 1.3242,
"step": 3747
},
{
"epoch": 0.6660446932338175,
"grad_norm": 0.35988457310294947,
"learning_rate": 1.0218964516095182e-05,
"loss": 1.3457,
"step": 3748
},
{
"epoch": 0.6662223999289173,
"grad_norm": 0.3426065982364232,
"learning_rate": 1.0209126685888684e-05,
"loss": 1.3068,
"step": 3749
},
{
"epoch": 0.666400106624017,
"grad_norm": 0.3509704802958005,
"learning_rate": 1.019929197034311e-05,
"loss": 1.3232,
"step": 3750
},
{
"epoch": 0.6665778133191168,
"grad_norm": 0.35380883786579376,
"learning_rate": 1.0189460372587066e-05,
"loss": 1.3465,
"step": 3751
},
{
"epoch": 0.6667555200142166,
"grad_norm": 0.3638205732214967,
"learning_rate": 1.0179631895748182e-05,
"loss": 1.2868,
"step": 3752
},
{
"epoch": 0.6669332267093163,
"grad_norm": 0.33920235529542114,
"learning_rate": 1.0169806542953066e-05,
"loss": 1.2887,
"step": 3753
},
{
"epoch": 0.667110933404416,
"grad_norm": 0.35342580407844676,
"learning_rate": 1.015998431732736e-05,
"loss": 1.3416,
"step": 3754
},
{
"epoch": 0.6672886400995157,
"grad_norm": 0.3449369121225887,
"learning_rate": 1.0150165221995698e-05,
"loss": 1.3123,
"step": 3755
},
{
"epoch": 0.6674663467946155,
"grad_norm": 0.35557088887945765,
"learning_rate": 1.01403492600817e-05,
"loss": 1.3956,
"step": 3756
},
{
"epoch": 0.6676440534897152,
"grad_norm": 0.34408306382962855,
"learning_rate": 1.0130536434708024e-05,
"loss": 1.3027,
"step": 3757
},
{
"epoch": 0.667821760184815,
"grad_norm": 0.3632565654352166,
"learning_rate": 1.0120726748996316e-05,
"loss": 1.2915,
"step": 3758
},
{
"epoch": 0.6679994668799147,
"grad_norm": 0.3501142147172689,
"learning_rate": 1.0110920206067214e-05,
"loss": 1.3153,
"step": 3759
},
{
"epoch": 0.6681771735750145,
"grad_norm": 0.3933698275205193,
"learning_rate": 1.010111680904037e-05,
"loss": 1.3068,
"step": 3760
},
{
"epoch": 0.6683548802701141,
"grad_norm": 0.3419709278988747,
"learning_rate": 1.0091316561034419e-05,
"loss": 1.3211,
"step": 3761
},
{
"epoch": 0.6685325869652139,
"grad_norm": 0.34478644818617654,
"learning_rate": 1.0081519465167022e-05,
"loss": 1.3187,
"step": 3762
},
{
"epoch": 0.6687102936603136,
"grad_norm": 0.3495158630032051,
"learning_rate": 1.0071725524554803e-05,
"loss": 1.3256,
"step": 3763
},
{
"epoch": 0.6688880003554134,
"grad_norm": 0.3506032762168772,
"learning_rate": 1.0061934742313406e-05,
"loss": 1.3257,
"step": 3764
},
{
"epoch": 0.6690657070505132,
"grad_norm": 0.349651829457587,
"learning_rate": 1.0052147121557451e-05,
"loss": 1.3579,
"step": 3765
},
{
"epoch": 0.6692434137456129,
"grad_norm": 0.35788659447741256,
"learning_rate": 1.0042362665400584e-05,
"loss": 1.3481,
"step": 3766
},
{
"epoch": 0.6694211204407126,
"grad_norm": 0.33918981298407347,
"learning_rate": 1.0032581376955416e-05,
"loss": 1.2971,
"step": 3767
},
{
"epoch": 0.6695988271358123,
"grad_norm": 0.34813670927718626,
"learning_rate": 1.0022803259333553e-05,
"loss": 1.319,
"step": 3768
},
{
"epoch": 0.6697765338309121,
"grad_norm": 0.34449118351977087,
"learning_rate": 1.0013028315645607e-05,
"loss": 1.3123,
"step": 3769
},
{
"epoch": 0.6699542405260118,
"grad_norm": 0.3438958265061071,
"learning_rate": 1.0003256549001165e-05,
"loss": 1.3043,
"step": 3770
},
{
"epoch": 0.6701319472211116,
"grad_norm": 0.3375014339966034,
"learning_rate": 9.993487962508815e-06,
"loss": 1.2339,
"step": 3771
},
{
"epoch": 0.6703096539162113,
"grad_norm": 0.35182798482816785,
"learning_rate": 9.983722559276122e-06,
"loss": 1.3449,
"step": 3772
},
{
"epoch": 0.670487360611311,
"grad_norm": 0.3511503896850602,
"learning_rate": 9.973960342409647e-06,
"loss": 1.3271,
"step": 3773
},
{
"epoch": 0.6706650673064107,
"grad_norm": 0.3638382031611361,
"learning_rate": 9.96420131501494e-06,
"loss": 1.393,
"step": 3774
},
{
"epoch": 0.6708427740015105,
"grad_norm": 0.35860176029015717,
"learning_rate": 9.954445480196512e-06,
"loss": 1.313,
"step": 3775
},
{
"epoch": 0.6710204806966102,
"grad_norm": 0.3744752254676168,
"learning_rate": 9.944692841057904e-06,
"loss": 1.3139,
"step": 3776
},
{
"epoch": 0.67119818739171,
"grad_norm": 0.3508376833014186,
"learning_rate": 9.934943400701609e-06,
"loss": 1.3405,
"step": 3777
},
{
"epoch": 0.6713758940868098,
"grad_norm": 0.3463289049433646,
"learning_rate": 9.925197162229093e-06,
"loss": 1.3111,
"step": 3778
},
{
"epoch": 0.6715536007819095,
"grad_norm": 0.3507675524005522,
"learning_rate": 9.915454128740813e-06,
"loss": 1.3204,
"step": 3779
},
{
"epoch": 0.6717313074770092,
"grad_norm": 0.3767013166388892,
"learning_rate": 9.905714303336236e-06,
"loss": 1.3166,
"step": 3780
},
{
"epoch": 0.6719090141721089,
"grad_norm": 0.35275093881827113,
"learning_rate": 9.895977689113766e-06,
"loss": 1.3206,
"step": 3781
},
{
"epoch": 0.6720867208672087,
"grad_norm": 0.3482172444120543,
"learning_rate": 9.886244289170811e-06,
"loss": 1.299,
"step": 3782
},
{
"epoch": 0.6722644275623084,
"grad_norm": 0.3464450319084066,
"learning_rate": 9.876514106603744e-06,
"loss": 1.3029,
"step": 3783
},
{
"epoch": 0.6724421342574082,
"grad_norm": 0.3459529817413587,
"learning_rate": 9.866787144507922e-06,
"loss": 1.3013,
"step": 3784
},
{
"epoch": 0.6726198409525079,
"grad_norm": 0.3549209544420543,
"learning_rate": 9.857063405977672e-06,
"loss": 1.3146,
"step": 3785
},
{
"epoch": 0.6727975476476076,
"grad_norm": 0.3562575867785443,
"learning_rate": 9.847342894106298e-06,
"loss": 1.2985,
"step": 3786
},
{
"epoch": 0.6729752543427073,
"grad_norm": 0.35158670251474794,
"learning_rate": 9.837625611986079e-06,
"loss": 1.3219,
"step": 3787
},
{
"epoch": 0.6731529610378071,
"grad_norm": 0.3487477320111606,
"learning_rate": 9.827911562708266e-06,
"loss": 1.2901,
"step": 3788
},
{
"epoch": 0.6733306677329068,
"grad_norm": 0.33706100674861095,
"learning_rate": 9.818200749363071e-06,
"loss": 1.2711,
"step": 3789
},
{
"epoch": 0.6735083744280066,
"grad_norm": 0.3541971645387061,
"learning_rate": 9.808493175039704e-06,
"loss": 1.3064,
"step": 3790
},
{
"epoch": 0.6736860811231064,
"grad_norm": 0.37072759124769167,
"learning_rate": 9.798788842826316e-06,
"loss": 1.34,
"step": 3791
},
{
"epoch": 0.6738637878182061,
"grad_norm": 0.34771497982926697,
"learning_rate": 9.78908775581004e-06,
"loss": 1.3344,
"step": 3792
},
{
"epoch": 0.6740414945133058,
"grad_norm": 0.35928663799730987,
"learning_rate": 9.779389917076976e-06,
"loss": 1.3567,
"step": 3793
},
{
"epoch": 0.6742192012084055,
"grad_norm": 0.3450947667235526,
"learning_rate": 9.769695329712183e-06,
"loss": 1.2974,
"step": 3794
},
{
"epoch": 0.6743969079035053,
"grad_norm": 0.35506179991366366,
"learning_rate": 9.760003996799698e-06,
"loss": 1.3228,
"step": 3795
},
{
"epoch": 0.674574614598605,
"grad_norm": 0.3500866972256888,
"learning_rate": 9.750315921422513e-06,
"loss": 1.2817,
"step": 3796
},
{
"epoch": 0.6747523212937048,
"grad_norm": 0.34776051677143127,
"learning_rate": 9.740631106662586e-06,
"loss": 1.3543,
"step": 3797
},
{
"epoch": 0.6749300279888045,
"grad_norm": 0.3529223896447972,
"learning_rate": 9.730949555600832e-06,
"loss": 1.3474,
"step": 3798
},
{
"epoch": 0.6751077346839042,
"grad_norm": 0.342802478837947,
"learning_rate": 9.721271271317159e-06,
"loss": 1.323,
"step": 3799
},
{
"epoch": 0.6752854413790039,
"grad_norm": 0.3401879199538994,
"learning_rate": 9.711596256890388e-06,
"loss": 1.3077,
"step": 3800
},
{
"epoch": 0.6754631480741037,
"grad_norm": 0.3453509964499724,
"learning_rate": 9.701924515398329e-06,
"loss": 1.328,
"step": 3801
},
{
"epoch": 0.6756408547692034,
"grad_norm": 0.3462947112932736,
"learning_rate": 9.692256049917745e-06,
"loss": 1.3207,
"step": 3802
},
{
"epoch": 0.6758185614643032,
"grad_norm": 0.34302742268084707,
"learning_rate": 9.68259086352435e-06,
"loss": 1.3044,
"step": 3803
},
{
"epoch": 0.675996268159403,
"grad_norm": 0.34341197492284836,
"learning_rate": 9.672928959292836e-06,
"loss": 1.2974,
"step": 3804
},
{
"epoch": 0.6761739748545026,
"grad_norm": 0.34533980256249763,
"learning_rate": 9.66327034029683e-06,
"loss": 1.3287,
"step": 3805
},
{
"epoch": 0.6763516815496023,
"grad_norm": 0.33867777180255004,
"learning_rate": 9.653615009608921e-06,
"loss": 1.2855,
"step": 3806
},
{
"epoch": 0.6765293882447021,
"grad_norm": 0.38211701835295925,
"learning_rate": 9.643962970300646e-06,
"loss": 1.3536,
"step": 3807
},
{
"epoch": 0.6767070949398019,
"grad_norm": 0.3481213999351934,
"learning_rate": 9.63431422544251e-06,
"loss": 1.2958,
"step": 3808
},
{
"epoch": 0.6768848016349016,
"grad_norm": 0.3468897160122548,
"learning_rate": 9.624668778103949e-06,
"loss": 1.3008,
"step": 3809
},
{
"epoch": 0.6770625083300014,
"grad_norm": 0.35069294245080834,
"learning_rate": 9.61502663135337e-06,
"loss": 1.2923,
"step": 3810
},
{
"epoch": 0.6772402150251011,
"grad_norm": 0.3540759426806464,
"learning_rate": 9.605387788258116e-06,
"loss": 1.3426,
"step": 3811
},
{
"epoch": 0.6774179217202008,
"grad_norm": 0.3460268658073583,
"learning_rate": 9.595752251884479e-06,
"loss": 1.3208,
"step": 3812
},
{
"epoch": 0.6775956284153005,
"grad_norm": 0.3536999674421588,
"learning_rate": 9.586120025297719e-06,
"loss": 1.3187,
"step": 3813
},
{
"epoch": 0.6777733351104003,
"grad_norm": 0.3415819582103319,
"learning_rate": 9.576491111562021e-06,
"loss": 1.3165,
"step": 3814
},
{
"epoch": 0.6779510418055,
"grad_norm": 0.3499569054713927,
"learning_rate": 9.566865513740528e-06,
"loss": 1.3194,
"step": 3815
},
{
"epoch": 0.6781287485005998,
"grad_norm": 0.34520727978431087,
"learning_rate": 9.557243234895314e-06,
"loss": 1.3076,
"step": 3816
},
{
"epoch": 0.6783064551956995,
"grad_norm": 0.3541586011526474,
"learning_rate": 9.547624278087405e-06,
"loss": 1.3225,
"step": 3817
},
{
"epoch": 0.6784841618907992,
"grad_norm": 0.34599349959414766,
"learning_rate": 9.538008646376786e-06,
"loss": 1.3165,
"step": 3818
},
{
"epoch": 0.6786618685858989,
"grad_norm": 0.3408889969586033,
"learning_rate": 9.528396342822363e-06,
"loss": 1.3019,
"step": 3819
},
{
"epoch": 0.6788395752809987,
"grad_norm": 0.34808137353755114,
"learning_rate": 9.51878737048199e-06,
"loss": 1.3208,
"step": 3820
},
{
"epoch": 0.6790172819760985,
"grad_norm": 0.34836505243671495,
"learning_rate": 9.509181732412462e-06,
"loss": 1.3338,
"step": 3821
},
{
"epoch": 0.6791949886711982,
"grad_norm": 0.3461827343094303,
"learning_rate": 9.499579431669517e-06,
"loss": 1.2834,
"step": 3822
},
{
"epoch": 0.679372695366298,
"grad_norm": 0.3535691304104468,
"learning_rate": 9.48998047130782e-06,
"loss": 1.3324,
"step": 3823
},
{
"epoch": 0.6795504020613977,
"grad_norm": 0.35212202904188833,
"learning_rate": 9.480384854380988e-06,
"loss": 1.3267,
"step": 3824
},
{
"epoch": 0.6797281087564974,
"grad_norm": 0.3521307448609341,
"learning_rate": 9.470792583941562e-06,
"loss": 1.319,
"step": 3825
},
{
"epoch": 0.6799058154515971,
"grad_norm": 0.3537148523706328,
"learning_rate": 9.461203663041018e-06,
"loss": 1.3094,
"step": 3826
},
{
"epoch": 0.6800835221466969,
"grad_norm": 0.35467440171585674,
"learning_rate": 9.451618094729788e-06,
"loss": 1.3428,
"step": 3827
},
{
"epoch": 0.6802612288417966,
"grad_norm": 0.34694156198339976,
"learning_rate": 9.442035882057214e-06,
"loss": 1.2758,
"step": 3828
},
{
"epoch": 0.6804389355368964,
"grad_norm": 0.34817260300649083,
"learning_rate": 9.432457028071577e-06,
"loss": 1.3008,
"step": 3829
},
{
"epoch": 0.6806166422319961,
"grad_norm": 0.34983431942021,
"learning_rate": 9.422881535820099e-06,
"loss": 1.3191,
"step": 3830
},
{
"epoch": 0.6807943489270958,
"grad_norm": 0.34791399600667977,
"learning_rate": 9.413309408348898e-06,
"loss": 1.3247,
"step": 3831
},
{
"epoch": 0.6809720556221955,
"grad_norm": 0.3765051064566119,
"learning_rate": 9.403740648703077e-06,
"loss": 1.3147,
"step": 3832
},
{
"epoch": 0.6811497623172953,
"grad_norm": 0.34263540995834074,
"learning_rate": 9.394175259926626e-06,
"loss": 1.3122,
"step": 3833
},
{
"epoch": 0.681327469012395,
"grad_norm": 0.347564617752683,
"learning_rate": 9.384613245062475e-06,
"loss": 1.2957,
"step": 3834
},
{
"epoch": 0.6815051757074948,
"grad_norm": 0.35047269701360595,
"learning_rate": 9.375054607152477e-06,
"loss": 1.3547,
"step": 3835
},
{
"epoch": 0.6816828824025946,
"grad_norm": 0.35166755056633603,
"learning_rate": 9.365499349237426e-06,
"loss": 1.3384,
"step": 3836
},
{
"epoch": 0.6818605890976942,
"grad_norm": 0.3542020560803107,
"learning_rate": 9.35594747435703e-06,
"loss": 1.3538,
"step": 3837
},
{
"epoch": 0.682038295792794,
"grad_norm": 0.3624485981994534,
"learning_rate": 9.346398985549906e-06,
"loss": 1.3461,
"step": 3838
},
{
"epoch": 0.6822160024878937,
"grad_norm": 0.35254253350476805,
"learning_rate": 9.336853885853613e-06,
"loss": 1.3291,
"step": 3839
},
{
"epoch": 0.6823937091829935,
"grad_norm": 0.3405852410742013,
"learning_rate": 9.327312178304622e-06,
"loss": 1.3174,
"step": 3840
},
{
"epoch": 0.6825714158780932,
"grad_norm": 0.3479091212341918,
"learning_rate": 9.317773865938342e-06,
"loss": 1.3182,
"step": 3841
},
{
"epoch": 0.682749122573193,
"grad_norm": 0.3496765962577703,
"learning_rate": 9.308238951789085e-06,
"loss": 1.3015,
"step": 3842
},
{
"epoch": 0.6829268292682927,
"grad_norm": 0.34455688269038826,
"learning_rate": 9.298707438890086e-06,
"loss": 1.2895,
"step": 3843
},
{
"epoch": 0.6831045359633924,
"grad_norm": 0.3433143760268825,
"learning_rate": 9.289179330273496e-06,
"loss": 1.3031,
"step": 3844
},
{
"epoch": 0.6832822426584921,
"grad_norm": 0.3420259873247715,
"learning_rate": 9.279654628970388e-06,
"loss": 1.309,
"step": 3845
},
{
"epoch": 0.6834599493535919,
"grad_norm": 0.3494045886258132,
"learning_rate": 9.270133338010747e-06,
"loss": 1.3271,
"step": 3846
},
{
"epoch": 0.6836376560486916,
"grad_norm": 0.35279539491095696,
"learning_rate": 9.260615460423475e-06,
"loss": 1.3563,
"step": 3847
},
{
"epoch": 0.6838153627437914,
"grad_norm": 0.34808211755404067,
"learning_rate": 9.25110099923639e-06,
"loss": 1.3345,
"step": 3848
},
{
"epoch": 0.6839930694388912,
"grad_norm": 0.34346597210547863,
"learning_rate": 9.24158995747621e-06,
"loss": 1.3144,
"step": 3849
},
{
"epoch": 0.6841707761339908,
"grad_norm": 0.3442393985684423,
"learning_rate": 9.232082338168594e-06,
"loss": 1.3305,
"step": 3850
},
{
"epoch": 0.6843484828290906,
"grad_norm": 0.337024842366552,
"learning_rate": 9.222578144338086e-06,
"loss": 1.309,
"step": 3851
},
{
"epoch": 0.6845261895241903,
"grad_norm": 0.3394711888358347,
"learning_rate": 9.213077379008155e-06,
"loss": 1.3093,
"step": 3852
},
{
"epoch": 0.6847038962192901,
"grad_norm": 0.40892558322884853,
"learning_rate": 9.203580045201159e-06,
"loss": 1.3028,
"step": 3853
},
{
"epoch": 0.6848816029143898,
"grad_norm": 0.33905524032718554,
"learning_rate": 9.194086145938382e-06,
"loss": 1.2712,
"step": 3854
},
{
"epoch": 0.6850593096094896,
"grad_norm": 0.33497931167987577,
"learning_rate": 9.184595684240021e-06,
"loss": 1.2981,
"step": 3855
},
{
"epoch": 0.6852370163045893,
"grad_norm": 0.35227319116127603,
"learning_rate": 9.175108663125167e-06,
"loss": 1.3129,
"step": 3856
},
{
"epoch": 0.685414722999689,
"grad_norm": 0.3497427138089859,
"learning_rate": 9.165625085611818e-06,
"loss": 1.3236,
"step": 3857
},
{
"epoch": 0.6855924296947887,
"grad_norm": 0.3486067003403602,
"learning_rate": 9.156144954716878e-06,
"loss": 1.3306,
"step": 3858
},
{
"epoch": 0.6857701363898885,
"grad_norm": 0.3411661407635252,
"learning_rate": 9.146668273456158e-06,
"loss": 1.2847,
"step": 3859
},
{
"epoch": 0.6859478430849882,
"grad_norm": 0.3546296536913926,
"learning_rate": 9.137195044844365e-06,
"loss": 1.3429,
"step": 3860
},
{
"epoch": 0.686125549780088,
"grad_norm": 0.35156977152162283,
"learning_rate": 9.127725271895114e-06,
"loss": 1.3055,
"step": 3861
},
{
"epoch": 0.6863032564751878,
"grad_norm": 0.40526076592896804,
"learning_rate": 9.118258957620914e-06,
"loss": 1.3681,
"step": 3862
},
{
"epoch": 0.6864809631702874,
"grad_norm": 0.3412053101762149,
"learning_rate": 9.10879610503318e-06,
"loss": 1.2832,
"step": 3863
},
{
"epoch": 0.6866586698653872,
"grad_norm": 0.3526232078960032,
"learning_rate": 9.099336717142218e-06,
"loss": 1.2877,
"step": 3864
},
{
"epoch": 0.6868363765604869,
"grad_norm": 0.34741992507027797,
"learning_rate": 9.089880796957247e-06,
"loss": 1.3074,
"step": 3865
},
{
"epoch": 0.6870140832555867,
"grad_norm": 0.350807357831264,
"learning_rate": 9.08042834748637e-06,
"loss": 1.3199,
"step": 3866
},
{
"epoch": 0.6871917899506864,
"grad_norm": 0.3465907242452997,
"learning_rate": 9.070979371736588e-06,
"loss": 1.3209,
"step": 3867
},
{
"epoch": 0.6873694966457862,
"grad_norm": 0.3492554524679477,
"learning_rate": 9.061533872713797e-06,
"loss": 1.2898,
"step": 3868
},
{
"epoch": 0.6875472033408858,
"grad_norm": 0.3520787990429506,
"learning_rate": 9.052091853422789e-06,
"loss": 1.2978,
"step": 3869
},
{
"epoch": 0.6877249100359856,
"grad_norm": 0.3515819416325869,
"learning_rate": 9.042653316867245e-06,
"loss": 1.2691,
"step": 3870
},
{
"epoch": 0.6879026167310853,
"grad_norm": 0.3636882371897461,
"learning_rate": 9.033218266049743e-06,
"loss": 1.3462,
"step": 3871
},
{
"epoch": 0.6880803234261851,
"grad_norm": 0.3593152890861413,
"learning_rate": 9.023786703971752e-06,
"loss": 1.3208,
"step": 3872
},
{
"epoch": 0.6882580301212848,
"grad_norm": 0.3458349002267076,
"learning_rate": 9.01435863363362e-06,
"loss": 1.3032,
"step": 3873
},
{
"epoch": 0.6884357368163846,
"grad_norm": 0.36358408573780693,
"learning_rate": 9.004934058034614e-06,
"loss": 1.3861,
"step": 3874
},
{
"epoch": 0.6886134435114843,
"grad_norm": 0.3576962415044505,
"learning_rate": 8.995512980172849e-06,
"loss": 1.3152,
"step": 3875
},
{
"epoch": 0.688791150206584,
"grad_norm": 0.35522137847676755,
"learning_rate": 8.986095403045354e-06,
"loss": 1.3537,
"step": 3876
},
{
"epoch": 0.6889688569016837,
"grad_norm": 0.3550782216796053,
"learning_rate": 8.976681329648038e-06,
"loss": 1.3518,
"step": 3877
},
{
"epoch": 0.6891465635967835,
"grad_norm": 0.3441026674926366,
"learning_rate": 8.967270762975684e-06,
"loss": 1.2732,
"step": 3878
},
{
"epoch": 0.6893242702918833,
"grad_norm": 0.35125313565214344,
"learning_rate": 8.95786370602199e-06,
"loss": 1.3442,
"step": 3879
},
{
"epoch": 0.689501976986983,
"grad_norm": 0.3497812745100688,
"learning_rate": 8.948460161779506e-06,
"loss": 1.324,
"step": 3880
},
{
"epoch": 0.6896796836820828,
"grad_norm": 0.3424759127743685,
"learning_rate": 8.939060133239678e-06,
"loss": 1.3017,
"step": 3881
},
{
"epoch": 0.6898573903771824,
"grad_norm": 0.3438567276225371,
"learning_rate": 8.929663623392835e-06,
"loss": 1.3095,
"step": 3882
},
{
"epoch": 0.6900350970722822,
"grad_norm": 0.34587267663342874,
"learning_rate": 8.920270635228176e-06,
"loss": 1.3244,
"step": 3883
},
{
"epoch": 0.6902128037673819,
"grad_norm": 0.34610739874587604,
"learning_rate": 8.910881171733793e-06,
"loss": 1.3264,
"step": 3884
},
{
"epoch": 0.6903905104624817,
"grad_norm": 0.376653368887619,
"learning_rate": 8.901495235896654e-06,
"loss": 1.3153,
"step": 3885
},
{
"epoch": 0.6905682171575814,
"grad_norm": 0.35824236444933455,
"learning_rate": 8.892112830702593e-06,
"loss": 1.305,
"step": 3886
},
{
"epoch": 0.6907459238526812,
"grad_norm": 0.33991086498602324,
"learning_rate": 8.882733959136325e-06,
"loss": 1.2488,
"step": 3887
},
{
"epoch": 0.6909236305477809,
"grad_norm": 0.39287813235686503,
"learning_rate": 8.873358624181463e-06,
"loss": 1.3825,
"step": 3888
},
{
"epoch": 0.6911013372428806,
"grad_norm": 0.35582595169903597,
"learning_rate": 8.86398682882047e-06,
"loss": 1.2765,
"step": 3889
},
{
"epoch": 0.6912790439379803,
"grad_norm": 0.35312343985328937,
"learning_rate": 8.854618576034694e-06,
"loss": 1.3261,
"step": 3890
},
{
"epoch": 0.6914567506330801,
"grad_norm": 0.3458076960640693,
"learning_rate": 8.845253868804341e-06,
"loss": 1.3402,
"step": 3891
},
{
"epoch": 0.6916344573281799,
"grad_norm": 0.3467519080494391,
"learning_rate": 8.8358927101085e-06,
"loss": 1.2987,
"step": 3892
},
{
"epoch": 0.6918121640232796,
"grad_norm": 0.3495804512568741,
"learning_rate": 8.826535102925147e-06,
"loss": 1.3292,
"step": 3893
},
{
"epoch": 0.6919898707183794,
"grad_norm": 0.36314718981333227,
"learning_rate": 8.8171810502311e-06,
"loss": 1.3201,
"step": 3894
},
{
"epoch": 0.692167577413479,
"grad_norm": 0.3529063663094267,
"learning_rate": 8.807830555002068e-06,
"loss": 1.3698,
"step": 3895
},
{
"epoch": 0.6923452841085788,
"grad_norm": 0.3385510487763001,
"learning_rate": 8.798483620212612e-06,
"loss": 1.3039,
"step": 3896
},
{
"epoch": 0.6925229908036785,
"grad_norm": 0.35311072625994233,
"learning_rate": 8.78914024883617e-06,
"loss": 1.3165,
"step": 3897
},
{
"epoch": 0.6927006974987783,
"grad_norm": 0.3557320939412152,
"learning_rate": 8.779800443845046e-06,
"loss": 1.3217,
"step": 3898
},
{
"epoch": 0.692878404193878,
"grad_norm": 0.35310126778366924,
"learning_rate": 8.770464208210405e-06,
"loss": 1.3426,
"step": 3899
},
{
"epoch": 0.6930561108889778,
"grad_norm": 0.37139272367830295,
"learning_rate": 8.761131544902281e-06,
"loss": 1.316,
"step": 3900
},
{
"epoch": 0.6932338175840774,
"grad_norm": 0.3524010963975427,
"learning_rate": 8.751802456889562e-06,
"loss": 1.33,
"step": 3901
},
{
"epoch": 0.6934115242791772,
"grad_norm": 0.3382395308642985,
"learning_rate": 8.74247694714002e-06,
"loss": 1.2664,
"step": 3902
},
{
"epoch": 0.6935892309742769,
"grad_norm": 0.35332559372259437,
"learning_rate": 8.733155018620268e-06,
"loss": 1.3386,
"step": 3903
},
{
"epoch": 0.6937669376693767,
"grad_norm": 0.3448888004509563,
"learning_rate": 8.723836674295787e-06,
"loss": 1.3155,
"step": 3904
},
{
"epoch": 0.6939446443644764,
"grad_norm": 0.35373870513180455,
"learning_rate": 8.714521917130924e-06,
"loss": 1.3482,
"step": 3905
},
{
"epoch": 0.6941223510595762,
"grad_norm": 0.35049675791160956,
"learning_rate": 8.70521075008886e-06,
"loss": 1.3628,
"step": 3906
},
{
"epoch": 0.694300057754676,
"grad_norm": 0.3514838273498937,
"learning_rate": 8.695903176131671e-06,
"loss": 1.3319,
"step": 3907
},
{
"epoch": 0.6944777644497756,
"grad_norm": 0.3421308598660658,
"learning_rate": 8.686599198220265e-06,
"loss": 1.3002,
"step": 3908
},
{
"epoch": 0.6946554711448754,
"grad_norm": 0.34224169314493397,
"learning_rate": 8.677298819314411e-06,
"loss": 1.309,
"step": 3909
},
{
"epoch": 0.6948331778399751,
"grad_norm": 0.34039028040602154,
"learning_rate": 8.66800204237273e-06,
"loss": 1.2797,
"step": 3910
},
{
"epoch": 0.6950108845350749,
"grad_norm": 0.34837889703407493,
"learning_rate": 8.658708870352712e-06,
"loss": 1.3129,
"step": 3911
},
{
"epoch": 0.6951885912301746,
"grad_norm": 0.35351899423524386,
"learning_rate": 8.649419306210694e-06,
"loss": 1.3625,
"step": 3912
},
{
"epoch": 0.6953662979252744,
"grad_norm": 0.3513242458155008,
"learning_rate": 8.640133352901843e-06,
"loss": 1.3505,
"step": 3913
},
{
"epoch": 0.695544004620374,
"grad_norm": 0.5754239007237273,
"learning_rate": 8.630851013380207e-06,
"loss": 1.3266,
"step": 3914
},
{
"epoch": 0.6957217113154738,
"grad_norm": 0.3648997466926659,
"learning_rate": 8.621572290598662e-06,
"loss": 1.3338,
"step": 3915
},
{
"epoch": 0.6958994180105735,
"grad_norm": 0.35452825537346505,
"learning_rate": 8.612297187508958e-06,
"loss": 1.3683,
"step": 3916
},
{
"epoch": 0.6960771247056733,
"grad_norm": 0.34494993945443636,
"learning_rate": 8.603025707061675e-06,
"loss": 1.2809,
"step": 3917
},
{
"epoch": 0.696254831400773,
"grad_norm": 0.3568044832903926,
"learning_rate": 8.593757852206243e-06,
"loss": 1.3301,
"step": 3918
},
{
"epoch": 0.6964325380958728,
"grad_norm": 0.36008606097110607,
"learning_rate": 8.584493625890944e-06,
"loss": 1.3138,
"step": 3919
},
{
"epoch": 0.6966102447909726,
"grad_norm": 0.3449384137174798,
"learning_rate": 8.5752330310629e-06,
"loss": 1.2607,
"step": 3920
},
{
"epoch": 0.6967879514860722,
"grad_norm": 0.3632581491977349,
"learning_rate": 8.56597607066808e-06,
"loss": 1.3704,
"step": 3921
},
{
"epoch": 0.696965658181172,
"grad_norm": 0.34329757596997185,
"learning_rate": 8.5567227476513e-06,
"loss": 1.2915,
"step": 3922
},
{
"epoch": 0.6971433648762717,
"grad_norm": 0.3468110398761063,
"learning_rate": 8.547473064956216e-06,
"loss": 1.3032,
"step": 3923
},
{
"epoch": 0.6973210715713715,
"grad_norm": 0.35231656878723566,
"learning_rate": 8.538227025525314e-06,
"loss": 1.3259,
"step": 3924
},
{
"epoch": 0.6974987782664712,
"grad_norm": 0.3840620569645232,
"learning_rate": 8.528984632299953e-06,
"loss": 1.2972,
"step": 3925
},
{
"epoch": 0.697676484961571,
"grad_norm": 0.34263441309122117,
"learning_rate": 8.519745888220301e-06,
"loss": 1.3053,
"step": 3926
},
{
"epoch": 0.6978541916566706,
"grad_norm": 0.3504331525947938,
"learning_rate": 8.51051079622538e-06,
"loss": 1.3276,
"step": 3927
},
{
"epoch": 0.6980318983517704,
"grad_norm": 0.3502814897956183,
"learning_rate": 8.50127935925305e-06,
"loss": 1.2876,
"step": 3928
},
{
"epoch": 0.6982096050468701,
"grad_norm": 0.3484387996942799,
"learning_rate": 8.492051580239984e-06,
"loss": 1.324,
"step": 3929
},
{
"epoch": 0.6983873117419699,
"grad_norm": 0.35890679365830147,
"learning_rate": 8.482827462121735e-06,
"loss": 1.3487,
"step": 3930
},
{
"epoch": 0.6985650184370696,
"grad_norm": 0.35000140400661656,
"learning_rate": 8.47360700783266e-06,
"loss": 1.302,
"step": 3931
},
{
"epoch": 0.6987427251321694,
"grad_norm": 0.35448600152903265,
"learning_rate": 8.46439022030596e-06,
"loss": 1.327,
"step": 3932
},
{
"epoch": 0.698920431827269,
"grad_norm": 0.3477138758946661,
"learning_rate": 8.455177102473669e-06,
"loss": 1.3096,
"step": 3933
},
{
"epoch": 0.6990981385223688,
"grad_norm": 0.3465647607419958,
"learning_rate": 8.44596765726665e-06,
"loss": 1.3269,
"step": 3934
},
{
"epoch": 0.6992758452174686,
"grad_norm": 0.34931426068835386,
"learning_rate": 8.436761887614603e-06,
"loss": 1.3153,
"step": 3935
},
{
"epoch": 0.6994535519125683,
"grad_norm": 0.34968883667241074,
"learning_rate": 8.427559796446054e-06,
"loss": 1.3217,
"step": 3936
},
{
"epoch": 0.6996312586076681,
"grad_norm": 0.33757302005586765,
"learning_rate": 8.418361386688366e-06,
"loss": 1.2782,
"step": 3937
},
{
"epoch": 0.6998089653027678,
"grad_norm": 0.34935239853079925,
"learning_rate": 8.409166661267717e-06,
"loss": 1.3474,
"step": 3938
},
{
"epoch": 0.6999866719978676,
"grad_norm": 0.34392343144377996,
"learning_rate": 8.399975623109133e-06,
"loss": 1.3173,
"step": 3939
},
{
"epoch": 0.7001643786929672,
"grad_norm": 0.3522976694693785,
"learning_rate": 8.390788275136452e-06,
"loss": 1.3536,
"step": 3940
},
{
"epoch": 0.700342085388067,
"grad_norm": 0.3503867063301098,
"learning_rate": 8.381604620272343e-06,
"loss": 1.3431,
"step": 3941
},
{
"epoch": 0.7005197920831667,
"grad_norm": 0.3493524330599995,
"learning_rate": 8.372424661438296e-06,
"loss": 1.3083,
"step": 3942
},
{
"epoch": 0.7006974987782665,
"grad_norm": 0.3464932463128864,
"learning_rate": 8.363248401554633e-06,
"loss": 1.3329,
"step": 3943
},
{
"epoch": 0.7008752054733662,
"grad_norm": 0.3383381776211295,
"learning_rate": 8.35407584354049e-06,
"loss": 1.2819,
"step": 3944
},
{
"epoch": 0.701052912168466,
"grad_norm": 0.3498436196579954,
"learning_rate": 8.344906990313834e-06,
"loss": 1.3562,
"step": 3945
},
{
"epoch": 0.7012306188635656,
"grad_norm": 0.3425480591737586,
"learning_rate": 8.33574184479145e-06,
"loss": 1.3235,
"step": 3946
},
{
"epoch": 0.7014083255586654,
"grad_norm": 0.34241297181992103,
"learning_rate": 8.326580409888938e-06,
"loss": 1.3216,
"step": 3947
},
{
"epoch": 0.7015860322537651,
"grad_norm": 0.34229220872694405,
"learning_rate": 8.317422688520722e-06,
"loss": 1.3038,
"step": 3948
},
{
"epoch": 0.7017637389488649,
"grad_norm": 0.3411831955276525,
"learning_rate": 8.308268683600053e-06,
"loss": 1.3066,
"step": 3949
},
{
"epoch": 0.7019414456439647,
"grad_norm": 0.35000061774590346,
"learning_rate": 8.299118398038999e-06,
"loss": 1.3565,
"step": 3950
},
{
"epoch": 0.7021191523390644,
"grad_norm": 0.34654847879027195,
"learning_rate": 8.289971834748421e-06,
"loss": 1.2992,
"step": 3951
},
{
"epoch": 0.7022968590341642,
"grad_norm": 0.34117614456582507,
"learning_rate": 8.280828996638009e-06,
"loss": 1.2814,
"step": 3952
},
{
"epoch": 0.7024745657292638,
"grad_norm": 0.33929488411192116,
"learning_rate": 8.271689886616292e-06,
"loss": 1.3097,
"step": 3953
},
{
"epoch": 0.7026522724243636,
"grad_norm": 0.3441968601951114,
"learning_rate": 8.26255450759058e-06,
"loss": 1.2855,
"step": 3954
},
{
"epoch": 0.7028299791194633,
"grad_norm": 0.3599839202046591,
"learning_rate": 8.253422862467016e-06,
"loss": 1.3714,
"step": 3955
},
{
"epoch": 0.7030076858145631,
"grad_norm": 0.35639372494148774,
"learning_rate": 8.24429495415054e-06,
"loss": 1.3575,
"step": 3956
},
{
"epoch": 0.7031853925096628,
"grad_norm": 0.34173256125656676,
"learning_rate": 8.235170785544915e-06,
"loss": 1.3175,
"step": 3957
},
{
"epoch": 0.7033630992047626,
"grad_norm": 0.34572341584151767,
"learning_rate": 8.226050359552713e-06,
"loss": 1.2883,
"step": 3958
},
{
"epoch": 0.7035408058998622,
"grad_norm": 0.3622428039486531,
"learning_rate": 8.216933679075309e-06,
"loss": 1.3137,
"step": 3959
},
{
"epoch": 0.703718512594962,
"grad_norm": 0.3478312718082184,
"learning_rate": 8.207820747012894e-06,
"loss": 1.3191,
"step": 3960
},
{
"epoch": 0.7038962192900617,
"grad_norm": 0.3484189733944188,
"learning_rate": 8.19871156626446e-06,
"loss": 1.3476,
"step": 3961
},
{
"epoch": 0.7040739259851615,
"grad_norm": 0.3524979863696322,
"learning_rate": 8.189606139727802e-06,
"loss": 1.309,
"step": 3962
},
{
"epoch": 0.7042516326802613,
"grad_norm": 0.3412157333337828,
"learning_rate": 8.180504470299543e-06,
"loss": 1.268,
"step": 3963
},
{
"epoch": 0.704429339375361,
"grad_norm": 0.3613012298947117,
"learning_rate": 8.171406560875088e-06,
"loss": 1.3541,
"step": 3964
},
{
"epoch": 0.7046070460704607,
"grad_norm": 0.34073489822029107,
"learning_rate": 8.16231241434866e-06,
"loss": 1.3,
"step": 3965
},
{
"epoch": 0.7047847527655604,
"grad_norm": 0.341139284457699,
"learning_rate": 8.153222033613254e-06,
"loss": 1.2895,
"step": 3966
},
{
"epoch": 0.7049624594606602,
"grad_norm": 0.3540906811925608,
"learning_rate": 8.14413542156072e-06,
"loss": 1.3179,
"step": 3967
},
{
"epoch": 0.7051401661557599,
"grad_norm": 0.3372616665020818,
"learning_rate": 8.135052581081664e-06,
"loss": 1.25,
"step": 3968
},
{
"epoch": 0.7053178728508597,
"grad_norm": 0.3643193185813,
"learning_rate": 8.125973515065513e-06,
"loss": 1.345,
"step": 3969
},
{
"epoch": 0.7054955795459594,
"grad_norm": 0.3532096421003043,
"learning_rate": 8.116898226400488e-06,
"loss": 1.3209,
"step": 3970
},
{
"epoch": 0.7056732862410592,
"grad_norm": 0.3426036990977531,
"learning_rate": 8.107826717973603e-06,
"loss": 1.3178,
"step": 3971
},
{
"epoch": 0.7058509929361588,
"grad_norm": 0.34272341383217203,
"learning_rate": 8.098758992670694e-06,
"loss": 1.2719,
"step": 3972
},
{
"epoch": 0.7060286996312586,
"grad_norm": 0.3501486965647963,
"learning_rate": 8.089695053376357e-06,
"loss": 1.3424,
"step": 3973
},
{
"epoch": 0.7062064063263583,
"grad_norm": 0.34658752754345223,
"learning_rate": 8.080634902974005e-06,
"loss": 1.345,
"step": 3974
},
{
"epoch": 0.7063841130214581,
"grad_norm": 0.34054983774552056,
"learning_rate": 8.071578544345846e-06,
"loss": 1.3197,
"step": 3975
},
{
"epoch": 0.7065618197165578,
"grad_norm": 0.34590784400627356,
"learning_rate": 8.062525980372867e-06,
"loss": 1.315,
"step": 3976
},
{
"epoch": 0.7067395264116576,
"grad_norm": 0.351466989359575,
"learning_rate": 8.053477213934876e-06,
"loss": 1.3063,
"step": 3977
},
{
"epoch": 0.7069172331067572,
"grad_norm": 0.34081769016207203,
"learning_rate": 8.044432247910448e-06,
"loss": 1.3034,
"step": 3978
},
{
"epoch": 0.707094939801857,
"grad_norm": 0.346078724252139,
"learning_rate": 8.035391085176955e-06,
"loss": 1.3295,
"step": 3979
},
{
"epoch": 0.7072726464969568,
"grad_norm": 0.34587198639750866,
"learning_rate": 8.026353728610565e-06,
"loss": 1.3414,
"step": 3980
},
{
"epoch": 0.7074503531920565,
"grad_norm": 0.3474039466119413,
"learning_rate": 8.017320181086225e-06,
"loss": 1.3143,
"step": 3981
},
{
"epoch": 0.7076280598871563,
"grad_norm": 0.3577269611721558,
"learning_rate": 8.008290445477682e-06,
"loss": 1.3402,
"step": 3982
},
{
"epoch": 0.707805766582256,
"grad_norm": 0.34034665683387316,
"learning_rate": 7.999264524657464e-06,
"loss": 1.2999,
