e1_code_ms_phi / trainer_state.json
neginr's picture
End of training
989f57c verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.982278481012658,
"eval_steps": 500,
"global_step": 1230,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004050632911392405,
"grad_norm": 8.615641611522339,
"learning_rate": 3.2520325203252037e-07,
"loss": 1.987,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 8.680722097369141,
"learning_rate": 6.504065040650407e-07,
"loss": 1.984,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 8.597483955296141,
"learning_rate": 9.75609756097561e-07,
"loss": 1.992,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 8.288269993861656,
"learning_rate": 1.3008130081300815e-06,
"loss": 1.9532,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 8.045860714987628,
"learning_rate": 1.6260162601626018e-06,
"loss": 1.9353,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 7.688014827883297,
"learning_rate": 1.951219512195122e-06,
"loss": 1.9188,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 6.472187294228624,
"learning_rate": 2.2764227642276426e-06,
"loss": 1.9525,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 5.832143524962857,
"learning_rate": 2.601626016260163e-06,
"loss": 1.905,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 3.8156991182587108,
"learning_rate": 2.926829268292683e-06,
"loss": 1.874,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 3.2223455370260403,
"learning_rate": 3.2520325203252037e-06,
"loss": 1.838,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 3.1855963996934333,
"learning_rate": 3.577235772357724e-06,
"loss": 1.8869,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 5.844887779868974,
"learning_rate": 3.902439024390244e-06,
"loss": 1.8501,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 6.00064965828111,
"learning_rate": 4.227642276422765e-06,
"loss": 1.8409,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 5.797102465427963,
"learning_rate": 4.552845528455285e-06,
"loss": 1.8142,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 5.593743435982958,
"learning_rate": 4.8780487804878055e-06,
"loss": 1.8274,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 3.5077808800928443,
"learning_rate": 5.203252032520326e-06,
"loss": 1.7748,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 2.889373206837541,
"learning_rate": 5.528455284552846e-06,
"loss": 1.7432,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.386165882324008,
"learning_rate": 5.853658536585366e-06,
"loss": 1.7669,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 1.9189386859543247,
"learning_rate": 6.178861788617887e-06,
"loss": 1.723,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.7321441859494309,
"learning_rate": 6.504065040650407e-06,
"loss": 1.6598,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.5827639202299555,
"learning_rate": 6.829268292682928e-06,
"loss": 1.658,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.7327152143082374,
"learning_rate": 7.154471544715448e-06,
"loss": 1.645,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.6801142050427733,
"learning_rate": 7.4796747967479676e-06,
"loss": 1.6346,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 1.516344138264732,
"learning_rate": 7.804878048780489e-06,
"loss": 1.6567,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 1.3794646497390177,
"learning_rate": 8.130081300813009e-06,
"loss": 1.6413,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 1.2616520852467525,
"learning_rate": 8.45528455284553e-06,
"loss": 1.5848,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 1.4353631935321454,
"learning_rate": 8.78048780487805e-06,
"loss": 1.6111,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 1.3171590730641414,
"learning_rate": 9.10569105691057e-06,
"loss": 1.5918,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 1.1965399490686643,
"learning_rate": 9.43089430894309e-06,
"loss": 1.6224,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 1.3468296066209429,
"learning_rate": 9.756097560975611e-06,
"loss": 1.6222,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 1.1881853954407895,
"learning_rate": 1.008130081300813e-05,
"loss": 1.592,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 1.104729956430697,
"learning_rate": 1.0406504065040652e-05,
"loss": 1.5933,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 1.0298453339552416,
"learning_rate": 1.0731707317073172e-05,
"loss": 1.5781,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.9958935729735326,
"learning_rate": 1.1056910569105692e-05,
"loss": 1.5424,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.9784980131461827,
"learning_rate": 1.1382113821138213e-05,
"loss": 1.5744,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.8560510509742679,
"learning_rate": 1.1707317073170731e-05,
"loss": 1.52,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 1.047394736326449,
"learning_rate": 1.2032520325203254e-05,
"loss": 1.5565,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.8680747839076585,
"learning_rate": 1.2357723577235774e-05,
"loss": 1.526,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.9746405188495155,
"learning_rate": 1.2682926829268294e-05,
"loss": 1.5671,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.8806989786895044,
"learning_rate": 1.3008130081300815e-05,
"loss": 1.5333,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.9715111735633324,
"learning_rate": 1.3333333333333333e-05,
"loss": 1.5741,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.8712782467605749,
"learning_rate": 1.3658536585365855e-05,
"loss": 1.5312,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 1.0530129024380095,
"learning_rate": 1.3983739837398376e-05,
"loss": 1.5634,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.8947921218617071,
"learning_rate": 1.4308943089430896e-05,
"loss": 1.499,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.9943850304866259,
"learning_rate": 1.4634146341463415e-05,
"loss": 1.5315,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.9002953049497533,
"learning_rate": 1.4959349593495935e-05,
"loss": 1.5135,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.8366454324074247,
"learning_rate": 1.528455284552846e-05,
"loss": 1.5464,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.930901052511549,
"learning_rate": 1.5609756097560978e-05,
"loss": 1.5587,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.8717862219579288,
"learning_rate": 1.5934959349593496e-05,
"loss": 1.5169,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.8435298904522474,
"learning_rate": 1.6260162601626018e-05,
"loss": 1.5318,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.9152689033364321,
"learning_rate": 1.6585365853658537e-05,
"loss": 1.515,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.8960106381257938,
"learning_rate": 1.691056910569106e-05,
"loss": 1.5011,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.9232908659978061,
"learning_rate": 1.7235772357723578e-05,
"loss": 1.5338,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.9317623764636531,
"learning_rate": 1.75609756097561e-05,
"loss": 1.5215,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.9205343935703529,
"learning_rate": 1.788617886178862e-05,
"loss": 1.5195,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.8866208729054518,
"learning_rate": 1.821138211382114e-05,
"loss": 1.5107,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 1.1581188426187847,
"learning_rate": 1.8536585365853663e-05,
"loss": 1.5351,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 1.0945053620140552,
"learning_rate": 1.886178861788618e-05,
"loss": 1.5046,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 1.0646836258078336,
"learning_rate": 1.91869918699187e-05,
"loss": 1.555,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 1.1259318144265071,
"learning_rate": 1.9512195121951222e-05,
"loss": 1.5298,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 1.251106519186625,
"learning_rate": 1.983739837398374e-05,
"loss": 1.5118,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.9528227533760406,
"learning_rate": 2.016260162601626e-05,
"loss": 1.4928,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 1.48888449344474,
"learning_rate": 2.048780487804878e-05,
"loss": 1.5457,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 1.2924381227430288,
"learning_rate": 2.0813008130081303e-05,
"loss": 1.5089,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 1.0744120527867034,
"learning_rate": 2.1138211382113822e-05,
"loss": 1.5138,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 1.5751084271957794,
"learning_rate": 2.1463414634146344e-05,
"loss": 1.4965,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 1.0482444237639876,
"learning_rate": 2.1788617886178863e-05,
"loss": 1.5121,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 1.591471196894525,
"learning_rate": 2.2113821138211385e-05,
"loss": 1.5033,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 1.1124141509546173,
"learning_rate": 2.2439024390243907e-05,
"loss": 1.4854,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 1.5887253009200497,
"learning_rate": 2.2764227642276426e-05,
"loss": 1.4928,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 1.1248153154427611,
"learning_rate": 2.3089430894308948e-05,
"loss": 1.4937,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 1.3597941379022749,
"learning_rate": 2.3414634146341463e-05,
"loss": 1.4994,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 1.545174631666826,
"learning_rate": 2.3739837398373985e-05,
"loss": 1.5224,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 1.4643999004055572,
"learning_rate": 2.4065040650406507e-05,
"loss": 1.5326,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 1.6120296887743768,
"learning_rate": 2.4390243902439026e-05,
"loss": 1.484,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 1.5937685209473895,
"learning_rate": 2.4715447154471548e-05,
"loss": 1.5153,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 1.5117636049151093,
"learning_rate": 2.5040650406504066e-05,
"loss": 1.5331,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 1.3113677655870812,
"learning_rate": 2.536585365853659e-05,
"loss": 1.4933,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 1.41971460971697,
"learning_rate": 2.569105691056911e-05,
"loss": 1.4621,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 1.0689546047417156,
"learning_rate": 2.601626016260163e-05,
"loss": 1.4884,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 1.3626631789843966,
"learning_rate": 2.634146341463415e-05,
"loss": 1.4999,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 0.9783814355763274,
"learning_rate": 2.6666666666666667e-05,
"loss": 1.4723,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 1.31176860660927,
"learning_rate": 2.699186991869919e-05,
"loss": 1.5211,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 1.3481950438403842,
"learning_rate": 2.731707317073171e-05,
"loss": 1.4818,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 1.1932182424999254,
"learning_rate": 2.764227642276423e-05,
"loss": 1.484,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 1.2905939148932706,
"learning_rate": 2.796747967479675e-05,
"loss": 1.4587,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 1.402589600818668,
"learning_rate": 2.829268292682927e-05,
"loss": 1.4941,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 1.6934868685507727,
"learning_rate": 2.8617886178861792e-05,
"loss": 1.4906,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 0.9872550858652401,
"learning_rate": 2.8943089430894314e-05,
"loss": 1.4816,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 1.6605422026537375,
"learning_rate": 2.926829268292683e-05,
"loss": 1.5206,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 1.5027842735407257,
"learning_rate": 2.959349593495935e-05,
"loss": 1.5117,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 1.1480085694807045,
"learning_rate": 2.991869918699187e-05,
"loss": 1.5078,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 1.6035342068952223,
"learning_rate": 3.0243902439024392e-05,
"loss": 1.4927,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 1.4824877506874063,
"learning_rate": 3.056910569105692e-05,
"loss": 1.497,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 1.344187553334793,
"learning_rate": 3.089430894308943e-05,
"loss": 1.4591,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 1.6981593019403471,
"learning_rate": 3.1219512195121955e-05,
"loss": 1.4534,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 1.0701303510755522,
"learning_rate": 3.154471544715447e-05,
"loss": 1.478,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 1.685560208811993,
"learning_rate": 3.186991869918699e-05,
"loss": 1.4831,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 1.1567194166273684,
"learning_rate": 3.2195121951219514e-05,
"loss": 1.4866,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 1.5799127007579434,
"learning_rate": 3.2520325203252037e-05,
"loss": 1.5197,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 1.3414442345108473,
"learning_rate": 3.284552845528456e-05,
"loss": 1.4747,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 1.448704224456761,
"learning_rate": 3.3170731707317074e-05,
"loss": 1.4938,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 1.6652261280762952,
"learning_rate": 3.3495934959349596e-05,
"loss": 1.5069,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 1.2561843261348615,
"learning_rate": 3.382113821138212e-05,
"loss": 1.4734,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 1.5453279734707444,
"learning_rate": 3.414634146341463e-05,
"loss": 1.5125,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 1.3332884285131663,
"learning_rate": 3.4471544715447155e-05,
"loss": 1.4844,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 1.1390774829720374,
"learning_rate": 3.479674796747968e-05,
"loss": 1.4744,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 2.150639870889838,
"learning_rate": 3.51219512195122e-05,
"loss": 1.4648,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 1.446762438795497,
"learning_rate": 3.544715447154472e-05,
"loss": 1.486,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 2.5008896952927855,
"learning_rate": 3.577235772357724e-05,
"loss": 1.4648,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 2.300550990286915,
"learning_rate": 3.609756097560976e-05,
"loss": 1.5273,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 2.114865915443387,
"learning_rate": 3.642276422764228e-05,
"loss": 1.5276,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 2.2609853004156397,
"learning_rate": 3.67479674796748e-05,
"loss": 1.512,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 1.5148258193751845,
"learning_rate": 3.7073170731707325e-05,
"loss": 1.4756,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 1.7365538462257004,
"learning_rate": 3.739837398373984e-05,
"loss": 1.4787,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 1.2749342024056973,
"learning_rate": 3.772357723577236e-05,
"loss": 1.497,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 0.9197025737446735,
"learning_rate": 3.804878048780488e-05,
"loss": 1.4784,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 1.1951218850480763,
"learning_rate": 3.83739837398374e-05,
"loss": 1.4624,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 1.3704164676546033,
"learning_rate": 3.869918699186992e-05,
"loss": 1.4785,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 1.4633416931791972,
"learning_rate": 3.9024390243902444e-05,
"loss": 1.5098,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 1.1379943140277364,
"learning_rate": 3.9349593495934966e-05,
"loss": 1.461,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 1.6898147196565907,
"learning_rate": 3.967479674796748e-05,
"loss": 1.5006,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 1.2605370702553997,
"learning_rate": 4e-05,
"loss": 1.4786,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 1.5043418375847386,
"learning_rate": 3.999991946137476e-05,
"loss": 1.499,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 1.527308311778344,
"learning_rate": 3.999967784614766e-05,
"loss": 1.5574,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 1.1397854906308342,
"learning_rate": 3.9999275156264656e-05,
"loss": 1.4933,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 1.8215483436106379,
"learning_rate": 3.999871139496895e-05,
"loss": 1.4983,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 1.219552237495796,
"learning_rate": 3.9997986566800995e-05,
"loss": 1.5018,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 1.7338796108372396,
"learning_rate": 3.999710067759846e-05,
"loss": 1.4802,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 1.7221622339306504,
"learning_rate": 3.999605373449617e-05,
"loss": 1.4848,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 1.8118627770302638,
"learning_rate": 3.9994845745926075e-05,
"loss": 1.5071,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 1.4625278501200494,
"learning_rate": 3.999347672161713e-05,
"loss": 1.473,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 2.264986507274312,
"learning_rate": 3.999194667259528e-05,
"loss": 1.4653,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 1.5480595297160644,
"learning_rate": 3.999025561118334e-05,
"loss": 1.4944,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 1.8007566806811748,
"learning_rate": 3.998840355100086e-05,
"loss": 1.4307,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 1.4905430758693674,
"learning_rate": 3.998639050696409e-05,
"loss": 1.4331,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 1.505935242898502,
"learning_rate": 3.998421649528582e-05,
"loss": 1.4508,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 1.3227527530718692,
"learning_rate": 3.9981881533475234e-05,
"loss": 1.5164,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 1.272068196092492,
"learning_rate": 3.997938564033779e-05,
"loss": 1.4535,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 1.897550273735802,
"learning_rate": 3.9976728835975064e-05,
"loss": 1.4735,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 1.396435168030463,
"learning_rate": 3.9973911141784605e-05,
"loss": 1.4918,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 1.481303939475835,
"learning_rate": 3.997093258045973e-05,
"loss": 1.4244,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 2.032408023355157,
"learning_rate": 3.996779317598936e-05,
"loss": 1.4361,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 1.2790432481033003,
"learning_rate": 3.996449295365782e-05,
"loss": 1.4909,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 1.6504450818808427,
"learning_rate": 3.996103194004467e-05,
"loss": 1.4993,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 1.3877982968340845,
"learning_rate": 3.995741016302441e-05,
"loss": 1.5148,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 1.1016531153041684,
"learning_rate": 3.9953627651766364e-05,
"loss": 1.493,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 1.5944829898704165,
"learning_rate": 3.9949684436734325e-05,
"loss": 1.4669,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 1.0931546341044185,
"learning_rate": 3.994558054968643e-05,
"loss": 1.4571,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 1.619517289871332,
"learning_rate": 3.994131602367481e-05,
"loss": 1.4395,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 1.5488751620763512,
"learning_rate": 3.9936890893045376e-05,
"loss": 1.5016,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 1.4626709227997978,
"learning_rate": 3.993230519343752e-05,
"loss": 1.4471,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 1.8623032693382766,
"learning_rate": 3.992755896178383e-05,
"loss": 1.4861,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 1.4931777642391326,
"learning_rate": 3.992265223630981e-05,
"loss": 1.4941,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 1.5312400454211808,
"learning_rate": 3.991758505653355e-05,
"loss": 1.449,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 1.277835879593366,
"learning_rate": 3.991235746326543e-05,
"loss": 1.439,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 1.4746093082874043,
"learning_rate": 3.9906969498607745e-05,
"loss": 1.4803,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 1.1954945350609507,
"learning_rate": 3.990142120595444e-05,
"loss": 1.4558,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 2.11059551757457,
"learning_rate": 3.98957126299907e-05,
