a1_math_big_math / trainer_state.json
ryanmarten's picture
End of training
8518f5f verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.988354430379747,
"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": 7.5356330688640805,
"learning_rate": 3.2520325203252037e-07,
"loss": 0.8354,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 7.201400965875826,
"learning_rate": 6.504065040650407e-07,
"loss": 0.8013,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 7.253727824970738,
"learning_rate": 9.75609756097561e-07,
"loss": 0.8069,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 7.297382595114797,
"learning_rate": 1.3008130081300815e-06,
"loss": 0.8186,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 6.732986928496777,
"learning_rate": 1.6260162601626018e-06,
"loss": 0.7895,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 6.805880425160705,
"learning_rate": 1.951219512195122e-06,
"loss": 0.8053,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 5.6585903421896315,
"learning_rate": 2.2764227642276426e-06,
"loss": 0.7761,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 5.04441241490745,
"learning_rate": 2.601626016260163e-06,
"loss": 0.7335,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 3.1689218354354782,
"learning_rate": 2.926829268292683e-06,
"loss": 0.6811,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 2.8349449077696147,
"learning_rate": 3.2520325203252037e-06,
"loss": 0.677,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 2.6532055439201834,
"learning_rate": 3.577235772357724e-06,
"loss": 0.6607,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 3.3627074652991893,
"learning_rate": 3.902439024390244e-06,
"loss": 0.6419,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 3.242863449519781,
"learning_rate": 4.227642276422765e-06,
"loss": 0.6249,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 3.2427406595860524,
"learning_rate": 4.552845528455285e-06,
"loss": 0.6061,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 2.7785357778664985,
"learning_rate": 4.8780487804878055e-06,
"loss": 0.5691,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 2.6776909220101115,
"learning_rate": 5.203252032520326e-06,
"loss": 0.5395,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 2.514958644517776,
"learning_rate": 5.528455284552846e-06,
"loss": 0.5385,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.0074016680944746,
"learning_rate": 5.853658536585366e-06,
"loss": 0.5289,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 1.806358079501236,
"learning_rate": 6.178861788617887e-06,
"loss": 0.5118,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.5256644883732995,
"learning_rate": 6.504065040650407e-06,
"loss": 0.5137,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.526669889602467,
"learning_rate": 6.829268292682928e-06,
"loss": 0.492,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.6399240321458548,
"learning_rate": 7.154471544715448e-06,
"loss": 0.4955,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.5492840112591806,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.4866,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 1.030880249581923,
"learning_rate": 7.804878048780489e-06,
"loss": 0.4758,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 0.8341903320917818,
"learning_rate": 8.130081300813009e-06,
"loss": 0.4563,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 1.1293555599517782,
"learning_rate": 8.45528455284553e-06,
"loss": 0.4539,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 1.0046105633880038,
"learning_rate": 8.78048780487805e-06,
"loss": 0.4318,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 0.7354882282705718,
"learning_rate": 9.10569105691057e-06,
"loss": 0.4559,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 0.682230588469553,
"learning_rate": 9.43089430894309e-06,
"loss": 0.4322,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 0.8063326092259202,
"learning_rate": 9.756097560975611e-06,
"loss": 0.4386,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 0.844237915696608,
"learning_rate": 1.008130081300813e-05,
"loss": 0.4527,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.7123647977061031,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.4077,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.6032281596714956,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.434,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.7779852157894015,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.45,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.6836451662078461,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.4206,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.5689778011201945,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.4345,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.6637654337227472,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.4297,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.6123138391314188,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.4182,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.5941466930229284,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.4205,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.5332489464120281,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.4094,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.6152995532991368,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.417,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.5381264093023036,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.4263,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.5369689631813412,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.3961,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.635257975085662,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.409,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.6454681358588018,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.4239,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.5491271787139909,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.412,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.5317354372604716,
"learning_rate": 1.528455284552846e-05,
"loss": 0.4023,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.5124064925407114,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.4014,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.5546819556573757,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.4028,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.5803666973871974,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.4156,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.585352877937866,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.3798,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.6144533428977065,
"learning_rate": 1.691056910569106e-05,
"loss": 0.4048,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.5584971642287446,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.3921,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.6486244966653673,
"learning_rate": 1.75609756097561e-05,
"loss": 0.3943,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.6694461578734215,
"learning_rate": 1.788617886178862e-05,