"step": 3983
},
{
"epoch": 0.7079834732773558,
"grad_norm": 0.34747637923791336,
"learning_rate": 7.990242421496883e-06,
"loss": 1.3347,
"step": 3984
},
{
"epoch": 0.7081611799724554,
"grad_norm": 0.34619691701462574,
"learning_rate": 7.981224138866032e-06,
"loss": 1.3287,
"step": 3985
},
{
"epoch": 0.7083388866675552,
"grad_norm": 0.3422313452976539,
"learning_rate": 7.972209679633815e-06,
"loss": 1.2921,
"step": 3986
},
{
"epoch": 0.7085165933626549,
"grad_norm": 0.3464733736640523,
"learning_rate": 7.9631990466679e-06,
"loss": 1.316,
"step": 3987
},
{
"epoch": 0.7086943000577547,
"grad_norm": 0.3423705818410509,
"learning_rate": 7.954192242834723e-06,
"loss": 1.2908,
"step": 3988
},
{
"epoch": 0.7088720067528544,
"grad_norm": 0.34020800221406955,
"learning_rate": 7.945189270999523e-06,
"loss": 1.3079,
"step": 3989
},
{
"epoch": 0.7090497134479542,
"grad_norm": 0.34698100448206765,
"learning_rate": 7.936190134026311e-06,
"loss": 1.3264,
"step": 3990
},
{
"epoch": 0.7092274201430538,
"grad_norm": 0.3443937743124649,
"learning_rate": 7.927194834777895e-06,
"loss": 1.3312,
"step": 3991
},
{
"epoch": 0.7094051268381536,
"grad_norm": 0.33673194734875,
"learning_rate": 7.91820337611584e-06,
"loss": 1.3038,
"step": 3992
},
{
"epoch": 0.7095828335332534,
"grad_norm": 0.3422663540043533,
"learning_rate": 7.909215760900501e-06,
"loss": 1.334,
"step": 3993
},
{
"epoch": 0.7097605402283531,
"grad_norm": 0.3498199072685063,
"learning_rate": 7.900231991991006e-06,
"loss": 1.3254,
"step": 3994
},
{
"epoch": 0.7099382469234529,
"grad_norm": 0.3360262833600118,
"learning_rate": 7.891252072245258e-06,
"loss": 1.2845,
"step": 3995
},
{
"epoch": 0.7101159536185526,
"grad_norm": 0.34796067265217623,
"learning_rate": 7.882276004519944e-06,
"loss": 1.3279,
"step": 3996
},
{
"epoch": 0.7102936603136523,
"grad_norm": 0.33110660065948805,
"learning_rate": 7.873303791670518e-06,
"loss": 1.2562,
"step": 3997
},
{
"epoch": 0.710471367008752,
"grad_norm": 0.3420231646079241,
"learning_rate": 7.864335436551205e-06,
"loss": 1.2929,
"step": 3998
},
{
"epoch": 0.7106490737038518,
"grad_norm": 0.3410278854590192,
"learning_rate": 7.855370942015006e-06,
"loss": 1.2791,
"step": 3999
},
{
"epoch": 0.7108267803989515,
"grad_norm": 0.34675208935634877,
"learning_rate": 7.846410310913707e-06,
"loss": 1.3168,
"step": 4000
},
{
"epoch": 0.7110044870940513,
"grad_norm": 0.34183227330689736,
"learning_rate": 7.837453546097846e-06,
"loss": 1.2804,
"step": 4001
},
{
"epoch": 0.711182193789151,
"grad_norm": 0.33850664161704874,
"learning_rate": 7.828500650416739e-06,
"loss": 1.3008,
"step": 4002
},
{
"epoch": 0.7113599004842508,
"grad_norm": 0.34036328946114275,
"learning_rate": 7.819551626718478e-06,
"loss": 1.3185,
"step": 4003
},
{
"epoch": 0.7115376071793504,
"grad_norm": 0.34141692013695923,
"learning_rate": 7.810606477849894e-06,
"loss": 1.3143,
"step": 4004
},
{
"epoch": 0.7117153138744502,
"grad_norm": 0.3481182488340219,
"learning_rate": 7.801665206656628e-06,
"loss": 1.3439,
"step": 4005
},
{
"epoch": 0.71189302056955,
"grad_norm": 0.3431330395920456,
"learning_rate": 7.79272781598306e-06,
"loss": 1.3019,
"step": 4006
},
{
"epoch": 0.7120707272646497,
"grad_norm": 0.3416949547630586,
"learning_rate": 7.783794308672343e-06,
"loss": 1.2939,
"step": 4007
},
{
"epoch": 0.7122484339597495,
"grad_norm": 0.34532845708064197,
"learning_rate": 7.774864687566383e-06,
"loss": 1.3105,
"step": 4008
},
{
"epoch": 0.7124261406548492,
"grad_norm": 0.34823823614235877,
"learning_rate": 7.765938955505887e-06,
"loss": 1.3195,
"step": 4009
},
{
"epoch": 0.7126038473499489,
"grad_norm": 0.352630760831812,
"learning_rate": 7.757017115330272e-06,
"loss": 1.3538,
"step": 4010
},
{
"epoch": 0.7127815540450486,
"grad_norm": 0.347704388946326,
"learning_rate": 7.748099169877752e-06,
"loss": 1.3023,
"step": 4011
},
{
"epoch": 0.7129592607401484,
"grad_norm": 0.35236081601923486,
"learning_rate": 7.739185121985295e-06,
"loss": 1.3146,
"step": 4012
},
{
"epoch": 0.7131369674352481,
"grad_norm": 0.3432301896677569,
"learning_rate": 7.730274974488616e-06,
"loss": 1.3184,
"step": 4013
},
{
"epoch": 0.7133146741303479,
"grad_norm": 0.36597400709637956,
"learning_rate": 7.721368730222221e-06,
"loss": 1.2544,
"step": 4014
},
{
"epoch": 0.7134923808254476,
"grad_norm": 0.3566758927394462,
"learning_rate": 7.71246639201934e-06,
"loss": 1.3168,
"step": 4015
},
{
"epoch": 0.7136700875205474,
"grad_norm": 0.339243959835646,
"learning_rate": 7.703567962711978e-06,
"loss": 1.3074,
"step": 4016
},
{
"epoch": 0.713847794215647,
"grad_norm": 0.34177152975985786,
"learning_rate": 7.694673445130891e-06,
"loss": 1.3383,
"step": 4017
},
{
"epoch": 0.7140255009107468,
"grad_norm": 0.34138406396791965,
"learning_rate": 7.685782842105593e-06,
"loss": 1.3008,
"step": 4018
},
{
"epoch": 0.7142032076058465,
"grad_norm": 0.3382462719100332,
"learning_rate": 7.676896156464355e-06,
"loss": 1.3059,
"step": 4019
},
{
"epoch": 0.7143809143009463,
"grad_norm": 0.34249480498420126,
"learning_rate": 7.668013391034194e-06,
"loss": 1.2883,
"step": 4020
},
{
"epoch": 0.7145586209960461,
"grad_norm": 0.3466544654579431,
"learning_rate": 7.659134548640888e-06,
"loss": 1.3185,
"step": 4021
},
{
"epoch": 0.7147363276911458,
"grad_norm": 0.3494239541877539,
"learning_rate": 7.650259632108953e-06,
"loss": 1.3409,
"step": 4022
},
{
"epoch": 0.7149140343862455,
"grad_norm": 0.35331316732504076,
"learning_rate": 7.641388644261684e-06,
"loss": 1.3346,
"step": 4023
},
{
"epoch": 0.7150917410813452,
"grad_norm": 0.3445995224850807,
"learning_rate": 7.632521587921102e-06,
"loss": 1.3024,
"step": 4024
},
{
"epoch": 0.715269447776445,
"grad_norm": 0.3381822813699869,
"learning_rate": 7.6236584659079905e-06,
"loss": 1.2844,
"step": 4025
},
{
"epoch": 0.7154471544715447,
"grad_norm": 0.3472168123410575,
"learning_rate": 7.614799281041863e-06,
"loss": 1.3421,
"step": 4026
},
{
"epoch": 0.7156248611666445,
"grad_norm": 0.347430143457322,
"learning_rate": 7.605944036140991e-06,
"loss": 1.3513,
"step": 4027
},
{
"epoch": 0.7158025678617442,
"grad_norm": 0.33979226368048715,
"learning_rate": 7.597092734022406e-06,
"loss": 1.3064,
"step": 4028
},
{
"epoch": 0.7159802745568439,
"grad_norm": 0.34501772968276934,
"learning_rate": 7.588245377501872e-06,
"loss": 1.3024,
"step": 4029
},
{
"epoch": 0.7161579812519436,
"grad_norm": 0.3518169192338613,
"learning_rate": 7.579401969393898e-06,
"loss": 1.3407,
"step": 4030
},
{
"epoch": 0.7163356879470434,
"grad_norm": 0.3420862864297147,
"learning_rate": 7.5705625125117345e-06,
"loss": 1.3177,
"step": 4031
},
{
"epoch": 0.7165133946421431,
"grad_norm": 0.3463352084097385,
"learning_rate": 7.561727009667383e-06,
"loss": 1.3183,
"step": 4032
},
{
"epoch": 0.7166911013372429,
"grad_norm": 0.34358667301230644,
"learning_rate": 7.552895463671583e-06,
"loss": 1.3044,
"step": 4033
},
{
"epoch": 0.7168688080323427,
"grad_norm": 0.34285658528137114,
"learning_rate": 7.544067877333814e-06,
"loss": 1.2963,
"step": 4034
},
{
"epoch": 0.7170465147274424,
"grad_norm": 0.3353324955239236,
"learning_rate": 7.535244253462295e-06,
"loss": 1.3215,
"step": 4035
},
{
"epoch": 0.717224221422542,
"grad_norm": 0.3647432128022017,
"learning_rate": 7.526424594863986e-06,
"loss": 1.2674,
"step": 4036
},
{
"epoch": 0.7174019281176418,
"grad_norm": 0.3469022434969301,
"learning_rate": 7.517608904344593e-06,
"loss": 1.3374,
"step": 4037
},
{
"epoch": 0.7175796348127416,
"grad_norm": 0.35854988713924324,
"learning_rate": 7.5087971847085515e-06,
"loss": 1.322,
"step": 4038
},
{
"epoch": 0.7177573415078413,
"grad_norm": 0.3368566693610481,
"learning_rate": 7.499989438759032e-06,
"loss": 1.2736,
"step": 4039
},
{
"epoch": 0.7179350482029411,
"grad_norm": 0.347420294139603,
"learning_rate": 7.491185669297953e-06,
"loss": 1.3267,
"step": 4040
},
{
"epoch": 0.7181127548980408,
"grad_norm": 0.33982651985085394,
"learning_rate": 7.482385879125939e-06,
"loss": 1.2984,
"step": 4041
},
{
"epoch": 0.7182904615931405,
"grad_norm": 0.3475523515893918,
"learning_rate": 7.473590071042387e-06,
"loss": 1.3341,
"step": 4042
},
{
"epoch": 0.7184681682882402,
"grad_norm": 0.3381308300075107,
"learning_rate": 7.464798247845402e-06,
"loss": 1.2497,
"step": 4043
},
{
"epoch": 0.71864587498334,
"grad_norm": 0.3429143723225144,
"learning_rate": 7.4560104123318314e-06,
"loss": 1.2854,
"step": 4044
},
{
"epoch": 0.7188235816784397,
"grad_norm": 0.3431059885096514,
"learning_rate": 7.447226567297246e-06,
"loss": 1.3307,
"step": 4045
},
{
"epoch": 0.7190012883735395,
"grad_norm": 0.3465359403232201,
"learning_rate": 7.43844671553595e-06,
"loss": 1.3172,
"step": 4046
},
{
"epoch": 0.7191789950686392,
"grad_norm": 0.33063716837556145,
"learning_rate": 7.429670859840998e-06,
"loss": 1.2282,
"step": 4047
},
{
"epoch": 0.719356701763739,
"grad_norm": 0.3367392043821428,
"learning_rate": 7.420899003004134e-06,
"loss": 1.3208,
"step": 4048
},
{
"epoch": 0.7195344084588386,
"grad_norm": 0.3630552362231714,
"learning_rate": 7.41213114781586e-06,
"loss": 1.3438,
"step": 4049
},
{
"epoch": 0.7197121151539384,
"grad_norm": 0.35012777203967266,
"learning_rate": 7.403367297065383e-06,
"loss": 1.355,
"step": 4050
},
{
"epoch": 0.7198898218490382,
"grad_norm": 0.3367638795089637,
"learning_rate": 7.394607453540667e-06,
"loss": 1.2879,
"step": 4051
},
{
"epoch": 0.7200675285441379,
"grad_norm": 0.34522156270100673,
"learning_rate": 7.385851620028377e-06,
"loss": 1.2753,
"step": 4052
},
{
"epoch": 0.7202452352392377,
"grad_norm": 0.3469652716402909,
"learning_rate": 7.377099799313905e-06,
"loss": 1.3168,
"step": 4053
},
{
"epoch": 0.7204229419343374,
"grad_norm": 0.3470029865285727,
"learning_rate": 7.368351994181371e-06,
"loss": 1.3088,
"step": 4054
},
{
"epoch": 0.7206006486294371,
"grad_norm": 0.34324610403960204,
"learning_rate": 7.359608207413615e-06,
"loss": 1.2985,
"step": 4055
},
{
"epoch": 0.7207783553245368,
"grad_norm": 0.3528790223246071,
"learning_rate": 7.350868441792205e-06,
"loss": 1.3156,
"step": 4056
},
{
"epoch": 0.7209560620196366,
"grad_norm": 0.3866584183753084,
"learning_rate": 7.34213270009742e-06,
"loss": 1.2976,
"step": 4057
},
{
"epoch": 0.7211337687147363,
"grad_norm": 0.3582539544670879,
"learning_rate": 7.333400985108263e-06,
"loss": 1.3414,
"step": 4058
},
{
"epoch": 0.7213114754098361,
"grad_norm": 0.34666141012363766,
"learning_rate": 7.324673299602461e-06,
"loss": 1.3381,
"step": 4059
},
{
"epoch": 0.7214891821049358,
"grad_norm": 0.34182546927180485,
"learning_rate": 7.315949646356444e-06,
"loss": 1.3111,
"step": 4060
},
{
"epoch": 0.7216668888000355,
"grad_norm": 0.34655302734998517,
"learning_rate": 7.307230028145387e-06,
"loss": 1.2995,
"step": 4061
},
{
"epoch": 0.7218445954951352,
"grad_norm": 0.34892818950818083,
"learning_rate": 7.298514447743161e-06,
"loss": 1.301,
"step": 4062
},
{
"epoch": 0.722022302190235,
"grad_norm": 0.33798553604141823,
"learning_rate": 7.2898029079223474e-06,
"loss": 1.2718,
"step": 4063
},
{
"epoch": 0.7222000088853348,
"grad_norm": 0.34614319835874274,
"learning_rate": 7.281095411454247e-06,
"loss": 1.2934,
"step": 4064
},
{
"epoch": 0.7223777155804345,
"grad_norm": 0.3470620820615626,
"learning_rate": 7.272391961108891e-06,
"loss": 1.3114,
"step": 4065
},
{
"epoch": 0.7225554222755343,
"grad_norm": 0.34946085445134856,
"learning_rate": 7.263692559655009e-06,
"loss": 1.3305,
"step": 4066
},
{
"epoch": 0.722733128970634,
"grad_norm": 0.3434813536069956,
"learning_rate": 7.254997209860038e-06,
"loss": 1.3119,
"step": 4067
},
{
"epoch": 0.7229108356657337,
"grad_norm": 0.34061548703717626,
"learning_rate": 7.246305914490137e-06,
"loss": 1.3081,
"step": 4068
},
{
"epoch": 0.7230885423608334,
"grad_norm": 0.451485966902142,
"learning_rate": 7.237618676310168e-06,
"loss": 1.3142,
"step": 4069
},
{
"epoch": 0.7232662490559332,
"grad_norm": 0.3552065709311755,
"learning_rate": 7.228935498083705e-06,
"loss": 1.3262,
"step": 4070
},
{
"epoch": 0.7234439557510329,
"grad_norm": 0.3549082974820173,
"learning_rate": 7.22025638257303e-06,
"loss": 1.3303,
"step": 4071
},
{
"epoch": 0.7236216624461327,
"grad_norm": 0.42694901708429145,
"learning_rate": 7.211581332539132e-06,
"loss": 1.3277,
"step": 4072
},
{
"epoch": 0.7237993691412324,
"grad_norm": 0.34716486715440426,
"learning_rate": 7.202910350741712e-06,
"loss": 1.3069,
"step": 4073
},
{
"epoch": 0.7239770758363321,
"grad_norm": 0.35786957430164473,
"learning_rate": 7.194243439939163e-06,
"loss": 1.3364,
"step": 4074
},
{
"epoch": 0.7241547825314318,
"grad_norm": 0.34837529444815096,
"learning_rate": 7.1855806028886045e-06,
"loss": 1.3249,
"step": 4075
},
{
"epoch": 0.7243324892265316,
"grad_norm": 0.35712499276661647,
"learning_rate": 7.176921842345843e-06,
"loss": 1.3715,
"step": 4076
},
{
"epoch": 0.7245101959216314,
"grad_norm": 0.34561545388645853,
"learning_rate": 7.168267161065392e-06,
"loss": 1.2739,
"step": 4077
},
{
"epoch": 0.7246879026167311,
"grad_norm": 0.34708008230067033,
"learning_rate": 7.159616561800467e-06,
"loss": 1.3014,
"step": 4078
},
{
"epoch": 0.7248656093118309,
"grad_norm": 0.36259132305698893,
"learning_rate": 7.15097004730299e-06,
"loss": 1.32,
"step": 4079
},
{
"epoch": 0.7250433160069306,
"grad_norm": 0.3470138069793676,
"learning_rate": 7.142327620323577e-06,
"loss": 1.2867,
"step": 4080
},
{
"epoch": 0.7252210227020303,
"grad_norm": 0.3497425940483375,
"learning_rate": 7.133689283611547e-06,
"loss": 1.3321,
"step": 4081
},
{
"epoch": 0.72539872939713,
"grad_norm": 0.41960512271340883,
"learning_rate": 7.125055039914919e-06,
"loss": 1.3049,
"step": 4082
},
{
"epoch": 0.7255764360922298,
"grad_norm": 0.344144185430658,
"learning_rate": 7.116424891980398e-06,
"loss": 1.2849,
"step": 4083
},
{
"epoch": 0.7257541427873295,
"grad_norm": 0.34913856932352094,
"learning_rate": 7.107798842553415e-06,
"loss": 1.3384,
"step": 4084
},
{
"epoch": 0.7259318494824293,
"grad_norm": 0.3563746669445148,
"learning_rate": 7.099176894378072e-06,
"loss": 1.3434,
"step": 4085
},
{
"epoch": 0.726109556177529,
"grad_norm": 0.3469724219845676,
"learning_rate": 7.090559050197165e-06,
"loss": 1.3028,
"step": 4086
},
{
"epoch": 0.7262872628726287,
"grad_norm": 0.34028358149857274,
"learning_rate": 7.081945312752198e-06,
"loss": 1.3105,
"step": 4087
},
{
"epoch": 0.7264649695677284,
"grad_norm": 0.348351351215965,
"learning_rate": 7.0733356847833555e-06,
"loss": 1.3176,
"step": 4088
},
{
"epoch": 0.7266426762628282,
"grad_norm": 0.360722530438116,
"learning_rate": 7.064730169029534e-06,
"loss": 1.3691,
"step": 4089
},
{
"epoch": 0.726820382957928,
"grad_norm": 0.34727566210409055,
"learning_rate": 7.056128768228305e-06,
"loss": 1.3009,
"step": 4090
},
{
"epoch": 0.7269980896530277,
"grad_norm": 0.34916300716344684,
"learning_rate": 7.047531485115935e-06,
"loss": 1.3229,
"step": 4091
},
{
"epoch": 0.7271757963481275,
"grad_norm": 0.34149868971740843,
"learning_rate": 7.03893832242738e-06,
"loss": 1.3002,
"step": 4092
},
{
"epoch": 0.7273535030432271,
"grad_norm": 0.3468503418429682,
"learning_rate": 7.030349282896291e-06,
"loss": 1.2819,
"step": 4093
},
{
"epoch": 0.7275312097383269,
"grad_norm": 0.3431632074066352,
"learning_rate": 7.021764369254999e-06,
"loss": 1.3127,
"step": 4094
},
{
"epoch": 0.7277089164334266,
"grad_norm": 0.3542849150466704,
"learning_rate": 7.013183584234529e-06,
"loss": 1.3067,
"step": 4095
},
{
"epoch": 0.7278866231285264,
"grad_norm": 0.3525862608494627,
"learning_rate": 7.004606930564588e-06,
"loss": 1.3544,
"step": 4096
},
{
"epoch": 0.7280643298236261,
"grad_norm": 0.3465448671792795,
"learning_rate": 6.996034410973564e-06,
"loss": 1.3169,
"step": 4097
},
{
"epoch": 0.7282420365187259,
"grad_norm": 0.34832229496300704,
"learning_rate": 6.987466028188552e-06,
"loss": 1.3274,
"step": 4098
},
{
"epoch": 0.7284197432138256,
"grad_norm": 0.34380558206787776,
"learning_rate": 6.978901784935308e-06,
"loss": 1.3274,
"step": 4099
},
{
"epoch": 0.7285974499089253,
"grad_norm": 0.35106469732476286,
"learning_rate": 6.970341683938287e-06,
"loss": 1.3288,
"step": 4100
},
{
"epoch": 0.728775156604025,
"grad_norm": 0.34420146349304,
"learning_rate": 6.961785727920602e-06,
"loss": 1.3052,
"step": 4101
},
{
"epoch": 0.7289528632991248,
"grad_norm": 0.34610023070689433,
"learning_rate": 6.9532339196040655e-06,
"loss": 1.3085,
"step": 4102
},
{
"epoch": 0.7291305699942245,
"grad_norm": 0.3448137343818155,
"learning_rate": 6.9446862617091815e-06,
"loss": 1.2863,
"step": 4103
},
{
"epoch": 0.7293082766893243,
"grad_norm": 0.40140533271955176,
"learning_rate": 6.9361427569551136e-06,
"loss": 1.3476,
"step": 4104
},
{
"epoch": 0.729485983384424,
"grad_norm": 0.34531078420635725,
"learning_rate": 6.927603408059711e-06,
"loss": 1.3068,
"step": 4105
},
{
"epoch": 0.7296636900795237,
"grad_norm": 0.352143152771808,
"learning_rate": 6.919068217739495e-06,
"loss": 1.3085,
"step": 4106
},
{
"epoch": 0.7298413967746235,
"grad_norm": 0.34309217694914285,
"learning_rate": 6.91053718870969e-06,
"loss": 1.337,
"step": 4107
},
{
"epoch": 0.7300191034697232,
"grad_norm": 0.3476586223839192,
"learning_rate": 6.902010323684158e-06,
"loss": 1.3247,
"step": 4108
},
{
"epoch": 0.730196810164823,
"grad_norm": 0.3482077589799202,
"learning_rate": 6.893487625375461e-06,
"loss": 1.3128,
"step": 4109
},
{
"epoch": 0.7303745168599227,
"grad_norm": 0.3431787292057853,
"learning_rate": 6.884969096494829e-06,
"loss": 1.3171,
"step": 4110
},
{
"epoch": 0.7305522235550225,
"grad_norm": 0.34465355469375686,
"learning_rate": 6.876454739752159e-06,
"loss": 1.3314,
"step": 4111
},
{
"epoch": 0.7307299302501222,
"grad_norm": 0.34698593914063824,
"learning_rate": 6.867944557856043e-06,
"loss": 1.2894,
"step": 4112
},
{
"epoch": 0.7309076369452219,
"grad_norm": 0.341698239442693,
"learning_rate": 6.859438553513724e-06,
"loss": 1.3354,
"step": 4113
},
{
"epoch": 0.7310853436403216,
"grad_norm": 0.34405011118853907,
"learning_rate": 6.850936729431119e-06,
"loss": 1.3026,
"step": 4114
},
{
"epoch": 0.7312630503354214,
"grad_norm": 0.3502559848729948,
"learning_rate": 6.84243908831282e-06,
"loss": 1.3304,
"step": 4115
},
{
"epoch": 0.7314407570305211,
"grad_norm": 0.3429752841644757,
"learning_rate": 6.833945632862084e-06,
"loss": 1.3057,
"step": 4116
},
{
"epoch": 0.7316184637256209,
"grad_norm": 0.344253757788228,
"learning_rate": 6.825456365780845e-06,
"loss": 1.3089,
"step": 4117
},
{
"epoch": 0.7317961704207206,
"grad_norm": 0.3425500911071719,
"learning_rate": 6.816971289769692e-06,
"loss": 1.3062,
"step": 4118
},
{
"epoch": 0.7319738771158203,
"grad_norm": 0.3428687776664972,
"learning_rate": 6.80849040752789e-06,
"loss": 1.3224,
"step": 4119
},
{
"epoch": 0.73215158381092,
"grad_norm": 0.35522480102697357,
"learning_rate": 6.800013721753367e-06,
"loss": 1.3496,
"step": 4120
},
{
"epoch": 0.7323292905060198,
"grad_norm": 0.34078073076737286,
"learning_rate": 6.791541235142709e-06,
"loss": 1.3146,
"step": 4121
},
{
"epoch": 0.7325069972011196,
"grad_norm": 0.34288660776948277,
"learning_rate": 6.783072950391194e-06,
"loss": 1.3275,
"step": 4122
},
{
"epoch": 0.7326847038962193,
"grad_norm": 0.3550551755858433,
"learning_rate": 6.77460887019272e-06,
"loss": 1.3691,
"step": 4123
},
{
"epoch": 0.7328624105913191,
"grad_norm": 0.34255439066181315,
"learning_rate": 6.766148997239883e-06,
"loss": 1.3111,
"step": 4124
},
{
"epoch": 0.7330401172864187,
"grad_norm": 0.3375395770129814,
"learning_rate": 6.7576933342239134e-06,
"loss": 1.2864,
"step": 4125
},
{
"epoch": 0.7332178239815185,
"grad_norm": 0.34675218625351384,
"learning_rate": 6.749241883834736e-06,
"loss": 1.3107,
"step": 4126
},
{
"epoch": 0.7333955306766182,
"grad_norm": 0.33870254613389145,
"learning_rate": 6.740794648760907e-06,
"loss": 1.2837,
"step": 4127
},
{
"epoch": 0.733573237371718,
"grad_norm": 0.35487052404645986,
"learning_rate": 6.7323516316896505e-06,
"loss": 1.3572,
"step": 4128
},
{
"epoch": 0.7337509440668177,
"grad_norm": 0.3400765151220762,
"learning_rate": 6.72391283530685e-06,
"loss": 1.2947,
"step": 4129
},
{
"epoch": 0.7339286507619175,
"grad_norm": 0.3349964524944202,
"learning_rate": 6.715478262297041e-06,
"loss": 1.2889,
"step": 4130
},
{
"epoch": 0.7341063574570172,
"grad_norm": 0.3518392883260942,
"learning_rate": 6.7070479153434276e-06,
"loss": 1.3081,
"step": 4131
},
{
"epoch": 0.7342840641521169,
"grad_norm": 0.343256862727128,
"learning_rate": 6.698621797127855e-06,
"loss": 1.3041,
"step": 4132
},
{
"epoch": 0.7344617708472166,
"grad_norm": 0.35001328060577014,
"learning_rate": 6.690199910330835e-06,
"loss": 1.3662,
"step": 4133
},
{
"epoch": 0.7346394775423164,
"grad_norm": 0.3450226558279936,
"learning_rate": 6.681782257631524e-06,
"loss": 1.3272,
"step": 4134
},
{
"epoch": 0.7348171842374162,
"grad_norm": 0.34468549784106733,
"learning_rate": 6.67336884170773e-06,
"loss": 1.2803,
"step": 4135
},
{
"epoch": 0.7349948909325159,
"grad_norm": 0.35170449522098957,
"learning_rate": 6.664959665235933e-06,
"loss": 1.3417,
"step": 4136
},
{
"epoch": 0.7351725976276157,
"grad_norm": 0.35005520135578194,
"learning_rate": 6.656554730891243e-06,
"loss": 1.3121,
"step": 4137
},
{
"epoch": 0.7353503043227153,
"grad_norm": 0.3437948200028736,
"learning_rate": 6.648154041347434e-06,
"loss": 1.3178,
"step": 4138
},
{
"epoch": 0.7355280110178151,
"grad_norm": 0.33601008714015856,
"learning_rate": 6.639757599276906e-06,
"loss": 1.2809,
"step": 4139
},
{
"epoch": 0.7357057177129148,
"grad_norm": 0.34753246075457067,
"learning_rate": 6.63136540735074e-06,
"loss": 1.3127,
"step": 4140
},
{
"epoch": 0.7358834244080146,
"grad_norm": 0.34418264092724643,
"learning_rate": 6.62297746823865e-06,
"loss": 1.3079,
"step": 4141
},
{
"epoch": 0.7360611311031143,
"grad_norm": 0.3460007104826968,
"learning_rate": 6.614593784608992e-06,
"loss": 1.3052,
"step": 4142
},
{
"epoch": 0.7362388377982141,
"grad_norm": 0.3481975426007932,
"learning_rate": 6.606214359128773e-06,
"loss": 1.3197,
"step": 4143
},
{
"epoch": 0.7364165444933138,
"grad_norm": 0.3480790838829178,
"learning_rate": 6.597839194463649e-06,
"loss": 1.298,
"step": 4144
},
{
"epoch": 0.7365942511884135,
"grad_norm": 0.3451661639011644,
"learning_rate": 6.5894682932779185e-06,
"loss": 1.2876,
"step": 4145
},
{
"epoch": 0.7367719578835132,
"grad_norm": 0.354601227106467,
"learning_rate": 6.581101658234517e-06,
"loss": 1.3366,
"step": 4146
},
{
"epoch": 0.736949664578613,
"grad_norm": 0.34746547588835225,
"learning_rate": 6.572739291995034e-06,
"loss": 1.3103,
"step": 4147
},
{
"epoch": 0.7371273712737128,
"grad_norm": 0.3434305974687531,
"learning_rate": 6.564381197219691e-06,
"loss": 1.2687,
"step": 4148
},
{
"epoch": 0.7373050779688125,
"grad_norm": 0.3475142540294302,
"learning_rate": 6.5560273765673535e-06,
"loss": 1.3013,
"step": 4149
},
{
"epoch": 0.7374827846639123,
"grad_norm": 0.3473330191284544,
"learning_rate": 6.547677832695538e-06,
"loss": 1.3032,
"step": 4150
},
{
"epoch": 0.7376604913590119,
"grad_norm": 0.33904701635539963,
"learning_rate": 6.539332568260386e-06,
"loss": 1.2812,
"step": 4151
},
{
"epoch": 0.7378381980541117,
"grad_norm": 0.33637753628908146,
"learning_rate": 6.530991585916682e-06,
"loss": 1.2948,
"step": 4152
},
{
"epoch": 0.7380159047492114,
"grad_norm": 0.3573513108238807,
"learning_rate": 6.522654888317852e-06,
"loss": 1.3197,
"step": 4153
},
{
"epoch": 0.7381936114443112,
"grad_norm": 0.3479655408244483,
"learning_rate": 6.5143224781159555e-06,
"loss": 1.3092,
"step": 4154
},
{
"epoch": 0.7383713181394109,
"grad_norm": 0.34681327946701324,
"learning_rate": 6.505994357961687e-06,
"loss": 1.3041,
"step": 4155
},
{
"epoch": 0.7385490248345107,
"grad_norm": 0.33959346865628043,
"learning_rate": 6.497670530504381e-06,
"loss": 1.2912,
"step": 4156
},
{
"epoch": 0.7387267315296103,
"grad_norm": 0.34369201097865865,
"learning_rate": 6.489350998392001e-06,
"loss": 1.3265,
"step": 4157
},
{
"epoch": 0.7389044382247101,
"grad_norm": 0.3329993588818506,
"learning_rate": 6.4810357642711395e-06,
"loss": 1.2937,
"step": 4158
},
{
"epoch": 0.7390821449198098,
"grad_norm": 0.34085429294795494,
"learning_rate": 6.472724830787047e-06,
"loss": 1.2992,
"step": 4159
},
{
"epoch": 0.7392598516149096,
"grad_norm": 0.341246518207478,
"learning_rate": 6.464418200583582e-06,
"loss": 1.3069,
"step": 4160
},
{
"epoch": 0.7394375583100093,
"grad_norm": 0.34629844722005126,
"learning_rate": 6.456115876303228e-06,
"loss": 1.3425,
"step": 4161
},
{
"epoch": 0.7396152650051091,
"grad_norm": 0.33917393956664726,
"learning_rate": 6.44781786058712e-06,
"loss": 1.3002,
"step": 4162
},
{
"epoch": 0.7397929717002089,
"grad_norm": 0.33721662443593003,
"learning_rate": 6.439524156075003e-06,
"loss": 1.301,
"step": 4163
},
{
"epoch": 0.7399706783953085,
"grad_norm": 0.3446203669520131,
"learning_rate": 6.431234765405274e-06,
"loss": 1.3255,
"step": 4164
},
{
"epoch": 0.7401483850904083,
"grad_norm": 0.3427442481444144,
"learning_rate": 6.42294969121494e-06,
"loss": 1.3031,
"step": 4165
},
{
"epoch": 0.740326091785508,
"grad_norm": 0.34081922589884395,
"learning_rate": 6.4146689361396365e-06,
"loss": 1.291,
"step": 4166
},
{
"epoch": 0.7405037984806078,
"grad_norm": 0.3420515423379779,
"learning_rate": 6.406392502813628e-06,
"loss": 1.28,
"step": 4167
},
{
"epoch": 0.7406815051757075,
"grad_norm": 0.3376121942977207,
"learning_rate": 6.398120393869802e-06,
"loss": 1.3008,
"step": 4168
},
{
"epoch": 0.7408592118708073,
"grad_norm": 0.34089501757010254,
"learning_rate": 6.389852611939675e-06,
"loss": 1.3075,
"step": 4169
},
{
"epoch": 0.7410369185659069,
"grad_norm": 0.3516506916576127,
"learning_rate": 6.381589159653383e-06,
"loss": 1.3364,
"step": 4170
},
{
"epoch": 0.7412146252610067,
"grad_norm": 0.34450102632024016,
"learning_rate": 6.373330039639685e-06,
"loss": 1.3042,
"step": 4171
},
{
"epoch": 0.7413923319561064,
"grad_norm": 0.34440654631013323,
"learning_rate": 6.365075254525955e-06,
"loss": 1.3187,
"step": 4172
},
{
"epoch": 0.7415700386512062,
"grad_norm": 0.34649406706813474,
"learning_rate": 6.356824806938209e-06,
"loss": 1.3067,
"step": 4173
},
{
"epoch": 0.7417477453463059,
"grad_norm": 0.357320400073252,
"learning_rate": 6.3485786995010645e-06,
"loss": 1.3414,
"step": 4174
},
{
"epoch": 0.7419254520414057,
"grad_norm": 0.35615568586301494,
"learning_rate": 6.340336934837768e-06,
"loss": 1.3561,
"step": 4175
},
{
"epoch": 0.7421031587365055,
"grad_norm": 0.3422689820889564,
"learning_rate": 6.332099515570169e-06,
"loss": 1.3127,
"step": 4176
},
{
"epoch": 0.7422808654316051,
"grad_norm": 0.36676353524989613,
"learning_rate": 6.323866444318742e-06,
"loss": 1.326,
"step": 4177
},
{
"epoch": 0.7424585721267049,
"grad_norm": 0.3431304277473577,
"learning_rate": 6.315637723702597e-06,
"loss": 1.3586,
"step": 4178
},
{
"epoch": 0.7426362788218046,
"grad_norm": 0.34255853247565354,
"learning_rate": 6.3074133563394405e-06,
"loss": 1.3369,
"step": 4179
},
{
"epoch": 0.7428139855169044,
"grad_norm": 0.34708823341322126,
"learning_rate": 6.299193344845593e-06,
"loss": 1.295,
"step": 4180
},
{
"epoch": 0.7429916922120041,
"grad_norm": 0.34364661551696146,
"learning_rate": 6.290977691835994e-06,
"loss": 1.3593,
"step": 4181
},
{
"epoch": 0.7431693989071039,
"grad_norm": 0.3556159283705207,
"learning_rate": 6.282766399924212e-06,
"loss": 1.3341,
"step": 4182
},
{
"epoch": 0.7433471056022035,
"grad_norm": 0.34672116131426706,
"learning_rate": 6.274559471722395e-06,
"loss": 1.2797,
"step": 4183
},
{
"epoch": 0.7435248122973033,
"grad_norm": 0.3532486345446146,
"learning_rate": 6.266356909841329e-06,
"loss": 1.3549,
"step": 4184
},
{
"epoch": 0.743702518992403,
"grad_norm": 0.33542873312900595,
"learning_rate": 6.258158716890404e-06,
"loss": 1.2674,
"step": 4185
},
{
"epoch": 0.7438802256875028,
"grad_norm": 0.3409228798584469,
"learning_rate": 6.249964895477612e-06,
"loss": 1.2837,
"step": 4186
},
{
"epoch": 0.7440579323826025,
"grad_norm": 0.3471932536703681,
"learning_rate": 6.241775448209573e-06,
"loss": 1.3201,
"step": 4187
},
{
"epoch": 0.7442356390777023,
"grad_norm": 0.3486543948318935,
"learning_rate": 6.233590377691498e-06,
"loss": 1.3407,
"step": 4188
},
{
"epoch": 0.7444133457728019,
"grad_norm": 0.3443679092752075,
"learning_rate": 6.225409686527215e-06,
"loss": 1.328,
"step": 4189
},
{
"epoch": 0.7445910524679017,
"grad_norm": 0.35064025915412356,
"learning_rate": 6.217233377319152e-06,
"loss": 1.3401,
"step": 4190
},
{
"epoch": 0.7447687591630014,
"grad_norm": 0.34185511968528753,
"learning_rate": 6.20906145266835e-06,
"loss": 1.3111,
"step": 4191
},
{
"epoch": 0.7449464658581012,
"grad_norm": 0.3443027773547935,
"learning_rate": 6.200893915174448e-06,
"loss": 1.3489,
"step": 4192
},
{
"epoch": 0.745124172553201,
"grad_norm": 0.350858393275898,
"learning_rate": 6.192730767435697e-06,
"loss": 1.3003,
"step": 4193
},
{
"epoch": 0.7453018792483007,
"grad_norm": 0.34174998556426345,
"learning_rate": 6.184572012048946e-06,
"loss": 1.2992,
"step": 4194
},
{
"epoch": 0.7454795859434005,
"grad_norm": 0.3455927572572645,
"learning_rate": 6.176417651609643e-06,
"loss": 1.3021,
"step": 4195
},
{
"epoch": 0.7456572926385001,
"grad_norm": 0.36480773836167274,
"learning_rate": 6.168267688711853e-06,
"loss": 1.3193,
"step": 4196
},
{
"epoch": 0.7458349993335999,
"grad_norm": 0.3406081595479604,
"learning_rate": 6.160122125948238e-06,
"loss": 1.2968,
"step": 4197
},
{
"epoch": 0.7460127060286996,
"grad_norm": 0.3480640091370051,
"learning_rate": 6.151980965910036e-06,
"loss": 1.2943,
"step": 4198
},
{
"epoch": 0.7461904127237994,
"grad_norm": 0.3445684181633426,
"learning_rate": 6.143844211187115e-06,
"loss": 1.3149,
"step": 4199
},
{
"epoch": 0.7463681194188991,
"grad_norm": 0.34933971342692427,
"learning_rate": 6.135711864367919e-06,
"loss": 1.3249,
"step": 4200
},
{
"epoch": 0.7465458261139989,
"grad_norm": 0.3425332369888758,
"learning_rate": 6.1275839280395155e-06,
"loss": 1.3255,
"step": 4201
},
{
"epoch": 0.7467235328090985,
"grad_norm": 0.34901764254149575,
"learning_rate": 6.119460404787547e-06,
"loss": 1.3686,
"step": 4202
},
{
"epoch": 0.7469012395041983,
"grad_norm": 0.3339388571804678,
"learning_rate": 6.1113412971962605e-06,
"loss": 1.2683,
"step": 4203
},
{
"epoch": 0.747078946199298,
"grad_norm": 0.43094455873264953,
"learning_rate": 6.103226607848494e-06,
"loss": 1.31,
"step": 4204
},
{
"epoch": 0.7472566528943978,
"grad_norm": 0.3465526518101947,
"learning_rate": 6.095116339325684e-06,
"loss": 1.3256,
"step": 4205
},
{
"epoch": 0.7474343595894976,
"grad_norm": 0.3415905248260003,
"learning_rate": 6.087010494207859e-06,
"loss": 1.3259,
"step": 4206
},
{
"epoch": 0.7476120662845973,
"grad_norm": 0.33992081096383225,
"learning_rate": 6.078909075073642e-06,
"loss": 1.2931,
"step": 4207
},
{
"epoch": 0.7477897729796971,
"grad_norm": 0.34883550562117266,
"learning_rate": 6.070812084500246e-06,
"loss": 1.3475,
"step": 4208
},
{
"epoch": 0.7479674796747967,
"grad_norm": 0.3477842862093228,
"learning_rate": 6.0627195250634716e-06,
"loss": 1.3305,
"step": 4209
},
{
"epoch": 0.7481451863698965,
"grad_norm": 0.34897778997251844,
"learning_rate": 6.054631399337723e-06,
"loss": 1.3415,
"step": 4210
},
{
"epoch": 0.7483228930649962,
"grad_norm": 0.34624002009071125,
"learning_rate": 6.04654770989598e-06,
"loss": 1.322,
"step": 4211
},
{
"epoch": 0.748500599760096,
"grad_norm": 0.35346077678049553,
"learning_rate": 6.038468459309818e-06,
"loss": 1.3554,
"step": 4212
},
{
"epoch": 0.7486783064551957,
"grad_norm": 0.3517435746242705,
"learning_rate": 6.030393650149404e-06,
"loss": 1.3241,
"step": 4213
},
{
"epoch": 0.7488560131502955,
"grad_norm": 0.3516304662158467,
"learning_rate": 6.022323284983466e-06,
"loss": 1.297,
"step": 4214
},
{
"epoch": 0.7490337198453951,
"grad_norm": 0.3420430085423513,
"learning_rate": 6.014257366379361e-06,
"loss": 1.3103,
"step": 4215
},
{
"epoch": 0.7492114265404949,
"grad_norm": 0.340190992086526,
"learning_rate": 6.006195896903002e-06,
"loss": 1.3098,
"step": 4216
},
{
"epoch": 0.7493891332355946,
"grad_norm": 0.35183823301870254,
"learning_rate": 5.998138879118891e-06,
"loss": 1.3049,
"step": 4217
},
{
"epoch": 0.7495668399306944,
"grad_norm": 0.3423032145289341,
"learning_rate": 5.990086315590122e-06,
"loss": 1.3017,
"step": 4218
},
{
"epoch": 0.7497445466257942,
"grad_norm": 0.3485555319176679,
"learning_rate": 5.982038208878362e-06,