"loss": 1.468,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 1.1651611577368415,
"learning_rate": 3.9889843816692596e-05,
"loss": 1.4691,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 2.373971705632944,
"learning_rate": 3.9883814813326766e-05,
"loss": 1.4537,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 1.8065586474292117,
"learning_rate": 3.9877625668449956e-05,
"loss": 1.4767,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 2.0307336637931495,
"learning_rate": 3.98712764319087e-05,
"loss": 1.4576,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 2.026503086789421,
"learning_rate": 3.9864767154838864e-05,
"loss": 1.4415,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 1.6158846432370393,
"learning_rate": 3.9858097889665277e-05,
"loss": 1.4657,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 1.5793292816087992,
"learning_rate": 3.985126869010129e-05,
"loss": 1.442,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 1.3821510694777055,
"learning_rate": 3.984427961114833e-05,
"loss": 1.4793,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 1.664108613008087,
"learning_rate": 3.9837130709095475e-05,
"loss": 1.4904,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 1.4836079417085863,
"learning_rate": 3.982982204151901e-05,
"loss": 1.4301,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 1.0551526190752498,
"learning_rate": 3.982235366728193e-05,
"loss": 1.4669,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 1.4499525176778651,
"learning_rate": 3.9814725646533505e-05,
"loss": 1.487,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 1.4169850704078746,
"learning_rate": 3.9806938040708746e-05,
"loss": 1.4684,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 1.1227727970593668,
"learning_rate": 3.9798990912527976e-05,
"loss": 1.4806,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 1.503480186359298,
"learning_rate": 3.979088432599627e-05,
"loss": 1.4766,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 1.1805398462979901,
"learning_rate": 3.9782618346402964e-05,
"loss": 1.46,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 1.274156002452071,
"learning_rate": 3.977419304032111e-05,
"loss": 1.4763,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 1.3587997224907882,
"learning_rate": 3.976560847560697e-05,
"loss": 1.4661,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 1.0982579140007032,
"learning_rate": 3.9756864721399456e-05,
"loss": 1.4853,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 1.1943409022683902,
"learning_rate": 3.974796184811956e-05,
"loss": 1.4898,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 1.3348203319417475,
"learning_rate": 3.973889992746979e-05,
"loss": 1.4492,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 1.238005528243546,
"learning_rate": 3.972967903243361e-05,
"loss": 1.4517,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.9097644149918148,
"learning_rate": 3.972029923727486e-05,
"loss": 1.4579,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 1.577576364225581,
"learning_rate": 3.971076061753709e-05,
"loss": 1.4695,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 1.1066124826856585,
"learning_rate": 3.9701063250043066e-05,
"loss": 1.4861,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 1.2548656144722827,
"learning_rate": 3.969120721289402e-05,
"loss": 1.4504,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 1.241595286673754,
"learning_rate": 3.9681192585469146e-05,
"loss": 1.4585,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 1.2459597163307097,
"learning_rate": 3.9671019448424865e-05,
"loss": 1.4784,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 1.5316041404469969,
"learning_rate": 3.966068788369422e-05,
"loss": 1.4611,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.8195429421808759,
"learning_rate": 3.965019797448622e-05,
"loss": 1.4708,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 1.4396198202772457,
"learning_rate": 3.963954980528515e-05,
"loss": 1.4742,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 1.3999642349651722,
"learning_rate": 3.9628743461849905e-05,
"loss": 1.4483,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.8654173530962698,
"learning_rate": 3.961777903121329e-05,
"loss": 1.4114,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 1.845798482659176,
"learning_rate": 3.960665660168131e-05,
"loss": 1.451,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 1.2581451590555,
"learning_rate": 3.9595376262832485e-05,
"loss": 1.4528,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 1.4716280935390378,
"learning_rate": 3.9583938105517127e-05,
"loss": 1.4881,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 1.4488985121677107,
"learning_rate": 3.957234222185657e-05,
"loss": 1.4887,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 1.2175413383253368,
"learning_rate": 3.9560588705242474e-05,
"loss": 1.4654,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 1.0934231264577776,
"learning_rate": 3.954867765033605e-05,
"loss": 1.4426,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 1.5404575917723593,
"learning_rate": 3.953660915306728e-05,
"loss": 1.4718,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 1.009697661669436,
"learning_rate": 3.952438331063419e-05,
"loss": 1.4822,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 1.3150451886076506,
"learning_rate": 3.951200022150205e-05,
"loss": 1.4598,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.9202235908380316,
"learning_rate": 3.949945998540253e-05,
"loss": 1.4766,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 1.058539355886223,
"learning_rate": 3.9486762703332993e-05,
"loss": 1.4483,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.9520292423980145,
"learning_rate": 3.947390847755559e-05,
"loss": 1.4915,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.9360254453418689,
"learning_rate": 3.946089741159648e-05,
"loss": 1.4412,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.8127347145942394,
"learning_rate": 3.944772961024501e-05,
"loss": 1.4717,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 1.0957619888918548,
"learning_rate": 3.943440517955285e-05,
"loss": 1.4411,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.7053747893461345,
"learning_rate": 3.9420924226833126e-05,
"loss": 1.4502,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.9253728092451594,
"learning_rate": 3.9407286860659566e-05,
"loss": 1.4567,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 1.0003575814323384,
"learning_rate": 3.9393493190865657e-05,
"loss": 1.4448,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 1.0983891847892189,
"learning_rate": 3.937954332854371e-05,
"loss": 1.4707,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 1.2414480147455542,
"learning_rate": 3.9365437386044016e-05,
"loss": 1.4654,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.9935235631578035,
"learning_rate": 3.935117547697387e-05,
"loss": 1.4493,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 1.1375279958511724,
"learning_rate": 3.933675771619675e-05,
"loss": 1.4854,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.9442574558658872,
"learning_rate": 3.932218421983131e-05,
"loss": 1.5073,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 1.0535920034929074,
"learning_rate": 3.9307455105250484e-05,
"loss": 1.473,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.8624888961285828,
"learning_rate": 3.929257049108054e-05,
"loss": 1.5161,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 1.261799091536763,
"learning_rate": 3.927753049720011e-05,
"loss": 1.4312,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.7309526981523612,
"learning_rate": 3.9262335244739234e-05,
"loss": 1.4406,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 1.1061275711193936,
"learning_rate": 3.92469848560784e-05,
"loss": 1.4395,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.8593717976297521,
"learning_rate": 3.923147945484751e-05,
"loss": 1.4274,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 1.001993939887989,
"learning_rate": 3.9215819165924956e-05,
"loss": 1.4277,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.9821562089733628,
"learning_rate": 3.920000411543654e-05,
"loss": 1.4589,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 1.1300202838213798,
"learning_rate": 3.9184034430754495e-05,
"loss": 1.4328,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.8972108496910549,
"learning_rate": 3.916791024049648e-05,
"loss": 1.4745,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.8206436717789091,
"learning_rate": 3.91516316745245e-05,
"loss": 1.4286,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 1.1305877832397286,
"learning_rate": 3.913519886394389e-05,
"loss": 1.4758,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 1.235514508560542,
"learning_rate": 3.911861194110225e-05,
"loss": 1.4515,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.7925143312753444,
"learning_rate": 3.910187103958837e-05,
"loss": 1.4903,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 1.0694846741500705,
"learning_rate": 3.908497629423117e-05,
"loss": 1.4383,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 1.1334036294111147,
"learning_rate": 3.9067927841098614e-05,
"loss": 1.458,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 1.1692512309263925,
"learning_rate": 3.9050725817496594e-05,
"loss": 1.4626,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.8156552196400554,
"learning_rate": 3.9033370361967844e-05,
"loss": 1.455,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.9265491113315321,
"learning_rate": 3.901586161429081e-05,
"loss": 1.4442,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 1.2419347644320744,
"learning_rate": 3.8998199715478545e-05,
"loss": 1.4589,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.8240140025055316,
"learning_rate": 3.8980384807777564e-05,
"loss": 1.4645,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.9279144950698436,
"learning_rate": 3.896241703466667e-05,
"loss": 1.4729,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 1.1126843095403756,
"learning_rate": 3.894429654085585e-05,
"loss": 1.4554,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.9752192635319107,
"learning_rate": 3.892602347228505e-05,
"loss": 1.4514,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.9584990089462958,
"learning_rate": 3.890759797612307e-05,
"loss": 1.4589,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.9117828446590077,
"learning_rate": 3.888902020076632e-05,
"loss": 1.4915,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 1.2158513094448236,
"learning_rate": 3.887029029583764e-05,
"loss": 1.493,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 1.0847410958419745,
"learning_rate": 3.8851408412185125e-05,
"loss": 1.4374,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.8910215600930377,
"learning_rate": 3.8832374701880855e-05,
"loss": 1.4718,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.7353985517622208,
"learning_rate": 3.881318931821972e-05,
"loss": 1.4838,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.9881939495122468,
"learning_rate": 3.879385241571817e-05,
"loss": 1.4994,
"step": 246
},
{
"epoch": 1.000506329113924,
"grad_norm": 1.1150336521938389,
"learning_rate": 3.8774364150112955e-05,
"loss": 1.4156,
"step": 247
},
{
"epoch": 1.0045569620253165,
"grad_norm": 0.8891182544886729,
"learning_rate": 3.8754724678359884e-05,
"loss": 1.4132,
"step": 248
},
{
"epoch": 1.0086075949367088,
"grad_norm": 0.7986511348681001,
"learning_rate": 3.873493415863256e-05,
"loss": 1.4119,
"step": 249
},
{
"epoch": 1.0126582278481013,
"grad_norm": 0.90663344935251,
"learning_rate": 3.871499275032111e-05,
"loss": 1.3945,
"step": 250
},
{
"epoch": 1.0167088607594936,
"grad_norm": 0.7606348119552222,
"learning_rate": 3.869490061403091e-05,
"loss": 1.3456,
"step": 251
},
{
"epoch": 1.0207594936708861,
"grad_norm": 1.0096793246465448,
"learning_rate": 3.867465791158124e-05,
"loss": 1.4259,
"step": 252
},
{
"epoch": 1.0248101265822784,
"grad_norm": 0.7890935885371549,
"learning_rate": 3.865426480600407e-05,
"loss": 1.3777,
"step": 253
},
{
"epoch": 1.028860759493671,
"grad_norm": 0.8362021816778373,
"learning_rate": 3.863372146154264e-05,
"loss": 1.38,
"step": 254
},
{
"epoch": 1.0329113924050632,
"grad_norm": 0.9029626820304667,
"learning_rate": 3.861302804365024e-05,
"loss": 1.3628,
"step": 255
},
{
"epoch": 1.0369620253164558,
"grad_norm": 0.8513743015291878,
"learning_rate": 3.85921847189888e-05,
"loss": 1.3566,
"step": 256
},
{
"epoch": 1.041012658227848,
"grad_norm": 0.9611368165013687,
"learning_rate": 3.85711916554276e-05,
"loss": 1.3935,
"step": 257
},
{
"epoch": 1.0450632911392406,
"grad_norm": 0.8087046841258464,
"learning_rate": 3.85500490220419e-05,
"loss": 1.4321,
"step": 258
},
{
"epoch": 1.0491139240506329,
"grad_norm": 0.9026073488176928,
"learning_rate": 3.852875698911154e-05,
"loss": 1.4204,
"step": 259
},
{
"epoch": 1.0531645569620254,
"grad_norm": 0.8201640193949115,
"learning_rate": 3.850731572811963e-05,
"loss": 1.3771,
"step": 260
},
{
"epoch": 1.0572151898734177,
"grad_norm": 0.8292605498609534,
"learning_rate": 3.848572541175116e-05,
"loss": 1.3412,
"step": 261
},
{
"epoch": 1.0612658227848102,
"grad_norm": 0.8183163011181238,
"learning_rate": 3.846398621389154e-05,
"loss": 1.3953,
"step": 262
},
{
"epoch": 1.0653164556962025,
"grad_norm": 0.6369210281482819,
"learning_rate": 3.84420983096253e-05,
"loss": 1.3666,
"step": 263
},
{
"epoch": 1.069367088607595,
"grad_norm": 1.0285381489596133,
"learning_rate": 3.8420061875234606e-05,
"loss": 1.3675,
"step": 264
},
{
"epoch": 1.0734177215189873,
"grad_norm": 0.9130416700435867,
"learning_rate": 3.839787708819787e-05,
"loss": 1.339,
"step": 265
},
{
"epoch": 1.0774683544303798,
"grad_norm": 0.8563599275294592,
"learning_rate": 3.8375544127188325e-05,
"loss": 1.3802,
"step": 266
},
{
"epoch": 1.081518987341772,
"grad_norm": 0.8268104605255455,
"learning_rate": 3.8353063172072564e-05,
"loss": 1.4192,
"step": 267
},
{
"epoch": 1.0855696202531646,
"grad_norm": 0.8402121321261657,
"learning_rate": 3.8330434403909105e-05,
"loss": 1.3815,
"step": 268
},
{
"epoch": 1.089620253164557,
"grad_norm": 1.0917294392093133,
"learning_rate": 3.8307658004946934e-05,
"loss": 1.3566,
"step": 269
},
{
"epoch": 1.0936708860759494,
"grad_norm": 0.8406854011014969,
"learning_rate": 3.8284734158624046e-05,
"loss": 1.3805,
"step": 270
},
{
"epoch": 1.0977215189873417,
"grad_norm": 0.7858361875986336,
"learning_rate": 3.826166304956594e-05,
"loss": 1.3902,
"step": 271
},
{
"epoch": 1.1017721518987342,
"grad_norm": 1.0264261793325862,
"learning_rate": 3.8238444863584164e-05,
"loss": 1.3793,
"step": 272
},
{
"epoch": 1.1058227848101265,
"grad_norm": 0.8619702941457591,
"learning_rate": 3.821507978767479e-05,
"loss": 1.4164,
"step": 273
},
{
"epoch": 1.109873417721519,
"grad_norm": 0.7511177617560583,
"learning_rate": 3.819156801001693e-05,
"loss": 1.3265,
"step": 274
},
{
"epoch": 1.1139240506329113,
"grad_norm": 1.018906333523581,
"learning_rate": 3.816790971997121e-05,
"loss": 1.3485,
"step": 275
},
{
"epoch": 1.1179746835443038,
"grad_norm": 0.8910813682507854,
"learning_rate": 3.8144105108078246e-05,
"loss": 1.3526,
"step": 276
},
{
"epoch": 1.1220253164556961,
"grad_norm": 1.0228185081627172,
"learning_rate": 3.81201543660571e-05,
"loss": 1.4017,
"step": 277
},
{
"epoch": 1.1260759493670887,
"grad_norm": 0.8076275426535322,
"learning_rate": 3.809605768680377e-05,
"loss": 1.3559,
"step": 278
},
{
"epoch": 1.130126582278481,
"grad_norm": 0.98952969957489,
"learning_rate": 3.807181526438958e-05,
"loss": 1.3543,
"step": 279
},
{
"epoch": 1.1341772151898735,
"grad_norm": 0.820532910352288,
"learning_rate": 3.8047427294059697e-05,
"loss": 1.3599,
"step": 280
},
{
"epoch": 1.1382278481012658,
"grad_norm": 0.8690507681533287,
"learning_rate": 3.802289397223145e-05,
"loss": 1.3732,
"step": 281
},
{
"epoch": 1.1422784810126583,
"grad_norm": 0.9189763221152858,
"learning_rate": 3.7998215496492854e-05,
"loss": 1.4171,
"step": 282
},
{
"epoch": 1.1463291139240506,
"grad_norm": 0.6130332828472955,
"learning_rate": 3.797339206560096e-05,
"loss": 1.3876,
"step": 283
},
{
"epoch": 1.150379746835443,
"grad_norm": 0.8479366603584081,
"learning_rate": 3.794842387948027e-05,
"loss": 1.4199,
"step": 284
},
{
"epoch": 1.1544303797468354,
"grad_norm": 0.8350400452459558,
"learning_rate": 3.7923311139221114e-05,
"loss": 1.3512,
"step": 285
},
{
"epoch": 1.158481012658228,
"grad_norm": 1.0557570212760365,
"learning_rate": 3.7898054047078054e-05,
"loss": 1.3989,
"step": 286
},
{
"epoch": 1.1625316455696202,
"grad_norm": 0.8938289980727977,
"learning_rate": 3.787265280646825e-05,
"loss": 1.3849,
"step": 287
},
{
"epoch": 1.1665822784810127,
"grad_norm": 0.8062164424700141,
"learning_rate": 3.7847107621969786e-05,
"loss": 1.3875,
"step": 288
},
{
"epoch": 1.170632911392405,
"grad_norm": 0.9929532724027101,
"learning_rate": 3.7821418699320064e-05,
"loss": 1.3828,
"step": 289
},
{
"epoch": 1.1746835443037975,
"grad_norm": 1.3238935722687084,
"learning_rate": 3.7795586245414145e-05,
"loss": 1.4101,
"step": 290
},
{
"epoch": 1.1787341772151898,
"grad_norm": 0.986423193240161,
"learning_rate": 3.776961046830306e-05,
"loss": 1.3609,
"step": 291
},
{
"epoch": 1.1827848101265823,
"grad_norm": 1.4271742744917026,
"learning_rate": 3.774349157719215e-05,
"loss": 1.3864,
"step": 292
},
{
"epoch": 1.1868354430379746,
"grad_norm": 0.9757598685333242,
"learning_rate": 3.7717229782439365e-05,
"loss": 1.3895,
"step": 293
},
{
"epoch": 1.1908860759493671,
"grad_norm": 1.3095877658547743,
"learning_rate": 3.769082529555359e-05,
"loss": 1.3531,
"step": 294
},
{
"epoch": 1.1949367088607594,
"grad_norm": 0.8102134131619252,
"learning_rate": 3.766427832919294e-05,
"loss": 1.4205,
"step": 295
},
{
"epoch": 1.198987341772152,
"grad_norm": 1.0346502256234198,
"learning_rate": 3.7637589097163024e-05,
"loss": 1.4027,
"step": 296
},
{
"epoch": 1.2030379746835442,
"grad_norm": 1.1116438444158967,
"learning_rate": 3.761075781441526e-05,
"loss": 1.4137,
"step": 297
},
{
"epoch": 1.2070886075949367,
"grad_norm": 1.1647677710581892,
"learning_rate": 3.75837846970451e-05,
"loss": 1.4179,
"step": 298
},
{
"epoch": 1.211139240506329,
"grad_norm": 0.733018494811617,
"learning_rate": 3.755666996229032e-05,
"loss": 1.3444,
"step": 299
},
{
"epoch": 1.2151898734177216,
"grad_norm": 1.6289401762168263,
"learning_rate": 3.752941382852927e-05,
"loss": 1.4042,
"step": 300
},
{
"epoch": 1.2192405063291138,
"grad_norm": 0.8382105111833787,
"learning_rate": 3.7502016515279115e-05,
"loss": 1.3406,
"step": 301
},
{
"epoch": 1.2232911392405064,
"grad_norm": 1.2686219238725254,
"learning_rate": 3.7474478243194043e-05,
"loss": 1.3875,
"step": 302
},
{
"epoch": 1.2273417721518987,
"grad_norm": 0.8636943032382878,
"learning_rate": 3.744679923406351e-05,
"loss": 1.3559,
"step": 303
},
{
"epoch": 1.2313924050632912,
"grad_norm": 0.8666911565508588,
"learning_rate": 3.741897971081043e-05,
"loss": 1.3762,
"step": 304
},
{
"epoch": 1.2354430379746835,
"grad_norm": 0.9391024160804152,
"learning_rate": 3.739101989748946e-05,
"loss": 1.3883,
"step": 305
},
{
"epoch": 1.239493670886076,
"grad_norm": 0.7796433997209938,
"learning_rate": 3.7362920019285066e-05,
"loss": 1.3686,
"step": 306
},
{
"epoch": 1.2435443037974683,
"grad_norm": 0.6730275103757214,
"learning_rate": 3.73346803025098e-05,
"loss": 1.362,
"step": 307
},
{
"epoch": 1.2475949367088608,
"grad_norm": 1.9375516106796957,
"learning_rate": 3.730630097460247e-05,
"loss": 1.4416,
"step": 308
},
{
"epoch": 1.251645569620253,
"grad_norm": 0.8991883575545008,
"learning_rate": 3.727778226412628e-05,
"loss": 1.4114,
"step": 309
},
{
"epoch": 1.2556962025316456,
"grad_norm": 1.1271101462236117,
"learning_rate": 3.7249124400767006e-05,
"loss": 1.4018,
"step": 310
},
{
"epoch": 1.259746835443038,
"grad_norm": 0.9953031540237265,
"learning_rate": 3.722032761533114e-05,
"loss": 1.3909,
"step": 311
},
{
"epoch": 1.2637974683544304,
"grad_norm": 0.9889420818798841,
"learning_rate": 3.719139213974403e-05,
"loss": 1.4066,
"step": 312
},
{
"epoch": 1.2678481012658227,
"grad_norm": 1.0314921938301156,
"learning_rate": 3.7162318207048006e-05,
"loss": 1.3951,
"step": 313
},
{
"epoch": 1.2718987341772152,
"grad_norm": 0.7343456510914778,
"learning_rate": 3.713310605140055e-05,
"loss": 1.3595,
"step": 314
},
{
"epoch": 1.2759493670886077,
"grad_norm": 0.7691574242817147,
"learning_rate": 3.710375590807233e-05,
"loss": 1.3443,
"step": 315
},
{
"epoch": 1.28,
"grad_norm": 0.9288660378192503,
"learning_rate": 3.7074268013445365e-05,
"loss": 1.3842,
"step": 316
},
{
"epoch": 1.2840506329113923,
"grad_norm": 0.8241755768649719,