"loss": 0.3983,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.599484262280129,
"learning_rate": 1.821138211382114e-05,
"loss": 0.4028,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.6762737798732859,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.3873,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.5904465869046148,
"learning_rate": 1.886178861788618e-05,
"loss": 0.4041,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.5756089601667397,
"learning_rate": 1.91869918699187e-05,
"loss": 0.3877,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.5334684582348354,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.3651,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.7582990348936732,
"learning_rate": 1.983739837398374e-05,
"loss": 0.4039,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.5835543352147674,
"learning_rate": 2.016260162601626e-05,
"loss": 0.3829,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.5050162103531988,
"learning_rate": 2.048780487804878e-05,
"loss": 0.3673,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.5708805448041003,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.3937,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.5473477844666208,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.3962,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.5767772579981055,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.3782,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.6721211856703776,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.3983,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 0.5109692195980667,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.3763,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 0.6471369381350736,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.3855,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 0.669001857155192,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.3827,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 0.6342679659107424,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.3625,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.6498058237965413,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.3771,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 0.5930372367079505,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.3915,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.7589385792900072,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.405,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.5495164135719603,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.3643,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 0.6866484852152177,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.3867,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.6207235473314903,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.3551,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 0.6098777444783098,
"learning_rate": 2.536585365853659e-05,
"loss": 0.3605,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 0.6444780911658265,
"learning_rate": 2.569105691056911e-05,
"loss": 0.3871,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 0.6608120096281637,
"learning_rate": 2.601626016260163e-05,
"loss": 0.3678,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 0.7580397714596296,
"learning_rate": 2.634146341463415e-05,
"loss": 0.3719,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 0.6671122125361024,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.3854,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 0.7970632024787427,
"learning_rate": 2.699186991869919e-05,
"loss": 0.3957,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 0.7366992107025959,
"learning_rate": 2.731707317073171e-05,
"loss": 0.3785,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 0.6288543215052098,
"learning_rate": 2.764227642276423e-05,
"loss": 0.3878,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 0.5834126244424173,
"learning_rate": 2.796747967479675e-05,
"loss": 0.3835,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 0.6401497417368652,
"learning_rate": 2.829268292682927e-05,
"loss": 0.3637,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 0.6070090270029983,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.4018,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 0.6985817141145747,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.3769,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 0.7545525691964186,
"learning_rate": 2.926829268292683e-05,
"loss": 0.3784,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 0.7083359144627455,
"learning_rate": 2.959349593495935e-05,
"loss": 0.3874,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 0.7332172174776229,
"learning_rate": 2.991869918699187e-05,
"loss": 0.3894,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 0.6987275879718177,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.3728,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 0.8102953983472229,
"learning_rate": 3.056910569105692e-05,
"loss": 0.3889,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 0.6925964413515279,
"learning_rate": 3.089430894308943e-05,
"loss": 0.3622,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 0.8327587220436943,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.3616,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 0.6173938234993211,
"learning_rate": 3.154471544715447e-05,
"loss": 0.3799,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 0.7528270432871829,
"learning_rate": 3.186991869918699e-05,
"loss": 0.3841,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 0.6740374592797312,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.3903,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.6950148937524135,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.3792,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 0.5706932654512163,
"learning_rate": 3.284552845528456e-05,
"loss": 0.3674,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 0.6656191686364515,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.3813,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 0.6300300781564973,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.4064,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 0.5740962561835604,
"learning_rate": 3.382113821138212e-05,
"loss": 0.3716,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 0.64662348721664,
"learning_rate": 3.414634146341463e-05,
"loss": 0.3595,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 0.6117574144328706,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.3756,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 0.678920642516966,
"learning_rate": 3.479674796747968e-05,
"loss": 0.3852,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 0.7001197281636697,
"learning_rate": 3.51219512195122e-05,
"loss": 0.3853,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 0.5476111672578697,
"learning_rate": 3.544715447154472e-05,
"loss": 0.373,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 0.5415501111966172,
"learning_rate": 3.577235772357724e-05,
"loss": 0.3895,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 0.644495905517219,