"loss": 1.3657,
"step": 4219
},
{
"epoch": 0.7499222533208939,
"grad_norm": 0.33612620563063583,
"learning_rate": 5.97399456154387e-06,
"loss": 1.2792,
"step": 4220
},
{
"epoch": 0.7500999600159935,
"grad_norm": 0.3433118871544173,
"learning_rate": 5.965955376145475e-06,
"loss": 1.2796,
"step": 4221
},
{
"epoch": 0.7502776667110933,
"grad_norm": 0.3495395908544088,
"learning_rate": 5.957920655240601e-06,
"loss": 1.2695,
"step": 4222
},
{
"epoch": 0.7504553734061931,
"grad_norm": 0.3442796673646616,
"learning_rate": 5.949890401385232e-06,
"loss": 1.3207,
"step": 4223
},
{
"epoch": 0.7506330801012928,
"grad_norm": 0.3953482913819814,
"learning_rate": 5.941864617133957e-06,
"loss": 1.297,
"step": 4224
},
{
"epoch": 0.7508107867963926,
"grad_norm": 0.37307989463211394,
"learning_rate": 5.933843305039921e-06,
"loss": 1.3382,
"step": 4225
},
{
"epoch": 0.7509884934914923,
"grad_norm": 0.3472294018535373,
"learning_rate": 5.925826467654856e-06,
"loss": 1.3408,
"step": 4226
},
{
"epoch": 0.7511662001865921,
"grad_norm": 0.34734555888522534,
"learning_rate": 5.917814107529069e-06,
"loss": 1.3291,
"step": 4227
},
{
"epoch": 0.7513439068816917,
"grad_norm": 0.34929686428594564,
"learning_rate": 5.909806227211441e-06,
"loss": 1.3049,
"step": 4228
},
{
"epoch": 0.7515216135767915,
"grad_norm": 0.34832618203934246,
"learning_rate": 5.9018028292494325e-06,
"loss": 1.3127,
"step": 4229
},
{
"epoch": 0.7516993202718912,
"grad_norm": 0.3523166163622038,
"learning_rate": 5.893803916189069e-06,
"loss": 1.3276,
"step": 4230
},
{
"epoch": 0.751877026966991,
"grad_norm": 0.33890749981469376,
"learning_rate": 5.885809490574961e-06,
"loss": 1.2785,
"step": 4231
},
{
"epoch": 0.7520547336620907,
"grad_norm": 0.3448647313551246,
"learning_rate": 5.877819554950284e-06,
"loss": 1.2937,
"step": 4232
},
{
"epoch": 0.7522324403571905,
"grad_norm": 0.4311517751458175,
"learning_rate": 5.869834111856778e-06,
"loss": 1.3506,
"step": 4233
},
{
"epoch": 0.7524101470522901,
"grad_norm": 0.3393258014004162,
"learning_rate": 5.8618531638347766e-06,
"loss": 1.2851,
"step": 4234
},
{
"epoch": 0.7525878537473899,
"grad_norm": 0.33824341789537293,
"learning_rate": 5.853876713423172e-06,
"loss": 1.2958,
"step": 4235
},
{
"epoch": 0.7527655604424897,
"grad_norm": 0.33909717383149385,
"learning_rate": 5.845904763159407e-06,
"loss": 1.2857,
"step": 4236
},
{
"epoch": 0.7529432671375894,
"grad_norm": 0.3469472126638024,
"learning_rate": 5.837937315579509e-06,
"loss": 1.3389,
"step": 4237
},
{
"epoch": 0.7531209738326892,
"grad_norm": 0.34513103455169336,
"learning_rate": 5.82997437321809e-06,
"loss": 1.2947,
"step": 4238
},
{
"epoch": 0.7532986805277889,
"grad_norm": 0.3383875977925546,
"learning_rate": 5.8220159386083004e-06,
"loss": 1.3148,
"step": 4239
},
{
"epoch": 0.7534763872228887,
"grad_norm": 0.3340379281689894,
"learning_rate": 5.814062014281869e-06,
"loss": 1.2673,
"step": 4240
},
{
"epoch": 0.7536540939179883,
"grad_norm": 0.3395233596828416,
"learning_rate": 5.8061126027690915e-06,
"loss": 1.3061,
"step": 4241
},
{
"epoch": 0.7538318006130881,
"grad_norm": 0.33462768957862127,
"learning_rate": 5.79816770659882e-06,
"loss": 1.2687,
"step": 4242
},
{
"epoch": 0.7540095073081878,
"grad_norm": 0.3398603451200291,
"learning_rate": 5.790227328298481e-06,
"loss": 1.3055,
"step": 4243
},
{
"epoch": 0.7541872140032876,
"grad_norm": 0.34169257423264643,
"learning_rate": 5.782291470394054e-06,
"loss": 1.3212,
"step": 4244
},
{
"epoch": 0.7543649206983873,
"grad_norm": 0.6704105433310065,
"learning_rate": 5.7743601354100885e-06,
"loss": 1.3208,
"step": 4245
},
{
"epoch": 0.7545426273934871,
"grad_norm": 0.34099853485204484,
"learning_rate": 5.766433325869687e-06,
"loss": 1.3221,
"step": 4246
},
{
"epoch": 0.7547203340885867,
"grad_norm": 0.3408685337579333,
"learning_rate": 5.758511044294515e-06,
"loss": 1.3363,
"step": 4247
},
{
"epoch": 0.7548980407836865,
"grad_norm": 0.34021587918433066,
"learning_rate": 5.750593293204807e-06,
"loss": 1.2972,
"step": 4248
},
{
"epoch": 0.7550757474787863,
"grad_norm": 0.3422513255756088,
"learning_rate": 5.742680075119344e-06,
"loss": 1.2807,
"step": 4249
},
{
"epoch": 0.755253454173886,
"grad_norm": 0.34375059414270154,
"learning_rate": 5.734771392555472e-06,
"loss": 1.312,
"step": 4250
},
{
"epoch": 0.7554311608689858,
"grad_norm": 0.348198755240176,
"learning_rate": 5.726867248029089e-06,
"loss": 1.3464,
"step": 4251
},
{
"epoch": 0.7556088675640855,
"grad_norm": 0.3434719012140069,
"learning_rate": 5.718967644054651e-06,
"loss": 1.3016,
"step": 4252
},
{
"epoch": 0.7557865742591852,
"grad_norm": 0.33794114617290866,
"learning_rate": 5.711072583145174e-06,
"loss": 1.2988,
"step": 4253
},
{
"epoch": 0.7559642809542849,
"grad_norm": 0.343778774717607,
"learning_rate": 5.703182067812225e-06,
"loss": 1.3178,
"step": 4254
},
{
"epoch": 0.7561419876493847,
"grad_norm": 0.34302638724337053,
"learning_rate": 5.6952961005659215e-06,
"loss": 1.3107,
"step": 4255
},
{
"epoch": 0.7563196943444844,
"grad_norm": 0.350200636249512,
"learning_rate": 5.687414683914936e-06,
"loss": 1.3334,
"step": 4256
},
{
"epoch": 0.7564974010395842,
"grad_norm": 0.3456518912574501,
"learning_rate": 5.679537820366512e-06,
"loss": 1.3088,
"step": 4257
},
{
"epoch": 0.7566751077346839,
"grad_norm": 0.3496742128568309,
"learning_rate": 5.671665512426408e-06,
"loss": 1.3314,
"step": 4258
},
{
"epoch": 0.7568528144297837,
"grad_norm": 0.3432504229528284,
"learning_rate": 5.663797762598962e-06,
"loss": 1.3293,
"step": 4259
},
{
"epoch": 0.7570305211248833,
"grad_norm": 0.3520977848034954,
"learning_rate": 5.655934573387052e-06,
"loss": 1.331,
"step": 4260
},
{
"epoch": 0.7572082278199831,
"grad_norm": 0.3481823817144091,
"learning_rate": 5.6480759472921e-06,
"loss": 1.3181,
"step": 4261
},
{
"epoch": 0.7573859345150828,
"grad_norm": 0.3526653715755676,
"learning_rate": 5.640221886814097e-06,
"loss": 1.3019,
"step": 4262
},
{
"epoch": 0.7575636412101826,
"grad_norm": 0.342821450459445,
"learning_rate": 5.632372394451558e-06,
"loss": 1.3331,
"step": 4263
},
{
"epoch": 0.7577413479052824,
"grad_norm": 0.3432940384506741,
"learning_rate": 5.624527472701556e-06,
"loss": 1.2891,
"step": 4264
},
{
"epoch": 0.7579190546003821,
"grad_norm": 0.33382835127012533,
"learning_rate": 5.616687124059708e-06,
"loss": 1.2695,
"step": 4265
},
{
"epoch": 0.7580967612954818,
"grad_norm": 0.33998697907823316,
"learning_rate": 5.608851351020175e-06,
"loss": 1.305,
"step": 4266
},
{
"epoch": 0.7582744679905815,
"grad_norm": 0.3472473383809747,
"learning_rate": 5.601020156075665e-06,
"loss": 1.3238,
"step": 4267
},
{
"epoch": 0.7584521746856813,
"grad_norm": 0.35202694940340284,
"learning_rate": 5.5931935417174295e-06,
"loss": 1.356,
"step": 4268
},
{
"epoch": 0.758629881380781,
"grad_norm": 0.34654132500759016,
"learning_rate": 5.58537151043526e-06,
"loss": 1.3169,
"step": 4269
},
{
"epoch": 0.7588075880758808,
"grad_norm": 0.34212747347541117,
"learning_rate": 5.577554064717488e-06,
"loss": 1.3205,
"step": 4270
},
{
"epoch": 0.7589852947709805,
"grad_norm": 0.3443944916670033,
"learning_rate": 5.5697412070509985e-06,
"loss": 1.3632,
"step": 4271
},
{
"epoch": 0.7591630014660803,
"grad_norm": 0.34709134158454263,
"learning_rate": 5.5619329399212045e-06,
"loss": 1.2888,
"step": 4272
},
{
"epoch": 0.7593407081611799,
"grad_norm": 0.3422973427107207,
"learning_rate": 5.55412926581207e-06,
"loss": 1.308,
"step": 4273
},
{
"epoch": 0.7595184148562797,
"grad_norm": 0.3427578412007267,
"learning_rate": 5.546330187206073e-06,
"loss": 1.2808,
"step": 4274
},
{
"epoch": 0.7596961215513794,
"grad_norm": 0.3410404374492175,
"learning_rate": 5.538535706584254e-06,
"loss": 1.3175,
"step": 4275
},
{
"epoch": 0.7598738282464792,
"grad_norm": 0.34324319213596627,
"learning_rate": 5.530745826426192e-06,
"loss": 1.3083,
"step": 4276
},
{
"epoch": 0.760051534941579,
"grad_norm": 0.38069031313621465,
"learning_rate": 5.522960549209988e-06,
"loss": 1.3401,
"step": 4277
},
{
"epoch": 0.7602292416366787,
"grad_norm": 0.33926660370580486,
"learning_rate": 5.515179877412289e-06,
"loss": 1.2866,
"step": 4278
},
{
"epoch": 0.7604069483317784,
"grad_norm": 0.34765085510557775,
"learning_rate": 5.507403813508267e-06,
"loss": 1.2938,
"step": 4279
},
{
"epoch": 0.7605846550268781,
"grad_norm": 0.3600103948986757,
"learning_rate": 5.499632359971641e-06,
"loss": 1.3109,
"step": 4280
},
{
"epoch": 0.7607623617219779,
"grad_norm": 0.34882656098882087,
"learning_rate": 5.49186551927465e-06,
"loss": 1.3035,
"step": 4281
},
{
"epoch": 0.7609400684170776,
"grad_norm": 0.35342172660051463,
"learning_rate": 5.4841032938880765e-06,
"loss": 1.3559,
"step": 4282
},
{
"epoch": 0.7611177751121774,
"grad_norm": 0.3447296897275748,
"learning_rate": 5.476345686281228e-06,
"loss": 1.3098,
"step": 4283
},
{
"epoch": 0.7612954818072771,
"grad_norm": 0.3458327661078427,
"learning_rate": 5.468592698921942e-06,
"loss": 1.2939,
"step": 4284
},
{
"epoch": 0.7614731885023768,
"grad_norm": 0.347520949901532,
"learning_rate": 5.460844334276598e-06,
"loss": 1.3233,
"step": 4285
},
{
"epoch": 0.7616508951974765,
"grad_norm": 0.35632219158098793,
"learning_rate": 5.453100594810093e-06,
"loss": 1.2937,
"step": 4286
},
{
"epoch": 0.7618286018925763,
"grad_norm": 0.344253513214666,
"learning_rate": 5.445361482985856e-06,
"loss": 1.2928,
"step": 4287
},
{
"epoch": 0.762006308587676,
"grad_norm": 0.34240984489878123,
"learning_rate": 5.437627001265848e-06,
"loss": 1.3164,
"step": 4288
},
{
"epoch": 0.7621840152827758,
"grad_norm": 0.33308431141064654,
"learning_rate": 5.42989715211054e-06,
"loss": 1.2659,
"step": 4289
},
{
"epoch": 0.7623617219778756,
"grad_norm": 0.3461476027768379,
"learning_rate": 5.422171937978953e-06,
"loss": 1.3514,
"step": 4290
},
{
"epoch": 0.7625394286729753,
"grad_norm": 0.34931443081294994,
"learning_rate": 5.414451361328623e-06,
"loss": 1.2861,
"step": 4291
},
{
"epoch": 0.762717135368075,
"grad_norm": 0.3643940669972285,
"learning_rate": 5.4067354246156075e-06,
"loss": 1.2999,
"step": 4292
},
{
"epoch": 0.7628948420631747,
"grad_norm": 0.34878787860820115,
"learning_rate": 5.3990241302944875e-06,
"loss": 1.3661,
"step": 4293
},
{
"epoch": 0.7630725487582745,
"grad_norm": 0.34120808842784583,
"learning_rate": 5.391317480818379e-06,
"loss": 1.3144,
"step": 4294
},
{
"epoch": 0.7632502554533742,
"grad_norm": 0.3469626067880268,
"learning_rate": 5.383615478638917e-06,
"loss": 1.3348,
"step": 4295
},
{
"epoch": 0.763427962148474,
"grad_norm": 0.3521537967587144,
"learning_rate": 5.3759181262062385e-06,
"loss": 1.3293,
"step": 4296
},
{
"epoch": 0.7636056688435737,
"grad_norm": 0.3504318662479647,
"learning_rate": 5.36822542596902e-06,
"loss": 1.303,
"step": 4297
},
{
"epoch": 0.7637833755386734,
"grad_norm": 0.3489323951333383,
"learning_rate": 5.360537380374453e-06,
"loss": 1.341,
"step": 4298
},
{
"epoch": 0.7639610822337731,
"grad_norm": 0.34121182386745313,
"learning_rate": 5.352853991868257e-06,
"loss": 1.285,
"step": 4299
},
{
"epoch": 0.7641387889288729,
"grad_norm": 0.34023501887894053,
"learning_rate": 5.345175262894659e-06,
"loss": 1.2994,
"step": 4300
},
{
"epoch": 0.7643164956239726,
"grad_norm": 0.3584208317716767,
"learning_rate": 5.337501195896406e-06,
"loss": 1.3241,
"step": 4301
},
{
"epoch": 0.7644942023190724,
"grad_norm": 0.3390070491129489,
"learning_rate": 5.329831793314764e-06,
"loss": 1.2872,
"step": 4302
},
{
"epoch": 0.7646719090141721,
"grad_norm": 0.34668332075915337,
"learning_rate": 5.322167057589511e-06,
"loss": 1.3409,
"step": 4303
},
{
"epoch": 0.7648496157092719,
"grad_norm": 0.3511759745357536,
"learning_rate": 5.314506991158948e-06,
"loss": 1.291,
"step": 4304
},
{
"epoch": 0.7650273224043715,
"grad_norm": 0.34534505981854907,
"learning_rate": 5.306851596459886e-06,
"loss": 1.3308,
"step": 4305
},
{
"epoch": 0.7652050290994713,
"grad_norm": 0.3463629116389305,
"learning_rate": 5.299200875927643e-06,
"loss": 1.3269,
"step": 4306
},
{
"epoch": 0.7653827357945711,
"grad_norm": 0.3366302856681844,
"learning_rate": 5.291554831996062e-06,
"loss": 1.275,
"step": 4307
},
{
"epoch": 0.7655604424896708,
"grad_norm": 0.3481468604621738,
"learning_rate": 5.283913467097497e-06,
"loss": 1.2906,
"step": 4308
},
{
"epoch": 0.7657381491847706,
"grad_norm": 0.341492871642623,
"learning_rate": 5.276276783662806e-06,
"loss": 1.2812,
"step": 4309
},
{
"epoch": 0.7659158558798703,
"grad_norm": 0.3317725396908999,
"learning_rate": 5.2686447841213685e-06,
"loss": 1.2427,
"step": 4310
},
{
"epoch": 0.76609356257497,
"grad_norm": 0.34463353209474684,
"learning_rate": 5.261017470901055e-06,
"loss": 1.3174,
"step": 4311
},
{
"epoch": 0.7662712692700697,
"grad_norm": 0.34290802624451444,
"learning_rate": 5.253394846428257e-06,
"loss": 1.3176,
"step": 4312
},
{
"epoch": 0.7664489759651695,
"grad_norm": 0.3405673936328139,
"learning_rate": 5.245776913127887e-06,
"loss": 1.3019,
"step": 4313
},
{
"epoch": 0.7666266826602692,
"grad_norm": 0.3390665902713169,
"learning_rate": 5.238163673423346e-06,
"loss": 1.2876,
"step": 4314
},
{
"epoch": 0.766804389355369,
"grad_norm": 0.33858236150085624,
"learning_rate": 5.23055512973655e-06,
"loss": 1.2952,
"step": 4315
},
{
"epoch": 0.7669820960504687,
"grad_norm": 0.3460333766208056,
"learning_rate": 5.22295128448792e-06,
"loss": 1.3222,
"step": 4316
},
{
"epoch": 0.7671598027455684,
"grad_norm": 0.34787751829228863,
"learning_rate": 5.215352140096379e-06,
"loss": 1.3352,
"step": 4317
},
{
"epoch": 0.7673375094406681,
"grad_norm": 1.108191783409848,
"learning_rate": 5.207757698979361e-06,
"loss": 1.308,
"step": 4318
},
{
"epoch": 0.7675152161357679,
"grad_norm": 0.34814797244124146,
"learning_rate": 5.2001679635528005e-06,
"loss": 1.3222,
"step": 4319
},
{
"epoch": 0.7676929228308677,
"grad_norm": 0.33869997375836786,
"learning_rate": 5.192582936231134e-06,
"loss": 1.2972,
"step": 4320
},
{
"epoch": 0.7678706295259674,
"grad_norm": 0.3426795421673565,
"learning_rate": 5.185002619427295e-06,
"loss": 1.2903,
"step": 4321
},
{
"epoch": 0.7680483362210672,
"grad_norm": 0.35846335504785243,
"learning_rate": 5.1774270155527365e-06,
"loss": 1.3206,
"step": 4322
},
{
"epoch": 0.7682260429161669,
"grad_norm": 0.36385801898537745,
"learning_rate": 5.169856127017396e-06,
"loss": 1.3153,
"step": 4323
},
{
"epoch": 0.7684037496112666,
"grad_norm": 0.3501146354430717,
"learning_rate": 5.162289956229714e-06,
"loss": 1.3422,
"step": 4324
},
{
"epoch": 0.7685814563063663,
"grad_norm": 0.3490321970964417,
"learning_rate": 5.154728505596633e-06,
"loss": 1.3397,
"step": 4325
},
{
"epoch": 0.7687591630014661,
"grad_norm": 0.34585442258606913,
"learning_rate": 5.147171777523594e-06,
"loss": 1.3475,
"step": 4326
},
{
"epoch": 0.7689368696965658,
"grad_norm": 0.35658993085406454,
"learning_rate": 5.13961977441453e-06,
"loss": 1.3167,
"step": 4327
},
{
"epoch": 0.7691145763916656,
"grad_norm": 0.3418028177693534,
"learning_rate": 5.1320724986718804e-06,
"loss": 1.3296,
"step": 4328
},
{
"epoch": 0.7692922830867653,
"grad_norm": 0.34276316429037645,
"learning_rate": 5.124529952696571e-06,
"loss": 1.2702,
"step": 4329
},
{
"epoch": 0.769469989781865,
"grad_norm": 0.3426566794746893,
"learning_rate": 5.1169921388880306e-06,
"loss": 1.3011,
"step": 4330
},
{
"epoch": 0.7696476964769647,
"grad_norm": 0.34411175100579783,
"learning_rate": 5.109459059644171e-06,
"loss": 1.3168,
"step": 4331
},
{
"epoch": 0.7698254031720645,
"grad_norm": 0.3377523484264508,
"learning_rate": 5.101930717361425e-06,
"loss": 1.2706,
"step": 4332
},
{
"epoch": 0.7700031098671642,
"grad_norm": 0.3419754668086324,
"learning_rate": 5.0944071144346855e-06,
"loss": 1.3365,
"step": 4333
},
{
"epoch": 0.770180816562264,
"grad_norm": 0.3352192722556796,
"learning_rate": 5.086888253257354e-06,
"loss": 1.3088,
"step": 4334
},
{
"epoch": 0.7703585232573638,
"grad_norm": 0.3547938781387917,
"learning_rate": 5.0793741362213155e-06,
"loss": 1.338,
"step": 4335
},
{
"epoch": 0.7705362299524635,
"grad_norm": 0.34998063592570755,
"learning_rate": 5.071864765716967e-06,
"loss": 1.3301,
"step": 4336
},
{
"epoch": 0.7707139366475632,
"grad_norm": 0.3437921759232778,
"learning_rate": 5.064360144133171e-06,
"loss": 1.2902,
"step": 4337
},
{
"epoch": 0.7708916433426629,
"grad_norm": 0.3557598958548118,
"learning_rate": 5.056860273857291e-06,
"loss": 1.3313,
"step": 4338
},
{
"epoch": 0.7710693500377627,
"grad_norm": 0.3476892216098733,
"learning_rate": 5.0493651572751765e-06,
"loss": 1.3221,
"step": 4339
},
{
"epoch": 0.7712470567328624,
"grad_norm": 0.33726831094703746,
"learning_rate": 5.041874796771165e-06,
"loss": 1.2591,
"step": 4340
},
{
"epoch": 0.7714247634279622,
"grad_norm": 0.3413760079127435,
"learning_rate": 5.034389194728082e-06,
"loss": 1.2777,
"step": 4341
},
{
"epoch": 0.7716024701230619,
"grad_norm": 0.33664761454343034,
"learning_rate": 5.026908353527236e-06,
"loss": 1.2684,
"step": 4342
},
{
"epoch": 0.7717801768181616,
"grad_norm": 0.34629212311128854,
"learning_rate": 5.019432275548423e-06,
"loss": 1.3459,
"step": 4343
},
{
"epoch": 0.7719578835132613,
"grad_norm": 0.340147967209719,
"learning_rate": 5.011960963169926e-06,
"loss": 1.2962,
"step": 4344
},
{
"epoch": 0.7721355902083611,
"grad_norm": 0.35264049109682183,
"learning_rate": 5.004494418768504e-06,
"loss": 1.2767,
"step": 4345
},
{
"epoch": 0.7723132969034608,
"grad_norm": 0.3598436517740002,
"learning_rate": 4.997032644719417e-06,
"loss": 1.3176,
"step": 4346
},
{
"epoch": 0.7724910035985606,
"grad_norm": 0.3445991369194026,
"learning_rate": 4.98957564339639e-06,
"loss": 1.3218,
"step": 4347
},
{
"epoch": 0.7726687102936604,
"grad_norm": 0.38363854081509435,
"learning_rate": 4.982123417171638e-06,
"loss": 1.3004,
"step": 4348
},
{
"epoch": 0.77284641698876,
"grad_norm": 0.33670178661994565,
"learning_rate": 4.974675968415841e-06,
"loss": 1.27,
"step": 4349
},
{
"epoch": 0.7730241236838598,
"grad_norm": 0.34872860509525555,
"learning_rate": 4.967233299498186e-06,
"loss": 1.3517,
"step": 4350
},
{
"epoch": 0.7732018303789595,
"grad_norm": 0.3456858015287268,
"learning_rate": 4.959795412786324e-06,
"loss": 1.312,
"step": 4351
},
{
"epoch": 0.7733795370740593,
"grad_norm": 0.34677210873998093,
"learning_rate": 4.952362310646384e-06,
"loss": 1.3076,
"step": 4352
},
{
"epoch": 0.773557243769159,
"grad_norm": 0.38321542732835506,
"learning_rate": 4.9449339954429775e-06,
"loss": 1.3102,
"step": 4353
},
{
"epoch": 0.7737349504642588,
"grad_norm": 0.3447687780931403,
"learning_rate": 4.937510469539191e-06,
"loss": 1.2984,
"step": 4354
},
{
"epoch": 0.7739126571593585,
"grad_norm": 0.35225290496011186,
"learning_rate": 4.930091735296585e-06,
"loss": 1.2954,
"step": 4355
},
{
"epoch": 0.7740903638544582,
"grad_norm": 0.34508955873160263,
"learning_rate": 4.922677795075202e-06,
"loss": 1.3011,
"step": 4356
},
{
"epoch": 0.7742680705495579,
"grad_norm": 0.3383548309628456,
"learning_rate": 4.915268651233553e-06,
"loss": 1.2853,
"step": 4357
},
{
"epoch": 0.7744457772446577,
"grad_norm": 0.33819298507395745,
"learning_rate": 4.907864306128627e-06,
"loss": 1.2594,
"step": 4358
},
{
"epoch": 0.7746234839397574,
"grad_norm": 0.3399706120877467,
"learning_rate": 4.90046476211588e-06,
"loss": 1.2826,
"step": 4359
},
{
"epoch": 0.7748011906348572,
"grad_norm": 0.3354998848924768,
"learning_rate": 4.893070021549258e-06,
"loss": 1.2809,
"step": 4360
},
{
"epoch": 0.774978897329957,
"grad_norm": 0.3461581246059562,
"learning_rate": 4.885680086781161e-06,
"loss": 1.3445,
"step": 4361
},
{
"epoch": 0.7751566040250566,
"grad_norm": 0.340401287205169,
"learning_rate": 4.878294960162466e-06,
"loss": 1.3143,
"step": 4362
},
{
"epoch": 0.7753343107201563,
"grad_norm": 0.34014079228182453,
"learning_rate": 4.870914644042526e-06,
"loss": 1.315,
"step": 4363
},
{
"epoch": 0.7755120174152561,
"grad_norm": 0.36258325465788754,
"learning_rate": 4.86353914076914e-06,
"loss": 1.3097,
"step": 4364
},
{
"epoch": 0.7756897241103559,
"grad_norm": 0.3384204140398828,
"learning_rate": 4.856168452688615e-06,
"loss": 1.3187,
"step": 4365
},
{
"epoch": 0.7758674308054556,
"grad_norm": 0.3343293368843455,
"learning_rate": 4.848802582145698e-06,
"loss": 1.2945,
"step": 4366
},
{
"epoch": 0.7760451375005554,
"grad_norm": 0.34471181338003637,
"learning_rate": 4.841441531483608e-06,
"loss": 1.2731,
"step": 4367
},
{
"epoch": 0.7762228441956551,
"grad_norm": 0.3413811888900859,
"learning_rate": 4.834085303044034e-06,
"loss": 1.2743,
"step": 4368
},
{
"epoch": 0.7764005508907548,
"grad_norm": 0.3450537582079213,
"learning_rate": 4.826733899167135e-06,
"loss": 1.345,
"step": 4369
},
{
"epoch": 0.7765782575858545,
"grad_norm": 0.34448622335993734,
"learning_rate": 4.819387322191537e-06,
"loss": 1.3002,
"step": 4370
},
{
"epoch": 0.7767559642809543,
"grad_norm": 0.3406847927066691,
"learning_rate": 4.812045574454311e-06,
"loss": 1.2946,
"step": 4371
},
{
"epoch": 0.776933670976054,
"grad_norm": 0.34045405004480395,
"learning_rate": 4.804708658291008e-06,
"loss": 1.3051,
"step": 4372
},
{
"epoch": 0.7771113776711538,
"grad_norm": 0.3446144080607388,
"learning_rate": 4.797376576035637e-06,
"loss": 1.2935,
"step": 4373
},
{
"epoch": 0.7772890843662535,
"grad_norm": 0.3457872428397815,
"learning_rate": 4.790049330020681e-06,
"loss": 1.3136,
"step": 4374
},
{
"epoch": 0.7774667910613532,
"grad_norm": 0.3413675178714204,
"learning_rate": 4.7827269225770675e-06,
"loss": 1.3252,
"step": 4375
},
{
"epoch": 0.777644497756453,
"grad_norm": 0.3396675436654211,
"learning_rate": 4.775409356034195e-06,
"loss": 1.3126,
"step": 4376
},
{
"epoch": 0.7778222044515527,
"grad_norm": 0.3489733568739252,
"learning_rate": 4.768096632719916e-06,
"loss": 1.3278,
"step": 4377
},
{
"epoch": 0.7779999111466525,
"grad_norm": 0.3406415346858529,
"learning_rate": 4.760788754960548e-06,
"loss": 1.3013,
"step": 4378
},
{
"epoch": 0.7781776178417522,
"grad_norm": 0.34397153121123497,
"learning_rate": 4.753485725080864e-06,
"loss": 1.3149,
"step": 4379
},
{
"epoch": 0.778355324536852,
"grad_norm": 0.34564506276034357,
"learning_rate": 4.746187545404093e-06,
"loss": 1.3433,
"step": 4380
},
{
"epoch": 0.7785330312319516,
"grad_norm": 0.34755484183538554,
"learning_rate": 4.738894218251926e-06,
"loss": 1.306,
"step": 4381
},
{
"epoch": 0.7787107379270514,
"grad_norm": 0.3472970443573358,
"learning_rate": 4.731605745944501e-06,
"loss": 1.3285,
"step": 4382
},
{
"epoch": 0.7788884446221511,
"grad_norm": 0.3391846899992639,
"learning_rate": 4.724322130800427e-06,
"loss": 1.3298,
"step": 4383
},
{
"epoch": 0.7790661513172509,
"grad_norm": 0.3495768680084348,
"learning_rate": 4.717043375136756e-06,
"loss": 1.286,
"step": 4384
},
{
"epoch": 0.7792438580123506,
"grad_norm": 0.3392408001201329,
"learning_rate": 4.709769481269002e-06,
"loss": 1.318,
"step": 4385
},
{
"epoch": 0.7794215647074504,
"grad_norm": 0.347808502693149,
"learning_rate": 4.702500451511116e-06,
"loss": 1.3508,
"step": 4386
},
{
"epoch": 0.7795992714025501,
"grad_norm": 0.33969540549758154,
"learning_rate": 4.695236288175513e-06,
"loss": 1.2733,
"step": 4387
},
{
"epoch": 0.7797769780976498,
"grad_norm": 0.34798957299794614,
"learning_rate": 4.687976993573071e-06,
"loss": 1.3066,
"step": 4388
},
{
"epoch": 0.7799546847927495,
"grad_norm": 0.44232983088154953,
"learning_rate": 4.680722570013103e-06,
"loss": 1.2929,
"step": 4389
},
{
"epoch": 0.7801323914878493,
"grad_norm": 0.341333435233234,
"learning_rate": 4.6734730198033784e-06,
"loss": 1.3317,
"step": 4390
},
{
"epoch": 0.780310098182949,
"grad_norm": 0.3476892301379565,
"learning_rate": 4.6662283452501145e-06,
"loss": 1.3396,
"step": 4391
},
{
"epoch": 0.7804878048780488,
"grad_norm": 0.34239468202027873,
"learning_rate": 4.658988548657977e-06,
"loss": 1.3018,
"step": 4392
},
{
"epoch": 0.7806655115731486,
"grad_norm": 0.35134600925004034,
"learning_rate": 4.651753632330085e-06,
"loss": 1.317,
"step": 4393
},
{
"epoch": 0.7808432182682482,
"grad_norm": 0.33337035984081675,
"learning_rate": 4.644523598567998e-06,
"loss": 1.2562,
"step": 4394
},
{
"epoch": 0.781020924963348,
"grad_norm": 0.3406414933112748,
"learning_rate": 4.637298449671728e-06,
"loss": 1.2982,
"step": 4395
},
{
"epoch": 0.7811986316584477,
"grad_norm": 0.3485421497188995,
"learning_rate": 4.630078187939722e-06,
"loss": 1.3513,
"step": 4396
},
{
"epoch": 0.7813763383535475,
"grad_norm": 0.33709635745056604,
"learning_rate": 4.622862815668896e-06,
"loss": 1.3015,
"step": 4397
},
{
"epoch": 0.7815540450486472,
"grad_norm": 0.3471510224474482,
"learning_rate": 4.615652335154588e-06,
"loss": 1.3469,
"step": 4398
},
{
"epoch": 0.781731751743747,
"grad_norm": 0.3542169798221436,
"learning_rate": 4.608446748690587e-06,
"loss": 1.3472,
"step": 4399
},
{
"epoch": 0.7819094584388467,
"grad_norm": 0.34315817861467257,
"learning_rate": 4.601246058569127e-06,
"loss": 1.3061,
"step": 4400
},
{
"epoch": 0.7820871651339464,
"grad_norm": 0.35212317050244957,
"learning_rate": 4.594050267080883e-06,
"loss": 1.3247,
"step": 4401
},
{
"epoch": 0.7822648718290461,
"grad_norm": 0.39818591947714327,
"learning_rate": 4.586859376514969e-06,
"loss": 1.2913,
"step": 4402
},
{
"epoch": 0.7824425785241459,
"grad_norm": 0.3410470062638385,
"learning_rate": 4.579673389158948e-06,
"loss": 1.2937,
"step": 4403
},
{
"epoch": 0.7826202852192456,
"grad_norm": 0.347135713225281,
"learning_rate": 4.572492307298813e-06,
"loss": 1.3213,
"step": 4404
},
{
"epoch": 0.7827979919143454,
"grad_norm": 0.34801677481824445,
"learning_rate": 4.565316133219e-06,
"loss": 1.3118,
"step": 4405
},
{
"epoch": 0.7829756986094452,
"grad_norm": 0.3577182694441601,
"learning_rate": 4.558144869202383e-06,
"loss": 1.362,
"step": 4406
},
{
"epoch": 0.7831534053045448,
"grad_norm": 0.33077579544814856,
"learning_rate": 4.550978517530287e-06,
"loss": 1.2507,
"step": 4407
},
{
"epoch": 0.7833311119996446,
"grad_norm": 0.3377465018454749,
"learning_rate": 4.54381708048246e-06,
"loss": 1.2942,
"step": 4408
},
{
"epoch": 0.7835088186947443,
"grad_norm": 0.34108780271647604,
"learning_rate": 4.536660560337083e-06,
"loss": 1.2963,
"step": 4409
},
{
"epoch": 0.7836865253898441,
"grad_norm": 0.3449431314993055,
"learning_rate": 4.529508959370774e-06,
"loss": 1.3116,
"step": 4410
},
{
"epoch": 0.7838642320849438,
"grad_norm": 0.3400384612727309,
"learning_rate": 4.5223622798586095e-06,
"loss": 1.3012,
"step": 4411
},
{
"epoch": 0.7840419387800436,
"grad_norm": 0.3482450934356735,
"learning_rate": 4.515220524074073e-06,
"loss": 1.331,
"step": 4412
},
{
"epoch": 0.7842196454751432,
"grad_norm": 0.35067295600395226,
"learning_rate": 4.508083694289092e-06,
"loss": 1.347,
"step": 4413
},
{
"epoch": 0.784397352170243,
"grad_norm": 0.3348172343770928,
"learning_rate": 4.5009517927740266e-06,
"loss": 1.2833,
"step": 4414
},
{
"epoch": 0.7845750588653427,
"grad_norm": 0.3411811623867012,
"learning_rate": 4.493824821797667e-06,
"loss": 1.2952,
"step": 4415
},
{
"epoch": 0.7847527655604425,
"grad_norm": 0.34247189030525443,
"learning_rate": 4.486702783627239e-06,
"loss": 1.295,
"step": 4416
},
{
"epoch": 0.7849304722555422,
"grad_norm": 0.34570333913965967,
"learning_rate": 4.479585680528398e-06,
"loss": 1.3384,
"step": 4417
},
{
"epoch": 0.785108178950642,
"grad_norm": 0.34153121578826307,
"learning_rate": 4.472473514765225e-06,
"loss": 1.308,
"step": 4418
},
{
"epoch": 0.7852858856457418,
"grad_norm": 0.33224257341861474,
"learning_rate": 4.465366288600235e-06,
"loss": 1.2425,
"step": 4419
},
{
"epoch": 0.7854635923408414,
"grad_norm": 0.3451441820205841,
"learning_rate": 4.4582640042943656e-06,
"loss": 1.3337,
"step": 4420
},
{
"epoch": 0.7856412990359412,
"grad_norm": 0.3404681973547898,
"learning_rate": 4.451166664106996e-06,
"loss": 1.3115,
"step": 4421
},
{
"epoch": 0.7858190057310409,
"grad_norm": 0.3467782321745968,
"learning_rate": 4.4440742702959215e-06,
"loss": 1.315,
"step": 4422
},
{
"epoch": 0.7859967124261407,
"grad_norm": 0.34988024460789274,
"learning_rate": 4.436986825117368e-06,
"loss": 1.3316,
"step": 4423
},
{
"epoch": 0.7861744191212404,
"grad_norm": 0.34071226795879034,
"learning_rate": 4.4299043308259714e-06,
"loss": 1.2999,
"step": 4424
},
{
"epoch": 0.7863521258163402,
"grad_norm": 0.3465231848900846,
"learning_rate": 4.422826789674821e-06,
"loss": 1.3215,
"step": 4425
},
{
"epoch": 0.7865298325114398,
"grad_norm": 0.33976811605853496,
"learning_rate": 4.41575420391541e-06,
"loss": 1.2984,
"step": 4426
},
{
"epoch": 0.7867075392065396,
"grad_norm": 0.3407814614065495,
"learning_rate": 4.408686575797663e-06,
"loss": 1.2974,
"step": 4427
},
{
"epoch": 0.7868852459016393,
"grad_norm": 0.3452903983789818,
"learning_rate": 4.401623907569923e-06,
"loss": 1.3217,
"step": 4428
},
{
"epoch": 0.7870629525967391,
"grad_norm": 0.34276191909169645,
"learning_rate": 4.394566201478954e-06,
"loss": 1.3048,
"step": 4429
},
{
"epoch": 0.7872406592918388,
"grad_norm": 0.34200463010175347,
"learning_rate": 4.387513459769959e-06,
"loss": 1.32,
"step": 4430
},
{
"epoch": 0.7874183659869386,
"grad_norm": 0.34278742552746566,
"learning_rate": 4.380465684686535e-06,
"loss": 1.3329,
"step": 4431
},
{
"epoch": 0.7875960726820384,
"grad_norm": 0.34510068790585174,
"learning_rate": 4.373422878470712e-06,
"loss": 1.3031,
"step": 4432
},
{
"epoch": 0.787773779377138,
"grad_norm": 0.34625047931139213,
"learning_rate": 4.366385043362946e-06,
"loss": 1.308,
"step": 4433
},
{
"epoch": 0.7879514860722378,
"grad_norm": 0.3364699380366685,
"learning_rate": 4.359352181602094e-06,
"loss": 1.2698,
"step": 4434
},
{
"epoch": 0.7881291927673375,
"grad_norm": 0.33732533313659285,
"learning_rate": 4.352324295425454e-06,
"loss": 1.3072,
"step": 4435
},
{
"epoch": 0.7883068994624373,
"grad_norm": 0.33906394452680694,
"learning_rate": 4.345301387068723e-06,
"loss": 1.321,
"step": 4436
},
{
"epoch": 0.788484606157537,
"grad_norm": 0.3366030519719437,
"learning_rate": 4.338283458766021e-06,
"loss": 1.2991,
"step": 4437
},
{
"epoch": 0.7886623128526368,
"grad_norm": 0.34509386743464004,
"learning_rate": 4.3312705127498835e-06,
"loss": 1.3068,
"step": 4438
},
{
"epoch": 0.7888400195477364,
"grad_norm": 0.3334507743276015,
"learning_rate": 4.324262551251259e-06,
"loss": 1.2878,
"step": 4439
},
{
"epoch": 0.7890177262428362,
"grad_norm": 0.3363287051261396,
"learning_rate": 4.317259576499513e-06,
"loss": 1.293,
"step": 4440
},
{
"epoch": 0.7891954329379359,
"grad_norm": 0.34416102993957376,
"learning_rate": 4.310261590722422e-06,
"loss": 1.2922,
"step": 4441
},
{
"epoch": 0.7893731396330357,
"grad_norm": 0.3453903768355734,
"learning_rate": 4.3032685961461775e-06,
"loss": 1.3582,
"step": 4442
},
{
"epoch": 0.7895508463281354,
"grad_norm": 0.33572545763901324,
"learning_rate": 4.296280594995377e-06,
"loss": 1.2758,
"step": 4443
},
{
"epoch": 0.7897285530232352,
"grad_norm": 0.34426042178289934,
"learning_rate": 4.289297589493046e-06,
"loss": 1.3135,
"step": 4444
},
{
"epoch": 0.7899062597183348,
"grad_norm": 0.33506855480443287,
"learning_rate": 4.282319581860612e-06,
"loss": 1.261,
"step": 4445
},
{
"epoch": 0.7900839664134346,
"grad_norm": 0.3341724523913581,
"learning_rate": 4.2753465743178975e-06,
"loss": 1.2694,
"step": 4446
},
{
"epoch": 0.7902616731085343,
"grad_norm": 0.3342819187216692,
"learning_rate": 4.268378569083153e-06,
"loss": 1.2731,
"step": 4447
},
{
"epoch": 0.7904393798036341,
"grad_norm": 0.3499600439242514,
"learning_rate": 4.261415568373027e-06,
"loss": 1.3287,
"step": 4448
},
{
"epoch": 0.7906170864987339,
"grad_norm": 0.34800759383263985,
"learning_rate": 4.254457574402591e-06,
"loss": 1.3337,
"step": 4449
},
{
"epoch": 0.7907947931938336,
"grad_norm": 0.34254073302888854,
"learning_rate": 4.247504589385309e-06,
"loss": 1.2933,
"step": 4450
},
{
"epoch": 0.7909724998889334,
"grad_norm": 0.34575458970044337,
"learning_rate": 4.240556615533056e-06,
"loss": 1.312,
"step": 4451
},
{
"epoch": 0.791150206584033,
"grad_norm": 0.3571608917298233,
"learning_rate": 4.233613655056112e-06,
"loss": 1.3609,
"step": 4452
},
{
"epoch": 0.7913279132791328,
"grad_norm": 0.34274111731790213,
"learning_rate": 4.226675710163166e-06,
"loss": 1.2859,
"step": 4453
},
{
"epoch": 0.7915056199742325,
"grad_norm": 0.35185831717416194,
"learning_rate": 4.219742783061307e-06,
"loss": 1.2949,
"step": 4454
},
{
"epoch": 0.7916833266693323,
"grad_norm": 0.34457729596245673,