"learning_rate": 3.7044642605011114e-05,
"loss": 1.3898,
"step": 317
},
{
"epoch": 1.2881012658227848,
"grad_norm": 0.7896281019627824,
"learning_rate": 3.701487992136854e-05,
"loss": 1.373,
"step": 318
},
{
"epoch": 1.2921518987341774,
"grad_norm": 1.0205233415250543,
"learning_rate": 3.69849802022222e-05,
"loss": 1.3896,
"step": 319
},
{
"epoch": 1.2962025316455696,
"grad_norm": 0.7102759332957402,
"learning_rate": 3.6954943688380334e-05,
"loss": 1.3343,
"step": 320
},
{
"epoch": 1.300253164556962,
"grad_norm": 0.9205995277557487,
"learning_rate": 3.692477062175289e-05,
"loss": 1.377,
"step": 321
},
{
"epoch": 1.3043037974683545,
"grad_norm": 0.7916142878186667,
"learning_rate": 3.689446124534958e-05,
"loss": 1.3546,
"step": 322
},
{
"epoch": 1.308354430379747,
"grad_norm": 0.7309106176439266,
"learning_rate": 3.686401580327799e-05,
"loss": 1.3612,
"step": 323
},
{
"epoch": 1.3124050632911393,
"grad_norm": 0.8747499475946591,
"learning_rate": 3.683343454074149e-05,
"loss": 1.3757,
"step": 324
},
{
"epoch": 1.3164556962025316,
"grad_norm": 0.9461928927088169,
"learning_rate": 3.6802717704037386e-05,
"loss": 1.3276,
"step": 325
},
{
"epoch": 1.320506329113924,
"grad_norm": 0.836894422334664,
"learning_rate": 3.6771865540554855e-05,
"loss": 1.3661,
"step": 326
},
{
"epoch": 1.3245569620253166,
"grad_norm": 0.7757044759058382,
"learning_rate": 3.674087829877297e-05,
"loss": 1.3997,
"step": 327
},
{
"epoch": 1.3286075949367089,
"grad_norm": 0.82857374917787,
"learning_rate": 3.6709756228258735e-05,
"loss": 1.4124,
"step": 328
},
{
"epoch": 1.3326582278481012,
"grad_norm": 0.7948990393799569,
"learning_rate": 3.667849957966501e-05,
"loss": 1.3998,
"step": 329
},
{
"epoch": 1.3367088607594937,
"grad_norm": 0.7426710263593549,
"learning_rate": 3.6647108604728546e-05,
"loss": 1.3919,
"step": 330
},
{
"epoch": 1.3407594936708862,
"grad_norm": 0.7704422445932746,
"learning_rate": 3.661558355626795e-05,
"loss": 1.3725,
"step": 331
},
{
"epoch": 1.3448101265822785,
"grad_norm": 0.7456277487977734,
"learning_rate": 3.658392468818163e-05,
"loss": 1.3943,
"step": 332
},
{
"epoch": 1.3488607594936708,
"grad_norm": 0.7041163501705238,
"learning_rate": 3.655213225544574e-05,
"loss": 1.3706,
"step": 333
},
{
"epoch": 1.3529113924050633,
"grad_norm": 0.8264839477147079,
"learning_rate": 3.652020651411218e-05,
"loss": 1.3874,
"step": 334
},
{
"epoch": 1.3569620253164558,
"grad_norm": 0.6237502169668947,
"learning_rate": 3.6488147721306474e-05,
"loss": 1.3947,
"step": 335
},
{
"epoch": 1.3610126582278481,
"grad_norm": 0.7656228828540372,
"learning_rate": 3.645595613522574e-05,
"loss": 1.3531,
"step": 336
},
{
"epoch": 1.3650632911392404,
"grad_norm": 0.6911694254474141,
"learning_rate": 3.642363201513657e-05,
"loss": 1.4073,
"step": 337
},
{
"epoch": 1.369113924050633,
"grad_norm": 0.7082857987081713,
"learning_rate": 3.6391175621373006e-05,
"loss": 1.4388,
"step": 338
},
{
"epoch": 1.3731645569620254,
"grad_norm": 0.6921131135595878,
"learning_rate": 3.6358587215334355e-05,
"loss": 1.3573,
"step": 339
},
{
"epoch": 1.3772151898734177,
"grad_norm": 0.7050485227949534,
"learning_rate": 3.632586705948318e-05,
"loss": 1.3671,
"step": 340
},
{
"epoch": 1.38126582278481,
"grad_norm": 0.8527702426455152,
"learning_rate": 3.629301541734311e-05,
"loss": 1.3809,
"step": 341
},
{
"epoch": 1.3853164556962025,
"grad_norm": 0.7132578719224759,
"learning_rate": 3.626003255349676e-05,
"loss": 1.4102,
"step": 342
},
{
"epoch": 1.389367088607595,
"grad_norm": 0.9651623388900192,
"learning_rate": 3.622691873358357e-05,
"loss": 1.4089,
"step": 343
},
{
"epoch": 1.3934177215189874,
"grad_norm": 0.8161551887678881,
"learning_rate": 3.61936742242977e-05,
"loss": 1.3437,
"step": 344
},
{
"epoch": 1.3974683544303796,
"grad_norm": 0.8629914133193918,
"learning_rate": 3.6160299293385864e-05,
"loss": 1.3819,
"step": 345
},
{
"epoch": 1.4015189873417722,
"grad_norm": 0.73047922283957,
"learning_rate": 3.612679420964516e-05,
"loss": 1.4068,
"step": 346
},
{
"epoch": 1.4055696202531647,
"grad_norm": 0.7575903699708684,
"learning_rate": 3.609315924292092e-05,
"loss": 1.3909,
"step": 347
},
{
"epoch": 1.409620253164557,
"grad_norm": 0.7245487119103745,
"learning_rate": 3.6059394664104554e-05,
"loss": 1.3956,
"step": 348
},
{
"epoch": 1.4136708860759493,
"grad_norm": 0.9430218884746455,
"learning_rate": 3.602550074513133e-05,
"loss": 1.3935,
"step": 349
},
{
"epoch": 1.4177215189873418,
"grad_norm": 1.4652055078799102,
"learning_rate": 3.599147775897822e-05,
"loss": 1.387,
"step": 350
},
{
"epoch": 1.4217721518987343,
"grad_norm": 1.0035638861098293,
"learning_rate": 3.595732597966167e-05,
"loss": 1.3874,
"step": 351
},
{
"epoch": 1.4258227848101266,
"grad_norm": 0.6503932573332342,
"learning_rate": 3.592304568223542e-05,
"loss": 1.3886,
"step": 352
},
{
"epoch": 1.4298734177215189,
"grad_norm": 0.7391213027522376,
"learning_rate": 3.588863714278826e-05,
"loss": 1.3513,
"step": 353
},
{
"epoch": 1.4339240506329114,
"grad_norm": 0.68463942767065,
"learning_rate": 3.585410063844186e-05,
"loss": 1.3745,
"step": 354
},
{
"epoch": 1.437974683544304,
"grad_norm": 0.6264161009183246,
"learning_rate": 3.581943644734846e-05,
"loss": 1.3679,
"step": 355
},
{
"epoch": 1.4420253164556962,
"grad_norm": 0.6836065613779252,
"learning_rate": 3.578464484868869e-05,
"loss": 1.3713,
"step": 356
},
{
"epoch": 1.4460759493670885,
"grad_norm": 0.8176353963305067,
"learning_rate": 3.5749726122669316e-05,
"loss": 1.3781,
"step": 357
},
{
"epoch": 1.450126582278481,
"grad_norm": 0.8676773754543644,
"learning_rate": 3.5714680550520943e-05,
"loss": 1.3765,
"step": 358
},
{
"epoch": 1.4541772151898735,
"grad_norm": 0.7055909075289973,
"learning_rate": 3.5679508414495794e-05,
"loss": 1.4035,
"step": 359
},
{
"epoch": 1.4582278481012658,
"grad_norm": 0.8552296651591919,
"learning_rate": 3.564420999786543e-05,
"loss": 1.3749,
"step": 360
},
{
"epoch": 1.4622784810126581,
"grad_norm": 0.7446844163413139,
"learning_rate": 3.560878558491842e-05,
"loss": 1.3831,
"step": 361
},
{
"epoch": 1.4663291139240506,
"grad_norm": 0.6879253054401387,
"learning_rate": 3.5573235460958145e-05,
"loss": 1.3933,
"step": 362
},
{
"epoch": 1.4703797468354431,
"grad_norm": 0.548877812112283,
"learning_rate": 3.553755991230039e-05,
"loss": 1.3714,
"step": 363
},
{
"epoch": 1.4744303797468354,
"grad_norm": 0.8058798242434794,
"learning_rate": 3.5501759226271144e-05,
"loss": 1.3972,
"step": 364
},
{
"epoch": 1.4784810126582277,
"grad_norm": 0.7330399791402602,
"learning_rate": 3.546583369120419e-05,
"loss": 1.3723,
"step": 365
},
{
"epoch": 1.4825316455696202,
"grad_norm": 0.7569922197182551,
"learning_rate": 3.5429783596438864e-05,
"loss": 1.3954,
"step": 366
},
{
"epoch": 1.4865822784810128,
"grad_norm": 0.6576864873241295,
"learning_rate": 3.539360923231766e-05,
"loss": 1.3601,
"step": 367
},
{
"epoch": 1.490632911392405,
"grad_norm": 0.7687790003867498,
"learning_rate": 3.535731089018394e-05,
"loss": 1.43,
"step": 368
},
{
"epoch": 1.4946835443037974,
"grad_norm": 0.761091154531718,
"learning_rate": 3.532088886237956e-05,
"loss": 1.379,
"step": 369
},
{
"epoch": 1.4987341772151899,
"grad_norm": 0.6458525637503512,
"learning_rate": 3.528434344224253e-05,
"loss": 1.355,
"step": 370
},
{
"epoch": 1.5027848101265824,
"grad_norm": 0.8346025092898166,
"learning_rate": 3.524767492410464e-05,
"loss": 1.3878,
"step": 371
},
{
"epoch": 1.5068354430379747,
"grad_norm": 0.6821522275276006,
"learning_rate": 3.521088360328908e-05,
"loss": 1.3761,
"step": 372
},
{
"epoch": 1.510886075949367,
"grad_norm": 0.6995863758901321,
"learning_rate": 3.517396977610811e-05,
"loss": 1.3492,
"step": 373
},
{
"epoch": 1.5149367088607595,
"grad_norm": 0.642155262589243,
"learning_rate": 3.5136933739860595e-05,
"loss": 1.3778,
"step": 374
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.8862275249033215,
"learning_rate": 3.509977579282971e-05,
"loss": 1.3517,
"step": 375
},
{
"epoch": 1.5230379746835443,
"grad_norm": 0.7978927648515889,
"learning_rate": 3.5062496234280424e-05,
"loss": 1.3854,
"step": 376
},
{
"epoch": 1.5270886075949366,
"grad_norm": 0.6353524968696146,
"learning_rate": 3.502509536445719e-05,
"loss": 1.3345,
"step": 377
},
{
"epoch": 1.531139240506329,
"grad_norm": 0.6655044576739494,
"learning_rate": 3.498757348458147e-05,
"loss": 1.3761,
"step": 378
},
{
"epoch": 1.5351898734177216,
"grad_norm": 0.6097689855427814,
"learning_rate": 3.4949930896849324e-05,
"loss": 1.4127,
"step": 379
},
{
"epoch": 1.539240506329114,
"grad_norm": 0.6740383702363372,
"learning_rate": 3.491216790442899e-05,
"loss": 1.3653,
"step": 380
},
{
"epoch": 1.5432911392405062,
"grad_norm": 0.6463814496926872,
"learning_rate": 3.487428481145839e-05,
"loss": 1.3942,
"step": 381
},
{
"epoch": 1.5473417721518987,
"grad_norm": 0.682957932799813,
"learning_rate": 3.483628192304278e-05,
"loss": 1.3764,
"step": 382
},
{
"epoch": 1.5513924050632912,
"grad_norm": 0.659834246912329,
"learning_rate": 3.479815954525219e-05,
"loss": 1.3914,
"step": 383
},
{
"epoch": 1.5554430379746835,
"grad_norm": 0.6818296554221766,
"learning_rate": 3.475991798511899e-05,
"loss": 1.3731,
"step": 384
},
{
"epoch": 1.5594936708860758,
"grad_norm": 0.7829749176166423,
"learning_rate": 3.4721557550635464e-05,
"loss": 1.4228,
"step": 385
},
{
"epoch": 1.5635443037974683,
"grad_norm": 0.7694395699942849,
"learning_rate": 3.468307855075128e-05,
"loss": 1.4165,
"step": 386
},
{
"epoch": 1.5675949367088609,
"grad_norm": 0.6061851167661072,
"learning_rate": 3.4644481295371005e-05,
"loss": 1.3654,
"step": 387
},
{
"epoch": 1.5716455696202531,
"grad_norm": 0.6287743135426461,
"learning_rate": 3.460576609535163e-05,
"loss": 1.3815,
"step": 388
},
{
"epoch": 1.5756962025316454,
"grad_norm": 0.7396003175423872,
"learning_rate": 3.456693326250006e-05,
"loss": 1.3602,
"step": 389
},
{
"epoch": 1.579746835443038,
"grad_norm": 0.6603842196299613,
"learning_rate": 3.452798310957058e-05,
"loss": 1.363,
"step": 390
},
{
"epoch": 1.5837974683544305,
"grad_norm": 0.6819942587065012,
"learning_rate": 3.4488915950262386e-05,
"loss": 1.3944,
"step": 391
},
{
"epoch": 1.5878481012658228,
"grad_norm": 0.5924485540400647,
"learning_rate": 3.4449732099216985e-05,
"loss": 1.418,
"step": 392
},
{
"epoch": 1.591898734177215,
"grad_norm": 0.7992458057362248,
"learning_rate": 3.441043187201574e-05,
"loss": 1.4098,
"step": 393
},
{
"epoch": 1.5959493670886076,
"grad_norm": 0.7657852600851169,
"learning_rate": 3.437101558517728e-05,
"loss": 1.3875,
"step": 394
},
{
"epoch": 1.6,
"grad_norm": 0.7728128050474146,
"learning_rate": 3.433148355615496e-05,
"loss": 1.3785,
"step": 395
},
{
"epoch": 1.6040506329113924,
"grad_norm": 0.9590900000478725,
"learning_rate": 3.4291836103334294e-05,
"loss": 1.3478,
"step": 396
},
{
"epoch": 1.6081012658227847,
"grad_norm": 0.8026123991737428,
"learning_rate": 3.425207354603043e-05,
"loss": 1.374,
"step": 397
},
{
"epoch": 1.6121518987341772,
"grad_norm": 0.8297515516755958,
"learning_rate": 3.421219620448553e-05,
"loss": 1.3838,
"step": 398
},
{
"epoch": 1.6162025316455697,
"grad_norm": 0.9807501480678043,
"learning_rate": 3.417220439986623e-05,
"loss": 1.4123,
"step": 399
},
{
"epoch": 1.620253164556962,
"grad_norm": 0.8855930795012977,
"learning_rate": 3.4132098454261024e-05,
"loss": 1.3731,
"step": 400
},
{
"epoch": 1.6243037974683543,
"grad_norm": 0.7162685633591221,
"learning_rate": 3.4091878690677676e-05,
"loss": 1.407,
"step": 401
},
{
"epoch": 1.6283544303797468,
"grad_norm": 0.7786680328668478,
"learning_rate": 3.405154543304065e-05,
"loss": 1.3718,
"step": 402
},
{
"epoch": 1.6324050632911393,
"grad_norm": 0.8645386053524851,
"learning_rate": 3.401109900618843e-05,
"loss": 1.3593,
"step": 403
},
{
"epoch": 1.6364556962025316,
"grad_norm": 1.7833607430820724,
"learning_rate": 3.3970539735870996e-05,
"loss": 1.3855,
"step": 404
},
{
"epoch": 1.640506329113924,
"grad_norm": 0.8717500286313808,
"learning_rate": 3.392986794874714e-05,
"loss": 1.426,
"step": 405
},
{
"epoch": 1.6445569620253164,
"grad_norm": 0.6974081237836177,
"learning_rate": 3.388908397238184e-05,
"loss": 1.3529,
"step": 406
},
{
"epoch": 1.648607594936709,
"grad_norm": 0.7081030784532533,
"learning_rate": 3.384818813524362e-05,
"loss": 1.3991,
"step": 407
},
{
"epoch": 1.6526582278481012,
"grad_norm": 0.7673347236398408,
"learning_rate": 3.380718076670195e-05,
"loss": 1.4052,
"step": 408
},
{
"epoch": 1.6567088607594935,
"grad_norm": 0.6317876052777615,
"learning_rate": 3.376606219702454e-05,
"loss": 1.3735,
"step": 409
},
{
"epoch": 1.660759493670886,
"grad_norm": 0.6681557340770994,
"learning_rate": 3.372483275737468e-05,
"loss": 1.3543,
"step": 410
},
{
"epoch": 1.6648101265822786,
"grad_norm": 0.7472515471667174,
"learning_rate": 3.368349277980861e-05,
"loss": 1.3877,
"step": 411
},
{
"epoch": 1.6688607594936709,
"grad_norm": 0.8379773992309224,
"learning_rate": 3.3642042597272844e-05,
"loss": 1.3835,
"step": 412
},
{
"epoch": 1.6729113924050631,
"grad_norm": 0.6013382158480429,
"learning_rate": 3.360048254360144e-05,
"loss": 1.3721,
"step": 413
},
{
"epoch": 1.6769620253164557,
"grad_norm": 0.7826281370246876,
"learning_rate": 3.355881295351336e-05,
"loss": 1.3951,
"step": 414
},
{
"epoch": 1.6810126582278482,
"grad_norm": 0.6671387128858535,
"learning_rate": 3.351703416260975e-05,
"loss": 1.3964,
"step": 415
},
{
"epoch": 1.6850632911392405,
"grad_norm": 0.7089499834321381,
"learning_rate": 3.347514650737126e-05,
"loss": 1.391,
"step": 416
},
{
"epoch": 1.6891139240506328,
"grad_norm": 0.7488960797339547,
"learning_rate": 3.3433150325155295e-05,
"loss": 1.3863,
"step": 417
},
{
"epoch": 1.6931645569620253,
"grad_norm": 0.6784387783471314,
"learning_rate": 3.339104595419334e-05,
"loss": 1.364,
"step": 418
},
{
"epoch": 1.6972151898734178,
"grad_norm": 0.7144974207240755,
"learning_rate": 3.3348833733588204e-05,
"loss": 1.3964,
"step": 419
},
{
"epoch": 1.70126582278481,
"grad_norm": 0.6663950836203119,
"learning_rate": 3.3306514003311305e-05,
"loss": 1.3891,
"step": 420
},
{
"epoch": 1.7053164556962024,
"grad_norm": 0.852420270971833,
"learning_rate": 3.326408710419996e-05,
"loss": 1.359,
"step": 421
},
{
"epoch": 1.709367088607595,
"grad_norm": 0.7130197835845776,
"learning_rate": 3.322155337795454e-05,
"loss": 1.3918,
"step": 422
},
{
"epoch": 1.7134177215189874,
"grad_norm": 0.6976741517461801,
"learning_rate": 3.317891316713587e-05,
"loss": 1.3884,
"step": 423
},
{
"epoch": 1.7174683544303797,
"grad_norm": 0.6867107839973946,
"learning_rate": 3.313616681516231e-05,
"loss": 1.365,
"step": 424
},
{
"epoch": 1.721518987341772,
"grad_norm": 0.565924378434859,
"learning_rate": 3.309331466630713e-05,
"loss": 1.3978,
"step": 425
},
{
"epoch": 1.7255696202531645,
"grad_norm": 0.7264288224146752,
"learning_rate": 3.305035706569563e-05,
"loss": 1.3436,
"step": 426
},
{
"epoch": 1.729620253164557,
"grad_norm": 0.6065488298407818,
"learning_rate": 3.3007294359302433e-05,
"loss": 1.3424,
"step": 427
},
{
"epoch": 1.7336708860759493,
"grad_norm": 0.7328862889149912,
"learning_rate": 3.296412689394864e-05,
"loss": 1.3679,
"step": 428
},
{
"epoch": 1.7377215189873416,
"grad_norm": 0.6340934177451167,
"learning_rate": 3.292085501729909e-05,
"loss": 1.3909,
"step": 429
},
{
"epoch": 1.7417721518987341,
"grad_norm": 0.7446579550655156,
"learning_rate": 3.2877479077859534e-05,
"loss": 1.3824,
"step": 430
},
{
"epoch": 1.7458227848101266,
"grad_norm": 0.6717415403892307,
"learning_rate": 3.283399942497381e-05,
"loss": 1.3573,
"step": 431
},
{
"epoch": 1.749873417721519,
"grad_norm": 0.6199007051585469,
"learning_rate": 3.279041640882108e-05,
"loss": 1.3907,
"step": 432
},
{
"epoch": 1.7539240506329112,
"grad_norm": 0.7902270699223696,
"learning_rate": 3.2746730380412964e-05,
"loss": 1.4279,
"step": 433
},
{
"epoch": 1.7579746835443038,
"grad_norm": 0.763793223525106,
"learning_rate": 3.2702941691590726e-05,
"loss": 1.3861,
"step": 434
},
{
"epoch": 1.7620253164556963,
"grad_norm": 0.6484901507054464,
"learning_rate": 3.265905069502244e-05,
"loss": 1.3994,
"step": 435
},
{
"epoch": 1.7660759493670886,
"grad_norm": 0.7507899684891405,
"learning_rate": 3.261505774420016e-05,
"loss": 1.3433,
"step": 436
},
{
"epoch": 1.7701265822784809,
"grad_norm": 0.6623362683089783,
"learning_rate": 3.257096319343707e-05,
"loss": 1.3671,
"step": 437
},
{
"epoch": 1.7741772151898734,
"grad_norm": 0.6260658700564133,
"learning_rate": 3.2526767397864614e-05,
"loss": 1.3707,
"step": 438
},
{
"epoch": 1.7782278481012659,
"grad_norm": 0.6487376864039962,
"learning_rate": 3.248247071342966e-05,
"loss": 1.3459,
"step": 439
},
{
"epoch": 1.7822784810126582,
"grad_norm": 0.6421067083020247,
"learning_rate": 3.243807349689161e-05,
"loss": 1.3936,
"step": 440
},
{
"epoch": 1.7863291139240505,
"grad_norm": 0.7815207930407434,
"learning_rate": 3.2393576105819544e-05,
"loss": 1.3812,
"step": 441
},
{
"epoch": 1.790379746835443,
"grad_norm": 0.7637478285841881,
"learning_rate": 3.2348978898589333e-05,
"loss": 1.3851,
"step": 442
},
{
"epoch": 1.7944303797468355,
"grad_norm": 0.650773586422324,
"learning_rate": 3.230428223438075e-05,
"loss": 1.3852,
"step": 443
},
{
"epoch": 1.7984810126582278,
"grad_norm": 0.7991517868515811,
"learning_rate": 3.225948647317459e-05,
"loss": 1.3552,
"step": 444
},
{
"epoch": 1.80253164556962,
"grad_norm": 0.5721096452119956,
"learning_rate": 3.2214591975749745e-05,
"loss": 1.3591,
"step": 445
},
{
"epoch": 1.8065822784810126,
"grad_norm": 0.5909508714971661,
"learning_rate": 3.216959910368034e-05,
"loss": 1.411,
"step": 446
},
{
"epoch": 1.8106329113924051,
"grad_norm": 0.5993363565656248,
"learning_rate": 3.212450821933277e-05,
"loss": 1.3478,
"step": 447
},
{
"epoch": 1.8146835443037974,
"grad_norm": 0.6201721102641272,
"learning_rate": 3.207931968586281e-05,
"loss": 1.396,
"step": 448
},
{
"epoch": 1.81873417721519,
"grad_norm": 0.5624801141277199,
"learning_rate": 3.203403386721272e-05,
"loss": 1.3486,
"step": 449
},
{
"epoch": 1.8227848101265822,
"grad_norm": 0.6030633459207573,
"learning_rate": 3.1988651128108245e-05,
"loss": 1.3592,
"step": 450
},
{
"epoch": 1.8268354430379747,
"grad_norm": 0.6626378937324846,
"learning_rate": 3.194317183405573e-05,
"loss": 1.3882,
"step": 451
},
{
"epoch": 1.8308860759493673,
"grad_norm": 0.6013847392667573,
"learning_rate": 3.189759635133914e-05,
"loss": 1.3873,
"step": 452
},
{
"epoch": 1.8349367088607595,
"grad_norm": 0.708027742307106,
"learning_rate": 3.185192504701718e-05,
"loss": 1.4051,
"step": 453
},
{
"epoch": 1.8389873417721518,
"grad_norm": 0.5772081524631106,
"learning_rate": 3.1806158288920234e-05,
"loss": 1.3874,
"step": 454
},
{
"epoch": 1.8430379746835444,
"grad_norm": 0.5695140381670215,
"learning_rate": 3.1760296445647477e-05,
"loss": 1.3776,
"step": 455
},
{
"epoch": 1.8470886075949369,
"grad_norm": 0.7619809669137889,
"learning_rate": 3.1714339886563896e-05,
"loss": 1.3757,
"step": 456
},
{
"epoch": 1.8511392405063292,
"grad_norm": 0.6776598217527482,
"learning_rate": 3.166828898179731e-05,
"loss": 1.3797,
"step": 457
},
{
"epoch": 1.8551898734177215,
"grad_norm": 0.7647267446195253,
"learning_rate": 3.162214410223536e-05,
"loss": 1.363,
"step": 458
},
{
"epoch": 1.859240506329114,
"grad_norm": 0.7618107383087158,
"learning_rate": 3.157590561952257e-05,
"loss": 1.3698,
"step": 459
},
{
"epoch": 1.8632911392405065,
"grad_norm": 0.5916517389742075,
"learning_rate": 3.152957390605732e-05,
"loss": 1.3749,
"step": 460
},
{
"epoch": 1.8673417721518988,
"grad_norm": 0.6215539591948732,
"learning_rate": 3.148314933498886e-05,
"loss": 1.4039,
"step": 461
},
{
"epoch": 1.871392405063291,
"grad_norm": 0.689286497613316,
"learning_rate": 3.143663228021431e-05,
"loss": 1.3792,
"step": 462
},
{
"epoch": 1.8754430379746836,
"grad_norm": 0.6504280995623329,
"learning_rate": 3.1390023116375624e-05,
"loss": 1.389,
"step": 463
},
{
"epoch": 1.879493670886076,
"grad_norm": 0.7535925712066275,
"learning_rate": 3.134332221885661e-05,
"loss": 1.3965,
"step": 464
},
{
"epoch": 1.8835443037974684,
"grad_norm": 0.7793589955996199,
"learning_rate": 3.129652996377987e-05,
"loss": 1.364,
"step": 465
},
{
"epoch": 1.8875949367088607,
"grad_norm": 0.7131343357857655,
"learning_rate": 3.12496467280038e-05,
"loss": 1.3568,
"step": 466
},
{
"epoch": 1.8916455696202532,
"grad_norm": 0.8566408164963178,
"learning_rate": 3.120267288911952e-05,
"loss": 1.3796,
"step": 467
},
{
"epoch": 1.8956962025316457,
"grad_norm": 0.705266350657106,
"learning_rate": 3.11556088254479e-05,
"loss": 1.3851,
"step": 468
},
{
"epoch": 1.899746835443038,
"grad_norm": 0.7734555925028461,
"learning_rate": 3.11084549160364e-05,
"loss": 1.3524,
"step": 469
},
{
"epoch": 1.9037974683544303,
"grad_norm": 0.7287001688124572,
"learning_rate": 3.106121154065615e-05,
