a1_code_apps_buggy / trainer_state.json
sedrickkeh's picture
End of training
2910690 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": 6.564570271110209,
"learning_rate": 3.2520325203252037e-07,
"loss": 1.0203,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 6.650738783471983,
"learning_rate": 6.504065040650407e-07,
"loss": 1.0385,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 6.614147504931167,
"learning_rate": 9.75609756097561e-07,
"loss": 1.0402,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 6.530291403302496,
"learning_rate": 1.3008130081300815e-06,
"loss": 1.0334,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 6.41749333801591,
"learning_rate": 1.6260162601626018e-06,
"loss": 1.0435,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 6.056787850413258,
"learning_rate": 1.951219512195122e-06,
"loss": 1.028,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 4.897869670249938,
"learning_rate": 2.2764227642276426e-06,
"loss": 1.0025,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 4.697564918952781,
"learning_rate": 2.601626016260163e-06,
"loss": 0.9975,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 2.954844452612212,
"learning_rate": 2.926829268292683e-06,
"loss": 0.9469,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 2.639122444841373,
"learning_rate": 3.2520325203252037e-06,
"loss": 0.9277,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 2.5418602310962704,
"learning_rate": 3.577235772357724e-06,
"loss": 0.9417,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 4.03718159096835,
"learning_rate": 3.902439024390244e-06,
"loss": 0.9237,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 4.366849210843851,
"learning_rate": 4.227642276422765e-06,
"loss": 0.9113,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 4.096043815356621,
"learning_rate": 4.552845528455285e-06,
"loss": 0.9028,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 3.839143138149361,
"learning_rate": 4.8780487804878055e-06,
"loss": 0.8915,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 2.8406882053236946,
"learning_rate": 5.203252032520326e-06,
"loss": 0.8502,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 2.4159627534861863,
"learning_rate": 5.528455284552846e-06,
"loss": 0.8289,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 1.9600624679606649,
"learning_rate": 5.853658536585366e-06,
"loss": 0.8145,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 1.5683530933764893,
"learning_rate": 6.178861788617887e-06,
"loss": 0.7996,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.3701709902159083,
"learning_rate": 6.504065040650407e-06,
"loss": 0.7741,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.2789785992498734,
"learning_rate": 6.829268292682928e-06,
"loss": 0.7817,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.3743306709085863,
"learning_rate": 7.154471544715448e-06,
"loss": 0.765,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.187434805695583,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.7555,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 0.9570925173398591,
"learning_rate": 7.804878048780489e-06,
"loss": 0.753,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 1.0520681923697377,
"learning_rate": 8.130081300813009e-06,
"loss": 0.7387,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 0.9197682174374587,
"learning_rate": 8.45528455284553e-06,
"loss": 0.7223,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 1.1042359633245316,
"learning_rate": 8.78048780487805e-06,
"loss": 0.7306,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 0.8572592600414636,
"learning_rate": 9.10569105691057e-06,
"loss": 0.7216,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 0.9818438812246167,
"learning_rate": 9.43089430894309e-06,
"loss": 0.7071,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 0.9419480353346157,
"learning_rate": 9.756097560975611e-06,
"loss": 0.6946,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 0.8313149262943869,
"learning_rate": 1.008130081300813e-05,
"loss": 0.7126,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.8298860235308718,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.7089,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.8134047753003946,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.6982,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.7769855162605798,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.7092,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.820193998635746,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.6886,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.6819296155154736,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.6907,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.7388097370641948,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.6951,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.7605240771699983,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.6767,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.5810444149164746,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.6791,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.6869956346792637,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.6799,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.5959562649272534,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.6749,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.655740083095977,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.6721,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.6639807700183801,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.6744,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.7322596523087247,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.6654,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.6644531798869938,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.6606,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.6076817053663623,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.6607,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.8385624830938764,
"learning_rate": 1.528455284552846e-05,
"loss": 0.6603,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.6363488768112967,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.6462,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.6651135428820405,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.6481,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.6389919323004513,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.6527,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.7015386599631557,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.6522,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.6407726858205297,
"learning_rate": 1.691056910569106e-05,
"loss": 0.6502,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.7924631256939718,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.6441,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.5812483203020653,
"learning_rate": 1.75609756097561e-05,
"loss": 0.6471,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.7102071019127485,
"learning_rate": 1.788617886178862e-05,
"loss": 0.6436,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.5100149533077052,
"learning_rate": 1.821138211382114e-05,
"loss": 0.6362,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.6168560658065412,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.6471,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.7653036039130323,
"learning_rate": 1.886178861788618e-05,
"loss": 0.655,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.705043471818363,
"learning_rate": 1.91869918699187e-05,
"loss": 0.6308,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.6729138055257656,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.6291,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.7335014121637716,
"learning_rate": 1.983739837398374e-05,
"loss": 0.6414,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.6947102253540197,
"learning_rate": 2.016260162601626e-05,
"loss": 0.6374,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.7462248126147633,
"learning_rate": 2.048780487804878e-05,
"loss": 0.6319,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.7518940595111343,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.6385,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.7085643731608854,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.6365,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.8256116180445547,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.6254,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.642525439331126,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.6193,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 0.6997893378520257,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.6306,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 1.0823586106527845,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.6293,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 0.9106013098618413,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.6377,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 0.8840028562136617,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.6215,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.8305036585322215,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.6279,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 0.6717306331702845,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.6169,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.6559549815414804,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.6262,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.8398936247381757,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.6168,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 1.1425645471268258,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.6223,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.8108651206952732,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.6242,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 0.8479442375249004,
"learning_rate": 2.536585365853659e-05,
"loss": 0.6264,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 0.7866533454891944,
"learning_rate": 2.569105691056911e-05,
"loss": 0.6101,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 0.6517303611522073,
"learning_rate": 2.601626016260163e-05,
"loss": 0.6151,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 0.6244955633683347,
"learning_rate": 2.634146341463415e-05,
"loss": 0.6364,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 0.6386355200294336,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.6089,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 1.000362220085523,
"learning_rate": 2.699186991869919e-05,
"loss": 0.616,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 1.3291714652172975,
"learning_rate": 2.731707317073171e-05,
"loss": 0.6173,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 0.853005453417322,
"learning_rate": 2.764227642276423e-05,
"loss": 0.6215,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 1.0901793889525349,
"learning_rate": 2.796747967479675e-05,
"loss": 0.6223,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 0.8855129805961502,
"learning_rate": 2.829268292682927e-05,
"loss": 0.6148,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 0.8278427128224868,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.6279,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 1.0100881856590687,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.6302,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 1.049905647305169,
"learning_rate": 2.926829268292683e-05,
"loss": 0.6224,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 0.7468359418032106,
"learning_rate": 2.959349593495935e-05,
"loss": 0.6081,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 0.8394481514919352,
"learning_rate": 2.991869918699187e-05,
"loss": 0.6178,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 1.086004041558417,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.6172,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 0.874124656339068,
"learning_rate": 3.056910569105692e-05,
"loss": 0.6191,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 1.0049715677780109,
"learning_rate": 3.089430894308943e-05,
"loss": 0.6398,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 1.1106261768795067,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.614,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 1.0701383877142479,
"learning_rate": 3.154471544715447e-05,
"loss": 0.5972,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 1.3058303691016806,
"learning_rate": 3.186991869918699e-05,
"loss": 0.5967,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 0.8945880057599052,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.6001,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 1.505815254468162,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.6177,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 0.7896788945852753,
"learning_rate": 3.284552845528456e-05,
"loss": 0.6074,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 1.5063459236203187,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.6192,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 0.9622056360939301,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.6231,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 1.377960227346876,
"learning_rate": 3.382113821138212e-05,
"loss": 0.6151,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 1.111431540708072,
"learning_rate": 3.414634146341463e-05,
"loss": 0.6141,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 1.0896888978753358,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.6085,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 0.9006121368683369,
"learning_rate": 3.479674796747968e-05,
"loss": 0.615,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 1.3373945431728094,
"learning_rate": 3.51219512195122e-05,
"loss": 0.6219,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 1.4405241089551615,
"learning_rate": 3.544715447154472e-05,
"loss": 0.5953,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 0.727069658355342,
"learning_rate": 3.577235772357724e-05,
"loss": 0.585,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 1.0725849722920129,
"learning_rate": 3.609756097560976e-05,
"loss": 0.6111,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 1.1832209498458466,
"learning_rate": 3.642276422764228e-05,
"loss": 0.6175,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 1.1714348795670066,
"learning_rate": 3.67479674796748e-05,
"loss": 0.6122,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 0.7621246929343158,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.6107,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 1.0197188601327734,
"learning_rate": 3.739837398373984e-05,
"loss": 0.6163,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 0.8006747266107718,
"learning_rate": 3.772357723577236e-05,
"loss": 0.5912,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 0.7556341205057592,
"learning_rate": 3.804878048780488e-05,
"loss": 0.607,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 0.6081987583020876,
"learning_rate": 3.83739837398374e-05,
"loss": 0.6002,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 0.8503645028535891,
"learning_rate": 3.869918699186992e-05,
"loss": 0.5946,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 1.3073028094497097,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.6118,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 1.3644003014429726,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.6102,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 1.127839773126872,
"learning_rate": 3.967479674796748e-05,
"loss": 0.6065,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 1.0831017380990702,
"learning_rate": 4e-05,
"loss": 0.606,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 1.5853420422926923,
"learning_rate": 3.999991946137476e-05,
"loss": 0.5912,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.793390955079785,
"learning_rate": 3.999967784614766e-05,
"loss": 0.6022,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 1.2622879493484478,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.6158,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 1.3192026250394528,
"learning_rate": 3.999871139496895e-05,
"loss": 0.6016,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 0.6936197494802387,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.5894,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 1.2120941519444297,
"learning_rate": 3.999710067759846e-05,
"loss": 0.6017,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 1.0809286793562785,
"learning_rate": 3.999605373449617e-05,
"loss": 0.5971,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.9287302934794337,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.6025,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 0.8721772906231454,
"learning_rate": 3.999347672161713e-05,
"loss": 0.5892,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 0.8565777922953347,
"learning_rate": 3.999194667259528e-05,
"loss": 0.5885,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 0.6328534981204801,
"learning_rate": 3.999025561118334e-05,
"loss": 0.5878,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 1.083745038031027,
"learning_rate": 3.998840355100086e-05,
"loss": 0.609,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 0.7777475515163781,
"learning_rate": 3.998639050696409e-05,
"loss": 0.5891,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 0.8295426885260478,
"learning_rate": 3.998421649528582e-05,
"loss": 0.6008,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 0.9225377078990974,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.601,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.7287269412173871,
"learning_rate": 3.997938564033779e-05,
"loss": 0.5975,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 0.7698478560474523,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.6027,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 1.0180190048982491,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.6072,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 1.0718164729945796,
"learning_rate": 3.997093258045973e-05,
"loss": 0.6059,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.9062956904845919,
"learning_rate": 3.996779317598936e-05,
"loss": 0.6048,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 0.7206658817394783,
"learning_rate": 3.996449295365782e-05,
"loss": 0.6039,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.6905864519471087,
"learning_rate": 3.996103194004467e-05,
"loss": 0.5965,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.6631960368028624,
"learning_rate": 3.995741016302441e-05,
"loss": 0.5944,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 0.6486839552139754,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.5799,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 0.9001676889707041,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.6097,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 0.8652406748699969,
"learning_rate": 3.994558054968643e-05,
"loss": 0.5907,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.7803695595136731,
"learning_rate": 3.994131602367481e-05,
"loss": 0.6095,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 0.8286718821108976,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.6005,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.8622179171607819,
"learning_rate": 3.993230519343752e-05,
"loss": 0.5852,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 0.801228850055872,
"learning_rate": 3.992755896178383e-05,
"loss": 0.601,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.6632700642330694,
"learning_rate": 3.992265223630981e-05,
"loss": 0.6036,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 0.6220947139941152,
"learning_rate": 3.991758505653355e-05,
"loss": 0.5954,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 0.9378539046441428,
"learning_rate": 3.991235746326543e-05,
"loss": 0.6139,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.9627811102857091,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.5928,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 0.6700951268146137,
"learning_rate": 3.990142120595444e-05,
"loss": 0.5809,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 0.5691426787063265,
"learning_rate": 3.98957126299907e-05,