"learning_rate": 4.21281487595603e-06,
"loss": 1.3242,
"step": 4455
},
{
"epoch": 0.791861033364432,
"grad_norm": 0.33723401241683165,
"learning_rate": 4.205891991051232e-06,
"loss": 1.3068,
"step": 4456
},
{
"epoch": 0.7920387400595318,
"grad_norm": 0.3462430757313181,
"learning_rate": 4.198974130549209e-06,
"loss": 1.3724,
"step": 4457
},
{
"epoch": 0.7922164467546314,
"grad_norm": 0.3541317621901335,
"learning_rate": 4.1920612966506715e-06,
"loss": 1.3107,
"step": 4458
},
{
"epoch": 0.7923941534497312,
"grad_norm": 0.34770970808264046,
"learning_rate": 4.185153491554717e-06,
"loss": 1.3354,
"step": 4459
},
{
"epoch": 0.7925718601448309,
"grad_norm": 0.34075074831307134,
"learning_rate": 4.178250717458847e-06,
"loss": 1.3008,
"step": 4460
},
{
"epoch": 0.7927495668399307,
"grad_norm": 0.3403705862042727,
"learning_rate": 4.171352976558971e-06,
"loss": 1.2594,
"step": 4461
},
{
"epoch": 0.7929272735350305,
"grad_norm": 0.33817581806598873,
"learning_rate": 4.164460271049375e-06,
"loss": 1.3043,
"step": 4462
},
{
"epoch": 0.7931049802301302,
"grad_norm": 0.3422852784681905,
"learning_rate": 4.15757260312277e-06,
"loss": 1.304,
"step": 4463
},
{
"epoch": 0.79328268692523,
"grad_norm": 0.34121681129810966,
"learning_rate": 4.150689974970252e-06,
"loss": 1.2904,
"step": 4464
},
{
"epoch": 0.7934603936203296,
"grad_norm": 0.34418360762274874,
"learning_rate": 4.143812388781314e-06,
"loss": 1.2657,
"step": 4465
},
{
"epoch": 0.7936381003154294,
"grad_norm": 0.35061539755159443,
"learning_rate": 4.136939846743837e-06,
"loss": 1.3096,
"step": 4466
},
{
"epoch": 0.7938158070105291,
"grad_norm": 0.34079886099495893,
"learning_rate": 4.130072351044125e-06,
"loss": 1.338,
"step": 4467
},
{
"epoch": 0.7939935137056289,
"grad_norm": 0.3382027965471477,
"learning_rate": 4.123209903866838e-06,
"loss": 1.2832,
"step": 4468
},
{
"epoch": 0.7941712204007286,
"grad_norm": 0.34381154793702284,
"learning_rate": 4.1163525073950586e-06,
"loss": 1.3232,
"step": 4469
},
{
"epoch": 0.7943489270958284,
"grad_norm": 0.3397366271026268,
"learning_rate": 4.1095001638102514e-06,
"loss": 1.2842,
"step": 4470
},
{
"epoch": 0.794526633790928,
"grad_norm": 0.3385351768658032,
"learning_rate": 4.102652875292272e-06,
"loss": 1.2929,
"step": 4471
},
{
"epoch": 0.7947043404860278,
"grad_norm": 0.35155724138581873,
"learning_rate": 4.09581064401938e-06,
"loss": 1.2842,
"step": 4472
},
{
"epoch": 0.7948820471811275,
"grad_norm": 0.33889754777870085,
"learning_rate": 4.088973472168216e-06,
"loss": 1.2999,
"step": 4473
},
{
"epoch": 0.7950597538762273,
"grad_norm": 0.3411531708523102,
"learning_rate": 4.0821413619138095e-06,
"loss": 1.2767,
"step": 4474
},
{
"epoch": 0.795237460571327,
"grad_norm": 0.34554886152784803,
"learning_rate": 4.075314315429584e-06,
"loss": 1.3075,
"step": 4475
},
{
"epoch": 0.7954151672664268,
"grad_norm": 0.34325467897694173,
"learning_rate": 4.068492334887353e-06,
"loss": 1.3252,
"step": 4476
},
{
"epoch": 0.7955928739615264,
"grad_norm": 0.3488551737418395,
"learning_rate": 4.061675422457314e-06,
"loss": 1.3633,
"step": 4477
},
{
"epoch": 0.7957705806566262,
"grad_norm": 0.3440719428795472,
"learning_rate": 4.054863580308057e-06,
"loss": 1.2993,
"step": 4478
},
{
"epoch": 0.795948287351726,
"grad_norm": 0.3362892646547857,
"learning_rate": 4.048056810606558e-06,
"loss": 1.2741,
"step": 4479
},
{
"epoch": 0.7961259940468257,
"grad_norm": 0.3507754543182098,
"learning_rate": 4.041255115518172e-06,
"loss": 1.3432,
"step": 4480
},
{
"epoch": 0.7963037007419255,
"grad_norm": 0.3402751009480102,
"learning_rate": 4.034458497206657e-06,
"loss": 1.3084,
"step": 4481
},
{
"epoch": 0.7964814074370252,
"grad_norm": 0.34544150781143546,
"learning_rate": 4.02766695783414e-06,
"loss": 1.3183,
"step": 4482
},
{
"epoch": 0.796659114132125,
"grad_norm": 0.3421615192611735,
"learning_rate": 4.020880499561139e-06,
"loss": 1.3057,
"step": 4483
},
{
"epoch": 0.7968368208272246,
"grad_norm": 0.3386349168466458,
"learning_rate": 4.014099124546549e-06,
"loss": 1.3288,
"step": 4484
},
{
"epoch": 0.7970145275223244,
"grad_norm": 0.34157181292428035,
"learning_rate": 4.007322834947651e-06,
"loss": 1.2916,
"step": 4485
},
{
"epoch": 0.7971922342174241,
"grad_norm": 0.3398169525132576,
"learning_rate": 4.00055163292012e-06,
"loss": 1.2999,
"step": 4486
},
{
"epoch": 0.7973699409125239,
"grad_norm": 0.3411339499646046,
"learning_rate": 3.993785520617997e-06,
"loss": 1.3058,
"step": 4487
},
{
"epoch": 0.7975476476076236,
"grad_norm": 0.3491218206997587,
"learning_rate": 3.98702450019371e-06,
"loss": 1.3589,
"step": 4488
},
{
"epoch": 0.7977253543027234,
"grad_norm": 0.33737871231219024,
"learning_rate": 3.9802685737980694e-06,
"loss": 1.323,
"step": 4489
},
{
"epoch": 0.797903060997823,
"grad_norm": 0.34209066065910815,
"learning_rate": 3.973517743580257e-06,
"loss": 1.3063,
"step": 4490
},
{
"epoch": 0.7980807676929228,
"grad_norm": 0.3419612968075547,
"learning_rate": 3.9667720116878425e-06,
"loss": 1.2889,
"step": 4491
},
{
"epoch": 0.7982584743880226,
"grad_norm": 0.34219153464049895,
"learning_rate": 3.9600313802667714e-06,
"loss": 1.3356,
"step": 4492
},
{
"epoch": 0.7984361810831223,
"grad_norm": 0.3357706107214097,
"learning_rate": 3.953295851461363e-06,
"loss": 1.2844,
"step": 4493
},
{
"epoch": 0.7986138877782221,
"grad_norm": 0.3344231589239494,
"learning_rate": 3.946565427414308e-06,
"loss": 1.2908,
"step": 4494
},
{
"epoch": 0.7987915944733218,
"grad_norm": 0.3325520842151161,
"learning_rate": 3.939840110266698e-06,
"loss": 1.2719,
"step": 4495
},
{
"epoch": 0.7989693011684216,
"grad_norm": 0.3394600493438445,
"learning_rate": 3.933119902157972e-06,
"loss": 1.3086,
"step": 4496
},
{
"epoch": 0.7991470078635212,
"grad_norm": 0.3564249077549884,
"learning_rate": 3.926404805225956e-06,
"loss": 1.3586,
"step": 4497
},
{
"epoch": 0.799324714558621,
"grad_norm": 0.3387923052521522,
"learning_rate": 3.919694821606854e-06,
"loss": 1.314,
"step": 4498
},
{
"epoch": 0.7995024212537207,
"grad_norm": 0.3425383638820576,
"learning_rate": 3.912989953435224e-06,
"loss": 1.3173,
"step": 4499
},
{
"epoch": 0.7996801279488205,
"grad_norm": 0.3397593153855081,
"learning_rate": 3.9062902028440235e-06,
"loss": 1.2989,
"step": 4500
},
{
"epoch": 0.7998578346439202,
"grad_norm": 0.3396965303278776,
"learning_rate": 3.899595571964565e-06,
"loss": 1.3172,
"step": 4501
},
{
"epoch": 0.80003554133902,
"grad_norm": 0.3400071596574887,
"learning_rate": 3.892906062926538e-06,
"loss": 1.2742,
"step": 4502
},
{
"epoch": 0.8002132480341196,
"grad_norm": 0.36242878920303223,
"learning_rate": 3.886221677857999e-06,
"loss": 1.292,
"step": 4503
},
{
"epoch": 0.8003909547292194,
"grad_norm": 0.3431723828850276,
"learning_rate": 3.879542418885373e-06,
"loss": 1.3158,
"step": 4504
},
{
"epoch": 0.8005686614243192,
"grad_norm": 0.34135950100918805,
"learning_rate": 3.872868288133474e-06,
"loss": 1.3094,
"step": 4505
},
{
"epoch": 0.8007463681194189,
"grad_norm": 0.338038106962566,
"learning_rate": 3.86619928772545e-06,
"loss": 1.3062,
"step": 4506
},
{
"epoch": 0.8009240748145187,
"grad_norm": 0.34364580887872187,
"learning_rate": 3.8595354197828405e-06,
"loss": 1.3114,
"step": 4507
},
{
"epoch": 0.8011017815096184,
"grad_norm": 0.34994721663057726,
"learning_rate": 3.852876686425546e-06,
"loss": 1.3239,
"step": 4508
},
{
"epoch": 0.8012794882047181,
"grad_norm": 0.34078110684254814,
"learning_rate": 3.8462230897718434e-06,
"loss": 1.3089,
"step": 4509
},
{
"epoch": 0.8014571948998178,
"grad_norm": 0.33957054539911474,
"learning_rate": 3.8395746319383605e-06,
"loss": 1.2975,
"step": 4510
},
{
"epoch": 0.8016349015949176,
"grad_norm": 0.34174544502247606,
"learning_rate": 3.832931315040098e-06,
"loss": 1.3402,
"step": 4511
},
{
"epoch": 0.8018126082900173,
"grad_norm": 0.33839907983642276,
"learning_rate": 3.82629314119042e-06,
"loss": 1.3151,
"step": 4512
},
{
"epoch": 0.8019903149851171,
"grad_norm": 0.347983260310792,
"learning_rate": 3.819660112501053e-06,
"loss": 1.35,
"step": 4513
},
{
"epoch": 0.8021680216802168,
"grad_norm": 0.339757639610529,
"learning_rate": 3.81303223108209e-06,
"loss": 1.2871,
"step": 4514
},
{
"epoch": 0.8023457283753166,
"grad_norm": 0.33612759515870605,
"learning_rate": 3.806409499041983e-06,
"loss": 1.2789,
"step": 4515
},
{
"epoch": 0.8025234350704162,
"grad_norm": 0.3379299441287126,
"learning_rate": 3.7997919184875498e-06,
"loss": 1.3355,
"step": 4516
},
{
"epoch": 0.802701141765516,
"grad_norm": 0.34787015294104406,
"learning_rate": 3.7931794915239638e-06,
"loss": 1.3334,
"step": 4517
},
{
"epoch": 0.8028788484606157,
"grad_norm": 0.3463743049493097,
"learning_rate": 3.7865722202547607e-06,
"loss": 1.3048,
"step": 4518
},
{
"epoch": 0.8030565551557155,
"grad_norm": 0.33939455588400136,
"learning_rate": 3.7799701067818474e-06,
"loss": 1.2747,
"step": 4519
},
{
"epoch": 0.8032342618508153,
"grad_norm": 0.33465398244956673,
"learning_rate": 3.7733731532054797e-06,
"loss": 1.3079,
"step": 4520
},
{
"epoch": 0.803411968545915,
"grad_norm": 0.33896152676227176,
"learning_rate": 3.766781361624261e-06,
"loss": 1.294,
"step": 4521
},
{
"epoch": 0.8035896752410147,
"grad_norm": 0.3389367003171283,
"learning_rate": 3.760194734135165e-06,
"loss": 1.3108,
"step": 4522
},
{
"epoch": 0.8037673819361144,
"grad_norm": 0.3367381566840238,
"learning_rate": 3.753613272833534e-06,
"loss": 1.2939,
"step": 4523
},
{
"epoch": 0.8039450886312142,
"grad_norm": 0.3484964404239962,
"learning_rate": 3.7470369798130477e-06,
"loss": 1.3334,
"step": 4524
},
{
"epoch": 0.8041227953263139,
"grad_norm": 0.3380458104798637,
"learning_rate": 3.740465857165747e-06,
"loss": 1.2708,
"step": 4525
},
{
"epoch": 0.8043005020214137,
"grad_norm": 0.3425439967680597,
"learning_rate": 3.7338999069820346e-06,
"loss": 1.3107,
"step": 4526
},
{
"epoch": 0.8044782087165134,
"grad_norm": 0.3452113081251813,
"learning_rate": 3.7273391313506556e-06,
"loss": 1.3294,
"step": 4527
},
{
"epoch": 0.8046559154116132,
"grad_norm": 0.3332841265322128,
"learning_rate": 3.7207835323587227e-06,
"loss": 1.2785,
"step": 4528
},
{
"epoch": 0.8048336221067128,
"grad_norm": 0.3383154381617818,
"learning_rate": 3.714233112091692e-06,
"loss": 1.2835,
"step": 4529
},
{
"epoch": 0.8050113288018126,
"grad_norm": 0.33854184865209397,
"learning_rate": 3.7076878726333767e-06,
"loss": 1.2817,
"step": 4530
},
{
"epoch": 0.8051890354969123,
"grad_norm": 0.3361391204116564,
"learning_rate": 3.7011478160659397e-06,
"loss": 1.2938,
"step": 4531
},
{
"epoch": 0.8053667421920121,
"grad_norm": 0.34369635889787575,
"learning_rate": 3.694612944469891e-06,
"loss": 1.3023,
"step": 4532
},
{
"epoch": 0.8055444488871119,
"grad_norm": 0.3361868335300112,
"learning_rate": 3.6880832599241047e-06,
"loss": 1.2979,
"step": 4533
},
{
"epoch": 0.8057221555822116,
"grad_norm": 0.34286384140466636,
"learning_rate": 3.6815587645057947e-06,
"loss": 1.3466,
"step": 4534
},
{
"epoch": 0.8058998622773113,
"grad_norm": 0.3419551163492004,
"learning_rate": 3.6750394602905217e-06,
"loss": 1.2988,
"step": 4535
},
{
"epoch": 0.806077568972411,
"grad_norm": 0.3422858897596777,
"learning_rate": 3.668525349352203e-06,
"loss": 1.2892,
"step": 4536
},
{
"epoch": 0.8062552756675108,
"grad_norm": 0.3532789426885383,
"learning_rate": 3.662016433763096e-06,
"loss": 1.3519,
"step": 4537
},
{
"epoch": 0.8064329823626105,
"grad_norm": 0.3407908976966685,
"learning_rate": 3.655512715593812e-06,
"loss": 1.3351,
"step": 4538
},
{
"epoch": 0.8066106890577103,
"grad_norm": 0.34573747879252265,
"learning_rate": 3.649014196913305e-06,
"loss": 1.3052,
"step": 4539
},
{
"epoch": 0.80678839575281,
"grad_norm": 0.3442947547800284,
"learning_rate": 3.6425208797888777e-06,
"loss": 1.3285,
"step": 4540
},
{
"epoch": 0.8069661024479097,
"grad_norm": 0.3405215131348537,
"learning_rate": 3.6360327662861684e-06,
"loss": 1.296,
"step": 4541
},
{
"epoch": 0.8071438091430094,
"grad_norm": 0.3403618044155951,
"learning_rate": 3.6295498584691834e-06,
"loss": 1.2666,
"step": 4542
},
{
"epoch": 0.8073215158381092,
"grad_norm": 0.3472728191179307,
"learning_rate": 3.6230721584002448e-06,
"loss": 1.3211,
"step": 4543
},
{
"epoch": 0.8074992225332089,
"grad_norm": 0.3375339884776245,
"learning_rate": 3.6165996681400326e-06,
"loss": 1.2959,
"step": 4544
},
{
"epoch": 0.8076769292283087,
"grad_norm": 0.34066746621807953,
"learning_rate": 3.6101323897475714e-06,
"loss": 1.3017,
"step": 4545
},
{
"epoch": 0.8078546359234084,
"grad_norm": 0.3488830438569876,
"learning_rate": 3.6036703252802173e-06,
"loss": 1.3178,
"step": 4546
},
{
"epoch": 0.8080323426185082,
"grad_norm": 0.3442532934514744,
"learning_rate": 3.597213476793686e-06,
"loss": 1.3226,
"step": 4547
},
{
"epoch": 0.8082100493136078,
"grad_norm": 0.34282199993155793,
"learning_rate": 3.590761846342015e-06,
"loss": 1.3062,
"step": 4548
},
{
"epoch": 0.8083877560087076,
"grad_norm": 0.3414317130151515,
"learning_rate": 3.5843154359775877e-06,
"loss": 1.3155,
"step": 4549
},
{
"epoch": 0.8085654627038074,
"grad_norm": 0.34083086459415146,
"learning_rate": 3.577874247751134e-06,
"loss": 1.3372,
"step": 4550
},
{
"epoch": 0.8087431693989071,
"grad_norm": 0.3465171309047562,
"learning_rate": 3.571438283711712e-06,
"loss": 1.3184,
"step": 4551
},
{
"epoch": 0.8089208760940069,
"grad_norm": 0.3445285944367498,
"learning_rate": 3.5650075459067267e-06,
"loss": 1.3073,
"step": 4552
},
{
"epoch": 0.8090985827891066,
"grad_norm": 0.34804025483215073,
"learning_rate": 3.5585820363819146e-06,
"loss": 1.3273,
"step": 4553
},
{
"epoch": 0.8092762894842063,
"grad_norm": 0.3531236025576362,
"learning_rate": 3.5521617571813495e-06,
"loss": 1.3155,
"step": 4554
},
{
"epoch": 0.809453996179306,
"grad_norm": 0.33194517499446063,
"learning_rate": 3.545746710347442e-06,
"loss": 1.2761,
"step": 4555
},
{
"epoch": 0.8096317028744058,
"grad_norm": 0.3628265038576701,
"learning_rate": 3.5393368979209464e-06,
"loss": 1.3447,
"step": 4556
},
{
"epoch": 0.8098094095695055,
"grad_norm": 0.3391487875470714,
"learning_rate": 3.5329323219409404e-06,
"loss": 1.3027,
"step": 4557
},
{
"epoch": 0.8099871162646053,
"grad_norm": 0.3404607299822333,
"learning_rate": 3.526532984444846e-06,
"loss": 1.332,
"step": 4558
},
{
"epoch": 0.810164822959705,
"grad_norm": 0.34015238816821153,
"learning_rate": 3.5201388874684004e-06,
"loss": 1.3167,
"step": 4559
},
{
"epoch": 0.8103425296548048,
"grad_norm": 0.3365985878458328,
"learning_rate": 3.5137500330456865e-06,
"loss": 1.3157,
"step": 4560
},
{
"epoch": 0.8105202363499044,
"grad_norm": 0.34050173064470235,
"learning_rate": 3.507366423209131e-06,
"loss": 1.3108,
"step": 4561
},
{
"epoch": 0.8106979430450042,
"grad_norm": 0.34411230647752616,
"learning_rate": 3.5009880599894743e-06,
"loss": 1.3176,
"step": 4562
},
{
"epoch": 0.810875649740104,
"grad_norm": 0.4942224020074281,
"learning_rate": 3.494614945415795e-06,
"loss": 1.328,
"step": 4563
},
{
"epoch": 0.8110533564352037,
"grad_norm": 0.34354631743723696,
"learning_rate": 3.4882470815154923e-06,
"loss": 1.3407,
"step": 4564
},
{
"epoch": 0.8112310631303035,
"grad_norm": 0.3372447362831482,
"learning_rate": 3.4818844703143206e-06,
"loss": 1.2866,
"step": 4565
},
{
"epoch": 0.8114087698254032,
"grad_norm": 0.3373438389413971,
"learning_rate": 3.4755271138363324e-06,
"loss": 1.3226,
"step": 4566
},
{
"epoch": 0.8115864765205029,
"grad_norm": 0.33586383223086913,
"learning_rate": 3.469175014103925e-06,
"loss": 1.3065,
"step": 4567
},
{
"epoch": 0.8117641832156026,
"grad_norm": 0.34101021541107496,
"learning_rate": 3.46282817313782e-06,
"loss": 1.3175,
"step": 4568
},
{
"epoch": 0.8119418899107024,
"grad_norm": 0.33513423683223004,
"learning_rate": 3.4564865929570667e-06,
"loss": 1.2868,
"step": 4569
},
{
"epoch": 0.8121195966058021,
"grad_norm": 0.3438560571977966,
"learning_rate": 3.450150275579045e-06,
"loss": 1.3349,
"step": 4570
},
{
"epoch": 0.8122973033009019,
"grad_norm": 0.33702666520408814,
"learning_rate": 3.4438192230194554e-06,
"loss": 1.2478,
"step": 4571
},
{
"epoch": 0.8124750099960016,
"grad_norm": 0.33055426225309154,
"learning_rate": 3.4374934372923254e-06,
"loss": 1.247,
"step": 4572
},
{
"epoch": 0.8126527166911013,
"grad_norm": 0.36239477915185836,
"learning_rate": 3.431172920410004e-06,
"loss": 1.2734,
"step": 4573
},
{
"epoch": 0.812830423386201,
"grad_norm": 0.34781864186861444,
"learning_rate": 3.4248576743831685e-06,
"loss": 1.3409,
"step": 4574
},
{
"epoch": 0.8130081300813008,
"grad_norm": 0.3455964869635224,
"learning_rate": 3.4185477012208135e-06,
"loss": 1.3339,
"step": 4575
},
{
"epoch": 0.8131858367764006,
"grad_norm": 0.3339922288531123,
"learning_rate": 3.4122430029302667e-06,
"loss": 1.283,
"step": 4576
},
{
"epoch": 0.8133635434715003,
"grad_norm": 0.33611232912750255,
"learning_rate": 3.405943581517164e-06,
"loss": 1.273,
"step": 4577
},
{
"epoch": 0.8135412501666001,
"grad_norm": 0.3487866972986766,
"learning_rate": 3.3996494389854707e-06,
"loss": 1.342,
"step": 4578
},
{
"epoch": 0.8137189568616998,
"grad_norm": 0.33451702658238675,
"learning_rate": 3.393360577337479e-06,
"loss": 1.2904,
"step": 4579
},
{
"epoch": 0.8138966635567995,
"grad_norm": 0.34591246691324984,
"learning_rate": 3.3870769985737928e-06,
"loss": 1.2875,
"step": 4580
},
{
"epoch": 0.8140743702518992,
"grad_norm": 0.34185974436527483,
"learning_rate": 3.380798704693329e-06,
"loss": 1.2979,
"step": 4581
},
{
"epoch": 0.814252076946999,
"grad_norm": 0.41296896547967277,
"learning_rate": 3.374525697693336e-06,
"loss": 1.3097,
"step": 4582
},
{
"epoch": 0.8144297836420987,
"grad_norm": 0.34119324438490345,
"learning_rate": 3.3682579795693715e-06,
"loss": 1.313,
"step": 4583
},
{
"epoch": 0.8146074903371985,
"grad_norm": 0.3401606429251715,
"learning_rate": 3.3619955523153204e-06,
"loss": 1.32,
"step": 4584
},
{
"epoch": 0.8147851970322982,
"grad_norm": 0.3410265500547769,
"learning_rate": 3.3557384179233754e-06,
"loss": 1.2976,
"step": 4585
},
{
"epoch": 0.8149629037273979,
"grad_norm": 0.33748350257274357,
"learning_rate": 3.349486578384049e-06,
"loss": 1.304,
"step": 4586
},
{
"epoch": 0.8151406104224976,
"grad_norm": 0.3483221761957685,
"learning_rate": 3.3432400356861705e-06,
"loss": 1.3477,
"step": 4587
},
{
"epoch": 0.8153183171175974,
"grad_norm": 0.3491456999937691,
"learning_rate": 3.336998791816881e-06,
"loss": 1.3033,
"step": 4588
},
{
"epoch": 0.8154960238126971,
"grad_norm": 0.3380320083150273,
"learning_rate": 3.330762848761637e-06,
"loss": 1.286,
"step": 4589
},
{
"epoch": 0.8156737305077969,
"grad_norm": 0.33016195634195783,
"learning_rate": 3.32453220850421e-06,
"loss": 1.2596,
"step": 4590
},
{
"epoch": 0.8158514372028967,
"grad_norm": 0.33300415192036603,
"learning_rate": 3.3183068730266844e-06,
"loss": 1.315,
"step": 4591
},
{
"epoch": 0.8160291438979964,
"grad_norm": 0.3375529671543962,
"learning_rate": 3.31208684430945e-06,
"loss": 1.2894,
"step": 4592
},
{
"epoch": 0.8162068505930961,
"grad_norm": 0.3288722748329811,
"learning_rate": 3.3058721243312264e-06,
"loss": 1.2914,
"step": 4593
},
{
"epoch": 0.8163845572881958,
"grad_norm": 0.3466382939058354,
"learning_rate": 3.2996627150690255e-06,
"loss": 1.3247,
"step": 4594
},
{
"epoch": 0.8165622639832956,
"grad_norm": 0.3361842282128901,
"learning_rate": 3.29345861849818e-06,
"loss": 1.3271,
"step": 4595
},
{
"epoch": 0.8167399706783953,
"grad_norm": 0.34355999937661275,
"learning_rate": 3.287259836592334e-06,
"loss": 1.3083,
"step": 4596
},
{
"epoch": 0.8169176773734951,
"grad_norm": 0.38467830289832744,
"learning_rate": 3.2810663713234204e-06,
"loss": 1.2755,
"step": 4597
},
{
"epoch": 0.8170953840685948,
"grad_norm": 0.34247733137511005,
"learning_rate": 3.2748782246617127e-06,
"loss": 1.3084,
"step": 4598
},
{
"epoch": 0.8172730907636945,
"grad_norm": 0.339537165375921,
"learning_rate": 3.268695398575772e-06,
"loss": 1.2878,
"step": 4599
},
{
"epoch": 0.8174507974587942,
"grad_norm": 0.34769567663298645,
"learning_rate": 3.262517895032473e-06,
"loss": 1.2987,
"step": 4600
},
{
"epoch": 0.817628504153894,
"grad_norm": 0.3415884398917667,
"learning_rate": 3.2563457159969912e-06,
"loss": 1.3187,
"step": 4601
},
{
"epoch": 0.8178062108489937,
"grad_norm": 0.34295623577720813,
"learning_rate": 3.250178863432818e-06,
"loss": 1.3387,
"step": 4602
},
{
"epoch": 0.8179839175440935,
"grad_norm": 0.33877175299103385,
"learning_rate": 3.2440173393017416e-06,
"loss": 1.2893,
"step": 4603
},
{
"epoch": 0.8181616242391933,
"grad_norm": 0.34110018010334525,
"learning_rate": 3.237861145563861e-06,
"loss": 1.3091,
"step": 4604
},
{
"epoch": 0.8183393309342929,
"grad_norm": 0.3448085512986606,
"learning_rate": 3.231710284177576e-06,
"loss": 1.3506,
"step": 4605
},
{
"epoch": 0.8185170376293927,
"grad_norm": 0.34242916387322136,
"learning_rate": 3.225564757099586e-06,
"loss": 1.346,
"step": 4606
},
{
"epoch": 0.8186947443244924,
"grad_norm": 0.34443566718954827,
"learning_rate": 3.2194245662849076e-06,
"loss": 1.3158,
"step": 4607
},
{
"epoch": 0.8188724510195922,
"grad_norm": 0.345963384397402,
"learning_rate": 3.213289713686849e-06,
"loss": 1.3071,
"step": 4608
},
{
"epoch": 0.8190501577146919,
"grad_norm": 0.4768793786272897,
"learning_rate": 3.2071602012570223e-06,
"loss": 1.3074,
"step": 4609
},
{
"epoch": 0.8192278644097917,
"grad_norm": 0.3397927334754639,
"learning_rate": 3.201036030945337e-06,
"loss": 1.2669,
"step": 4610
},
{
"epoch": 0.8194055711048914,
"grad_norm": 0.3418396470908267,
"learning_rate": 3.19491720470001e-06,
"loss": 1.2954,
"step": 4611
},
{
"epoch": 0.8195832777999911,
"grad_norm": 0.33860986710532726,
"learning_rate": 3.188803724467553e-06,
"loss": 1.3023,
"step": 4612
},
{
"epoch": 0.8197609844950908,
"grad_norm": 0.3306234410120881,
"learning_rate": 3.1826955921927815e-06,
"loss": 1.2924,
"step": 4613
},
{
"epoch": 0.8199386911901906,
"grad_norm": 0.34300028812944067,
"learning_rate": 3.1765928098188037e-06,
"loss": 1.3226,
"step": 4614
},
{
"epoch": 0.8201163978852903,
"grad_norm": 0.3382359058742841,
"learning_rate": 3.170495379287033e-06,
"loss": 1.291,
"step": 4615
},
{
"epoch": 0.8202941045803901,
"grad_norm": 0.3522335230612715,
"learning_rate": 3.1644033025371714e-06,
"loss": 1.3182,
"step": 4616
},
{
"epoch": 0.8204718112754898,
"grad_norm": 0.3365252220281805,
"learning_rate": 3.1583165815072302e-06,
"loss": 1.283,
"step": 4617
},
{
"epoch": 0.8206495179705895,
"grad_norm": 0.34291280615126923,
"learning_rate": 3.1522352181335103e-06,
"loss": 1.3254,
"step": 4618
},
{
"epoch": 0.8208272246656892,
"grad_norm": 0.33492505608371015,
"learning_rate": 3.1461592143506015e-06,
"loss": 1.2838,
"step": 4619
},
{
"epoch": 0.821004931360789,
"grad_norm": 0.3386304505140558,
"learning_rate": 3.1400885720913956e-06,
"loss": 1.3217,
"step": 4620
},
{
"epoch": 0.8211826380558888,
"grad_norm": 0.3328846301703854,
"learning_rate": 3.134023293287076e-06,
"loss": 1.2805,
"step": 4621
},
{
"epoch": 0.8213603447509885,
"grad_norm": 0.33835243237486556,
"learning_rate": 3.1279633798671294e-06,
"loss": 1.328,
"step": 4622
},
{
"epoch": 0.8215380514460883,
"grad_norm": 0.3375540505576467,
"learning_rate": 3.121908833759324e-06,
"loss": 1.3142,
"step": 4623
},
{
"epoch": 0.821715758141188,
"grad_norm": 0.34803028201228386,
"learning_rate": 3.115859656889728e-06,
"loss": 1.3152,
"step": 4624
},
{
"epoch": 0.8218934648362877,
"grad_norm": 0.3414878141596167,
"learning_rate": 3.109815851182694e-06,
"loss": 1.3034,
"step": 4625
},
{
"epoch": 0.8220711715313874,
"grad_norm": 0.36413003877188294,
"learning_rate": 3.1037774185608716e-06,
"loss": 1.3448,
"step": 4626
},
{
"epoch": 0.8222488782264872,
"grad_norm": 0.3540050483406548,
"learning_rate": 3.0977443609452007e-06,
"loss": 1.2758,
"step": 4627
},
{
"epoch": 0.8224265849215869,
"grad_norm": 0.3388110594775293,
"learning_rate": 3.0917166802549103e-06,
"loss": 1.3156,
"step": 4628
},
{
"epoch": 0.8226042916166867,
"grad_norm": 0.3417001296813917,
"learning_rate": 3.085694378407518e-06,
"loss": 1.3305,
"step": 4629
},
{
"epoch": 0.8227819983117864,
"grad_norm": 0.34867505699947743,
"learning_rate": 3.079677457318826e-06,
"loss": 1.3341,
"step": 4630
},
{
"epoch": 0.8229597050068861,
"grad_norm": 0.34832142086671053,
"learning_rate": 3.0736659189029415e-06,
"loss": 1.3065,
"step": 4631
},
{
"epoch": 0.8231374117019858,
"grad_norm": 0.3409181742871657,
"learning_rate": 3.0676597650722417e-06,
"loss": 1.3142,
"step": 4632
},
{
"epoch": 0.8233151183970856,
"grad_norm": 0.34456190191272856,
"learning_rate": 3.0616589977374024e-06,
"loss": 1.3305,
"step": 4633
},
{
"epoch": 0.8234928250921854,
"grad_norm": 0.3574612047926084,
"learning_rate": 3.0556636188073717e-06,
"loss": 1.288,
"step": 4634
},
{
"epoch": 0.8236705317872851,
"grad_norm": 0.3314709985334294,
"learning_rate": 3.0496736301893915e-06,
"loss": 1.2792,
"step": 4635
},
{
"epoch": 0.8238482384823849,
"grad_norm": 0.3292329407718297,
"learning_rate": 3.043689033788999e-06,
"loss": 1.2734,
"step": 4636
},
{
"epoch": 0.8240259451774845,
"grad_norm": 0.33888135117334117,
"learning_rate": 3.0377098315100027e-06,
"loss": 1.305,
"step": 4637
},
{
"epoch": 0.8242036518725843,
"grad_norm": 0.338469154397661,
"learning_rate": 3.0317360252544994e-06,
"loss": 1.316,
"step": 4638
},
{
"epoch": 0.824381358567684,
"grad_norm": 0.3331707163501605,
"learning_rate": 3.0257676169228633e-06,
"loss": 1.2737,
"step": 4639
},
{
"epoch": 0.8245590652627838,
"grad_norm": 0.3524181062995475,
"learning_rate": 3.0198046084137744e-06,
"loss": 1.3497,
"step": 4640
},
{
"epoch": 0.8247367719578835,
"grad_norm": 0.32910498325993154,
"learning_rate": 3.0138470016241616e-06,
"loss": 1.2741,
"step": 4641
},
{
"epoch": 0.8249144786529833,
"grad_norm": 0.34152351157350247,
"learning_rate": 3.0078947984492557e-06,
"loss": 1.3215,
"step": 4642
},
{
"epoch": 0.825092185348083,
"grad_norm": 0.33892690881270693,
"learning_rate": 3.001948000782564e-06,
"loss": 1.2988,
"step": 4643
},
{
"epoch": 0.8252698920431827,
"grad_norm": 0.33989668852355664,
"learning_rate": 2.996006610515874e-06,
"loss": 1.3201,
"step": 4644
},
{
"epoch": 0.8254475987382824,
"grad_norm": 0.3422164927516207,
"learning_rate": 2.990070629539257e-06,
"loss": 1.3308,
"step": 4645
},
{
"epoch": 0.8256253054333822,
"grad_norm": 0.33811387161110074,
"learning_rate": 2.9841400597410607e-06,
"loss": 1.2845,
"step": 4646
},
{
"epoch": 0.825803012128482,
"grad_norm": 0.34564129675006894,
"learning_rate": 2.97821490300791e-06,
"loss": 1.339,
"step": 4647
},
{
"epoch": 0.8259807188235817,
"grad_norm": 0.34169122031608534,
"learning_rate": 2.972295161224705e-06,
"loss": 1.3096,
"step": 4648
},
{
"epoch": 0.8261584255186815,
"grad_norm": 0.34345966889686075,
"learning_rate": 2.9663808362746314e-06,
"loss": 1.289,
"step": 4649
},
{
"epoch": 0.8263361322137811,
"grad_norm": 0.3454421794827229,
"learning_rate": 2.960471930039146e-06,
"loss": 1.3358,
"step": 4650
},
{
"epoch": 0.8265138389088809,
"grad_norm": 0.3412632698631247,
"learning_rate": 2.9545684443979826e-06,
"loss": 1.3216,
"step": 4651
},
{
"epoch": 0.8266915456039806,
"grad_norm": 0.345640947554814,
"learning_rate": 2.9486703812291485e-06,
"loss": 1.3216,
"step": 4652
},
{
"epoch": 0.8268692522990804,
"grad_norm": 0.34243001490868175,
"learning_rate": 2.942777742408929e-06,
"loss": 1.32,
"step": 4653
},
{
"epoch": 0.8270469589941801,
"grad_norm": 0.34256299531669293,
"learning_rate": 2.9368905298118866e-06,
"loss": 1.3225,
"step": 4654
},
{
"epoch": 0.8272246656892799,
"grad_norm": 0.34307399091147317,
"learning_rate": 2.9310087453108592e-06,
"loss": 1.3205,
"step": 4655
},
{
"epoch": 0.8274023723843796,
"grad_norm": 0.3368645290034622,
"learning_rate": 2.9251323907769436e-06,
"loss": 1.2812,
"step": 4656
},
{
"epoch": 0.8275800790794793,
"grad_norm": 0.3438649364737481,
"learning_rate": 2.9192614680795196e-06,
"loss": 1.3257,
"step": 4657
},
{
"epoch": 0.827757785774579,
"grad_norm": 0.5612749418825967,
"learning_rate": 2.9133959790862354e-06,
"loss": 1.3156,
"step": 4658
},
{
"epoch": 0.8279354924696788,
"grad_norm": 0.3382197286658945,
"learning_rate": 2.9075359256630255e-06,
"loss": 1.2676,
"step": 4659
},
{
"epoch": 0.8281131991647785,
"grad_norm": 0.33838559716167177,
"learning_rate": 2.901681309674074e-06,
"loss": 1.2971,
"step": 4660
},
{
"epoch": 0.8282909058598783,
"grad_norm": 0.34402517279983946,
"learning_rate": 2.8958321329818463e-06,
"loss": 1.307,
"step": 4661
},
{
"epoch": 0.8284686125549781,
"grad_norm": 0.3395429499677862,
"learning_rate": 2.889988397447074e-06,
"loss": 1.3098,
"step": 4662
},
{
"epoch": 0.8286463192500777,
"grad_norm": 0.3477995337551216,
"learning_rate": 2.8841501049287624e-06,
"loss": 1.2811,
"step": 4663
},
{
"epoch": 0.8288240259451775,
"grad_norm": 0.340854066440371,
"learning_rate": 2.87831725728418e-06,
"loss": 1.2596,
"step": 4664
},
{
"epoch": 0.8290017326402772,
"grad_norm": 0.335800490335906,
"learning_rate": 2.8724898563688673e-06,
"loss": 1.2793,
"step": 4665
},
{
"epoch": 0.829179439335377,
"grad_norm": 0.34094551755510405,
"learning_rate": 2.866667904036626e-06,
"loss": 1.283,
"step": 4666
},
{
"epoch": 0.8293571460304767,
"grad_norm": 0.3391895562829231,
"learning_rate": 2.8608514021395285e-06,
"loss": 1.2944,
"step": 4667
},
{
"epoch": 0.8295348527255765,
"grad_norm": 0.34895945348518054,
"learning_rate": 2.85504035252792e-06,
"loss": 1.2936,
"step": 4668
},
{
"epoch": 0.8297125594206761,
"grad_norm": 0.3366546328129815,
"learning_rate": 2.849234757050401e-06,
"loss": 1.3045,
"step": 4669
},
{
"epoch": 0.8298902661157759,
"grad_norm": 0.3491893739695462,
"learning_rate": 2.8434346175538395e-06,
"loss": 1.3347,
"step": 4670
},
{
"epoch": 0.8300679728108756,
"grad_norm": 0.33714030149640645,
"learning_rate": 2.837639935883376e-06,
"loss": 1.2798,
"step": 4671
},
{
"epoch": 0.8302456795059754,
"grad_norm": 0.3337513547830566,
"learning_rate": 2.8318507138823913e-06,
"loss": 1.2672,
"step": 4672
},
{
"epoch": 0.8304233862010751,
"grad_norm": 0.3412962916767757,
"learning_rate": 2.826066953392561e-06,
"loss": 1.2993,
"step": 4673
},
{
"epoch": 0.8306010928961749,
"grad_norm": 0.3373514257612226,
"learning_rate": 2.8202886562538023e-06,
"loss": 1.2951,
"step": 4674
},
{
"epoch": 0.8307787995912747,
"grad_norm": 0.33733318584977623,
"learning_rate": 2.8145158243043026e-06,
"loss": 1.2771,
"step": 4675
},
{
"epoch": 0.8309565062863743,
"grad_norm": 0.3422733296025344,
"learning_rate": 2.808748459380506e-06,
"loss": 1.3074,
"step": 4676
},
{
"epoch": 0.831134212981474,
"grad_norm": 0.34029595532523726,
"learning_rate": 2.8029865633171204e-06,
"loss": 1.3131,
"step": 4677
},
{
"epoch": 0.8313119196765738,
"grad_norm": 0.34156993856254037,
"learning_rate": 2.7972301379471133e-06,
"loss": 1.321,
"step": 4678
},
{
"epoch": 0.8314896263716736,
"grad_norm": 0.337965609422608,
"learning_rate": 2.791479185101713e-06,
"loss": 1.2763,
"step": 4679
},
{
"epoch": 0.8316673330667733,
"grad_norm": 0.3404250406402814,
"learning_rate": 2.785733706610403e-06,
"loss": 1.2931,
"step": 4680
},
{
"epoch": 0.8318450397618731,
"grad_norm": 0.34506802226881056,
"learning_rate": 2.779993704300927e-06,
"loss": 1.3276,
"step": 4681
},
{
"epoch": 0.8320227464569727,
"grad_norm": 0.33144448351122996,
"learning_rate": 2.7742591799992923e-06,
"loss": 1.2505,
"step": 4682
},
{
"epoch": 0.8322004531520725,
"grad_norm": 0.3438951116593279,
"learning_rate": 2.768530135529759e-06,
"loss": 1.2789,
"step": 4683
},
{
"epoch": 0.8323781598471722,
"grad_norm": 0.34056170998610175,
"learning_rate": 2.762806572714842e-06,
"loss": 1.2788,
"step": 4684
},
{
"epoch": 0.832555866542272,
"grad_norm": 0.34424591274260313,
"learning_rate": 2.757088493375315e-06,
"loss": 1.3481,
"step": 4685
},
{
"epoch": 0.8327335732373717,
"grad_norm": 0.336954384646606,
"learning_rate": 2.7513758993302043e-06,
"loss": 1.3228,
"step": 4686
},
{
"epoch": 0.8329112799324715,
"grad_norm": 0.3415759560563202,
"learning_rate": 2.745668792396794e-06,
"loss": 1.3135,
"step": 4687
},
{
"epoch": 0.8330889866275712,
"grad_norm": 0.3374626157987603,
"learning_rate": 2.7399671743906255e-06,
"loss": 1.3044,
"step": 4688
},
{
"epoch": 0.8332666933226709,
"grad_norm": 0.3387937273146511,
"learning_rate": 2.7342710471254874e-06,
"loss": 1.3168,
"step": 4689
},
{
"epoch": 0.8334444000177706,
"grad_norm": 0.33601706863269465,