"learning_rate": 3.609756097560976e-05,
"loss": 0.4011,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 0.5911997400031409,
"learning_rate": 3.642276422764228e-05,
"loss": 0.3701,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 0.5728006150439562,
"learning_rate": 3.67479674796748e-05,
"loss": 0.3726,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 0.7587559998025868,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.3845,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 0.5861027603159441,
"learning_rate": 3.739837398373984e-05,
"loss": 0.3722,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 0.5694304139413606,
"learning_rate": 3.772357723577236e-05,
"loss": 0.3751,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 0.59322414682224,
"learning_rate": 3.804878048780488e-05,
"loss": 0.3662,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 0.5410941699127024,
"learning_rate": 3.83739837398374e-05,
"loss": 0.3737,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 0.5579219084076499,
"learning_rate": 3.869918699186992e-05,
"loss": 0.3821,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 0.7711014177119695,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.381,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 0.7000728877474965,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.3753,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 0.6254063089804005,
"learning_rate": 3.967479674796748e-05,
"loss": 0.3839,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 0.7502205396104371,
"learning_rate": 4e-05,
"loss": 0.3943,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 0.6581002120716252,
"learning_rate": 3.999991946137476e-05,
"loss": 0.3818,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.9040898811526893,
"learning_rate": 3.999967784614766e-05,
"loss": 0.377,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 0.9171063155702521,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.3598,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 0.7843732736842247,
"learning_rate": 3.999871139496895e-05,
"loss": 0.4028,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 0.6501685575774409,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.3575,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 0.8522028634369185,
"learning_rate": 3.999710067759846e-05,
"loss": 0.3469,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 0.5616688852511569,
"learning_rate": 3.999605373449617e-05,
"loss": 0.3872,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.9561945077435448,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.368,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 0.6385517624805415,
"learning_rate": 3.999347672161713e-05,
"loss": 0.3663,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 0.7184069530125493,
"learning_rate": 3.999194667259528e-05,
"loss": 0.3643,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 0.7262933879747613,
"learning_rate": 3.999025561118334e-05,
"loss": 0.3678,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 0.5802934418352055,
"learning_rate": 3.998840355100086e-05,
"loss": 0.3771,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 0.7777943504474174,
"learning_rate": 3.998639050696409e-05,
"loss": 0.3773,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 0.5682540335595604,
"learning_rate": 3.998421649528582e-05,
"loss": 0.3667,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 0.7492509194204291,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.3586,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.587983619214593,
"learning_rate": 3.997938564033779e-05,
"loss": 0.3799,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 0.6460470289974566,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.3779,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.5631370534418936,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.3737,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 0.6366617826052812,
"learning_rate": 3.997093258045973e-05,
"loss": 0.3739,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.5840560378759898,
"learning_rate": 3.996779317598936e-05,
"loss": 0.3696,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 0.5901310568922997,
"learning_rate": 3.996449295365782e-05,
"loss": 0.377,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.6230659173460267,
"learning_rate": 3.996103194004467e-05,
"loss": 0.3713,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.6286624944187134,
"learning_rate": 3.995741016302441e-05,
"loss": 0.3581,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 0.6474621671035757,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.3643,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 0.5863596203358714,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.376,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 0.6418662039481762,
"learning_rate": 3.994558054968643e-05,
"loss": 0.3843,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.5803195318307623,
"learning_rate": 3.994131602367481e-05,
"loss": 0.3768,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 0.6048259633916582,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.3633,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.5202464741175635,
"learning_rate": 3.993230519343752e-05,
"loss": 0.3664,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 0.5764148202431616,
"learning_rate": 3.992755896178383e-05,
"loss": 0.3924,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.5180507724826,
"learning_rate": 3.992265223630981e-05,
"loss": 0.3611,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 0.4919198326716956,
"learning_rate": 3.991758505653355e-05,
"loss": 0.3523,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 0.5123942734225929,
"learning_rate": 3.991235746326543e-05,
"loss": 0.3692,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.587010743251933,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.383,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 0.49529328867434325,
"learning_rate": 3.990142120595444e-05,
"loss": 0.3701,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 0.5453494262871824,
"learning_rate": 3.98957126299907e-05,
"loss": 0.3785,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 0.47036458942588805,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.3736,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 0.552068800657062,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.3783,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 0.4595435692151219,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.3705,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 0.5556330489738479,
"learning_rate": 3.98712764319087e-05,
"loss": 0.3906,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.47095922752615027,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.3607,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 0.4801546305449945,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.3816,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.46229735086404355,
"learning_rate": 3.985126869010129e-05,
"loss": 0.358,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 0.4713592585019107,
"learning_rate": 3.984427961114833e-05,