"loss": 0.6003,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 0.6546082926861614,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.6004,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 0.6351327048679135,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.591,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 0.5103220402433847,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.5809,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 0.5992590085986876,
"learning_rate": 3.98712764319087e-05,
"loss": 0.5881,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.5805664123964966,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.5821,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 0.637499086840973,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.5726,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.5193481787115525,
"learning_rate": 3.985126869010129e-05,
"loss": 0.5981,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 0.5917019873729756,
"learning_rate": 3.984427961114833e-05,
"loss": 0.582,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 0.562372949913346,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.5673,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 0.5394863761248704,
"learning_rate": 3.982982204151901e-05,
"loss": 0.5735,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.5378152288552683,
"learning_rate": 3.982235366728193e-05,
"loss": 0.5952,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.5384479511773729,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.5764,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 0.7351394847349977,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.5948,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.7806494463280336,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.5941,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.5535993573594569,
"learning_rate": 3.979088432599627e-05,
"loss": 0.5918,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.6276341166796098,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.563,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 0.5769963122919576,
"learning_rate": 3.977419304032111e-05,
"loss": 0.5873,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 0.5849096162431718,
"learning_rate": 3.976560847560697e-05,
"loss": 0.5879,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 0.6841253818050892,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.5769,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 0.5628086965251162,
"learning_rate": 3.974796184811956e-05,
"loss": 0.575,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.5692485390040511,
"learning_rate": 3.973889992746979e-05,
"loss": 0.59,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.697739254711223,
"learning_rate": 3.972967903243361e-05,
"loss": 0.5796,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.6514763901264384,
"learning_rate": 3.972029923727486e-05,
"loss": 0.5874,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.7507531110781857,
"learning_rate": 3.971076061753709e-05,
"loss": 0.5867,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.9375272225208995,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.5785,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.7915135079694513,
"learning_rate": 3.969120721289402e-05,
"loss": 0.5918,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.7131752862870969,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.5882,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.8806661178336984,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.5834,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.9392466417102341,
"learning_rate": 3.966068788369422e-05,
"loss": 0.5796,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.7268350319704158,
"learning_rate": 3.965019797448622e-05,
"loss": 0.5772,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.9309798440172284,
"learning_rate": 3.963954980528515e-05,
"loss": 0.5873,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.9483799090176203,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.5853,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.6435785166753079,
"learning_rate": 3.961777903121329e-05,
"loss": 0.5878,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.7409893830891758,
"learning_rate": 3.960665660168131e-05,
"loss": 0.5848,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.9045842280313247,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.5841,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.6989656259612773,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.5785,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 0.7808171556352567,
"learning_rate": 3.957234222185657e-05,
"loss": 0.5767,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.5497165981599593,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.5765,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 0.843619669398686,
"learning_rate": 3.954867765033605e-05,
"loss": 0.6011,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.6938688973478265,
"learning_rate": 3.953660915306728e-05,
"loss": 0.5704,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.7499912854011482,
"learning_rate": 3.952438331063419e-05,
"loss": 0.5922,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.7470769759183343,
"learning_rate": 3.951200022150205e-05,
"loss": 0.5778,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.575090784880628,
"learning_rate": 3.949945998540253e-05,
"loss": 0.5682,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.5693804508440627,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.5898,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.6242051903369166,
"learning_rate": 3.947390847755559e-05,
"loss": 0.585,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.5766577016406277,
"learning_rate": 3.946089741159648e-05,
"loss": 0.5729,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.5500534592192335,
"learning_rate": 3.944772961024501e-05,
"loss": 0.5724,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.5673605834944145,
"learning_rate": 3.943440517955285e-05,
"loss": 0.5828,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.5309460640703659,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.5637,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.477546672658377,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.5881,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.6374336706663287,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.5771,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.6702572896844226,
"learning_rate": 3.937954332854371e-05,
"loss": 0.5862,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.6869811618756635,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.5806,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.4822968557887995,
"learning_rate": 3.935117547697387e-05,
"loss": 0.5749,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.5485198171611837,
"learning_rate": 3.933675771619675e-05,
"loss": 0.5728,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.5558133061688061,
"learning_rate": 3.932218421983131e-05,
"loss": 0.5786,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.5577480442353234,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.5742,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.5309137727656644,
"learning_rate": 3.929257049108054e-05,
"loss": 0.5707,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.7145671598005969,
"learning_rate": 3.927753049720011e-05,
"loss": 0.573,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.5757466362979069,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.5801,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.5157033491950906,
"learning_rate": 3.92469848560784e-05,
"loss": 0.5834,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.5268121220311328,
"learning_rate": 3.923147945484751e-05,
"loss": 0.5712,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.5964143931886458,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.5453,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.570259606900189,
"learning_rate": 3.920000411543654e-05,
"loss": 0.5744,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.5799686353488258,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.5859,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.5193159233810167,
"learning_rate": 3.916791024049648e-05,
"loss": 0.5674,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.607738612688029,
"learning_rate": 3.91516316745245e-05,
"loss": 0.5768,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.6351616989896921,
"learning_rate": 3.913519886394389e-05,
"loss": 0.5748,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.5859319684684778,
"learning_rate": 3.911861194110225e-05,
"loss": 0.5702,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.6086856494234545,
"learning_rate": 3.910187103958837e-05,
"loss": 0.5761,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.6647679228916327,
"learning_rate": 3.908497629423117e-05,
"loss": 0.5596,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.46224376435435793,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.5638,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.6227162012969892,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.5679,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.6208518164169488,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.5774,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.4726307884464191,
"learning_rate": 3.901586161429081e-05,
"loss": 0.5665,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.4764631417308816,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.5725,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.5241723094240475,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.5674,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.49396596096330947,
"learning_rate": 3.896241703466667e-05,
"loss": 0.5621,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.45473897239117,
"learning_rate": 3.894429654085585e-05,
"loss": 0.5665,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.599234641911452,
"learning_rate": 3.892602347228505e-05,
"loss": 0.5606,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.4820305373496122,
"learning_rate": 3.890759797612307e-05,
"loss": 0.5494,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.4654685740033189,
"learning_rate": 3.888902020076632e-05,
"loss": 0.5614,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.3995218195338633,
"learning_rate": 3.887029029583764e-05,
"loss": 0.5716,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.538575274633003,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.5857,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.5304797129532196,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.5661,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.4450298420383671,
"learning_rate": 3.881318931821972e-05,
"loss": 0.5858,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.4965215825539208,
"learning_rate": 3.879385241571817e-05,
"loss": 0.5768,
"step": 246
},
{
"epoch": 1.000506329113924,
"grad_norm": 0.4911719503289595,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.642,
"step": 247
},
{
"epoch": 1.0045569620253165,
"grad_norm": 0.5462753665885739,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.5,
"step": 248
},
{
"epoch": 1.0086075949367088,
"grad_norm": 0.5076903569558483,
"learning_rate": 3.873493415863256e-05,
"loss": 0.4956,
"step": 249
},
{
"epoch": 1.0126582278481013,
"grad_norm": 0.6089778865245409,
"learning_rate": 3.871499275032111e-05,
"loss": 0.5349,
"step": 250
},
{
"epoch": 1.0167088607594936,
"grad_norm": 0.702169301061408,
"learning_rate": 3.869490061403091e-05,
"loss": 0.493,
"step": 251
},
{
"epoch": 1.0207594936708861,
"grad_norm": 0.8026148272696562,
"learning_rate": 3.867465791158124e-05,
"loss": 0.5141,
"step": 252
},
{
"epoch": 1.0248101265822784,
"grad_norm": 0.4939988696794215,
"learning_rate": 3.865426480600407e-05,
"loss": 0.5405,
"step": 253
},
{
"epoch": 1.028860759493671,
"grad_norm": 0.7095310574761051,
"learning_rate": 3.863372146154264e-05,
"loss": 0.4867,
"step": 254
},
{
"epoch": 1.0329113924050632,
"grad_norm": 0.6764438520786151,
"learning_rate": 3.861302804365024e-05,
"loss": 0.4793,
"step": 255
},
{
"epoch": 1.0369620253164558,
"grad_norm": 0.6479246758643384,
"learning_rate": 3.85921847189888e-05,
"loss": 0.5243,
"step": 256
},
{
"epoch": 1.041012658227848,
"grad_norm": 0.5822702638672428,
"learning_rate": 3.85711916554276e-05,
"loss": 0.4992,
"step": 257
},
{
"epoch": 1.0450632911392406,
"grad_norm": 0.4678047543662033,
"learning_rate": 3.85500490220419e-05,
"loss": 0.4708,
"step": 258
},
{
"epoch": 1.0491139240506329,
"grad_norm": 0.5046406845769702,
"learning_rate": 3.852875698911154e-05,
"loss": 0.5186,
"step": 259
},
{
"epoch": 1.0531645569620254,
"grad_norm": 0.6440144071397198,
"learning_rate": 3.850731572811963e-05,
"loss": 0.5229,
"step": 260
},
{
"epoch": 1.0572151898734177,
"grad_norm": 0.6145237371148581,
"learning_rate": 3.848572541175116e-05,
"loss": 0.4843,
"step": 261
},
{
"epoch": 1.0612658227848102,
"grad_norm": 0.5279448869618092,
"learning_rate": 3.846398621389154e-05,
"loss": 0.505,
"step": 262
},
{
"epoch": 1.0653164556962025,
"grad_norm": 0.534131638300486,
"learning_rate": 3.84420983096253e-05,
"loss": 0.516,
"step": 263
},
{
"epoch": 1.069367088607595,
"grad_norm": 0.5039416932162349,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.4968,
"step": 264
},
{
"epoch": 1.0734177215189873,
"grad_norm": 0.5403174075893369,
"learning_rate": 3.839787708819787e-05,
"loss": 0.527,
"step": 265
},
{
"epoch": 1.0774683544303798,
"grad_norm": 0.5315447439026343,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.5299,
"step": 266
},
{
"epoch": 1.081518987341772,
"grad_norm": 0.6367964693960507,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.5009,
"step": 267
},
{
"epoch": 1.0855696202531646,
"grad_norm": 0.5667181703889953,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.4863,
"step": 268
},
{
"epoch": 1.089620253164557,
"grad_norm": 0.605272338071571,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.4964,
"step": 269
},
{
"epoch": 1.0936708860759494,
"grad_norm": 0.4637038946288315,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.5062,
"step": 270
},
{
"epoch": 1.0977215189873417,
"grad_norm": 0.49352417290908995,
"learning_rate": 3.826166304956594e-05,
"loss": 0.4939,
"step": 271
},
{
"epoch": 1.1017721518987342,
"grad_norm": 0.5861347469085959,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.4979,
"step": 272
},
{
"epoch": 1.1058227848101265,
"grad_norm": 0.5932503533462071,
"learning_rate": 3.821507978767479e-05,
"loss": 0.5198,
"step": 273
},
{
"epoch": 1.109873417721519,
"grad_norm": 0.5988482547051569,
"learning_rate": 3.819156801001693e-05,
"loss": 0.5379,
"step": 274
},
{
"epoch": 1.1139240506329113,
"grad_norm": 0.5636960497240129,
"learning_rate": 3.816790971997121e-05,
"loss": 0.5,
"step": 275
},
{
"epoch": 1.1179746835443038,
"grad_norm": 0.4693541780595626,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.4944,
"step": 276
},
{
"epoch": 1.1220253164556961,
"grad_norm": 0.4498812277274408,
"learning_rate": 3.81201543660571e-05,
"loss": 0.5081,
"step": 277
},
{
"epoch": 1.1260759493670887,
"grad_norm": 0.48502018838516864,
"learning_rate": 3.809605768680377e-05,
"loss": 0.5138,
"step": 278
},
{
"epoch": 1.130126582278481,
"grad_norm": 0.5100782498582531,
"learning_rate": 3.807181526438958e-05,
"loss": 0.5017,
"step": 279
},
{
"epoch": 1.1341772151898735,
"grad_norm": 0.49575577864415676,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.5073,
"step": 280
},
{
"epoch": 1.1382278481012658,
"grad_norm": 0.4426542967162579,
"learning_rate": 3.802289397223145e-05,
"loss": 0.5053,
"step": 281
},
{
"epoch": 1.1422784810126583,
"grad_norm": 0.7325627303150797,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.4844,
"step": 282
},
{
"epoch": 1.1463291139240506,
"grad_norm": 0.5042618775149287,
"learning_rate": 3.797339206560096e-05,
"loss": 0.5137,
"step": 283
},
{
"epoch": 1.150379746835443,
"grad_norm": 0.5329043922123483,
"learning_rate": 3.794842387948027e-05,
"loss": 0.4924,
"step": 284
},
{
"epoch": 1.1544303797468354,
"grad_norm": 0.6683948512792193,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.5341,
"step": 285
},
{
"epoch": 1.158481012658228,
"grad_norm": 0.6502893438195617,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.4898,
"step": 286
},
{
"epoch": 1.1625316455696202,
"grad_norm": 0.44132106199151894,
"learning_rate": 3.787265280646825e-05,
"loss": 0.5006,
"step": 287
},
{
"epoch": 1.1665822784810127,
"grad_norm": 0.564320460737409,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.5157,
"step": 288
},
{
"epoch": 1.170632911392405,
"grad_norm": 0.761894141803819,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.4988,
"step": 289
},
{
"epoch": 1.1746835443037975,
"grad_norm": 0.597259237426647,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.5107,
"step": 290
},
{
"epoch": 1.1787341772151898,
"grad_norm": 0.6243553477601466,
"learning_rate": 3.776961046830306e-05,
"loss": 0.5247,
"step": 291
},
{
"epoch": 1.1827848101265823,
"grad_norm": 0.49136544761899364,
"learning_rate": 3.774349157719215e-05,
"loss": 0.4776,
"step": 292
},
{
"epoch": 1.1868354430379746,
"grad_norm": 0.57470710253024,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.5304,
"step": 293
},
{
"epoch": 1.1908860759493671,
"grad_norm": 0.576083976581777,
"learning_rate": 3.769082529555359e-05,
"loss": 0.4803,
"step": 294
},
{
"epoch": 1.1949367088607594,
"grad_norm": 0.4633584204581518,
"learning_rate": 3.766427832919294e-05,
"loss": 0.4728,
"step": 295
},
{
"epoch": 1.198987341772152,
"grad_norm": 0.6333179972119941,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.5247,
"step": 296
},
{
"epoch": 1.2030379746835442,
"grad_norm": 0.824330101409083,
"learning_rate": 3.761075781441526e-05,
"loss": 0.4719,
"step": 297
},
{
"epoch": 1.2070886075949367,
"grad_norm": 0.6431007647363081,
"learning_rate": 3.75837846970451e-05,
"loss": 0.5216,
"step": 298
},
{
"epoch": 1.211139240506329,
"grad_norm": 0.6151360446558434,
"learning_rate": 3.755666996229032e-05,
"loss": 0.5283,
"step": 299
},
{
"epoch": 1.2151898734177216,
"grad_norm": 0.5492730841154702,
"learning_rate": 3.752941382852927e-05,
"loss": 0.4741,
"step": 300
},
{
"epoch": 1.2192405063291138,
"grad_norm": 0.5291000560255255,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.5216,
"step": 301
},
{
"epoch": 1.2232911392405064,
"grad_norm": 0.5791154803753256,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.527,
"step": 302
},
{
"epoch": 1.2273417721518987,
"grad_norm": 0.6533348658745074,
"learning_rate": 3.744679923406351e-05,
"loss": 0.5108,
"step": 303
},
{
"epoch": 1.2313924050632912,
"grad_norm": 0.5728447147335255,
"learning_rate": 3.741897971081043e-05,
"loss": 0.5061,
"step": 304
},
{
"epoch": 1.2354430379746835,
"grad_norm": 0.6673968007477146,
"learning_rate": 3.739101989748946e-05,
"loss": 0.5298,
"step": 305
},
{
"epoch": 1.239493670886076,
"grad_norm": 0.5423297234591297,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.5258,
"step": 306
},
{
"epoch": 1.2435443037974683,
"grad_norm": 0.5174068860035399,
"learning_rate": 3.73346803025098e-05,
"loss": 0.4786,
"step": 307
},
{
"epoch": 1.2475949367088608,
"grad_norm": 0.5596704899526621,
"learning_rate": 3.730630097460247e-05,
"loss": 0.5212,
"step": 308
},
{
"epoch": 1.251645569620253,
"grad_norm": 0.46627120376462033,
"learning_rate": 3.727778226412628e-05,
"loss": 0.4951,
"step": 309
},
{
"epoch": 1.2556962025316456,
"grad_norm": 0.5966817153994897,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.5115,
"step": 310
},
{
"epoch": 1.259746835443038,
"grad_norm": 0.5480203470456343,
"learning_rate": 3.722032761533114e-05,
"loss": 0.4777,
"step": 311
},
{
"epoch": 1.2637974683544304,
"grad_norm": 0.41574559949664636,
"learning_rate": 3.719139213974403e-05,
"loss": 0.5138,
"step": 312
},
{
"epoch": 1.2678481012658227,
"grad_norm": 0.4836180173751796,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.4933,
"step": 313
},
{
"epoch": 1.2718987341772152,
"grad_norm": 0.46150626181817156,
"learning_rate": 3.713310605140055e-05,
"loss": 0.5468,
"step": 314
},
{
"epoch": 1.2759493670886077,
"grad_norm": 0.42552108821103884,
"learning_rate": 3.710375590807233e-05,
"loss": 0.5001,
"step": 315
},
{
"epoch": 1.28,
"grad_norm": 0.5018556644586456,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.4987,
"step": 316
},
{
"epoch": 1.2840506329113923,
"grad_norm": 0.4582604777163456,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.52,
"step": 317
},
{
"epoch": 1.2881012658227848,
"grad_norm": 0.5728156175462885,
"learning_rate": 3.701487992136854e-05,
"loss": 0.4715,
"step": 318
},
{