"learning_rate": 2.728580412413424e-06,
"loss": 1.2819,
"step": 4690
},
{
"epoch": 0.8336221067128704,
"grad_norm": 0.34781616188846537,
"learning_rate": 2.722895272064734e-06,
"loss": 1.3233,
"step": 4691
},
{
"epoch": 0.8337998134079702,
"grad_norm": 0.34086747324718725,
"learning_rate": 2.71721562788797e-06,
"loss": 1.3031,
"step": 4692
},
{
"epoch": 0.8339775201030699,
"grad_norm": 0.3420549017464182,
"learning_rate": 2.7115414816899386e-06,
"loss": 1.31,
"step": 4693
},
{
"epoch": 0.8341552267981697,
"grad_norm": 0.330119796327603,
"learning_rate": 2.70587283527568e-06,
"loss": 1.2841,
"step": 4694
},
{
"epoch": 0.8343329334932693,
"grad_norm": 0.3415978336651676,
"learning_rate": 2.7002096904484986e-06,
"loss": 1.3484,
"step": 4695
},
{
"epoch": 0.8345106401883691,
"grad_norm": 0.3413741951736629,
"learning_rate": 2.6945520490099573e-06,
"loss": 1.3234,
"step": 4696
},
{
"epoch": 0.8346883468834688,
"grad_norm": 0.3563264203406242,
"learning_rate": 2.6888999127598524e-06,
"loss": 1.2913,
"step": 4697
},
{
"epoch": 0.8348660535785686,
"grad_norm": 0.3365065350767785,
"learning_rate": 2.6832532834962366e-06,
"loss": 1.264,
"step": 4698
},
{
"epoch": 0.8350437602736683,
"grad_norm": 0.34021676227489284,
"learning_rate": 2.677612163015404e-06,
"loss": 1.3156,
"step": 4699
},
{
"epoch": 0.8352214669687681,
"grad_norm": 0.3306819346894872,
"learning_rate": 2.671976553111908e-06,
"loss": 1.2608,
"step": 4700
},
{
"epoch": 0.8353991736638677,
"grad_norm": 0.33972125871925746,
"learning_rate": 2.666346455578537e-06,
"loss": 1.3252,
"step": 4701
},
{
"epoch": 0.8355768803589675,
"grad_norm": 0.33822924649907526,
"learning_rate": 2.6607218722063312e-06,
"loss": 1.3072,
"step": 4702
},
{
"epoch": 0.8357545870540672,
"grad_norm": 0.33724281852524574,
"learning_rate": 2.6551028047845793e-06,
"loss": 1.2892,
"step": 4703
},
{
"epoch": 0.835932293749167,
"grad_norm": 0.33638155551554,
"learning_rate": 2.64948925510081e-06,
"loss": 1.2946,
"step": 4704
},
{
"epoch": 0.8361100004442668,
"grad_norm": 0.3371822196102848,
"learning_rate": 2.6438812249407964e-06,
"loss": 1.3457,
"step": 4705
},
{
"epoch": 0.8362877071393665,
"grad_norm": 0.34079059519716054,
"learning_rate": 2.6382787160885646e-06,
"loss": 1.3399,
"step": 4706
},
{
"epoch": 0.8364654138344663,
"grad_norm": 0.3348518281646178,
"learning_rate": 2.6326817303263764e-06,
"loss": 1.3218,
"step": 4707
},
{
"epoch": 0.8366431205295659,
"grad_norm": 0.3332606548921229,
"learning_rate": 2.62709026943474e-06,
"loss": 1.3045,
"step": 4708
},
{
"epoch": 0.8368208272246657,
"grad_norm": 0.33401092980621755,
"learning_rate": 2.621504335192393e-06,
"loss": 1.268,
"step": 4709
},
{
"epoch": 0.8369985339197654,
"grad_norm": 0.34395573180368255,
"learning_rate": 2.615923929376338e-06,
"loss": 1.3083,
"step": 4710
},
{
"epoch": 0.8371762406148652,
"grad_norm": 0.34238903648960284,
"learning_rate": 2.6103490537618026e-06,
"loss": 1.3321,
"step": 4711
},
{
"epoch": 0.8373539473099649,
"grad_norm": 0.3380220085091059,
"learning_rate": 2.6047797101222628e-06,
"loss": 1.2997,
"step": 4712
},
{
"epoch": 0.8375316540050647,
"grad_norm": 0.3360030507595287,
"learning_rate": 2.5992159002294283e-06,
"loss": 1.3157,
"step": 4713
},
{
"epoch": 0.8377093607001643,
"grad_norm": 0.3417176698891634,
"learning_rate": 2.5936576258532453e-06,
"loss": 1.3169,
"step": 4714
},
{
"epoch": 0.8378870673952641,
"grad_norm": 0.33159684439520093,
"learning_rate": 2.588104888761924e-06,
"loss": 1.2496,
"step": 4715
},
{
"epoch": 0.8380647740903638,
"grad_norm": 0.3519433739606262,
"learning_rate": 2.5825576907218784e-06,
"loss": 1.3345,
"step": 4716
},
{
"epoch": 0.8382424807854636,
"grad_norm": 0.3434747471068325,
"learning_rate": 2.577016033497781e-06,
"loss": 1.3449,
"step": 4717
},
{
"epoch": 0.8384201874805634,
"grad_norm": 0.3363161152212965,
"learning_rate": 2.5714799188525353e-06,
"loss": 1.3159,
"step": 4718
},
{
"epoch": 0.8385978941756631,
"grad_norm": 0.3292546600310894,
"learning_rate": 2.565949348547283e-06,
"loss": 1.2696,
"step": 4719
},
{
"epoch": 0.8387756008707629,
"grad_norm": 0.3298646227467006,
"learning_rate": 2.5604243243414083e-06,
"loss": 1.2736,
"step": 4720
},
{
"epoch": 0.8389533075658625,
"grad_norm": 0.33398555760834153,
"learning_rate": 2.5549048479925233e-06,
"loss": 1.2963,
"step": 4721
},
{
"epoch": 0.8391310142609623,
"grad_norm": 0.3345201953801166,
"learning_rate": 2.549390921256476e-06,
"loss": 1.2933,
"step": 4722
},
{
"epoch": 0.839308720956062,
"grad_norm": 0.3357660073265736,
"learning_rate": 2.5438825458873483e-06,
"loss": 1.2599,
"step": 4723
},
{
"epoch": 0.8394864276511618,
"grad_norm": 0.33772148061191015,
"learning_rate": 2.538379723637461e-06,
"loss": 1.3027,
"step": 4724
},
{
"epoch": 0.8396641343462615,
"grad_norm": 0.33766249759599215,
"learning_rate": 2.532882456257364e-06,
"loss": 1.2962,
"step": 4725
},
{
"epoch": 0.8398418410413613,
"grad_norm": 0.33036091224143055,
"learning_rate": 2.527390745495841e-06,
"loss": 1.2755,
"step": 4726
},
{
"epoch": 0.8400195477364609,
"grad_norm": 0.3407242895540329,
"learning_rate": 2.521904593099911e-06,
"loss": 1.2935,
"step": 4727
},
{
"epoch": 0.8401972544315607,
"grad_norm": 0.3426604887930078,
"learning_rate": 2.5164240008148143e-06,
"loss": 1.313,
"step": 4728
},
{
"epoch": 0.8403749611266604,
"grad_norm": 0.3434738572694736,
"learning_rate": 2.5109489703840396e-06,
"loss": 1.334,
"step": 4729
},
{
"epoch": 0.8405526678217602,
"grad_norm": 0.3371532509916858,
"learning_rate": 2.505479503549295e-06,
"loss": 1.268,
"step": 4730
},
{
"epoch": 0.84073037451686,
"grad_norm": 0.34538408984646873,
"learning_rate": 2.500015602050525e-06,
"loss": 1.3239,
"step": 4731
},
{
"epoch": 0.8409080812119597,
"grad_norm": 0.33213961299425715,
"learning_rate": 2.494557267625888e-06,
"loss": 1.2607,
"step": 4732
},
{
"epoch": 0.8410857879070593,
"grad_norm": 0.34251002856556206,
"learning_rate": 2.4891045020117852e-06,
"loss": 1.3193,
"step": 4733
},
{
"epoch": 0.8412634946021591,
"grad_norm": 0.34024424279937415,
"learning_rate": 2.4836573069428527e-06,
"loss": 1.3053,
"step": 4734
},
{
"epoch": 0.8414412012972589,
"grad_norm": 0.3380718315734486,
"learning_rate": 2.478215684151939e-06,
"loss": 1.3024,
"step": 4735
},
{
"epoch": 0.8416189079923586,
"grad_norm": 0.3360984960704525,
"learning_rate": 2.472779635370128e-06,
"loss": 1.2774,
"step": 4736
},
{
"epoch": 0.8417966146874584,
"grad_norm": 0.33035967014903805,
"learning_rate": 2.467349162326729e-06,
"loss": 1.2705,
"step": 4737
},
{
"epoch": 0.8419743213825581,
"grad_norm": 0.3390518259819538,
"learning_rate": 2.4619242667492784e-06,
"loss": 1.308,
"step": 4738
},
{
"epoch": 0.8421520280776579,
"grad_norm": 0.3412478745870477,
"learning_rate": 2.4565049503635386e-06,
"loss": 1.3077,
"step": 4739
},
{
"epoch": 0.8423297347727575,
"grad_norm": 0.34158249201963414,
"learning_rate": 2.451091214893493e-06,
"loss": 1.312,
"step": 4740
},
{
"epoch": 0.8425074414678573,
"grad_norm": 0.3386546601656368,
"learning_rate": 2.4456830620613526e-06,
"loss": 1.3101,
"step": 4741
},
{
"epoch": 0.842685148162957,
"grad_norm": 0.3535589762575256,
"learning_rate": 2.4402804935875504e-06,
"loss": 1.3113,
"step": 4742
},
{
"epoch": 0.8428628548580568,
"grad_norm": 0.3679993690584502,
"learning_rate": 2.4348835111907533e-06,
"loss": 1.328,
"step": 4743
},
{
"epoch": 0.8430405615531565,
"grad_norm": 0.3420638069995078,
"learning_rate": 2.429492116587839e-06,
"loss": 1.2884,
"step": 4744
},
{
"epoch": 0.8432182682482563,
"grad_norm": 0.34342008979012084,
"learning_rate": 2.424106311493908e-06,
"loss": 1.3437,
"step": 4745
},
{
"epoch": 0.8433959749433559,
"grad_norm": 0.3464501269505784,
"learning_rate": 2.4187260976222947e-06,
"loss": 1.3412,
"step": 4746
},
{
"epoch": 0.8435736816384557,
"grad_norm": 0.3415953435145099,
"learning_rate": 2.4133514766845333e-06,
"loss": 1.2932,
"step": 4747
},
{
"epoch": 0.8437513883335555,
"grad_norm": 0.34456369606561477,
"learning_rate": 2.4079824503904027e-06,
"loss": 1.3017,
"step": 4748
},
{
"epoch": 0.8439290950286552,
"grad_norm": 0.3438033028862634,
"learning_rate": 2.402619020447885e-06,
"loss": 1.3174,
"step": 4749
},
{
"epoch": 0.844106801723755,
"grad_norm": 0.34204532262449466,
"learning_rate": 2.3972611885631936e-06,
"loss": 1.2849,
"step": 4750
},
{
"epoch": 0.8442845084188547,
"grad_norm": 0.34542893571376343,
"learning_rate": 2.391908956440745e-06,
"loss": 1.3485,
"step": 4751
},
{
"epoch": 0.8444622151139545,
"grad_norm": 0.3391098006148434,
"learning_rate": 2.3865623257831995e-06,
"loss": 1.3142,
"step": 4752
},
{
"epoch": 0.8446399218090541,
"grad_norm": 0.33019986087943226,
"learning_rate": 2.3812212982914163e-06,
"loss": 1.2703,
"step": 4753
},
{
"epoch": 0.8448176285041539,
"grad_norm": 0.3380393045250622,
"learning_rate": 2.37588587566447e-06,
"loss": 1.2989,
"step": 4754
},
{
"epoch": 0.8449953351992536,
"grad_norm": 0.33694295788325546,
"learning_rate": 2.3705560595996648e-06,
"loss": 1.2966,
"step": 4755
},
{
"epoch": 0.8451730418943534,
"grad_norm": 0.33410679602278226,
"learning_rate": 2.3652318517925067e-06,
"loss": 1.3064,
"step": 4756
},
{
"epoch": 0.8453507485894531,
"grad_norm": 0.33282845454234206,
"learning_rate": 2.3599132539367386e-06,
"loss": 1.304,
"step": 4757
},
{
"epoch": 0.8455284552845529,
"grad_norm": 0.3400862131202978,
"learning_rate": 2.354600267724301e-06,
"loss": 1.3247,
"step": 4758
},
{
"epoch": 0.8457061619796525,
"grad_norm": 0.33737457064726223,
"learning_rate": 2.349292894845356e-06,
"loss": 1.2748,
"step": 4759
},
{
"epoch": 0.8458838686747523,
"grad_norm": 0.3426028120244511,
"learning_rate": 2.3439911369882773e-06,
"loss": 1.3437,
"step": 4760
},
{
"epoch": 0.846061575369852,
"grad_norm": 0.32909052912524583,
"learning_rate": 2.3386949958396522e-06,
"loss": 1.2685,
"step": 4761
},
{
"epoch": 0.8462392820649518,
"grad_norm": 0.35419371851317566,
"learning_rate": 2.3334044730842866e-06,
"loss": 1.3419,
"step": 4762
},
{
"epoch": 0.8464169887600516,
"grad_norm": 0.33531069711295775,
"learning_rate": 2.328119570405194e-06,
"loss": 1.3189,
"step": 4763
},
{
"epoch": 0.8465946954551513,
"grad_norm": 0.33720851008889485,
"learning_rate": 2.322840289483599e-06,
"loss": 1.3013,
"step": 4764
},
{
"epoch": 0.846772402150251,
"grad_norm": 0.34118593562895344,
"learning_rate": 2.3175666319989375e-06,
"loss": 1.3047,
"step": 4765
},
{
"epoch": 0.8469501088453507,
"grad_norm": 0.35672027421633973,
"learning_rate": 2.312298599628868e-06,
"loss": 1.2964,
"step": 4766
},
{
"epoch": 0.8471278155404505,
"grad_norm": 0.34104109822521456,
"learning_rate": 2.307036194049248e-06,
"loss": 1.3283,
"step": 4767
},
{
"epoch": 0.8473055222355502,
"grad_norm": 0.3412494121265298,
"learning_rate": 2.301779416934147e-06,
"loss": 1.3112,
"step": 4768
},
{
"epoch": 0.84748322893065,
"grad_norm": 0.37767034413193457,
"learning_rate": 2.2965282699558423e-06,
"loss": 1.3215,
"step": 4769
},
{
"epoch": 0.8476609356257497,
"grad_norm": 0.3359630101135187,
"learning_rate": 2.291282754784816e-06,
"loss": 1.2992,
"step": 4770
},
{
"epoch": 0.8478386423208495,
"grad_norm": 0.34077093554428606,
"learning_rate": 2.2860428730897798e-06,
"loss": 1.2941,
"step": 4771
},
{
"epoch": 0.8480163490159491,
"grad_norm": 0.3402563896919677,
"learning_rate": 2.2808086265376317e-06,
"loss": 1.2974,
"step": 4772
},
{
"epoch": 0.8481940557110489,
"grad_norm": 0.34113439236996573,
"learning_rate": 2.2755800167934816e-06,
"loss": 1.3087,
"step": 4773
},
{
"epoch": 0.8483717624061486,
"grad_norm": 0.34306233618650045,
"learning_rate": 2.2703570455206523e-06,
"loss": 1.3004,
"step": 4774
},
{
"epoch": 0.8485494691012484,
"grad_norm": 0.34265420891507103,
"learning_rate": 2.2651397143806663e-06,
"loss": 1.3049,
"step": 4775
},
{
"epoch": 0.8487271757963482,
"grad_norm": 0.3367660579431857,
"learning_rate": 2.259928025033258e-06,
"loss": 1.3213,
"step": 4776
},
{
"epoch": 0.8489048824914479,
"grad_norm": 0.3375196331542742,
"learning_rate": 2.254721979136363e-06,
"loss": 1.2826,
"step": 4777
},
{
"epoch": 0.8490825891865476,
"grad_norm": 0.33998335236883676,
"learning_rate": 2.2495215783461188e-06,
"loss": 1.3189,
"step": 4778
},
{
"epoch": 0.8492602958816473,
"grad_norm": 0.33648019525206607,
"learning_rate": 2.2443268243168693e-06,
"loss": 1.2992,
"step": 4779
},
{
"epoch": 0.8494380025767471,
"grad_norm": 0.3475136356155178,
"learning_rate": 2.239137718701172e-06,
"loss": 1.3233,
"step": 4780
},
{
"epoch": 0.8496157092718468,
"grad_norm": 0.3354619396849394,
"learning_rate": 2.2339542631497757e-06,
"loss": 1.2929,
"step": 4781
},
{
"epoch": 0.8497934159669466,
"grad_norm": 0.33515159117495874,
"learning_rate": 2.2287764593116323e-06,
"loss": 1.2749,
"step": 4782
},
{
"epoch": 0.8499711226620463,
"grad_norm": 0.3417600262662483,
"learning_rate": 2.2236043088339e-06,
"loss": 1.3066,
"step": 4783
},
{
"epoch": 0.8501488293571461,
"grad_norm": 0.3415360837561473,
"learning_rate": 2.2184378133619377e-06,
"loss": 1.3337,
"step": 4784
},
{
"epoch": 0.8503265360522457,
"grad_norm": 0.3445893661771799,
"learning_rate": 2.2132769745393048e-06,
"loss": 1.3275,
"step": 4785
},
{
"epoch": 0.8505042427473455,
"grad_norm": 0.33684606935225114,
"learning_rate": 2.2081217940077602e-06,
"loss": 1.2717,
"step": 4786
},
{
"epoch": 0.8506819494424452,
"grad_norm": 0.3339777031748798,
"learning_rate": 2.2029722734072645e-06,
"loss": 1.3171,
"step": 4787
},
{
"epoch": 0.850859656137545,
"grad_norm": 0.3353589258763285,
"learning_rate": 2.1978284143759754e-06,
"loss": 1.2806,
"step": 4788
},
{
"epoch": 0.8510373628326448,
"grad_norm": 0.34315756312228973,
"learning_rate": 2.192690218550251e-06,
"loss": 1.317,
"step": 4789
},
{
"epoch": 0.8512150695277445,
"grad_norm": 0.33799286540775597,
"learning_rate": 2.1875576875646567e-06,
"loss": 1.2594,
"step": 4790
},
{
"epoch": 0.8513927762228441,
"grad_norm": 0.3339687631302786,
"learning_rate": 2.182430823051935e-06,
"loss": 1.2744,
"step": 4791
},
{
"epoch": 0.8515704829179439,
"grad_norm": 0.3436477161679449,
"learning_rate": 2.1773096266430427e-06,
"loss": 1.3396,
"step": 4792
},
{
"epoch": 0.8517481896130437,
"grad_norm": 0.33454816046990765,
"learning_rate": 2.1721940999671266e-06,
"loss": 1.2653,
"step": 4793
},
{
"epoch": 0.8519258963081434,
"grad_norm": 0.34287241624493514,
"learning_rate": 2.167084244651536e-06,
"loss": 1.2938,
"step": 4794
},
{
"epoch": 0.8521036030032432,
"grad_norm": 0.3453561821954253,
"learning_rate": 2.1619800623218112e-06,
"loss": 1.3087,
"step": 4795
},
{
"epoch": 0.8522813096983429,
"grad_norm": 0.3427741691714065,
"learning_rate": 2.1568815546016884e-06,
"loss": 1.3035,
"step": 4796
},
{
"epoch": 0.8524590163934426,
"grad_norm": 0.34010220465760393,
"learning_rate": 2.1517887231130973e-06,
"loss": 1.3135,
"step": 4797
},
{
"epoch": 0.8526367230885423,
"grad_norm": 0.5456633108029658,
"learning_rate": 2.146701569476164e-06,
"loss": 1.2597,
"step": 4798
},
{
"epoch": 0.8528144297836421,
"grad_norm": 0.3500356214696735,
"learning_rate": 2.141620095309209e-06,
"loss": 1.2887,
"step": 4799
},
{
"epoch": 0.8529921364787418,
"grad_norm": 0.3394612534660109,
"learning_rate": 2.1365443022287423e-06,
"loss": 1.3161,
"step": 4800
},
{
"epoch": 0.8531698431738416,
"grad_norm": 0.3420763163638529,
"learning_rate": 2.1314741918494698e-06,
"loss": 1.3374,
"step": 4801
},
{
"epoch": 0.8533475498689413,
"grad_norm": 0.3375302380473684,
"learning_rate": 2.1264097657842918e-06,
"loss": 1.3253,
"step": 4802
},
{
"epoch": 0.8535252565640411,
"grad_norm": 0.34001552995968803,
"learning_rate": 2.121351025644289e-06,
"loss": 1.291,
"step": 4803
},
{
"epoch": 0.8537029632591407,
"grad_norm": 0.34416106538826596,
"learning_rate": 2.1162979730387544e-06,
"loss": 1.3186,
"step": 4804
},
{
"epoch": 0.8538806699542405,
"grad_norm": 0.334964332615023,
"learning_rate": 2.1112506095751505e-06,
"loss": 1.2803,
"step": 4805
},
{
"epoch": 0.8540583766493403,
"grad_norm": 0.33751875802122405,
"learning_rate": 2.1062089368591464e-06,
"loss": 1.301,
"step": 4806
},
{
"epoch": 0.85423608334444,
"grad_norm": 0.3385029655562725,
"learning_rate": 2.101172956494577e-06,
"loss": 1.2793,
"step": 4807
},
{
"epoch": 0.8544137900395398,
"grad_norm": 0.3429255380673374,
"learning_rate": 2.0961426700834985e-06,
"loss": 1.2696,
"step": 4808
},
{
"epoch": 0.8545914967346395,
"grad_norm": 0.34039736120447195,
"learning_rate": 2.091118079226133e-06,
"loss": 1.2955,
"step": 4809
},
{
"epoch": 0.8547692034297392,
"grad_norm": 0.33525600786197496,
"learning_rate": 2.0860991855209e-06,
"loss": 1.2952,
"step": 4810
},
{
"epoch": 0.8549469101248389,
"grad_norm": 0.3425355501518802,
"learning_rate": 2.0810859905643997e-06,
"loss": 1.3004,
"step": 4811
},
{
"epoch": 0.8551246168199387,
"grad_norm": 0.33403867968678874,
"learning_rate": 2.076078495951428e-06,
"loss": 1.2827,
"step": 4812
},
{
"epoch": 0.8553023235150384,
"grad_norm": 0.3497281457855066,
"learning_rate": 2.071076703274961e-06,
"loss": 1.308,
"step": 4813
},
{
"epoch": 0.8554800302101382,
"grad_norm": 0.33752738257795023,
"learning_rate": 2.0660806141261624e-06,
"loss": 1.3061,
"step": 4814
},
{
"epoch": 0.8556577369052379,
"grad_norm": 0.33536594934816644,
"learning_rate": 2.0610902300943823e-06,
"loss": 1.2868,
"step": 4815
},
{
"epoch": 0.8558354436003377,
"grad_norm": 0.34220883651645323,
"learning_rate": 2.056105552767158e-06,
"loss": 1.3451,
"step": 4816
},
{
"epoch": 0.8560131502954373,
"grad_norm": 0.3320180013146911,
"learning_rate": 2.051126583730203e-06,
"loss": 1.2748,
"step": 4817
},
{
"epoch": 0.8561908569905371,
"grad_norm": 0.4042297902548018,
"learning_rate": 2.0461533245674283e-06,
"loss": 1.3014,
"step": 4818
},
{
"epoch": 0.8563685636856369,
"grad_norm": 0.33894034668685147,
"learning_rate": 2.041185776860919e-06,
"loss": 1.2997,
"step": 4819
},
{
"epoch": 0.8565462703807366,
"grad_norm": 0.344749182536908,
"learning_rate": 2.036223942190945e-06,
"loss": 1.3222,
"step": 4820
},
{
"epoch": 0.8567239770758364,
"grad_norm": 0.33693717627928654,
"learning_rate": 2.0312678221359605e-06,
"loss": 1.2792,
"step": 4821
},
{
"epoch": 0.8569016837709361,
"grad_norm": 0.3401783891043378,
"learning_rate": 2.0263174182725962e-06,
"loss": 1.3499,
"step": 4822
},
{
"epoch": 0.8570793904660358,
"grad_norm": 0.33234095385501333,
"learning_rate": 2.0213727321756725e-06,
"loss": 1.2996,
"step": 4823
},
{
"epoch": 0.8572570971611355,
"grad_norm": 0.37942928797824665,
"learning_rate": 2.0164337654181864e-06,
"loss": 1.2991,
"step": 4824
},
{
"epoch": 0.8574348038562353,
"grad_norm": 0.3310699743334873,
"learning_rate": 2.0115005195713144e-06,
"loss": 1.3108,
"step": 4825
},
{
"epoch": 0.857612510551335,
"grad_norm": 0.3325932978631719,
"learning_rate": 2.0065729962044143e-06,
"loss": 1.2602,
"step": 4826
},
{
"epoch": 0.8577902172464348,
"grad_norm": 0.33911618791773457,
"learning_rate": 2.001651196885028e-06,
"loss": 1.3152,
"step": 4827
},
{
"epoch": 0.8579679239415345,
"grad_norm": 0.3422295713900585,
"learning_rate": 1.9967351231788746e-06,
"loss": 1.3127,
"step": 4828
},
{
"epoch": 0.8581456306366342,
"grad_norm": 0.3348063742404844,
"learning_rate": 1.99182477664984e-06,
"loss": 1.3132,
"step": 4829
},
{
"epoch": 0.8583233373317339,
"grad_norm": 0.3347637890307705,
"learning_rate": 1.986920158860004e-06,
"loss": 1.2959,
"step": 4830
},
{
"epoch": 0.8585010440268337,
"grad_norm": 0.3445718065935647,
"learning_rate": 1.9820212713696143e-06,
"loss": 1.3294,
"step": 4831
},
{
"epoch": 0.8586787507219334,
"grad_norm": 0.33634939967747285,
"learning_rate": 1.9771281157371034e-06,
"loss": 1.2651,
"step": 4832
},
{
"epoch": 0.8588564574170332,
"grad_norm": 0.35006815792153984,
"learning_rate": 1.972240693519074e-06,
"loss": 1.323,
"step": 4833
},
{
"epoch": 0.859034164112133,
"grad_norm": 0.3321274966802367,
"learning_rate": 1.9673590062703087e-06,
"loss": 1.2873,
"step": 4834
},
{
"epoch": 0.8592118708072327,
"grad_norm": 0.3405124415089642,
"learning_rate": 1.9624830555437603e-06,
"loss": 1.2786,
"step": 4835
},
{
"epoch": 0.8593895775023324,
"grad_norm": 0.34526570512905047,
"learning_rate": 1.957612842890564e-06,
"loss": 1.3381,
"step": 4836
},
{
"epoch": 0.8595672841974321,
"grad_norm": 0.3334271788426179,
"learning_rate": 1.9527483698600247e-06,
"loss": 1.2868,
"step": 4837
},
{
"epoch": 0.8597449908925319,
"grad_norm": 0.33039677467195233,
"learning_rate": 1.9478896379996226e-06,
"loss": 1.2988,
"step": 4838
},
{
"epoch": 0.8599226975876316,
"grad_norm": 0.3279848194948554,
"learning_rate": 1.9430366488550122e-06,
"loss": 1.2956,
"step": 4839
},
{
"epoch": 0.8601004042827314,
"grad_norm": 0.3402887086436454,
"learning_rate": 1.9381894039700168e-06,
"loss": 1.3346,
"step": 4840
},
{
"epoch": 0.8602781109778311,
"grad_norm": 0.33086574185684886,
"learning_rate": 1.9333479048866422e-06,
"loss": 1.2699,
"step": 4841
},
{
"epoch": 0.8604558176729308,
"grad_norm": 0.33566669369912555,
"learning_rate": 1.928512153145059e-06,
"loss": 1.3035,
"step": 4842
},
{
"epoch": 0.8606335243680305,
"grad_norm": 0.3377015891169754,
"learning_rate": 1.923682150283612e-06,
"loss": 1.314,
"step": 4843
},
{
"epoch": 0.8608112310631303,
"grad_norm": 0.34812085784556873,
"learning_rate": 1.918857897838811e-06,
"loss": 1.3292,
"step": 4844
},
{
"epoch": 0.86098893775823,
"grad_norm": 0.33651091474776684,
"learning_rate": 1.9140393973453373e-06,
"loss": 1.3029,
"step": 4845
},
{
"epoch": 0.8611666444533298,
"grad_norm": 0.3425981564103593,
"learning_rate": 1.90922665033606e-06,
"loss": 1.266,
"step": 4846
},
{
"epoch": 0.8613443511484296,
"grad_norm": 0.3434734711115433,
"learning_rate": 1.904419658341996e-06,
"loss": 1.2942,
"step": 4847
},
{
"epoch": 0.8615220578435293,
"grad_norm": 0.33631209400663464,
"learning_rate": 1.899618422892342e-06,
"loss": 1.2857,
"step": 4848
},
{
"epoch": 0.861699764538629,
"grad_norm": 0.33827664959146,
"learning_rate": 1.8948229455144562e-06,
"loss": 1.3342,
"step": 4849
},
{
"epoch": 0.8618774712337287,
"grad_norm": 0.33462815524576656,
"learning_rate": 1.890033227733883e-06,
"loss": 1.2882,
"step": 4850
},
{
"epoch": 0.8620551779288285,
"grad_norm": 0.33616705714191747,
"learning_rate": 1.8852492710743075e-06,
"loss": 1.2783,
"step": 4851
},
{
"epoch": 0.8622328846239282,
"grad_norm": 0.3358941173679607,
"learning_rate": 1.880471077057604e-06,
"loss": 1.3171,
"step": 4852
},
{
"epoch": 0.862410591319028,
"grad_norm": 0.3304086928862392,
"learning_rate": 1.875698647203803e-06,
"loss": 1.2971,
"step": 4853
},
{
"epoch": 0.8625882980141277,
"grad_norm": 0.3416630142390152,
"learning_rate": 1.8709319830311035e-06,
"loss": 1.2913,
"step": 4854
},
{
"epoch": 0.8627660047092274,
"grad_norm": 0.34363217595839474,
"learning_rate": 1.8661710860558747e-06,
"loss": 1.3365,
"step": 4855
},
{
"epoch": 0.8629437114043271,
"grad_norm": 0.3392433319360242,
"learning_rate": 1.861415957792645e-06,
"loss": 1.3264,
"step": 4856
},
{
"epoch": 0.8631214180994269,
"grad_norm": 0.33542499760864575,
"learning_rate": 1.8566665997541111e-06,
"loss": 1.3279,
"step": 4857
},
{
"epoch": 0.8632991247945266,
"grad_norm": 0.34187797735349285,
"learning_rate": 1.8519230134511312e-06,
"loss": 1.3559,
"step": 4858
},
{
"epoch": 0.8634768314896264,
"grad_norm": 0.34078377939257787,
"learning_rate": 1.8471852003927314e-06,
"loss": 1.3163,
"step": 4859
},
{
"epoch": 0.8636545381847262,
"grad_norm": 0.33441803823065597,
"learning_rate": 1.8424531620860997e-06,
"loss": 1.2737,
"step": 4860
},
{
"epoch": 0.8638322448798258,
"grad_norm": 0.33804253957139746,
"learning_rate": 1.837726900036585e-06,
"loss": 1.309,
"step": 4861
},
{
"epoch": 0.8640099515749255,
"grad_norm": 0.33265442706137865,
"learning_rate": 1.833006415747698e-06,
"loss": 1.2903,
"step": 4862
},
{
"epoch": 0.8641876582700253,
"grad_norm": 0.34575962804236243,
"learning_rate": 1.828291710721115e-06,
"loss": 1.3213,
"step": 4863
},
{
"epoch": 0.8643653649651251,
"grad_norm": 0.3326259466773653,
"learning_rate": 1.8235827864566747e-06,
"loss": 1.2917,
"step": 4864
},
{
"epoch": 0.8645430716602248,
"grad_norm": 0.3408685659988551,
"learning_rate": 1.8188796444523782e-06,
"loss": 1.3086,
"step": 4865
},
{
"epoch": 0.8647207783553246,
"grad_norm": 0.3370920711436664,
"learning_rate": 1.8141822862043734e-06,
"loss": 1.2912,
"step": 4866
},
{
"epoch": 0.8648984850504243,
"grad_norm": 0.3382123515175436,
"learning_rate": 1.8094907132069827e-06,
"loss": 1.3038,
"step": 4867
},
{
"epoch": 0.865076191745524,
"grad_norm": 0.3353988870834331,
"learning_rate": 1.8048049269526812e-06,
"loss": 1.3248,
"step": 4868
},
{
"epoch": 0.8652538984406237,
"grad_norm": 0.3339296393260738,
"learning_rate": 1.800124928932112e-06,
"loss": 1.2687,
"step": 4869
},
{
"epoch": 0.8654316051357235,
"grad_norm": 0.34098775242557317,
"learning_rate": 1.7954507206340666e-06,
"loss": 1.3094,
"step": 4870
},
{
"epoch": 0.8656093118308232,
"grad_norm": 0.34312210157074124,
"learning_rate": 1.7907823035454974e-06,
"loss": 1.3386,
"step": 4871
},
{
"epoch": 0.865787018525923,
"grad_norm": 0.34366032319454254,
"learning_rate": 1.786119679151519e-06,
"loss": 1.3182,
"step": 4872
},
{
"epoch": 0.8659647252210227,
"grad_norm": 0.33742547814850427,
"learning_rate": 1.781462848935398e-06,
"loss": 1.3409,
"step": 4873
},
{
"epoch": 0.8661424319161224,
"grad_norm": 0.3326391722247485,
"learning_rate": 1.7768118143785606e-06,
"loss": 1.3111,
"step": 4874
},
{
"epoch": 0.8663201386112221,
"grad_norm": 0.34084154057601485,
"learning_rate": 1.7721665769605856e-06,
"loss": 1.2866,
"step": 4875
},
{
"epoch": 0.8664978453063219,
"grad_norm": 0.3417574659612621,
"learning_rate": 1.767527138159213e-06,
"loss": 1.3099,
"step": 4876
},
{
"epoch": 0.8666755520014217,
"grad_norm": 0.33847827672015146,
"learning_rate": 1.7628934994503356e-06,
"loss": 1.2691,
"step": 4877
},
{
"epoch": 0.8668532586965214,
"grad_norm": 0.34006243998718516,
"learning_rate": 1.7582656623079963e-06,
"loss": 1.3104,
"step": 4878
},
{
"epoch": 0.8670309653916212,
"grad_norm": 0.33796929892083105,
"learning_rate": 1.7536436282044023e-06,
"loss": 1.3075,
"step": 4879
},
{
"epoch": 0.8672086720867209,
"grad_norm": 0.3423844898610636,
"learning_rate": 1.7490273986099105e-06,
"loss": 1.3296,
"step": 4880
},
{
"epoch": 0.8673863787818206,
"grad_norm": 0.3335840165989184,
"learning_rate": 1.7444169749930328e-06,
"loss": 1.2781,
"step": 4881
},
{
"epoch": 0.8675640854769203,
"grad_norm": 0.33553098060952186,
"learning_rate": 1.7398123588204185e-06,
"loss": 1.3047,
"step": 4882
},
{
"epoch": 0.8677417921720201,
"grad_norm": 0.3361179438877957,
"learning_rate": 1.7352135515568935e-06,
"loss": 1.2868,
"step": 4883
},
{
"epoch": 0.8679194988671198,
"grad_norm": 0.34016110420007484,
"learning_rate": 1.7306205546654253e-06,
"loss": 1.297,
"step": 4884
},
{
"epoch": 0.8680972055622196,
"grad_norm": 0.33771429927545543,
"learning_rate": 1.7260333696071275e-06,
"loss": 1.3067,
"step": 4885
},
{
"epoch": 0.8682749122573193,
"grad_norm": 0.3353136452324088,
"learning_rate": 1.7214519978412725e-06,
"loss": 1.3002,
"step": 4886
},
{
"epoch": 0.868452618952419,
"grad_norm": 0.3272037908154217,
"learning_rate": 1.716876440825277e-06,
"loss": 1.274,
"step": 4887
},
{
"epoch": 0.8686303256475187,
"grad_norm": 0.342625148387587,
"learning_rate": 1.7123067000147232e-06,
"loss": 1.3245,
"step": 4888
},
{
"epoch": 0.8688080323426185,
"grad_norm": 0.33265258566191896,
"learning_rate": 1.7077427768633192e-06,
"loss": 1.2676,
"step": 4889
},
{
"epoch": 0.8689857390377183,
"grad_norm": 0.3392858757998415,
"learning_rate": 1.7031846728229395e-06,
"loss": 1.2868,
"step": 4890
},
{
"epoch": 0.869163445732818,
"grad_norm": 0.33619307010574323,
"learning_rate": 1.6986323893436019e-06,
"loss": 1.264,
"step": 4891
},
{
"epoch": 0.8693411524279178,
"grad_norm": 0.3916777806999602,
"learning_rate": 1.6940859278734723e-06,
"loss": 1.2992,
"step": 4892
},
{
"epoch": 0.8695188591230174,
"grad_norm": 0.3424943343008366,
"learning_rate": 1.6895452898588693e-06,
"loss": 1.3162,
"step": 4893
},
{
"epoch": 0.8696965658181172,
"grad_norm": 0.3324573518175218,
"learning_rate": 1.6850104767442532e-06,
"loss": 1.2936,
"step": 4894
},
{
"epoch": 0.8698742725132169,
"grad_norm": 0.3407538809929013,
"learning_rate": 1.6804814899722343e-06,
"loss": 1.3126,
"step": 4895
},
{
"epoch": 0.8700519792083167,
"grad_norm": 0.33564105217724716,
"learning_rate": 1.6759583309835647e-06,
"loss": 1.2783,
"step": 4896
},
{
"epoch": 0.8702296859034164,
"grad_norm": 0.33573758695838474,
"learning_rate": 1.671441001217151e-06,
"loss": 1.3065,
"step": 4897
},
{
"epoch": 0.8704073925985162,
"grad_norm": 0.3433787629109456,
"learning_rate": 1.6669295021100395e-06,
"loss": 1.3241,
"step": 4898
},
{
"epoch": 0.8705850992936159,
"grad_norm": 0.3398589267959003,
"learning_rate": 1.662423835097422e-06,
"loss": 1.2849,
"step": 4899
},
{
"epoch": 0.8707628059887156,
"grad_norm": 0.33825016623930754,
"learning_rate": 1.6579240016126362e-06,
"loss": 1.3462,
"step": 4900
},
{
"epoch": 0.8709405126838153,
"grad_norm": 0.3370658740053003,
"learning_rate": 1.6534300030871597e-06,
"loss": 1.3099,
"step": 4901
},
{
"epoch": 0.8711182193789151,
"grad_norm": 0.34037286658258836,
"learning_rate": 1.6489418409506242e-06,
"loss": 1.3371,
"step": 4902
},
{
"epoch": 0.8712959260740148,
"grad_norm": 0.3465611524412698,
"learning_rate": 1.644459516630803e-06,
"loss": 1.3047,
"step": 4903
},
{
"epoch": 0.8714736327691146,
"grad_norm": 0.3407281920628262,
"learning_rate": 1.6399830315535936e-06,
"loss": 1.2902,
"step": 4904
},
{
"epoch": 0.8716513394642144,
"grad_norm": 0.3322288539093778,
"learning_rate": 1.635512387143061e-06,
"loss": 1.2771,
"step": 4905
},
{
"epoch": 0.871829046159314,
"grad_norm": 0.3392050587379412,
"learning_rate": 1.6310475848213924e-06,
"loss": 1.327,
"step": 4906
},
{
"epoch": 0.8720067528544138,
"grad_norm": 0.33546798335883604,
"learning_rate": 1.6265886260089337e-06,
"loss": 1.3117,
"step": 4907
},
{
"epoch": 0.8721844595495135,
"grad_norm": 0.3334953171352217,
"learning_rate": 1.622135512124161e-06,
"loss": 1.3023,
"step": 4908
},
{
"epoch": 0.8723621662446133,
"grad_norm": 0.33557024775461997,
"learning_rate": 1.617688244583695e-06,
"loss": 1.2973,
"step": 4909
},
{
"epoch": 0.872539872939713,
"grad_norm": 0.34039096943692543,
"learning_rate": 1.6132468248022926e-06,
"loss": 1.3246,
"step": 4910
},
{
"epoch": 0.8727175796348128,
"grad_norm": 0.3338196083268901,
"learning_rate": 1.6088112541928524e-06,
"loss": 1.2852,
"step": 4911
},
{
"epoch": 0.8728952863299125,
"grad_norm": 0.3378227262247297,
"learning_rate": 1.6043815341664148e-06,
"loss": 1.2705,
"step": 4912
},
{
"epoch": 0.8730729930250122,
"grad_norm": 0.33787801325061284,
"learning_rate": 1.5999576661321548e-06,
"loss": 1.3019,
"step": 4913
},
{
"epoch": 0.8732506997201119,
"grad_norm": 0.34054723493952266,
"learning_rate": 1.5955396514973908e-06,
"loss": 1.2873,
"step": 4914
},
{
"epoch": 0.8734284064152117,
"grad_norm": 0.33760796041122776,
"learning_rate": 1.5911274916675723e-06,
"loss": 1.3125,
"step": 4915
},
{
"epoch": 0.8736061131103114,
"grad_norm": 0.33030108153916715,
"learning_rate": 1.5867211880462963e-06,
"loss": 1.2753,
"step": 4916
},
{
"epoch": 0.8737838198054112,
"grad_norm": 0.3375258651147281,
"learning_rate": 1.5823207420352882e-06,
"loss": 1.3106,
"step": 4917
},
{
"epoch": 0.873961526500511,
"grad_norm": 0.3429316591692246,
"learning_rate": 1.5779261550344106e-06,
"loss": 1.2906,
"step": 4918
},
{
"epoch": 0.8741392331956106,
"grad_norm": 0.3512007760280824,
"learning_rate": 1.5735374284416693e-06,
"loss": 1.2947,
"step": 4919
},
{
"epoch": 0.8743169398907104,
"grad_norm": 0.33728083834735495,
"learning_rate": 1.5691545636531903e-06,
"loss": 1.255,
"step": 4920
},
{
"epoch": 0.8744946465858101,
"grad_norm": 0.33945290536956674,
"learning_rate": 1.5647775620632555e-06,
"loss": 1.3479,
"step": 4921
},
{
"epoch": 0.8746723532809099,
"grad_norm": 0.3980407418588491,
"learning_rate": 1.5604064250642693e-06,
"loss": 1.322,
"step": 4922
},
{
"epoch": 0.8748500599760096,
"grad_norm": 0.33824828649326716,
"learning_rate": 1.5560411540467723e-06,
"loss": 1.3272,
"step": 4923
},
{
"epoch": 0.8750277666711094,
"grad_norm": 0.3650093015962154,
"learning_rate": 1.551681750399432e-06,
"loss": 1.2908,
"step": 4924
},
{
"epoch": 0.875205473366209,