"loss": 0.3576,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 0.5227027345945428,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.3843,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 0.5476256217538504,
"learning_rate": 3.982982204151901e-05,
"loss": 0.3759,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.4155539654245211,
"learning_rate": 3.982235366728193e-05,
"loss": 0.3535,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.5495085406493314,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.3604,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 0.42801950483840934,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.3563,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.5339338987309228,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.36,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.50649009582474,
"learning_rate": 3.979088432599627e-05,
"loss": 0.3823,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.4788371551876081,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.3569,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 0.47432108736524436,
"learning_rate": 3.977419304032111e-05,
"loss": 0.369,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 0.4745544551429114,
"learning_rate": 3.976560847560697e-05,
"loss": 0.3502,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 0.45975597630177323,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.3599,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 0.4579240242442755,
"learning_rate": 3.974796184811956e-05,
"loss": 0.3501,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.5426034503995587,
"learning_rate": 3.973889992746979e-05,
"loss": 0.3688,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.5154770610426235,
"learning_rate": 3.972967903243361e-05,
"loss": 0.363,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.5498754044973335,
"learning_rate": 3.972029923727486e-05,
"loss": 0.3752,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.4997604330240588,
"learning_rate": 3.971076061753709e-05,
"loss": 0.3478,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.5178019263413225,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.3706,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.5158071614452495,
"learning_rate": 3.969120721289402e-05,
"loss": 0.3714,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.5910117825960008,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.3717,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.45691157632662743,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.3719,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.5208133234800578,
"learning_rate": 3.966068788369422e-05,
"loss": 0.358,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.44895216678643307,
"learning_rate": 3.965019797448622e-05,
"loss": 0.3524,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.4644916544721236,
"learning_rate": 3.963954980528515e-05,
"loss": 0.3509,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.5013657055678457,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.3503,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.48518504652876304,
"learning_rate": 3.961777903121329e-05,
"loss": 0.3649,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.4796529821435585,
"learning_rate": 3.960665660168131e-05,
"loss": 0.3595,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.5328854757928938,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.3618,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.4858951249388405,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.3613,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 0.43752033885451547,
"learning_rate": 3.957234222185657e-05,
"loss": 0.3763,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.4653787622238782,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.3541,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 0.4443446041874942,
"learning_rate": 3.954867765033605e-05,
"loss": 0.3748,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.4422863822613439,
"learning_rate": 3.953660915306728e-05,
"loss": 0.3625,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.48137081840574636,
"learning_rate": 3.952438331063419e-05,
"loss": 0.3604,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.48504543343760176,
"learning_rate": 3.951200022150205e-05,
"loss": 0.3546,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.47707477304944806,
"learning_rate": 3.949945998540253e-05,
"loss": 0.3575,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.4819247400407336,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.3614,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.5133140014723326,
"learning_rate": 3.947390847755559e-05,
"loss": 0.3494,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.5186529032867118,
"learning_rate": 3.946089741159648e-05,
"loss": 0.3888,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.4655057307775711,
"learning_rate": 3.944772961024501e-05,
"loss": 0.3748,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.5551675091994906,
"learning_rate": 3.943440517955285e-05,
"loss": 0.3627,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.42158027177522456,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.3589,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.4958091725595573,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.3705,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.557683298557716,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.391,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.44746433842510114,
"learning_rate": 3.937954332854371e-05,
"loss": 0.3708,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.5738230895559812,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.367,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.42152010717551414,
"learning_rate": 3.935117547697387e-05,
"loss": 0.3569,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.5071607171468698,
"learning_rate": 3.933675771619675e-05,
"loss": 0.366,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.4595914206634382,
"learning_rate": 3.932218421983131e-05,
"loss": 0.3725,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.4652316781328293,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.3611,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.4987521447109579,
"learning_rate": 3.929257049108054e-05,
"loss": 0.3553,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.4511408775948645,
"learning_rate": 3.927753049720011e-05,
"loss": 0.3493,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.4574140920133443,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.3624,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.5076223191821818,
"learning_rate": 3.92469848560784e-05,
"loss": 0.3697,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.4058225230210054,
"learning_rate": 3.923147945484751e-05,
"loss": 0.3461,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.5097164949976879,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.378,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.443447079419054,
"learning_rate": 3.920000411543654e-05,