"epoch": 1.2921518987341774,
"grad_norm": 0.526676482629563,
"learning_rate": 3.69849802022222e-05,
"loss": 0.5027,
"step": 319
},
{
"epoch": 1.2962025316455696,
"grad_norm": 0.5117111482253465,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.5204,
"step": 320
},
{
"epoch": 1.300253164556962,
"grad_norm": 0.6244225506214646,
"learning_rate": 3.692477062175289e-05,
"loss": 0.538,
"step": 321
},
{
"epoch": 1.3043037974683545,
"grad_norm": 0.5030090673395848,
"learning_rate": 3.689446124534958e-05,
"loss": 0.5033,
"step": 322
},
{
"epoch": 1.308354430379747,
"grad_norm": 0.5008884435507844,
"learning_rate": 3.686401580327799e-05,
"loss": 0.5131,
"step": 323
},
{
"epoch": 1.3124050632911393,
"grad_norm": 0.5216919921148138,
"learning_rate": 3.683343454074149e-05,
"loss": 0.4754,
"step": 324
},
{
"epoch": 1.3164556962025316,
"grad_norm": 0.5699312532891941,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.5054,
"step": 325
},
{
"epoch": 1.320506329113924,
"grad_norm": 0.567416945247353,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.535,
"step": 326
},
{
"epoch": 1.3245569620253166,
"grad_norm": 0.4789282651780934,
"learning_rate": 3.674087829877297e-05,
"loss": 0.4861,
"step": 327
},
{
"epoch": 1.3286075949367089,
"grad_norm": 0.5283050430698466,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.5181,
"step": 328
},
{
"epoch": 1.3326582278481012,
"grad_norm": 0.45681181517700054,
"learning_rate": 3.667849957966501e-05,
"loss": 0.4784,
"step": 329
},
{
"epoch": 1.3367088607594937,
"grad_norm": 0.53145561635262,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.5019,
"step": 330
},
{
"epoch": 1.3407594936708862,
"grad_norm": 0.5255722024879411,
"learning_rate": 3.661558355626795e-05,
"loss": 0.5326,
"step": 331
},
{
"epoch": 1.3448101265822785,
"grad_norm": 0.4254435305868729,
"learning_rate": 3.658392468818163e-05,
"loss": 0.4613,
"step": 332
},
{
"epoch": 1.3488607594936708,
"grad_norm": 0.505453621876707,
"learning_rate": 3.655213225544574e-05,
"loss": 0.5074,
"step": 333
},
{
"epoch": 1.3529113924050633,
"grad_norm": 0.4497640558088102,
"learning_rate": 3.652020651411218e-05,
"loss": 0.5323,
"step": 334
},
{
"epoch": 1.3569620253164558,
"grad_norm": 0.564363720010871,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.5431,
"step": 335
},
{
"epoch": 1.3610126582278481,
"grad_norm": 0.6168509814992285,
"learning_rate": 3.645595613522574e-05,
"loss": 0.4818,
"step": 336
},
{
"epoch": 1.3650632911392404,
"grad_norm": 0.47853070251120494,
"learning_rate": 3.642363201513657e-05,
"loss": 0.4899,
"step": 337
},
{
"epoch": 1.369113924050633,
"grad_norm": 0.5064919185478629,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.5019,
"step": 338
},
{
"epoch": 1.3731645569620254,
"grad_norm": 0.5614201629996941,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.4906,
"step": 339
},
{
"epoch": 1.3772151898734177,
"grad_norm": 0.4675607697195212,
"learning_rate": 3.632586705948318e-05,
"loss": 0.483,
"step": 340
},
{
"epoch": 1.38126582278481,
"grad_norm": 0.44089253622917535,
"learning_rate": 3.629301541734311e-05,
"loss": 0.4828,
"step": 341
},
{
"epoch": 1.3853164556962025,
"grad_norm": 0.5270632201524744,
"learning_rate": 3.626003255349676e-05,
"loss": 0.5269,
"step": 342
},
{
"epoch": 1.389367088607595,
"grad_norm": 0.46132316827319,
"learning_rate": 3.622691873358357e-05,
"loss": 0.477,
"step": 343
},
{
"epoch": 1.3934177215189874,
"grad_norm": 0.48056020256482535,
"learning_rate": 3.61936742242977e-05,
"loss": 0.4898,
"step": 344
},
{
"epoch": 1.3974683544303796,
"grad_norm": 0.5311020994200378,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.4998,
"step": 345
},
{
"epoch": 1.4015189873417722,
"grad_norm": 0.5660741633502374,
"learning_rate": 3.612679420964516e-05,
"loss": 0.5037,
"step": 346
},
{
"epoch": 1.4055696202531647,
"grad_norm": 0.40556823537769726,
"learning_rate": 3.609315924292092e-05,
"loss": 0.49,
"step": 347
},
{
"epoch": 1.409620253164557,
"grad_norm": 0.44105952317868735,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.5131,
"step": 348
},
{
"epoch": 1.4136708860759493,
"grad_norm": 0.5393012908464014,
"learning_rate": 3.602550074513133e-05,
"loss": 0.5023,
"step": 349
},
{
"epoch": 1.4177215189873418,
"grad_norm": 0.592204863780736,
"learning_rate": 3.599147775897822e-05,
"loss": 0.5214,
"step": 350
},
{
"epoch": 1.4217721518987343,
"grad_norm": 0.44143007537210976,
"learning_rate": 3.595732597966167e-05,
"loss": 0.5226,
"step": 351
},
{
"epoch": 1.4258227848101266,
"grad_norm": 0.4979312760374977,
"learning_rate": 3.592304568223542e-05,
"loss": 0.4968,
"step": 352
},
{
"epoch": 1.4298734177215189,
"grad_norm": 0.49140227630417566,
"learning_rate": 3.588863714278826e-05,
"loss": 0.5018,
"step": 353
},
{
"epoch": 1.4339240506329114,
"grad_norm": 0.4304005156074492,
"learning_rate": 3.585410063844186e-05,
"loss": 0.5056,
"step": 354
},
{
"epoch": 1.437974683544304,
"grad_norm": 0.4244518827258193,
"learning_rate": 3.581943644734846e-05,
"loss": 0.5098,
"step": 355
},
{
"epoch": 1.4420253164556962,
"grad_norm": 0.38329407233731744,
"learning_rate": 3.578464484868869e-05,
"loss": 0.4895,
"step": 356
},
{
"epoch": 1.4460759493670885,
"grad_norm": 0.4856059174239211,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.5394,
"step": 357
},
{
"epoch": 1.450126582278481,
"grad_norm": 0.48126858863591215,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.4895,
"step": 358
},
{
"epoch": 1.4541772151898735,
"grad_norm": 0.4143957810378929,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.5178,
"step": 359
},
{
"epoch": 1.4582278481012658,
"grad_norm": 0.4134635722529027,
"learning_rate": 3.564420999786543e-05,
"loss": 0.4883,
"step": 360
},
{
"epoch": 1.4622784810126581,
"grad_norm": 0.42979136559808745,
"learning_rate": 3.560878558491842e-05,
"loss": 0.4921,
"step": 361
},
{
"epoch": 1.4663291139240506,
"grad_norm": 0.4356007235273636,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.4782,
"step": 362
},
{
"epoch": 1.4703797468354431,
"grad_norm": 0.5945576821227525,
"learning_rate": 3.553755991230039e-05,
"loss": 0.5069,
"step": 363
},
{
"epoch": 1.4744303797468354,
"grad_norm": 0.4753291764258734,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.514,
"step": 364
},
{
"epoch": 1.4784810126582277,
"grad_norm": 0.6633757883059148,
"learning_rate": 3.546583369120419e-05,
"loss": 0.4864,
"step": 365
},
{
"epoch": 1.4825316455696202,
"grad_norm": 0.5178485147879625,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.5252,
"step": 366
},
{
"epoch": 1.4865822784810128,
"grad_norm": 0.4926409929233386,
"learning_rate": 3.539360923231766e-05,
"loss": 0.4736,
"step": 367
},
{
"epoch": 1.490632911392405,
"grad_norm": 0.41847691477684545,
"learning_rate": 3.535731089018394e-05,
"loss": 0.5194,
"step": 368
},
{
"epoch": 1.4946835443037974,
"grad_norm": 0.5593201411911041,
"learning_rate": 3.532088886237956e-05,
"loss": 0.5126,
"step": 369
},
{
"epoch": 1.4987341772151899,
"grad_norm": 0.4223818165491817,
"learning_rate": 3.528434344224253e-05,
"loss": 0.4885,
"step": 370
},
{
"epoch": 1.5027848101265824,
"grad_norm": 0.5204458180525342,
"learning_rate": 3.524767492410464e-05,
"loss": 0.5058,
"step": 371
},
{
"epoch": 1.5068354430379747,
"grad_norm": 0.45088267765374307,
"learning_rate": 3.521088360328908e-05,
"loss": 0.5033,
"step": 372
},
{
"epoch": 1.510886075949367,
"grad_norm": 0.4468049128568422,
"learning_rate": 3.517396977610811e-05,
"loss": 0.5033,
"step": 373
},
{
"epoch": 1.5149367088607595,
"grad_norm": 0.4734123677543831,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.4998,
"step": 374
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.5654084513571828,
"learning_rate": 3.509977579282971e-05,
"loss": 0.5408,
"step": 375
},
{
"epoch": 1.5230379746835443,
"grad_norm": 0.4568460479309204,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.4845,
"step": 376
},
{
"epoch": 1.5270886075949366,
"grad_norm": 0.4451424696824511,
"learning_rate": 3.502509536445719e-05,
"loss": 0.491,
"step": 377
},
{
"epoch": 1.531139240506329,
"grad_norm": 0.4485087455786973,
"learning_rate": 3.498757348458147e-05,
"loss": 0.5265,
"step": 378
},
{
"epoch": 1.5351898734177216,
"grad_norm": 0.4806987909381463,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.4821,
"step": 379
},
{
"epoch": 1.539240506329114,
"grad_norm": 0.47238814952985025,
"learning_rate": 3.491216790442899e-05,
"loss": 0.5053,
"step": 380
},
{
"epoch": 1.5432911392405062,
"grad_norm": 0.46409050899525434,
"learning_rate": 3.487428481145839e-05,
"loss": 0.4913,
"step": 381
},
{
"epoch": 1.5473417721518987,
"grad_norm": 0.45512285698970795,
"learning_rate": 3.483628192304278e-05,
"loss": 0.5007,
"step": 382
},
{
"epoch": 1.5513924050632912,
"grad_norm": 0.4917037449610245,
"learning_rate": 3.479815954525219e-05,
"loss": 0.5331,
"step": 383
},
{
"epoch": 1.5554430379746835,
"grad_norm": 0.4220142249529263,
"learning_rate": 3.475991798511899e-05,
"loss": 0.4676,
"step": 384
},
{
"epoch": 1.5594936708860758,
"grad_norm": 0.5339308897468177,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.544,
"step": 385
},
{
"epoch": 1.5635443037974683,
"grad_norm": 0.4542327084379918,
"learning_rate": 3.468307855075128e-05,
"loss": 0.4758,
"step": 386
},
{
"epoch": 1.5675949367088609,
"grad_norm": 0.497554816991106,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.474,
"step": 387
},
{
"epoch": 1.5716455696202531,
"grad_norm": 0.6496864156249647,
"learning_rate": 3.460576609535163e-05,
"loss": 0.5175,
"step": 388
},
{
"epoch": 1.5756962025316454,
"grad_norm": 0.4369495677372348,
"learning_rate": 3.456693326250006e-05,
"loss": 0.513,
"step": 389
},
{
"epoch": 1.579746835443038,
"grad_norm": 0.5110310226151287,
"learning_rate": 3.452798310957058e-05,
"loss": 0.4905,
"step": 390
},
{
"epoch": 1.5837974683544305,
"grad_norm": 0.450134673612493,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.5011,
"step": 391
},
{
"epoch": 1.5878481012658228,
"grad_norm": 0.44142132678917817,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.5172,
"step": 392
},
{
"epoch": 1.591898734177215,
"grad_norm": 0.4329283910929912,
"learning_rate": 3.441043187201574e-05,
"loss": 0.5031,
"step": 393
},
{
"epoch": 1.5959493670886076,
"grad_norm": 0.43462992853355464,
"learning_rate": 3.437101558517728e-05,
"loss": 0.4986,
"step": 394
},
{
"epoch": 1.6,
"grad_norm": 0.5653553419866736,
"learning_rate": 3.433148355615496e-05,
"loss": 0.4882,
"step": 395
},
{
"epoch": 1.6040506329113924,
"grad_norm": 0.49212146997196793,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.5253,
"step": 396
},
{
"epoch": 1.6081012658227847,
"grad_norm": 0.5447686209697618,
"learning_rate": 3.425207354603043e-05,
"loss": 0.4897,
"step": 397
},
{
"epoch": 1.6121518987341772,
"grad_norm": 0.42868054894994073,
"learning_rate": 3.421219620448553e-05,
"loss": 0.4931,
"step": 398
},
{
"epoch": 1.6162025316455697,
"grad_norm": 0.47308269958487,
"learning_rate": 3.417220439986623e-05,
"loss": 0.5074,
"step": 399
},
{
"epoch": 1.620253164556962,
"grad_norm": 0.45235875685735044,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.5071,
"step": 400
},
{
"epoch": 1.6243037974683543,
"grad_norm": 0.4643242658551944,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.5147,
"step": 401
},
{
"epoch": 1.6283544303797468,
"grad_norm": 0.41587265969198606,
"learning_rate": 3.405154543304065e-05,
"loss": 0.4794,
"step": 402
},
{
"epoch": 1.6324050632911393,
"grad_norm": 0.5455216715168784,
"learning_rate": 3.401109900618843e-05,
"loss": 0.5622,
"step": 403
},
{
"epoch": 1.6364556962025316,
"grad_norm": 0.4253153765117907,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.4915,
"step": 404
},
{
"epoch": 1.640506329113924,
"grad_norm": 0.5128191169297663,
"learning_rate": 3.392986794874714e-05,
"loss": 0.4869,
"step": 405
},
{
"epoch": 1.6445569620253164,
"grad_norm": 0.4374625911467672,
"learning_rate": 3.388908397238184e-05,
"loss": 0.4977,
"step": 406
},
{
"epoch": 1.648607594936709,
"grad_norm": 0.5255904779117861,
"learning_rate": 3.384818813524362e-05,
"loss": 0.4941,
"step": 407
},
{
"epoch": 1.6526582278481012,
"grad_norm": 0.3863828312401065,
"learning_rate": 3.380718076670195e-05,
"loss": 0.4915,
"step": 408
},
{
"epoch": 1.6567088607594935,
"grad_norm": 0.42761360593438513,
"learning_rate": 3.376606219702454e-05,
"loss": 0.5007,
"step": 409
},
{
"epoch": 1.660759493670886,
"grad_norm": 0.4110434013071151,
"learning_rate": 3.372483275737468e-05,
"loss": 0.4668,
"step": 410
},
{
"epoch": 1.6648101265822786,
"grad_norm": 0.42449426360642295,
"learning_rate": 3.368349277980861e-05,
"loss": 0.5214,
"step": 411
},
{
"epoch": 1.6688607594936709,
"grad_norm": 0.44527085497352425,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.5026,
"step": 412
},
{
"epoch": 1.6729113924050631,
"grad_norm": 0.4173026949754567,
"learning_rate": 3.360048254360144e-05,
"loss": 0.5276,
"step": 413
},
{
"epoch": 1.6769620253164557,
"grad_norm": 0.39656579482904925,
"learning_rate": 3.355881295351336e-05,
"loss": 0.4821,
"step": 414
},
{
"epoch": 1.6810126582278482,
"grad_norm": 0.3698486526438957,
"learning_rate": 3.351703416260975e-05,
"loss": 0.4751,
"step": 415
},
{
"epoch": 1.6850632911392405,
"grad_norm": 0.4541776129443969,
"learning_rate": 3.347514650737126e-05,
"loss": 0.5239,
"step": 416
},
{
"epoch": 1.6891139240506328,
"grad_norm": 0.41531812389405637,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.4813,
"step": 417
},
{
"epoch": 1.6931645569620253,
"grad_norm": 0.41680329008809497,
"learning_rate": 3.339104595419334e-05,
"loss": 0.4921,
"step": 418
},
{
"epoch": 1.6972151898734178,
"grad_norm": 0.4342459332967542,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.5144,
"step": 419
},
{
"epoch": 1.70126582278481,
"grad_norm": 0.4296487092306945,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.5038,
"step": 420
},
{
"epoch": 1.7053164556962024,
"grad_norm": 0.4704515680755863,
"learning_rate": 3.326408710419996e-05,
"loss": 0.5185,
"step": 421
},
{
"epoch": 1.709367088607595,
"grad_norm": 0.4709628115096382,
"learning_rate": 3.322155337795454e-05,
"loss": 0.4632,
"step": 422
},
{
"epoch": 1.7134177215189874,
"grad_norm": 0.4364734816458604,
"learning_rate": 3.317891316713587e-05,
"loss": 0.5096,
"step": 423
},
{
"epoch": 1.7174683544303797,
"grad_norm": 0.5815248102236298,
"learning_rate": 3.313616681516231e-05,
"loss": 0.4975,
"step": 424
},
{
"epoch": 1.721518987341772,
"grad_norm": 0.4284767320397114,
"learning_rate": 3.309331466630713e-05,
"loss": 0.5048,
"step": 425
},
{
"epoch": 1.7255696202531645,
"grad_norm": 0.5004901943813579,
"learning_rate": 3.305035706569563e-05,
"loss": 0.5171,
"step": 426
},
{
"epoch": 1.729620253164557,
"grad_norm": 0.36310704633857405,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.4624,
"step": 427
},
{
"epoch": 1.7336708860759493,
"grad_norm": 0.4280219608882381,
"learning_rate": 3.296412689394864e-05,
"loss": 0.5281,
"step": 428
},
{
"epoch": 1.7377215189873416,
"grad_norm": 0.49487081440373953,
"learning_rate": 3.292085501729909e-05,
"loss": 0.4982,
"step": 429
},
{
"epoch": 1.7417721518987341,
"grad_norm": 0.42693358784262275,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.4746,
"step": 430
},
{
"epoch": 1.7458227848101266,
"grad_norm": 0.5299974278319718,
"learning_rate": 3.283399942497381e-05,
"loss": 0.4856,
"step": 431
},
{
"epoch": 1.749873417721519,
"grad_norm": 0.4880643998408143,
"learning_rate": 3.279041640882108e-05,
"loss": 0.5302,
"step": 432
},
{
"epoch": 1.7539240506329112,
"grad_norm": 0.49812165777117373,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.4915,
"step": 433
},
{
"epoch": 1.7579746835443038,
"grad_norm": 0.5454136822628279,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.5009,
"step": 434
},
{
"epoch": 1.7620253164556963,
"grad_norm": 0.4149064579181973,
"learning_rate": 3.265905069502244e-05,
"loss": 0.4714,
"step": 435
},
{
"epoch": 1.7660759493670886,
"grad_norm": 0.5554146232529701,
"learning_rate": 3.261505774420016e-05,
"loss": 0.5229,
"step": 436
},
{
"epoch": 1.7701265822784809,
"grad_norm": 0.6295198093317546,
"learning_rate": 3.257096319343707e-05,
"loss": 0.5103,
"step": 437
},
{
"epoch": 1.7741772151898734,
"grad_norm": 0.4883653461144971,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.5301,
"step": 438
},
{
"epoch": 1.7782278481012659,
"grad_norm": 0.6900941901893376,
"learning_rate": 3.248247071342966e-05,
"loss": 0.4887,
"step": 439
},
{
"epoch": 1.7822784810126582,
"grad_norm": 0.3877407884565803,
"learning_rate": 3.243807349689161e-05,
"loss": 0.4883,
"step": 440
},
{
"epoch": 1.7863291139240505,
"grad_norm": 0.7057201214336966,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.5172,
"step": 441
},
{
"epoch": 1.790379746835443,
"grad_norm": 0.3975238095817153,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.4844,
"step": 442
},
{
"epoch": 1.7944303797468355,
"grad_norm": 0.5956449523580888,
"learning_rate": 3.230428223438075e-05,
"loss": 0.5053,
"step": 443
},
{
"epoch": 1.7984810126582278,
"grad_norm": 0.4305112960899675,
"learning_rate": 3.225948647317459e-05,
"loss": 0.4608,
"step": 444
},
{
"epoch": 1.80253164556962,
"grad_norm": 0.526203355378657,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.5055,
"step": 445
},
{
"epoch": 1.8065822784810126,
"grad_norm": 0.43428906648803267,
"learning_rate": 3.216959910368034e-05,
"loss": 0.4895,
"step": 446
},
{
"epoch": 1.8106329113924051,
"grad_norm": 0.5316541598026316,
"learning_rate": 3.212450821933277e-05,
"loss": 0.5324,
"step": 447
},
{
"epoch": 1.8146835443037974,
"grad_norm": 0.43250593020434236,
"learning_rate": 3.207931968586281e-05,
"loss": 0.4917,
"step": 448
},
{
"epoch": 1.81873417721519,
"grad_norm": 0.4617393079817778,
"learning_rate": 3.203403386721272e-05,
"loss": 0.4872,
"step": 449
},
{
"epoch": 1.8227848101265822,
"grad_norm": 0.46236821887999274,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.5179,
"step": 450
},
{
"epoch": 1.8268354430379747,
"grad_norm": 0.4956499848558462,
"learning_rate": 3.194317183405573e-05,
"loss": 0.4827,
"step": 451
},
{
"epoch": 1.8308860759493673,
"grad_norm": 0.4306737937442977,
"learning_rate": 3.189759635133914e-05,
"loss": 0.4878,
"step": 452
},
{
"epoch": 1.8349367088607595,
"grad_norm": 0.5238714701263991,
"learning_rate": 3.185192504701718e-05,
"loss": 0.5169,
"step": 453
},
{
"epoch": 1.8389873417721518,
"grad_norm": 0.4299539082043518,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.4836,
"step": 454
},
{
"epoch": 1.8430379746835444,
"grad_norm": 0.4573867757546762,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.4736,
"step": 455
},
{
"epoch": 1.8470886075949369,
"grad_norm": 0.457977062444919,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.4916,
"step": 456
},
{
"epoch": 1.8511392405063292,
"grad_norm": 0.43243019126099863,
"learning_rate": 3.166828898179731e-05,
"loss": 0.5203,
"step": 457
},
{
"epoch": 1.8551898734177215,
"grad_norm": 0.4417273516748031,
"learning_rate": 3.162214410223536e-05,
"loss": 0.4866,
"step": 458
},
{
"epoch": 1.859240506329114,
"grad_norm": 0.42992706726995317,
"learning_rate": 3.157590561952257e-05,
"loss": 0.5071,
"step": 459
},
{
"epoch": 1.8632911392405065,
"grad_norm": 0.5088870957593753,
"learning_rate": 3.152957390605732e-05,
"loss": 0.5018,
"step": 460
},
{
"epoch": 1.8673417721518988,
"grad_norm": 0.3766836930609599,
"learning_rate": 3.148314933498886e-05,
"loss": 0.5154,
"step": 461
},
{
"epoch": 1.871392405063291,
"grad_norm": 0.4334823826569883,
"learning_rate": 3.143663228021431e-05,
"loss": 0.481,
"step": 462
},
{
"epoch": 1.8754430379746836,
"grad_norm": 0.37549671597507805,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.5082,
"step": 463
},
{
"epoch": 1.879493670886076,
"grad_norm": 0.3877802456451501,
"learning_rate": 3.134332221885661e-05,
"loss": 0.4488,
"step": 464
},
{
"epoch": 1.8835443037974684,
"grad_norm": 0.44838707238004666,
"learning_rate": 3.129652996377987e-05,
"loss": 0.5544,
"step": 465
},
{
"epoch": 1.8875949367088607,
"grad_norm": 0.3777437616095098,
"learning_rate": 3.12496467280038e-05,
"loss": 0.4255,
"step": 466
},
{
"epoch": 1.8916455696202532,
"grad_norm": 0.43064552295301106,
"learning_rate": 3.120267288911952e-05,
"loss": 0.5028,
"step": 467
},
{
"epoch": 1.8956962025316457,
"grad_norm": 0.3645638264753111,
"learning_rate": 3.11556088254479e-05,
"loss": 0.51,
"step": 468
},
{
"epoch": 1.899746835443038,
"grad_norm": 0.39781650133198887,
"learning_rate": 3.11084549160364e-05,
"loss": 0.5243,
"step": 469
},
{
"epoch": 1.9037974683544303,
"grad_norm": 0.41424123442286026,
"learning_rate": 3.106121154065615e-05,
"loss": 0.4702,
"step": 470
},
{
"epoch": 1.9078481012658228,
"grad_norm": 0.41462962050915836,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.5061,
"step": 471
},
{
"epoch": 1.9118987341772153,