"grad_norm": 0.336530889729427,
"learning_rate": 1.5473282155090718e-06,
"loss": 1.2964,
"step": 4925
},
{
"epoch": 0.8753831800613088,
"grad_norm": 0.35366308091748894,
"learning_rate": 1.54298055076062e-06,
"loss": 1.2771,
"step": 4926
},
{
"epoch": 0.8755608867564085,
"grad_norm": 0.3440198993463506,
"learning_rate": 1.5386387575371564e-06,
"loss": 1.3454,
"step": 4927
},
{
"epoch": 0.8757385934515083,
"grad_norm": 0.33137670434685634,
"learning_rate": 1.5343028372198854e-06,
"loss": 1.239,
"step": 4928
},
{
"epoch": 0.875916300146608,
"grad_norm": 0.3431680473363984,
"learning_rate": 1.529972791188139e-06,
"loss": 1.3264,
"step": 4929
},
{
"epoch": 0.8760940068417078,
"grad_norm": 0.33612447540479234,
"learning_rate": 1.5256486208193977e-06,
"loss": 1.2719,
"step": 4930
},
{
"epoch": 0.8762717135368076,
"grad_norm": 0.34185946597682226,
"learning_rate": 1.5213303274892566e-06,
"loss": 1.337,
"step": 4931
},
{
"epoch": 0.8764494202319072,
"grad_norm": 0.3423321919027742,
"learning_rate": 1.5170179125714436e-06,
"loss": 1.328,
"step": 4932
},
{
"epoch": 0.876627126927007,
"grad_norm": 0.34363334416762087,
"learning_rate": 1.512711377437821e-06,
"loss": 1.3278,
"step": 4933
},
{
"epoch": 0.8768048336221067,
"grad_norm": 0.3409723193122846,
"learning_rate": 1.5084107234583779e-06,
"loss": 1.301,
"step": 4934
},
{
"epoch": 0.8769825403172065,
"grad_norm": 0.34056608577481495,
"learning_rate": 1.504115952001235e-06,
"loss": 1.3109,
"step": 4935
},
{
"epoch": 0.8771602470123062,
"grad_norm": 0.33560913777688217,
"learning_rate": 1.4998270644326373e-06,
"loss": 1.2951,
"step": 4936
},
{
"epoch": 0.877337953707406,
"grad_norm": 0.33696404946567415,
"learning_rate": 1.4955440621169626e-06,
"loss": 1.3231,
"step": 4937
},
{
"epoch": 0.8775156604025056,
"grad_norm": 0.33729977457134513,
"learning_rate": 1.4912669464167095e-06,
"loss": 1.309,
"step": 4938
},
{
"epoch": 0.8776933670976054,
"grad_norm": 0.33427079806786014,
"learning_rate": 1.4869957186925187e-06,
"loss": 1.2983,
"step": 4939
},
{
"epoch": 0.8778710737927051,
"grad_norm": 0.3361565294215759,
"learning_rate": 1.482730380303139e-06,
"loss": 1.3069,
"step": 4940
},
{
"epoch": 0.8780487804878049,
"grad_norm": 0.3601237811023612,
"learning_rate": 1.4784709326054648e-06,
"loss": 1.3076,
"step": 4941
},
{
"epoch": 0.8782264871829046,
"grad_norm": 0.3454655138478688,
"learning_rate": 1.4742173769544943e-06,
"loss": 1.2985,
"step": 4942
},
{
"epoch": 0.8784041938780044,
"grad_norm": 0.34230127714873676,
"learning_rate": 1.4699697147033676e-06,
"loss": 1.3099,
"step": 4943
},
{
"epoch": 0.8785819005731041,
"grad_norm": 0.3382965490057003,
"learning_rate": 1.465727947203348e-06,
"loss": 1.3236,
"step": 4944
},
{
"epoch": 0.8787596072682038,
"grad_norm": 0.3349516789256831,
"learning_rate": 1.461492075803823e-06,
"loss": 1.3194,
"step": 4945
},
{
"epoch": 0.8789373139633035,
"grad_norm": 0.34505359861319856,
"learning_rate": 1.4572621018523013e-06,
"loss": 1.3552,
"step": 4946
},
{
"epoch": 0.8791150206584033,
"grad_norm": 0.3317439346243464,
"learning_rate": 1.4530380266944177e-06,
"loss": 1.2905,
"step": 4947
},
{
"epoch": 0.8792927273535031,
"grad_norm": 0.33909906479453317,
"learning_rate": 1.448819851673926e-06,
"loss": 1.2942,
"step": 4948
},
{
"epoch": 0.8794704340486028,
"grad_norm": 0.3419858408642446,
"learning_rate": 1.444607578132713e-06,
"loss": 1.303,
"step": 4949
},
{
"epoch": 0.8796481407437026,
"grad_norm": 0.34275313947630576,
"learning_rate": 1.4404012074107776e-06,
"loss": 1.2957,
"step": 4950
},
{
"epoch": 0.8798258474388022,
"grad_norm": 0.3408195379387933,
"learning_rate": 1.4362007408462476e-06,
"loss": 1.3226,
"step": 4951
},
{
"epoch": 0.880003554133902,
"grad_norm": 0.33801741485903786,
"learning_rate": 1.432006179775367e-06,
"loss": 1.3085,
"step": 4952
},
{
"epoch": 0.8801812608290017,
"grad_norm": 0.338466666924121,
"learning_rate": 1.427817525532511e-06,
"loss": 1.2852,
"step": 4953
},
{
"epoch": 0.8803589675241015,
"grad_norm": 0.3311454527042409,
"learning_rate": 1.4236347794501648e-06,
"loss": 1.2618,
"step": 4954
},
{
"epoch": 0.8805366742192012,
"grad_norm": 0.3309684187017016,
"learning_rate": 1.4194579428589395e-06,
"loss": 1.2797,
"step": 4955
},
{
"epoch": 0.880714380914301,
"grad_norm": 0.3365226763194636,
"learning_rate": 1.4152870170875676e-06,
"loss": 1.2978,
"step": 4956
},
{
"epoch": 0.8808920876094006,
"grad_norm": 0.33955660479341365,
"learning_rate": 1.4111220034628925e-06,
"loss": 1.3279,
"step": 4957
},
{
"epoch": 0.8810697943045004,
"grad_norm": 0.3314393093620107,
"learning_rate": 1.4069629033098898e-06,
"loss": 1.2963,
"step": 4958
},
{
"epoch": 0.8812475009996001,
"grad_norm": 0.3681172841263416,
"learning_rate": 1.4028097179516453e-06,
"loss": 1.3253,
"step": 4959
},
{
"epoch": 0.8814252076946999,
"grad_norm": 0.3363585472382271,
"learning_rate": 1.3986624487093647e-06,
"loss": 1.2933,
"step": 4960
},
{
"epoch": 0.8816029143897997,
"grad_norm": 0.33338803008039236,
"learning_rate": 1.3945210969023747e-06,
"loss": 1.3081,
"step": 4961
},
{
"epoch": 0.8817806210848994,
"grad_norm": 0.33977083189928003,
"learning_rate": 1.3903856638481106e-06,
"loss": 1.3339,
"step": 4962
},
{
"epoch": 0.8819583277799992,
"grad_norm": 0.3361101808727305,
"learning_rate": 1.3862561508621442e-06,
"loss": 1.2815,
"step": 4963
},
{
"epoch": 0.8821360344750988,
"grad_norm": 0.33054367059156714,
"learning_rate": 1.3821325592581402e-06,
"loss": 1.2873,
"step": 4964
},
{
"epoch": 0.8823137411701986,
"grad_norm": 0.33785081161422137,
"learning_rate": 1.3780148903478919e-06,
"loss": 1.2878,
"step": 4965
},
{
"epoch": 0.8824914478652983,
"grad_norm": 0.3389732879552767,
"learning_rate": 1.3739031454413088e-06,
"loss": 1.3218,
"step": 4966
},
{
"epoch": 0.8826691545603981,
"grad_norm": 0.3469386464862643,
"learning_rate": 1.3697973258464158e-06,
"loss": 1.3007,
"step": 4967
},
{
"epoch": 0.8828468612554978,
"grad_norm": 0.3359955625235351,
"learning_rate": 1.3656974328693507e-06,
"loss": 1.3017,
"step": 4968
},
{
"epoch": 0.8830245679505976,
"grad_norm": 0.3392413854022774,
"learning_rate": 1.3616034678143652e-06,
"loss": 1.3473,
"step": 4969
},
{
"epoch": 0.8832022746456972,
"grad_norm": 0.3350766941680935,
"learning_rate": 1.357515431983829e-06,
"loss": 1.2902,
"step": 4970
},
{
"epoch": 0.883379981340797,
"grad_norm": 0.33905708366419296,
"learning_rate": 1.3534333266782195e-06,
"loss": 1.2911,
"step": 4971
},
{
"epoch": 0.8835576880358967,
"grad_norm": 0.3341775773274519,
"learning_rate": 1.3493571531961358e-06,
"loss": 1.309,
"step": 4972
},
{
"epoch": 0.8837353947309965,
"grad_norm": 0.3356019934096407,
"learning_rate": 1.3452869128342805e-06,
"loss": 1.2864,
"step": 4973
},
{
"epoch": 0.8839131014260962,
"grad_norm": 0.3358525994468507,
"learning_rate": 1.3412226068874756e-06,
"loss": 1.2847,
"step": 4974
},
{
"epoch": 0.884090808121196,
"grad_norm": 0.3362979083775878,
"learning_rate": 1.3371642366486559e-06,
"loss": 1.285,
"step": 4975
},
{
"epoch": 0.8842685148162958,
"grad_norm": 0.33480941307741596,
"learning_rate": 1.3331118034088574e-06,
"loss": 1.3016,
"step": 4976
},
{
"epoch": 0.8844462215113954,
"grad_norm": 0.33847829718027733,
"learning_rate": 1.3290653084572447e-06,
"loss": 1.2965,
"step": 4977
},
{
"epoch": 0.8846239282064952,
"grad_norm": 0.3410010261041247,
"learning_rate": 1.3250247530810834e-06,
"loss": 1.3596,
"step": 4978
},
{
"epoch": 0.8848016349015949,
"grad_norm": 0.3322527574209985,
"learning_rate": 1.3209901385657453e-06,
"loss": 1.2754,
"step": 4979
},
{
"epoch": 0.8849793415966947,
"grad_norm": 0.3355345280944348,
"learning_rate": 1.3169614661947128e-06,
"loss": 1.3133,
"step": 4980
},
{
"epoch": 0.8851570482917944,
"grad_norm": 0.3391638436886404,
"learning_rate": 1.312938737249594e-06,
"loss": 1.3151,
"step": 4981
},
{
"epoch": 0.8853347549868942,
"grad_norm": 0.3353781916457287,
"learning_rate": 1.308921953010087e-06,
"loss": 1.2846,
"step": 4982
},
{
"epoch": 0.8855124616819938,
"grad_norm": 0.3360063264861376,
"learning_rate": 1.3049111147540083e-06,
"loss": 1.2726,
"step": 4983
},
{
"epoch": 0.8856901683770936,
"grad_norm": 0.3349067117377192,
"learning_rate": 1.300906223757281e-06,
"loss": 1.309,
"step": 4984
},
{
"epoch": 0.8858678750721933,
"grad_norm": 0.3327688989055137,
"learning_rate": 1.2969072812939377e-06,
"loss": 1.2853,
"step": 4985
},
{
"epoch": 0.8860455817672931,
"grad_norm": 0.3437299866512817,
"learning_rate": 1.2929142886361134e-06,
"loss": 1.3212,
"step": 4986
},
{
"epoch": 0.8862232884623928,
"grad_norm": 0.3406621023415861,
"learning_rate": 1.2889272470540571e-06,
"loss": 1.3172,
"step": 4987
},
{
"epoch": 0.8864009951574926,
"grad_norm": 0.34372859180376025,
"learning_rate": 1.2849461578161226e-06,
"loss": 1.3385,
"step": 4988
},
{
"epoch": 0.8865787018525922,
"grad_norm": 0.33091955826954383,
"learning_rate": 1.2809710221887662e-06,
"loss": 1.288,
"step": 4989
},
{
"epoch": 0.886756408547692,
"grad_norm": 0.3373319299352474,
"learning_rate": 1.2770018414365515e-06,
"loss": 1.3215,
"step": 4990
},
{
"epoch": 0.8869341152427918,
"grad_norm": 0.33290944252179067,
"learning_rate": 1.2730386168221575e-06,
"loss": 1.2953,
"step": 4991
},
{
"epoch": 0.8871118219378915,
"grad_norm": 0.3403107767632461,
"learning_rate": 1.2690813496063537e-06,
"loss": 1.3029,
"step": 4992
},
{
"epoch": 0.8872895286329913,
"grad_norm": 0.3455754728310711,
"learning_rate": 1.2651300410480261e-06,
"loss": 1.2904,
"step": 4993
},
{
"epoch": 0.887467235328091,
"grad_norm": 0.340611714387933,
"learning_rate": 1.2611846924041538e-06,
"loss": 1.3057,
"step": 4994
},
{
"epoch": 0.8876449420231908,
"grad_norm": 0.3375041133022778,
"learning_rate": 1.2572453049298328e-06,
"loss": 1.2989,
"step": 4995
},
{
"epoch": 0.8878226487182904,
"grad_norm": 0.33281074686845874,
"learning_rate": 1.253311879878254e-06,
"loss": 1.3032,
"step": 4996
},
{
"epoch": 0.8880003554133902,
"grad_norm": 0.3596426100353275,
"learning_rate": 1.249384418500712e-06,
"loss": 1.2914,
"step": 4997
},
{
"epoch": 0.8881780621084899,
"grad_norm": 0.3393698869479232,
"learning_rate": 1.2454629220466075e-06,
"loss": 1.3325,
"step": 4998
},
{
"epoch": 0.8883557688035897,
"grad_norm": 0.4632593591452712,
"learning_rate": 1.2415473917634403e-06,
"loss": 1.3247,
"step": 4999
},
{
"epoch": 0.8885334754986894,
"grad_norm": 0.33090151166024834,
"learning_rate": 1.2376378288968226e-06,
"loss": 1.2593,
"step": 5000
},
{
"epoch": 0.8887111821937892,
"grad_norm": 0.3378695530735783,
"learning_rate": 1.233734234690449e-06,
"loss": 1.2943,
"step": 5001
},
{
"epoch": 0.8888888888888888,
"grad_norm": 0.34611562162562953,
"learning_rate": 1.2298366103861326e-06,
"loss": 1.3441,
"step": 5002
},
{
"epoch": 0.8890665955839886,
"grad_norm": 0.3364288048362215,
"learning_rate": 1.2259449572237792e-06,
"loss": 1.3482,
"step": 5003
},
{
"epoch": 0.8892443022790884,
"grad_norm": 0.34005973731691036,
"learning_rate": 1.2220592764413918e-06,
"loss": 1.3148,
"step": 5004
},
{
"epoch": 0.8894220089741881,
"grad_norm": 0.3443908010911508,
"learning_rate": 1.2181795692750887e-06,
"loss": 1.3254,
"step": 5005
},
{
"epoch": 0.8895997156692879,
"grad_norm": 0.3342193288331521,
"learning_rate": 1.214305836959071e-06,
"loss": 1.2871,
"step": 5006
},
{
"epoch": 0.8897774223643876,
"grad_norm": 0.33494340433977604,
"learning_rate": 1.2104380807256488e-06,
"loss": 1.2934,
"step": 5007
},
{
"epoch": 0.8899551290594874,
"grad_norm": 0.3374600035867566,
"learning_rate": 1.2065763018052267e-06,
"loss": 1.3039,
"step": 5008
},
{
"epoch": 0.890132835754587,
"grad_norm": 0.3358426810157898,
"learning_rate": 1.2027205014263088e-06,
"loss": 1.2973,
"step": 5009
},
{
"epoch": 0.8903105424496868,
"grad_norm": 0.3344374695593696,
"learning_rate": 1.198870680815496e-06,
"loss": 1.3026,
"step": 5010
},
{
"epoch": 0.8904882491447865,
"grad_norm": 0.34366334292895406,
"learning_rate": 1.195026841197493e-06,
"loss": 1.3492,
"step": 5011
},
{
"epoch": 0.8906659558398863,
"grad_norm": 0.3348502604297981,
"learning_rate": 1.191188983795095e-06,
"loss": 1.2922,
"step": 5012
},
{
"epoch": 0.890843662534986,
"grad_norm": 0.33962833868769976,
"learning_rate": 1.1873571098291947e-06,
"loss": 1.3076,
"step": 5013
},
{
"epoch": 0.8910213692300858,
"grad_norm": 0.34504549340657514,
"learning_rate": 1.1835312205187877e-06,
"loss": 1.3324,
"step": 5014
},
{
"epoch": 0.8911990759251854,
"grad_norm": 0.33732990910007354,
"learning_rate": 1.1797113170809581e-06,
"loss": 1.31,
"step": 5015
},
{
"epoch": 0.8913767826202852,
"grad_norm": 0.3410777908158039,
"learning_rate": 1.1758974007308943e-06,
"loss": 1.3241,
"step": 5016
},
{
"epoch": 0.891554489315385,
"grad_norm": 0.3365905394586732,
"learning_rate": 1.1720894726818654e-06,
"loss": 1.3376,
"step": 5017
},
{
"epoch": 0.8917321960104847,
"grad_norm": 0.34705115702279254,
"learning_rate": 1.1682875341452494e-06,
"loss": 1.3202,
"step": 5018
},
{
"epoch": 0.8919099027055845,
"grad_norm": 0.33639870681132666,
"learning_rate": 1.1644915863305163e-06,
"loss": 1.303,
"step": 5019
},
{
"epoch": 0.8920876094006842,
"grad_norm": 0.3393326306530512,
"learning_rate": 1.160701630445229e-06,
"loss": 1.3212,
"step": 5020
},
{
"epoch": 0.8922653160957839,
"grad_norm": 0.3279676023597697,
"learning_rate": 1.156917667695041e-06,
"loss": 1.2753,
"step": 5021
},
{
"epoch": 0.8924430227908836,
"grad_norm": 0.33757744394021727,
"learning_rate": 1.153139699283703e-06,
"loss": 1.3165,
"step": 5022
},
{
"epoch": 0.8926207294859834,
"grad_norm": 0.3367051085190104,
"learning_rate": 1.1493677264130575e-06,
"loss": 1.317,
"step": 5023
},
{
"epoch": 0.8927984361810831,
"grad_norm": 0.32859898000545484,
"learning_rate": 1.1456017502830408e-06,
"loss": 1.2633,
"step": 5024
},
{
"epoch": 0.8929761428761829,
"grad_norm": 0.3364812149219591,
"learning_rate": 1.1418417720916785e-06,
"loss": 1.3142,
"step": 5025
},
{
"epoch": 0.8931538495712826,
"grad_norm": 0.3328176446616962,
"learning_rate": 1.1380877930350943e-06,
"loss": 1.2837,
"step": 5026
},
{
"epoch": 0.8933315562663824,
"grad_norm": 0.3320497639717848,
"learning_rate": 1.1343398143074947e-06,
"loss": 1.2969,
"step": 5027
},
{
"epoch": 0.893509262961482,
"grad_norm": 0.3383980038643938,
"learning_rate": 1.130597837101186e-06,
"loss": 1.2986,
"step": 5028
},
{
"epoch": 0.8936869696565818,
"grad_norm": 0.3523648989172742,
"learning_rate": 1.1268618626065608e-06,
"loss": 1.3115,
"step": 5029
},
{
"epoch": 0.8938646763516815,
"grad_norm": 0.33410622431237214,
"learning_rate": 1.1231318920121015e-06,
"loss": 1.2836,
"step": 5030
},
{
"epoch": 0.8940423830467813,
"grad_norm": 0.3316546981955704,
"learning_rate": 1.1194079265043878e-06,
"loss": 1.2508,
"step": 5031
},
{
"epoch": 0.894220089741881,
"grad_norm": 0.33308445713697243,
"learning_rate": 1.115689967268072e-06,
"loss": 1.3229,
"step": 5032
},
{
"epoch": 0.8943977964369808,
"grad_norm": 0.3400415356327592,
"learning_rate": 1.111978015485915e-06,
"loss": 1.2877,
"step": 5033
},
{
"epoch": 0.8945755031320805,
"grad_norm": 0.3352372206881174,
"learning_rate": 1.1082720723387564e-06,
"loss": 1.2675,
"step": 5034
},
{
"epoch": 0.8947532098271802,
"grad_norm": 0.3382856304763251,
"learning_rate": 1.1045721390055265e-06,
"loss": 1.2991,
"step": 5035
},
{
"epoch": 0.89493091652228,
"grad_norm": 0.33151217422310986,
"learning_rate": 1.1008782166632415e-06,
"loss": 1.293,
"step": 5036
},
{
"epoch": 0.8951086232173797,
"grad_norm": 0.3431201810421127,
"learning_rate": 1.0971903064870126e-06,
"loss": 1.3229,
"step": 5037
},
{
"epoch": 0.8952863299124795,
"grad_norm": 0.33294437108233216,
"learning_rate": 1.0935084096500327e-06,
"loss": 1.2788,
"step": 5038
},
{
"epoch": 0.8954640366075792,
"grad_norm": 0.33421337771505605,
"learning_rate": 1.0898325273235777e-06,
"loss": 1.2907,
"step": 5039
},
{
"epoch": 0.895641743302679,
"grad_norm": 0.3696488759082839,
"learning_rate": 1.086162660677017e-06,
"loss": 1.2628,
"step": 5040
},
{
"epoch": 0.8958194499977786,
"grad_norm": 0.3322602358338896,
"learning_rate": 1.0824988108778035e-06,
"loss": 1.3013,
"step": 5041
},
{
"epoch": 0.8959971566928784,
"grad_norm": 0.34674448417074666,
"learning_rate": 1.07884097909148e-06,
"loss": 1.3527,
"step": 5042
},
{
"epoch": 0.8961748633879781,
"grad_norm": 0.3345631805347794,
"learning_rate": 1.0751891664816672e-06,
"loss": 1.2846,
"step": 5043
},
{
"epoch": 0.8963525700830779,
"grad_norm": 0.36846350052943105,
"learning_rate": 1.07154337421008e-06,
"loss": 1.2981,
"step": 5044
},
{
"epoch": 0.8965302767781776,
"grad_norm": 0.32892781152518447,
"learning_rate": 1.0679036034365108e-06,
"loss": 1.2487,
"step": 5045
},
{
"epoch": 0.8967079834732774,
"grad_norm": 0.3330087153362224,
"learning_rate": 1.064269855318838e-06,
"loss": 1.3097,
"step": 5046
},
{
"epoch": 0.896885690168377,
"grad_norm": 0.4645849538924861,
"learning_rate": 1.0606421310130277e-06,
"loss": 1.2765,
"step": 5047
},
{
"epoch": 0.8970633968634768,
"grad_norm": 0.33586115119951615,
"learning_rate": 1.0570204316731258e-06,
"loss": 1.2737,
"step": 5048
},
{
"epoch": 0.8972411035585766,
"grad_norm": 0.334345964313032,
"learning_rate": 1.053404758451264e-06,
"loss": 1.2725,
"step": 5049
},
{
"epoch": 0.8974188102536763,
"grad_norm": 0.3373894169702928,
"learning_rate": 1.0497951124976513e-06,
"loss": 1.2974,
"step": 5050
},
{
"epoch": 0.8975965169487761,
"grad_norm": 0.3359603669490781,
"learning_rate": 1.0461914949605912e-06,
"loss": 1.2915,
"step": 5051
},
{
"epoch": 0.8977742236438758,
"grad_norm": 0.34040534511239845,
"learning_rate": 1.04259390698646e-06,
"loss": 1.3191,
"step": 5052
},
{
"epoch": 0.8979519303389755,
"grad_norm": 0.3369290971814123,
"learning_rate": 1.03900234971972e-06,
"loss": 1.328,
"step": 5053
},
{
"epoch": 0.8981296370340752,
"grad_norm": 0.3327108397880067,
"learning_rate": 1.035416824302906e-06,
"loss": 1.2825,
"step": 5054
},
{
"epoch": 0.898307343729175,
"grad_norm": 0.33507387333260813,
"learning_rate": 1.0318373318766416e-06,
"loss": 1.2826,
"step": 5055
},
{
"epoch": 0.8984850504242747,
"grad_norm": 0.3417434624161221,
"learning_rate": 1.0282638735796379e-06,
"loss": 1.3009,
"step": 5056
},
{
"epoch": 0.8986627571193745,
"grad_norm": 0.33755111306803476,
"learning_rate": 1.0246964505486768e-06,
"loss": 1.3092,
"step": 5057
},
{
"epoch": 0.8988404638144742,
"grad_norm": 0.3331171482999626,
"learning_rate": 1.021135063918619e-06,
"loss": 1.2846,
"step": 5058
},
{
"epoch": 0.899018170509574,
"grad_norm": 0.33177477975331376,
"learning_rate": 1.017579714822412e-06,
"loss": 1.2809,
"step": 5059
},
{
"epoch": 0.8991958772046736,
"grad_norm": 0.33214213090922806,
"learning_rate": 1.01403040439108e-06,
"loss": 1.2982,
"step": 5060
},
{
"epoch": 0.8993735838997734,
"grad_norm": 0.34050319303541504,
"learning_rate": 1.0104871337537214e-06,
"loss": 1.3125,
"step": 5061
},
{
"epoch": 0.8995512905948732,
"grad_norm": 0.33557697789342145,
"learning_rate": 1.0069499040375198e-06,
"loss": 1.2961,
"step": 5062
},
{
"epoch": 0.8997289972899729,
"grad_norm": 0.344485439090265,
"learning_rate": 1.0034187163677344e-06,
"loss": 1.3438,
"step": 5063
},
{
"epoch": 0.8999067039850727,
"grad_norm": 0.34185759389303927,
"learning_rate": 9.998935718676982e-07,
"loss": 1.3125,
"step": 5064
},
{
"epoch": 0.9000844106801724,
"grad_norm": 0.3378197794350579,
"learning_rate": 9.963744716588342e-07,
"loss": 1.3057,
"step": 5065
},
{
"epoch": 0.9002621173752721,
"grad_norm": 0.3438106797208994,
"learning_rate": 9.928614168606287e-07,
"loss": 1.3159,
"step": 5066
},
{
"epoch": 0.9004398240703718,
"grad_norm": 0.3427487923610736,
"learning_rate": 9.893544085906526e-07,
"loss": 1.3207,
"step": 5067
},
{
"epoch": 0.9006175307654716,
"grad_norm": 0.34105761954295766,
"learning_rate": 9.85853447964551e-07,
"loss": 1.3397,
"step": 5068
},
{
"epoch": 0.9007952374605713,
"grad_norm": 0.3368328687846125,
"learning_rate": 9.82358536096044e-07,
"loss": 1.3044,
"step": 5069
},
{
"epoch": 0.9009729441556711,
"grad_norm": 0.3327699504497123,
"learning_rate": 9.788696740969295e-07,
"loss": 1.2631,
"step": 5070
},
{
"epoch": 0.9011506508507708,
"grad_norm": 0.3386639386140282,
"learning_rate": 9.75386863077079e-07,
"loss": 1.2989,
"step": 5071
},
{
"epoch": 0.9013283575458706,
"grad_norm": 0.3457375190410628,
"learning_rate": 9.719101041444424e-07,
"loss": 1.3108,
"step": 5072
},
{
"epoch": 0.9015060642409702,
"grad_norm": 0.3381255419509287,
"learning_rate": 9.68439398405041e-07,
"loss": 1.292,
"step": 5073
},
{
"epoch": 0.90168377093607,
"grad_norm": 0.33829550368796457,
"learning_rate": 9.6497474696297e-07,
"loss": 1.2939,
"step": 5074
},
{
"epoch": 0.9018614776311698,
"grad_norm": 0.468792305381327,
"learning_rate": 9.61516150920403e-07,
"loss": 1.3427,
"step": 5075
},
{
"epoch": 0.9020391843262695,
"grad_norm": 0.33288249034263573,
"learning_rate": 9.580636113775842e-07,
"loss": 1.288,
"step": 5076
},
{
"epoch": 0.9022168910213693,
"grad_norm": 0.3387745615053534,
"learning_rate": 9.54617129432831e-07,
"loss": 1.3005,
"step": 5077
},
{
"epoch": 0.902394597716469,
"grad_norm": 0.33524365519801075,
"learning_rate": 9.511767061825283e-07,
"loss": 1.2741,
"step": 5078
},
{
"epoch": 0.9025723044115687,
"grad_norm": 0.33613078142547076,
"learning_rate": 9.477423427211475e-07,
"loss": 1.3105,
"step": 5079
},
{
"epoch": 0.9027500111066684,
"grad_norm": 0.3358237933837281,
"learning_rate": 9.443140401412232e-07,
"loss": 1.3054,
"step": 5080
},
{
"epoch": 0.9029277178017682,
"grad_norm": 0.33328280870564914,
"learning_rate": 9.408917995333588e-07,
"loss": 1.3095,
"step": 5081
},
{
"epoch": 0.9031054244968679,
"grad_norm": 0.3449151501119699,
"learning_rate": 9.374756219862369e-07,
"loss": 1.3346,
"step": 5082
},
{
"epoch": 0.9032831311919677,
"grad_norm": 0.3376135107083614,
"learning_rate": 9.340655085866057e-07,
"loss": 1.3035,
"step": 5083
},
{
"epoch": 0.9034608378870674,
"grad_norm": 0.33848496221030716,
"learning_rate": 9.306614604192865e-07,
"loss": 1.2943,
"step": 5084
},
{
"epoch": 0.9036385445821671,
"grad_norm": 0.33174111944571416,
"learning_rate": 9.27263478567173e-07,
"loss": 1.2741,
"step": 5085
},
{
"epoch": 0.9038162512772668,
"grad_norm": 0.34308831049315436,
"learning_rate": 9.238715641112272e-07,
"loss": 1.3514,
"step": 5086
},
{
"epoch": 0.9039939579723666,
"grad_norm": 0.33495478620134506,
"learning_rate": 9.204857181304772e-07,
"loss": 1.3078,
"step": 5087
},
{
"epoch": 0.9041716646674663,
"grad_norm": 0.337956165073301,
"learning_rate": 9.171059417020256e-07,
"loss": 1.3168,
"step": 5088
},
{
"epoch": 0.9043493713625661,
"grad_norm": 0.3366567063251527,
"learning_rate": 9.137322359010459e-07,
"loss": 1.3161,
"step": 5089
},
{
"epoch": 0.9045270780576659,
"grad_norm": 0.33670776196921365,
"learning_rate": 9.10364601800775e-07,
"loss": 1.297,
"step": 5090
},
{
"epoch": 0.9047047847527656,
"grad_norm": 0.33174485722431946,
"learning_rate": 9.070030404725227e-07,
"loss": 1.29,
"step": 5091
},
{
"epoch": 0.9048824914478653,
"grad_norm": 0.34222568136440523,
"learning_rate": 9.036475529856603e-07,
"loss": 1.2893,
"step": 5092
},
{
"epoch": 0.905060198142965,
"grad_norm": 0.3426580384670296,
"learning_rate": 9.002981404076361e-07,
"loss": 1.3234,
"step": 5093
},
{
"epoch": 0.9052379048380648,
"grad_norm": 0.34711881856761106,
"learning_rate": 8.969548038039577e-07,
"loss": 1.3328,
"step": 5094
},
{
"epoch": 0.9054156115331645,
"grad_norm": 0.34475968755153513,
"learning_rate": 8.936175442382078e-07,
"loss": 1.3031,
"step": 5095
},
{
"epoch": 0.9055933182282643,
"grad_norm": 0.3411742592508703,
"learning_rate": 8.90286362772026e-07,
"loss": 1.3316,
"step": 5096
},
{
"epoch": 0.905771024923364,
"grad_norm": 0.34264420231169745,
"learning_rate": 8.869612604651268e-07,
"loss": 1.306,
"step": 5097
},
{
"epoch": 0.9059487316184637,
"grad_norm": 0.32773147516178014,
"learning_rate": 8.836422383752908e-07,
"loss": 1.2744,
"step": 5098
},
{
"epoch": 0.9061264383135634,
"grad_norm": 0.33739953491737523,
"learning_rate": 8.803292975583555e-07,
"loss": 1.2858,
"step": 5099
},
{
"epoch": 0.9063041450086632,
"grad_norm": 0.3371289790027248,
"learning_rate": 8.770224390682314e-07,
"loss": 1.3084,
"step": 5100
},
{
"epoch": 0.9064818517037629,
"grad_norm": 0.3544289511034341,
"learning_rate": 8.737216639568946e-07,
"loss": 1.321,
"step": 5101
},
{
"epoch": 0.9066595583988627,
"grad_norm": 0.33616678584277265,
"learning_rate": 8.704269732743808e-07,
"loss": 1.2978,
"step": 5102
},
{
"epoch": 0.9068372650939625,
"grad_norm": 0.3420416044887154,
"learning_rate": 8.671383680687961e-07,
"loss": 1.3042,
"step": 5103
},
{
"epoch": 0.9070149717890622,
"grad_norm": 0.35133306503575457,
"learning_rate": 8.638558493863058e-07,
"loss": 1.3371,
"step": 5104
},
{
"epoch": 0.9071926784841619,
"grad_norm": 0.3351229624203168,
"learning_rate": 8.605794182711413e-07,
"loss": 1.2686,
"step": 5105
},
{
"epoch": 0.9073703851792616,
"grad_norm": 0.33817131578245696,
"learning_rate": 8.573090757655955e-07,
"loss": 1.2771,
"step": 5106
},
{
"epoch": 0.9075480918743614,
"grad_norm": 0.3356464612645768,
"learning_rate": 8.540448229100295e-07,
"loss": 1.3045,
"step": 5107
},
{
"epoch": 0.9077257985694611,
"grad_norm": 0.3313399780379022,
"learning_rate": 8.507866607428594e-07,
"loss": 1.2914,
"step": 5108
},
{
"epoch": 0.9079035052645609,
"grad_norm": 0.3319005292034756,
"learning_rate": 8.475345903005694e-07,
"loss": 1.2848,
"step": 5109
},
{
"epoch": 0.9080812119596606,
"grad_norm": 0.3321233337754627,
"learning_rate": 8.442886126177052e-07,
"loss": 1.2558,
"step": 5110
},
{
"epoch": 0.9082589186547603,
"grad_norm": 0.3434572625789376,
"learning_rate": 8.41048728726872e-07,
"loss": 1.2926,
"step": 5111
},
{
"epoch": 0.90843662534986,
"grad_norm": 0.3373733157684014,
"learning_rate": 8.378149396587387e-07,
"loss": 1.3083,
"step": 5112
},
{
"epoch": 0.9086143320449598,
"grad_norm": 0.3367541069959196,
"learning_rate": 8.345872464420379e-07,
"loss": 1.2876,
"step": 5113
},
{
"epoch": 0.9087920387400595,
"grad_norm": 0.3320424793677119,
"learning_rate": 8.313656501035528e-07,
"loss": 1.2882,
"step": 5114
},
{
"epoch": 0.9089697454351593,
"grad_norm": 0.3370132472099522,
"learning_rate": 8.281501516681367e-07,
"loss": 1.3126,
"step": 5115
},
{
"epoch": 0.909147452130259,
"grad_norm": 0.3369610533213377,
"learning_rate": 8.24940752158696e-07,
"loss": 1.3079,
"step": 5116
},
{
"epoch": 0.9093251588253587,
"grad_norm": 0.340862036873464,
"learning_rate": 8.217374525962097e-07,
"loss": 1.3318,
"step": 5117
},
{
"epoch": 0.9095028655204584,
"grad_norm": 0.33404214384809594,
"learning_rate": 8.185402539997023e-07,
"loss": 1.2603,
"step": 5118
},
{
"epoch": 0.9096805722155582,
"grad_norm": 0.334055039852909,
"learning_rate": 8.153491573862649e-07,
"loss": 1.3128,
"step": 5119
},
{
"epoch": 0.909858278910658,
"grad_norm": 0.32570171286290317,
"learning_rate": 8.121641637710431e-07,
"loss": 1.2686,
"step": 5120
},
{
"epoch": 0.9100359856057577,
"grad_norm": 0.3291621571292636,
"learning_rate": 8.089852741672444e-07,
"loss": 1.2552,
"step": 5121
},
{
"epoch": 0.9102136923008575,
"grad_norm": 0.3413231887529552,
"learning_rate": 8.05812489586133e-07,
"loss": 1.3486,
"step": 5122
},
{
"epoch": 0.9103913989959572,
"grad_norm": 0.33102533049208654,
"learning_rate": 8.026458110370328e-07,
"loss": 1.2825,
"step": 5123
},
{
"epoch": 0.9105691056910569,
"grad_norm": 0.3334001180566168,
"learning_rate": 7.994852395273222e-07,
"loss": 1.3118,
"step": 5124
},
{
"epoch": 0.9107468123861566,
"grad_norm": 0.3413804617841322,
"learning_rate": 7.963307760624351e-07,
"loss": 1.3172,
"step": 5125
},
{
"epoch": 0.9109245190812564,
"grad_norm": 0.34612651643838194,
"learning_rate": 7.931824216458727e-07,
"loss": 1.3418,
"step": 5126
},
{
"epoch": 0.9111022257763561,
"grad_norm": 0.34208794036585816,
"learning_rate": 7.900401772791832e-07,
"loss": 1.2981,
"step": 5127
},
{
"epoch": 0.9112799324714559,
"grad_norm": 0.33979280139284124,
"learning_rate": 7.869040439619713e-07,
"loss": 1.3146,
"step": 5128
},
{
"epoch": 0.9114576391665556,
"grad_norm": 0.32825738361129514,
"learning_rate": 7.837740226919033e-07,
"loss": 1.2776,
"step": 5129
},
{
"epoch": 0.9116353458616553,
"grad_norm": 0.3383615181528766,
"learning_rate": 7.806501144646939e-07,
"loss": 1.3005,
"step": 5130
},
{
"epoch": 0.911813052556755,
"grad_norm": 0.33443644438965614,
"learning_rate": 7.775323202741191e-07,
"loss": 1.292,
"step": 5131
},
{
"epoch": 0.9119907592518548,
"grad_norm": 0.3370399103848601,
"learning_rate": 7.744206411120103e-07,
"loss": 1.2591,
"step": 5132
},
{
"epoch": 0.9121684659469546,
"grad_norm": 0.331130667860708,
"learning_rate": 7.713150779682465e-07,
"loss": 1.2728,
"step": 5133
},
{
"epoch": 0.9123461726420543,
"grad_norm": 0.3331972505294611,
"learning_rate": 7.682156318307687e-07,
"loss": 1.2755,
"step": 5134
},
{
"epoch": 0.9125238793371541,
"grad_norm": 0.3378094680987226,
"learning_rate": 7.651223036855681e-07,
"loss": 1.2986,
"step": 5135
},
{
"epoch": 0.9127015860322538,
"grad_norm": 0.3379485582873529,
"learning_rate": 7.620350945166932e-07,
"loss": 1.3319,
"step": 5136
},
{
"epoch": 0.9128792927273535,
"grad_norm": 0.3633273694576243,
"learning_rate": 7.589540053062383e-07,
"loss": 1.3034,
"step": 5137
},
{
"epoch": 0.9130569994224532,
"grad_norm": 0.34613892038397376,
"learning_rate": 7.558790370343594e-07,
"loss": 1.3436,
"step": 5138
},
{
"epoch": 0.913234706117553,
"grad_norm": 0.3339712711851095,
"learning_rate": 7.528101906792584e-07,
"loss": 1.3014,
"step": 5139
},
{
"epoch": 0.9134124128126527,
"grad_norm": 0.33886872160383386,
"learning_rate": 7.497474672171967e-07,
"loss": 1.2949,
"step": 5140
},
{
"epoch": 0.9135901195077525,
"grad_norm": 0.34060677741846945,
"learning_rate": 7.466908676224838e-07,
"loss": 1.3301,
"step": 5141
},
{
"epoch": 0.9137678262028522,
"grad_norm": 0.33348669378403045,
"learning_rate": 7.436403928674818e-07,
"loss": 1.2977,
"step": 5142
},
{
"epoch": 0.9139455328979519,
"grad_norm": 0.33235816217026537,
"learning_rate": 7.405960439226012e-07,
"loss": 1.2965,
"step": 5143
},
{
"epoch": 0.9141232395930516,
"grad_norm": 0.33198904428807097,
"learning_rate": 7.375578217563095e-07,
"loss": 1.3127,
"step": 5144
},
{
"epoch": 0.9143009462881514,
"grad_norm": 0.3339832442500648,
"learning_rate": 7.345257273351203e-07,
"loss": 1.3082,
"step": 5145
},
{
"epoch": 0.9144786529832512,
"grad_norm": 0.3374312732131295,
"learning_rate": 7.314997616236019e-07,
"loss": 1.3229,
"step": 5146
},
{
"epoch": 0.9146563596783509,
"grad_norm": 0.33705612320144585,
"learning_rate": 7.284799255843689e-07,
"loss": 1.3354,
"step": 5147
},
{
"epoch": 0.9148340663734507,
"grad_norm": 0.33780537158837975,
"learning_rate": 7.254662201780882e-07,
"loss": 1.2852,
"step": 5148
},
{
"epoch": 0.9150117730685503,
"grad_norm": 0.3365069395792473,
"learning_rate": 7.224586463634753e-07,
"loss": 1.2825,
"step": 5149
},
{
"epoch": 0.9151894797636501,
"grad_norm": 0.3348165086814598,
"learning_rate": 7.194572050973003e-07,
"loss": 1.2979,
"step": 5150
},
{
"epoch": 0.9153671864587498,
"grad_norm": 0.32852758848000885,
"learning_rate": 7.164618973343774e-07,
"loss": 1.251,
"step": 5151
},
{
"epoch": 0.9155448931538496,
"grad_norm": 0.33812062124090264,
"learning_rate": 7.13472724027564e-07,
"loss": 1.3036,
"step": 5152
},
{
"epoch": 0.9157225998489493,
"grad_norm": 0.3393798330478529,
"learning_rate": 7.104896861277754e-07,
"loss": 1.3292,
"step": 5153
},
{
"epoch": 0.9159003065440491,
"grad_norm": 0.3403104395254294,
"learning_rate": 7.075127845839769e-07,
"loss": 1.3164,
"step": 5154
},
{
"epoch": 0.9160780132391488,
"grad_norm": 0.3288007653805369,
"learning_rate": 7.045420203431708e-07,
"loss": 1.2767,
"step": 5155
},
{
"epoch": 0.9162557199342485,
"grad_norm": 0.3309835289574933,
"learning_rate": 7.015773943504167e-07,
"loss": 1.2925,
"step": 5156
},
{
"epoch": 0.9164334266293482,
"grad_norm": 0.3348626890654266,
"learning_rate": 6.986189075488159e-07,
"loss": 1.2924,
"step": 5157
},
{
"epoch": 0.916611133324448,
"grad_norm": 0.33295910407223533,
"learning_rate": 6.956665608795199e-07,
"loss": 1.2754,
"step": 5158
},
{
"epoch": 0.9167888400195477,
"grad_norm": 0.36327590202146653,
"learning_rate": 6.927203552817263e-07,
"loss": 1.324,
"step": 5159
},
{
"epoch": 0.9169665467146475,
"grad_norm": 0.3364467189285343,
"learning_rate": 6.897802916926766e-07,