"loss": 0.355,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.45933858845144765,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.3571,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.5530316786026653,
"learning_rate": 3.916791024049648e-05,
"loss": 0.3715,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.5032318030571665,
"learning_rate": 3.91516316745245e-05,
"loss": 0.3432,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.4658587143242956,
"learning_rate": 3.913519886394389e-05,
"loss": 0.3448,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.6026920769491775,
"learning_rate": 3.911861194110225e-05,
"loss": 0.3679,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.43048936719381603,
"learning_rate": 3.910187103958837e-05,
"loss": 0.3515,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.6224978506420007,
"learning_rate": 3.908497629423117e-05,
"loss": 0.3649,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.465094813115337,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.3745,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.5362309405996935,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.352,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.4883389151790975,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.3514,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.5223655981450401,
"learning_rate": 3.901586161429081e-05,
"loss": 0.3805,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.4299611005304316,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.3667,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.45139858089707,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.3739,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.4906806061253534,
"learning_rate": 3.896241703466667e-05,
"loss": 0.3431,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.3650637149783543,
"learning_rate": 3.894429654085585e-05,
"loss": 0.3445,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.5008925669663145,
"learning_rate": 3.892602347228505e-05,
"loss": 0.3659,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.4351809051584318,
"learning_rate": 3.890759797612307e-05,
"loss": 0.3494,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.42680957572095346,
"learning_rate": 3.888902020076632e-05,
"loss": 0.3582,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.4055304192985378,
"learning_rate": 3.887029029583764e-05,
"loss": 0.3501,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.4422797898673829,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.3727,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.40027054759755626,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.3513,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.4549553252777807,
"learning_rate": 3.881318931821972e-05,
"loss": 0.3504,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.43985334836436624,
"learning_rate": 3.879385241571817e-05,
"loss": 0.3586,
"step": 246
},
{
"epoch": 1.0020253164556963,
"grad_norm": 0.48257534238509464,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.3284,
"step": 247
},
{
"epoch": 1.0060759493670886,
"grad_norm": 0.41348849874506727,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.2822,
"step": 248
},
{
"epoch": 1.010126582278481,
"grad_norm": 0.4079789599550021,
"learning_rate": 3.873493415863256e-05,
"loss": 0.2831,
"step": 249
},
{
"epoch": 1.0141772151898734,
"grad_norm": 0.4543889592234004,
"learning_rate": 3.871499275032111e-05,
"loss": 0.2864,
"step": 250
},
{
"epoch": 1.0182278481012659,
"grad_norm": 0.5349793908394375,
"learning_rate": 3.869490061403091e-05,
"loss": 0.2781,
"step": 251
},
{
"epoch": 1.0222784810126582,
"grad_norm": 0.5188209667842602,
"learning_rate": 3.867465791158124e-05,
"loss": 0.2784,
"step": 252
},
{
"epoch": 1.0263291139240507,
"grad_norm": 0.4799554440439683,
"learning_rate": 3.865426480600407e-05,
"loss": 0.2641,
"step": 253
},
{
"epoch": 1.030379746835443,
"grad_norm": 0.4849306986692548,
"learning_rate": 3.863372146154264e-05,
"loss": 0.2938,
"step": 254
},
{
"epoch": 1.0344303797468355,
"grad_norm": 0.44568366667765424,
"learning_rate": 3.861302804365024e-05,
"loss": 0.2839,
"step": 255
},
{
"epoch": 1.0384810126582278,
"grad_norm": 0.48802627730339926,
"learning_rate": 3.85921847189888e-05,
"loss": 0.2862,
"step": 256
},
{
"epoch": 1.0425316455696203,
"grad_norm": 0.5203493982887369,
"learning_rate": 3.85711916554276e-05,
"loss": 0.2716,
"step": 257
},
{
"epoch": 1.0465822784810126,
"grad_norm": 0.4490686627925636,
"learning_rate": 3.85500490220419e-05,
"loss": 0.2815,
"step": 258
},
{
"epoch": 1.0506329113924051,
"grad_norm": 0.496617455601101,
"learning_rate": 3.852875698911154e-05,
"loss": 0.2883,
"step": 259
},
{
"epoch": 1.0546835443037974,
"grad_norm": 0.4989003155394961,
"learning_rate": 3.850731572811963e-05,
"loss": 0.2823,
"step": 260
},
{
"epoch": 1.05873417721519,
"grad_norm": 0.4257426890094703,
"learning_rate": 3.848572541175116e-05,
"loss": 0.2664,
"step": 261
},
{
"epoch": 1.0627848101265822,
"grad_norm": 0.5573403146594285,
"learning_rate": 3.846398621389154e-05,
"loss": 0.2934,
"step": 262
},
{
"epoch": 1.0668354430379747,
"grad_norm": 0.43144726981060955,
"learning_rate": 3.84420983096253e-05,
"loss": 0.2811,
"step": 263
},
{
"epoch": 1.070886075949367,
"grad_norm": 0.5216254168508768,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.2824,
"step": 264
},
{
"epoch": 1.0749367088607595,
"grad_norm": 0.4543516391411195,
"learning_rate": 3.839787708819787e-05,
"loss": 0.2862,
"step": 265
},
{
"epoch": 1.0789873417721518,
"grad_norm": 0.47713358919059207,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.2801,
"step": 266
},
{
"epoch": 1.0830379746835443,
"grad_norm": 0.4075804903169975,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.2801,
"step": 267
},
{
"epoch": 1.0870886075949366,
"grad_norm": 0.4177950684483678,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.2739,
"step": 268
},
{
"epoch": 1.0911392405063292,
"grad_norm": 0.4190715034505677,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.2984,
"step": 269
},
{
"epoch": 1.0951898734177214,
"grad_norm": 0.4007971561978126,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.2923,
"step": 270
},
{
"epoch": 1.099240506329114,
"grad_norm": 0.4416681943981449,
"learning_rate": 3.826166304956594e-05,
"loss": 0.2741,
"step": 271
},
{
"epoch": 1.1032911392405063,
"grad_norm": 0.447167925005286,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.2916,
"step": 272
},
{
"epoch": 1.1073417721518988,
"grad_norm": 0.42005039975795455,
"learning_rate": 3.821507978767479e-05,
"loss": 0.2888,
"step": 273
},
{
"epoch": 1.111392405063291,
"grad_norm": 0.5562362295269725,
"learning_rate": 3.819156801001693e-05,
"loss": 0.2817,
"step": 274
},
{
"epoch": 1.1154430379746836,
"grad_norm": 0.5184541211850773,
"learning_rate": 3.816790971997121e-05,
"loss": 0.288,
"step": 275
},
{
"epoch": 1.1194936708860759,
"grad_norm": 0.47262644197864695,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.2614,
"step": 276
},
{
"epoch": 1.1235443037974684,
"grad_norm": 0.4930260238012457,
"learning_rate": 3.81201543660571e-05,
"loss": 0.2868,
"step": 277
},
{
"epoch": 1.1275949367088607,
"grad_norm": 0.5868022783325034,