"grad_norm": 0.34995812582994806,
"learning_rate": 3.096645791467348e-05,
"loss": 0.469,
"step": 472
},
{
"epoch": 1.9159493670886076,
"grad_norm": 0.5685020572151849,
"learning_rate": 3.091894842720373e-05,
"loss": 0.5075,
"step": 473
},
{
"epoch": 1.92,
"grad_norm": 0.40257660630248393,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.4856,
"step": 474
},
{
"epoch": 1.9240506329113924,
"grad_norm": 0.5275254142034348,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.4751,
"step": 475
},
{
"epoch": 1.928101265822785,
"grad_norm": 0.4464980047172513,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.5318,
"step": 476
},
{
"epoch": 1.9321518987341773,
"grad_norm": 0.4926371556184767,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.4908,
"step": 477
},
{
"epoch": 1.9362025316455695,
"grad_norm": 0.4687051798351903,
"learning_rate": 3.06800895716334e-05,
"loss": 0.4921,
"step": 478
},
{
"epoch": 1.940253164556962,
"grad_norm": 0.39854625948034167,
"learning_rate": 3.063205821011029e-05,
"loss": 0.4874,
"step": 479
},
{
"epoch": 1.9443037974683546,
"grad_norm": 0.4315223204682622,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.5156,
"step": 480
},
{
"epoch": 1.9483544303797469,
"grad_norm": 0.4613372462389797,
"learning_rate": 3.053573898718618e-05,
"loss": 0.4829,
"step": 481
},
{
"epoch": 1.9524050632911392,
"grad_norm": 0.36254431128945397,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.4901,
"step": 482
},
{
"epoch": 1.9564556962025317,
"grad_norm": 0.39752231849443725,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.4829,
"step": 483
},
{
"epoch": 1.9605063291139242,
"grad_norm": 0.36362262435948706,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.5509,
"step": 484
},
{
"epoch": 1.9645569620253165,
"grad_norm": 0.3618089707724395,
"learning_rate": 3.034208541656255e-05,
"loss": 0.4661,
"step": 485
},
{
"epoch": 1.9686075949367088,
"grad_norm": 0.4007956930019114,
"learning_rate": 3.029346281309218e-05,
"loss": 0.5024,
"step": 486
},
{
"epoch": 1.9726582278481013,
"grad_norm": 0.4186326628891383,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.5176,
"step": 487
},
{
"epoch": 1.9767088607594938,
"grad_norm": 0.44694807881473575,
"learning_rate": 3.019596929201569e-05,
"loss": 0.4787,
"step": 488
},
{
"epoch": 1.980759493670886,
"grad_norm": 0.4428435501972263,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.4948,
"step": 489
},
{
"epoch": 1.9848101265822784,
"grad_norm": 0.3735522817575865,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.476,
"step": 490
},
{
"epoch": 1.988860759493671,
"grad_norm": 0.40768103988634,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.5093,
"step": 491
},
{
"epoch": 1.9929113924050634,
"grad_norm": 0.490613396086442,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.5016,
"step": 492
},
{
"epoch": 1.9969620253164557,
"grad_norm": 0.3725019830676128,
"learning_rate": 2.995080534235264e-05,
"loss": 0.5297,
"step": 493
},
{
"epoch": 2.001012658227848,
"grad_norm": 0.6304844297220933,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.5435,
"step": 494
},
{
"epoch": 2.0050632911392405,
"grad_norm": 0.4934948036875176,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.4076,
"step": 495
},
{
"epoch": 2.009113924050633,
"grad_norm": 0.9154410989999222,
"learning_rate": 2.980274210295326e-05,
"loss": 0.4037,
"step": 496
},
{
"epoch": 2.013164556962025,
"grad_norm": 0.6508525991953026,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.4038,
"step": 497
},
{
"epoch": 2.0172151898734176,
"grad_norm": 0.4955238061232409,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.4112,
"step": 498
},
{
"epoch": 2.02126582278481,
"grad_norm": 0.4229864133123068,
"learning_rate": 2.965396831793362e-05,
"loss": 0.4168,
"step": 499
},
{
"epoch": 2.0253164556962027,
"grad_norm": 0.5574391056041166,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.404,
"step": 500
},
{
"epoch": 2.0293670886075947,
"grad_norm": 0.42360316900593,
"learning_rate": 2.955439637024526e-05,
"loss": 0.3915,
"step": 501
},
{
"epoch": 2.0334177215189873,
"grad_norm": 0.5967079787862879,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.4207,
"step": 502
},
{
"epoch": 2.0374683544303798,
"grad_norm": 0.6007550143597615,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.379,
"step": 503
},
{
"epoch": 2.0415189873417723,
"grad_norm": 0.47784044716439156,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.3859,
"step": 504
},
{
"epoch": 2.0455696202531644,
"grad_norm": 0.5016674329260242,
"learning_rate": 2.93543322968538e-05,
"loss": 0.386,
"step": 505
},
{
"epoch": 2.049620253164557,
"grad_norm": 0.48515783787511924,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.4037,
"step": 506
},
{
"epoch": 2.0536708860759494,
"grad_norm": 0.49450169798833765,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.4254,
"step": 507
},
{
"epoch": 2.057721518987342,
"grad_norm": 0.4038634927399572,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.386,
"step": 508
},
{
"epoch": 2.061772151898734,
"grad_norm": 0.5077147252253258,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.418,
"step": 509
},
{
"epoch": 2.0658227848101265,
"grad_norm": 0.44676906832141833,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.4118,
"step": 510
},
{
"epoch": 2.069873417721519,
"grad_norm": 0.4520580395782798,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.4265,
"step": 511
},
{
"epoch": 2.0739240506329115,
"grad_norm": 0.46166561679891166,
"learning_rate": 2.900133526131588e-05,
"loss": 0.4043,
"step": 512
},
{
"epoch": 2.0779746835443036,
"grad_norm": 0.4256699627938866,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.4062,
"step": 513
},
{
"epoch": 2.082025316455696,
"grad_norm": 0.5315254410646818,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.4233,
"step": 514
},
{
"epoch": 2.0860759493670886,
"grad_norm": 0.3645776822375198,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.3655,
"step": 515
},
{
"epoch": 2.090126582278481,
"grad_norm": 0.47972404141366165,
"learning_rate": 2.879801882631847e-05,
"loss": 0.3997,
"step": 516
},
{
"epoch": 2.094177215189873,
"grad_norm": 0.43873777315819124,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.427,
"step": 517
},
{
"epoch": 2.0982278481012657,
"grad_norm": 0.4479088629812752,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.4083,
"step": 518
},
{
"epoch": 2.1022784810126582,
"grad_norm": 0.42049198410253585,
"learning_rate": 2.864478605989494e-05,
"loss": 0.4062,
"step": 519
},
{
"epoch": 2.1063291139240508,
"grad_norm": 0.38941722234091797,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.407,
"step": 520
},
{
"epoch": 2.110379746835443,
"grad_norm": 0.40960209218666577,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.4387,
"step": 521
},
{
"epoch": 2.1144303797468353,
"grad_norm": 0.3591304830357265,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.3552,
"step": 522
},
{
"epoch": 2.118481012658228,
"grad_norm": 0.4040647090846734,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.417,
"step": 523
},
{
"epoch": 2.1225316455696204,
"grad_norm": 0.35127425356148323,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.3656,
"step": 524
},
{
"epoch": 2.1265822784810124,
"grad_norm": 0.43251639851019685,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.435,
"step": 525
},
{
"epoch": 2.130632911392405,
"grad_norm": 0.42123287330611925,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.4061,
"step": 526
},
{
"epoch": 2.1346835443037975,
"grad_norm": 0.3939685029003428,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.3986,
"step": 527
},
{
"epoch": 2.13873417721519,
"grad_norm": 0.39111801478919844,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.3819,
"step": 528
},
{
"epoch": 2.1427848101265825,
"grad_norm": 0.40798281690224736,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.3997,
"step": 529
},
{
"epoch": 2.1468354430379746,
"grad_norm": 0.3952949428909774,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.3645,
"step": 530
},
{
"epoch": 2.150886075949367,
"grad_norm": 0.3500859649556749,
"learning_rate": 2.802570062143278e-05,
"loss": 0.4178,
"step": 531
},
{
"epoch": 2.1549367088607596,
"grad_norm": 0.4095486757067151,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.3982,
"step": 532
},
{
"epoch": 2.1589873417721517,
"grad_norm": 0.39916775188604664,
"learning_rate": 2.792159532078314e-05,
"loss": 0.4379,
"step": 533
},
{
"epoch": 2.163037974683544,
"grad_norm": 0.5216301780603962,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.4195,
"step": 534
},
{
"epoch": 2.1670886075949367,
"grad_norm": 0.40415651993351526,
"learning_rate": 2.781723482288857e-05,
"loss": 0.4112,
"step": 535
},
{
"epoch": 2.1711392405063292,
"grad_norm": 0.4681556343070256,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.4236,
"step": 536
},
{
"epoch": 2.1751898734177217,
"grad_norm": 0.4447771441391858,
"learning_rate": 2.771262248976272e-05,
"loss": 0.3752,
"step": 537
},
{
"epoch": 2.179240506329114,
"grad_norm": 0.4134025570588484,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.4043,
"step": 538
},
{
"epoch": 2.1832911392405063,
"grad_norm": 0.4086962082074082,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.3825,
"step": 539
},
{
"epoch": 2.187341772151899,
"grad_norm": 0.41175163333916365,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.44,
"step": 540
},
{
"epoch": 2.191392405063291,
"grad_norm": 0.4000447807505925,
"learning_rate": 2.7502655806328e-05,
"loss": 0.4007,
"step": 541
},
{
"epoch": 2.1954430379746834,
"grad_norm": 0.3974987146843925,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.3769,
"step": 542
},
{
"epoch": 2.199493670886076,
"grad_norm": 0.48252743967494455,
"learning_rate": 2.739730822017673e-05,
"loss": 0.4351,
"step": 543
},
{
"epoch": 2.2035443037974685,
"grad_norm": 0.3461098134131447,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.3877,
"step": 544
},
{
"epoch": 2.207594936708861,
"grad_norm": 0.4661868049305004,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.4237,
"step": 545
},
{
"epoch": 2.211645569620253,
"grad_norm": 0.35714348291165066,
"learning_rate": 2.723884107774236e-05,
"loss": 0.396,
"step": 546
},
{
"epoch": 2.2156962025316456,
"grad_norm": 0.3850410801366739,
"learning_rate": 2.718590152796239e-05,
"loss": 0.3993,
"step": 547
},
{
"epoch": 2.219746835443038,
"grad_norm": 0.40725807540651243,
"learning_rate": 2.71329041039194e-05,
"loss": 0.4052,
"step": 548
},
{
"epoch": 2.22379746835443,
"grad_norm": 0.34907683778178583,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.3879,
"step": 549
},
{
"epoch": 2.2278481012658227,
"grad_norm": 0.42375894427066013,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.397,
"step": 550
},
{
"epoch": 2.231898734177215,
"grad_norm": 0.3919765052560397,
"learning_rate": 2.697356885686189e-05,
"loss": 0.4387,
"step": 551
},
{
"epoch": 2.2359493670886077,
"grad_norm": 0.44787205378634626,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.3642,
"step": 552
},
{
"epoch": 2.24,
"grad_norm": 0.41312378139087536,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.4236,
"step": 553
},
{
"epoch": 2.2440506329113923,
"grad_norm": 0.4742868271037675,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.4263,
"step": 554
},
{
"epoch": 2.248101265822785,
"grad_norm": 0.3818740413931032,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.3995,
"step": 555
},
{
"epoch": 2.2521518987341773,
"grad_norm": 0.3675831867292822,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.3839,
"step": 556
},
{
"epoch": 2.2562025316455694,
"grad_norm": 0.4162180620514688,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.4049,
"step": 557
},
{
"epoch": 2.260253164556962,
"grad_norm": 0.40598377421269677,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.4173,
"step": 558
},
{
"epoch": 2.2643037974683544,
"grad_norm": 0.3679999516335007,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.4023,
"step": 559
},
{
"epoch": 2.268354430379747,
"grad_norm": 0.35689471981935206,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.413,
"step": 560
},
{
"epoch": 2.2724050632911394,
"grad_norm": 0.37885231385121465,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.4066,
"step": 561
},
{
"epoch": 2.2764556962025315,
"grad_norm": 0.3663266609339282,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.3918,
"step": 562
},
{
"epoch": 2.280506329113924,
"grad_norm": 0.4034397641591289,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.41,
"step": 563
},
{
"epoch": 2.2845569620253166,
"grad_norm": 0.40955626339409634,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.4269,
"step": 564
},
{
"epoch": 2.2886075949367086,
"grad_norm": 0.35407627677997544,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.4235,
"step": 565
},
{
"epoch": 2.292658227848101,
"grad_norm": 0.3549428788794233,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.3646,
"step": 566
},
{
"epoch": 2.2967088607594937,
"grad_norm": 0.4224730734672626,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.424,
"step": 567
},
{
"epoch": 2.300759493670886,
"grad_norm": 0.35075003032296437,
"learning_rate": 2.606146264569603e-05,
"loss": 0.3786,
"step": 568
},
{
"epoch": 2.3048101265822787,
"grad_norm": 0.4558193484882467,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.4546,
"step": 569
},
{
"epoch": 2.3088607594936708,
"grad_norm": 0.3988709779103802,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.4044,
"step": 570
},
{
"epoch": 2.3129113924050633,
"grad_norm": 0.4187297355369692,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.4029,
"step": 571
},
{
"epoch": 2.316962025316456,
"grad_norm": 0.43210938770580215,
"learning_rate": 2.584472010606015e-05,
"loss": 0.3992,
"step": 572
},
{
"epoch": 2.321012658227848,
"grad_norm": 0.38905747829521753,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.4089,
"step": 573
},
{
"epoch": 2.3250632911392404,
"grad_norm": 0.49549008230527225,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.4009,
"step": 574
},
{
"epoch": 2.329113924050633,
"grad_norm": 0.3378770245205281,
"learning_rate": 2.568166741322048e-05,
"loss": 0.3941,
"step": 575
},
{
"epoch": 2.3331645569620254,
"grad_norm": 0.503833733937564,
"learning_rate": 2.56272244128509e-05,
"loss": 0.4445,
"step": 576
},
{
"epoch": 2.337215189873418,
"grad_norm": 0.37620476547321036,
"learning_rate": 2.55727360915895e-05,
"loss": 0.4099,
"step": 577
},
{
"epoch": 2.34126582278481,
"grad_norm": 0.45890207853335224,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.4328,
"step": 578
},
{
"epoch": 2.3453164556962025,
"grad_norm": 0.3558605894406628,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.3727,
"step": 579
},
{
"epoch": 2.349367088607595,
"grad_norm": 0.3848076669137592,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.4194,
"step": 580
},
{
"epoch": 2.353417721518987,
"grad_norm": 0.40870068213316446,
"learning_rate": 2.535433837985559e-05,
"loss": 0.4336,
"step": 581
},
{
"epoch": 2.3574683544303796,
"grad_norm": 0.37423007811771475,
"learning_rate": 2.529963004393324e-05,
"loss": 0.4203,
"step": 582
},
{
"epoch": 2.361518987341772,
"grad_norm": 0.3454700934587644,
"learning_rate": 2.524487902551908e-05,
"loss": 0.3966,
"step": 583
},
{
"epoch": 2.3655696202531646,
"grad_norm": 0.36541769638380167,
"learning_rate": 2.519008576557029e-05,
"loss": 0.4319,
"step": 584
},
{
"epoch": 2.369620253164557,
"grad_norm": 0.40723472278733674,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.4027,
"step": 585
},
{
"epoch": 2.3736708860759492,
"grad_norm": 0.3053929085706871,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.3797,
"step": 586
},
{
"epoch": 2.3777215189873417,
"grad_norm": 0.45344419616534937,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.4383,
"step": 587
},
{
"epoch": 2.3817721518987343,
"grad_norm": 0.37257476019278707,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.4049,
"step": 588
},
{
"epoch": 2.3858227848101268,
"grad_norm": 0.4385455050563881,
"learning_rate": 2.491550129992345e-05,
"loss": 0.4171,
"step": 589
},
{
"epoch": 2.389873417721519,
"grad_norm": 0.3822053078426103,
"learning_rate": 2.486046386967024e-05,
"loss": 0.3963,
"step": 590
},
{
"epoch": 2.3939240506329114,
"grad_norm": 0.4661976003078043,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.4248,
"step": 591
},
{
"epoch": 2.397974683544304,
"grad_norm": 0.3668296003109733,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.3836,
"step": 592
},
{
"epoch": 2.4020253164556964,
"grad_norm": 0.44172060119590956,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.3985,
"step": 593
},
{
"epoch": 2.4060759493670885,
"grad_norm": 0.37702022248927824,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.4186,
"step": 594
},
{
"epoch": 2.410126582278481,
"grad_norm": 0.41219340588166303,
"learning_rate": 2.458469841202444e-05,
"loss": 0.4135,
"step": 595
},
{
"epoch": 2.4141772151898735,
"grad_norm": 0.35222272172745955,
"learning_rate": 2.452943276859226e-05,
"loss": 0.3799,
"step": 596
},
{
"epoch": 2.418227848101266,
"grad_norm": 0.3792707043241185,
"learning_rate": 2.447413064573125e-05,
"loss": 0.4062,
"step": 597
},
{
"epoch": 2.422278481012658,
"grad_norm": 0.3652588130638372,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.4138,
"step": 598
},
{
"epoch": 2.4263291139240506,
"grad_norm": 0.3822057050568963,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.4084,
"step": 599
},
{
"epoch": 2.430379746835443,
"grad_norm": 0.38157558911967315,
"learning_rate": 2.430800985597963e-05,
"loss": 0.4134,
"step": 600
},
{
"epoch": 2.4344303797468356,
"grad_norm": 0.4440396425390895,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.4006,
"step": 601
},
{
"epoch": 2.4384810126582277,
"grad_norm": 0.3871248072755634,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.4,
"step": 602
},
{
"epoch": 2.44253164556962,
"grad_norm": 0.3465413183469448,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.3944,
"step": 603
},
{
"epoch": 2.4465822784810127,
"grad_norm": 0.3449066033964096,
"learning_rate": 2.408603181104957e-05,
"loss": 0.4081,
"step": 604
},
{
"epoch": 2.4506329113924052,
"grad_norm": 0.34073892359699137,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.3926,
"step": 605
},
{
"epoch": 2.4546835443037973,
"grad_norm": 0.4130281773661855,
"learning_rate": 2.397484355008521e-05,
"loss": 0.4098,
"step": 606
},
{
"epoch": 2.45873417721519,
"grad_norm": 0.3539936165113313,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.4137,
"step": 607
},
{
"epoch": 2.4627848101265823,
"grad_norm": 0.38780597850352266,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.4152,
"step": 608
},
{
"epoch": 2.466835443037975,
"grad_norm": 0.34385778336465583,
"learning_rate": 2.380782218329337e-05,
"loss": 0.4057,
"step": 609
},
{
"epoch": 2.470886075949367,
"grad_norm": 0.3571964843522031,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.3506,
"step": 610
},
{
"epoch": 2.4749367088607594,
"grad_norm": 0.38555052964827186,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.4272,
"step": 611
},
{
"epoch": 2.478987341772152,
"grad_norm": 0.37707066803630485,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.3951,
"step": 612
},
{
"epoch": 2.4830379746835445,
"grad_norm": 0.3935846202258573,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.4418,
"step": 613
},
{
"epoch": 2.4870886075949366,
"grad_norm": 0.3794487470136585,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.3837,
"step": 614
},
{
"epoch": 2.491139240506329,
"grad_norm": 0.3690724928773584,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.4054,
"step": 615
},
{
"epoch": 2.4951898734177216,
"grad_norm": 0.4132142657161311,
"learning_rate": 2.341705325992901e-05,
"loss": 0.4363,
"step": 616
},
{
"epoch": 2.499240506329114,
"grad_norm": 0.41171972153191055,
"learning_rate": 2.336111544604222e-05,
"loss": 0.3828,
"step": 617
},
{
"epoch": 2.503291139240506,
"grad_norm": 0.3746690658043282,
"learning_rate": 2.33051505621937e-05,
"loss": 0.3713,
"step": 618
},
{
"epoch": 2.5073417721518987,
"grad_norm": 0.4203488464882228,
"learning_rate": 2.324915905911693e-05,
"loss": 0.433,
"step": 619
},
{
"epoch": 2.511392405063291,
"grad_norm": 0.39100172466968824,
"learning_rate": 2.319314138775977e-05,
"loss": 0.3939,
"step": 620
},
{
"epoch": 2.5154430379746833,
"grad_norm": 0.40204448686448035,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.422,
"step": 621
},
{
"epoch": 2.519493670886076,
"grad_norm": 0.3713683374552082,
"learning_rate": 2.308102934504593e-05,
"loss": 0.3882,
"step": 622
},
{
"epoch": 2.5235443037974683,
"grad_norm": 0.4003220895532632,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.4155,
"step": 623
},
{
"epoch": 2.527594936708861,