"loss": 1.2918,
"step": 5160
},
{
"epoch": 0.9171442534097473,
"grad_norm": 0.33282180718443355,
"learning_rate": 6.868463710476603e-07,
"loss": 1.2671,
"step": 5161
},
{
"epoch": 0.9173219601048469,
"grad_norm": 0.3688651563681204,
"learning_rate": 6.83918594280013e-07,
"loss": 1.3129,
"step": 5162
},
{
"epoch": 0.9174996667999467,
"grad_norm": 0.33735509036527017,
"learning_rate": 6.809969623211143e-07,
"loss": 1.2968,
"step": 5163
},
{
"epoch": 0.9176773734950464,
"grad_norm": 0.336977934516834,
"learning_rate": 6.780814761003962e-07,
"loss": 1.3091,
"step": 5164
},
{
"epoch": 0.9178550801901462,
"grad_norm": 0.3724713154551324,
"learning_rate": 6.751721365453235e-07,
"loss": 1.2859,
"step": 5165
},
{
"epoch": 0.9180327868852459,
"grad_norm": 0.3342241321137182,
"learning_rate": 6.722689445814179e-07,
"loss": 1.2753,
"step": 5166
},
{
"epoch": 0.9182104935803457,
"grad_norm": 0.3380615918067299,
"learning_rate": 6.693719011322275e-07,
"loss": 1.2826,
"step": 5167
},
{
"epoch": 0.9183882002754454,
"grad_norm": 0.4421557068706892,
"learning_rate": 6.664810071193706e-07,
"loss": 1.3552,
"step": 5168
},
{
"epoch": 0.9185659069705451,
"grad_norm": 0.33789153071882544,
"learning_rate": 6.635962634624848e-07,
"loss": 1.3066,
"step": 5169
},
{
"epoch": 0.9187436136656448,
"grad_norm": 0.3440278578130502,
"learning_rate": 6.607176710792673e-07,
"loss": 1.3158,
"step": 5170
},
{
"epoch": 0.9189213203607446,
"grad_norm": 0.33664160463439763,
"learning_rate": 6.5784523088545e-07,
"loss": 1.2985,
"step": 5171
},
{
"epoch": 0.9190990270558443,
"grad_norm": 0.34047717466592087,
"learning_rate": 6.549789437948062e-07,
"loss": 1.3505,
"step": 5172
},
{
"epoch": 0.9192767337509441,
"grad_norm": 0.3511418178697144,
"learning_rate": 6.521188107191667e-07,
"loss": 1.3825,
"step": 5173
},
{
"epoch": 0.9194544404460439,
"grad_norm": 0.32779405708617615,
"learning_rate": 6.492648325683837e-07,
"loss": 1.2629,
"step": 5174
},
{
"epoch": 0.9196321471411435,
"grad_norm": 0.36287131174393444,
"learning_rate": 6.464170102503641e-07,
"loss": 1.2843,
"step": 5175
},
{
"epoch": 0.9198098538362433,
"grad_norm": 0.33351038795451554,
"learning_rate": 6.435753446710546e-07,
"loss": 1.3016,
"step": 5176
},
{
"epoch": 0.919987560531343,
"grad_norm": 0.3362376799148404,
"learning_rate": 6.407398367344386e-07,
"loss": 1.3074,
"step": 5177
},
{
"epoch": 0.9201652672264428,
"grad_norm": 0.33043581691761825,
"learning_rate": 6.379104873425502e-07,
"loss": 1.2813,
"step": 5178
},
{
"epoch": 0.9203429739215425,
"grad_norm": 0.33882124121220536,
"learning_rate": 6.35087297395458e-07,
"loss": 1.3043,
"step": 5179
},
{
"epoch": 0.9205206806166423,
"grad_norm": 0.33441148998869236,
"learning_rate": 6.322702677912684e-07,
"loss": 1.2492,
"step": 5180
},
{
"epoch": 0.9206983873117419,
"grad_norm": 0.3297528849155827,
"learning_rate": 6.294593994261333e-07,
"loss": 1.2611,
"step": 5181
},
{
"epoch": 0.9208760940068417,
"grad_norm": 0.33790667880875574,
"learning_rate": 6.266546931942419e-07,
"loss": 1.2975,
"step": 5182
},
{
"epoch": 0.9210538007019414,
"grad_norm": 0.3359834156978389,
"learning_rate": 6.238561499878249e-07,
"loss": 1.3016,
"step": 5183
},
{
"epoch": 0.9212315073970412,
"grad_norm": 0.34139885660717656,
"learning_rate": 6.210637706971523e-07,
"loss": 1.3148,
"step": 5184
},
{
"epoch": 0.9214092140921409,
"grad_norm": 0.3392468284994007,
"learning_rate": 6.182775562105314e-07,
"loss": 1.3384,
"step": 5185
},
{
"epoch": 0.9215869207872407,
"grad_norm": 0.3330033265988267,
"learning_rate": 6.154975074143066e-07,
"loss": 1.3169,
"step": 5186
},
{
"epoch": 0.9217646274823404,
"grad_norm": 0.3367272056460735,
"learning_rate": 6.127236251928703e-07,
"loss": 1.3316,
"step": 5187
},
{
"epoch": 0.9219423341774401,
"grad_norm": 0.33874051660029286,
"learning_rate": 6.099559104286435e-07,
"loss": 1.3345,
"step": 5188
},
{
"epoch": 0.9221200408725398,
"grad_norm": 0.33429270596390304,
"learning_rate": 6.071943640020861e-07,
"loss": 1.3211,
"step": 5189
},
{
"epoch": 0.9222977475676396,
"grad_norm": 0.3511243101092831,
"learning_rate": 6.044389867916999e-07,
"loss": 1.3654,
"step": 5190
},
{
"epoch": 0.9224754542627394,
"grad_norm": 0.3326917417715172,
"learning_rate": 6.016897796740196e-07,
"loss": 1.2678,
"step": 5191
},
{
"epoch": 0.9226531609578391,
"grad_norm": 0.3327662934580919,
"learning_rate": 5.989467435236229e-07,
"loss": 1.298,
"step": 5192
},
{
"epoch": 0.9228308676529389,
"grad_norm": 0.3408594774559458,
"learning_rate": 5.962098792131233e-07,
"loss": 1.3283,
"step": 5193
},
{
"epoch": 0.9230085743480385,
"grad_norm": 0.3392139839158109,
"learning_rate": 5.93479187613164e-07,
"loss": 1.3344,
"step": 5194
},
{
"epoch": 0.9231862810431383,
"grad_norm": 0.34298945709735057,
"learning_rate": 5.907546695924304e-07,
"loss": 1.3315,
"step": 5195
},
{
"epoch": 0.923363987738238,
"grad_norm": 0.34839012281885234,
"learning_rate": 5.880363260176447e-07,
"loss": 1.3288,
"step": 5196
},
{
"epoch": 0.9235416944333378,
"grad_norm": 0.33586549145328065,
"learning_rate": 5.853241577535618e-07,
"loss": 1.3025,
"step": 5197
},
{
"epoch": 0.9237194011284375,
"grad_norm": 0.33728289503506,
"learning_rate": 5.826181656629737e-07,
"loss": 1.3073,
"step": 5198
},
{
"epoch": 0.9238971078235373,
"grad_norm": 0.3362525338975974,
"learning_rate": 5.799183506067074e-07,
"loss": 1.2981,
"step": 5199
},
{
"epoch": 0.924074814518637,
"grad_norm": 0.33936363757564386,
"learning_rate": 5.772247134436204e-07,
"loss": 1.2873,
"step": 5200
},
{
"epoch": 0.9242525212137367,
"grad_norm": 0.3294116338614067,
"learning_rate": 5.745372550306183e-07,
"loss": 1.2492,
"step": 5201
},
{
"epoch": 0.9244302279088364,
"grad_norm": 0.3355051858640874,
"learning_rate": 5.718559762226284e-07,
"loss": 1.289,
"step": 5202
},
{
"epoch": 0.9246079346039362,
"grad_norm": 0.3317076140120103,
"learning_rate": 5.691808778726127e-07,
"loss": 1.2596,
"step": 5203
},
{
"epoch": 0.924785641299036,
"grad_norm": 0.34531658916559793,
"learning_rate": 5.665119608315772e-07,
"loss": 1.3126,
"step": 5204
},
{
"epoch": 0.9249633479941357,
"grad_norm": 0.3358499138084227,
"learning_rate": 5.63849225948545e-07,
"loss": 1.2995,
"step": 5205
},
{
"epoch": 0.9251410546892355,
"grad_norm": 0.33521933388153535,
"learning_rate": 5.611926740705897e-07,
"loss": 1.2896,
"step": 5206
},
{
"epoch": 0.9253187613843351,
"grad_norm": 0.3321308974327001,
"learning_rate": 5.585423060428064e-07,
"loss": 1.275,
"step": 5207
},
{
"epoch": 0.9254964680794349,
"grad_norm": 0.34462054863522534,
"learning_rate": 5.558981227083293e-07,
"loss": 1.2927,
"step": 5208
},
{
"epoch": 0.9256741747745346,
"grad_norm": 0.337268043883938,
"learning_rate": 5.532601249083213e-07,
"loss": 1.3013,
"step": 5209
},
{
"epoch": 0.9258518814696344,
"grad_norm": 0.3389793861839343,
"learning_rate": 5.506283134819824e-07,
"loss": 1.3062,
"step": 5210
},
{
"epoch": 0.9260295881647341,
"grad_norm": 0.3345867467225901,
"learning_rate": 5.48002689266538e-07,
"loss": 1.2574,
"step": 5211
},
{
"epoch": 0.9262072948598339,
"grad_norm": 0.336054624355936,
"learning_rate": 5.453832530972491e-07,
"loss": 1.2939,
"step": 5212
},
{
"epoch": 0.9263850015549335,
"grad_norm": 0.333461061757694,
"learning_rate": 5.42770005807407e-07,
"loss": 1.3096,
"step": 5213
},
{
"epoch": 0.9265627082500333,
"grad_norm": 0.33720811045056676,
"learning_rate": 5.401629482283355e-07,
"loss": 1.3013,
"step": 5214
},
{
"epoch": 0.926740414945133,
"grad_norm": 0.33166806567972557,
"learning_rate": 5.375620811893889e-07,
"loss": 1.3068,
"step": 5215
},
{
"epoch": 0.9269181216402328,
"grad_norm": 0.3387396919526507,
"learning_rate": 5.349674055179522e-07,
"loss": 1.2744,
"step": 5216
},
{
"epoch": 0.9270958283353326,
"grad_norm": 0.34124097891316973,
"learning_rate": 5.323789220394427e-07,
"loss": 1.3144,
"step": 5217
},
{
"epoch": 0.9272735350304323,
"grad_norm": 0.34133509057736683,
"learning_rate": 5.297966315772996e-07,
"loss": 1.3483,
"step": 5218
},
{
"epoch": 0.9274512417255321,
"grad_norm": 0.34695361811520203,
"learning_rate": 5.272205349530035e-07,
"loss": 1.3694,
"step": 5219
},
{
"epoch": 0.9276289484206317,
"grad_norm": 0.3364864663291623,
"learning_rate": 5.246506329860568e-07,
"loss": 1.3083,
"step": 5220
},
{
"epoch": 0.9278066551157315,
"grad_norm": 0.3306316982932445,
"learning_rate": 5.220869264939943e-07,
"loss": 1.2607,
"step": 5221
},
{
"epoch": 0.9279843618108312,
"grad_norm": 0.3418904706087796,
"learning_rate": 5.195294162923792e-07,
"loss": 1.3266,
"step": 5222
},
{
"epoch": 0.928162068505931,
"grad_norm": 0.3426305542399539,
"learning_rate": 5.169781031948007e-07,
"loss": 1.2659,
"step": 5223
},
{
"epoch": 0.9283397752010307,
"grad_norm": 0.33676233339774037,
"learning_rate": 5.144329880128851e-07,
"loss": 1.3119,
"step": 5224
},
{
"epoch": 0.9285174818961305,
"grad_norm": 0.3346604083287368,
"learning_rate": 5.118940715562781e-07,
"loss": 1.3025,
"step": 5225
},
{
"epoch": 0.9286951885912301,
"grad_norm": 0.3407962335209748,
"learning_rate": 5.093613546326604e-07,
"loss": 1.3198,
"step": 5226
},
{
"epoch": 0.9288728952863299,
"grad_norm": 0.34157290351073866,
"learning_rate": 5.068348380477317e-07,
"loss": 1.3091,
"step": 5227
},
{
"epoch": 0.9290506019814296,
"grad_norm": 0.3309748609201002,
"learning_rate": 5.043145226052227e-07,
"loss": 1.2853,
"step": 5228
},
{
"epoch": 0.9292283086765294,
"grad_norm": 0.33126860355930476,
"learning_rate": 5.018004091069006e-07,
"loss": 1.271,
"step": 5229
},
{
"epoch": 0.9294060153716291,
"grad_norm": 0.33177479490650924,
"learning_rate": 4.99292498352546e-07,
"loss": 1.3012,
"step": 5230
},
{
"epoch": 0.9295837220667289,
"grad_norm": 0.3340161861707443,
"learning_rate": 4.967907911399783e-07,
"loss": 1.2868,
"step": 5231
},
{
"epoch": 0.9297614287618287,
"grad_norm": 0.3345184255943346,
"learning_rate": 4.942952882650321e-07,
"loss": 1.2653,
"step": 5232
},
{
"epoch": 0.9299391354569283,
"grad_norm": 0.33255228908096535,
"learning_rate": 4.918059905215767e-07,
"loss": 1.274,
"step": 5233
},
{
"epoch": 0.9301168421520281,
"grad_norm": 0.3355326393034995,
"learning_rate": 4.89322898701503e-07,
"loss": 1.3091,
"step": 5234
},
{
"epoch": 0.9302945488471278,
"grad_norm": 0.33695977868517396,
"learning_rate": 4.868460135947306e-07,
"loss": 1.2917,
"step": 5235
},
{
"epoch": 0.9304722555422276,
"grad_norm": 0.36687619870112503,
"learning_rate": 4.843753359892023e-07,
"loss": 1.3246,
"step": 5236
},
{
"epoch": 0.9306499622373273,
"grad_norm": 0.3380737634377051,
"learning_rate": 4.81910866670885e-07,
"loss": 1.3434,
"step": 5237
},
{
"epoch": 0.9308276689324271,
"grad_norm": 0.32920313617381564,
"learning_rate": 4.794526064237782e-07,
"loss": 1.255,
"step": 5238
},
{
"epoch": 0.9310053756275267,
"grad_norm": 0.3308445423000373,
"learning_rate": 4.770005560298963e-07,
"loss": 1.2768,
"step": 5239
},
{
"epoch": 0.9311830823226265,
"grad_norm": 0.3292882450720255,
"learning_rate": 4.745547162692865e-07,
"loss": 1.2716,
"step": 5240
},
{
"epoch": 0.9313607890177262,
"grad_norm": 0.3315690956991991,
"learning_rate": 4.721150879200109e-07,
"loss": 1.3053,
"step": 5241
},
{
"epoch": 0.931538495712826,
"grad_norm": 0.3492147842392085,
"learning_rate": 4.6968167175816647e-07,
"loss": 1.3409,
"step": 5242
},
{
"epoch": 0.9317162024079257,
"grad_norm": 0.4944942630281119,
"learning_rate": 4.6725446855786507e-07,
"loss": 1.3191,
"step": 5243
},
{
"epoch": 0.9318939091030255,
"grad_norm": 0.33485679715529926,
"learning_rate": 4.64833479091249e-07,
"loss": 1.2973,
"step": 5244
},
{
"epoch": 0.9320716157981251,
"grad_norm": 0.4567945448685379,
"learning_rate": 4.6241870412847557e-07,
"loss": 1.2951,
"step": 5245
},
{
"epoch": 0.9322493224932249,
"grad_norm": 0.3333145799696461,
"learning_rate": 4.600101444377347e-07,
"loss": 1.3035,
"step": 5246
},
{
"epoch": 0.9324270291883247,
"grad_norm": 0.33061749076052194,
"learning_rate": 4.5760780078523135e-07,
"loss": 1.2846,
"step": 5247
},
{
"epoch": 0.9326047358834244,
"grad_norm": 0.3353098196179165,
"learning_rate": 4.552116739352008e-07,
"loss": 1.3119,
"step": 5248
},
{
"epoch": 0.9327824425785242,
"grad_norm": 0.3321930555806189,
"learning_rate": 4.5282176464989116e-07,
"loss": 1.2455,
"step": 5249
},
{
"epoch": 0.9329601492736239,
"grad_norm": 0.32920223548583155,
"learning_rate": 4.504380736895808e-07,
"loss": 1.3079,
"step": 5250
},
{
"epoch": 0.9331378559687237,
"grad_norm": 0.33418110680316204,
"learning_rate": 4.4806060181256105e-07,
"loss": 1.2888,
"step": 5251
},
{
"epoch": 0.9333155626638233,
"grad_norm": 0.34209775043126267,
"learning_rate": 4.45689349775158e-07,
"loss": 1.3171,
"step": 5252
},
{
"epoch": 0.9334932693589231,
"grad_norm": 0.3345935581016497,
"learning_rate": 4.433243183317082e-07,
"loss": 1.3007,
"step": 5253
},
{
"epoch": 0.9336709760540228,
"grad_norm": 0.3481740849884764,
"learning_rate": 4.409655082345721e-07,
"loss": 1.2997,
"step": 5254
},
{
"epoch": 0.9338486827491226,
"grad_norm": 0.3468127159369682,
"learning_rate": 4.386129202341316e-07,
"loss": 1.3014,
"step": 5255
},
{
"epoch": 0.9340263894442223,
"grad_norm": 0.3354416703241512,
"learning_rate": 4.3626655507879034e-07,
"loss": 1.3085,
"step": 5256
},
{
"epoch": 0.9342040961393221,
"grad_norm": 0.38977093503742055,
"learning_rate": 4.33926413514969e-07,
"loss": 1.2957,
"step": 5257
},
{
"epoch": 0.9343818028344217,
"grad_norm": 0.33781482734805296,
"learning_rate": 4.3159249628711433e-07,
"loss": 1.293,
"step": 5258
},
{
"epoch": 0.9345595095295215,
"grad_norm": 0.3318386698373112,
"learning_rate": 4.2926480413768566e-07,
"loss": 1.29,
"step": 5259
},
{
"epoch": 0.9347372162246212,
"grad_norm": 0.3363434967491569,
"learning_rate": 4.2694333780716635e-07,
"loss": 1.3121,
"step": 5260
},
{
"epoch": 0.934914922919721,
"grad_norm": 0.33117434743036933,
"learning_rate": 4.246280980340589e-07,
"loss": 1.2804,
"step": 5261
},
{
"epoch": 0.9350926296148208,
"grad_norm": 0.3311479586701653,
"learning_rate": 4.223190855548853e-07,
"loss": 1.2619,
"step": 5262
},
{
"epoch": 0.9352703363099205,
"grad_norm": 0.33214321988985057,
"learning_rate": 4.200163011041869e-07,
"loss": 1.284,
"step": 5263
},
{
"epoch": 0.9354480430050203,
"grad_norm": 0.33253764484413995,
"learning_rate": 4.177197454145221e-07,
"loss": 1.2948,
"step": 5264
},
{
"epoch": 0.9356257497001199,
"grad_norm": 0.3303609842721058,
"learning_rate": 4.154294192164621e-07,
"loss": 1.2638,
"step": 5265
},
{
"epoch": 0.9358034563952197,
"grad_norm": 0.3334376576480305,
"learning_rate": 4.131453232386129e-07,
"loss": 1.3053,
"step": 5266
},
{
"epoch": 0.9359811630903194,
"grad_norm": 0.3326756021789061,
"learning_rate": 4.108674582075822e-07,
"loss": 1.3162,
"step": 5267
},
{
"epoch": 0.9361588697854192,
"grad_norm": 0.33814852722126715,
"learning_rate": 4.0859582484800374e-07,
"loss": 1.3299,
"step": 5268
},
{
"epoch": 0.9363365764805189,
"grad_norm": 0.3356921063340142,
"learning_rate": 4.063304238825261e-07,
"loss": 1.3258,
"step": 5269
},
{
"epoch": 0.9365142831756187,
"grad_norm": 0.33906186356647233,
"learning_rate": 4.040712560318172e-07,
"loss": 1.3285,
"step": 5270
},
{
"epoch": 0.9366919898707183,
"grad_norm": 0.3340254320669047,
"learning_rate": 4.0181832201455995e-07,
"loss": 1.2977,
"step": 5271
},
{
"epoch": 0.9368696965658181,
"grad_norm": 0.33040408694873435,
"learning_rate": 3.995716225474522e-07,
"loss": 1.2616,
"step": 5272
},
{
"epoch": 0.9370474032609178,
"grad_norm": 0.3387648720395594,
"learning_rate": 3.973311583452155e-07,
"loss": 1.3133,
"step": 5273
},
{
"epoch": 0.9372251099560176,
"grad_norm": 0.32878045081332674,
"learning_rate": 3.9509693012058204e-07,
"loss": 1.2454,
"step": 5274
},
{
"epoch": 0.9374028166511174,
"grad_norm": 0.3342778636439139,
"learning_rate": 3.9286893858430096e-07,
"loss": 1.2817,
"step": 5275
},
{
"epoch": 0.9375805233462171,
"grad_norm": 0.3326654321721221,
"learning_rate": 3.9064718444514093e-07,
"loss": 1.2883,
"step": 5276
},
{
"epoch": 0.9377582300413168,
"grad_norm": 0.3377124918098523,
"learning_rate": 3.88431668409881e-07,
"loss": 1.2815,
"step": 5277
},
{
"epoch": 0.9379359367364165,
"grad_norm": 0.3419251134838801,
"learning_rate": 3.862223911833196e-07,
"loss": 1.3287,
"step": 5278
},
{
"epoch": 0.9381136434315163,
"grad_norm": 0.3290949949877623,
"learning_rate": 3.8401935346826793e-07,
"loss": 1.2791,
"step": 5279
},
{
"epoch": 0.938291350126616,
"grad_norm": 0.34570522426842093,
"learning_rate": 3.818225559655564e-07,
"loss": 1.296,
"step": 5280
},
{
"epoch": 0.9384690568217158,
"grad_norm": 0.3380830599659484,
"learning_rate": 3.7963199937402605e-07,
"loss": 1.3258,
"step": 5281
},
{
"epoch": 0.9386467635168155,
"grad_norm": 0.33449648736363413,
"learning_rate": 3.77447684390535e-07,
"loss": 1.2424,
"step": 5282
},
{
"epoch": 0.9388244702119153,
"grad_norm": 0.34191587812848734,
"learning_rate": 3.7526961170995413e-07,
"loss": 1.2943,
"step": 5283
},
{
"epoch": 0.9390021769070149,
"grad_norm": 0.33888578987320067,
"learning_rate": 3.730977820251669e-07,
"loss": 1.3088,
"step": 5284
},
{
"epoch": 0.9391798836021147,
"grad_norm": 0.3290096616812628,
"learning_rate": 3.709321960270784e-07,
"loss": 1.3073,
"step": 5285
},
{
"epoch": 0.9393575902972144,
"grad_norm": 0.3363876450745388,
"learning_rate": 3.6877285440459986e-07,
"loss": 1.2829,
"step": 5286
},
{
"epoch": 0.9395352969923142,
"grad_norm": 0.33695766940857935,
"learning_rate": 3.666197578446573e-07,
"loss": 1.3119,
"step": 5287
},
{
"epoch": 0.939713003687414,
"grad_norm": 0.33191708808771636,
"learning_rate": 3.6447290703219174e-07,
"loss": 1.2834,
"step": 5288
},
{
"epoch": 0.9398907103825137,
"grad_norm": 0.3423429179161675,
"learning_rate": 3.623323026501546e-07,
"loss": 1.3242,
"step": 5289
},
{
"epoch": 0.9400684170776133,
"grad_norm": 0.3339460417134832,
"learning_rate": 3.60197945379519e-07,
"loss": 1.3021,
"step": 5290
},
{
"epoch": 0.9402461237727131,
"grad_norm": 0.352970198624759,
"learning_rate": 3.5806983589925736e-07,
"loss": 1.3048,
"step": 5291
},
{
"epoch": 0.9404238304678129,
"grad_norm": 0.334287695956641,
"learning_rate": 3.559479748863659e-07,
"loss": 1.3064,
"step": 5292
},
{
"epoch": 0.9406015371629126,
"grad_norm": 0.34217998132215927,
"learning_rate": 3.538323630158469e-07,
"loss": 1.3136,
"step": 5293
},
{
"epoch": 0.9407792438580124,
"grad_norm": 0.3491227678436001,
"learning_rate": 3.517230009607131e-07,
"loss": 1.2793,
"step": 5294
},
{
"epoch": 0.9409569505531121,
"grad_norm": 0.3380591156015529,
"learning_rate": 3.4961988939199885e-07,
"loss": 1.3108,
"step": 5295
},
{
"epoch": 0.9411346572482119,
"grad_norm": 0.33903649336658115,
"learning_rate": 3.475230289787379e-07,
"loss": 1.3197,
"step": 5296
},
{
"epoch": 0.9413123639433115,
"grad_norm": 0.33557888328671176,
"learning_rate": 3.454324203879833e-07,
"loss": 1.3107,
"step": 5297
},
{
"epoch": 0.9414900706384113,
"grad_norm": 0.33447589343300843,
"learning_rate": 3.4334806428479416e-07,
"loss": 1.3124,
"step": 5298
},
{
"epoch": 0.941667777333511,
"grad_norm": 0.33585680832822423,
"learning_rate": 3.4126996133224677e-07,
"loss": 1.2942,
"step": 5299
},
{
"epoch": 0.9418454840286108,
"grad_norm": 0.3380014892958177,
"learning_rate": 3.3919811219142563e-07,
"loss": 1.2993,
"step": 5300
},
{
"epoch": 0.9420231907237105,
"grad_norm": 0.33679142330569756,
"learning_rate": 3.371325175214235e-07,
"loss": 1.2913,
"step": 5301
},
{
"epoch": 0.9422008974188103,
"grad_norm": 0.3273791683055824,
"learning_rate": 3.350731779793415e-07,
"loss": 1.278,
"step": 5302
},
{
"epoch": 0.94237860411391,
"grad_norm": 0.33448295577675363,
"learning_rate": 3.330200942202977e-07,
"loss": 1.2905,
"step": 5303
},
{
"epoch": 0.9425563108090097,
"grad_norm": 0.32726824123555587,
"learning_rate": 3.3097326689741637e-07,
"loss": 1.26,
"step": 5304
},
{
"epoch": 0.9427340175041095,
"grad_norm": 0.33176423264759586,
"learning_rate": 3.2893269666183227e-07,
"loss": 1.2972,
"step": 5305
},
{
"epoch": 0.9429117241992092,
"grad_norm": 0.3324214166514102,
"learning_rate": 3.2689838416268825e-07,
"loss": 1.3032,
"step": 5306
},
{
"epoch": 0.943089430894309,
"grad_norm": 0.34028972467802404,
"learning_rate": 3.2487033004713564e-07,
"loss": 1.3221,
"step": 5307
},
{
"epoch": 0.9432671375894087,
"grad_norm": 0.3356597188590468,
"learning_rate": 3.2284853496034275e-07,
"loss": 1.3208,
"step": 5308
},
{
"epoch": 0.9434448442845084,
"grad_norm": 0.35019242443031307,
"learning_rate": 3.208329995454729e-07,
"loss": 1.3592,
"step": 5309
},
{
"epoch": 0.9436225509796081,
"grad_norm": 0.33679682673911765,
"learning_rate": 3.188237244437109e-07,
"loss": 1.3337,
"step": 5310
},
{
"epoch": 0.9438002576747079,
"grad_norm": 0.33109773676518667,
"learning_rate": 3.1682071029424335e-07,
"loss": 1.2911,
"step": 5311
},
{
"epoch": 0.9439779643698076,
"grad_norm": 0.33800182483216845,
"learning_rate": 3.148239577342649e-07,
"loss": 1.3159,
"step": 5312
},
{
"epoch": 0.9441556710649074,
"grad_norm": 0.3327922003266528,
"learning_rate": 3.12833467398983e-07,
"loss": 1.2872,
"step": 5313
},
{
"epoch": 0.9443333777600071,
"grad_norm": 0.3395299618979198,
"learning_rate": 3.108492399216068e-07,
"loss": 1.3111,
"step": 5314
},
{
"epoch": 0.9445110844551069,
"grad_norm": 0.3314234798608234,
"learning_rate": 3.088712759333623e-07,
"loss": 1.2585,
"step": 5315
},
{
"epoch": 0.9446887911502065,
"grad_norm": 0.3316830778018896,
"learning_rate": 3.0689957606347075e-07,
"loss": 1.2899,
"step": 5316
},
{
"epoch": 0.9448664978453063,
"grad_norm": 0.33439238739266525,
"learning_rate": 3.049341409391704e-07,
"loss": 1.3049,
"step": 5317
},
{
"epoch": 0.945044204540406,
"grad_norm": 0.3353087860908845,
"learning_rate": 3.0297497118570107e-07,
"loss": 1.2856,
"step": 5318
},
{
"epoch": 0.9452219112355058,
"grad_norm": 0.33387179691977453,
"learning_rate": 3.010220674263131e-07,
"loss": 1.3009,
"step": 5319
},
{
"epoch": 0.9453996179306056,
"grad_norm": 0.3279358876582797,
"learning_rate": 2.990754302822629e-07,
"loss": 1.2634,
"step": 5320
},
{
"epoch": 0.9455773246257053,
"grad_norm": 0.332719949983515,
"learning_rate": 2.971350603728085e-07,
"loss": 1.2875,
"step": 5321
},
{
"epoch": 0.945755031320805,
"grad_norm": 0.3353226457514404,
"learning_rate": 2.9520095831522043e-07,
"loss": 1.2874,
"step": 5322
},
{
"epoch": 0.9459327380159047,
"grad_norm": 0.33964054342015243,
"learning_rate": 2.9327312472477553e-07,
"loss": 1.3295,
"step": 5323
},
{
"epoch": 0.9461104447110045,
"grad_norm": 0.3407162703988188,
"learning_rate": 2.9135156021474987e-07,
"loss": 1.3131,
"step": 5324
},
{
"epoch": 0.9462881514061042,
"grad_norm": 0.3317221677525906,
"learning_rate": 2.8943626539643e-07,
"loss": 1.269,
"step": 5325
},
{
"epoch": 0.946465858101204,
"grad_norm": 0.33794509287020225,
"learning_rate": 2.875272408791085e-07,
"loss": 1.3161,
"step": 5326
},
{
"epoch": 0.9466435647963037,
"grad_norm": 0.334804169105494,
"learning_rate": 2.8562448727008197e-07,
"loss": 1.2618,
"step": 5327
},
{
"epoch": 0.9468212714914035,
"grad_norm": 0.33744850319211117,
"learning_rate": 2.837280051746505e-07,
"loss": 1.325,
"step": 5328
},
{
"epoch": 0.9469989781865031,
"grad_norm": 0.3519072846845854,
"learning_rate": 2.8183779519612263e-07,
"loss": 1.2926,
"step": 5329
},
{
"epoch": 0.9471766848816029,
"grad_norm": 0.3331509073945191,
"learning_rate": 2.7995385793580854e-07,
"loss": 1.2595,
"step": 5330
},
{
"epoch": 0.9473543915767026,
"grad_norm": 0.33122590919533546,
"learning_rate": 2.780761939930221e-07,
"loss": 1.3113,
"step": 5331
},
{
"epoch": 0.9475320982718024,
"grad_norm": 0.3367754143591216,
"learning_rate": 2.7620480396508997e-07,
"loss": 1.3253,
"step": 5332
},
{
"epoch": 0.9477098049669022,
"grad_norm": 0.3373396241286046,
"learning_rate": 2.7433968844732926e-07,
"loss": 1.3345,
"step": 5333
},
{
"epoch": 0.9478875116620019,
"grad_norm": 0.33833942188435234,
"learning_rate": 2.7248084803307205e-07,
"loss": 1.3059,
"step": 5334
},
{
"epoch": 0.9480652183571016,
"grad_norm": 0.33310951193928057,
"learning_rate": 2.706282833136498e-07,
"loss": 1.2903,
"step": 5335
},
{
"epoch": 0.9482429250522013,
"grad_norm": 0.3341475083853235,
"learning_rate": 2.687819948783976e-07,
"loss": 1.2708,
"step": 5336
},
{
"epoch": 0.9484206317473011,
"grad_norm": 0.3647666944376384,
"learning_rate": 2.669419833146547e-07,
"loss": 1.3032,
"step": 5337
},
{
"epoch": 0.9485983384424008,
"grad_norm": 0.4459489142531167,
"learning_rate": 2.6510824920776614e-07,
"loss": 1.3278,
"step": 5338
},
{
"epoch": 0.9487760451375006,
"grad_norm": 0.3294529327678566,
"learning_rate": 2.632807931410741e-07,
"loss": 1.2638,
"step": 5339
},
{
"epoch": 0.9489537518326003,
"grad_norm": 0.32656115980550926,
"learning_rate": 2.614596156959248e-07,
"loss": 1.2544,
"step": 5340
},
{
"epoch": 0.9491314585277,
"grad_norm": 0.3391544364919122,
"learning_rate": 2.5964471745167473e-07,
"loss": 1.3175,
"step": 5341
},
{
"epoch": 0.9493091652227997,
"grad_norm": 0.33354262198820567,
"learning_rate": 2.578360989856732e-07,
"loss": 1.28,
"step": 5342
},
{
"epoch": 0.9494868719178995,
"grad_norm": 0.34311106805230995,
"learning_rate": 2.560337608732755e-07,
"loss": 1.3012,
"step": 5343
},
{
"epoch": 0.9496645786129992,
"grad_norm": 0.3298462037120003,
"learning_rate": 2.5423770368784296e-07,
"loss": 1.3031,
"step": 5344
},
{
"epoch": 0.949842285308099,
"grad_norm": 0.33889718388635426,
"learning_rate": 2.524479280007297e-07,
"loss": 1.3346,
"step": 5345
},
{
"epoch": 0.9500199920031988,
"grad_norm": 0.33111615375090214,
"learning_rate": 2.506644343813025e-07,
"loss": 1.3025,
"step": 5346
},
{
"epoch": 0.9501976986982985,
"grad_norm": 0.3361844457330211,
"learning_rate": 2.4888722339692084e-07,
"loss": 1.2972,
"step": 5347
},
{
"epoch": 0.9503754053933982,
"grad_norm": 0.3327184719196878,
"learning_rate": 2.4711629561294805e-07,
"loss": 1.3028,
"step": 5348
},
{
"epoch": 0.9505531120884979,
"grad_norm": 0.3343425236582637,
"learning_rate": 2.453516515927512e-07,
"loss": 1.3113,
"step": 5349
},
{
"epoch": 0.9507308187835977,
"grad_norm": 0.3324200432155975,
"learning_rate": 2.4359329189769907e-07,
"loss": 1.3042,
"step": 5350
},
{
"epoch": 0.9509085254786974,
"grad_norm": 0.32830100997270395,
"learning_rate": 2.4184121708715537e-07,
"loss": 1.2831,
"step": 5351
},
{
"epoch": 0.9510862321737972,
"grad_norm": 0.3346542265322023,
"learning_rate": 2.400954277184897e-07,
"loss": 1.282,
"step": 5352
},
{
"epoch": 0.9512639388688969,
"grad_norm": 0.3310792846126255,
"learning_rate": 2.3835592434707123e-07,
"loss": 1.2778,
"step": 5353
},
{
"epoch": 0.9514416455639966,
"grad_norm": 0.330308335697489,
"learning_rate": 2.3662270752626616e-07,
"loss": 1.2997,
"step": 5354
},
{
"epoch": 0.9516193522590963,
"grad_norm": 0.33424574596148676,
"learning_rate": 2.3489577780744676e-07,
"loss": 1.3056,
"step": 5355
},
{
"epoch": 0.9517970589541961,
"grad_norm": 0.34010184521757525,
"learning_rate": 2.3317513573997808e-07,
"loss": 1.3516,
"step": 5356
},
{
"epoch": 0.9519747656492958,
"grad_norm": 0.34360640319239094,
"learning_rate": 2.314607818712311e-07,
"loss": 1.3278,
"step": 5357
},
{
"epoch": 0.9521524723443956,
"grad_norm": 0.33941917082555495,
"learning_rate": 2.2975271674657186e-07,
"loss": 1.3197,
"step": 5358
},
{
"epoch": 0.9523301790394954,
"grad_norm": 0.33580678203716313,
"learning_rate": 2.2805094090937007e-07,
"loss": 1.2482,
"step": 5359
},
{
"epoch": 0.9525078857345951,
"grad_norm": 0.33022252518576717,
"learning_rate": 2.2635545490099498e-07,
"loss": 1.2566,
"step": 5360
},
{
"epoch": 0.9526855924296947,
"grad_norm": 0.3376545333061651,
"learning_rate": 2.2466625926080843e-07,
"loss": 1.3017,
"step": 5361
},
{
"epoch": 0.9528632991247945,
"grad_norm": 0.33215672279417213,
"learning_rate": 2.2298335452617614e-07,
"loss": 1.3091,
"step": 5362
},
{
"epoch": 0.9530410058198943,
"grad_norm": 0.7457935765974009,
"learning_rate": 2.21306741232461e-07,
"loss": 1.3323,
"step": 5363
},
{
"epoch": 0.953218712514994,
"grad_norm": 0.3380241144932053,
"learning_rate": 2.1963641991302963e-07,
"loss": 1.3074,
"step": 5364
},
{
"epoch": 0.9533964192100938,
"grad_norm": 0.3351363437691952,
"learning_rate": 2.1797239109923706e-07,
"loss": 1.3069,
"step": 5365
},
{
"epoch": 0.9535741259051935,
"grad_norm": 0.3462040895378065,
"learning_rate": 2.1631465532044427e-07,
"loss": 1.2992,
"step": 5366
},
{
"epoch": 0.9537518326002932,
"grad_norm": 0.3282108695401426,
"learning_rate": 2.1466321310401162e-07,
"loss": 1.2751,
"step": 5367
},
{
"epoch": 0.9539295392953929,
"grad_norm": 0.33017528913573135,
"learning_rate": 2.1301806497528777e-07,
"loss": 1.287,
"step": 5368
},
{
"epoch": 0.9541072459904927,
"grad_norm": 0.32683738120332007,
"learning_rate": 2.1137921145762964e-07,
"loss": 1.2423,
"step": 5369
},
{
"epoch": 0.9542849526855924,
"grad_norm": 0.33724730709379497,
"learning_rate": 2.0974665307238684e-07,
"loss": 1.3258,
"step": 5370
},
{
"epoch": 0.9544626593806922,
"grad_norm": 0.33596819986369136,
"learning_rate": 2.0812039033890397e-07,
"loss": 1.2721,
"step": 5371
},
{
"epoch": 0.954640366075792,
"grad_norm": 0.33653990562279457,
"learning_rate": 2.065004237745294e-07,
"loss": 1.2792,
"step": 5372
},
{
"epoch": 0.9548180727708916,
"grad_norm": 0.3406236193550991,
"learning_rate": 2.04886753894602e-07,
"loss": 1.2935,
"step": 5373
},
{
"epoch": 0.9549957794659913,
"grad_norm": 0.33594036861405197,
"learning_rate": 2.0327938121246449e-07,
"loss": 1.2961,
"step": 5374
},
{
"epoch": 0.9551734861610911,
"grad_norm": 0.3289328468331801,
"learning_rate": 2.0167830623944784e-07,
"loss": 1.2906,
"step": 5375
},
{
"epoch": 0.9553511928561909,
"grad_norm": 0.33534294019711297,
"learning_rate": 2.0008352948488906e-07,
"loss": 1.3239,
"step": 5376
},
{
"epoch": 0.9555288995512906,
"grad_norm": 0.33080122871154166,
"learning_rate": 1.9849505145611126e-07,
"loss": 1.2954,
"step": 5377
},
{
"epoch": 0.9557066062463904,
"grad_norm": 0.33527143322298897,
"learning_rate": 1.9691287265844127e-07,
"loss": 1.2754,
"step": 5378
},
{
"epoch": 0.9558843129414901,
"grad_norm": 0.3395203197859448,
"learning_rate": 1.9533699359520097e-07,
"loss": 1.3037,
"step": 5379
},
{
"epoch": 0.9560620196365898,
"grad_norm": 0.33959059897331006,
"learning_rate": 1.9376741476770488e-07,
"loss": 1.3439,
"step": 5380
},
{
"epoch": 0.9562397263316895,
"grad_norm": 0.33179284379370616,
"learning_rate": 1.9220413667526915e-07,
"loss": 1.2832,
"step": 5381
},
{
"epoch": 0.9564174330267893,
"grad_norm": 0.3363045440063136,
"learning_rate": 1.9064715981519821e-07,
"loss": 1.3037,
"step": 5382
},
{
"epoch": 0.956595139721889,
"grad_norm": 0.3395834966942603,
"learning_rate": 1.890964846828003e-07,
"loss": 1.319,
"step": 5383
},
{
"epoch": 0.9567728464169888,
"grad_norm": 0.3317783703337158,
"learning_rate": 1.8755211177136968e-07,
"loss": 1.2839,
"step": 5384
},
{
"epoch": 0.9569505531120885,
"grad_norm": 0.33700883379700824,
"learning_rate": 1.8601404157220226e-07,
"loss": 1.2822,
"step": 5385
},
{
"epoch": 0.9571282598071882,
"grad_norm": 0.33409885598182876,
"learning_rate": 1.8448227457458666e-07,
"loss": 1.3014,
"step": 5386
},
{
"epoch": 0.9573059665022879,
"grad_norm": 0.3355455533869002,
"learning_rate": 1.8295681126580645e-07,
"loss": 1.3189,
"step": 5387
},
{
"epoch": 0.9574836731973877,
"grad_norm": 0.3275030887075174,
"learning_rate": 1.8143765213114007e-07,
"loss": 1.2833,
"step": 5388
},
{
"epoch": 0.9576613798924875,
"grad_norm": 0.33283730156491004,
"learning_rate": 1.7992479765386316e-07,
"loss": 1.2882,
"step": 5389
},
{
"epoch": 0.9578390865875872,
"grad_norm": 0.33166920301523045,
"learning_rate": 1.784182483152419e-07,
"loss": 1.311,
"step": 5390
},
{
"epoch": 0.958016793282687,
"grad_norm": 0.3327900004619429,
"learning_rate": 1.769180045945351e-07,
"loss": 1.2606,
"step": 5391
},
{
"epoch": 0.9581944999777867,
"grad_norm": 0.3494186822296372,
"learning_rate": 1.7542406696900328e-07,
"loss": 1.3356,
"step": 5392
},
{
"epoch": 0.9583722066728864,
"grad_norm": 0.3368617402397969,
"learning_rate": 1.7393643591389288e-07,
"loss": 1.2989,
"step": 5393
},
{
"epoch": 0.9585499133679861,
"grad_norm": 0.3293906433290475,
"learning_rate": 1.7245511190244756e-07,
"loss": 1.2804,
"step": 5394
},
{
"epoch": 0.9587276200630859,
"grad_norm": 0.335145078155273,
"learning_rate": 1.7098009540590376e-07,
"loss": 1.2853,