"loss": 1.3927,
"step": 470
},
{
"epoch": 1.9078481012658228,
"grad_norm": 0.711876238048519,
"learning_rate": 3.1013879079798805e-05,
"loss": 1.3733,
"step": 471
},
{
"epoch": 1.9118987341772153,
"grad_norm": 0.6998932902187173,
"learning_rate": 3.096645791467348e-05,
"loss": 1.4062,
"step": 472
},
{
"epoch": 1.9159493670886076,
"grad_norm": 0.7551877889866037,
"learning_rate": 3.091894842720373e-05,
"loss": 1.3619,
"step": 473
},
{
"epoch": 1.92,
"grad_norm": 0.7607735997715396,
"learning_rate": 3.0871351000024425e-05,
"loss": 1.3711,
"step": 474
},
{
"epoch": 1.9240506329113924,
"grad_norm": 0.8268272813841067,
"learning_rate": 3.0823666016478716e-05,
"loss": 1.3575,
"step": 475
},
{
"epoch": 1.928101265822785,
"grad_norm": 0.6629164666389126,
"learning_rate": 3.0775893860614896e-05,
"loss": 1.4111,
"step": 476
},
{
"epoch": 1.9321518987341773,
"grad_norm": 0.7003341874024968,
"learning_rate": 3.0728034917183336e-05,
"loss": 1.3639,
"step": 477
},
{
"epoch": 1.9362025316455695,
"grad_norm": 0.7204909429275707,
"learning_rate": 3.06800895716334e-05,
"loss": 1.3632,
"step": 478
},
{
"epoch": 1.940253164556962,
"grad_norm": 0.6981436063660695,
"learning_rate": 3.063205821011029e-05,
"loss": 1.3867,
"step": 479
},
{
"epoch": 1.9443037974683546,
"grad_norm": 0.7470305701866339,
"learning_rate": 3.0583941219452016e-05,
"loss": 1.3929,
"step": 480
},
{
"epoch": 1.9483544303797469,
"grad_norm": 0.7302295132142407,
"learning_rate": 3.053573898718618e-05,
"loss": 1.3429,
"step": 481
},
{
"epoch": 1.9524050632911392,
"grad_norm": 0.6505596371931905,
"learning_rate": 3.0487451901526956e-05,
"loss": 1.3884,
"step": 482
},
{
"epoch": 1.9564556962025317,
"grad_norm": 0.6999308736501885,
"learning_rate": 3.0439080351371875e-05,
"loss": 1.3695,
"step": 483
},
{
"epoch": 1.9605063291139242,
"grad_norm": 0.7222648760640005,
"learning_rate": 3.0390624726298764e-05,
"loss": 1.3684,
"step": 484
},
{
"epoch": 1.9645569620253165,
"grad_norm": 0.6074414964462902,
"learning_rate": 3.034208541656255e-05,
"loss": 1.4113,
"step": 485
},
{
"epoch": 1.9686075949367088,
"grad_norm": 0.7457536250436803,
"learning_rate": 3.029346281309218e-05,
"loss": 1.3541,
"step": 486
},
{
"epoch": 1.9726582278481013,
"grad_norm": 0.5436709248072691,
"learning_rate": 3.0244757307487415e-05,
"loss": 1.3967,
"step": 487
},
{
"epoch": 1.9767088607594938,
"grad_norm": 0.7413079819842553,
"learning_rate": 3.019596929201569e-05,
"loss": 1.3802,
"step": 488
},
{
"epoch": 1.980759493670886,
"grad_norm": 0.6585817055922382,
"learning_rate": 3.0147099159608985e-05,
"loss": 1.3571,
"step": 489
},
{
"epoch": 1.9848101265822784,
"grad_norm": 0.794367554529059,
"learning_rate": 3.0098147303860616e-05,
"loss": 1.3784,
"step": 490
},
{
"epoch": 1.988860759493671,
"grad_norm": 0.6913975972927425,
"learning_rate": 3.0049114119022117e-05,
"loss": 1.3534,
"step": 491
},
{
"epoch": 1.9929113924050634,
"grad_norm": 0.8131123029659822,
"learning_rate": 3.0000000000000004e-05,
"loss": 1.3325,
"step": 492
},
{
"epoch": 1.9969620253164557,
"grad_norm": 0.642672413744278,
"learning_rate": 2.995080534235264e-05,
"loss": 1.3431,
"step": 493
},
{
"epoch": 2.001012658227848,
"grad_norm": 0.8091449638909065,
"learning_rate": 2.9901530542287044e-05,
"loss": 1.3426,
"step": 494
},
{
"epoch": 2.0050632911392405,
"grad_norm": 0.7768758357304671,
"learning_rate": 2.9852175996655676e-05,
"loss": 1.2636,
"step": 495
},
{
"epoch": 2.009113924050633,
"grad_norm": 1.1494253189473456,
"learning_rate": 2.980274210295326e-05,
"loss": 1.2368,
"step": 496
},
{
"epoch": 2.013164556962025,
"grad_norm": 1.0813551167605018,
"learning_rate": 2.9753229259313578e-05,
"loss": 1.273,
"step": 497
},
{
"epoch": 2.0172151898734176,
"grad_norm": 0.8194237497175929,
"learning_rate": 2.9703637864506274e-05,
"loss": 1.231,
"step": 498
},
{
"epoch": 2.02126582278481,
"grad_norm": 1.162249437171599,
"learning_rate": 2.965396831793362e-05,
"loss": 1.2214,
"step": 499
},
{
"epoch": 2.0253164556962027,
"grad_norm": 0.8136942858497734,
"learning_rate": 2.9604221019627316e-05,
"loss": 1.2991,
"step": 500
},
{
"epoch": 2.0293670886075947,
"grad_norm": 1.1058230992978715,
"learning_rate": 2.955439637024526e-05,
"loss": 1.2251,
"step": 501
},
{
"epoch": 2.0334177215189873,
"grad_norm": 1.0196854020367383,
"learning_rate": 2.9504494771068334e-05,
"loss": 1.2651,
"step": 502
},
{
"epoch": 2.0374683544303798,
"grad_norm": 0.7501525204546852,
"learning_rate": 2.9454516623997156e-05,
"loss": 1.2179,
"step": 503
},
{
"epoch": 2.0415189873417723,
"grad_norm": 1.081600194280537,
"learning_rate": 2.9404462331548847e-05,
"loss": 1.2233,
"step": 504
},
{
"epoch": 2.0455696202531644,
"grad_norm": 0.5792055057873841,
"learning_rate": 2.93543322968538e-05,
"loss": 1.216,
"step": 505
},
{
"epoch": 2.049620253164557,
"grad_norm": 0.8825043077674071,
"learning_rate": 2.9304126923652428e-05,
"loss": 1.254,
"step": 506
},
{
"epoch": 2.0536708860759494,
"grad_norm": 0.7880729292486276,
"learning_rate": 2.9253846616291896e-05,
"loss": 1.2442,
"step": 507
},
{
"epoch": 2.057721518987342,
"grad_norm": 0.7682897068134563,
"learning_rate": 2.9203491779722896e-05,
"loss": 1.2119,
"step": 508
},
{
"epoch": 2.061772151898734,
"grad_norm": 0.9272051708451627,
"learning_rate": 2.9153062819496357e-05,
"loss": 1.2574,
"step": 509
},
{
"epoch": 2.0658227848101265,
"grad_norm": 0.6883745504542177,
"learning_rate": 2.9102560141760178e-05,
"loss": 1.2894,
"step": 510
},
{
"epoch": 2.069873417721519,
"grad_norm": 0.7242602246324404,
"learning_rate": 2.9051984153256004e-05,
"loss": 1.2183,
"step": 511
},
{
"epoch": 2.0739240506329115,
"grad_norm": 0.6962543877316004,
"learning_rate": 2.900133526131588e-05,
"loss": 1.2167,
"step": 512
},
{
"epoch": 2.0779746835443036,
"grad_norm": 0.6630481282809939,
"learning_rate": 2.8950613873859025e-05,
"loss": 1.2547,
"step": 513
},
{
"epoch": 2.082025316455696,
"grad_norm": 0.7058028873193685,
"learning_rate": 2.8899820399388515e-05,
"loss": 1.2606,
"step": 514
},
{
"epoch": 2.0860759493670886,
"grad_norm": 0.642202830264344,
"learning_rate": 2.8848955246988012e-05,
"loss": 1.2885,
"step": 515
},
{
"epoch": 2.090126582278481,
"grad_norm": 0.8464399153684956,
"learning_rate": 2.879801882631847e-05,
"loss": 1.2384,
"step": 516
},
{
"epoch": 2.094177215189873,
"grad_norm": 0.5985987864379472,
"learning_rate": 2.8747011547614808e-05,
"loss": 1.2626,
"step": 517
},
{
"epoch": 2.0982278481012657,
"grad_norm": 0.8209782495019733,
"learning_rate": 2.8695933821682635e-05,
"loss": 1.2167,
"step": 518
},
{
"epoch": 2.1022784810126582,
"grad_norm": 0.656880749402127,
"learning_rate": 2.864478605989494e-05,
"loss": 1.231,
"step": 519
},
{
"epoch": 2.1063291139240508,
"grad_norm": 0.6129656228392018,
"learning_rate": 2.8593568674188765e-05,
"loss": 1.2246,
"step": 520
},
{
"epoch": 2.110379746835443,
"grad_norm": 0.6516389026029068,
"learning_rate": 2.8542282077061892e-05,
"loss": 1.2397,
"step": 521
},
{
"epoch": 2.1144303797468353,
"grad_norm": 0.5914845312907084,
"learning_rate": 2.8490926681569523e-05,
"loss": 1.2436,
"step": 522
},
{
"epoch": 2.118481012658228,
"grad_norm": 0.5861366276190385,
"learning_rate": 2.8439502901320956e-05,
"loss": 1.2327,
"step": 523
},
{
"epoch": 2.1225316455696204,
"grad_norm": 0.614229486623933,
"learning_rate": 2.8388011150476237e-05,
"loss": 1.2266,
"step": 524
},
{
"epoch": 2.1265822784810124,
"grad_norm": 0.5683280056559875,
"learning_rate": 2.8336451843742866e-05,
"loss": 1.2094,
"step": 525
},
{
"epoch": 2.130632911392405,
"grad_norm": 0.6174840862696923,
"learning_rate": 2.8284825396372387e-05,
"loss": 1.237,
"step": 526
},
{
"epoch": 2.1346835443037975,
"grad_norm": 0.6401411129552411,
"learning_rate": 2.8233132224157132e-05,
"loss": 1.2322,
"step": 527
},
{
"epoch": 2.13873417721519,
"grad_norm": 0.6809977050307494,
"learning_rate": 2.8181372743426805e-05,
"loss": 1.2468,
"step": 528
},
{
"epoch": 2.1427848101265825,
"grad_norm": 0.5447585442078468,
"learning_rate": 2.8129547371045128e-05,
"loss": 1.201,
"step": 529
},
{
"epoch": 2.1468354430379746,
"grad_norm": 0.6002393708412256,
"learning_rate": 2.8077656524406534e-05,
"loss": 1.2124,
"step": 530
},
{
"epoch": 2.150886075949367,
"grad_norm": 0.5773640914503281,
"learning_rate": 2.802570062143278e-05,
"loss": 1.2588,
"step": 531
},
{
"epoch": 2.1549367088607596,
"grad_norm": 0.5813425260127274,
"learning_rate": 2.7973680080569555e-05,
"loss": 1.2665,
"step": 532
},
{
"epoch": 2.1589873417721517,
"grad_norm": 0.5689340544186497,
"learning_rate": 2.792159532078314e-05,
"loss": 1.2348,
"step": 533
},
{
"epoch": 2.163037974683544,
"grad_norm": 0.5640146097243055,
"learning_rate": 2.7869446761557033e-05,
"loss": 1.2381,
"step": 534
},
{
"epoch": 2.1670886075949367,
"grad_norm": 0.5145076125906114,
"learning_rate": 2.781723482288857e-05,
"loss": 1.2819,
"step": 535
},
{
"epoch": 2.1711392405063292,
"grad_norm": 0.8614303397181456,
"learning_rate": 2.7764959925285517e-05,
"loss": 1.2615,
"step": 536
},
{
"epoch": 2.1751898734177217,
"grad_norm": 0.6974803011671088,
"learning_rate": 2.771262248976272e-05,
"loss": 1.245,
"step": 537
},
{
"epoch": 2.179240506329114,
"grad_norm": 0.5615337474788794,
"learning_rate": 2.7660222937838677e-05,
"loss": 1.2618,
"step": 538
},
{
"epoch": 2.1832911392405063,
"grad_norm": 0.6058967613944483,
"learning_rate": 2.7607761691532186e-05,
"loss": 1.2525,
"step": 539
},
{
"epoch": 2.187341772151899,
"grad_norm": 0.5303002938840129,
"learning_rate": 2.7555239173358916e-05,
"loss": 1.2411,
"step": 540
},
{
"epoch": 2.191392405063291,
"grad_norm": 0.5402230628698986,
"learning_rate": 2.7502655806328e-05,
"loss": 1.2348,
"step": 541
},
{
"epoch": 2.1954430379746834,
"grad_norm": 0.5661626374792779,
"learning_rate": 2.7450012013938648e-05,
"loss": 1.2488,
"step": 542
},
{
"epoch": 2.199493670886076,
"grad_norm": 0.5504043023694448,
"learning_rate": 2.739730822017673e-05,
"loss": 1.2509,
"step": 543
},
{
"epoch": 2.2035443037974685,
"grad_norm": 0.6262899653243039,
"learning_rate": 2.7344544849511355e-05,
"loss": 1.2417,
"step": 544
},
{
"epoch": 2.207594936708861,
"grad_norm": 0.5545437561864398,
"learning_rate": 2.7291722326891456e-05,
"loss": 1.2201,
"step": 545
},
{
"epoch": 2.211645569620253,
"grad_norm": 0.6037310439109309,
"learning_rate": 2.723884107774236e-05,
"loss": 1.2236,
"step": 546
},
{
"epoch": 2.2156962025316456,
"grad_norm": 0.5374350432364119,
"learning_rate": 2.718590152796239e-05,
"loss": 1.2724,
"step": 547
},
{
"epoch": 2.219746835443038,
"grad_norm": 0.5900615627862229,
"learning_rate": 2.71329041039194e-05,
"loss": 1.2065,
"step": 548
},
{
"epoch": 2.22379746835443,
"grad_norm": 0.6507726370044351,
"learning_rate": 2.7079849232447357e-05,
"loss": 1.2642,
"step": 549
},
{
"epoch": 2.2278481012658227,
"grad_norm": 0.7151603904509916,
"learning_rate": 2.7026737340842895e-05,
"loss": 1.2339,
"step": 550
},
{
"epoch": 2.231898734177215,
"grad_norm": 0.6270150786270356,
"learning_rate": 2.697356885686189e-05,
"loss": 1.2553,
"step": 551
},
{
"epoch": 2.2359493670886077,
"grad_norm": 0.6979948348074577,
"learning_rate": 2.6920344208716014e-05,
"loss": 1.2471,
"step": 552
},
{
"epoch": 2.24,
"grad_norm": 0.6649298871397636,
"learning_rate": 2.6867063825069252e-05,
"loss": 1.2717,
"step": 553
},
{
"epoch": 2.2440506329113923,
"grad_norm": 0.6041844036528765,
"learning_rate": 2.6813728135034494e-05,
"loss": 1.2621,
"step": 554
},
{
"epoch": 2.248101265822785,
"grad_norm": 0.6025256507630476,
"learning_rate": 2.6760337568170056e-05,
"loss": 1.2498,
"step": 555
},
{
"epoch": 2.2521518987341773,
"grad_norm": 0.5717609546814847,
"learning_rate": 2.6706892554476226e-05,
"loss": 1.2314,
"step": 556
},
{
"epoch": 2.2562025316455694,
"grad_norm": 0.6156859658824004,
"learning_rate": 2.6653393524391795e-05,
"loss": 1.276,
"step": 557
},
{
"epoch": 2.260253164556962,
"grad_norm": 0.551267081906443,
"learning_rate": 2.6599840908790592e-05,
"loss": 1.234,
"step": 558
},
{
"epoch": 2.2643037974683544,
"grad_norm": 0.6559927986909337,
"learning_rate": 2.6546235138978028e-05,
"loss": 1.2793,
"step": 559
},
{
"epoch": 2.268354430379747,
"grad_norm": 0.6256342680705126,
"learning_rate": 2.6492576646687597e-05,
"loss": 1.2536,
"step": 560
},
{
"epoch": 2.2724050632911394,
"grad_norm": 0.6080101842634819,
"learning_rate": 2.6438865864077425e-05,
"loss": 1.2736,
"step": 561
},
{
"epoch": 2.2764556962025315,
"grad_norm": 0.5746259008406501,
"learning_rate": 2.6385103223726766e-05,
"loss": 1.2662,
"step": 562
},
{
"epoch": 2.280506329113924,
"grad_norm": 0.6084855193950818,
"learning_rate": 2.6331289158632537e-05,
"loss": 1.2791,
"step": 563
},
{
"epoch": 2.2845569620253166,
"grad_norm": 0.5822788310943046,
"learning_rate": 2.6277424102205817e-05,
"loss": 1.2636,
"step": 564
},
{
"epoch": 2.2886075949367086,
"grad_norm": 0.6538299465764322,
"learning_rate": 2.6223508488268374e-05,
"loss": 1.2511,
"step": 565
},
{
"epoch": 2.292658227848101,
"grad_norm": 0.5415534720936008,
"learning_rate": 2.6169542751049148e-05,
"loss": 1.2559,
"step": 566
},
{
"epoch": 2.2967088607594937,
"grad_norm": 0.6131203911295449,
"learning_rate": 2.6115527325180754e-05,
"loss": 1.2347,
"step": 567
},
{
"epoch": 2.300759493670886,
"grad_norm": 0.5728238220793456,
"learning_rate": 2.606146264569603e-05,
"loss": 1.2444,
"step": 568
},
{
"epoch": 2.3048101265822787,
"grad_norm": 0.5477763488881782,
"learning_rate": 2.6007349148024447e-05,
"loss": 1.2898,
"step": 569
},
{
"epoch": 2.3088607594936708,
"grad_norm": 0.5947759775491406,
"learning_rate": 2.5953187267988694e-05,
"loss": 1.262,
"step": 570
},
{
"epoch": 2.3129113924050633,
"grad_norm": 0.6084875156410368,
"learning_rate": 2.5898977441801097e-05,
"loss": 1.2287,
"step": 571
},
{
"epoch": 2.316962025316456,
"grad_norm": 0.5749918047927587,
"learning_rate": 2.584472010606015e-05,
"loss": 1.2264,
"step": 572
},
{
"epoch": 2.321012658227848,
"grad_norm": 0.6343000630785216,
"learning_rate": 2.5790415697746976e-05,
"loss": 1.2553,
"step": 573
},
{
"epoch": 2.3250632911392404,
"grad_norm": 0.5735795788431596,
"learning_rate": 2.5736064654221808e-05,
"loss": 1.2186,
"step": 574
},
{
"epoch": 2.329113924050633,
"grad_norm": 0.5543918426149519,
"learning_rate": 2.568166741322048e-05,
"loss": 1.2756,
"step": 575
},
{
"epoch": 2.3331645569620254,
"grad_norm": 0.5651553418815803,
"learning_rate": 2.56272244128509e-05,
"loss": 1.2583,
"step": 576
},
{
"epoch": 2.337215189873418,
"grad_norm": 0.547007377417892,
"learning_rate": 2.55727360915895e-05,
"loss": 1.2606,
"step": 577
},
{
"epoch": 2.34126582278481,
"grad_norm": 0.5849756836004916,
"learning_rate": 2.5518202888277734e-05,
"loss": 1.2381,
"step": 578
},
{
"epoch": 2.3453164556962025,
"grad_norm": 0.6391157487952199,
"learning_rate": 2.5463625242118523e-05,
"loss": 1.2621,
"step": 579
},
{
"epoch": 2.349367088607595,
"grad_norm": 0.6733023543922348,
"learning_rate": 2.5409003592672723e-05,
"loss": 1.2891,
"step": 580
},
{
"epoch": 2.353417721518987,
"grad_norm": 0.5973591276936419,
"learning_rate": 2.535433837985559e-05,
"loss": 1.2855,
"step": 581
},
{
"epoch": 2.3574683544303796,
"grad_norm": 0.6818027525124967,
"learning_rate": 2.529963004393324e-05,
"loss": 1.2447,
"step": 582
},
{
"epoch": 2.361518987341772,
"grad_norm": 0.6232459793673589,
"learning_rate": 2.524487902551908e-05,
"loss": 1.2541,
"step": 583
},
{
"epoch": 2.3655696202531646,
"grad_norm": 0.49944948515694404,
"learning_rate": 2.519008576557029e-05,
"loss": 1.2314,
"step": 584
},
{
"epoch": 2.369620253164557,
"grad_norm": 0.6560528765991742,
"learning_rate": 2.5135250705384254e-05,
"loss": 1.264,
"step": 585
},
{
"epoch": 2.3736708860759492,
"grad_norm": 0.7047146622174902,
"learning_rate": 2.5080374286595007e-05,
"loss": 1.2299,
"step": 586
},
{
"epoch": 2.3777215189873417,
"grad_norm": 0.5058974488535803,
"learning_rate": 2.5025456951169677e-05,
"loss": 1.2375,
"step": 587
},
{
"epoch": 2.3817721518987343,
"grad_norm": 0.7391197630305786,
"learning_rate": 2.4970499141404942e-05,
"loss": 1.2759,
"step": 588
},
{
"epoch": 2.3858227848101268,
"grad_norm": 0.5832611201897978,
"learning_rate": 2.491550129992345e-05,
"loss": 1.2479,
"step": 589
},
{
"epoch": 2.389873417721519,
"grad_norm": 0.6903244380547484,
"learning_rate": 2.486046386967024e-05,
"loss": 1.2409,
"step": 590
},
{
"epoch": 2.3939240506329114,
"grad_norm": 0.5472602889002922,
"learning_rate": 2.4805387293909214e-05,
"loss": 1.2337,
"step": 591
},
{
"epoch": 2.397974683544304,
"grad_norm": 0.6842725057634984,
"learning_rate": 2.4750272016219552e-05,
"loss": 1.2713,
"step": 592
},
{
"epoch": 2.4020253164556964,
"grad_norm": 0.5974362505060102,
"learning_rate": 2.4695118480492114e-05,
"loss": 1.257,
"step": 593
},
{
"epoch": 2.4060759493670885,
"grad_norm": 0.5402688701629411,
"learning_rate": 2.4639927130925898e-05,
"loss": 1.2701,
"step": 594
},
{
"epoch": 2.410126582278481,
"grad_norm": 0.570910151720766,
"learning_rate": 2.458469841202444e-05,
"loss": 1.2599,
"step": 595
},
{
"epoch": 2.4141772151898735,
"grad_norm": 0.6169391698579857,
"learning_rate": 2.452943276859226e-05,
"loss": 1.2849,
"step": 596
},
{
"epoch": 2.418227848101266,
"grad_norm": 0.6028454558736777,
"learning_rate": 2.447413064573125e-05,
"loss": 1.2397,
"step": 597
},
{
"epoch": 2.422278481012658,
"grad_norm": 0.5810737920034796,
"learning_rate": 2.4418792488837095e-05,
"loss": 1.2207,
"step": 598
},
{
"epoch": 2.4263291139240506,
"grad_norm": 0.6051293918861065,
"learning_rate": 2.4363418743595713e-05,
"loss": 1.2553,
"step": 599
},
{
"epoch": 2.430379746835443,
"grad_norm": 0.5920020760830135,
"learning_rate": 2.430800985597963e-05,
"loss": 1.2114,
"step": 600
},
{
"epoch": 2.4344303797468356,
"grad_norm": 0.551473543514109,
"learning_rate": 2.4252566272244415e-05,
"loss": 1.2441,
"step": 601
},
{
"epoch": 2.4384810126582277,
"grad_norm": 0.5257989740813591,
"learning_rate": 2.4197088438925063e-05,
"loss": 1.2167,
"step": 602
},
{
"epoch": 2.44253164556962,
"grad_norm": 0.5506786090960738,
"learning_rate": 2.4141576802832417e-05,
"loss": 1.2127,
"step": 603
},
{
"epoch": 2.4465822784810127,
"grad_norm": 0.5107879454710212,
"learning_rate": 2.408603181104957e-05,
"loss": 1.2475,
"step": 604
},
{
"epoch": 2.4506329113924052,
"grad_norm": 0.6177020125514245,
"learning_rate": 2.4030453910928245e-05,
"loss": 1.2478,
"step": 605
},
{
"epoch": 2.4546835443037973,
"grad_norm": 0.4826952261288295,
"learning_rate": 2.397484355008521e-05,
"loss": 1.2238,
"step": 606
},
{
"epoch": 2.45873417721519,
"grad_norm": 0.6567403500642346,
"learning_rate": 2.3919201176398662e-05,
"loss": 1.2349,
"step": 607
},
{
"epoch": 2.4627848101265823,
"grad_norm": 0.5193524847687772,
"learning_rate": 2.3863527238004633e-05,
"loss": 1.2423,
"step": 608
},
{
"epoch": 2.466835443037975,
"grad_norm": 0.6586307429532519,
"learning_rate": 2.380782218329337e-05,
"loss": 1.2621,
"step": 609
},
{
"epoch": 2.470886075949367,
"grad_norm": 0.5462175654421538,
"learning_rate": 2.3752086460905725e-05,
"loss": 1.2835,
"step": 610
},
{
"epoch": 2.4749367088607594,
"grad_norm": 0.6194893958893987,
"learning_rate": 2.3696320519729544e-05,
"loss": 1.2234,
"step": 611
},
{
"epoch": 2.478987341772152,
"grad_norm": 0.6886466156962265,
"learning_rate": 2.3640524808896045e-05,
"loss": 1.2568,
"step": 612
},
{
"epoch": 2.4830379746835445,
"grad_norm": 0.5671750523554887,
"learning_rate": 2.3584699777776222e-05,
"loss": 1.2988,
"step": 613
},
{
"epoch": 2.4870886075949366,
"grad_norm": 0.6217078057791634,
"learning_rate": 2.3528845875977195e-05,
"loss": 1.2376,
"step": 614
},
{
"epoch": 2.491139240506329,
"grad_norm": 0.5619748253425767,
"learning_rate": 2.3472963553338614e-05,
"loss": 1.2814,
"step": 615
},
{
"epoch": 2.4951898734177216,
"grad_norm": 0.6630574316375283,
"learning_rate": 2.341705325992901e-05,
"loss": 1.255,
"step": 616
},
{
"epoch": 2.499240506329114,
"grad_norm": 0.5007768392136278,
"learning_rate": 2.336111544604222e-05,
"loss": 1.2349,
"step": 617
},
{
"epoch": 2.503291139240506,
"grad_norm": 0.5766587285871668,
"learning_rate": 2.33051505621937e-05,
"loss": 1.2672,
"step": 618
},
{
"epoch": 2.5073417721518987,
"grad_norm": 0.5800104285892583,
"learning_rate": 2.324915905911693e-05,
"loss": 1.2541,
"step": 619
},
{
"epoch": 2.511392405063291,
"grad_norm": 0.5654957792766403,
"learning_rate": 2.319314138775977e-05,
"loss": 1.2379,
"step": 620
},
{
"epoch": 2.5154430379746833,
"grad_norm": 0.5997270662064565,
"learning_rate": 2.3137097999280856e-05,
"loss": 1.2595,
"step": 621
},
{
"epoch": 2.519493670886076,
"grad_norm": 0.5492554321357106,
"learning_rate": 2.308102934504593e-05,
"loss": 1.2513,
"step": 622
},
{
"epoch": 2.5235443037974683,
"grad_norm": 0.5841025333399301,
"learning_rate": 2.3024935876624222e-05,
"loss": 1.2283,
"step": 623