"grad_norm": 0.4170968538089543,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.446,
"step": 624
},
{
"epoch": 2.5316455696202533,
"grad_norm": 0.38284581183443106,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.3842,
"step": 625
},
{
"epoch": 2.5356962025316454,
"grad_norm": 0.38386723923902855,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.4142,
"step": 626
},
{
"epoch": 2.539746835443038,
"grad_norm": 0.3987429822537067,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.3968,
"step": 627
},
{
"epoch": 2.5437974683544304,
"grad_norm": 0.35740124927928735,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.4136,
"step": 628
},
{
"epoch": 2.547848101265823,
"grad_norm": 0.4086772624285927,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.4414,
"step": 629
},
{
"epoch": 2.5518987341772155,
"grad_norm": 0.3101503628157208,
"learning_rate": 2.26316247733093e-05,
"loss": 0.3597,
"step": 630
},
{
"epoch": 2.5559493670886075,
"grad_norm": 0.4095380463604193,
"learning_rate": 2.257534907110328e-05,
"loss": 0.4275,
"step": 631
},
{
"epoch": 2.56,
"grad_norm": 0.3552161517024676,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.3976,
"step": 632
},
{
"epoch": 2.5640506329113926,
"grad_norm": 0.3775415803467563,
"learning_rate": 2.246273589557294e-05,
"loss": 0.4108,
"step": 633
},
{
"epoch": 2.5681012658227846,
"grad_norm": 0.34857757482684293,
"learning_rate": 2.240639932921966e-05,
"loss": 0.3655,
"step": 634
},
{
"epoch": 2.572151898734177,
"grad_norm": 0.40296231899768875,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.4175,
"step": 635
},
{
"epoch": 2.5762025316455697,
"grad_norm": 0.36131881479105415,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.425,
"step": 636
},
{
"epoch": 2.580253164556962,
"grad_norm": 0.41235148104813407,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.4024,
"step": 637
},
{
"epoch": 2.5843037974683547,
"grad_norm": 0.3487238376758643,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.4149,
"step": 638
},
{
"epoch": 2.5883544303797468,
"grad_norm": 0.3878471330408811,
"learning_rate": 2.212443486523819e-05,
"loss": 0.3581,
"step": 639
},
{
"epoch": 2.5924050632911393,
"grad_norm": 0.417826115801736,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.4546,
"step": 640
},
{
"epoch": 2.596455696202532,
"grad_norm": 0.325402944333296,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.3553,
"step": 641
},
{
"epoch": 2.600506329113924,
"grad_norm": 0.3788290286433297,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.4072,
"step": 642
},
{
"epoch": 2.6045569620253164,
"grad_norm": 0.3409235665866428,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.3947,
"step": 643
},
{
"epoch": 2.608607594936709,
"grad_norm": 0.36610157795004983,
"learning_rate": 2.184204266048803e-05,
"loss": 0.4159,
"step": 644
},
{
"epoch": 2.6126582278481014,
"grad_norm": 0.3089455138605491,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.3865,
"step": 645
},
{
"epoch": 2.616708860759494,
"grad_norm": 0.3877114868667849,
"learning_rate": 2.17289787446841e-05,
"loss": 0.4129,
"step": 646
},
{
"epoch": 2.620759493670886,
"grad_norm": 0.3363315983565323,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.3942,
"step": 647
},
{
"epoch": 2.6248101265822785,
"grad_norm": 0.3783599660689252,
"learning_rate": 2.161585912916385e-05,
"loss": 0.4039,
"step": 648
},
{
"epoch": 2.628860759493671,
"grad_norm": 0.3364311178238321,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.4201,
"step": 649
},
{
"epoch": 2.632911392405063,
"grad_norm": 0.41450680921676025,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.3916,
"step": 650
},
{
"epoch": 2.6369620253164556,
"grad_norm": 0.3298742282732143,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.3813,
"step": 651
},
{
"epoch": 2.641012658227848,
"grad_norm": 0.37190347359239956,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.4022,
"step": 652
},
{
"epoch": 2.6450632911392407,
"grad_norm": 0.3636796928820619,
"learning_rate": 2.133284032321232e-05,
"loss": 0.4204,
"step": 653
},
{
"epoch": 2.649113924050633,
"grad_norm": 0.32037948337361444,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.3999,
"step": 654
},
{
"epoch": 2.6531645569620252,
"grad_norm": 0.3531333184627574,
"learning_rate": 2.121955446741306e-05,
"loss": 0.3981,
"step": 655
},
{
"epoch": 2.6572151898734178,
"grad_norm": 0.36470192070432345,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.4276,
"step": 656
},
{
"epoch": 2.6612658227848103,
"grad_norm": 0.3243594923266069,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.3735,
"step": 657
},
{
"epoch": 2.6653164556962023,
"grad_norm": 0.3687551959893952,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.3991,
"step": 658
},
{
"epoch": 2.669367088607595,
"grad_norm": 0.3153076088697761,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.3947,
"step": 659
},
{
"epoch": 2.6734177215189874,
"grad_norm": 0.3366423799295819,
"learning_rate": 2.093617592756158e-05,
"loss": 0.408,
"step": 660
},
{
"epoch": 2.67746835443038,
"grad_norm": 0.3783416051604966,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.4244,
"step": 661
},
{
"epoch": 2.6815189873417724,
"grad_norm": 0.35499998013350254,
"learning_rate": 2.082276853708978e-05,
"loss": 0.4054,
"step": 662
},
{
"epoch": 2.6855696202531645,
"grad_norm": 0.3686558451881377,
"learning_rate": 2.076605467380071e-05,
"loss": 0.4485,
"step": 663
},
{
"epoch": 2.689620253164557,
"grad_norm": 0.35170196014865857,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.3905,
"step": 664
},
{
"epoch": 2.6936708860759495,
"grad_norm": 0.3872617174747687,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.4325,
"step": 665
},
{
"epoch": 2.6977215189873416,
"grad_norm": 0.3302651972664198,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.3672,
"step": 666
},
{
"epoch": 2.701772151898734,
"grad_norm": 0.3879811653203523,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.4156,
"step": 667
},
{
"epoch": 2.7058227848101266,
"grad_norm": 0.3347308275251131,
"learning_rate": 2.048240194884496e-05,
"loss": 0.4075,
"step": 668
},
{
"epoch": 2.709873417721519,
"grad_norm": 0.40421925137166675,
"learning_rate": 2.042565792045695e-05,
"loss": 0.4059,
"step": 669
},
{
"epoch": 2.7139240506329116,
"grad_norm": 0.3093742256877643,
"learning_rate": 2.036891046387857e-05,
"loss": 0.379,
"step": 670
},
{
"epoch": 2.7179746835443037,
"grad_norm": 0.41389616985824607,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.4279,
"step": 671
},
{
"epoch": 2.7220253164556962,
"grad_norm": 0.38091468748391394,
"learning_rate": 2.025540709431948e-05,
"loss": 0.415,
"step": 672
},
{
"epoch": 2.7260759493670887,
"grad_norm": 0.3574182561586124,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.4154,
"step": 673
},
{
"epoch": 2.730126582278481,
"grad_norm": 0.32935090522685473,
"learning_rate": 2.014189549672245e-05,
"loss": 0.4218,
"step": 674
},
{
"epoch": 2.7341772151898733,
"grad_norm": 0.3375955238768204,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.3975,
"step": 675
},
{
"epoch": 2.738227848101266,
"grad_norm": 0.3169469936860583,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.3617,
"step": 676
},
{
"epoch": 2.7422784810126584,
"grad_norm": 0.33801858028806625,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.4377,
"step": 677
},
{
"epoch": 2.746329113924051,
"grad_norm": 0.3459809930160941,
"learning_rate": 1.991486224484123e-05,
"loss": 0.4372,
"step": 678
},
{
"epoch": 2.750379746835443,
"grad_norm": 0.3569447689948828,
"learning_rate": 1.985810450327756e-05,
"loss": 0.3792,
"step": 679
},
{
"epoch": 2.7544303797468355,
"grad_norm": 0.3891436395259758,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.4226,
"step": 680
},
{
"epoch": 2.758481012658228,
"grad_norm": 0.3851661465957586,
"learning_rate": 1.974459290568053e-05,
"loss": 0.397,
"step": 681
},
{
"epoch": 2.76253164556962,
"grad_norm": 0.4015879766922774,
"learning_rate": 1.968783996385397e-05,
"loss": 0.4133,
"step": 682
},
{
"epoch": 2.7665822784810126,
"grad_norm": 0.3536953202571622,
"learning_rate": 1.963108953612143e-05,
"loss": 0.4064,
"step": 683
},
{
"epoch": 2.770632911392405,
"grad_norm": 0.37841700338233,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.3838,
"step": 684
},
{
"epoch": 2.7746835443037976,
"grad_norm": 0.3612591596401444,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.4046,
"step": 685
},
{
"epoch": 2.77873417721519,
"grad_norm": 0.37063539262584855,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.418,
"step": 686
},
{
"epoch": 2.782784810126582,
"grad_norm": 0.3337665261902088,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.3786,
"step": 687
},
{
"epoch": 2.7868354430379747,
"grad_norm": 0.3576173940035445,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.3969,
"step": 688
},
{
"epoch": 2.790886075949367,
"grad_norm": 0.38614188603141647,
"learning_rate": 1.92906653591874e-05,
"loss": 0.4285,
"step": 689
},
{
"epoch": 2.7949367088607593,
"grad_norm": 0.37176121466458034,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.4204,
"step": 690
},
{
"epoch": 2.798987341772152,
"grad_norm": 0.36866189473315625,
"learning_rate": 1.917723146291022e-05,
"loss": 0.3938,
"step": 691
},
{
"epoch": 2.8030379746835443,
"grad_norm": 0.3760468574956014,
"learning_rate": 1.912052422608584e-05,
"loss": 0.4099,
"step": 692
},
{
"epoch": 2.807088607594937,
"grad_norm": 0.36242613191303635,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.3758,
"step": 693
},
{
"epoch": 2.8111392405063294,
"grad_norm": 0.42620323776156965,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.4453,
"step": 694
},
{
"epoch": 2.8151898734177214,
"grad_norm": 0.31939219047693623,
"learning_rate": 1.895044684123479e-05,
"loss": 0.3825,
"step": 695
},
{
"epoch": 2.819240506329114,
"grad_norm": 0.35929020832737774,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.3918,
"step": 696
},
{
"epoch": 2.8232911392405065,
"grad_norm": 0.40812393513129436,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.4519,
"step": 697
},
{
"epoch": 2.8273417721518985,
"grad_norm": 0.33850956714818214,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.3923,
"step": 698
},
{
"epoch": 2.831392405063291,
"grad_norm": 0.38986587073262824,
"learning_rate": 1.872379746550743e-05,
"loss": 0.3937,
"step": 699
},
{
"epoch": 2.8354430379746836,
"grad_norm": 0.3573192240570744,
"learning_rate": 1.866715967678769e-05,
"loss": 0.4235,
"step": 700
},
{
"epoch": 2.839493670886076,
"grad_norm": 0.37199611211185046,
"learning_rate": 1.861053262258067e-05,
"loss": 0.3895,
"step": 701
},
{
"epoch": 2.8435443037974686,
"grad_norm": 0.4097142133560977,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.413,
"step": 702
},
{
"epoch": 2.8475949367088607,
"grad_norm": 0.3617913503862617,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.3806,
"step": 703
},
{
"epoch": 2.851645569620253,
"grad_norm": 0.39884725901887663,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.4546,
"step": 704
},
{
"epoch": 2.8556962025316457,
"grad_norm": 0.39426584287182753,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.4109,
"step": 705
},
{
"epoch": 2.8597468354430378,
"grad_norm": 0.3404811819877716,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.3808,
"step": 706
},
{
"epoch": 2.8637974683544303,
"grad_norm": 0.38497452809807065,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.4236,
"step": 707
},
{
"epoch": 2.867848101265823,
"grad_norm": 0.34082056299081137,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.3781,
"step": 708
},
{
"epoch": 2.8718987341772153,
"grad_norm": 0.37137796121959676,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.4115,
"step": 709
},
{
"epoch": 2.875949367088608,
"grad_norm": 0.36297916130466273,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.4102,
"step": 710
},
{
"epoch": 2.88,
"grad_norm": 0.3872854329026497,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.4245,
"step": 711
},
{
"epoch": 2.8840506329113924,
"grad_norm": 0.307878757804808,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.3736,
"step": 712
},
{
"epoch": 2.888101265822785,
"grad_norm": 0.4095040084766943,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.4271,
"step": 713
},
{
"epoch": 2.892151898734177,
"grad_norm": 0.37364347417326055,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.4353,
"step": 714
},
{
"epoch": 2.8962025316455695,
"grad_norm": 0.34418839923409006,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.4171,
"step": 715
},
{
"epoch": 2.900253164556962,
"grad_norm": 0.3475288886361682,
"learning_rate": 1.776272483901182e-05,
"loss": 0.3754,
"step": 716
},
{
"epoch": 2.9043037974683545,
"grad_norm": 0.3503160309610593,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.4327,
"step": 717
},
{
"epoch": 2.908354430379747,
"grad_norm": 0.34715281887527716,
"learning_rate": 1.764995661794301e-05,
"loss": 0.3634,
"step": 718
},
{
"epoch": 2.912405063291139,
"grad_norm": 0.39703557483930085,
"learning_rate": 1.759360067078035e-05,
"loss": 0.4151,
"step": 719
},
{
"epoch": 2.9164556962025316,
"grad_norm": 0.3237704086539288,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.3863,
"step": 720
},
{
"epoch": 2.920506329113924,
"grad_norm": 0.3664591031389774,
"learning_rate": 1.748094737261012e-05,
"loss": 0.4392,
"step": 721
},
{
"epoch": 2.9245569620253162,
"grad_norm": 0.34102253802493965,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.3802,
"step": 722
},
{
"epoch": 2.9286075949367087,
"grad_norm": 0.3537733862912376,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.4199,
"step": 723
},
{
"epoch": 2.9326582278481013,
"grad_norm": 0.3653629519003497,
"learning_rate": 1.731212071922883e-05,
"loss": 0.4289,
"step": 724
},
{
"epoch": 2.9367088607594938,
"grad_norm": 0.3142065584351328,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.3929,
"step": 725
},
{
"epoch": 2.9407594936708863,
"grad_norm": 0.3727966665639162,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.391,
"step": 726
},
{
"epoch": 2.9448101265822784,
"grad_norm": 0.3332381396137547,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.4055,
"step": 727
},
{
"epoch": 2.948860759493671,
"grad_norm": 0.32465826843404083,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.4006,
"step": 728
},
{
"epoch": 2.9529113924050634,
"grad_norm": 0.32953641221114177,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.3913,
"step": 729
},
{
"epoch": 2.9569620253164555,
"grad_norm": 0.31513928231768,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.3731,
"step": 730
},
{
"epoch": 2.961012658227848,
"grad_norm": 0.35443017453692444,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.4336,
"step": 731
},
{
"epoch": 2.9650632911392405,
"grad_norm": 0.30803709737248847,
"learning_rate": 1.686290200071915e-05,
"loss": 0.3849,
"step": 732
},
{
"epoch": 2.969113924050633,
"grad_norm": 0.3347294542251129,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.3881,
"step": 733
},
{
"epoch": 2.9731645569620255,
"grad_norm": 0.30311856127123005,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.3787,
"step": 734
},
{
"epoch": 2.9772151898734176,
"grad_norm": 0.37378814989541576,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.469,
"step": 735
},
{
"epoch": 2.98126582278481,
"grad_norm": 0.3342133780666599,
"learning_rate": 1.663888455395778e-05,
"loss": 0.3923,
"step": 736
},
{
"epoch": 2.9853164556962026,
"grad_norm": 0.35005745167039787,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.3905,
"step": 737
},
{
"epoch": 2.9893670886075947,
"grad_norm": 0.37607745284416977,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.42,
"step": 738
},
{
"epoch": 2.993417721518987,
"grad_norm": 0.3315687298236548,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.4072,
"step": 739
},
{
"epoch": 2.9974683544303797,
"grad_norm": 0.3292731936373325,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.4381,
"step": 740
},
{
"epoch": 3.0015189873417722,
"grad_norm": 0.35545719290302735,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.3882,
"step": 741
},
{
"epoch": 3.0055696202531648,
"grad_norm": 0.3736341698712185,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.2865,
"step": 742
},
{
"epoch": 3.009620253164557,
"grad_norm": 0.7292622281149919,
"learning_rate": 1.624791353909428e-05,
"loss": 0.3286,
"step": 743
},
{
"epoch": 3.0136708860759494,
"grad_norm": 0.7288463540031005,
"learning_rate": 1.619217781670663e-05,
"loss": 0.3336,
"step": 744
},
{
"epoch": 3.017721518987342,
"grad_norm": 0.4657568396026094,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.3066,
"step": 745
},
{
"epoch": 3.0217721518987344,
"grad_norm": 0.4778654491973135,
"learning_rate": 1.608079882360134e-05,
"loss": 0.2855,
"step": 746
},
{
"epoch": 3.0258227848101265,
"grad_norm": 0.5402135035843545,
"learning_rate": 1.60251564499148e-05,
"loss": 0.33,
"step": 747
},
{
"epoch": 3.029873417721519,
"grad_norm": 0.38715427317885476,
"learning_rate": 1.596954608907176e-05,
"loss": 0.3175,
"step": 748
},
{
"epoch": 3.0339240506329115,
"grad_norm": 0.3990751321789323,
"learning_rate": 1.591396818895043e-05,
"loss": 0.2778,
"step": 749
},
{
"epoch": 3.037974683544304,
"grad_norm": 0.44376036039624767,
"learning_rate": 1.585842319716759e-05,
"loss": 0.3104,
"step": 750
},
{
"epoch": 3.042025316455696,
"grad_norm": 0.369857361086485,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.3195,
"step": 751
},
{
"epoch": 3.0460759493670886,
"grad_norm": 0.38605444201336564,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.2817,
"step": 752
},
{
"epoch": 3.050126582278481,
"grad_norm": 0.41193949379348305,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.2974,
"step": 753
},
{
"epoch": 3.0541772151898736,
"grad_norm": 0.39434925723612574,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.2941,
"step": 754
},
{
"epoch": 3.0582278481012657,
"grad_norm": 0.43319703074931143,
"learning_rate": 1.558120751116291e-05,
"loss": 0.3429,
"step": 755
},
{
"epoch": 3.062278481012658,
"grad_norm": 0.41052422978090675,
"learning_rate": 1.552586935426876e-05,
"loss": 0.3047,
"step": 756
},
{
"epoch": 3.0663291139240507,
"grad_norm": 0.3690100647020218,
"learning_rate": 1.547056723140774e-05,
"loss": 0.3046,
"step": 757
},
{
"epoch": 3.0703797468354432,
"grad_norm": 0.3797485749360148,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.304,
"step": 758
},
{
"epoch": 3.0744303797468353,
"grad_norm": 0.39779561828363075,
"learning_rate": 1.536007286907411e-05,
"loss": 0.2928,
"step": 759
},
{
"epoch": 3.078481012658228,
"grad_norm": 0.3396021439259308,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.2671,
"step": 760
},
{
"epoch": 3.0825316455696203,
"grad_norm": 0.4418866701716418,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.3355,
"step": 761
},
{
"epoch": 3.086582278481013,
"grad_norm": 0.34397982875730543,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.2939,
"step": 762
},
{
"epoch": 3.090632911392405,
"grad_norm": 0.3766891957459184,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.2826,
"step": 763
},
{
"epoch": 3.0946835443037974,
"grad_norm": 0.37404051371747354,
"learning_rate": 1.508449870007656e-05,
"loss": 0.3025,
"step": 764
},
{
"epoch": 3.09873417721519,
"grad_norm": 0.3687946165415446,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.3051,
"step": 765
},
{
"epoch": 3.1027848101265825,
"grad_norm": 0.3624794104299201,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.2952,
"step": 766
},
{
"epoch": 3.1068354430379745,
"grad_norm": 0.3910690965314667,
"learning_rate": 1.4919625713405e-05,
"loss": 0.2967,
"step": 767
},
{
"epoch": 3.110886075949367,
"grad_norm": 0.384824157701629,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.3089,
"step": 768
},
{
"epoch": 3.1149367088607596,
"grad_norm": 0.41683631721157366,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.3014,
"step": 769
},
{
"epoch": 3.118987341772152,
"grad_norm": 0.33459084753148866,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.2796,
"step": 770
},
{
"epoch": 3.123037974683544,
"grad_norm": 0.36506392236705065,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.3185,
"step": 771
},
{
"epoch": 3.1270886075949367,
"grad_norm": 0.33853836028260975,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.3119,
"step": 772
},
{
"epoch": 3.131139240506329,
"grad_norm": 0.36365760626204624,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.3193,
"step": 773
},
{
"epoch": 3.1351898734177217,
"grad_norm": 0.30814572996099626,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.2803,
"step": 774
},
{
"epoch": 3.1392405063291138,
"grad_norm": 0.36139995960798493,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.2883,
"step": 775
},
{
"epoch": 3.1432911392405063,