"learning_rate": 3.809605768680377e-05,
"loss": 0.2798,
"step": 278
},
{
"epoch": 1.1316455696202532,
"grad_norm": 0.44795975017085926,
"learning_rate": 3.807181526438958e-05,
"loss": 0.2751,
"step": 279
},
{
"epoch": 1.1356962025316455,
"grad_norm": 0.6072674660790324,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.2802,
"step": 280
},
{
"epoch": 1.139746835443038,
"grad_norm": 0.44326179518799297,
"learning_rate": 3.802289397223145e-05,
"loss": 0.2877,
"step": 281
},
{
"epoch": 1.1437974683544303,
"grad_norm": 0.5998590819192737,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.2939,
"step": 282
},
{
"epoch": 1.1478481012658228,
"grad_norm": 0.4795928226957979,
"learning_rate": 3.797339206560096e-05,
"loss": 0.2939,
"step": 283
},
{
"epoch": 1.1518987341772151,
"grad_norm": 0.4447962992143998,
"learning_rate": 3.794842387948027e-05,
"loss": 0.2857,
"step": 284
},
{
"epoch": 1.1559493670886076,
"grad_norm": 0.5259315515833463,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.2733,
"step": 285
},
{
"epoch": 1.16,
"grad_norm": 0.44602320173286425,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.2853,
"step": 286
},
{
"epoch": 1.1640506329113924,
"grad_norm": 0.47055871340058425,
"learning_rate": 3.787265280646825e-05,
"loss": 0.2989,
"step": 287
},
{
"epoch": 1.1681012658227847,
"grad_norm": 0.5113845260075098,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.2755,
"step": 288
},
{
"epoch": 1.1721518987341772,
"grad_norm": 0.42863275604526263,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.2953,
"step": 289
},
{
"epoch": 1.1762025316455695,
"grad_norm": 0.5510398668597969,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.2753,
"step": 290
},
{
"epoch": 1.180253164556962,
"grad_norm": 0.5039205927592333,
"learning_rate": 3.776961046830306e-05,
"loss": 0.2879,
"step": 291
},
{
"epoch": 1.1843037974683543,
"grad_norm": 0.5541006517329731,
"learning_rate": 3.774349157719215e-05,
"loss": 0.283,
"step": 292
},
{
"epoch": 1.1883544303797469,
"grad_norm": 0.5531323033184927,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.2818,
"step": 293
},
{
"epoch": 1.1924050632911392,
"grad_norm": 0.5454248262328826,
"learning_rate": 3.769082529555359e-05,
"loss": 0.302,
"step": 294
},
{
"epoch": 1.1964556962025317,
"grad_norm": 0.49241048960807515,
"learning_rate": 3.766427832919294e-05,
"loss": 0.2733,
"step": 295
},
{
"epoch": 1.200506329113924,
"grad_norm": 0.41380058960045357,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.2818,
"step": 296
},
{
"epoch": 1.2045569620253165,
"grad_norm": 0.466485883199489,
"learning_rate": 3.761075781441526e-05,
"loss": 0.2897,
"step": 297
},
{
"epoch": 1.2086075949367088,
"grad_norm": 0.3943683422562435,
"learning_rate": 3.75837846970451e-05,
"loss": 0.294,
"step": 298
},
{
"epoch": 1.2126582278481013,
"grad_norm": 0.45176532080716986,
"learning_rate": 3.755666996229032e-05,
"loss": 0.2875,
"step": 299
},
{
"epoch": 1.2167088607594936,
"grad_norm": 0.3993630903970663,
"learning_rate": 3.752941382852927e-05,
"loss": 0.2989,
"step": 300
},
{
"epoch": 1.220759493670886,
"grad_norm": 0.467590829335534,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.294,
"step": 301
},
{
"epoch": 1.2248101265822784,
"grad_norm": 0.42839245768494705,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.2974,
"step": 302
},
{
"epoch": 1.228860759493671,
"grad_norm": 0.4178181782395724,
"learning_rate": 3.744679923406351e-05,
"loss": 0.2927,
"step": 303
},
{
"epoch": 1.2329113924050632,
"grad_norm": 0.4145693787956526,
"learning_rate": 3.741897971081043e-05,
"loss": 0.2985,
"step": 304
},
{
"epoch": 1.2369620253164557,
"grad_norm": 0.4444976274025904,
"learning_rate": 3.739101989748946e-05,
"loss": 0.2797,
"step": 305
},
{
"epoch": 1.241012658227848,
"grad_norm": 0.3628393082382017,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.2768,
"step": 306
},
{
"epoch": 1.2450632911392405,
"grad_norm": 0.43020585557427776,
"learning_rate": 3.73346803025098e-05,
"loss": 0.2889,
"step": 307
},
{
"epoch": 1.2491139240506328,
"grad_norm": 0.4040797532813833,
"learning_rate": 3.730630097460247e-05,
"loss": 0.2852,
"step": 308
},
{
"epoch": 1.2531645569620253,
"grad_norm": 0.5097416434559024,
"learning_rate": 3.727778226412628e-05,
"loss": 0.2754,
"step": 309
},
{
"epoch": 1.2572151898734178,
"grad_norm": 0.39310204463739146,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.277,
"step": 310
},
{
"epoch": 1.2612658227848101,
"grad_norm": 0.459049209285664,
"learning_rate": 3.722032761533114e-05,
"loss": 0.2642,
"step": 311
},
{
"epoch": 1.2653164556962024,
"grad_norm": 0.3897450276226379,
"learning_rate": 3.719139213974403e-05,
"loss": 0.2926,
"step": 312
},
{
"epoch": 1.269367088607595,
"grad_norm": 0.5185341299001212,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.2757,
"step": 313
},
{
"epoch": 1.2734177215189875,
"grad_norm": 0.43158185378152875,
"learning_rate": 3.713310605140055e-05,
"loss": 0.302,
"step": 314
},
{
"epoch": 1.2774683544303798,
"grad_norm": 0.4534274185095166,
"learning_rate": 3.710375590807233e-05,
"loss": 0.308,
"step": 315
},
{
"epoch": 1.281518987341772,
"grad_norm": 0.4586713279052157,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.3055,
"step": 316
},
{
"epoch": 1.2855696202531646,
"grad_norm": 0.4429912339740587,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.2874,
"step": 317
},
{
"epoch": 1.289620253164557,
"grad_norm": 0.39323219414183463,
"learning_rate": 3.701487992136854e-05,
"loss": 0.2867,
"step": 318
},
{
"epoch": 1.2936708860759494,
"grad_norm": 0.4262569559366195,
"learning_rate": 3.69849802022222e-05,
"loss": 0.2961,
"step": 319
},
{
"epoch": 1.2977215189873417,
"grad_norm": 0.3915245542304271,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.2702,
"step": 320
},
{
"epoch": 1.3017721518987342,
"grad_norm": 0.4865183429422377,
"learning_rate": 3.692477062175289e-05,
"loss": 0.2965,
"step": 321
},
{
"epoch": 1.3058227848101267,
"grad_norm": 0.43073033773459357,
"learning_rate": 3.689446124534958e-05,
"loss": 0.2822,
"step": 322
},
{
"epoch": 1.309873417721519,
"grad_norm": 0.44716237684226523,
"learning_rate": 3.686401580327799e-05,
"loss": 0.2863,
"step": 323
},
{
"epoch": 1.3139240506329113,
"grad_norm": 0.3954056077832622,
"learning_rate": 3.683343454074149e-05,
"loss": 0.298,
"step": 324
},
{
"epoch": 1.3179746835443038,
"grad_norm": 0.45600203985738835,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.2712,
"step": 325
},
{
"epoch": 1.3220253164556963,
"grad_norm": 0.4353709774655904,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.2992,
"step": 326
},
{
"epoch": 1.3260759493670886,
"grad_norm": 0.4989420969168605,
"learning_rate": 3.674087829877297e-05,
"loss": 0.2998,
"step": 327
},
{
"epoch": 1.330126582278481,
"grad_norm": 0.42006702572874205,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.2892,
"step": 328
},
{
"epoch": 1.3341772151898734,
"grad_norm": 0.45680381438864237,
"learning_rate": 3.667849957966501e-05,
"loss": 0.273,
"step": 329
},
{
"epoch": 1.338227848101266,
"grad_norm": 0.4298311945658629,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.293,
"step": 330
},
{
"epoch": 1.3422784810126582,
"grad_norm": 0.3940164594567533,
"learning_rate": 3.661558355626795e-05,