"step": 5395
},
{
"epoch": 0.9589053267581856,
"grad_norm": 0.3342358111820173,
"learning_rate": 1.69511386893495e-07,
"loss": 1.2962,
"step": 5396
},
{
"epoch": 0.9590830334532854,
"grad_norm": 0.33643669177156377,
"learning_rate": 1.680489868324431e-07,
"loss": 1.282,
"step": 5397
},
{
"epoch": 0.9592607401483851,
"grad_norm": 0.3332196065178126,
"learning_rate": 1.6659289568796255e-07,
"loss": 1.3122,
"step": 5398
},
{
"epoch": 0.9594384468434848,
"grad_norm": 0.3415816328300186,
"learning_rate": 1.6514311392326954e-07,
"loss": 1.3296,
"step": 5399
},
{
"epoch": 0.9596161535385845,
"grad_norm": 0.33506073361109456,
"learning_rate": 1.6369964199956178e-07,
"loss": 1.2922,
"step": 5400
},
{
"epoch": 0.9597938602336843,
"grad_norm": 0.3292968573630102,
"learning_rate": 1.622624803760342e-07,
"loss": 1.2862,
"step": 5401
},
{
"epoch": 0.959971566928784,
"grad_norm": 0.32943957713724065,
"learning_rate": 1.6083162950987884e-07,
"loss": 1.2853,
"step": 5402
},
{
"epoch": 0.9601492736238838,
"grad_norm": 0.3407050189194684,
"learning_rate": 1.5940708985627606e-07,
"loss": 1.3256,
"step": 5403
},
{
"epoch": 0.9603269803189836,
"grad_norm": 0.34081834549741674,
"learning_rate": 1.5798886186839445e-07,
"loss": 1.3097,
"step": 5404
},
{
"epoch": 0.9605046870140832,
"grad_norm": 0.33394879515565296,
"learning_rate": 1.5657694599740424e-07,
"loss": 1.3295,
"step": 5405
},
{
"epoch": 0.960682393709183,
"grad_norm": 0.3413759594301907,
"learning_rate": 1.5517134269245727e-07,
"loss": 1.3178,
"step": 5406
},
{
"epoch": 0.9608601004042827,
"grad_norm": 0.3339957264717283,
"learning_rate": 1.5377205240070692e-07,
"loss": 1.2992,
"step": 5407
},
{
"epoch": 0.9610378070993825,
"grad_norm": 0.3351725774982221,
"learning_rate": 1.5237907556729047e-07,
"loss": 1.292,
"step": 5408
},
{
"epoch": 0.9612155137944822,
"grad_norm": 0.3384415410294281,
"learning_rate": 1.5099241263534236e-07,
"loss": 1.2846,
"step": 5409
},
{
"epoch": 0.961393220489582,
"grad_norm": 0.32825914789203064,
"learning_rate": 1.4961206404598306e-07,
"loss": 1.2508,
"step": 5410
},
{
"epoch": 0.9615709271846817,
"grad_norm": 0.3332888882329536,
"learning_rate": 1.4823803023833017e-07,
"loss": 1.2738,
"step": 5411
},
{
"epoch": 0.9617486338797814,
"grad_norm": 0.34414749313140414,
"learning_rate": 1.4687031164948962e-07,
"loss": 1.3589,
"step": 5412
},
{
"epoch": 0.9619263405748811,
"grad_norm": 0.33221655680847106,
"learning_rate": 1.455089087145578e-07,
"loss": 1.3317,
"step": 5413
},
{
"epoch": 0.9621040472699809,
"grad_norm": 0.3369303093338457,
"learning_rate": 1.4415382186662386e-07,
"loss": 1.2874,
"step": 5414
},
{
"epoch": 0.9622817539650806,
"grad_norm": 0.3373898606115465,
"learning_rate": 1.4280505153676294e-07,
"loss": 1.3057,
"step": 5415
},
{
"epoch": 0.9624594606601804,
"grad_norm": 0.3409579382692791,
"learning_rate": 1.4146259815404962e-07,
"loss": 1.2925,
"step": 5416
},
{
"epoch": 0.9626371673552802,
"grad_norm": 0.3391811701760967,
"learning_rate": 1.401264621455378e-07,
"loss": 1.33,
"step": 5417
},
{
"epoch": 0.9628148740503798,
"grad_norm": 0.3350469361183545,
"learning_rate": 1.387966439362809e-07,
"loss": 1.2963,
"step": 5418
},
{
"epoch": 0.9629925807454796,
"grad_norm": 0.332028027251155,
"learning_rate": 1.374731439493182e-07,
"loss": 1.2758,
"step": 5419
},
{
"epoch": 0.9631702874405793,
"grad_norm": 0.3413757338526801,
"learning_rate": 1.3615596260567743e-07,
"loss": 1.3299,
"step": 5420
},
{
"epoch": 0.9633479941356791,
"grad_norm": 0.33785773530835034,
"learning_rate": 1.348451003243856e-07,
"loss": 1.2938,
"step": 5421
},
{
"epoch": 0.9635257008307788,
"grad_norm": 0.3402198060451595,
"learning_rate": 1.3354055752244688e-07,
"loss": 1.3174,
"step": 5422
},
{
"epoch": 0.9637034075258786,
"grad_norm": 0.3320578219686115,
"learning_rate": 1.3224233461486047e-07,
"loss": 1.2561,
"step": 5423
},
{
"epoch": 0.9638811142209783,
"grad_norm": 0.3366184059375613,
"learning_rate": 1.3095043201461822e-07,
"loss": 1.2802,
"step": 5424
},
{
"epoch": 0.964058820916078,
"grad_norm": 0.328543583729006,
"learning_rate": 1.2966485013269804e-07,
"loss": 1.2385,
"step": 5425
},
{
"epoch": 0.9642365276111777,
"grad_norm": 0.3316808277773567,
"learning_rate": 1.2838558937806833e-07,
"loss": 1.2885,
"step": 5426
},
{
"epoch": 0.9644142343062775,
"grad_norm": 0.3315045380084273,
"learning_rate": 1.271126501576858e-07,
"loss": 1.2986,
"step": 5427
},
{
"epoch": 0.9645919410013772,
"grad_norm": 0.3392580089895331,
"learning_rate": 1.2584603287649321e-07,
"loss": 1.3135,
"step": 5428
},
{
"epoch": 0.964769647696477,
"grad_norm": 0.33149015098302187,
"learning_rate": 1.2458573793743045e-07,
"loss": 1.2933,
"step": 5429
},
{
"epoch": 0.9649473543915768,
"grad_norm": 0.33178808783708824,
"learning_rate": 1.233317657414168e-07,
"loss": 1.2872,
"step": 5430
},
{
"epoch": 0.9651250610866764,
"grad_norm": 0.33797297541554683,
"learning_rate": 1.2208411668736652e-07,
"loss": 1.318,
"step": 5431
},
{
"epoch": 0.9653027677817761,
"grad_norm": 0.3379926099566494,
"learning_rate": 1.208427911721799e-07,
"loss": 1.298,
"step": 5432
},
{
"epoch": 0.9654804744768759,
"grad_norm": 0.3373301563374417,
"learning_rate": 1.1960778959074547e-07,
"loss": 1.3254,
"step": 5433
},
{
"epoch": 0.9656581811719757,
"grad_norm": 0.3339198184452578,
"learning_rate": 1.1837911233593791e-07,
"loss": 1.2724,
"step": 5434
},
{
"epoch": 0.9658358878670754,
"grad_norm": 0.33669460454127126,
"learning_rate": 1.1715675979862895e-07,
"loss": 1.3325,
"step": 5435
},
{
"epoch": 0.9660135945621752,
"grad_norm": 0.335276734985783,
"learning_rate": 1.1594073236766757e-07,
"loss": 1.3113,
"step": 5436
},
{
"epoch": 0.9661913012572748,
"grad_norm": 0.336295637886606,
"learning_rate": 1.1473103042989764e-07,
"loss": 1.2984,
"step": 5437
},
{
"epoch": 0.9663690079523746,
"grad_norm": 0.3317316274026581,
"learning_rate": 1.1352765437014246e-07,
"loss": 1.2946,
"step": 5438
},
{
"epoch": 0.9665467146474743,
"grad_norm": 0.3446751709451749,
"learning_rate": 1.123306045712247e-07,
"loss": 1.3597,
"step": 5439
},
{
"epoch": 0.9667244213425741,
"grad_norm": 0.3368891718052307,
"learning_rate": 1.1113988141394416e-07,
"loss": 1.3246,
"step": 5440
},
{
"epoch": 0.9669021280376738,
"grad_norm": 0.3455178272720978,
"learning_rate": 1.0995548527709565e-07,
"loss": 1.3254,
"step": 5441
},
{
"epoch": 0.9670798347327736,
"grad_norm": 0.3325045173084782,
"learning_rate": 1.0877741653745554e-07,
"loss": 1.2961,
"step": 5442
},
{
"epoch": 0.9672575414278733,
"grad_norm": 0.3331173862442826,
"learning_rate": 1.0760567556979295e-07,
"loss": 1.2721,
"step": 5443
},
{
"epoch": 0.967435248122973,
"grad_norm": 0.33806433183517737,
"learning_rate": 1.0644026274685638e-07,
"loss": 1.3224,
"step": 5444
},
{
"epoch": 0.9676129548180727,
"grad_norm": 0.3302660154019278,
"learning_rate": 1.0528117843938701e-07,
"loss": 1.2712,
"step": 5445
},
{
"epoch": 0.9677906615131725,
"grad_norm": 0.3321388656833224,
"learning_rate": 1.0412842301611215e-07,
"loss": 1.314,
"step": 5446
},
{
"epoch": 0.9679683682082723,
"grad_norm": 0.332617891998742,
"learning_rate": 1.029819968437451e-07,
"loss": 1.2749,
"step": 5447
},
{
"epoch": 0.968146074903372,
"grad_norm": 0.33896413170439227,
"learning_rate": 1.0184190028698305e-07,
"loss": 1.3016,
"step": 5448
},
{
"epoch": 0.9683237815984718,
"grad_norm": 0.33615074046941207,
"learning_rate": 1.0070813370851585e-07,
"loss": 1.3114,
"step": 5449
},
{
"epoch": 0.9685014882935714,
"grad_norm": 0.3358921663186439,
"learning_rate": 9.9580697469015e-08,
"loss": 1.2992,
"step": 5450
},
{
"epoch": 0.9686791949886712,
"grad_norm": 0.3326395576602177,
"learning_rate": 9.845959192713583e-08,
"loss": 1.2795,
"step": 5451
},
{
"epoch": 0.9688569016837709,
"grad_norm": 0.3373089484627442,
"learning_rate": 9.734481743952861e-08,
"loss": 1.3006,
"step": 5452
},
{
"epoch": 0.9690346083788707,
"grad_norm": 0.33847321441893696,
"learning_rate": 9.623637436082078e-08,
"loss": 1.3031,
"step": 5453
},
{
"epoch": 0.9692123150739704,
"grad_norm": 0.3417744839990038,
"learning_rate": 9.513426304362804e-08,
"loss": 1.3327,
"step": 5454
},
{
"epoch": 0.9693900217690702,
"grad_norm": 0.33430671839566883,
"learning_rate": 9.40384838385544e-08,
"loss": 1.3053,
"step": 5455
},
{
"epoch": 0.9695677284641699,
"grad_norm": 0.34006379638154577,
"learning_rate": 9.294903709418768e-08,
"loss": 1.3393,
"step": 5456
},
{
"epoch": 0.9697454351592696,
"grad_norm": 0.33597619561779085,
"learning_rate": 9.186592315710175e-08,
"loss": 1.2883,
"step": 5457
},
{
"epoch": 0.9699231418543693,
"grad_norm": 0.3270349270023118,
"learning_rate": 9.078914237185432e-08,
"loss": 1.2823,
"step": 5458
},
{
"epoch": 0.9701008485494691,
"grad_norm": 0.3362791538056177,
"learning_rate": 8.97186950809914e-08,
"loss": 1.3273,
"step": 5459
},
{
"epoch": 0.9702785552445689,
"grad_norm": 0.3391275154764027,
"learning_rate": 8.865458162504059e-08,
"loss": 1.3248,
"step": 5460
},
{
"epoch": 0.9704562619396686,
"grad_norm": 0.3326116947512282,
"learning_rate": 8.759680234251556e-08,
"loss": 1.3005,
"step": 5461
},
{
"epoch": 0.9706339686347684,
"grad_norm": 0.33666946483418986,
"learning_rate": 8.654535756991821e-08,
"loss": 1.3293,
"step": 5462
},
{
"epoch": 0.970811675329868,
"grad_norm": 0.33740591815669274,
"learning_rate": 8.550024764173215e-08,
"loss": 1.3203,
"step": 5463
},
{
"epoch": 0.9709893820249678,
"grad_norm": 0.3349140615956207,
"learning_rate": 8.446147289042694e-08,
"loss": 1.3006,
"step": 5464
},
{
"epoch": 0.9711670887200675,
"grad_norm": 0.33874296719561303,
"learning_rate": 8.342903364645382e-08,
"loss": 1.3338,
"step": 5465
},
{
"epoch": 0.9713447954151673,
"grad_norm": 0.347925021577498,
"learning_rate": 8.240293023825452e-08,
"loss": 1.3484,
"step": 5466
},
{
"epoch": 0.971522502110267,
"grad_norm": 0.3441172032935066,
"learning_rate": 8.138316299225013e-08,
"loss": 1.3601,
"step": 5467
},
{
"epoch": 0.9717002088053668,
"grad_norm": 0.3316851614974511,
"learning_rate": 8.036973223284783e-08,
"loss": 1.2971,
"step": 5468
},
{
"epoch": 0.9718779155004664,
"grad_norm": 0.3327525541733529,
"learning_rate": 7.936263828243861e-08,
"loss": 1.2931,
"step": 5469
},
{
"epoch": 0.9720556221955662,
"grad_norm": 0.32983939617089825,
"learning_rate": 7.836188146139956e-08,
"loss": 1.2876,
"step": 5470
},
{
"epoch": 0.9722333288906659,
"grad_norm": 0.33626364964631983,
"learning_rate": 7.736746208808932e-08,
"loss": 1.2989,
"step": 5471
},
{
"epoch": 0.9724110355857657,
"grad_norm": 0.3286820834687168,
"learning_rate": 7.637938047885041e-08,
"loss": 1.2544,
"step": 5472
},
{
"epoch": 0.9725887422808654,
"grad_norm": 0.3504225531462737,
"learning_rate": 7.539763694801139e-08,
"loss": 1.3098,
"step": 5473
},
{
"epoch": 0.9727664489759652,
"grad_norm": 0.3384509660428817,
"learning_rate": 7.442223180788465e-08,
"loss": 1.3505,
"step": 5474
},
{
"epoch": 0.972944155671065,
"grad_norm": 0.3355878188891053,
"learning_rate": 7.345316536876202e-08,
"loss": 1.299,
"step": 5475
},
{
"epoch": 0.9731218623661646,
"grad_norm": 0.3338709797455537,
"learning_rate": 7.24904379389213e-08,
"loss": 1.2922,
"step": 5476
},
{
"epoch": 0.9732995690612644,
"grad_norm": 0.3361998311078078,
"learning_rate": 7.153404982462864e-08,
"loss": 1.3191,
"step": 5477
},
{
"epoch": 0.9734772757563641,
"grad_norm": 0.33655276248061883,
"learning_rate": 7.05840013301251e-08,
"loss": 1.3226,
"step": 5478
},
{
"epoch": 0.9736549824514639,
"grad_norm": 0.3290054576864593,
"learning_rate": 6.964029275764006e-08,
"loss": 1.2914,
"step": 5479
},
{
"epoch": 0.9738326891465636,
"grad_norm": 0.3936778479373302,
"learning_rate": 6.870292440738446e-08,
"loss": 1.2671,
"step": 5480
},
{
"epoch": 0.9740103958416634,
"grad_norm": 0.3369558952406521,
"learning_rate": 6.777189657755534e-08,
"loss": 1.2873,
"step": 5481
},
{
"epoch": 0.974188102536763,
"grad_norm": 0.33123224234168125,
"learning_rate": 6.684720956432689e-08,
"loss": 1.2678,
"step": 5482
},
{
"epoch": 0.9743658092318628,
"grad_norm": 0.3330404745900061,
"learning_rate": 6.592886366186158e-08,
"loss": 1.2814,
"step": 5483
},
{
"epoch": 0.9745435159269625,
"grad_norm": 0.3344681472935706,
"learning_rate": 6.501685916230128e-08,
"loss": 1.318,
"step": 5484
},
{
"epoch": 0.9747212226220623,
"grad_norm": 0.337812257756538,
"learning_rate": 6.41111963557739e-08,
"loss": 1.2889,
"step": 5485
},
{
"epoch": 0.974898929317162,
"grad_norm": 0.3301149795262486,
"learning_rate": 6.321187553038455e-08,
"loss": 1.293,
"step": 5486
},
{
"epoch": 0.9750766360122618,
"grad_norm": 0.32719083155617334,
"learning_rate": 6.23188969722266e-08,
"loss": 1.2812,
"step": 5487
},
{
"epoch": 0.9752543427073616,
"grad_norm": 0.33287140985433816,
"learning_rate": 6.143226096537058e-08,
"loss": 1.3015,
"step": 5488
},
{
"epoch": 0.9754320494024612,
"grad_norm": 0.3391701955894426,
"learning_rate": 6.055196779187534e-08,
"loss": 1.3161,
"step": 5489
},
{
"epoch": 0.975609756097561,
"grad_norm": 0.3358753685135224,
"learning_rate": 5.967801773177684e-08,
"loss": 1.3018,
"step": 5490
},
{
"epoch": 0.9757874627926607,
"grad_norm": 0.3367128132940395,
"learning_rate": 5.881041106309715e-08,
"loss": 1.3252,
"step": 5491
},
{
"epoch": 0.9759651694877605,
"grad_norm": 0.3304377751612406,
"learning_rate": 5.794914806183549e-08,
"loss": 1.2652,
"step": 5492
},
{
"epoch": 0.9761428761828602,
"grad_norm": 0.3352607218018879,
"learning_rate": 5.709422900197714e-08,
"loss": 1.2986,
"step": 5493
},
{
"epoch": 0.97632058287796,
"grad_norm": 0.3354054577305746,
"learning_rate": 5.6245654155486776e-08,
"loss": 1.3312,
"step": 5494
},
{
"epoch": 0.9764982895730596,
"grad_norm": 0.34991782186606724,
"learning_rate": 5.5403423792312894e-08,
"loss": 1.3186,
"step": 5495
},
{
"epoch": 0.9766759962681594,
"grad_norm": 0.33457653651966157,
"learning_rate": 5.4567538180385626e-08,
"loss": 1.3212,
"step": 5496
},
{
"epoch": 0.9768537029632591,
"grad_norm": 0.32913042013512733,
"learning_rate": 5.3737997585616706e-08,
"loss": 1.298,
"step": 5497
},
{
"epoch": 0.9770314096583589,
"grad_norm": 0.3314647508962754,
"learning_rate": 5.291480227189505e-08,
"loss": 1.269,
"step": 5498
},
{
"epoch": 0.9772091163534586,
"grad_norm": 0.37550324295989174,
"learning_rate": 5.209795250109562e-08,
"loss": 1.2732,
"step": 5499
},
{
"epoch": 0.9773868230485584,
"grad_norm": 0.33712717332587105,
"learning_rate": 5.128744853307721e-08,
"loss": 1.291,
"step": 5500
},
{
"epoch": 0.977564529743658,
"grad_norm": 0.3285974097091128,
"learning_rate": 5.0483290625671364e-08,
"loss": 1.2527,
"step": 5501
},
{
"epoch": 0.9777422364387578,
"grad_norm": 0.3394711576141235,
"learning_rate": 4.968547903470011e-08,
"loss": 1.3218,
"step": 5502
},
{
"epoch": 0.9779199431338575,
"grad_norm": 0.3383779653665115,
"learning_rate": 4.889401401396043e-08,
"loss": 1.3248,
"step": 5503
},
{
"epoch": 0.9780976498289573,
"grad_norm": 0.3332201684933822,
"learning_rate": 4.810889581523093e-08,
"loss": 1.299,
"step": 5504
},
{
"epoch": 0.9782753565240571,
"grad_norm": 0.33484954251167376,
"learning_rate": 4.733012468827625e-08,
"loss": 1.2825,
"step": 5505
},
{
"epoch": 0.9784530632191568,
"grad_norm": 0.32986678526246915,
"learning_rate": 4.655770088083378e-08,
"loss": 1.2801,
"step": 5506
},
{
"epoch": 0.9786307699142566,
"grad_norm": 0.33932175075649096,
"learning_rate": 4.5791624638626966e-08,
"loss": 1.332,
"step": 5507
},
{
"epoch": 0.9788084766093562,
"grad_norm": 0.35584269986907746,
"learning_rate": 4.503189620536086e-08,
"loss": 1.3399,
"step": 5508
},
{
"epoch": 0.978986183304456,
"grad_norm": 0.3328831259348096,
"learning_rate": 4.4278515822719915e-08,
"loss": 1.2899,
"step": 5509
},
{
"epoch": 0.9791638899995557,
"grad_norm": 0.3353546495844903,
"learning_rate": 4.3531483730367976e-08,
"loss": 1.2897,
"step": 5510
},
{
"epoch": 0.9793415966946555,
"grad_norm": 0.32802326052137515,
"learning_rate": 4.279080016594828e-08,
"loss": 1.2728,
"step": 5511
},
{
"epoch": 0.9795193033897552,
"grad_norm": 0.33448861210925984,
"learning_rate": 4.2056465365085675e-08,
"loss": 1.3177,
"step": 5512
},
{
"epoch": 0.979697010084855,
"grad_norm": 0.332002358870622,
"learning_rate": 4.1328479561388855e-08,
"loss": 1.278,
"step": 5513
},
{
"epoch": 0.9798747167799546,
"grad_norm": 0.34810219157581457,
"learning_rate": 4.0606842986441465e-08,
"loss": 1.3072,
"step": 5514
},
{
"epoch": 0.9800524234750544,
"grad_norm": 0.32780409222417123,
"learning_rate": 3.989155586981097e-08,
"loss": 1.3021,
"step": 5515
},
{
"epoch": 0.9802301301701541,
"grad_norm": 0.3414522926433442,
"learning_rate": 3.9182618439044253e-08,
"loss": 1.3679,
"step": 5516
},
{
"epoch": 0.9804078368652539,
"grad_norm": 0.34332927785934064,
"learning_rate": 3.848003091966535e-08,
"loss": 1.3161,
"step": 5517
},
{
"epoch": 0.9805855435603537,
"grad_norm": 0.33236748708803787,
"learning_rate": 3.778379353518214e-08,
"loss": 1.2724,
"step": 5518
},
{
"epoch": 0.9807632502554534,
"grad_norm": 0.33452254550441124,
"learning_rate": 3.709390650708189e-08,
"loss": 1.2955,
"step": 5519
},
{
"epoch": 0.9809409569505532,
"grad_norm": 0.34510070616931116,
"learning_rate": 3.641037005482906e-08,
"loss": 1.343,
"step": 5520
},
{
"epoch": 0.9811186636456528,
"grad_norm": 0.3299536358805788,
"learning_rate": 3.5733184395867484e-08,
"loss": 1.2714,
"step": 5521
},
{
"epoch": 0.9812963703407526,
"grad_norm": 0.33433186740874976,
"learning_rate": 3.506234974562706e-08,
"loss": 1.3126,
"step": 5522
},
{
"epoch": 0.9814740770358523,
"grad_norm": 0.34172303023060296,
"learning_rate": 3.439786631750819e-08,
"loss": 1.2842,
"step": 5523
},
{
"epoch": 0.9816517837309521,
"grad_norm": 0.3610909829155115,
"learning_rate": 3.3739734322899565e-08,
"loss": 1.2801,
"step": 5524
},
{
"epoch": 0.9818294904260518,
"grad_norm": 0.3299613263182237,
"learning_rate": 3.308795397116482e-08,
"loss": 1.2603,
"step": 5525
},
{
"epoch": 0.9820071971211516,
"grad_norm": 0.3422836341405969,
"learning_rate": 3.244252546964699e-08,
"loss": 1.3213,
"step": 5526
},
{
"epoch": 0.9821849038162512,
"grad_norm": 0.33327171798432637,
"learning_rate": 3.180344902366628e-08,
"loss": 1.307,
"step": 5527
},
{
"epoch": 0.982362610511351,
"grad_norm": 0.3328758421141553,
"learning_rate": 3.1170724836528944e-08,
"loss": 1.2695,
"step": 5528
},
{
"epoch": 0.9825403172064507,
"grad_norm": 0.3769113729689146,
"learning_rate": 3.054435310951398e-08,
"loss": 1.2869,
"step": 5529
},
{
"epoch": 0.9827180239015505,
"grad_norm": 0.33692186121422363,
"learning_rate": 2.9924334041882e-08,
"loss": 1.3068,
"step": 5530
},
{
"epoch": 0.9828957305966503,
"grad_norm": 0.33650265603991925,
"learning_rate": 2.9310667830875218e-08,
"loss": 1.3134,
"step": 5531
},
{
"epoch": 0.98307343729175,
"grad_norm": 0.33188285987256877,
"learning_rate": 2.8703354671708595e-08,
"loss": 1.2776,
"step": 5532
},
{
"epoch": 0.9832511439868497,
"grad_norm": 0.33756575005832656,
"learning_rate": 2.810239475758314e-08,
"loss": 1.2884,
"step": 5533
},
{
"epoch": 0.9834288506819494,
"grad_norm": 0.3319976096469046,
"learning_rate": 2.750778827967482e-08,
"loss": 1.2811,
"step": 5534
},
{
"epoch": 0.9836065573770492,
"grad_norm": 0.3391952674963217,
"learning_rate": 2.6919535427138988e-08,
"loss": 1.2865,
"step": 5535
},
{
"epoch": 0.9837842640721489,
"grad_norm": 0.33135369899155037,
"learning_rate": 2.633763638710818e-08,
"loss": 1.2932,
"step": 5536
},
{
"epoch": 0.9839619707672487,
"grad_norm": 0.3236997442278936,
"learning_rate": 2.576209134469654e-08,
"loss": 1.2246,
"step": 5537
},
{
"epoch": 0.9841396774623484,
"grad_norm": 0.3319862988494194,
"learning_rate": 2.5192900482997606e-08,
"loss": 1.2809,
"step": 5538
},
{
"epoch": 0.9843173841574482,
"grad_norm": 0.33525873381825516,
"learning_rate": 2.463006398307988e-08,
"loss": 1.2921,
"step": 5539
},
{
"epoch": 0.9844950908525478,
"grad_norm": 0.33256548132500036,
"learning_rate": 2.407358202399124e-08,
"loss": 1.2979,
"step": 5540
},
{
"epoch": 0.9846727975476476,
"grad_norm": 0.34503330503621443,
"learning_rate": 2.352345478276119e-08,
"loss": 1.3062,
"step": 5541
},
{
"epoch": 0.9848505042427473,
"grad_norm": 0.3418211847848679,
"learning_rate": 2.297968243439419e-08,
"loss": 1.3149,
"step": 5542
},
{
"epoch": 0.9850282109378471,
"grad_norm": 0.3452129260993286,
"learning_rate": 2.2442265151876308e-08,
"loss": 1.3389,
"step": 5543
},
{
"epoch": 0.9852059176329468,
"grad_norm": 0.3319408533130281,
"learning_rate": 2.1911203106168567e-08,
"loss": 1.2942,
"step": 5544
},
{
"epoch": 0.9853836243280466,
"grad_norm": 0.3449612322229831,
"learning_rate": 2.138649646621138e-08,
"loss": 1.2869,
"step": 5545
},
{
"epoch": 0.9855613310231462,
"grad_norm": 0.3288789244407863,
"learning_rate": 2.0868145398922344e-08,
"loss": 1.2704,
"step": 5546
},
{
"epoch": 0.985739037718246,
"grad_norm": 0.3325328922511324,
"learning_rate": 2.0356150069202885e-08,
"loss": 1.307,
"step": 5547
},
{
"epoch": 0.9859167444133458,
"grad_norm": 0.3334476495808856,
"learning_rate": 1.9850510639927158e-08,
"loss": 1.2823,
"step": 5548
},
{
"epoch": 0.9860944511084455,
"grad_norm": 0.33879744099335635,
"learning_rate": 1.9351227271946494e-08,
"loss": 1.3209,
"step": 5549
},
{
"epoch": 0.9862721578035453,
"grad_norm": 0.3346800994282828,
"learning_rate": 1.8858300124091623e-08,
"loss": 1.2897,
"step": 5550
},
{
"epoch": 0.986449864498645,
"grad_norm": 0.33734498017922665,
"learning_rate": 1.8371729353174884e-08,
"loss": 1.3129,
"step": 5551
},
{
"epoch": 0.9866275711937448,
"grad_norm": 0.3345560686604493,
"learning_rate": 1.789151511398357e-08,
"loss": 1.2972,
"step": 5552
},
{
"epoch": 0.9868052778888444,
"grad_norm": 0.3362552287939946,
"learning_rate": 1.7417657559282154e-08,
"loss": 1.3059,
"step": 5553
},
{
"epoch": 0.9869829845839442,
"grad_norm": 0.35067487929225155,
"learning_rate": 1.6950156839812272e-08,
"loss": 1.3303,
"step": 5554
},
{
"epoch": 0.9871606912790439,
"grad_norm": 0.33371944506577755,
"learning_rate": 1.648901310429496e-08,
"loss": 1.3043,
"step": 5555
},
{
"epoch": 0.9873383979741437,
"grad_norm": 0.3331758562998296,
"learning_rate": 1.603422649942843e-08,
"loss": 1.2998,
"step": 5556
},
{
"epoch": 0.9875161046692434,
"grad_norm": 0.3349133337136182,
"learning_rate": 1.55857971698925e-08,
"loss": 1.3061,
"step": 5557
},
{
"epoch": 0.9876938113643432,
"grad_norm": 0.33703989318676,
"learning_rate": 1.5143725258337516e-08,
"loss": 1.2875,
"step": 5558
},
{
"epoch": 0.9878715180594428,
"grad_norm": 0.33995535887202005,
"learning_rate": 1.47080109053932e-08,
"loss": 1.2878,
"step": 5559
},
{
"epoch": 0.9880492247545426,
"grad_norm": 0.3352135729890313,
"learning_rate": 1.4278654249673118e-08,
"loss": 1.2957,
"step": 5560
},
{
"epoch": 0.9882269314496424,
"grad_norm": 0.332983293429903,
"learning_rate": 1.385565542776135e-08,
"loss": 1.2991,
"step": 5561
},
{
"epoch": 0.9884046381447421,
"grad_norm": 0.33405046060739696,
"learning_rate": 1.3439014574221365e-08,
"loss": 1.3025,
"step": 5562
},
{
"epoch": 0.9885823448398419,
"grad_norm": 0.3370012177746043,
"learning_rate": 1.302873182159603e-08,
"loss": 1.3072,
"step": 5563
},
{
"epoch": 0.9887600515349416,
"grad_norm": 0.33510100405903104,
"learning_rate": 1.2624807300403163e-08,
"loss": 1.3268,
"step": 5564
},
{
"epoch": 0.9889377582300413,
"grad_norm": 0.32957897465763736,
"learning_rate": 1.2227241139137758e-08,
"loss": 1.2816,
"step": 5565
},
{
"epoch": 0.989115464925141,
"grad_norm": 0.33064766525449285,
"learning_rate": 1.1836033464271978e-08,
"loss": 1.2714,
"step": 5566
},
{
"epoch": 0.9892931716202408,
"grad_norm": 0.3328892110605918,
"learning_rate": 1.1451184400261828e-08,
"loss": 1.3082,
"step": 5567
},
{
"epoch": 0.9894708783153405,
"grad_norm": 0.3401175708954232,
"learning_rate": 1.1072694069529377e-08,
"loss": 1.3246,
"step": 5568
},
{
"epoch": 0.9896485850104403,
"grad_norm": 0.3295430954153633,
"learning_rate": 1.0700562592480535e-08,
"loss": 1.2848,
"step": 5569
},
{
"epoch": 0.98982629170554,
"grad_norm": 0.33775847071635434,
"learning_rate": 1.0334790087500602e-08,
"loss": 1.3056,
"step": 5570
},
{
"epoch": 0.9900039984006398,
"grad_norm": 0.3357244109353607,
"learning_rate": 9.975376670945391e-09,
"loss": 1.3131,
"step": 5571
},
{
"epoch": 0.9901817050957394,
"grad_norm": 0.3545317514702473,
"learning_rate": 9.622322457152334e-09,
"loss": 1.3107,
"step": 5572
},
{
"epoch": 0.9903594117908392,
"grad_norm": 0.3408421730396278,
"learning_rate": 9.275627558436029e-09,
"loss": 1.3154,
"step": 5573
},
{
"epoch": 0.990537118485939,
"grad_norm": 0.3394148605531864,
"learning_rate": 8.935292085083813e-09,
"loss": 1.3632,
"step": 5574
},
{
"epoch": 0.9907148251810387,
"grad_norm": 0.3330486809773884,
"learning_rate": 8.601316145362415e-09,
"loss": 1.2933,
"step": 5575
},
{
"epoch": 0.9908925318761385,
"grad_norm": 0.330304851645659,
"learning_rate": 8.273699845520178e-09,
"loss": 1.2736,
"step": 5576
},
{
"epoch": 0.9910702385712382,
"grad_norm": 0.3371388560971867,
"learning_rate": 7.952443289773736e-09,
"loss": 1.2966,
"step": 5577
},
{
"epoch": 0.9912479452663379,
"grad_norm": 0.33295926372675677,
"learning_rate": 7.637546580323563e-09,
"loss": 1.2867,
"step": 5578
},
{
"epoch": 0.9914256519614376,
"grad_norm": 0.3327902744721392,
"learning_rate": 7.329009817340638e-09,
"loss": 1.2927,
"step": 5579
},
{
"epoch": 0.9916033586565374,
"grad_norm": 0.3318071351677232,
"learning_rate": 7.026833098982e-09,
"loss": 1.2815,
"step": 5580
},
{
"epoch": 0.9917810653516371,
"grad_norm": 0.3380948737124673,
"learning_rate": 6.731016521370759e-09,
"loss": 1.3166,
"step": 5581
},
{
"epoch": 0.9919587720467369,
"grad_norm": 0.3355982348752432,
"learning_rate": 6.441560178613859e-09,
"loss": 1.3125,
"step": 5582
},
{
"epoch": 0.9921364787418366,
"grad_norm": 0.34445764578961857,
"learning_rate": 6.158464162790978e-09,
"loss": 1.3362,
"step": 5583
},
{
"epoch": 0.9923141854369364,
"grad_norm": 0.3357308445662798,
"learning_rate": 5.881728563963407e-09,
"loss": 1.3062,
"step": 5584
},
{
"epoch": 0.992491892132036,
"grad_norm": 0.32860436051123854,
"learning_rate": 5.61135347016295e-09,
"loss": 1.2735,
"step": 5585
},
{
"epoch": 0.9926695988271358,
"grad_norm": 0.33541514637400555,
"learning_rate": 5.347338967403026e-09,
"loss": 1.2814,
"step": 5586
},
{
"epoch": 0.9928473055222355,
"grad_norm": 0.3354504833983017,
"learning_rate": 5.0896851396720075e-09,
"loss": 1.2925,
"step": 5587
},
{
"epoch": 0.9930250122173353,
"grad_norm": 0.3290619966230952,
"learning_rate": 4.838392068930997e-09,
"loss": 1.2981,
"step": 5588
},
{
"epoch": 0.9932027189124351,
"grad_norm": 0.3376145256957276,
"learning_rate": 4.593459835124936e-09,
"loss": 1.3221,
"step": 5589
},
{
"epoch": 0.9933804256075348,
"grad_norm": 0.33919557771086134,
"learning_rate": 4.354888516169276e-09,
"loss": 1.3181,
"step": 5590
},
{
"epoch": 0.9935581323026345,
"grad_norm": 0.34783061253735675,
"learning_rate": 4.122678187958862e-09,
"loss": 1.2917,
"step": 5591
},
{
"epoch": 0.9937358389977342,
"grad_norm": 0.34111500886830387,
"learning_rate": 3.896828924363494e-09,
"loss": 1.3352,
"step": 5592
},
{
"epoch": 0.993913545692834,
"grad_norm": 0.32600318321380806,
"learning_rate": 3.677340797232365e-09,
"loss": 1.2595,
"step": 5593
},
{
"epoch": 0.9940912523879337,
"grad_norm": 0.33495407713116027,
"learning_rate": 3.4642138763851805e-09,
"loss": 1.3079,
"step": 5594
},
{
"epoch": 0.9942689590830335,
"grad_norm": 0.3414480894291363,
"learning_rate": 3.2574482296232613e-09,
"loss": 1.3453,
"step": 5595
},
{
"epoch": 0.9944466657781332,
"grad_norm": 0.3347591725568227,
"learning_rate": 3.0570439227228798e-09,
"loss": 1.2976,
"step": 5596
},
{
"epoch": 0.9946243724732329,
"grad_norm": 0.33915538633044634,
"learning_rate": 2.8630010194352633e-09,
"loss": 1.2972,
"step": 5597
},
{
"epoch": 0.9948020791683326,
"grad_norm": 0.3360781620147765,
"learning_rate": 2.6753195814910315e-09,
"loss": 1.3047,
"step": 5598
},
{
"epoch": 0.9949797858634324,
"grad_norm": 0.33433905067445796,
"learning_rate": 2.493999668595759e-09,
"loss": 1.3313,
"step": 5599
},
{
"epoch": 0.9951574925585321,
"grad_norm": 0.3395116395315644,
"learning_rate": 2.3190413384277522e-09,
"loss": 1.3158,
"step": 5600
},
{
"epoch": 0.9953351992536319,
"grad_norm": 0.338757517847559,
"learning_rate": 2.1504446466447115e-09,
"loss": 1.3198,
"step": 5601
},
{
"epoch": 0.9955129059487317,
"grad_norm": 0.33723656881776337,
"learning_rate": 1.988209646883732e-09,
"loss": 1.2775,
"step": 5602
},
{
"epoch": 0.9956906126438314,
"grad_norm": 0.3366514977370329,
"learning_rate": 1.8323363907524206e-09,
"loss": 1.278,
"step": 5603
},
{
"epoch": 0.995868319338931,
"grad_norm": 0.3377955856574426,
"learning_rate": 1.6828249278355579e-09,
"loss": 1.3043,
"step": 5604
},
{
"epoch": 0.9960460260340308,
"grad_norm": 0.32689575357260514,
"learning_rate": 1.5396753056995395e-09,
"loss": 1.2809,
"step": 5605
},
{
"epoch": 0.9962237327291306,
"grad_norm": 0.40647351124180825,
"learning_rate": 1.4028875698790524e-09,
"loss": 1.3272,
"step": 5606
},
{
"epoch": 0.9964014394242303,
"grad_norm": 0.3339080305134657,
"learning_rate": 1.272461763890398e-09,
"loss": 1.2833,
"step": 5607
},
{
"epoch": 0.9965791461193301,
"grad_norm": 0.33526640895605775,
"learning_rate": 1.148397929227052e-09,
"loss": 1.2949,
"step": 5608
},
{
"epoch": 0.9967568528144298,
"grad_norm": 0.3287574105995458,
"learning_rate": 1.0306961053507813e-09,
"loss": 1.2585,
"step": 5609
},
{
"epoch": 0.9969345595095295,
"grad_norm": 0.33981451186689343,
"learning_rate": 9.193563297094088e-10,
"loss": 1.3166,
"step": 5610
},
{
"epoch": 0.9971122662046292,
"grad_norm": 0.33549208274907316,
"learning_rate": 8.143786377190488e-10,
"loss": 1.3155,
"step": 5611
},
{
"epoch": 0.997289972899729,
"grad_norm": 0.33946821539604854,
"learning_rate": 7.157630627774303e-10,
"loss": 1.3398,
"step": 5612
},
{
"epoch": 0.9974676795948287,
"grad_norm": 0.33224934006903234,
"learning_rate": 6.235096362550153e-10,
"loss": 1.2682,
"step": 5613
},
{
"epoch": 0.9976453862899285,
"grad_norm": 0.3393511063877655,
"learning_rate": 5.376183874994389e-10,
"loss": 1.3134,
"step": 5614
},
{
"epoch": 0.9978230929850282,
"grad_norm": 0.3377013826730708,
"learning_rate": 4.5808934383329007e-10,
"loss": 1.3194,
"step": 5615
},
{
"epoch": 0.998000799680128,
"grad_norm": 0.33008719471156145,
"learning_rate": 3.8492253055855133e-10,
"loss": 1.301,
"step": 5616
},
{
"epoch": 0.9981785063752276,
"grad_norm": 0.33392257143602777,
"learning_rate": 3.1811797094993824e-10,
"loss": 1.2711,
"step": 5617
},
{
"epoch": 0.9983562130703274,
"grad_norm": 0.33466621838158467,
"learning_rate": 2.5767568625711946e-10,
"loss": 1.2779,
"step": 5618
},
{
"epoch": 0.9985339197654272,
"grad_norm": 0.32980081087954677,
"learning_rate": 2.0359569570915781e-10,
"loss": 1.2756,
"step": 5619
},
{
"epoch": 0.9987116264605269,
"grad_norm": 0.33372437129361626,
"learning_rate": 1.5587801651228973e-10,
"loss": 1.2977,
"step": 5620
},
{
"epoch": 0.9988893331556267,
"grad_norm": 0.32632112295014554,
"learning_rate": 1.1452266384548439e-10,
"loss": 1.2365,
"step": 5621
},
{
"epoch": 0.9990670398507264,
"grad_norm": 0.35918998607978136,
"learning_rate": 7.952965086044373e-11,
"loss": 1.3766,
"step": 5622
},
{
"epoch": 0.9992447465458261,
"grad_norm": 0.3376545850127059,
"learning_rate": 5.089898869492516e-11,
"loss": 1.3173,
"step": 5623
},
{
"epoch": 0.9994224532409258,
"grad_norm": 0.3387514237338692,
"learning_rate": 2.8630686454977908e-11,
"loss": 1.3245,
"step": 5624
},
{
"epoch": 0.9996001599360256,
"grad_norm": 0.3315692788753281,
"learning_rate": 1.2724751221604436e-11,
"loss": 1.2789,
"step": 5625
},
{
"epoch": 0.9997778666311253,
"grad_norm": 0.3339396334376731,
"learning_rate": 3.1811880574217357e-12,
"loss": 1.3026,
"step": 5626
},
{
"epoch": 0.9999555733262251,
"grad_norm": 0.3290399403888501,
"learning_rate": 0.0,
"loss": 1.2749,
"step": 5627
},
{
"epoch": 0.9999555733262251,
"step": 5627,
"total_flos": 6423200346931200.0,
"train_loss": 1.3842069712459713,
"train_runtime": 74502.9134,
"train_samples_per_second": 14.502,
"train_steps_per_second": 0.076
}
],
"logging_steps": 1,
"max_steps": 5627,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 400,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6423200346931200.0,
"train_batch_size": 3,
"trial_name": null,
"trial_params": null
}