"grad_norm": 0.3755724035727792,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.337,
"step": 776
},
{
"epoch": 3.147341772151899,
"grad_norm": 0.3457169713904709,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.2714,
"step": 777
},
{
"epoch": 3.1513924050632913,
"grad_norm": 0.36693738865687775,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.3251,
"step": 778
},
{
"epoch": 3.1554430379746834,
"grad_norm": 0.3108953250207747,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.2841,
"step": 779
},
{
"epoch": 3.159493670886076,
"grad_norm": 0.3734993122582448,
"learning_rate": 1.420958430225303e-05,
"loss": 0.3232,
"step": 780
},
{
"epoch": 3.1635443037974684,
"grad_norm": 0.31593848587887224,
"learning_rate": 1.415527989393985e-05,
"loss": 0.317,
"step": 781
},
{
"epoch": 3.167594936708861,
"grad_norm": 0.3276718118857601,
"learning_rate": 1.410102255819891e-05,
"loss": 0.2767,
"step": 782
},
{
"epoch": 3.171645569620253,
"grad_norm": 0.3545851641990169,
"learning_rate": 1.404681273201131e-05,
"loss": 0.3172,
"step": 783
},
{
"epoch": 3.1756962025316455,
"grad_norm": 0.3406559285285585,
"learning_rate": 1.399265085197556e-05,
"loss": 0.2923,
"step": 784
},
{
"epoch": 3.179746835443038,
"grad_norm": 0.3450005685324494,
"learning_rate": 1.393853735430398e-05,
"loss": 0.2892,
"step": 785
},
{
"epoch": 3.1837974683544306,
"grad_norm": 0.3439892751574166,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.3171,
"step": 786
},
{
"epoch": 3.1878481012658226,
"grad_norm": 0.3513171802045518,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.2778,
"step": 787
},
{
"epoch": 3.191898734177215,
"grad_norm": 0.3913693086824694,
"learning_rate": 1.377649151173163e-05,
"loss": 0.3575,
"step": 788
},
{
"epoch": 3.1959493670886077,
"grad_norm": 0.3153713519657333,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.2669,
"step": 789
},
{
"epoch": 3.2,
"grad_norm": 0.4047410345260184,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.3216,
"step": 790
},
{
"epoch": 3.2040506329113922,
"grad_norm": 0.32408273243516045,
"learning_rate": 1.361489677627324e-05,
"loss": 0.3014,
"step": 791
},
{
"epoch": 3.2081012658227848,
"grad_norm": 0.3688706308397694,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.3203,
"step": 792
},
{
"epoch": 3.2121518987341773,
"grad_norm": 0.3569945498418179,
"learning_rate": 1.350742335331241e-05,
"loss": 0.3067,
"step": 793
},
{
"epoch": 3.21620253164557,
"grad_norm": 0.32112325165917066,
"learning_rate": 1.345376486102198e-05,
"loss": 0.3206,
"step": 794
},
{
"epoch": 3.220253164556962,
"grad_norm": 0.35728628435900667,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.2779,
"step": 795
},
{
"epoch": 3.2243037974683544,
"grad_norm": 0.38471472434341625,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.3224,
"step": 796
},
{
"epoch": 3.228354430379747,
"grad_norm": 0.36408576609463167,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.3266,
"step": 797
},
{
"epoch": 3.2324050632911394,
"grad_norm": 0.4073904572888445,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.3323,
"step": 798
},
{
"epoch": 3.2364556962025315,
"grad_norm": 0.3294145730208109,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.2656,
"step": 799
},
{
"epoch": 3.240506329113924,
"grad_norm": 0.3446555096962436,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.3141,
"step": 800
},
{
"epoch": 3.2445569620253165,
"grad_norm": 0.3315531598043646,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.2977,
"step": 801
},
{
"epoch": 3.248607594936709,
"grad_norm": 0.3801369399918843,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.3075,
"step": 802
},
{
"epoch": 3.252658227848101,
"grad_norm": 0.34373635870310193,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.2959,
"step": 803
},
{
"epoch": 3.2567088607594936,
"grad_norm": 0.3522280874448191,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.3018,
"step": 804
},
{
"epoch": 3.260759493670886,
"grad_norm": 0.3455948946471353,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.3111,
"step": 805
},
{
"epoch": 3.2648101265822786,
"grad_norm": 0.3115112790874253,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.3,
"step": 806
},
{
"epoch": 3.2688607594936707,
"grad_norm": 0.34694469404964845,
"learning_rate": 1.276115892225764e-05,
"loss": 0.3155,
"step": 807
},
{
"epoch": 3.2729113924050632,
"grad_norm": 0.32959060819636926,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.3103,
"step": 808
},
{
"epoch": 3.2769620253164558,
"grad_norm": 0.34471829561644185,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.313,
"step": 809
},
{
"epoch": 3.2810126582278483,
"grad_norm": 0.34618067425513654,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.3146,
"step": 810
},
{
"epoch": 3.2850632911392403,
"grad_norm": 0.36130133942051057,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.3002,
"step": 811
},
{
"epoch": 3.289113924050633,
"grad_norm": 0.32652849968352143,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.3052,
"step": 812
},
{
"epoch": 3.2931645569620254,
"grad_norm": 0.360130966828178,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.3215,
"step": 813
},
{
"epoch": 3.297215189873418,
"grad_norm": 0.3136300408349466,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.2889,
"step": 814
},
{
"epoch": 3.30126582278481,
"grad_norm": 0.34008073809012385,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.3112,
"step": 815
},
{
"epoch": 3.3053164556962025,
"grad_norm": 0.33564708830791806,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.3285,
"step": 816
},
{
"epoch": 3.309367088607595,
"grad_norm": 0.3760753780780767,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.2963,
"step": 817
},
{
"epoch": 3.3134177215189875,
"grad_norm": 0.32046607516672027,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.3096,
"step": 818
},
{
"epoch": 3.3174683544303796,
"grad_norm": 0.3420425396850801,
"learning_rate": 1.213055323844297e-05,
"loss": 0.3071,
"step": 819
},
{
"epoch": 3.321518987341772,
"grad_norm": 0.32755677685357626,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.2655,
"step": 820
},
{
"epoch": 3.3255696202531646,
"grad_norm": 0.33119963933059815,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.2985,
"step": 821
},
{
"epoch": 3.329620253164557,
"grad_norm": 0.34765336125731566,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.3077,
"step": 822
},
{
"epoch": 3.333670886075949,
"grad_norm": 0.34022233023384907,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.306,
"step": 823
},
{
"epoch": 3.3377215189873417,
"grad_norm": 0.3203107532517401,
"learning_rate": 1.187045262895488e-05,
"loss": 0.3129,
"step": 824
},
{
"epoch": 3.3417721518987342,
"grad_norm": 0.3508388014041473,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.3264,
"step": 825
},
{
"epoch": 3.3458227848101267,
"grad_norm": 0.2881271547298497,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.2717,
"step": 826
},
{
"epoch": 3.349873417721519,
"grad_norm": 0.33609270405993663,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.3093,
"step": 827
},
{
"epoch": 3.3539240506329113,
"grad_norm": 0.3557206632977488,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.3587,
"step": 828
},
{
"epoch": 3.357974683544304,
"grad_norm": 0.2894360124546518,
"learning_rate": 1.161198884952377e-05,
"loss": 0.2528,
"step": 829
},
{
"epoch": 3.3620253164556964,
"grad_norm": 0.32212042780261674,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.2836,
"step": 830
},
{
"epoch": 3.3660759493670884,
"grad_norm": 0.3208239634702613,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.3078,
"step": 831
},
{
"epoch": 3.370126582278481,
"grad_norm": 0.3190607262192318,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.3159,
"step": 832
},
{
"epoch": 3.3741772151898735,
"grad_norm": 0.3419113162657851,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.3054,
"step": 833
},
{
"epoch": 3.378227848101266,
"grad_norm": 0.3151917177504675,
"learning_rate": 1.135521394010506e-05,
"loss": 0.3074,
"step": 834
},
{
"epoch": 3.382278481012658,
"grad_norm": 0.34049974394292876,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.346,
"step": 835
},
{
"epoch": 3.3863291139240506,
"grad_norm": 0.2990413763146066,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.2824,
"step": 836
},
{
"epoch": 3.390379746835443,
"grad_norm": 0.3282011979863411,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.3233,
"step": 837
},
{
"epoch": 3.3944303797468356,
"grad_norm": 0.3283512078096152,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.3059,
"step": 838
},
{
"epoch": 3.3984810126582277,
"grad_norm": 0.27873251712231323,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.2984,
"step": 839
},
{
"epoch": 3.40253164556962,
"grad_norm": 0.345166046216817,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.3423,
"step": 840
},
{
"epoch": 3.4065822784810127,
"grad_norm": 0.3036988149768153,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.2975,
"step": 841
},
{
"epoch": 3.410632911392405,
"grad_norm": 0.3033156845886753,
"learning_rate": 1.0948015846744e-05,
"loss": 0.2893,
"step": 842
},
{
"epoch": 3.4146835443037973,
"grad_norm": 0.3334049182086911,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.3317,
"step": 843
},
{
"epoch": 3.41873417721519,
"grad_norm": 0.3105437189699824,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.3315,
"step": 844
},
{
"epoch": 3.4227848101265823,
"grad_norm": 0.323508239140487,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.3124,
"step": 845
},
{
"epoch": 3.426835443037975,
"grad_norm": 0.2863298679654191,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.2727,
"step": 846
},
{
"epoch": 3.430886075949367,
"grad_norm": 0.3499745981428249,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.3473,
"step": 847
},
{
"epoch": 3.4349367088607594,
"grad_norm": 0.317479786420863,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.2998,
"step": 848
},
{
"epoch": 3.438987341772152,
"grad_norm": 0.2800861832750057,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.2862,
"step": 849
},
{
"epoch": 3.4430379746835444,
"grad_norm": 0.3105554502570675,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.2988,
"step": 850
},
{
"epoch": 3.4470886075949365,
"grad_norm": 0.29820053981401085,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.3004,
"step": 851
},
{
"epoch": 3.451139240506329,
"grad_norm": 0.33298981182853415,
"learning_rate": 1.044560362975474e-05,
"loss": 0.3316,
"step": 852
},
{
"epoch": 3.4551898734177215,
"grad_norm": 0.29511715843378067,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.3012,
"step": 853
},
{
"epoch": 3.459240506329114,
"grad_norm": 0.30982309203746633,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.305,
"step": 854
},
{
"epoch": 3.463291139240506,
"grad_norm": 0.3268303114107129,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.3044,
"step": 855
},
{
"epoch": 3.4673417721518986,
"grad_norm": 0.3041481885207742,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.3322,
"step": 856
},
{
"epoch": 3.471392405063291,
"grad_norm": 0.29359803162882436,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.3161,
"step": 857
},
{
"epoch": 3.4754430379746837,
"grad_norm": 0.33547959309472525,
"learning_rate": 1.014782400334433e-05,
"loss": 0.3304,
"step": 858
},
{
"epoch": 3.479493670886076,
"grad_norm": 0.2854414558818818,
"learning_rate": 1.009846945771296e-05,
"loss": 0.2765,
"step": 859
},
{
"epoch": 3.4835443037974683,
"grad_norm": 0.3153419076232269,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.319,
"step": 860
},
{
"epoch": 3.487594936708861,
"grad_norm": 0.28937503194380115,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.2821,
"step": 861
},
{
"epoch": 3.4916455696202533,
"grad_norm": 0.3118946369729269,
"learning_rate": 9.950885880977891e-06,
"loss": 0.2919,
"step": 862
},
{
"epoch": 3.4956962025316454,
"grad_norm": 0.33812243250474633,
"learning_rate": 9.901852696139382e-06,
"loss": 0.3227,
"step": 863
},
{
"epoch": 3.499746835443038,
"grad_norm": 0.29119758578192556,
"learning_rate": 9.852900840391027e-06,
"loss": 0.303,
"step": 864
},
{
"epoch": 3.5037974683544304,
"grad_norm": 0.3220567372997527,
"learning_rate": 9.804030707984313e-06,
"loss": 0.3204,
"step": 865
},
{
"epoch": 3.507848101265823,
"grad_norm": 0.2954472088528195,
"learning_rate": 9.755242692512599e-06,
"loss": 0.2967,
"step": 866
},
{
"epoch": 3.5118987341772154,
"grad_norm": 0.3081130479604762,
"learning_rate": 9.70653718690782e-06,
"loss": 0.2952,
"step": 867
},
{
"epoch": 3.5159493670886075,
"grad_norm": 0.3161065062016706,
"learning_rate": 9.657914583437454e-06,
"loss": 0.3152,
"step": 868
},
{
"epoch": 3.52,
"grad_norm": 0.2917534431993787,
"learning_rate": 9.609375273701246e-06,
"loss": 0.3079,
"step": 869
},
{
"epoch": 3.5240506329113925,
"grad_norm": 0.3047855901999337,
"learning_rate": 9.560919648628133e-06,
"loss": 0.3053,
"step": 870
},
{
"epoch": 3.5281012658227846,
"grad_norm": 0.2949892434007892,
"learning_rate": 9.512548098473047e-06,
"loss": 0.3021,
"step": 871
},
{
"epoch": 3.532151898734177,
"grad_norm": 0.29795237246533735,
"learning_rate": 9.464261012813825e-06,
"loss": 0.3096,
"step": 872
},
{
"epoch": 3.5362025316455696,
"grad_norm": 0.30690726648655003,
"learning_rate": 9.416058780547987e-06,
"loss": 0.3178,
"step": 873
},
{
"epoch": 3.540253164556962,
"grad_norm": 0.3060320219546213,
"learning_rate": 9.367941789889714e-06,
"loss": 0.2947,
"step": 874
},
{
"epoch": 3.5443037974683547,
"grad_norm": 0.324389002710288,
"learning_rate": 9.319910428366607e-06,
"loss": 0.3228,
"step": 875
},
{
"epoch": 3.5483544303797467,
"grad_norm": 0.29581976139815386,
"learning_rate": 9.271965082816667e-06,
"loss": 0.2805,
"step": 876
},
{
"epoch": 3.5524050632911393,
"grad_norm": 0.31620267845966676,
"learning_rate": 9.224106139385111e-06,
"loss": 0.3237,
"step": 877
},
{
"epoch": 3.5564556962025318,
"grad_norm": 0.3055851029000954,
"learning_rate": 9.176333983521291e-06,
"loss": 0.3309,
"step": 878
},
{
"epoch": 3.560506329113924,
"grad_norm": 0.314328456335115,
"learning_rate": 9.12864899997558e-06,
"loss": 0.3277,
"step": 879
},
{
"epoch": 3.5645569620253164,
"grad_norm": 0.28753022126416977,
"learning_rate": 9.08105157279628e-06,
"loss": 0.2715,
"step": 880
},
{
"epoch": 3.568607594936709,
"grad_norm": 0.3137419147599129,
"learning_rate": 9.03354208532653e-06,
"loss": 0.3192,
"step": 881
},
{
"epoch": 3.5726582278481014,
"grad_norm": 0.31446824545497654,
"learning_rate": 8.986120920201205e-06,
"loss": 0.3302,
"step": 882
},
{
"epoch": 3.576708860759494,
"grad_norm": 0.2965754404570372,
"learning_rate": 8.938788459343852e-06,
"loss": 0.2812,
"step": 883
},
{
"epoch": 3.580759493670886,
"grad_norm": 0.32087429560221115,
"learning_rate": 8.8915450839636e-06,
"loss": 0.3356,
"step": 884
},
{
"epoch": 3.5848101265822785,
"grad_norm": 0.3124157974402817,
"learning_rate": 8.844391174552116e-06,
"loss": 0.2912,
"step": 885
},
{
"epoch": 3.588860759493671,
"grad_norm": 0.27894715997950303,
"learning_rate": 8.797327110880479e-06,
"loss": 0.2856,
"step": 886
},
{
"epoch": 3.592911392405063,
"grad_norm": 0.31917150074511585,
"learning_rate": 8.750353271996206e-06,
"loss": 0.3246,
"step": 887
},
{
"epoch": 3.5969620253164556,
"grad_norm": 0.3006640708234061,
"learning_rate": 8.703470036220132e-06,
"loss": 0.2888,
"step": 888
},
{
"epoch": 3.601012658227848,
"grad_norm": 0.2925557945843118,
"learning_rate": 8.656677781143394e-06,
"loss": 0.2843,
"step": 889
},
{
"epoch": 3.6050632911392406,
"grad_norm": 0.30190767222540416,
"learning_rate": 8.609976883624377e-06,
"loss": 0.3026,
"step": 890
},
{
"epoch": 3.609113924050633,
"grad_norm": 0.32202338015234727,
"learning_rate": 8.563367719785698e-06,
"loss": 0.3104,
"step": 891
},
{
"epoch": 3.613164556962025,
"grad_norm": 0.30686352722078103,
"learning_rate": 8.516850665011138e-06,
"loss": 0.2733,
"step": 892
},
{
"epoch": 3.6172151898734177,
"grad_norm": 0.3342596476681076,
"learning_rate": 8.47042609394269e-06,
"loss": 0.303,
"step": 893
},
{
"epoch": 3.6212658227848102,
"grad_norm": 0.29359613088392944,
"learning_rate": 8.424094380477432e-06,
"loss": 0.2914,
"step": 894
},
{
"epoch": 3.6253164556962023,
"grad_norm": 0.3235333910732101,
"learning_rate": 8.37785589776465e-06,
"loss": 0.3324,
"step": 895
},
{
"epoch": 3.629367088607595,
"grad_norm": 0.30244325623446444,
"learning_rate": 8.331711018202694e-06,
"loss": 0.2831,
"step": 896
},
{
"epoch": 3.6334177215189873,
"grad_norm": 0.3163316815514907,
"learning_rate": 8.285660113436104e-06,
"loss": 0.3148,
"step": 897
},
{
"epoch": 3.63746835443038,
"grad_norm": 0.2962403296702216,
"learning_rate": 8.239703554352527e-06,
"loss": 0.3027,
"step": 898
},
{
"epoch": 3.6415189873417724,
"grad_norm": 0.3006844046840596,
"learning_rate": 8.193841711079775e-06,
"loss": 0.3012,
"step": 899
},
{
"epoch": 3.6455696202531644,
"grad_norm": 0.29949269371433795,
"learning_rate": 8.148074952982828e-06,
"loss": 0.2771,
"step": 900
},
{
"epoch": 3.649620253164557,
"grad_norm": 0.31515533767148507,
"learning_rate": 8.102403648660859e-06,
"loss": 0.2992,
"step": 901
},
{
"epoch": 3.6536708860759495,
"grad_norm": 0.3191208880663791,
"learning_rate": 8.056828165944282e-06,
"loss": 0.2861,
"step": 902
},
{
"epoch": 3.6577215189873415,
"grad_norm": 0.3410184023016313,
"learning_rate": 8.011348871891762e-06,
"loss": 0.3423,
"step": 903
},
{
"epoch": 3.661772151898734,
"grad_norm": 0.3150330099158935,
"learning_rate": 7.965966132787287e-06,
"loss": 0.3133,
"step": 904
},
{
"epoch": 3.6658227848101266,
"grad_norm": 0.28707458957366655,
"learning_rate": 7.920680314137189e-06,
"loss": 0.3038,
"step": 905
},
{
"epoch": 3.669873417721519,
"grad_norm": 0.2906858922804771,
"learning_rate": 7.875491780667246e-06,
"loss": 0.3047,
"step": 906
},
{
"epoch": 3.6739240506329116,
"grad_norm": 0.3221229291640142,
"learning_rate": 7.830400896319667e-06,
"loss": 0.3109,
"step": 907
},
{
"epoch": 3.6779746835443037,
"grad_norm": 0.32058177758259654,
"learning_rate": 7.785408024250259e-06,
"loss": 0.3014,
"step": 908
},
{
"epoch": 3.682025316455696,
"grad_norm": 0.3141503649939848,
"learning_rate": 7.74051352682542e-06,
"loss": 0.3121,
"step": 909
},
{
"epoch": 3.6860759493670887,
"grad_norm": 0.28053197985982453,
"learning_rate": 7.695717765619257e-06,
"loss": 0.293,
"step": 910
},
{
"epoch": 3.690126582278481,
"grad_norm": 0.3246638799881899,
"learning_rate": 7.651021101410673e-06,
"loss": 0.3,
"step": 911
},
{
"epoch": 3.6941772151898733,
"grad_norm": 0.32332347091975094,
"learning_rate": 7.606423894180464e-06,
"loss": 0.323,
"step": 912
},
{
"epoch": 3.698227848101266,
"grad_norm": 0.2851442032734005,
"learning_rate": 7.56192650310839e-06,
"loss": 0.2572,
"step": 913
},
{
"epoch": 3.7022784810126583,
"grad_norm": 0.30841162887076623,
"learning_rate": 7.517529286570349e-06,
"loss": 0.3067,
"step": 914
},
{
"epoch": 3.706329113924051,
"grad_norm": 0.3048328954400383,
"learning_rate": 7.473232602135387e-06,
"loss": 0.2921,
"step": 915
},
{
"epoch": 3.710379746835443,
"grad_norm": 0.30245747089816366,
"learning_rate": 7.429036806562935e-06,
"loss": 0.3141,
"step": 916
},
{
"epoch": 3.7144303797468354,
"grad_norm": 0.2982971588164874,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.2889,
"step": 917
},
{
"epoch": 3.718481012658228,
"grad_norm": 0.3231706532187127,
"learning_rate": 7.340949304977567e-06,
"loss": 0.3435,
"step": 918
},
{
"epoch": 3.72253164556962,
"grad_norm": 0.29420991132476276,
"learning_rate": 7.297058308409282e-06,
"loss": 0.2802,
"step": 919
},
{
"epoch": 3.7265822784810125,
"grad_norm": 0.29290198013137597,
"learning_rate": 7.25326961958704e-06,
"loss": 0.296,
"step": 920
},
{
"epoch": 3.730632911392405,
"grad_norm": 0.32861657323294113,
"learning_rate": 7.209583591178921e-06,
"loss": 0.3177,
"step": 921
},
{
"epoch": 3.7346835443037976,
"grad_norm": 0.2906745999110606,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.2869,
"step": 922
},
{
"epoch": 3.73873417721519,
"grad_norm": 0.3304025371941078,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.3386,
"step": 923
},
{
"epoch": 3.742784810126582,
"grad_norm": 0.28228553123191036,
"learning_rate": 7.079144982700909e-06,
"loss": 0.2528,
"step": 924
},
{
"epoch": 3.7468354430379747,
"grad_norm": 0.35000400115764246,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.3421,
"step": 925
},
{
"epoch": 3.750886075949367,
"grad_norm": 0.30136888340616913,
"learning_rate": 6.99270564069757e-06,
"loss": 0.2893,
"step": 926
},
{
"epoch": 3.7549367088607593,
"grad_norm": 0.30718401477872587,
"learning_rate": 6.949642934304375e-06,
"loss": 0.2992,
"step": 927
},
{
"epoch": 3.7589873417721518,
"grad_norm": 0.29859027962971385,
"learning_rate": 6.906685333692871e-06,
"loss": 0.2803,
"step": 928
},
{
"epoch": 3.7630379746835443,