},
{
"epoch": 2.527594936708861,
"grad_norm": 0.6746226790286795,
"learning_rate": 2.2968818045784813e-05,
"loss": 1.2262,
"step": 624
},
{
"epoch": 2.5316455696202533,
"grad_norm": 0.6709221505144375,
"learning_rate": 2.2912676304493006e-05,
"loss": 1.2357,
"step": 625
},
{
"epoch": 2.5356962025316454,
"grad_norm": 0.6031217929007998,
"learning_rate": 2.2856511104906668e-05,
"loss": 1.2554,
"step": 626
},
{
"epoch": 2.539746835443038,
"grad_norm": 0.6191497505061805,
"learning_rate": 2.2800322899372586e-05,
"loss": 1.241,
"step": 627
},
{
"epoch": 2.5437974683544304,
"grad_norm": 0.6438302982215388,
"learning_rate": 2.2744112140422844e-05,
"loss": 1.2321,
"step": 628
},
{
"epoch": 2.547848101265823,
"grad_norm": 0.5973790571824865,
"learning_rate": 2.2687879280771177e-05,
"loss": 1.2742,
"step": 629
},
{
"epoch": 2.5518987341772155,
"grad_norm": 0.5619748131934528,
"learning_rate": 2.26316247733093e-05,
"loss": 1.2987,
"step": 630
},
{
"epoch": 2.5559493670886075,
"grad_norm": 0.6403346945231175,
"learning_rate": 2.257534907110328e-05,
"loss": 1.2229,
"step": 631
},
{
"epoch": 2.56,
"grad_norm": 0.5676942210194226,
"learning_rate": 2.2519052627389882e-05,
"loss": 1.2834,
"step": 632
},
{
"epoch": 2.5640506329113926,
"grad_norm": 0.6334761094233862,
"learning_rate": 2.246273589557294e-05,
"loss": 1.306,
"step": 633
},
{
"epoch": 2.5681012658227846,
"grad_norm": 0.6387216893544679,
"learning_rate": 2.240639932921966e-05,
"loss": 1.2556,
"step": 634
},
{
"epoch": 2.572151898734177,
"grad_norm": 0.6880886975217463,
"learning_rate": 2.2350043382056995e-05,
"loss": 1.2338,
"step": 635
},
{
"epoch": 2.5762025316455697,
"grad_norm": 0.6480464328513991,
"learning_rate": 2.2293668507968015e-05,
"loss": 1.2793,
"step": 636
},
{
"epoch": 2.580253164556962,
"grad_norm": 0.7351559085947702,
"learning_rate": 2.2237275160988186e-05,
"loss": 1.2325,
"step": 637
},
{
"epoch": 2.5843037974683547,
"grad_norm": 0.48856148736811533,
"learning_rate": 2.2180863795301787e-05,
"loss": 1.2369,
"step": 638
},
{
"epoch": 2.5883544303797468,
"grad_norm": 0.7851827380695394,
"learning_rate": 2.212443486523819e-05,
"loss": 1.2713,
"step": 639
},
{
"epoch": 2.5924050632911393,
"grad_norm": 0.6021784291347317,
"learning_rate": 2.2067988825268243e-05,
"loss": 1.2122,
"step": 640
},
{
"epoch": 2.596455696202532,
"grad_norm": 0.6326216361779337,
"learning_rate": 2.2011526130000596e-05,
"loss": 1.2898,
"step": 641
},
{
"epoch": 2.600506329113924,
"grad_norm": 0.6405529975534646,
"learning_rate": 2.1955047234178038e-05,
"loss": 1.2322,
"step": 642
},
{
"epoch": 2.6045569620253164,
"grad_norm": 0.5799399852116268,
"learning_rate": 2.1898552592673825e-05,
"loss": 1.2741,
"step": 643
},
{
"epoch": 2.608607594936709,
"grad_norm": 0.6970527340163166,
"learning_rate": 2.184204266048803e-05,
"loss": 1.2248,
"step": 644
},
{
"epoch": 2.6126582278481014,
"grad_norm": 0.6113898067002624,
"learning_rate": 2.1785517892743887e-05,
"loss": 1.2649,
"step": 645
},
{
"epoch": 2.616708860759494,
"grad_norm": 0.7080844538824044,
"learning_rate": 2.17289787446841e-05,
"loss": 1.2351,
"step": 646
},
{
"epoch": 2.620759493670886,
"grad_norm": 0.5730967407301544,
"learning_rate": 2.1672425671667198e-05,
"loss": 1.2619,
"step": 647
},
{
"epoch": 2.6248101265822785,
"grad_norm": 0.729675124845355,
"learning_rate": 2.161585912916385e-05,
"loss": 1.2455,
"step": 648
},
{
"epoch": 2.628860759493671,
"grad_norm": 0.624247023508424,
"learning_rate": 2.1559279572753214e-05,
"loss": 1.2229,
"step": 649
},
{
"epoch": 2.632911392405063,
"grad_norm": 0.6174979801623363,
"learning_rate": 2.1502687458119268e-05,
"loss": 1.2464,
"step": 650
},
{
"epoch": 2.6369620253164556,
"grad_norm": 0.5372339903973433,
"learning_rate": 2.1446083241047116e-05,
"loss": 1.2683,
"step": 651
},
{
"epoch": 2.641012658227848,
"grad_norm": 0.6808767299391796,
"learning_rate": 2.1389467377419333e-05,
"loss": 1.2934,
"step": 652
},
{
"epoch": 2.6450632911392407,
"grad_norm": 0.6438327228504789,
"learning_rate": 2.133284032321232e-05,
"loss": 1.2566,
"step": 653
},
{
"epoch": 2.649113924050633,
"grad_norm": 0.6068123355882361,
"learning_rate": 2.1276202534492566e-05,
"loss": 1.2864,
"step": 654
},
{
"epoch": 2.6531645569620252,
"grad_norm": 0.6031720728015478,
"learning_rate": 2.121955446741306e-05,
"loss": 1.2067,
"step": 655
},
{
"epoch": 2.6572151898734178,
"grad_norm": 0.5499112573103984,
"learning_rate": 2.1162896578209517e-05,
"loss": 1.2474,
"step": 656
},
{
"epoch": 2.6612658227848103,
"grad_norm": 0.5013772259633164,
"learning_rate": 2.1106229323196813e-05,
"loss": 1.2398,
"step": 657
},
{
"epoch": 2.6653164556962023,
"grad_norm": 0.6260491724685034,
"learning_rate": 2.1049553158765214e-05,
"loss": 1.27,
"step": 658
},
{
"epoch": 2.669367088607595,
"grad_norm": 0.4998149885166331,
"learning_rate": 2.0992868541376764e-05,
"loss": 1.245,
"step": 659
},
{
"epoch": 2.6734177215189874,
"grad_norm": 0.5511438928447903,
"learning_rate": 2.093617592756158e-05,
"loss": 1.2802,
"step": 660
},
{
"epoch": 2.67746835443038,
"grad_norm": 0.6428620655180537,
"learning_rate": 2.0879475773914167e-05,
"loss": 1.285,
"step": 661
},
{
"epoch": 2.6815189873417724,
"grad_norm": 0.5825979888552961,
"learning_rate": 2.082276853708978e-05,
"loss": 1.2683,
"step": 662
},
{
"epoch": 2.6855696202531645,
"grad_norm": 0.5636037812558418,
"learning_rate": 2.076605467380071e-05,
"loss": 1.2571,
"step": 663
},
{
"epoch": 2.689620253164557,
"grad_norm": 0.6024740487447374,
"learning_rate": 2.0709334640812613e-05,
"loss": 1.2531,
"step": 664
},
{
"epoch": 2.6936708860759495,
"grad_norm": 0.49818304532633756,
"learning_rate": 2.0652608894940824e-05,
"loss": 1.2103,
"step": 665
},
{
"epoch": 2.6977215189873416,
"grad_norm": 0.53584003715082,
"learning_rate": 2.0595877893046722e-05,
"loss": 1.2766,
"step": 666
},
{
"epoch": 2.701772151898734,
"grad_norm": 0.5687194446768747,
"learning_rate": 2.0539142092033985e-05,
"loss": 1.2713,
"step": 667
},
{
"epoch": 2.7058227848101266,
"grad_norm": 0.5057362171708076,
"learning_rate": 2.048240194884496e-05,
"loss": 1.2502,
"step": 668
},
{
"epoch": 2.709873417721519,
"grad_norm": 0.5245849383718579,
"learning_rate": 2.042565792045695e-05,
"loss": 1.2281,
"step": 669
},
{
"epoch": 2.7139240506329116,
"grad_norm": 0.4814376255103471,
"learning_rate": 2.036891046387857e-05,
"loss": 1.2676,
"step": 670
},
{
"epoch": 2.7179746835443037,
"grad_norm": 0.5160669830003197,
"learning_rate": 2.0312160036146036e-05,
"loss": 1.2733,
"step": 671
},
{
"epoch": 2.7220253164556962,
"grad_norm": 0.4788407856616411,
"learning_rate": 2.025540709431948e-05,
"loss": 1.2141,
"step": 672
},
{
"epoch": 2.7260759493670887,
"grad_norm": 0.5377547332723315,
"learning_rate": 2.0198652095479298e-05,
"loss": 1.2647,
"step": 673
},
{
"epoch": 2.730126582278481,
"grad_norm": 0.5457719318333892,
"learning_rate": 2.014189549672245e-05,
"loss": 1.276,
"step": 674
},
{
"epoch": 2.7341772151898733,
"grad_norm": 0.4943201503276657,
"learning_rate": 2.0085137755158776e-05,
"loss": 1.2413,
"step": 675
},
{
"epoch": 2.738227848101266,
"grad_norm": 0.578636126272599,
"learning_rate": 2.0028379327907327e-05,
"loss": 1.239,
"step": 676
},
{
"epoch": 2.7422784810126584,
"grad_norm": 0.4923189641106431,
"learning_rate": 1.9971620672092676e-05,
"loss": 1.2108,
"step": 677
},
{
"epoch": 2.746329113924051,
"grad_norm": 0.5278868124207688,
"learning_rate": 1.991486224484123e-05,
"loss": 1.2503,
"step": 678
},
{
"epoch": 2.750379746835443,
"grad_norm": 0.5450656910000219,
"learning_rate": 1.985810450327756e-05,
"loss": 1.2279,
"step": 679
},
{
"epoch": 2.7544303797468355,
"grad_norm": 0.49768554748811433,
"learning_rate": 1.9801347904520706e-05,
"loss": 1.2685,
"step": 680
},
{
"epoch": 2.758481012658228,
"grad_norm": 0.5480917135731539,
"learning_rate": 1.974459290568053e-05,
"loss": 1.276,
"step": 681
},
{
"epoch": 2.76253164556962,
"grad_norm": 0.607867137192176,
"learning_rate": 1.968783996385397e-05,
"loss": 1.2681,
"step": 682
},
{
"epoch": 2.7665822784810126,
"grad_norm": 0.5267925567505597,
"learning_rate": 1.963108953612143e-05,
"loss": 1.2542,
"step": 683
},
{
"epoch": 2.770632911392405,
"grad_norm": 0.4854525871332087,
"learning_rate": 1.9574342079543056e-05,
"loss": 1.2202,
"step": 684
},
{
"epoch": 2.7746835443037976,
"grad_norm": 0.5271265705719412,
"learning_rate": 1.9517598051155046e-05,
"loss": 1.2371,
"step": 685
},
{
"epoch": 2.77873417721519,
"grad_norm": 0.4953595608196222,
"learning_rate": 1.9460857907966025e-05,
"loss": 1.2617,
"step": 686
},
{
"epoch": 2.782784810126582,
"grad_norm": 0.48894544914353966,
"learning_rate": 1.9404122106953285e-05,
"loss": 1.2618,
"step": 687
},
{
"epoch": 2.7868354430379747,
"grad_norm": 0.5370988587477935,
"learning_rate": 1.9347391105059176e-05,
"loss": 1.2619,
"step": 688
},
{
"epoch": 2.790886075949367,
"grad_norm": 0.4897080896753555,
"learning_rate": 1.92906653591874e-05,
"loss": 1.2426,
"step": 689
},
{
"epoch": 2.7949367088607593,
"grad_norm": 0.5748831040885557,
"learning_rate": 1.9233945326199295e-05,
"loss": 1.2809,
"step": 690
},
{
"epoch": 2.798987341772152,
"grad_norm": 0.5158934853434705,
"learning_rate": 1.917723146291022e-05,
"loss": 1.2652,
"step": 691
},
{
"epoch": 2.8030379746835443,
"grad_norm": 0.5645165160720521,
"learning_rate": 1.912052422608584e-05,
"loss": 1.274,
"step": 692
},
{
"epoch": 2.807088607594937,
"grad_norm": 0.4849470229960467,
"learning_rate": 1.9063824072438428e-05,
"loss": 1.2483,
"step": 693
},
{
"epoch": 2.8111392405063294,
"grad_norm": 0.487342385651646,
"learning_rate": 1.9007131458623246e-05,
"loss": 1.234,
"step": 694
},
{
"epoch": 2.8151898734177214,
"grad_norm": 0.4971539794601758,
"learning_rate": 1.895044684123479e-05,
"loss": 1.2428,
"step": 695
},
{
"epoch": 2.819240506329114,
"grad_norm": 0.49846278544566036,
"learning_rate": 1.8893770676803194e-05,
"loss": 1.2685,
"step": 696
},
{
"epoch": 2.8232911392405065,
"grad_norm": 0.5735552451353678,
"learning_rate": 1.8837103421790486e-05,
"loss": 1.273,
"step": 697
},
{
"epoch": 2.8273417721518985,
"grad_norm": 0.4683367797546678,
"learning_rate": 1.8780445532586952e-05,
"loss": 1.2779,
"step": 698
},
{
"epoch": 2.831392405063291,
"grad_norm": 0.5779065972423325,
"learning_rate": 1.872379746550743e-05,
"loss": 1.2613,
"step": 699
},
{
"epoch": 2.8354430379746836,
"grad_norm": 0.5232124489594128,
"learning_rate": 1.866715967678769e-05,
"loss": 1.2871,
"step": 700
},
{
"epoch": 2.839493670886076,
"grad_norm": 0.5406604756922285,
"learning_rate": 1.861053262258067e-05,
"loss": 1.2119,
"step": 701
},
{
"epoch": 2.8435443037974686,
"grad_norm": 0.4890555469280865,
"learning_rate": 1.8553916758952897e-05,
"loss": 1.2184,
"step": 702
},
{
"epoch": 2.8475949367088607,
"grad_norm": 0.5395462105085046,
"learning_rate": 1.8497312541880735e-05,
"loss": 1.2493,
"step": 703
},
{
"epoch": 2.851645569620253,
"grad_norm": 0.5231946888063219,
"learning_rate": 1.8440720427246786e-05,
"loss": 1.2657,
"step": 704
},
{
"epoch": 2.8556962025316457,
"grad_norm": 0.45896783435951366,
"learning_rate": 1.8384140870836157e-05,
"loss": 1.2703,
"step": 705
},
{
"epoch": 2.8597468354430378,
"grad_norm": 0.551497707240939,
"learning_rate": 1.8327574328332806e-05,
"loss": 1.2707,
"step": 706
},
{
"epoch": 2.8637974683544303,
"grad_norm": 0.5039676172331792,
"learning_rate": 1.8271021255315906e-05,
"loss": 1.2536,
"step": 707
},
{
"epoch": 2.867848101265823,
"grad_norm": 0.5225019301373928,
"learning_rate": 1.8214482107256117e-05,
"loss": 1.2658,
"step": 708
},
{
"epoch": 2.8718987341772153,
"grad_norm": 0.4727318721163185,
"learning_rate": 1.8157957339511968e-05,
"loss": 1.2433,
"step": 709
},
{
"epoch": 2.875949367088608,
"grad_norm": 0.5548752492015276,
"learning_rate": 1.8101447407326182e-05,
"loss": 1.2417,
"step": 710
},
{
"epoch": 2.88,
"grad_norm": 0.5953176538070245,
"learning_rate": 1.8044952765821966e-05,
"loss": 1.2879,
"step": 711
},
{
"epoch": 2.8840506329113924,
"grad_norm": 0.49601465778635595,
"learning_rate": 1.7988473869999407e-05,
"loss": 1.2435,
"step": 712
},
{
"epoch": 2.888101265822785,
"grad_norm": 0.5953594807705938,
"learning_rate": 1.7932011174731764e-05,
"loss": 1.2545,
"step": 713
},
{
"epoch": 2.892151898734177,
"grad_norm": 0.5447211814878272,
"learning_rate": 1.7875565134761817e-05,
"loss": 1.2689,
"step": 714
},
{
"epoch": 2.8962025316455695,
"grad_norm": 0.5299267164369422,
"learning_rate": 1.7819136204698226e-05,
"loss": 1.2427,
"step": 715
},
{
"epoch": 2.900253164556962,
"grad_norm": 0.4575793868871151,
"learning_rate": 1.776272483901182e-05,
"loss": 1.2609,
"step": 716
},
{
"epoch": 2.9043037974683545,
"grad_norm": 0.5171706179625555,
"learning_rate": 1.7706331492031995e-05,
"loss": 1.2548,
"step": 717
},
{
"epoch": 2.908354430379747,
"grad_norm": 0.45051789617938665,
"learning_rate": 1.764995661794301e-05,
"loss": 1.2536,
"step": 718
},
{
"epoch": 2.912405063291139,
"grad_norm": 0.5210087979184251,
"learning_rate": 1.759360067078035e-05,
"loss": 1.2117,
"step": 719
},
{
"epoch": 2.9164556962025316,
"grad_norm": 0.5338644831217907,
"learning_rate": 1.7537264104427064e-05,
"loss": 1.2487,
"step": 720
},
{
"epoch": 2.920506329113924,
"grad_norm": 0.5358377516892175,
"learning_rate": 1.748094737261012e-05,
"loss": 1.2799,
"step": 721
},
{
"epoch": 2.9245569620253162,
"grad_norm": 0.48870305306252465,
"learning_rate": 1.7424650928896726e-05,
"loss": 1.2483,
"step": 722
},
{
"epoch": 2.9286075949367087,
"grad_norm": 0.566522630557182,
"learning_rate": 1.7368375226690712e-05,
"loss": 1.2487,
"step": 723
},
{
"epoch": 2.9326582278481013,
"grad_norm": 0.47528716738240223,
"learning_rate": 1.731212071922883e-05,
"loss": 1.2651,
"step": 724
},
{
"epoch": 2.9367088607594938,
"grad_norm": 0.5293292852435424,
"learning_rate": 1.7255887859577156e-05,
"loss": 1.2606,
"step": 725
},
{
"epoch": 2.9407594936708863,
"grad_norm": 0.5407973849390401,
"learning_rate": 1.7199677100627427e-05,
"loss": 1.27,
"step": 726
},
{
"epoch": 2.9448101265822784,
"grad_norm": 0.5915609067984887,
"learning_rate": 1.7143488895093343e-05,
"loss": 1.2211,
"step": 727
},
{
"epoch": 2.948860759493671,
"grad_norm": 0.528447713265915,
"learning_rate": 1.7087323695506994e-05,
"loss": 1.2638,
"step": 728
},
{
"epoch": 2.9529113924050634,
"grad_norm": 0.559352199627487,
"learning_rate": 1.7031181954215194e-05,
"loss": 1.2463,
"step": 729
},
{
"epoch": 2.9569620253164555,
"grad_norm": 0.6913377657282361,
"learning_rate": 1.6975064123375788e-05,
"loss": 1.2707,
"step": 730
},
{
"epoch": 2.961012658227848,
"grad_norm": 0.5397501108249637,
"learning_rate": 1.6918970654954084e-05,
"loss": 1.2282,
"step": 731
},
{
"epoch": 2.9650632911392405,
"grad_norm": 0.6452570644968646,
"learning_rate": 1.686290200071915e-05,
"loss": 1.2686,
"step": 732
},
{
"epoch": 2.969113924050633,
"grad_norm": 0.5303985140839327,
"learning_rate": 1.6806858612240234e-05,
"loss": 1.2127,
"step": 733
},
{
"epoch": 2.9731645569620255,
"grad_norm": 0.5951985011835034,
"learning_rate": 1.6750840940883078e-05,
"loss": 1.2821,
"step": 734
},
{
"epoch": 2.9772151898734176,
"grad_norm": 0.536563110242829,
"learning_rate": 1.6694849437806305e-05,
"loss": 1.2375,
"step": 735
},
{
"epoch": 2.98126582278481,
"grad_norm": 0.5676433346159683,
"learning_rate": 1.663888455395778e-05,
"loss": 1.2383,
"step": 736
},
{
"epoch": 2.9853164556962026,
"grad_norm": 0.599767173560357,
"learning_rate": 1.6582946740070995e-05,
"loss": 1.2699,
"step": 737
},
{
"epoch": 2.9893670886075947,
"grad_norm": 0.5602603316435527,
"learning_rate": 1.6527036446661396e-05,
"loss": 1.2471,
"step": 738
},
{
"epoch": 2.993417721518987,
"grad_norm": 0.5683892762887245,
"learning_rate": 1.6471154124022818e-05,
"loss": 1.2349,
"step": 739
},
{
"epoch": 2.9974683544303797,
"grad_norm": 0.5890836225681715,
"learning_rate": 1.6415300222223788e-05,
"loss": 1.2391,
"step": 740
},
{
"epoch": 3.0015189873417722,
"grad_norm": 0.6208202319010283,
"learning_rate": 1.6359475191103958e-05,
"loss": 1.1901,
"step": 741
},
{
"epoch": 3.0055696202531648,
"grad_norm": 0.7510303406014098,
"learning_rate": 1.6303679480270466e-05,
"loss": 1.0915,
"step": 742
},
{
"epoch": 3.009620253164557,
"grad_norm": 0.926321357800325,
"learning_rate": 1.624791353909428e-05,
"loss": 1.1402,
"step": 743
},
{
"epoch": 3.0136708860759494,
"grad_norm": 0.8586214061890556,
"learning_rate": 1.619217781670663e-05,
"loss": 1.1011,
"step": 744
},
{
"epoch": 3.017721518987342,
"grad_norm": 0.9934187031578188,
"learning_rate": 1.6136472761995373e-05,
"loss": 1.0465,
"step": 745
},
{
"epoch": 3.0217721518987344,
"grad_norm": 0.8870441522496558,
"learning_rate": 1.608079882360134e-05,
"loss": 1.1052,
"step": 746
},
{
"epoch": 3.0258227848101265,
"grad_norm": 0.702287607724995,
"learning_rate": 1.60251564499148e-05,
"loss": 1.0809,
"step": 747
},
{
"epoch": 3.029873417721519,
"grad_norm": 0.9978168192598335,
"learning_rate": 1.596954608907176e-05,
"loss": 1.1022,
"step": 748
},
{
"epoch": 3.0339240506329115,
"grad_norm": 0.8814499350017357,
"learning_rate": 1.591396818895043e-05,
"loss": 1.09,
"step": 749
},
{
"epoch": 3.037974683544304,
"grad_norm": 0.8727806954737071,
"learning_rate": 1.585842319716759e-05,
"loss": 1.0624,
"step": 750
},
{
"epoch": 3.042025316455696,
"grad_norm": 0.7939607053192566,
"learning_rate": 1.5802911561074944e-05,
"loss": 1.0963,
"step": 751
},
{
"epoch": 3.0460759493670886,
"grad_norm": 0.9410079791090099,
"learning_rate": 1.5747433727755595e-05,
"loss": 1.1102,
"step": 752
},
{
"epoch": 3.050126582278481,
"grad_norm": 0.8723055859540828,
"learning_rate": 1.5691990144020376e-05,
"loss": 1.0613,
"step": 753
},
{
"epoch": 3.0541772151898736,
"grad_norm": 0.801624650891849,
"learning_rate": 1.5636581256404297e-05,
"loss": 1.0587,
"step": 754
},
{
"epoch": 3.0582278481012657,
"grad_norm": 0.9207407573769101,
"learning_rate": 1.558120751116291e-05,
"loss": 1.093,
"step": 755
},
{
"epoch": 3.062278481012658,
"grad_norm": 0.7399775760734916,
"learning_rate": 1.552586935426876e-05,
"loss": 1.1157,
"step": 756
},
{
"epoch": 3.0663291139240507,
"grad_norm": 0.7988335136019642,
"learning_rate": 1.547056723140774e-05,
"loss": 1.07,
"step": 757
},
{
"epoch": 3.0703797468354432,
"grad_norm": 0.7690914934778039,
"learning_rate": 1.5415301587975565e-05,
"loss": 1.0787,
"step": 758
},
{
"epoch": 3.0744303797468353,
"grad_norm": 0.6643284150485221,
"learning_rate": 1.536007286907411e-05,
"loss": 1.0899,
"step": 759
},
{
"epoch": 3.078481012658228,
"grad_norm": 0.7098213085812929,
"learning_rate": 1.5304881519507896e-05,
"loss": 1.0625,
"step": 760
},
{
"epoch": 3.0825316455696203,
"grad_norm": 0.8289433600419651,
"learning_rate": 1.5249727983780453e-05,
"loss": 1.1265,
"step": 761
},
{
"epoch": 3.086582278481013,
"grad_norm": 0.5710708225124373,
"learning_rate": 1.5194612706090786e-05,
"loss": 1.0654,
"step": 762
},
{
"epoch": 3.090632911392405,
"grad_norm": 0.665716263252891,
"learning_rate": 1.5139536130329771e-05,
"loss": 1.1019,
"step": 763
},
{
"epoch": 3.0946835443037974,
"grad_norm": 0.7379462582298566,
"learning_rate": 1.508449870007656e-05,
"loss": 1.0942,
"step": 764
},
{
"epoch": 3.09873417721519,
"grad_norm": 0.5934542683610728,
"learning_rate": 1.5029500858595056e-05,
"loss": 1.1071,
"step": 765
},
{
"epoch": 3.1027848101265825,
"grad_norm": 0.6033060200633277,
"learning_rate": 1.4974543048830328e-05,
"loss": 1.1092,
"step": 766
},
{
"epoch": 3.1068354430379745,
"grad_norm": 0.6492772254030309,
"learning_rate": 1.4919625713405e-05,
"loss": 1.0578,
"step": 767
},
{
"epoch": 3.110886075949367,
"grad_norm": 0.6474286002023519,
"learning_rate": 1.4864749294615756e-05,
"loss": 1.092,
"step": 768
},
{
"epoch": 3.1149367088607596,
"grad_norm": 0.5778631204408031,
"learning_rate": 1.4809914234429716e-05,
"loss": 1.1108,
"step": 769
},
{
"epoch": 3.118987341772152,
"grad_norm": 0.6787444230747895,
"learning_rate": 1.4755120974480923e-05,
"loss": 1.0673,
"step": 770
},
{
"epoch": 3.123037974683544,
"grad_norm": 0.6490865475297352,
"learning_rate": 1.4700369956066771e-05,
"loss": 1.1065,
"step": 771
},
{
"epoch": 3.1270886075949367,
"grad_norm": 0.5637974628443285,
"learning_rate": 1.4645661620144413e-05,
"loss": 1.0838,
"step": 772
},
{
"epoch": 3.131139240506329,
"grad_norm": 0.6282495878345422,
"learning_rate": 1.4590996407327284e-05,
"loss": 1.1104,
"step": 773
},
{
"epoch": 3.1351898734177217,
"grad_norm": 0.5459306644499987,
"learning_rate": 1.4536374757881487e-05,
"loss": 1.1048,
"step": 774
},
{
"epoch": 3.1392405063291138,
"grad_norm": 0.600800613648533,
"learning_rate": 1.4481797111722271e-05,
"loss": 1.0982,
"step": 775
},
{
"epoch": 3.1432911392405063,
"grad_norm": 0.5144989694828929,
"learning_rate": 1.4427263908410507e-05,
"loss": 1.068,
"step": 776
},
{
"epoch": 3.147341772151899,