"loss": 0.2902,
"step": 331
},
{
"epoch": 1.3463291139240505,
"grad_norm": 0.4115856706011703,
"learning_rate": 3.658392468818163e-05,
"loss": 0.2791,
"step": 332
},
{
"epoch": 1.350379746835443,
"grad_norm": 0.38652062216942906,
"learning_rate": 3.655213225544574e-05,
"loss": 0.3082,
"step": 333
},
{
"epoch": 1.3544303797468356,
"grad_norm": 0.4074694097947701,
"learning_rate": 3.652020651411218e-05,
"loss": 0.2967,
"step": 334
},
{
"epoch": 1.3584810126582278,
"grad_norm": 0.42399137197621306,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.3068,
"step": 335
},
{
"epoch": 1.3625316455696201,
"grad_norm": 0.3665346937913575,
"learning_rate": 3.645595613522574e-05,
"loss": 0.2861,
"step": 336
},
{
"epoch": 1.3665822784810127,
"grad_norm": 0.3646668637395103,
"learning_rate": 3.642363201513657e-05,
"loss": 0.2879,
"step": 337
},
{
"epoch": 1.3706329113924052,
"grad_norm": 0.4616507230660206,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.2896,
"step": 338
},
{
"epoch": 1.3746835443037975,
"grad_norm": 0.38218897174903593,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.2814,
"step": 339
},
{
"epoch": 1.3787341772151898,
"grad_norm": 0.39182682495969823,
"learning_rate": 3.632586705948318e-05,
"loss": 0.2856,
"step": 340
},
{
"epoch": 1.3827848101265823,
"grad_norm": 0.48236149530968975,
"learning_rate": 3.629301541734311e-05,
"loss": 0.2784,
"step": 341
},
{
"epoch": 1.3868354430379748,
"grad_norm": 0.40819947738000484,
"learning_rate": 3.626003255349676e-05,
"loss": 0.2868,
"step": 342
},
{
"epoch": 1.390886075949367,
"grad_norm": 0.4738303186693813,
"learning_rate": 3.622691873358357e-05,
"loss": 0.2823,
"step": 343
},
{
"epoch": 1.3949367088607594,
"grad_norm": 0.43047996398325195,
"learning_rate": 3.61936742242977e-05,
"loss": 0.2746,
"step": 344
},
{
"epoch": 1.398987341772152,
"grad_norm": 0.4700613487552194,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.2855,
"step": 345
},
{
"epoch": 1.4030379746835444,
"grad_norm": 0.4064986485633765,
"learning_rate": 3.612679420964516e-05,
"loss": 0.289,
"step": 346
},
{
"epoch": 1.4070886075949367,
"grad_norm": 0.4637741993198083,
"learning_rate": 3.609315924292092e-05,
"loss": 0.3081,
"step": 347
},
{
"epoch": 1.411139240506329,
"grad_norm": 0.42824788528271346,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.2897,
"step": 348
},
{
"epoch": 1.4151898734177215,
"grad_norm": 0.40688276430727394,
"learning_rate": 3.602550074513133e-05,
"loss": 0.2947,
"step": 349
},
{
"epoch": 1.419240506329114,
"grad_norm": 0.360640959067316,
"learning_rate": 3.599147775897822e-05,
"loss": 0.2897,
"step": 350
},
{
"epoch": 1.4232911392405063,
"grad_norm": 0.48038098010040714,
"learning_rate": 3.595732597966167e-05,
"loss": 0.2999,
"step": 351
},
{
"epoch": 1.4273417721518986,
"grad_norm": 0.43860115295034297,
"learning_rate": 3.592304568223542e-05,
"loss": 0.3015,
"step": 352
},
{
"epoch": 1.4313924050632911,
"grad_norm": 0.44952811373730306,
"learning_rate": 3.588863714278826e-05,
"loss": 0.2985,
"step": 353
},
{
"epoch": 1.4354430379746836,
"grad_norm": 0.4880039266046244,
"learning_rate": 3.585410063844186e-05,
"loss": 0.2876,
"step": 354
},
{
"epoch": 1.439493670886076,
"grad_norm": 0.4116762914809638,
"learning_rate": 3.581943644734846e-05,
"loss": 0.2838,
"step": 355
},
{
"epoch": 1.4435443037974682,
"grad_norm": 0.411955924682326,
"learning_rate": 3.578464484868869e-05,
"loss": 0.2784,
"step": 356
},
{
"epoch": 1.4475949367088607,
"grad_norm": 0.39267445985844457,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.2667,
"step": 357
},
{
"epoch": 1.4516455696202533,
"grad_norm": 0.40268584399533325,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.2878,
"step": 358
},
{
"epoch": 1.4556962025316456,
"grad_norm": 0.3566111500373363,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.3173,
"step": 359
},
{
"epoch": 1.4597468354430378,
"grad_norm": 0.39989157801890446,
"learning_rate": 3.564420999786543e-05,
"loss": 0.2852,
"step": 360
},
{
"epoch": 1.4637974683544304,
"grad_norm": 0.34783691702949565,
"learning_rate": 3.560878558491842e-05,
"loss": 0.3037,
"step": 361
},
{
"epoch": 1.4678481012658229,
"grad_norm": 0.4761314145031566,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.3031,
"step": 362
},
{
"epoch": 1.4718987341772152,
"grad_norm": 0.37623543707432294,
"learning_rate": 3.553755991230039e-05,
"loss": 0.2866,
"step": 363
},
{
"epoch": 1.4759493670886075,
"grad_norm": 0.4011390700086861,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.2896,
"step": 364
},
{
"epoch": 1.48,
"grad_norm": 0.37665471254250593,
"learning_rate": 3.546583369120419e-05,
"loss": 0.2872,
"step": 365
},
{
"epoch": 1.4840506329113925,
"grad_norm": 0.39850657758082936,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.2801,
"step": 366
},
{
"epoch": 1.4881012658227848,
"grad_norm": 0.35127524579942193,
"learning_rate": 3.539360923231766e-05,
"loss": 0.2809,
"step": 367
},
{
"epoch": 1.492151898734177,
"grad_norm": 0.38422905380748357,
"learning_rate": 3.535731089018394e-05,
"loss": 0.2924,
"step": 368
},
{
"epoch": 1.4962025316455696,
"grad_norm": 0.3608860713705904,
"learning_rate": 3.532088886237956e-05,
"loss": 0.293,
"step": 369
},
{
"epoch": 1.5002531645569621,
"grad_norm": 0.371255322353779,
"learning_rate": 3.528434344224253e-05,
"loss": 0.2823,
"step": 370
},
{
"epoch": 1.5043037974683544,
"grad_norm": 0.36648328673621633,
"learning_rate": 3.524767492410464e-05,
"loss": 0.2864,
"step": 371
},
{
"epoch": 1.5083544303797467,
"grad_norm": 0.3537846909412916,
"learning_rate": 3.521088360328908e-05,
"loss": 0.2874,
"step": 372
},
{
"epoch": 1.5124050632911392,
"grad_norm": 0.40062627398747536,
"learning_rate": 3.517396977610811e-05,
"loss": 0.2826,
"step": 373
},
{
"epoch": 1.5164556962025317,
"grad_norm": 0.4119824376439357,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.2804,
"step": 374
},
{
"epoch": 1.520506329113924,
"grad_norm": 0.4027640439732389,
"learning_rate": 3.509977579282971e-05,
"loss": 0.2983,
"step": 375
},
{
"epoch": 1.5245569620253163,
"grad_norm": 0.34955369963115784,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.2964,
"step": 376
},
{
"epoch": 1.5286075949367088,
"grad_norm": 0.3827102180844901,
"learning_rate": 3.502509536445719e-05,
"loss": 0.2786,
"step": 377
},
{
"epoch": 1.5326582278481014,
"grad_norm": 0.3681161165201001,
"learning_rate": 3.498757348458147e-05,
"loss": 0.2864,
"step": 378
},
{
"epoch": 1.5367088607594936,
"grad_norm": 0.3766281821528207,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.2806,
"step": 379
},
{
"epoch": 1.540759493670886,
"grad_norm": 0.40565369882526964,
"learning_rate": 3.491216790442899e-05,
"loss": 0.2893,
"step": 380
},
{
"epoch": 1.5448101265822785,
"grad_norm": 0.3638951306666875,
"learning_rate": 3.487428481145839e-05,
"loss": 0.2691,
"step": 381
},
{
"epoch": 1.548860759493671,
"grad_norm": 0.4276065997650786,
"learning_rate": 3.483628192304278e-05,
"loss": 0.2818,
"step": 382
},
{
"epoch": 1.5529113924050633,
"grad_norm": 0.39828166584094216,
"learning_rate": 3.479815954525219e-05,
"loss": 0.271,
"step": 383
},
{
"epoch": 1.5569620253164556,
"grad_norm": 0.42484683829763253,
"learning_rate": 3.475991798511899e-05,
"loss": 0.291,
"step": 384
},
{