"grad_norm": 0.29825407448031976,
"learning_rate": 6.86383318483769e-06,
"loss": 0.2919,
"step": 929
},
{
"epoch": 3.767088607594937,
"grad_norm": 0.3133417441587745,
"learning_rate": 6.821086832864139e-06,
"loss": 0.3091,
"step": 930
},
{
"epoch": 3.7711392405063293,
"grad_norm": 0.32727524975060984,
"learning_rate": 6.77844662204546e-06,
"loss": 0.3185,
"step": 931
},
{
"epoch": 3.7751898734177214,
"grad_norm": 0.29719040755056664,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.2934,
"step": 932
},
{
"epoch": 3.779240506329114,
"grad_norm": 0.3075800618279458,
"learning_rate": 6.693485996688695e-06,
"loss": 0.2929,
"step": 933
},
{
"epoch": 3.7832911392405064,
"grad_norm": 0.33342046143388815,
"learning_rate": 6.651166266411801e-06,
"loss": 0.2815,
"step": 934
},
{
"epoch": 3.7873417721518985,
"grad_norm": 0.3051006578135973,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.3179,
"step": 935
},
{
"epoch": 3.791392405063291,
"grad_norm": 0.31382044739309506,
"learning_rate": 6.566849674844711e-06,
"loss": 0.2885,
"step": 936
},
{
"epoch": 3.7954430379746835,
"grad_norm": 0.3156058941620484,
"learning_rate": 6.524853492628747e-06,
"loss": 0.2908,
"step": 937
},
{
"epoch": 3.799493670886076,
"grad_norm": 0.31283499175666163,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.3028,
"step": 938
},
{
"epoch": 3.8035443037974686,
"grad_norm": 0.2948743137197648,
"learning_rate": 6.441187046486648e-06,
"loss": 0.2916,
"step": 939
},
{
"epoch": 3.8075949367088606,
"grad_norm": 0.331392753444647,
"learning_rate": 6.399517456398567e-06,
"loss": 0.3054,
"step": 940
},
{
"epoch": 3.811645569620253,
"grad_norm": 0.3252245741298627,
"learning_rate": 6.357957402727164e-06,
"loss": 0.284,
"step": 941
},
{
"epoch": 3.8156962025316457,
"grad_norm": 0.3107654668856293,
"learning_rate": 6.316507220191395e-06,
"loss": 0.329,
"step": 942
},
{
"epoch": 3.8197468354430377,
"grad_norm": 0.2660145410625429,
"learning_rate": 6.275167242625331e-06,
"loss": 0.2921,
"step": 943
},
{
"epoch": 3.8237974683544302,
"grad_norm": 0.3514817371192165,
"learning_rate": 6.233937802975471e-06,
"loss": 0.3348,
"step": 944
},
{
"epoch": 3.8278481012658228,
"grad_norm": 0.3200331242389006,
"learning_rate": 6.192819233298046e-06,
"loss": 0.2907,
"step": 945
},
{
"epoch": 3.8318987341772153,
"grad_norm": 0.29331748891004805,
"learning_rate": 6.151811864756383e-06,
"loss": 0.2881,
"step": 946
},
{
"epoch": 3.835949367088608,
"grad_norm": 0.3199940881224829,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.3284,
"step": 947
},
{
"epoch": 3.84,
"grad_norm": 0.31224872873727205,
"learning_rate": 6.070132051252868e-06,
"loss": 0.3103,
"step": 948
},
{
"epoch": 3.8440506329113924,
"grad_norm": 0.31222552176566476,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.2719,
"step": 949
},
{
"epoch": 3.848101265822785,
"grad_norm": 0.3004480665170546,
"learning_rate": 5.988900993811575e-06,
"loss": 0.3274,
"step": 950
},
{
"epoch": 3.852151898734177,
"grad_norm": 0.3145877981146191,
"learning_rate": 5.948454566959363e-06,
"loss": 0.3248,
"step": 951
},
{
"epoch": 3.8562025316455695,
"grad_norm": 0.29172027946764323,
"learning_rate": 5.908121309322328e-06,
"loss": 0.3064,
"step": 952
},
{
"epoch": 3.860253164556962,
"grad_norm": 0.30774672901081923,
"learning_rate": 5.867901545738976e-06,
"loss": 0.3058,
"step": 953
},
{
"epoch": 3.8643037974683545,
"grad_norm": 0.28571751280757784,
"learning_rate": 5.827795600133774e-06,
"loss": 0.2747,
"step": 954
},
{
"epoch": 3.868354430379747,
"grad_norm": 0.31511925235471416,
"learning_rate": 5.787803795514466e-06,
"loss": 0.3453,
"step": 955
},
{
"epoch": 3.872405063291139,
"grad_norm": 0.2710031276348199,
"learning_rate": 5.747926453969576e-06,
"loss": 0.2908,
"step": 956
},
{
"epoch": 3.8764556962025316,
"grad_norm": 0.2972170849396098,
"learning_rate": 5.708163896665708e-06,
"loss": 0.2848,
"step": 957
},
{
"epoch": 3.880506329113924,
"grad_norm": 0.29853961006090424,
"learning_rate": 5.668516443845047e-06,
"loss": 0.2866,
"step": 958
},
{
"epoch": 3.884556962025316,
"grad_norm": 0.31110346019685386,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.327,
"step": 959
},
{
"epoch": 3.8886075949367087,
"grad_norm": 0.3138426740579346,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.2929,
"step": 960
},
{
"epoch": 3.8926582278481012,
"grad_norm": 0.28253884692125153,
"learning_rate": 5.550267900783019e-06,
"loss": 0.2986,
"step": 961
},
{
"epoch": 3.8967088607594937,
"grad_norm": 0.31246558944307107,
"learning_rate": 5.511084049737623e-06,
"loss": 0.3367,
"step": 962
},
{
"epoch": 3.9007594936708863,
"grad_norm": 0.32469264786769253,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.3078,
"step": 963
},
{
"epoch": 3.9048101265822783,
"grad_norm": 0.29590405525136926,
"learning_rate": 5.433066737499948e-06,
"loss": 0.2985,
"step": 964
},
{
"epoch": 3.908860759493671,
"grad_norm": 0.2973657810863817,
"learning_rate": 5.394233904648376e-06,
"loss": 0.2708,
"step": 965
},
{
"epoch": 3.9129113924050634,
"grad_norm": 0.3026558729076128,
"learning_rate": 5.355518704628997e-06,
"loss": 0.3309,
"step": 966
},
{
"epoch": 3.9169620253164554,
"grad_norm": 0.2923715873208451,
"learning_rate": 5.316921449248731e-06,
"loss": 0.2979,
"step": 967
},
{
"epoch": 3.921012658227848,
"grad_norm": 0.2946020526853441,
"learning_rate": 5.278442449364538e-06,
"loss": 0.3001,
"step": 968
},
{
"epoch": 3.9250632911392405,
"grad_norm": 0.2953565894359084,
"learning_rate": 5.240082014881016e-06,
"loss": 0.2959,
"step": 969
},
{
"epoch": 3.929113924050633,
"grad_norm": 0.31768546719123686,
"learning_rate": 5.201840454747822e-06,
"loss": 0.3099,
"step": 970
},
{
"epoch": 3.9331645569620255,
"grad_norm": 0.28911389703836504,
"learning_rate": 5.163718076957223e-06,
"loss": 0.3026,
"step": 971
},
{
"epoch": 3.9372151898734176,
"grad_norm": 0.31969625051612316,
"learning_rate": 5.125715188541609e-06,
"loss": 0.3032,
"step": 972
},
{
"epoch": 3.94126582278481,
"grad_norm": 0.27696520303086686,
"learning_rate": 5.087832095571021e-06,
"loss": 0.2885,
"step": 973
},
{
"epoch": 3.9453164556962026,
"grad_norm": 0.28024027082014996,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.2984,
"step": 974
},
{
"epoch": 3.9493670886075947,
"grad_norm": 0.2766181036147888,
"learning_rate": 5.01242651541854e-06,
"loss": 0.2892,
"step": 975
},
{
"epoch": 3.953417721518987,
"grad_norm": 0.30089518883457333,
"learning_rate": 4.974904635542815e-06,
"loss": 0.3221,
"step": 976
},
{
"epoch": 3.9574683544303797,
"grad_norm": 0.2963032272041155,
"learning_rate": 4.937503765719582e-06,
"loss": 0.2858,
"step": 977
},
{
"epoch": 3.961518987341772,
"grad_norm": 0.2989299685008238,
"learning_rate": 4.900224207170299e-06,
"loss": 0.3058,
"step": 978
},
{
"epoch": 3.9655696202531647,
"grad_norm": 0.28778430135709554,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.2862,
"step": 979
},
{
"epoch": 3.969620253164557,
"grad_norm": 0.30238031795623715,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.3175,
"step": 980
},
{
"epoch": 3.9736708860759493,
"grad_norm": 0.2901373896585849,
"learning_rate": 4.789116396710924e-06,
"loss": 0.2808,
"step": 981
},
{
"epoch": 3.977721518987342,
"grad_norm": 0.2745605121063818,
"learning_rate": 4.752325075895368e-06,
"loss": 0.2833,
"step": 982
},
{
"epoch": 3.981772151898734,
"grad_norm": 0.276192780156961,
"learning_rate": 4.715656557757473e-06,
"loss": 0.3004,
"step": 983
},
{
"epoch": 3.9858227848101264,
"grad_norm": 0.2871749739397835,
"learning_rate": 4.679111137620442e-06,
"loss": 0.3002,
"step": 984
},
{
"epoch": 3.989873417721519,
"grad_norm": 0.2971961781290793,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.2787,
"step": 985
},
{
"epoch": 3.9939240506329114,
"grad_norm": 0.28309720615137624,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.2708,
"step": 986
},
{
"epoch": 3.997974683544304,
"grad_norm": 0.3318791103666527,
"learning_rate": 4.570216403561141e-06,
"loss": 0.3321,
"step": 987
},
{
"epoch": 4.002025316455696,
"grad_norm": 0.4098356822836525,
"learning_rate": 4.534166308795815e-06,
"loss": 0.2867,
"step": 988
},
{
"epoch": 4.006075949367089,
"grad_norm": 0.49639143360673627,
"learning_rate": 4.498240773728859e-06,
"loss": 0.2332,
"step": 989
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.41178455875739367,
"learning_rate": 4.462440087699609e-06,
"loss": 0.2563,
"step": 990
},
{
"epoch": 4.014177215189873,
"grad_norm": 0.5189184213534569,
"learning_rate": 4.426764539041861e-06,
"loss": 0.2313,
"step": 991
},
{
"epoch": 4.018227848101266,
"grad_norm": 0.49637572305955996,
"learning_rate": 4.391214415081582e-06,
"loss": 0.2121,
"step": 992
},
{
"epoch": 4.022278481012658,
"grad_norm": 0.3712845171516279,
"learning_rate": 4.355790002134579e-06,
"loss": 0.2125,
"step": 993
},
{
"epoch": 4.02632911392405,
"grad_norm": 0.5197116781102222,
"learning_rate": 4.320491585504207e-06,
"loss": 0.2646,
"step": 994
},
{
"epoch": 4.030379746835443,
"grad_norm": 0.4417038791759217,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.205,
"step": 995
},
{
"epoch": 4.034430379746835,
"grad_norm": 0.4147104865557197,
"learning_rate": 4.250273877330691e-06,
"loss": 0.2454,
"step": 996
},
{
"epoch": 4.038481012658228,
"grad_norm": 0.3656288657437056,
"learning_rate": 4.215355151311313e-06,
"loss": 0.2391,
"step": 997
},
{
"epoch": 4.04253164556962,
"grad_norm": 0.4034328338860195,
"learning_rate": 4.180563552651542e-06,
"loss": 0.2135,
"step": 998
},
{
"epoch": 4.046582278481012,
"grad_norm": 0.4161356942607432,
"learning_rate": 4.145899361558147e-06,
"loss": 0.2403,
"step": 999
},
{
"epoch": 4.050632911392405,
"grad_norm": 0.3370715039425024,
"learning_rate": 4.111362857211738e-06,
"loss": 0.2282,
"step": 1000
},
{
"epoch": 4.054683544303797,
"grad_norm": 0.33035955630023056,
"learning_rate": 4.076954317764592e-06,
"loss": 0.2418,
"step": 1001
},
{
"epoch": 4.0587341772151895,
"grad_norm": 0.3348140446625731,
"learning_rate": 4.042674020338335e-06,
"loss": 0.2353,
"step": 1002
},
{
"epoch": 4.062784810126582,
"grad_norm": 0.36153705807964637,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.24,
"step": 1003
},
{
"epoch": 4.0668354430379745,
"grad_norm": 0.3209073272501053,
"learning_rate": 3.974499254868674e-06,
"loss": 0.2367,
"step": 1004
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.32109922644077427,
"learning_rate": 3.940605335895451e-06,
"loss": 0.2382,
"step": 1005
},
{
"epoch": 4.0749367088607595,
"grad_norm": 0.32114976297039843,
"learning_rate": 3.90684075707908e-06,
"loss": 0.2174,
"step": 1006
},
{
"epoch": 4.078987341772152,
"grad_norm": 0.38344402803801353,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.247,
"step": 1007
},
{
"epoch": 4.083037974683545,
"grad_norm": 0.3419113360643892,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.2396,
"step": 1008
},
{
"epoch": 4.087088607594937,
"grad_norm": 0.28697361818333794,
"learning_rate": 3.806325775702304e-06,
"loss": 0.2027,
"step": 1009
},
{
"epoch": 4.091139240506329,
"grad_norm": 0.3036866837416376,
"learning_rate": 3.773081266416434e-06,
"loss": 0.2124,
"step": 1010
},
{
"epoch": 4.095189873417722,
"grad_norm": 0.3138089090750777,
"learning_rate": 3.739967446503245e-06,
"loss": 0.2337,
"step": 1011
},
{
"epoch": 4.099240506329114,
"grad_norm": 0.3315105224801696,
"learning_rate": 3.706984582656894e-06,
"loss": 0.2236,
"step": 1012
},
{
"epoch": 4.103291139240507,
"grad_norm": 0.30242835807261625,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.2442,
"step": 1013
},
{
"epoch": 4.107341772151899,
"grad_norm": 0.2918355985380491,
"learning_rate": 3.641412784665648e-06,
"loss": 0.2242,
"step": 1014
},
{
"epoch": 4.111392405063291,
"grad_norm": 0.3307248873641473,
"learning_rate": 3.608824378627005e-06,
"loss": 0.2156,
"step": 1015
},
{
"epoch": 4.115443037974684,
"grad_norm": 0.33013489533279206,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.2583,
"step": 1016
},
{
"epoch": 4.119493670886076,
"grad_norm": 0.28214144489761234,
"learning_rate": 3.544043864774269e-06,
"loss": 0.2165,
"step": 1017
},
{
"epoch": 4.123544303797468,
"grad_norm": 0.3190270541188394,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.2415,
"step": 1018
},
{
"epoch": 4.127594936708861,
"grad_norm": 0.31524491822614564,
"learning_rate": 3.479793485887819e-06,
"loss": 0.2468,
"step": 1019
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.3288308437366631,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.2106,
"step": 1020
},
{
"epoch": 4.135696202531646,
"grad_norm": 0.29097703873980013,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.2186,
"step": 1021
},
{
"epoch": 4.139746835443038,
"grad_norm": 0.30687366291071083,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.2326,
"step": 1022
},
{
"epoch": 4.14379746835443,
"grad_norm": 0.31608019010996485,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.2378,
"step": 1023
},
{
"epoch": 4.147848101265823,
"grad_norm": 0.291004879980413,
"learning_rate": 3.321500420335e-06,
"loss": 0.2091,
"step": 1024
},
{
"epoch": 4.151898734177215,
"grad_norm": 0.3164743577683457,
"learning_rate": 3.290243771741275e-06,
"loss": 0.2127,
"step": 1025
},
{
"epoch": 4.155949367088607,
"grad_norm": 0.32592226650954476,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.244,
"step": 1026
},
{
"epoch": 4.16,
"grad_norm": 0.2953419568487332,
"learning_rate": 3.228134459445149e-06,
"loss": 0.2131,
"step": 1027
},
{
"epoch": 4.164050632911392,
"grad_norm": 0.3185108685055801,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.2413,
"step": 1028
},
{
"epoch": 4.168101265822785,
"grad_norm": 0.2883018798764031,
"learning_rate": 3.166565459258513e-06,
"loss": 0.2016,
"step": 1029
},
{
"epoch": 4.172151898734177,
"grad_norm": 0.2906227051100642,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.2379,
"step": 1030
},
{
"epoch": 4.176202531645569,
"grad_norm": 0.29100401024179173,
"learning_rate": 3.105538754650419e-06,
"loss": 0.2515,
"step": 1031
},
{
"epoch": 4.180253164556962,
"grad_norm": 0.2703576731839679,
"learning_rate": 3.07522937824712e-06,
"loss": 0.1993,
"step": 1032
},
{
"epoch": 4.184303797468354,
"grad_norm": 0.3007873835579283,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.2326,
"step": 1033
},
{
"epoch": 4.188354430379746,
"grad_norm": 0.3106689439175919,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.2485,
"step": 1034
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.284270147445994,
"learning_rate": 2.985120078631465e-06,
"loss": 0.207,
"step": 1035
},
{
"epoch": 4.1964556962025314,
"grad_norm": 0.29660335442004376,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.2223,
"step": 1036
},
{
"epoch": 4.200506329113924,
"grad_norm": 0.2986488947077349,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.237,
"step": 1037
},
{
"epoch": 4.2045569620253165,
"grad_norm": 0.30071147024816247,
"learning_rate": 2.896244091927678e-06,
"loss": 0.2459,
"step": 1038
},
{
"epoch": 4.2086075949367086,
"grad_norm": 0.2756482622630843,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.211,
"step": 1039
},
{
"epoch": 4.2126582278481015,
"grad_norm": 0.2905667812389659,
"learning_rate": 2.837681792951994e-06,
"loss": 0.2352,
"step": 1040
},
{
"epoch": 4.216708860759494,
"grad_norm": 0.29614868573856873,
"learning_rate": 2.808607860255981e-06,
"loss": 0.2338,
"step": 1041
},
{
"epoch": 4.220759493670886,
"grad_norm": 0.3288716760853001,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.2332,
"step": 1042
},
{
"epoch": 4.224810126582279,
"grad_norm": 0.30696068894287404,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.2186,
"step": 1043
},
{
"epoch": 4.228860759493671,
"grad_norm": 0.2870121476432383,
"learning_rate": 2.722217735873718e-06,
"loss": 0.2195,
"step": 1044
},
{
"epoch": 4.232911392405064,
"grad_norm": 0.29446578946175395,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.2261,
"step": 1045
},
{
"epoch": 4.236962025316456,
"grad_norm": 0.28562583784210216,
"learning_rate": 2.665319697490205e-06,
"loss": 0.2144,
"step": 1046
},
{
"epoch": 4.241012658227848,
"grad_norm": 0.2971732629463712,
"learning_rate": 2.637079980714945e-06,
"loss": 0.2298,
"step": 1047
},
{
"epoch": 4.245063291139241,
"grad_norm": 0.28474722784761014,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.2025,
"step": 1048
},
{
"epoch": 4.249113924050633,
"grad_norm": 0.31017818902756883,
"learning_rate": 2.581020289189571e-06,
"loss": 0.244,
"step": 1049
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.3057541566833319,
"learning_rate": 2.553200765936501e-06,
"loss": 0.2387,
"step": 1050
},
{
"epoch": 4.257215189873418,
"grad_norm": 0.28587612298782433,
"learning_rate": 2.525521756805962e-06,
"loss": 0.2167,
"step": 1051
},
{
"epoch": 4.26126582278481,
"grad_norm": 0.28731559654731836,
"learning_rate": 2.497983484720885e-06,
"loss": 0.2275,
"step": 1052
},
{
"epoch": 4.265316455696203,
"grad_norm": 0.304092298956458,
"learning_rate": 2.470586171470728e-06,
"loss": 0.2437,
"step": 1053
},
{
"epoch": 4.269367088607595,
"grad_norm": 0.2765274999097936,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.2251,
"step": 1054
},
{
"epoch": 4.273417721518987,
"grad_norm": 0.30802636785397436,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.256,
"step": 1055
},
{
"epoch": 4.27746835443038,
"grad_norm": 0.32073281598041087,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.2191,
"step": 1056
},
{
"epoch": 4.281518987341772,
"grad_norm": 0.2715838407718697,
"learning_rate": 2.362410902836978e-06,
"loss": 0.1867,
"step": 1057
},
{
"epoch": 4.285569620253165,
"grad_norm": 0.28122965603108374,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.2085,
"step": 1058
},
{
"epoch": 4.289620253164557,
"grad_norm": 0.32013290523283583,
"learning_rate": 2.309174704446411e-06,
"loss": 0.2353,
"step": 1059
},
{
"epoch": 4.293670886075949,
"grad_norm": 0.2980128455648876,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.2134,
"step": 1060
},
{
"epoch": 4.297721518987342,
"grad_norm": 0.2863157085052083,
"learning_rate": 2.256508422807855e-06,
"loss": 0.2082,
"step": 1061
},
{
"epoch": 4.301772151898734,
"grad_norm": 0.3129474674017205,
"learning_rate": 2.230389531696946e-06,
"loss": 0.213,
"step": 1062
},
{
"epoch": 4.305822784810126,
"grad_norm": 0.29629262304019083,
"learning_rate": 2.204413754585857e-06,
"loss": 0.2202,
"step": 1063
},
{
"epoch": 4.309873417721519,
"grad_norm": 0.3157588117919042,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.2395,
"step": 1064
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.28822998434737446,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.2033,
"step": 1065
},
{
"epoch": 4.317974683544303,
"grad_norm": 0.3065900184889734,
"learning_rate": 2.127347193531757e-06,
"loss": 0.2407,
"step": 1066
},
{
"epoch": 4.322025316455696,
"grad_norm": 0.2777959597328875,
"learning_rate": 2.101945952921942e-06,
"loss": 0.2202,
"step": 1067
},
{
"epoch": 4.326075949367088,
"grad_norm": 0.29457621104459303,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.2329,
"step": 1068
},
{
"epoch": 4.330126582278481,
"grad_norm": 0.282726070329837,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.2143,
"step": 1069
},
{
"epoch": 4.334177215189873,
"grad_norm": 0.3075325235224783,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.2369,
"step": 1070
},
{
"epoch": 4.3382278481012655,
"grad_norm": 0.3110870710294005,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.2248,
"step": 1071
},
{
"epoch": 4.3422784810126585,
"grad_norm": 0.285475347700405,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.1933,
"step": 1072
},
{
"epoch": 4.3463291139240505,
"grad_norm": 0.3011375291221876,
"learning_rate": 1.95257270594031e-06,
"loss": 0.2434,
"step": 1073
},
{
"epoch": 4.3503797468354435,
"grad_norm": 0.3032881065212222,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.2307,
"step": 1074
},
{
"epoch": 4.3544303797468356,
"grad_norm": 0.2916660198733973,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.1955,
"step": 1075
},
{
"epoch": 4.358481012658228,
"grad_norm": 0.3065397242763573,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.2387,
"step": 1076
},
{
"epoch": 4.362531645569621,
"grad_norm": 0.29712809111674354,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.2373,
"step": 1077
},
{
"epoch": 4.366582278481013,
"grad_norm": 0.297007266260377,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.2153,
"step": 1078
},
{
"epoch": 4.370632911392405,
"grad_norm": 0.29967416538768077,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.2294,
"step": 1079
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.2831200217417031,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.2128,
"step": 1080
},
{
"epoch": 4.37873417721519,
"grad_norm": 0.29791491779723184,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.2506,