"grad_norm": 0.6272640602916604,
"learning_rate": 1.4372775587149108e-05,
"loss": 1.1057,
"step": 777
},
{
"epoch": 3.1513924050632913,
"grad_norm": 0.49438709847584444,
"learning_rate": 1.4318332586779522e-05,
"loss": 1.0721,
"step": 778
},
{
"epoch": 3.1554430379746834,
"grad_norm": 0.5981434561320972,
"learning_rate": 1.4263935345778202e-05,
"loss": 1.1064,
"step": 779
},
{
"epoch": 3.159493670886076,
"grad_norm": 0.544716161488482,
"learning_rate": 1.420958430225303e-05,
"loss": 1.0852,
"step": 780
},
{
"epoch": 3.1635443037974684,
"grad_norm": 0.6191079650381538,
"learning_rate": 1.415527989393985e-05,
"loss": 1.1002,
"step": 781
},
{
"epoch": 3.167594936708861,
"grad_norm": 0.5585664976706268,
"learning_rate": 1.410102255819891e-05,
"loss": 1.0925,
"step": 782
},
{
"epoch": 3.171645569620253,
"grad_norm": 0.5689487728889705,
"learning_rate": 1.404681273201131e-05,
"loss": 1.0867,
"step": 783
},
{
"epoch": 3.1756962025316455,
"grad_norm": 0.5903068113602556,
"learning_rate": 1.399265085197556e-05,
"loss": 1.1354,
"step": 784
},
{
"epoch": 3.179746835443038,
"grad_norm": 0.4967774141155372,
"learning_rate": 1.393853735430398e-05,
"loss": 1.0805,
"step": 785
},
{
"epoch": 3.1837974683544306,
"grad_norm": 0.5247575082247564,
"learning_rate": 1.3884472674819246e-05,
"loss": 1.0993,
"step": 786
},
{
"epoch": 3.1878481012658226,
"grad_norm": 0.6474317688947145,
"learning_rate": 1.3830457248950864e-05,
"loss": 1.0588,
"step": 787
},
{
"epoch": 3.191898734177215,
"grad_norm": 0.5747890244566953,
"learning_rate": 1.377649151173163e-05,
"loss": 1.0874,
"step": 788
},
{
"epoch": 3.1959493670886077,
"grad_norm": 0.5459020767143781,
"learning_rate": 1.3722575897794181e-05,
"loss": 1.0791,
"step": 789
},
{
"epoch": 3.2,
"grad_norm": 0.5694550798790898,
"learning_rate": 1.3668710841367472e-05,
"loss": 1.0756,
"step": 790
},
{
"epoch": 3.2040506329113922,
"grad_norm": 0.5252195523685619,
"learning_rate": 1.361489677627324e-05,
"loss": 1.0589,
"step": 791
},
{
"epoch": 3.2081012658227848,
"grad_norm": 0.5180153027261849,
"learning_rate": 1.3561134135922585e-05,
"loss": 1.1148,
"step": 792
},
{
"epoch": 3.2121518987341773,
"grad_norm": 0.6468292562170497,
"learning_rate": 1.350742335331241e-05,
"loss": 1.1385,
"step": 793
},
{
"epoch": 3.21620253164557,
"grad_norm": 0.5390570218055659,
"learning_rate": 1.345376486102198e-05,
"loss": 1.0855,
"step": 794
},
{
"epoch": 3.220253164556962,
"grad_norm": 0.6174059502276985,
"learning_rate": 1.3400159091209414e-05,
"loss": 1.0779,
"step": 795
},
{
"epoch": 3.2243037974683544,
"grad_norm": 0.5505443201524844,
"learning_rate": 1.3346606475608216e-05,
"loss": 1.0659,
"step": 796
},
{
"epoch": 3.228354430379747,
"grad_norm": 0.5451659622655888,
"learning_rate": 1.3293107445523781e-05,
"loss": 1.0792,
"step": 797
},
{
"epoch": 3.2324050632911394,
"grad_norm": 0.6325856053987682,
"learning_rate": 1.3239662431829949e-05,
"loss": 1.076,
"step": 798
},
{
"epoch": 3.2364556962025315,
"grad_norm": 0.5095359890479392,
"learning_rate": 1.3186271864965509e-05,
"loss": 1.093,
"step": 799
},
{
"epoch": 3.240506329113924,
"grad_norm": 0.6190087012160116,
"learning_rate": 1.3132936174930756e-05,
"loss": 1.0895,
"step": 800
},
{
"epoch": 3.2445569620253165,
"grad_norm": 0.5508715058512142,
"learning_rate": 1.3079655791283995e-05,
"loss": 1.0764,
"step": 801
},
{
"epoch": 3.248607594936709,
"grad_norm": 0.5734883343498642,
"learning_rate": 1.3026431143138108e-05,
"loss": 1.0495,
"step": 802
},
{
"epoch": 3.252658227848101,
"grad_norm": 0.5082943385621608,
"learning_rate": 1.2973262659157114e-05,
"loss": 1.0724,
"step": 803
},
{
"epoch": 3.2567088607594936,
"grad_norm": 0.6283087476072808,
"learning_rate": 1.2920150767552651e-05,
"loss": 1.0807,
"step": 804
},
{
"epoch": 3.260759493670886,
"grad_norm": 0.5337438708034233,
"learning_rate": 1.2867095896080607e-05,
"loss": 1.078,
"step": 805
},
{
"epoch": 3.2648101265822786,
"grad_norm": 0.6135646510351138,
"learning_rate": 1.2814098472037612e-05,
"loss": 1.1091,
"step": 806
},
{
"epoch": 3.2688607594936707,
"grad_norm": 0.6349034197188677,
"learning_rate": 1.276115892225764e-05,
"loss": 1.0814,
"step": 807
},
{
"epoch": 3.2729113924050632,
"grad_norm": 0.5033664628455571,
"learning_rate": 1.2708277673108555e-05,
"loss": 1.054,
"step": 808
},
{
"epoch": 3.2769620253164558,
"grad_norm": 0.6105898666137843,
"learning_rate": 1.2655455150488649e-05,
"loss": 1.0957,
"step": 809
},
{
"epoch": 3.2810126582278483,
"grad_norm": 0.5365736856008991,
"learning_rate": 1.2602691779823272e-05,
"loss": 1.0875,
"step": 810
},
{
"epoch": 3.2850632911392403,
"grad_norm": 0.5167484549630172,
"learning_rate": 1.2549987986061355e-05,
"loss": 1.0564,
"step": 811
},
{
"epoch": 3.289113924050633,
"grad_norm": 0.5420018699148835,
"learning_rate": 1.2497344193672005e-05,
"loss": 1.0776,
"step": 812
},
{
"epoch": 3.2931645569620254,
"grad_norm": 0.5107062101315082,
"learning_rate": 1.2444760826641092e-05,
"loss": 1.0943,
"step": 813
},
{
"epoch": 3.297215189873418,
"grad_norm": 0.5458309908919269,
"learning_rate": 1.2392238308467817e-05,
"loss": 1.0736,
"step": 814
},
{
"epoch": 3.30126582278481,
"grad_norm": 0.5085413046681425,
"learning_rate": 1.2339777062161326e-05,
"loss": 1.1188,
"step": 815
},
{
"epoch": 3.3053164556962025,
"grad_norm": 0.5145621645731109,
"learning_rate": 1.2287377510237293e-05,
"loss": 1.0697,
"step": 816
},
{
"epoch": 3.309367088607595,
"grad_norm": 0.5279236675122893,
"learning_rate": 1.2235040074714488e-05,
"loss": 1.0827,
"step": 817
},
{
"epoch": 3.3134177215189875,
"grad_norm": 0.5817921902005823,
"learning_rate": 1.2182765177111434e-05,
"loss": 1.056,
"step": 818
},
{
"epoch": 3.3174683544303796,
"grad_norm": 0.5207035518341144,
"learning_rate": 1.213055323844297e-05,
"loss": 1.0978,
"step": 819
},
{
"epoch": 3.321518987341772,
"grad_norm": 0.5596636548639724,
"learning_rate": 1.2078404679216864e-05,
"loss": 1.068,
"step": 820
},
{
"epoch": 3.3255696202531646,
"grad_norm": 0.628931110557474,
"learning_rate": 1.2026319919430458e-05,
"loss": 1.0634,
"step": 821
},
{
"epoch": 3.329620253164557,
"grad_norm": 0.5081750500537711,
"learning_rate": 1.1974299378567227e-05,
"loss": 1.0915,
"step": 822
},
{
"epoch": 3.333670886075949,
"grad_norm": 0.5526694909482114,
"learning_rate": 1.1922343475593462e-05,
"loss": 1.0644,
"step": 823
},
{
"epoch": 3.3377215189873417,
"grad_norm": 0.5321992488955425,
"learning_rate": 1.187045262895488e-05,
"loss": 1.0809,
"step": 824
},
{
"epoch": 3.3417721518987342,
"grad_norm": 0.5353051788314538,
"learning_rate": 1.1818627256573203e-05,
"loss": 1.0874,
"step": 825
},
{
"epoch": 3.3458227848101267,
"grad_norm": 0.5665917472554869,
"learning_rate": 1.1766867775842864e-05,
"loss": 1.1094,
"step": 826
},
{
"epoch": 3.349873417721519,
"grad_norm": 0.48691774491304723,
"learning_rate": 1.1715174603627615e-05,
"loss": 1.1094,
"step": 827
},
{
"epoch": 3.3539240506329113,
"grad_norm": 0.5801788802089181,
"learning_rate": 1.1663548156257147e-05,
"loss": 1.0919,
"step": 828
},
{
"epoch": 3.357974683544304,
"grad_norm": 0.5326475433756642,
"learning_rate": 1.161198884952377e-05,
"loss": 1.1035,
"step": 829
},
{
"epoch": 3.3620253164556964,
"grad_norm": 0.5287452985765041,
"learning_rate": 1.1560497098679056e-05,
"loss": 1.1206,
"step": 830
},
{
"epoch": 3.3660759493670884,
"grad_norm": 0.5151651723568196,
"learning_rate": 1.1509073318430479e-05,
"loss": 1.0863,
"step": 831
},
{
"epoch": 3.370126582278481,
"grad_norm": 0.5246535262801125,
"learning_rate": 1.1457717922938116e-05,
"loss": 1.0917,
"step": 832
},
{
"epoch": 3.3741772151898735,
"grad_norm": 0.5085223206266836,
"learning_rate": 1.1406431325811233e-05,
"loss": 1.0925,
"step": 833
},
{
"epoch": 3.378227848101266,
"grad_norm": 0.5101752790282125,
"learning_rate": 1.135521394010506e-05,
"loss": 1.1075,
"step": 834
},
{
"epoch": 3.382278481012658,
"grad_norm": 0.4961360692393899,
"learning_rate": 1.1304066178317367e-05,
"loss": 1.1082,
"step": 835
},
{
"epoch": 3.3863291139240506,
"grad_norm": 0.5184897475423347,
"learning_rate": 1.1252988452385199e-05,
"loss": 1.1004,
"step": 836
},
{
"epoch": 3.390379746835443,
"grad_norm": 0.5121200184577955,
"learning_rate": 1.1201981173681536e-05,
"loss": 1.113,
"step": 837
},
{
"epoch": 3.3944303797468356,
"grad_norm": 0.49721487818119003,
"learning_rate": 1.1151044753011991e-05,
"loss": 1.07,
"step": 838
},
{
"epoch": 3.3984810126582277,
"grad_norm": 0.4926109648560399,
"learning_rate": 1.1100179600611491e-05,
"loss": 1.0581,
"step": 839
},
{
"epoch": 3.40253164556962,
"grad_norm": 0.4981879904911693,
"learning_rate": 1.1049386126140985e-05,
"loss": 1.1118,
"step": 840
},
{
"epoch": 3.4065822784810127,
"grad_norm": 0.520095934151107,
"learning_rate": 1.0998664738684128e-05,
"loss": 1.109,
"step": 841
},
{
"epoch": 3.410632911392405,
"grad_norm": 0.47360325413264626,
"learning_rate": 1.0948015846744e-05,
"loss": 1.0768,
"step": 842
},
{
"epoch": 3.4146835443037973,
"grad_norm": 0.5064977286965664,
"learning_rate": 1.0897439858239832e-05,
"loss": 1.0851,
"step": 843
},
{
"epoch": 3.41873417721519,
"grad_norm": 0.5702414250215233,
"learning_rate": 1.0846937180503652e-05,
"loss": 1.1038,
"step": 844
},
{
"epoch": 3.4227848101265823,
"grad_norm": 0.5838131168521612,
"learning_rate": 1.0796508220277117e-05,
"loss": 1.0636,
"step": 845
},
{
"epoch": 3.426835443037975,
"grad_norm": 0.4851470951144798,
"learning_rate": 1.0746153383708107e-05,
"loss": 1.102,
"step": 846
},
{
"epoch": 3.430886075949367,
"grad_norm": 0.5357978079144178,
"learning_rate": 1.0695873076347579e-05,
"loss": 1.0914,
"step": 847
},
{
"epoch": 3.4349367088607594,
"grad_norm": 0.5120452114868796,
"learning_rate": 1.0645667703146205e-05,
"loss": 1.1181,
"step": 848
},
{
"epoch": 3.438987341772152,
"grad_norm": 0.5122217314259747,
"learning_rate": 1.0595537668451161e-05,
"loss": 1.1031,
"step": 849
},
{
"epoch": 3.4430379746835444,
"grad_norm": 0.5394389569576447,
"learning_rate": 1.0545483376002854e-05,
"loss": 1.0855,
"step": 850
},
{
"epoch": 3.4470886075949365,
"grad_norm": 0.5356413994416376,
"learning_rate": 1.0495505228931676e-05,
"loss": 1.0675,
"step": 851
},
{
"epoch": 3.451139240506329,
"grad_norm": 0.5294288472516849,
"learning_rate": 1.044560362975474e-05,
"loss": 1.1125,
"step": 852
},
{
"epoch": 3.4551898734177215,
"grad_norm": 0.5150225339160343,
"learning_rate": 1.0395778980372695e-05,
"loss": 1.054,
"step": 853
},
{
"epoch": 3.459240506329114,
"grad_norm": 0.5070471097188434,
"learning_rate": 1.0346031682066381e-05,
"loss": 1.0649,
"step": 854
},
{
"epoch": 3.463291139240506,
"grad_norm": 0.5422006665055134,
"learning_rate": 1.0296362135493724e-05,
"loss": 1.1034,
"step": 855
},
{
"epoch": 3.4673417721518986,
"grad_norm": 0.5339166625595353,
"learning_rate": 1.0246770740686422e-05,
"loss": 1.0785,
"step": 856
},
{
"epoch": 3.471392405063291,
"grad_norm": 0.520672945461222,
"learning_rate": 1.0197257897046743e-05,
"loss": 1.0616,
"step": 857
},
{
"epoch": 3.4754430379746837,
"grad_norm": 0.5082632539874644,
"learning_rate": 1.014782400334433e-05,
"loss": 1.0636,
"step": 858
},
{
"epoch": 3.479493670886076,
"grad_norm": 0.5215829781946013,
"learning_rate": 1.009846945771296e-05,
"loss": 1.1136,
"step": 859
},
{
"epoch": 3.4835443037974683,
"grad_norm": 0.46122445281421093,
"learning_rate": 1.0049194657647363e-05,
"loss": 1.0786,
"step": 860
},
{
"epoch": 3.487594936708861,
"grad_norm": 0.5319387655362979,
"learning_rate": 1.0000000000000006e-05,
"loss": 1.0999,
"step": 861
},
{
"epoch": 3.4916455696202533,
"grad_norm": 0.5253816433551323,
"learning_rate": 9.950885880977891e-06,
"loss": 1.0837,
"step": 862
},
{
"epoch": 3.4956962025316454,
"grad_norm": 0.47820161495066665,
"learning_rate": 9.901852696139382e-06,
"loss": 1.0828,
"step": 863
},
{
"epoch": 3.499746835443038,
"grad_norm": 0.4853622982420876,
"learning_rate": 9.852900840391027e-06,
"loss": 1.0683,
"step": 864
},
{
"epoch": 3.5037974683544304,
"grad_norm": 0.5408501074806961,
"learning_rate": 9.804030707984313e-06,
"loss": 1.0954,
"step": 865
},
{
"epoch": 3.507848101265823,
"grad_norm": 0.5154985891311936,
"learning_rate": 9.755242692512599e-06,
"loss": 1.0964,
"step": 866
},
{
"epoch": 3.5118987341772154,
"grad_norm": 0.4646727260533976,
"learning_rate": 9.70653718690782e-06,
"loss": 1.0518,
"step": 867
},
{
"epoch": 3.5159493670886075,
"grad_norm": 0.5106557483257654,
"learning_rate": 9.657914583437454e-06,
"loss": 1.1189,
"step": 868
},
{
"epoch": 3.52,
"grad_norm": 0.5190795872777753,
"learning_rate": 9.609375273701246e-06,
"loss": 1.0787,
"step": 869
},
{
"epoch": 3.5240506329113925,
"grad_norm": 0.46936436359453343,
"learning_rate": 9.560919648628133e-06,
"loss": 1.0683,
"step": 870
},
{
"epoch": 3.5281012658227846,
"grad_norm": 0.47031947142491337,
"learning_rate": 9.512548098473047e-06,
"loss": 1.1158,
"step": 871
},
{
"epoch": 3.532151898734177,
"grad_norm": 0.5093080643775301,
"learning_rate": 9.464261012813825e-06,
"loss": 1.0742,
"step": 872
},
{
"epoch": 3.5362025316455696,
"grad_norm": 0.4768797386454964,
"learning_rate": 9.416058780547987e-06,
"loss": 1.088,
"step": 873
},
{
"epoch": 3.540253164556962,
"grad_norm": 0.47843542688390284,
"learning_rate": 9.367941789889714e-06,
"loss": 1.0954,
"step": 874
},
{
"epoch": 3.5443037974683547,
"grad_norm": 0.5166980183020622,
"learning_rate": 9.319910428366607e-06,
"loss": 1.1303,
"step": 875
},
{
"epoch": 3.5483544303797467,
"grad_norm": 0.4496249221041384,
"learning_rate": 9.271965082816667e-06,
"loss": 1.1046,
"step": 876
},
{
"epoch": 3.5524050632911393,
"grad_norm": 0.48223623905543284,
"learning_rate": 9.224106139385111e-06,
"loss": 1.0833,
"step": 877
},
{
"epoch": 3.5564556962025318,
"grad_norm": 0.566987708322667,
"learning_rate": 9.176333983521291e-06,
"loss": 1.0796,
"step": 878
},
{
"epoch": 3.560506329113924,
"grad_norm": 0.49844219226966907,
"learning_rate": 9.12864899997558e-06,
"loss": 1.0886,
"step": 879
},
{
"epoch": 3.5645569620253164,
"grad_norm": 0.48048905808696035,
"learning_rate": 9.08105157279628e-06,
"loss": 1.0679,
"step": 880
},
{
"epoch": 3.568607594936709,
"grad_norm": 0.5480224347370678,
"learning_rate": 9.03354208532653e-06,
"loss": 1.0788,
"step": 881
},
{
"epoch": 3.5726582278481014,
"grad_norm": 0.5125944479300717,
"learning_rate": 8.986120920201205e-06,
"loss": 1.0656,
"step": 882
},
{
"epoch": 3.576708860759494,
"grad_norm": 0.47957059171351474,
"learning_rate": 8.938788459343852e-06,
"loss": 1.0642,
"step": 883
},
{
"epoch": 3.580759493670886,
"grad_norm": 0.4734670283585952,
"learning_rate": 8.8915450839636e-06,
"loss": 1.092,
"step": 884
},
{
"epoch": 3.5848101265822785,
"grad_norm": 0.4833046270507588,
"learning_rate": 8.844391174552116e-06,
"loss": 1.1048,
"step": 885
},
{
"epoch": 3.588860759493671,
"grad_norm": 0.4549220191873579,
"learning_rate": 8.797327110880479e-06,
"loss": 1.0727,
"step": 886
},
{
"epoch": 3.592911392405063,
"grad_norm": 0.4690141664149051,
"learning_rate": 8.750353271996206e-06,
"loss": 1.0737,
"step": 887
},
{
"epoch": 3.5969620253164556,
"grad_norm": 0.4675944056690804,
"learning_rate": 8.703470036220132e-06,
"loss": 1.1001,
"step": 888
},
{
"epoch": 3.601012658227848,
"grad_norm": 0.4782801478132864,
"learning_rate": 8.656677781143394e-06,
"loss": 1.0777,
"step": 889
},
{
"epoch": 3.6050632911392406,
"grad_norm": 0.4771090846906864,
"learning_rate": 8.609976883624377e-06,
"loss": 1.1098,
"step": 890
},
{
"epoch": 3.609113924050633,
"grad_norm": 0.46245265685533393,
"learning_rate": 8.563367719785698e-06,
"loss": 1.0953,
"step": 891
},
{
"epoch": 3.613164556962025,
"grad_norm": 0.4724628266927542,
"learning_rate": 8.516850665011138e-06,
"loss": 1.0842,
"step": 892
},
{
"epoch": 3.6172151898734177,
"grad_norm": 0.48504245078496666,
"learning_rate": 8.47042609394269e-06,
"loss": 1.0788,
"step": 893
},
{
"epoch": 3.6212658227848102,
"grad_norm": 0.4986936851141963,
"learning_rate": 8.424094380477432e-06,
"loss": 1.0645,
"step": 894
},
{
"epoch": 3.6253164556962023,
"grad_norm": 0.4893289483903203,
"learning_rate": 8.37785589776465e-06,
"loss": 1.0803,
"step": 895
},
{
"epoch": 3.629367088607595,
"grad_norm": 0.47505249620792156,
"learning_rate": 8.331711018202694e-06,
"loss": 1.1231,
"step": 896
},
{
"epoch": 3.6334177215189873,
"grad_norm": 0.5198256456351684,
"learning_rate": 8.285660113436104e-06,
"loss": 1.1075,
"step": 897
},
{
"epoch": 3.63746835443038,
"grad_norm": 0.47682023750943486,
"learning_rate": 8.239703554352527e-06,
"loss": 1.0938,
"step": 898
},
{
"epoch": 3.6415189873417724,
"grad_norm": 0.5142104388789442,
"learning_rate": 8.193841711079775e-06,
"loss": 1.0912,
"step": 899
},
{
"epoch": 3.6455696202531644,
"grad_norm": 0.506786942085681,
"learning_rate": 8.148074952982828e-06,
"loss": 1.0869,
"step": 900
},
{
"epoch": 3.649620253164557,
"grad_norm": 0.4877616902924065,
"learning_rate": 8.102403648660859e-06,
"loss": 1.0984,
"step": 901
},
{
"epoch": 3.6536708860759495,
"grad_norm": 0.5081939991081638,
"learning_rate": 8.056828165944282e-06,
"loss": 1.0919,
"step": 902
},
{
"epoch": 3.6577215189873415,
"grad_norm": 0.6098761688640817,
"learning_rate": 8.011348871891762e-06,
"loss": 1.1057,
"step": 903
},
{
"epoch": 3.661772151898734,
"grad_norm": 0.47439931544093344,
"learning_rate": 7.965966132787287e-06,
"loss": 1.1079,
"step": 904
},
{
"epoch": 3.6658227848101266,
"grad_norm": 0.4992344135407126,
"learning_rate": 7.920680314137189e-06,
"loss": 1.089,
"step": 905
},
{
"epoch": 3.669873417721519,
"grad_norm": 0.5021970860826235,
"learning_rate": 7.875491780667246e-06,
"loss": 1.0621,
"step": 906
},
{
"epoch": 3.6739240506329116,
"grad_norm": 0.4907862338362437,
"learning_rate": 7.830400896319667e-06,
"loss": 1.0892,
"step": 907
},
{
"epoch": 3.6779746835443037,
"grad_norm": 0.5254657619262012,
"learning_rate": 7.785408024250259e-06,
"loss": 1.1267,
"step": 908
},
{
"epoch": 3.682025316455696,
"grad_norm": 0.4977503626782785,
"learning_rate": 7.74051352682542e-06,
"loss": 1.0788,
"step": 909
},
{
"epoch": 3.6860759493670887,
"grad_norm": 0.45569216683982877,
"learning_rate": 7.695717765619257e-06,
"loss": 1.1194,
"step": 910
},
{
"epoch": 3.690126582278481,
"grad_norm": 0.49501657996933585,
"learning_rate": 7.651021101410673e-06,
"loss": 1.1127,
"step": 911
},
{
"epoch": 3.6941772151898733,
"grad_norm": 0.5073028518508228,
"learning_rate": 7.606423894180464e-06,
"loss": 1.0696,
"step": 912
},
{
"epoch": 3.698227848101266,
"grad_norm": 0.5424098999233283,
"learning_rate": 7.56192650310839e-06,
"loss": 1.0743,
"step": 913
},
{
"epoch": 3.7022784810126583,
"grad_norm": 0.49040477078425876,
"learning_rate": 7.517529286570349e-06,
"loss": 1.0938,
"step": 914
},
{
"epoch": 3.706329113924051,
"grad_norm": 0.5083882792377802,
"learning_rate": 7.473232602135387e-06,
"loss": 1.1148,
"step": 915
},
{
"epoch": 3.710379746835443,
"grad_norm": 0.502606028508724,
"learning_rate": 7.429036806562935e-06,
"loss": 1.1024,
"step": 916
},
{
"epoch": 3.7144303797468354,
"grad_norm": 0.4971864840923049,
"learning_rate": 7.3849422557998455e-06,
"loss": 1.0879,
"step": 917
},
{
"epoch": 3.718481012658228,
"grad_norm": 0.5008277348096614,
"learning_rate": 7.340949304977567e-06,
"loss": 1.0824,
"step": 918
},
{
"epoch": 3.72253164556962,
"grad_norm": 0.4853111832060397,
"learning_rate": 7.297058308409282e-06,
"loss": 1.0988,
"step": 919
},
{
"epoch": 3.7265822784810125,
"grad_norm": 0.5365061550978811,
"learning_rate": 7.25326961958704e-06,
"loss": 1.1222,
"step": 920
},
{
"epoch": 3.730632911392405,
"grad_norm": 0.48891731185845405,
"learning_rate": 7.209583591178921e-06,
"loss": 1.1037,
"step": 921
},
{
"epoch": 3.7346835443037976,
"grad_norm": 0.5207233853400592,
"learning_rate": 7.1660005750261925e-06,
"loss": 1.1176,
"step": 922
},
{
"epoch": 3.73873417721519,
"grad_norm": 0.45607294790527314,
"learning_rate": 7.1225209221404765e-06,
"loss": 1.0868,
"step": 923
},
{
"epoch": 3.742784810126582,
"grad_norm": 0.49268476213657575,
"learning_rate": 7.079144982700909e-06,
"loss": 1.0679,
"step": 924
},
{
"epoch": 3.7468354430379747,
"grad_norm": 0.5022316591703319,
"learning_rate": 7.0358731060513695e-06,
"loss": 1.1089,
"step": 925
},
{
"epoch": 3.750886075949367,
"grad_norm": 0.5438497203803175,
"learning_rate": 6.99270564069757e-06,
"loss": 1.0747,
"step": 926
},
{
"epoch": 3.7549367088607593,
"grad_norm": 0.4661376619310044,
"learning_rate": 6.949642934304375e-06,
"loss": 1.0747,
"step": 927
},
{
"epoch": 3.7589873417721518,
"grad_norm": 0.482947321380622,
"learning_rate": 6.906685333692871e-06,
"loss": 1.0984,
"step": 928
},
{
"epoch": 3.7630379746835443,
"grad_norm": 0.465154055501484,
"learning_rate": 6.86383318483769e-06,
"loss": 1.0938,
"step": 929
},
{
"epoch": 3.767088607594937,
"grad_norm": 0.549296914923835,