"step": 1081
},
{
"epoch": 4.382784810126582,
"grad_norm": 0.3081307376472192,
"learning_rate": 1.738336950434061e-06,
"loss": 0.2366,
"step": 1082
},
{
"epoch": 4.386835443037975,
"grad_norm": 0.26946109070262464,
"learning_rate": 1.715265841375957e-06,
"loss": 0.1821,
"step": 1083
},
{
"epoch": 4.390886075949367,
"grad_norm": 0.3128231837600124,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.2578,
"step": 1084
},
{
"epoch": 4.39493670886076,
"grad_norm": 0.2723480890759205,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.2024,
"step": 1085
},
{
"epoch": 4.398987341772152,
"grad_norm": 0.3049345302727996,
"learning_rate": 1.646936827927441e-06,
"loss": 0.2551,
"step": 1086
},
{
"epoch": 4.403037974683544,
"grad_norm": 0.2945905517893415,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.2393,
"step": 1087
},
{
"epoch": 4.407088607594937,
"grad_norm": 0.28723600189125476,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.2249,
"step": 1088
},
{
"epoch": 4.411139240506329,
"grad_norm": 0.27580194742122993,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.2121,
"step": 1089
},
{
"epoch": 4.415189873417722,
"grad_norm": 0.2930840584010334,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.2232,
"step": 1090
},
{
"epoch": 4.419240506329114,
"grad_norm": 0.2937957264256722,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.2123,
"step": 1091
},
{
"epoch": 4.423291139240506,
"grad_norm": 0.3025747067938916,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.2344,
"step": 1092
},
{
"epoch": 4.427341772151899,
"grad_norm": 0.3018384383395,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.2498,
"step": 1093
},
{
"epoch": 4.431392405063291,
"grad_norm": 0.2893906214381865,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.2143,
"step": 1094
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.30952813323990175,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.2472,
"step": 1095
},
{
"epoch": 4.439493670886076,
"grad_norm": 0.2729645875900364,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.1976,
"step": 1096
},
{
"epoch": 4.443544303797468,
"grad_norm": 0.2962231521384023,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.2345,
"step": 1097
},
{
"epoch": 4.44759493670886,
"grad_norm": 0.3429287900540949,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.2633,
"step": 1098
},
{
"epoch": 4.451645569620253,
"grad_norm": 0.28915331401779737,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.224,
"step": 1099
},
{
"epoch": 4.455696202531645,
"grad_norm": 0.29280768710978555,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.2329,
"step": 1100
},
{
"epoch": 4.459746835443038,
"grad_norm": 0.2780559133851651,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.229,
"step": 1101
},
{
"epoch": 4.46379746835443,
"grad_norm": 0.2833727433919842,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.2303,
"step": 1102
},
{
"epoch": 4.467848101265822,
"grad_norm": 0.3042956847732379,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.2508,
"step": 1103
},
{
"epoch": 4.471898734177215,
"grad_norm": 0.2876220038795367,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.2303,
"step": 1104
},
{
"epoch": 4.4759493670886075,
"grad_norm": 0.2864707210136753,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.2329,
"step": 1105
},
{
"epoch": 4.48,
"grad_norm": 0.289840248057396,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.2338,
"step": 1106
},
{
"epoch": 4.4840506329113925,
"grad_norm": 0.2802851201999864,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.2187,
"step": 1107
},
{
"epoch": 4.488101265822785,
"grad_norm": 0.2907769984936369,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.2119,
"step": 1108
},
{
"epoch": 4.4921518987341775,
"grad_norm": 0.2927993461275801,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.2054,
"step": 1109
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.31034268879011406,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.2506,
"step": 1110
},
{
"epoch": 4.500253164556962,
"grad_norm": 0.2973469592645843,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.2273,
"step": 1111
},
{
"epoch": 4.504303797468355,
"grad_norm": 0.27573334735620125,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.2253,
"step": 1112
},
{
"epoch": 4.508354430379747,
"grad_norm": 0.28955387287272605,
"learning_rate": 1.092402023876933e-06,
"loss": 0.2317,
"step": 1113
},
{
"epoch": 4.512405063291139,
"grad_norm": 0.27728335133493176,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.2108,
"step": 1114
},
{
"epoch": 4.516455696202532,
"grad_norm": 0.304649103660127,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.2391,
"step": 1115
},
{
"epoch": 4.520506329113924,
"grad_norm": 0.286963013692076,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.2173,
"step": 1116
},
{
"epoch": 4.524556962025317,
"grad_norm": 0.2850872422729242,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.2186,
"step": 1117
},
{
"epoch": 4.528607594936709,
"grad_norm": 0.2728813193767907,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.2277,
"step": 1118
},
{
"epoch": 4.532658227848101,
"grad_norm": 0.29686640449264107,
"learning_rate": 9.841383857091947e-07,
"loss": 0.2415,
"step": 1119
},
{
"epoch": 4.536708860759494,
"grad_norm": 0.2825592047738526,
"learning_rate": 9.666296380321616e-07,
"loss": 0.2117,
"step": 1120
},
{
"epoch": 4.540759493670886,
"grad_norm": 0.2950835357121032,
"learning_rate": 9.492741825034124e-07,
"loss": 0.2498,
"step": 1121
},
{
"epoch": 4.544810126582279,
"grad_norm": 0.278409375019131,
"learning_rate": 9.320721589013892e-07,
"loss": 0.2199,
"step": 1122
},
{
"epoch": 4.548860759493671,
"grad_norm": 0.26422115515294603,
"learning_rate": 9.150237057688339e-07,
"loss": 0.194,
"step": 1123
},
{
"epoch": 4.552911392405063,
"grad_norm": 0.3154095837571298,
"learning_rate": 8.981289604116328e-07,
"loss": 0.2492,
"step": 1124
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.2773206365235178,
"learning_rate": 8.813880588977542e-07,
"loss": 0.2176,
"step": 1125
},
{
"epoch": 4.561012658227848,
"grad_norm": 0.2921778331243518,
"learning_rate": 8.648011360561126e-07,
"loss": 0.2466,
"step": 1126
},
{
"epoch": 4.56506329113924,
"grad_norm": 0.27642288209111693,
"learning_rate": 8.483683254755037e-07,
"loss": 0.2126,
"step": 1127
},
{
"epoch": 4.569113924050633,
"grad_norm": 0.2817120522711001,
"learning_rate": 8.320897595035227e-07,
"loss": 0.2135,
"step": 1128
},
{
"epoch": 4.573164556962025,
"grad_norm": 0.29629283803490386,
"learning_rate": 8.159655692455093e-07,
"loss": 0.2539,
"step": 1129
},
{
"epoch": 4.577215189873417,
"grad_norm": 0.27350296504921845,
"learning_rate": 7.999958845634648e-07,
"loss": 0.2225,
"step": 1130
},
{
"epoch": 4.58126582278481,
"grad_norm": 0.29497914858725144,
"learning_rate": 7.841808340750478e-07,
"loss": 0.2351,
"step": 1131
},
{
"epoch": 4.585316455696202,
"grad_norm": 0.3003583344471951,
"learning_rate": 7.685205451524869e-07,
"loss": 0.2616,
"step": 1132
},
{
"epoch": 4.589367088607595,
"grad_norm": 0.25712300212544575,
"learning_rate": 7.530151439216027e-07,
"loss": 0.2002,
"step": 1133
},
{
"epoch": 4.593417721518987,
"grad_norm": 0.2759202008132098,
"learning_rate": 7.376647552607675e-07,
"loss": 0.2291,
"step": 1134
},
{
"epoch": 4.597468354430379,
"grad_norm": 0.2865850663258399,
"learning_rate": 7.224695027998963e-07,
"loss": 0.2313,
"step": 1135
},
{
"epoch": 4.601518987341772,
"grad_norm": 0.28508706267548645,
"learning_rate": 7.07429508919466e-07,
"loss": 0.2254,
"step": 1136
},
{
"epoch": 4.605569620253164,
"grad_norm": 0.29456088337653913,
"learning_rate": 6.925448947495206e-07,
"loss": 0.2421,
"step": 1137
},
{
"epoch": 4.609620253164557,
"grad_norm": 0.286383232572337,
"learning_rate": 6.778157801686936e-07,
"loss": 0.226,
"step": 1138
},
{
"epoch": 4.613670886075949,
"grad_norm": 0.2874765053362072,
"learning_rate": 6.632422838032515e-07,
"loss": 0.2133,
"step": 1139
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.2894081465801748,
"learning_rate": 6.488245230261281e-07,
"loss": 0.2595,
"step": 1140
},
{
"epoch": 4.6217721518987345,
"grad_norm": 0.26426955949335845,
"learning_rate": 6.345626139559868e-07,
"loss": 0.2079,
"step": 1141
},
{
"epoch": 4.6258227848101265,
"grad_norm": 0.280913284320995,
"learning_rate": 6.204566714562866e-07,
"loss": 0.2207,
"step": 1142
},
{
"epoch": 4.629873417721519,
"grad_norm": 0.2812350580615992,
"learning_rate": 6.06506809134344e-07,
"loss": 0.2264,
"step": 1143
},
{
"epoch": 4.633924050632912,
"grad_norm": 0.29292273371919403,
"learning_rate": 5.927131393404373e-07,
"loss": 0.2407,
"step": 1144
},
{
"epoch": 4.637974683544304,
"grad_norm": 0.288852006508432,
"learning_rate": 5.790757731668817e-07,
"loss": 0.2298,
"step": 1145
},
{
"epoch": 4.642025316455696,
"grad_norm": 0.2747119874820988,
"learning_rate": 5.655948204471507e-07,
"loss": 0.2161,
"step": 1146
},
{
"epoch": 4.646075949367089,
"grad_norm": 0.3100050169487352,
"learning_rate": 5.522703897549875e-07,
"loss": 0.2338,
"step": 1147
},
{
"epoch": 4.650126582278481,
"grad_norm": 0.2807273885338933,
"learning_rate": 5.391025884035239e-07,
"loss": 0.2259,
"step": 1148
},
{
"epoch": 4.654177215189874,
"grad_norm": 0.32130818450537774,
"learning_rate": 5.260915224444207e-07,
"loss": 0.2591,
"step": 1149
},
{
"epoch": 4.658227848101266,
"grad_norm": 0.2779170961902077,
"learning_rate": 5.132372966670129e-07,
"loss": 0.1942,
"step": 1150
},
{
"epoch": 4.662278481012658,
"grad_norm": 0.313688704145969,
"learning_rate": 5.005400145974704e-07,
"loss": 0.2442,
"step": 1151
},
{
"epoch": 4.666329113924051,
"grad_norm": 0.277385928267181,
"learning_rate": 4.879997784979562e-07,
"loss": 0.2169,
"step": 1152
},
{
"epoch": 4.670379746835443,
"grad_norm": 0.2794520840117712,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.2208,
"step": 1153
},
{
"epoch": 4.674430379746836,
"grad_norm": 0.28049509993730753,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.1993,
"step": 1154
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.2936845877848516,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.2428,
"step": 1155
},
{
"epoch": 4.68253164556962,
"grad_norm": 0.28471848639677594,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.2166,
"step": 1156
},
{
"epoch": 4.686582278481013,
"grad_norm": 0.2968349253641284,
"learning_rate": 4.27657778143431e-07,
"loss": 0.2394,
"step": 1157
},
{
"epoch": 4.690632911392405,
"grad_norm": 0.28907736965089714,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.237,
"step": 1158
},
{
"epoch": 4.694683544303797,
"grad_norm": 0.29719507064787093,
"learning_rate": 4.046237371675177e-07,
"loss": 0.2377,
"step": 1159
},
{
"epoch": 4.69873417721519,
"grad_norm": 0.2908586850588327,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.2209,
"step": 1160
},
{
"epoch": 4.702784810126582,
"grad_norm": 0.30176752318124955,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.2404,
"step": 1161
},
{
"epoch": 4.706835443037974,
"grad_norm": 0.27631278798249503,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.1903,
"step": 1162
},
{
"epoch": 4.710886075949367,
"grad_norm": 0.3049517422495807,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.231,
"step": 1163
},
{
"epoch": 4.714936708860759,
"grad_norm": 0.3040946860199761,
"learning_rate": 3.498020255137813e-07,
"loss": 0.2344,
"step": 1164
},
{
"epoch": 4.718987341772152,
"grad_norm": 0.2816181302652299,
"learning_rate": 3.393121163057811e-07,
"loss": 0.2085,
"step": 1165
},
{
"epoch": 4.723037974683544,
"grad_norm": 0.32162691648490993,
"learning_rate": 3.289805515751399e-07,
"loss": 0.2505,
"step": 1166
},
{
"epoch": 4.727088607594936,
"grad_norm": 0.26911548022388154,
"learning_rate": 3.188074145308573e-07,
"loss": 0.1872,
"step": 1167
},
{
"epoch": 4.731139240506329,
"grad_norm": 0.30577957699777497,
"learning_rate": 3.087927871059804e-07,
"loss": 0.2373,
"step": 1168
},
{
"epoch": 4.735189873417721,
"grad_norm": 0.2903514592817847,
"learning_rate": 2.989367499569418e-07,
"loss": 0.2092,
"step": 1169
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.31031630888533834,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.2438,
"step": 1170
},
{
"epoch": 4.743291139240506,
"grad_norm": 0.2863546034469591,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.2059,
"step": 1171
},
{
"epoch": 4.7473417721518985,
"grad_norm": 0.29143641617190974,
"learning_rate": 2.703209675663887e-07,
"loss": 0.2236,
"step": 1172
},
{
"epoch": 4.751392405063291,
"grad_norm": 0.2809579713571302,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.2273,
"step": 1173
},
{
"epoch": 4.7554430379746835,
"grad_norm": 0.2790047823578926,
"learning_rate": 2.520381518804471e-07,
"loss": 0.2106,
"step": 1174
},
{
"epoch": 4.759493670886076,
"grad_norm": 0.2987887132782839,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.248,
"step": 1175
},
{
"epoch": 4.7635443037974685,
"grad_norm": 0.2952637357501775,
"learning_rate": 2.343915243930317e-07,
"loss": 0.2255,
"step": 1176
},
{
"epoch": 4.767594936708861,
"grad_norm": 0.26846097633177407,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.2126,
"step": 1177
},
{
"epoch": 4.7716455696202535,
"grad_norm": 0.27868897995038916,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.2181,
"step": 1178
},
{
"epoch": 4.775696202531646,
"grad_norm": 0.2901411077202586,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.2242,
"step": 1179
},
{
"epoch": 4.779746835443038,
"grad_norm": 0.29089165474589385,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.2332,
"step": 1180
},
{
"epoch": 4.783797468354431,
"grad_norm": 0.2837786693822112,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.2371,
"step": 1181
},
{
"epoch": 4.787848101265823,
"grad_norm": 0.2849551035102473,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.2147,
"step": 1182
},
{
"epoch": 4.791898734177215,
"grad_norm": 0.2968686211762889,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.2554,
"step": 1183
},
{
"epoch": 4.795949367088608,
"grad_norm": 0.29228047154367637,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.2371,
"step": 1184
},
{
"epoch": 4.8,
"grad_norm": 0.27762598155781393,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.2146,
"step": 1185
},
{
"epoch": 4.804050632911393,
"grad_norm": 0.2904285779419282,
"learning_rate": 1.557203888516745e-07,
"loss": 0.21,
"step": 1186
},
{
"epoch": 4.808101265822785,
"grad_norm": 0.28803368058074114,
"learning_rate": 1.487313098987131e-07,
"loss": 0.2453,
"step": 1187
},
{
"epoch": 4.812151898734177,
"grad_norm": 0.27388801708606764,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.2253,
"step": 1188
},
{
"epoch": 4.81620253164557,
"grad_norm": 0.27319381343680205,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.213,
"step": 1189
},
{
"epoch": 4.820253164556962,
"grad_norm": 0.2865789011721331,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.2223,
"step": 1190
},
{
"epoch": 4.824303797468355,
"grad_norm": 0.29777218628054086,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.2415,
"step": 1191
},
{
"epoch": 4.828354430379747,
"grad_norm": 0.2838636763238733,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.2335,
"step": 1192
},
{
"epoch": 4.832405063291139,
"grad_norm": 0.29875889084100865,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.2445,
"step": 1193
},
{
"epoch": 4.836455696202532,
"grad_norm": 0.3022404876727881,
"learning_rate": 1.042873700093061e-07,
"loss": 0.2406,
"step": 1194
},
{
"epoch": 4.840506329113924,
"grad_norm": 0.2733042107225577,
"learning_rate": 9.857879404556291e-08,
"loss": 0.1946,
"step": 1195
},
{
"epoch": 4.844556962025316,
"grad_norm": 0.3034938609657879,
"learning_rate": 9.303050139225722e-08,
"loss": 0.2655,
"step": 1196
},
{
"epoch": 4.848607594936709,
"grad_norm": 0.2637550632962007,
"learning_rate": 8.76425367345779e-08,
"loss": 0.2206,
"step": 1197
},
{
"epoch": 4.852658227848101,
"grad_norm": 0.377586840386621,
"learning_rate": 8.241494346644897e-08,
"loss": 0.2346,
"step": 1198
},
{
"epoch": 4.856708860759493,
"grad_norm": 0.27244621903507643,
"learning_rate": 7.734776369019204e-08,
"loss": 0.2128,
"step": 1199
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.3086856478014234,
"learning_rate": 7.244103821617332e-08,
"loss": 0.2484,
"step": 1200
},
{
"epoch": 4.864810126582278,
"grad_norm": 0.27546189648944186,
"learning_rate": 6.769480656248606e-08,
"loss": 0.2167,
"step": 1201
},
{
"epoch": 4.868860759493671,
"grad_norm": 0.26311147411135916,
"learning_rate": 6.310910695462635e-08,
"loss": 0.2115,
"step": 1202
},
{
"epoch": 4.872911392405063,
"grad_norm": 0.28902696634594194,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.2378,
"step": 1203
},
{
"epoch": 4.876962025316455,
"grad_norm": 0.2788992298638475,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.2181,
"step": 1204
},
{
"epoch": 4.881012658227848,
"grad_norm": 0.2758392492782826,
"learning_rate": 5.031556326567488e-08,
"loss": 0.2404,
"step": 1205
},
{
"epoch": 4.88506329113924,
"grad_norm": 0.26839955558163087,
"learning_rate": 4.637234823364312e-08,
"loss": 0.2224,
"step": 1206
},
{
"epoch": 4.889113924050633,
"grad_norm": 0.28117197738971966,
"learning_rate": 4.258983697558838e-08,
"loss": 0.2119,
"step": 1207
},
{
"epoch": 4.8931645569620255,
"grad_norm": 0.27792752865095083,
"learning_rate": 3.896805995533548e-08,
"loss": 0.2328,
"step": 1208
},
{
"epoch": 4.8972151898734175,
"grad_norm": 0.26828026989849285,
"learning_rate": 3.550704634218028e-08,
"loss": 0.206,
"step": 1209
},
{
"epoch": 4.9012658227848105,
"grad_norm": 0.2688059535516326,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.2137,
"step": 1210
},
{
"epoch": 4.905316455696203,
"grad_norm": 0.29265415405601714,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.2426,
"step": 1211
},
{
"epoch": 4.909367088607595,
"grad_norm": 0.27660575741950527,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.2278,
"step": 1212
},
{
"epoch": 4.913417721518988,
"grad_norm": 0.26711303561346983,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.2425,
"step": 1213
},
{
"epoch": 4.91746835443038,
"grad_norm": 0.277177137265781,
"learning_rate": 2.061435966221881e-08,
"loss": 0.2362,
"step": 1214
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.2770561961634842,
"learning_rate": 1.811846652477245e-08,
"loss": 0.22,
"step": 1215
},
{
"epoch": 4.925569620253165,
"grad_norm": 0.2667376329225243,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.2204,
"step": 1216
},
{
"epoch": 4.929620253164557,
"grad_norm": 0.28836009602817714,
"learning_rate": 1.360949303591097e-08,
"loss": 0.2513,
"step": 1217
},
{
"epoch": 4.93367088607595,
"grad_norm": 0.2766466703241091,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.2271,
"step": 1218
},
{
"epoch": 4.937721518987342,
"grad_norm": 0.27799828982610497,
"learning_rate": 9.744388816668172e-09,
"loss": 0.227,
"step": 1219
},
{
"epoch": 4.941772151898734,
"grad_norm": 0.2793381696369662,
"learning_rate": 8.05332740472009e-09,
"loss": 0.2318,
"step": 1220
},
{
"epoch": 4.945822784810127,
"grad_norm": 0.272388288255058,
"learning_rate": 6.523278382872811e-09,
"loss": 0.2137,
"step": 1221
},
{
"epoch": 4.949873417721519,
"grad_norm": 0.2790317633136994,
"learning_rate": 5.15425407393133e-09,
"loss": 0.23,
"step": 1222
},
{
"epoch": 4.953924050632912,
"grad_norm": 0.2834265462110359,
"learning_rate": 3.94626550383137e-09,
"loss": 0.2333,
"step": 1223
},
{
"epoch": 4.957974683544304,
"grad_norm": 0.2840150313451397,
"learning_rate": 2.899322401546112e-09,
"loss": 0.2186,
"step": 1224
},
{
"epoch": 4.962025316455696,
"grad_norm": 0.3103902968566873,
"learning_rate": 2.013433199010706e-09,
"loss": 0.2434,
"step": 1225
},
{
"epoch": 4.966075949367089,
"grad_norm": 0.2577952453762563,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.1935,
"step": 1226
},
{
"epoch": 4.970126582278481,
"grad_norm": 0.2873630501402775,
"learning_rate": 7.248437353468695e-10,
"loss": 0.2418,
"step": 1227
},
{
"epoch": 4.974177215189873,
"grad_norm": 0.2746073408238836,
"learning_rate": 3.221538523412449e-10,
"loss": 0.2011,
"step": 1228
},
{
"epoch": 4.978227848101266,
"grad_norm": 0.3027695572968247,
"learning_rate": 8.053862524670663e-11,
"loss": 0.2431,
"step": 1229
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.29000224099012994,
"learning_rate": 0.0,
"loss": 0.2463,
"step": 1230
},
{
"epoch": 4.982278481012658,
"step": 1230,
"total_flos": 3.008404639465865e+18,
"train_loss": 0.41607419755158387,
"train_runtime": 78317.3531,
"train_samples_per_second": 2.017,
"train_steps_per_second": 0.016
}
],
"logging_steps": 1.0,
"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": 3.008404639465865e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}