"learning_rate": 6.821086832864139e-06,
"loss": 1.0994,
"step": 930
},
{
"epoch": 3.7711392405063293,
"grad_norm": 0.5229025414350521,
"learning_rate": 6.77844662204546e-06,
"loss": 1.0833,
"step": 931
},
{
"epoch": 3.7751898734177214,
"grad_norm": 0.47373388765125823,
"learning_rate": 6.7359128958000455e-06,
"loss": 1.1018,
"step": 932
},
{
"epoch": 3.779240506329114,
"grad_norm": 0.4881470277620784,
"learning_rate": 6.693485996688695e-06,
"loss": 1.0472,
"step": 933
},
{
"epoch": 3.7832911392405064,
"grad_norm": 0.5658883711004198,
"learning_rate": 6.651166266411801e-06,
"loss": 1.0781,
"step": 934
},
{
"epoch": 3.7873417721518985,
"grad_norm": 0.49801301500328476,
"learning_rate": 6.6089540458066725e-06,
"loss": 1.0717,
"step": 935
},
{
"epoch": 3.791392405063291,
"grad_norm": 0.4853115815172323,
"learning_rate": 6.566849674844711e-06,
"loss": 1.1001,
"step": 936
},
{
"epoch": 3.7954430379746835,
"grad_norm": 0.4593783421143412,
"learning_rate": 6.524853492628747e-06,
"loss": 1.05,
"step": 937
},
{
"epoch": 3.799493670886076,
"grad_norm": 0.4952651905151486,
"learning_rate": 6.4829658373902536e-06,
"loss": 1.078,
"step": 938
},
{
"epoch": 3.8035443037974686,
"grad_norm": 0.5164126930414986,
"learning_rate": 6.441187046486648e-06,
"loss": 1.0763,
"step": 939
},
{
"epoch": 3.8075949367088606,
"grad_norm": 0.4745949113520698,
"learning_rate": 6.399517456398567e-06,
"loss": 1.0894,
"step": 940
},
{
"epoch": 3.811645569620253,
"grad_norm": 0.4749902527623226,
"learning_rate": 6.357957402727164e-06,
"loss": 1.0884,
"step": 941
},
{
"epoch": 3.8156962025316457,
"grad_norm": 0.47641713530721275,
"learning_rate": 6.316507220191395e-06,
"loss": 1.0862,
"step": 942
},
{
"epoch": 3.8197468354430377,
"grad_norm": 0.4976339376137511,
"learning_rate": 6.275167242625331e-06,
"loss": 1.0967,
"step": 943
},
{
"epoch": 3.8237974683544302,
"grad_norm": 0.4593847958173142,
"learning_rate": 6.233937802975471e-06,
"loss": 1.1049,
"step": 944
},
{
"epoch": 3.8278481012658228,
"grad_norm": 0.4837213249957769,
"learning_rate": 6.192819233298046e-06,
"loss": 1.1085,
"step": 945
},
{
"epoch": 3.8318987341772153,
"grad_norm": 0.4804607445767388,
"learning_rate": 6.151811864756383e-06,
"loss": 1.0741,
"step": 946
},
{
"epoch": 3.835949367088608,
"grad_norm": 0.4578035431897188,
"learning_rate": 6.1109160276181655e-06,
"loss": 1.1055,
"step": 947
},
{
"epoch": 3.84,
"grad_norm": 0.44780312314526405,
"learning_rate": 6.070132051252868e-06,
"loss": 1.1049,
"step": 948
},
{
"epoch": 3.8440506329113924,
"grad_norm": 0.4613514271469346,
"learning_rate": 6.0294602641290034e-06,
"loss": 1.0784,
"step": 949
},
{
"epoch": 3.848101265822785,
"grad_norm": 0.4897296186076931,
"learning_rate": 5.988900993811575e-06,
"loss": 1.0861,
"step": 950
},
{
"epoch": 3.852151898734177,
"grad_norm": 0.4850097737624442,
"learning_rate": 5.948454566959363e-06,
"loss": 1.1478,
"step": 951
},
{
"epoch": 3.8562025316455695,
"grad_norm": 0.4419939043177187,
"learning_rate": 5.908121309322328e-06,
"loss": 1.0698,
"step": 952
},
{
"epoch": 3.860253164556962,
"grad_norm": 0.46593043267997275,
"learning_rate": 5.867901545738976e-06,
"loss": 1.078,
"step": 953
},
{
"epoch": 3.8643037974683545,
"grad_norm": 0.4661571777549694,
"learning_rate": 5.827795600133774e-06,
"loss": 1.0964,
"step": 954
},
{
"epoch": 3.868354430379747,
"grad_norm": 0.45868472214110706,
"learning_rate": 5.787803795514466e-06,
"loss": 1.0849,
"step": 955
},
{
"epoch": 3.872405063291139,
"grad_norm": 0.473867384376204,
"learning_rate": 5.747926453969576e-06,
"loss": 1.0765,
"step": 956
},
{
"epoch": 3.8764556962025316,
"grad_norm": 0.47904236678740747,
"learning_rate": 5.708163896665708e-06,
"loss": 1.1042,
"step": 957
},
{
"epoch": 3.880506329113924,
"grad_norm": 0.46492650696044774,
"learning_rate": 5.668516443845047e-06,
"loss": 1.075,
"step": 958
},
{
"epoch": 3.884556962025316,
"grad_norm": 0.44449523944387836,
"learning_rate": 5.6289844148227225e-06,
"loss": 1.0787,
"step": 959
},
{
"epoch": 3.8886075949367087,
"grad_norm": 0.435268781845486,
"learning_rate": 5.5895681279842615e-06,
"loss": 1.0371,
"step": 960
},
{
"epoch": 3.8926582278481012,
"grad_norm": 0.4773215193084178,
"learning_rate": 5.550267900783019e-06,
"loss": 1.1075,
"step": 961
},
{
"epoch": 3.8967088607594937,
"grad_norm": 0.5246749483144946,
"learning_rate": 5.511084049737623e-06,
"loss": 1.0785,
"step": 962
},
{
"epoch": 3.9007594936708863,
"grad_norm": 0.4804719049681873,
"learning_rate": 5.4720168904294215e-06,
"loss": 1.118,
"step": 963
},
{
"epoch": 3.9048101265822783,
"grad_norm": 0.4657150665261349,
"learning_rate": 5.433066737499948e-06,
"loss": 1.1059,
"step": 964
},
{
"epoch": 3.908860759493671,
"grad_norm": 0.44695875421091436,
"learning_rate": 5.394233904648376e-06,
"loss": 1.0711,
"step": 965
},
{
"epoch": 3.9129113924050634,
"grad_norm": 0.4529501661622864,
"learning_rate": 5.355518704628997e-06,
"loss": 1.0797,
"step": 966
},
{
"epoch": 3.9169620253164554,
"grad_norm": 0.5016954390712551,
"learning_rate": 5.316921449248731e-06,
"loss": 1.0529,
"step": 967
},
{
"epoch": 3.921012658227848,
"grad_norm": 0.47656214824740945,
"learning_rate": 5.278442449364538e-06,
"loss": 1.11,
"step": 968
},
{
"epoch": 3.9250632911392405,
"grad_norm": 0.46024799414194806,
"learning_rate": 5.240082014881016e-06,
"loss": 1.0764,
"step": 969
},
{
"epoch": 3.929113924050633,
"grad_norm": 0.45189245489846463,
"learning_rate": 5.201840454747822e-06,
"loss": 1.0917,
"step": 970
},
{
"epoch": 3.9331645569620255,
"grad_norm": 0.45446748018276856,
"learning_rate": 5.163718076957223e-06,
"loss": 1.0909,
"step": 971
},
{
"epoch": 3.9372151898734176,
"grad_norm": 0.449561848664391,
"learning_rate": 5.125715188541609e-06,
"loss": 1.0914,
"step": 972
},
{
"epoch": 3.94126582278481,
"grad_norm": 0.4488151765777236,
"learning_rate": 5.087832095571021e-06,
"loss": 1.0851,
"step": 973
},
{
"epoch": 3.9453164556962026,
"grad_norm": 0.48030064093009006,
"learning_rate": 5.0500691031506766e-06,
"loss": 1.1057,
"step": 974
},
{
"epoch": 3.9493670886075947,
"grad_norm": 0.4567568994881793,
"learning_rate": 5.01242651541854e-06,
"loss": 1.0893,
"step": 975
},
{
"epoch": 3.953417721518987,
"grad_norm": 0.44655516214311325,
"learning_rate": 4.974904635542815e-06,
"loss": 1.0643,
"step": 976
},
{
"epoch": 3.9574683544303797,
"grad_norm": 0.4493193099088335,
"learning_rate": 4.937503765719582e-06,
"loss": 1.1169,
"step": 977
},
{
"epoch": 3.961518987341772,
"grad_norm": 0.4282706738035696,
"learning_rate": 4.900224207170299e-06,
"loss": 1.0675,
"step": 978
},
{
"epoch": 3.9655696202531647,
"grad_norm": 0.4554173986544851,
"learning_rate": 4.8630662601394065e-06,
"loss": 1.0825,
"step": 979
},
{
"epoch": 3.969620253164557,
"grad_norm": 0.4815601723537509,
"learning_rate": 4.8260302238918995e-06,
"loss": 1.0977,
"step": 980
},
{
"epoch": 3.9736708860759493,
"grad_norm": 0.4540132706448093,
"learning_rate": 4.789116396710924e-06,
"loss": 1.1053,
"step": 981
},
{
"epoch": 3.977721518987342,
"grad_norm": 0.4621935904351466,
"learning_rate": 4.752325075895368e-06,
"loss": 1.0883,
"step": 982
},
{
"epoch": 3.981772151898734,
"grad_norm": 0.4615840352230554,
"learning_rate": 4.715656557757473e-06,
"loss": 1.0743,
"step": 983
},
{
"epoch": 3.9858227848101264,
"grad_norm": 0.4617129293037875,
"learning_rate": 4.679111137620442e-06,
"loss": 1.0479,
"step": 984
},
{
"epoch": 3.989873417721519,
"grad_norm": 0.4680861435123717,
"learning_rate": 4.6426891098160585e-06,
"loss": 1.0863,
"step": 985
},
{
"epoch": 3.9939240506329114,
"grad_norm": 0.4463000587697989,
"learning_rate": 4.6063907676823474e-06,
"loss": 1.0619,
"step": 986
},
{
"epoch": 3.997974683544304,
"grad_norm": 0.46384889517387834,
"learning_rate": 4.570216403561141e-06,
"loss": 1.0612,
"step": 987
},
{
"epoch": 4.002025316455696,
"grad_norm": 0.6561480716258716,
"learning_rate": 4.534166308795815e-06,
"loss": 1.0173,
"step": 988
},
{
"epoch": 4.006075949367089,
"grad_norm": 0.9303546080150179,
"learning_rate": 4.498240773728859e-06,
"loss": 0.9729,
"step": 989
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.6921857694492077,
"learning_rate": 4.462440087699609e-06,
"loss": 0.9553,
"step": 990
},
{
"epoch": 4.014177215189873,
"grad_norm": 0.7332358506433847,
"learning_rate": 4.426764539041861e-06,
"loss": 0.9653,
"step": 991
},
{
"epoch": 4.018227848101266,
"grad_norm": 1.2622787076037467,
"learning_rate": 4.391214415081582e-06,
"loss": 0.8779,
"step": 992
},
{
"epoch": 4.022278481012658,
"grad_norm": 1.2164914714730921,
"learning_rate": 4.355790002134579e-06,
"loss": 0.9214,
"step": 993
},
{
"epoch": 4.02632911392405,
"grad_norm": 0.7752127359673404,
"learning_rate": 4.320491585504207e-06,
"loss": 0.9865,
"step": 994
},
{
"epoch": 4.030379746835443,
"grad_norm": 0.8068420257721804,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.9264,
"step": 995
},
{
"epoch": 4.034430379746835,
"grad_norm": 0.8351111470478945,
"learning_rate": 4.250273877330691e-06,
"loss": 0.9762,
"step": 996
},
{
"epoch": 4.038481012658228,
"grad_norm": 0.8007566084366775,
"learning_rate": 4.215355151311313e-06,
"loss": 0.9476,
"step": 997
},
{
"epoch": 4.04253164556962,
"grad_norm": 0.7210058373647178,
"learning_rate": 4.180563552651542e-06,
"loss": 0.9035,
"step": 998
},
{
"epoch": 4.046582278481012,
"grad_norm": 0.6701592236059128,
"learning_rate": 4.145899361558147e-06,
"loss": 0.9152,
"step": 999
},
{
"epoch": 4.050632911392405,
"grad_norm": 0.668992702013737,
"learning_rate": 4.111362857211738e-06,
"loss": 0.927,
"step": 1000
},
{
"epoch": 4.054683544303797,
"grad_norm": 0.7875082491558371,
"learning_rate": 4.076954317764592e-06,
"loss": 0.9348,
"step": 1001
},
{
"epoch": 4.0587341772151895,
"grad_norm": 0.7850220150297883,
"learning_rate": 4.042674020338335e-06,
"loss": 0.946,
"step": 1002
},
{
"epoch": 4.062784810126582,
"grad_norm": 0.724337659233419,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.922,
"step": 1003
},
{
"epoch": 4.0668354430379745,
"grad_norm": 0.6223055676228991,
"learning_rate": 3.974499254868674e-06,
"loss": 0.9321,
"step": 1004
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.7236035709126875,
"learning_rate": 3.940605335895451e-06,
"loss": 0.9106,
"step": 1005
},
{
"epoch": 4.0749367088607595,
"grad_norm": 0.6855598057723271,
"learning_rate": 3.90684075707908e-06,
"loss": 0.9437,
"step": 1006
},
{
"epoch": 4.078987341772152,
"grad_norm": 0.6988161437260564,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.9755,
"step": 1007
},
{
"epoch": 4.083037974683545,
"grad_norm": 0.6299764532464304,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.9402,
"step": 1008
},
{
"epoch": 4.087088607594937,
"grad_norm": 0.6739576995580662,
"learning_rate": 3.806325775702304e-06,
"loss": 0.9532,
"step": 1009
},
{
"epoch": 4.091139240506329,
"grad_norm": 0.6056043652766001,
"learning_rate": 3.773081266416434e-06,
"loss": 0.9565,
"step": 1010
},
{
"epoch": 4.095189873417722,
"grad_norm": 0.6218128495404703,
"learning_rate": 3.739967446503245e-06,
"loss": 0.898,
"step": 1011
},
{
"epoch": 4.099240506329114,
"grad_norm": 0.5740679918740786,
"learning_rate": 3.706984582656894e-06,
"loss": 0.9408,
"step": 1012
},
{
"epoch": 4.103291139240507,
"grad_norm": 0.5695033870942398,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.9403,
"step": 1013
},
{
"epoch": 4.107341772151899,
"grad_norm": 0.6159427528612437,
"learning_rate": 3.641412784665648e-06,
"loss": 0.9237,
"step": 1014
},
{
"epoch": 4.111392405063291,
"grad_norm": 0.5844488653779869,
"learning_rate": 3.608824378627005e-06,
"loss": 0.9155,
"step": 1015
},
{
"epoch": 4.115443037974684,
"grad_norm": 0.4988923098441874,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.9383,
"step": 1016
},
{
"epoch": 4.119493670886076,
"grad_norm": 0.5730769052391543,
"learning_rate": 3.544043864774269e-06,
"loss": 0.9614,
"step": 1017
},
{
"epoch": 4.123544303797468,
"grad_norm": 0.5929956894740648,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.9469,
"step": 1018
},
{
"epoch": 4.127594936708861,
"grad_norm": 0.5204099969803048,
"learning_rate": 3.479793485887819e-06,
"loss": 0.9259,
"step": 1019
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.5375416131716853,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.951,
"step": 1020
},
{
"epoch": 4.135696202531646,
"grad_norm": 0.5105820570498509,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.9145,
"step": 1021
},
{
"epoch": 4.139746835443038,
"grad_norm": 0.5274654065476356,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.901,
"step": 1022
},
{
"epoch": 4.14379746835443,
"grad_norm": 0.5373962847727779,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.903,
"step": 1023
},
{
"epoch": 4.147848101265823,
"grad_norm": 0.5106623382528704,
"learning_rate": 3.321500420335e-06,
"loss": 0.9278,
"step": 1024
},
{
"epoch": 4.151898734177215,
"grad_norm": 0.535990016883024,
"learning_rate": 3.290243771741275e-06,
"loss": 0.8943,
"step": 1025
},
{
"epoch": 4.155949367088607,
"grad_norm": 0.5208461481231389,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.9203,
"step": 1026
},
{
"epoch": 4.16,
"grad_norm": 0.5624715628321368,
"learning_rate": 3.228134459445149e-06,
"loss": 0.9238,
"step": 1027
},
{
"epoch": 4.164050632911392,
"grad_norm": 0.5277580453720315,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.959,
"step": 1028
},
{
"epoch": 4.168101265822785,
"grad_norm": 0.5045176311330415,
"learning_rate": 3.166565459258513e-06,
"loss": 0.9116,
"step": 1029
},
{
"epoch": 4.172151898734177,
"grad_norm": 0.5195352720297988,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.9597,
"step": 1030
},
{
"epoch": 4.176202531645569,
"grad_norm": 0.5175135639842248,
"learning_rate": 3.105538754650419e-06,
"loss": 0.9447,
"step": 1031
},
{
"epoch": 4.180253164556962,
"grad_norm": 0.4959262376620597,
"learning_rate": 3.07522937824712e-06,
"loss": 0.9312,
"step": 1032
},
{
"epoch": 4.184303797468354,
"grad_norm": 0.5261455839708709,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.9323,
"step": 1033
},
{
"epoch": 4.188354430379746,
"grad_norm": 0.5449309022762875,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.9627,
"step": 1034
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.5106190943297423,
"learning_rate": 2.985120078631465e-06,
"loss": 0.9174,
"step": 1035
},
{
"epoch": 4.1964556962025314,
"grad_norm": 0.5083427860499741,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.9265,
"step": 1036
},
{
"epoch": 4.200506329113924,
"grad_norm": 0.4951036912701695,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.9314,
"step": 1037
},
{
"epoch": 4.2045569620253165,
"grad_norm": 0.5313361897367099,
"learning_rate": 2.896244091927678e-06,
"loss": 0.958,
"step": 1038
},
{
"epoch": 4.2086075949367086,
"grad_norm": 0.510852984467588,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.9071,
"step": 1039
},
{
"epoch": 4.2126582278481015,
"grad_norm": 0.49187990785008456,
"learning_rate": 2.837681792951994e-06,
"loss": 0.9112,
"step": 1040
},
{
"epoch": 4.216708860759494,
"grad_norm": 0.4968459909688019,
"learning_rate": 2.808607860255981e-06,
"loss": 0.9413,
"step": 1041
},
{
"epoch": 4.220759493670886,
"grad_norm": 0.5080235236704547,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.9558,
"step": 1042
},
{
"epoch": 4.224810126582279,
"grad_norm": 0.5025137923235796,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.9461,
"step": 1043
},
{
"epoch": 4.228860759493671,
"grad_norm": 0.5108939358896738,
"learning_rate": 2.722217735873718e-06,
"loss": 0.9563,
"step": 1044
},
{
"epoch": 4.232911392405064,
"grad_norm": 0.5097432118397089,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.9886,
"step": 1045
},
{
"epoch": 4.236962025316456,
"grad_norm": 0.48045736959858387,
"learning_rate": 2.665319697490205e-06,
"loss": 0.9193,
"step": 1046
},
{
"epoch": 4.241012658227848,
"grad_norm": 0.4997177736379196,
"learning_rate": 2.637079980714945e-06,
"loss": 0.939,
"step": 1047
},
{
"epoch": 4.245063291139241,
"grad_norm": 0.4860409290152529,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.9254,
"step": 1048
},
{
"epoch": 4.249113924050633,
"grad_norm": 0.49532978448463694,
"learning_rate": 2.581020289189571e-06,
"loss": 0.9423,
"step": 1049
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.5085408324545386,
"learning_rate": 2.553200765936501e-06,
"loss": 0.9208,
"step": 1050
},
{
"epoch": 4.257215189873418,
"grad_norm": 0.49717739708498443,
"learning_rate": 2.525521756805962e-06,
"loss": 0.9684,
"step": 1051
},
{
"epoch": 4.26126582278481,
"grad_norm": 0.49939640487670794,
"learning_rate": 2.497983484720885e-06,
"loss": 0.9537,
"step": 1052
},
{
"epoch": 4.265316455696203,
"grad_norm": 0.47772452549375943,
"learning_rate": 2.470586171470728e-06,
"loss": 0.9159,
"step": 1053
},
{
"epoch": 4.269367088607595,
"grad_norm": 0.5142306032999792,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.9147,
"step": 1054
},
{
"epoch": 4.273417721518987,
"grad_norm": 0.4864685085170298,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.906,
"step": 1055
},
{
"epoch": 4.27746835443038,
"grad_norm": 0.48939159330490684,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.9197,
"step": 1056
},
{
"epoch": 4.281518987341772,
"grad_norm": 0.49795118573466784,
"learning_rate": 2.362410902836978e-06,
"loss": 0.9238,
"step": 1057
},
{
"epoch": 4.285569620253165,
"grad_norm": 0.5127222742798273,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.9555,
"step": 1058
},
{
"epoch": 4.289620253164557,
"grad_norm": 0.5204782496016115,
"learning_rate": 2.309174704446411e-06,
"loss": 0.9208,
"step": 1059
},
{
"epoch": 4.293670886075949,
"grad_norm": 0.4834093186572312,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.951,
"step": 1060
},
{
"epoch": 4.297721518987342,
"grad_norm": 0.49642314480288013,
"learning_rate": 2.256508422807855e-06,
"loss": 0.9504,
"step": 1061
},
{
"epoch": 4.301772151898734,
"grad_norm": 0.48925909298131925,
"learning_rate": 2.230389531696946e-06,
"loss": 0.927,
"step": 1062
},
{
"epoch": 4.305822784810126,
"grad_norm": 0.48551102777198657,
"learning_rate": 2.204413754585857e-06,
"loss": 0.9432,
"step": 1063
},
{
"epoch": 4.309873417721519,
"grad_norm": 0.5119613728257052,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.9261,
"step": 1064
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.5191592596338819,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.9244,
"step": 1065
},
{
"epoch": 4.317974683544303,
"grad_norm": 0.5251291953926087,
"learning_rate": 2.127347193531757e-06,
"loss": 0.9505,
"step": 1066
},
{
"epoch": 4.322025316455696,
"grad_norm": 0.4889075842589695,
"learning_rate": 2.101945952921942e-06,
"loss": 0.9688,
"step": 1067
},
{
"epoch": 4.326075949367088,
"grad_norm": 0.4915728800754518,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.914,
"step": 1068
},
{
"epoch": 4.330126582278481,
"grad_norm": 0.5053613362991279,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.9196,
"step": 1069
},
{
"epoch": 4.334177215189873,
"grad_norm": 0.4959516773756681,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.9302,
"step": 1070
},
{
"epoch": 4.3382278481012655,
"grad_norm": 0.5171205316985689,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.9363,
"step": 1071
},
{
"epoch": 4.3422784810126585,
"grad_norm": 0.5044391207689592,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.9441,
"step": 1072
},
{
"epoch": 4.3463291139240505,
"grad_norm": 0.4943449504820268,
"learning_rate": 1.95257270594031e-06,
"loss": 0.9508,
"step": 1073
},
{
"epoch": 4.3503797468354435,
"grad_norm": 0.4850896626156242,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.9286,
"step": 1074
},
{
"epoch": 4.3544303797468356,
"grad_norm": 0.49348828542452156,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.9106,
"step": 1075
},
{
"epoch": 4.358481012658228,
"grad_norm": 0.48100936780747994,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.894,
"step": 1076
},
{
"epoch": 4.362531645569621,
"grad_norm": 0.4800293142553574,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.9112,
"step": 1077
},
{
"epoch": 4.366582278481013,
"grad_norm": 0.49244069589846184,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.9392,
"step": 1078
},
{
"epoch": 4.370632911392405,
"grad_norm": 0.5046605011573989,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.949,
"step": 1079
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.492912414529253,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.9676,
"step": 1080
},
{
"epoch": 4.37873417721519,
"grad_norm": 0.48203789528993674,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.9327,
"step": 1081
},
{
"epoch": 4.382784810126582,
"grad_norm": 0.4936243109456259,
"learning_rate": 1.738336950434061e-06,
"loss": 0.8868,
"step": 1082
},
{
"epoch": 4.386835443037975,
"grad_norm": 0.48830691297679696,
"learning_rate": 1.715265841375957e-06,
"loss": 0.9201,
"step": 1083
},
{
"epoch": 4.390886075949367,
"grad_norm": 0.5174456642054066,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.9451,
"step": 1084
},
{
"epoch": 4.39493670886076,
"grad_norm": 0.48566182981447836,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.924,
"step": 1085
},
{
"epoch": 4.398987341772152,
"grad_norm": 0.5054411074147384,
"learning_rate": 1.646936827927441e-06,
"loss": 0.9392,
"step": 1086
},
{
"epoch": 4.403037974683544,
"grad_norm": 0.4808054042467124,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.933,
"step": 1087
},
{
"epoch": 4.407088607594937,
"grad_norm": 0.4761635279079274,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.9514,
"step": 1088
},
{
"epoch": 4.411139240506329,
"grad_norm": 0.4979186953578598,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.9338,
"step": 1089
},
{
"epoch": 4.415189873417722,
"grad_norm": 0.48284723782397326,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.9674,
"step": 1090
},
{
"epoch": 4.419240506329114,
"grad_norm": 0.48726536588702024,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.9423,
"step": 1091
},
{
"epoch": 4.423291139240506,
"grad_norm": 0.48014750527113187,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.9289,
"step": 1092
},
{
"epoch": 4.427341772151899,
"grad_norm": 0.48303748821366815,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.932,
"step": 1093
},
{
"epoch": 4.431392405063291,
"grad_norm": 0.48203767075833603,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.9274,
"step": 1094
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.48293160447431455,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.9512,
"step": 1095
},
{
"epoch": 4.439493670886076,
"grad_norm": 0.4921635958047356,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.9472,
"step": 1096
},
{
"epoch": 4.443544303797468,
"grad_norm": 0.4911107764447124,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.9434,
"step": 1097
},
{
"epoch": 4.44759493670886,
"grad_norm": 0.498538263527519,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.9364,
"step": 1098
},
{
"epoch": 4.451645569620253,
"grad_norm": 0.4978298376994617,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.9765,
"step": 1099
},
{
"epoch": 4.455696202531645,
"grad_norm": 0.4807522809932903,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.9479,
"step": 1100
},
{
"epoch": 4.459746835443038,
"grad_norm": 0.48802018775077843,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.9345,
"step": 1101
},
{
"epoch": 4.46379746835443,
"grad_norm": 0.49835703214943555,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.9254,
"step": 1102
},
{
"epoch": 4.467848101265822,
"grad_norm": 0.48831825214512287,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.943,
"step": 1103
},
{
"epoch": 4.471898734177215,
"grad_norm": 0.4916446241985829,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.9465,
"step": 1104
},
{
"epoch": 4.4759493670886075,
"grad_norm": 0.49301956855555157,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.9065,
"step": 1105
},
{
"epoch": 4.48,
"grad_norm": 0.4807994697200558,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.9388,
"step": 1106
},
{
"epoch": 4.4840506329113925,
"grad_norm": 0.5027130531220406,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.9087,
"step": 1107
},
{
"epoch": 4.488101265822785,
"grad_norm": 0.4960603939895552,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.9589,
"step": 1108
},
{
"epoch": 4.4921518987341775,
"grad_norm": 0.486239324187543,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.9319,
"step": 1109
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.49537713870186856,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.8953,
"step": 1110
},
{
"epoch": 4.500253164556962,
"grad_norm": 0.4818371524050594,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.9256,
"step": 1111
},
{
"epoch": 4.504303797468355,
"grad_norm": 0.49368146598567697,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.9198,
"step": 1112
},
{
"epoch": 4.508354430379747,
"grad_norm": 0.48671664050587576,
"learning_rate": 1.092402023876933e-06,
"loss": 0.9108,
"step": 1113
},
{
"epoch": 4.512405063291139,
"grad_norm": 0.49025403088476427,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.9405,
"step": 1114
},
{
"epoch": 4.516455696202532,
"grad_norm": 0.5158473493499364,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.9462,
"step": 1115
},
{
"epoch": 4.520506329113924,
"grad_norm": 0.5053425985862848,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.9906,
"step": 1116
},
{
"epoch": 4.524556962025317,
"grad_norm": 0.4787500490131169,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.9173,
"step": 1117
},
{
"epoch": 4.528607594936709,
"grad_norm": 0.49294390483707917,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.9405,
"step": 1118
},
{
"epoch": 4.532658227848101,
"grad_norm": 0.4989788680868574,
"learning_rate": 9.841383857091947e-07,
"loss": 0.9805,
"step": 1119
},
{
"epoch": 4.536708860759494,
"grad_norm": 0.48300161241173784,
"learning_rate": 9.666296380321616e-07,
"loss": 0.9436,
"step": 1120
},
{
"epoch": 4.540759493670886,
"grad_norm": 0.4792848191859614,
"learning_rate": 9.492741825034124e-07,
"loss": 0.9587,
"step": 1121
},
{
"epoch": 4.544810126582279,
"grad_norm": 0.48175530387839527,
"learning_rate": 9.320721589013892e-07,
"loss": 0.9204,
"step": 1122
},
{
"epoch": 4.548860759493671,
"grad_norm": 0.49982449170536253,
"learning_rate": 9.150237057688339e-07,
"loss": 0.954,
"step": 1123
},
{
"epoch": 4.552911392405063,
"grad_norm": 0.4792114547206421,
"learning_rate": 8.981289604116328e-07,
"loss": 0.913,
"step": 1124
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.4775213145882231,
"learning_rate": 8.813880588977542e-07,
"loss": 0.9593,
"step": 1125
},
{
"epoch": 4.561012658227848,
"grad_norm": 0.47393377539220544,
"learning_rate": 8.648011360561126e-07,
"loss": 0.925,
"step": 1126
},
{
"epoch": 4.56506329113924,
"grad_norm": 0.47552131232702327,
"learning_rate": 8.483683254755037e-07,
"loss": 0.9325,
"step": 1127
},
{
"epoch": 4.569113924050633,
"grad_norm": 0.4780622698366451,
"learning_rate": 8.320897595035227e-07,
"loss": 0.9406,
"step": 1128
},
{
"epoch": 4.573164556962025,
"grad_norm": 0.47799684106787527,
"learning_rate": 8.159655692455093e-07,
"loss": 0.9424,
"step": 1129
},
{
"epoch": 4.577215189873417,
"grad_norm": 0.48798877619907954,
"learning_rate": 7.999958845634648e-07,
"loss": 0.9209,
"step": 1130
},
{
"epoch": 4.58126582278481,
"grad_norm": 0.4908893482950025,
"learning_rate": 7.841808340750478e-07,
"loss": 0.9334,
"step": 1131
},
{
"epoch": 4.585316455696202,
"grad_norm": 0.46756001610164366,
"learning_rate": 7.685205451524869e-07,
"loss": 0.9159,
"step": 1132
},
{
"epoch": 4.589367088607595,
"grad_norm": 0.4846408297001695,
"learning_rate": 7.530151439216027e-07,
"loss": 0.9219,
"step": 1133
},
{
"epoch": 4.593417721518987,
"grad_norm": 0.4851700750972771,
"learning_rate": 7.376647552607675e-07,
"loss": 0.9312,
"step": 1134
},
{
"epoch": 4.597468354430379,
"grad_norm": 0.4762205354108853,
"learning_rate": 7.224695027998963e-07,
"loss": 0.9647,
"step": 1135
},
{
"epoch": 4.601518987341772,
"grad_norm": 0.47941342593528813,
"learning_rate": 7.07429508919466e-07,
"loss": 0.9197,
"step": 1136
},
{
"epoch": 4.605569620253164,
"grad_norm": 0.48676058973619135,
"learning_rate": 6.925448947495206e-07,
"loss": 0.9417,
"step": 1137
},
{
"epoch": 4.609620253164557,
"grad_norm": 0.47845031288285955,
"learning_rate": 6.778157801686936e-07,
"loss": 0.9651,
"step": 1138
},
{
"epoch": 4.613670886075949,
"grad_norm": 0.479870345175695,
"learning_rate": 6.632422838032515e-07,
"loss": 0.9216,
"step": 1139
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.48261626854757556,
"learning_rate": 6.488245230261281e-07,
"loss": 0.93,
"step": 1140
},
{
"epoch": 4.6217721518987345,
"grad_norm": 0.48963999113719264,
"learning_rate": 6.345626139559868e-07,
"loss": 0.9208,
"step": 1141
},
{
"epoch": 4.6258227848101265,
"grad_norm": 0.4728270050967934,
"learning_rate": 6.204566714562866e-07,
"loss": 0.9657,
"step": 1142
},
{
"epoch": 4.629873417721519,
"grad_norm": 0.49181619391567877,
"learning_rate": 6.06506809134344e-07,
"loss": 0.9136,
"step": 1143
},
{
"epoch": 4.633924050632912,
"grad_norm": 0.4907677663980347,
"learning_rate": 5.927131393404373e-07,
"loss": 0.9341,
"step": 1144
},
{
"epoch": 4.637974683544304,
"grad_norm": 0.4818202656623583,
"learning_rate": 5.790757731668817e-07,
"loss": 0.9052,
"step": 1145
},
{
"epoch": 4.642025316455696,
"grad_norm": 0.4679715443204932,
"learning_rate": 5.655948204471507e-07,
"loss": 0.8921,
"step": 1146
},
{
"epoch": 4.646075949367089,
"grad_norm": 0.4936700139102443,
"learning_rate": 5.522703897549875e-07,
"loss": 0.9539,
"step": 1147
},
{
"epoch": 4.650126582278481,
"grad_norm": 0.48370746040056184,
"learning_rate": 5.391025884035239e-07,
"loss": 0.9607,
"step": 1148
},
{
"epoch": 4.654177215189874,
"grad_norm": 0.47238863699404615,
"learning_rate": 5.260915224444207e-07,
"loss": 0.9371,
"step": 1149
},
{
"epoch": 4.658227848101266,
"grad_norm": 0.4746786846385125,
"learning_rate": 5.132372966670129e-07,
"loss": 0.9023,
"step": 1150
},
{
"epoch": 4.662278481012658,
"grad_norm": 0.4777743203167845,
"learning_rate": 5.005400145974704e-07,
"loss": 0.9412,
"step": 1151
},
{
"epoch": 4.666329113924051,
"grad_norm": 0.49131196824260076,
"learning_rate": 4.879997784979562e-07,
"loss": 0.9412,
"step": 1152
},
{
"epoch": 4.670379746835443,
"grad_norm": 0.4805490284056136,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.9209,
"step": 1153
},
{
"epoch": 4.674430379746836,
"grad_norm": 0.4676464440945373,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.9166,
"step": 1154
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.4846061444486225,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.9035,
"step": 1155
},
{
"epoch": 4.68253164556962,
"grad_norm": 0.4675909977013077,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.8714,
"step": 1156
},
{
"epoch": 4.686582278481013,
"grad_norm": 0.48764924945324206,
"learning_rate": 4.27657778143431e-07,
"loss": 0.9389,
"step": 1157
},
{
"epoch": 4.690632911392405,
"grad_norm": 0.4890939713308712,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.9156,
"step": 1158
},
{
"epoch": 4.694683544303797,
"grad_norm": 0.4781954370800776,
"learning_rate": 4.046237371675177e-07,
"loss": 0.9368,
"step": 1159
},
{
"epoch": 4.69873417721519,
"grad_norm": 0.4796558359921206,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.9205,
"step": 1160
},
{
"epoch": 4.702784810126582,
"grad_norm": 0.47182680285228257,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.9527,
"step": 1161
},
{
"epoch": 4.706835443037974,
"grad_norm": 0.490981831855253,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.9066,
"step": 1162
},
{
"epoch": 4.710886075949367,
"grad_norm": 0.4701721833045833,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.9491,
"step": 1163
},
{
"epoch": 4.714936708860759,
"grad_norm": 0.47271126764599375,
"learning_rate": 3.498020255137813e-07,
"loss": 0.9143,
"step": 1164
},
{
"epoch": 4.718987341772152,
"grad_norm": 0.4858026452459758,
"learning_rate": 3.393121163057811e-07,
"loss": 0.9482,
"step": 1165
},
{
"epoch": 4.723037974683544,
"grad_norm": 0.49120788823008715,
"learning_rate": 3.289805515751399e-07,
"loss": 0.962,
"step": 1166
},
{
"epoch": 4.727088607594936,
"grad_norm": 0.4798049216666636,
"learning_rate": 3.188074145308573e-07,
"loss": 0.9072,
"step": 1167
},
{
"epoch": 4.731139240506329,
"grad_norm": 0.4768317117937864,
"learning_rate": 3.087927871059804e-07,
"loss": 0.9385,
"step": 1168
},
{
"epoch": 4.735189873417721,
"grad_norm": 0.48851971055988136,
"learning_rate": 2.989367499569418e-07,
"loss": 0.9501,
"step": 1169
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.47384372830850924,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.9118,
"step": 1170
},
{
"epoch": 4.743291139240506,
"grad_norm": 0.4843776650760558,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.9545,
"step": 1171
},
{
"epoch": 4.7473417721518985,
"grad_norm": 0.48584063326244814,
"learning_rate": 2.703209675663887e-07,
"loss": 0.9403,
"step": 1172
},
{
"epoch": 4.751392405063291,
"grad_norm": 0.4938435516517681,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.9095,
"step": 1173
},
{
"epoch": 4.7554430379746835,
"grad_norm": 0.4691803107004478,
"learning_rate": 2.520381518804471e-07,
"loss": 0.9262,
"step": 1174
},
{
"epoch": 4.759493670886076,
"grad_norm": 0.47018366380385956,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.9048,
"step": 1175
},
{
"epoch": 4.7635443037974685,
"grad_norm": 0.4795392164619577,
"learning_rate": 2.343915243930317e-07,
"loss": 0.9587,
"step": 1176
},
{
"epoch": 4.767594936708861,
"grad_norm": 0.4835394318185762,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.9549,
"step": 1177
},
{
"epoch": 4.7716455696202535,
"grad_norm": 0.4860903509272599,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.9674,
"step": 1178
},
{
"epoch": 4.775696202531646,
"grad_norm": 0.4788550645219886,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.9457,
"step": 1179
},
{
"epoch": 4.779746835443038,
"grad_norm": 0.46740454028695766,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.9743,
"step": 1180
},
{
"epoch": 4.783797468354431,
"grad_norm": 0.473011630671819,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.9179,
"step": 1181
},
{
"epoch": 4.787848101265823,
"grad_norm": 0.48289218173069737,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.9442,
"step": 1182
},
{
"epoch": 4.791898734177215,
"grad_norm": 0.48050843030086904,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.9253,
"step": 1183
},
{
"epoch": 4.795949367088608,
"grad_norm": 0.47539179349160277,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.9398,
"step": 1184
},
{
"epoch": 4.8,
"grad_norm": 0.48285490373049045,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.9653,
"step": 1185
},
{
"epoch": 4.804050632911393,
"grad_norm": 0.5035968131676818,
"learning_rate": 1.557203888516745e-07,
"loss": 0.9455,
"step": 1186
},
{
"epoch": 4.808101265822785,
"grad_norm": 0.48792238816376865,
"learning_rate": 1.487313098987131e-07,
"loss": 0.9396,
"step": 1187
},
{
"epoch": 4.812151898734177,
"grad_norm": 0.4696678541207507,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.9067,
"step": 1188
},
{
"epoch": 4.81620253164557,
"grad_norm": 0.4638964066379355,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.9023,
"step": 1189
},
{
"epoch": 4.820253164556962,
"grad_norm": 0.4687403307106285,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.9209,
"step": 1190
},
{
"epoch": 4.824303797468355,
"grad_norm": 0.47590664211406536,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.9518,
"step": 1191
},
{
"epoch": 4.828354430379747,
"grad_norm": 0.4767140590447752,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.9304,
"step": 1192
},
{
"epoch": 4.832405063291139,
"grad_norm": 0.4835057794465881,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.9364,
"step": 1193
},
{
"epoch": 4.836455696202532,
"grad_norm": 0.4773343354321854,
"learning_rate": 1.042873700093061e-07,
"loss": 0.9687,
"step": 1194
},
{
"epoch": 4.840506329113924,
"grad_norm": 0.47766010652697916,
"learning_rate": 9.857879404556291e-08,
"loss": 0.9229,
"step": 1195
},
{
"epoch": 4.844556962025316,
"grad_norm": 0.46964928017130664,
"learning_rate": 9.303050139225722e-08,
"loss": 0.935,
"step": 1196
},
{
"epoch": 4.848607594936709,
"grad_norm": 0.4958105412317642,
"learning_rate": 8.76425367345779e-08,
"loss": 0.9296,
"step": 1197
},
{
"epoch": 4.852658227848101,
"grad_norm": 0.47769217954430016,
"learning_rate": 8.241494346644897e-08,
"loss": 0.9139,
"step": 1198
},
{
"epoch": 4.856708860759493,
"grad_norm": 0.48084069460056006,
"learning_rate": 7.734776369019204e-08,
"loss": 0.9217,
"step": 1199
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.4716229460529743,
"learning_rate": 7.244103821617332e-08,
"loss": 0.9307,
"step": 1200
},
{
"epoch": 4.864810126582278,
"grad_norm": 0.48101896446225584,
"learning_rate": 6.769480656248606e-08,
"loss": 0.9517,
"step": 1201
},
{
"epoch": 4.868860759493671,
"grad_norm": 0.49862786093335826,
"learning_rate": 6.310910695462635e-08,
"loss": 0.9233,
"step": 1202
},
{
"epoch": 4.872911392405063,
"grad_norm": 0.48064005100881124,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.9569,
"step": 1203
},
{
"epoch": 4.876962025316455,
"grad_norm": 0.4807453766605093,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.9256,
"step": 1204
},
{
"epoch": 4.881012658227848,
"grad_norm": 0.49389133111125944,
"learning_rate": 5.031556326567488e-08,
"loss": 0.9634,
"step": 1205
},
{
"epoch": 4.88506329113924,
"grad_norm": 0.4794395714684595,
"learning_rate": 4.637234823364312e-08,
"loss": 0.9257,
"step": 1206
},
{
"epoch": 4.889113924050633,
"grad_norm": 0.48998694987189567,
"learning_rate": 4.258983697558838e-08,
"loss": 0.9312,
"step": 1207
},
{
"epoch": 4.8931645569620255,
"grad_norm": 0.477834761455395,
"learning_rate": 3.896805995533548e-08,
"loss": 0.9527,
"step": 1208
},
{
"epoch": 4.8972151898734175,
"grad_norm": 0.46973403593858837,
"learning_rate": 3.550704634218028e-08,
"loss": 0.921,
"step": 1209
},
{
"epoch": 4.9012658227848105,
"grad_norm": 0.4927358973056418,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.953,
"step": 1210
},
{
"epoch": 4.905316455696203,
"grad_norm": 0.47813099120087227,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.9031,
"step": 1211
},
{
"epoch": 4.909367088607595,
"grad_norm": 0.4782560914870409,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.9074,
"step": 1212
},
{
"epoch": 4.913417721518988,
"grad_norm": 0.4808870333173331,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.9229,
"step": 1213
},
{
"epoch": 4.91746835443038,
"grad_norm": 0.4823011942118952,
"learning_rate": 2.061435966221881e-08,
"loss": 0.9205,
"step": 1214
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.46839810768443396,
"learning_rate": 1.811846652477245e-08,
"loss": 0.9684,
"step": 1215
},
{
"epoch": 4.925569620253165,
"grad_norm": 0.4633441667227147,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.9412,
"step": 1216
},
{
"epoch": 4.929620253164557,
"grad_norm": 0.48139227939089574,
"learning_rate": 1.360949303591097e-08,
"loss": 0.9507,
"step": 1217
},
{
"epoch": 4.93367088607595,
"grad_norm": 0.47258068078943627,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.9348,
"step": 1218
},
{
"epoch": 4.937721518987342,
"grad_norm": 0.4841724364668855,
"learning_rate": 9.744388816668172e-09,
"loss": 0.933,
"step": 1219
},
{
"epoch": 4.941772151898734,
"grad_norm": 0.459073098171973,
"learning_rate": 8.05332740472009e-09,
"loss": 0.9438,
"step": 1220
},
{
"epoch": 4.945822784810127,
"grad_norm": 0.4838952477287419,
"learning_rate": 6.523278382872811e-09,
"loss": 0.9227,
"step": 1221
},
{
"epoch": 4.949873417721519,
"grad_norm": 0.47564359291958236,
"learning_rate": 5.15425407393133e-09,
"loss": 0.9165,
"step": 1222
},
{
"epoch": 4.953924050632912,
"grad_norm": 0.4741419492012334,
"learning_rate": 3.94626550383137e-09,
"loss": 0.9473,
"step": 1223
},
{
"epoch": 4.957974683544304,
"grad_norm": 0.4888279516644087,
"learning_rate": 2.899322401546112e-09,
"loss": 0.9046,
"step": 1224
},
{
"epoch": 4.962025316455696,
"grad_norm": 0.4885213039631545,
"learning_rate": 2.013433199010706e-09,
"loss": 0.9226,
"step": 1225
},
{
"epoch": 4.966075949367089,
"grad_norm": 0.4689825290213745,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.9663,
"step": 1226
},
{
"epoch": 4.970126582278481,
"grad_norm": 0.48071422023048466,
"learning_rate": 7.248437353468695e-10,
"loss": 0.9345,
"step": 1227
},
{
"epoch": 4.974177215189873,
"grad_norm": 0.47579961197172027,
"learning_rate": 3.221538523412449e-10,
"loss": 0.9636,
"step": 1228
},
{
"epoch": 4.978227848101266,
"grad_norm": 0.4753620666320983,
"learning_rate": 8.053862524670663e-11,
"loss": 0.9289,
"step": 1229
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.4775392960291494,
"learning_rate": 0.0,
"loss": 0.9537,
"step": 1230
},
{
"epoch": 4.982278481012658,
"step": 1230,
"total_flos": 2.219643011007873e+18,
"train_loss": 1.2360469060215524,
"train_runtime": 17782.2852,
"train_samples_per_second": 8.885,
"train_steps_per_second": 0.069
}
],
"logging_steps": 1,
"max_steps": 1230,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.219643011007873e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}