"epoch": 1.561012658227848,
"grad_norm": 0.3686168346607197,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.3029,
"step": 385
},
{
"epoch": 1.5650632911392406,
"grad_norm": 0.3907971565000809,
"learning_rate": 3.468307855075128e-05,
"loss": 0.2919,
"step": 386
},
{
"epoch": 1.5691139240506329,
"grad_norm": 0.3921840947824225,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.2821,
"step": 387
},
{
"epoch": 1.5731645569620252,
"grad_norm": 0.3624537727634575,
"learning_rate": 3.460576609535163e-05,
"loss": 0.2967,
"step": 388
},
{
"epoch": 1.5772151898734177,
"grad_norm": 0.448098919874471,
"learning_rate": 3.456693326250006e-05,
"loss": 0.2905,
"step": 389
},
{
"epoch": 1.5812658227848102,
"grad_norm": 0.37674222704551324,
"learning_rate": 3.452798310957058e-05,
"loss": 0.2756,
"step": 390
},
{
"epoch": 1.5853164556962025,
"grad_norm": 0.3613794902886389,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.2811,
"step": 391
},
{
"epoch": 1.5893670886075948,
"grad_norm": 0.43431263575021284,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.2889,
"step": 392
},
{
"epoch": 1.5934177215189873,
"grad_norm": 0.3686405107250314,
"learning_rate": 3.441043187201574e-05,
"loss": 0.284,
"step": 393
},
{
"epoch": 1.5974683544303798,
"grad_norm": 0.4370525380747679,
"learning_rate": 3.437101558517728e-05,
"loss": 0.2852,
"step": 394
},
{
"epoch": 1.6015189873417721,
"grad_norm": 0.3776299843796782,
"learning_rate": 3.433148355615496e-05,
"loss": 0.2944,
"step": 395
},
{
"epoch": 1.6055696202531644,
"grad_norm": 0.3828636358129876,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.2953,
"step": 396
},
{
"epoch": 1.609620253164557,
"grad_norm": 0.3874580299221138,
"learning_rate": 3.425207354603043e-05,
"loss": 0.3016,
"step": 397
},
{
"epoch": 1.6136708860759494,
"grad_norm": 0.37534436712771674,
"learning_rate": 3.421219620448553e-05,
"loss": 0.2728,
"step": 398
},
{
"epoch": 1.6177215189873417,
"grad_norm": 0.37956253629041636,
"learning_rate": 3.417220439986623e-05,
"loss": 0.2987,
"step": 399
},
{
"epoch": 1.621772151898734,
"grad_norm": 0.3506398693326743,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.2934,
"step": 400
},
{
"epoch": 1.6258227848101265,
"grad_norm": 0.35090999559778885,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.2841,
"step": 401
},
{
"epoch": 1.629873417721519,
"grad_norm": 0.39690695033564577,
"learning_rate": 3.405154543304065e-05,
"loss": 0.2861,
"step": 402
},
{
"epoch": 1.6339240506329114,
"grad_norm": 0.35804954501387887,
"learning_rate": 3.401109900618843e-05,
"loss": 0.2913,
"step": 403
},
{
"epoch": 1.6379746835443036,
"grad_norm": 0.3827652706942494,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.2918,
"step": 404
},
{
"epoch": 1.6420253164556962,
"grad_norm": 0.37565428433060144,
"learning_rate": 3.392986794874714e-05,
"loss": 0.2907,
"step": 405
},
{
"epoch": 1.6460759493670887,
"grad_norm": 0.37760046950189724,
"learning_rate": 3.388908397238184e-05,
"loss": 0.2857,
"step": 406
},
{
"epoch": 1.650126582278481,
"grad_norm": 0.33966060612165627,
"learning_rate": 3.384818813524362e-05,
"loss": 0.2844,
"step": 407
},
{
"epoch": 1.6541772151898733,
"grad_norm": 0.36591461270267467,
"learning_rate": 3.380718076670195e-05,
"loss": 0.2827,
"step": 408
},
{
"epoch": 1.6582278481012658,
"grad_norm": 0.3876395604271008,
"learning_rate": 3.376606219702454e-05,
"loss": 0.2935,
"step": 409
},
{
"epoch": 1.6622784810126583,
"grad_norm": 0.33371939123979627,
"learning_rate": 3.372483275737468e-05,
"loss": 0.2878,
"step": 410
},
{
"epoch": 1.6663291139240506,
"grad_norm": 0.4017800362923474,
"learning_rate": 3.368349277980861e-05,
"loss": 0.3048,
"step": 411
},
{
"epoch": 1.6703797468354429,
"grad_norm": 0.4082370767958038,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.3063,
"step": 412
},
{
"epoch": 1.6744303797468354,
"grad_norm": 0.37664478760146225,
"learning_rate": 3.360048254360144e-05,
"loss": 0.304,
"step": 413
},
{
"epoch": 1.678481012658228,
"grad_norm": 0.41520922182223,
"learning_rate": 3.355881295351336e-05,
"loss": 0.3032,
"step": 414
},
{
"epoch": 1.6825316455696202,
"grad_norm": 0.37347962628062314,
"learning_rate": 3.351703416260975e-05,
"loss": 0.2865,
"step": 415
},
{
"epoch": 1.6865822784810125,
"grad_norm": 0.36127543606249685,
"learning_rate": 3.347514650737126e-05,
"loss": 0.2962,
"step": 416
},
{
"epoch": 1.690632911392405,
"grad_norm": 0.3729742184181311,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.2876,
"step": 417
},
{
"epoch": 1.6946835443037975,
"grad_norm": 0.3770011344278073,
"learning_rate": 3.339104595419334e-05,
"loss": 0.2909,
"step": 418
},
{
"epoch": 1.6987341772151898,
"grad_norm": 0.38342474644242786,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.2942,
"step": 419
},
{
"epoch": 1.7027848101265821,
"grad_norm": 0.4055167772676693,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.2922,
"step": 420
},
{
"epoch": 1.7068354430379746,
"grad_norm": 0.40300333775265046,
"learning_rate": 3.326408710419996e-05,
"loss": 0.2835,
"step": 421
},
{
"epoch": 1.7108860759493671,
"grad_norm": 0.34825784871478643,
"learning_rate": 3.322155337795454e-05,
"loss": 0.2685,
"step": 422
},
{
"epoch": 1.7149367088607594,
"grad_norm": 0.43092606671842243,
"learning_rate": 3.317891316713587e-05,
"loss": 0.2898,
"step": 423
},
{
"epoch": 1.7189873417721517,
"grad_norm": 0.38901099952570506,
"learning_rate": 3.313616681516231e-05,
"loss": 0.2962,
"step": 424
},
{
"epoch": 1.7230379746835442,
"grad_norm": 0.47177344027238133,
"learning_rate": 3.309331466630713e-05,
"loss": 0.2862,
"step": 425
},
{
"epoch": 1.7270886075949368,
"grad_norm": 0.4047356285573965,
"learning_rate": 3.305035706569563e-05,
"loss": 0.2898,
"step": 426
},
{
"epoch": 1.7311392405063293,
"grad_norm": 0.40841892760063164,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.2956,
"step": 427
},
{
"epoch": 1.7351898734177216,
"grad_norm": 0.4192719388462693,
"learning_rate": 3.296412689394864e-05,
"loss": 0.281,
"step": 428
},
{
"epoch": 1.7392405063291139,
"grad_norm": 0.410163872783481,
"learning_rate": 3.292085501729909e-05,
"loss": 0.2818,
"step": 429
},
{
"epoch": 1.7432911392405064,
"grad_norm": 0.38922417711281593,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.2869,
"step": 430
},
{
"epoch": 1.747341772151899,
"grad_norm": 0.3815528193242208,
"learning_rate": 3.283399942497381e-05,
"loss": 0.2786,
"step": 431
},
{
"epoch": 1.7513924050632912,
"grad_norm": 0.3828359725761322,
"learning_rate": 3.279041640882108e-05,
"loss": 0.274,
"step": 432
},
{
"epoch": 1.7554430379746835,
"grad_norm": 0.3972491009356625,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.283,
"step": 433
},
{
"epoch": 1.759493670886076,
"grad_norm": 0.39428804323647265,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.276,
"step": 434
},
{
"epoch": 1.7635443037974685,
"grad_norm": 0.4217587663899652,
"learning_rate": 3.265905069502244e-05,
"loss": 0.2826,
"step": 435
},
{
"epoch": 1.7675949367088608,
"grad_norm": 0.39756921189757033,
"learning_rate": 3.261505774420016e-05,
"loss": 0.2878,
"step": 436
},
{
"epoch": 1.771645569620253,
"grad_norm": 0.38945365781496827,
"learning_rate": 3.257096319343707e-05,
"loss": 0.2909,
"step": 437
},
{
"epoch": 1.7756962025316456,
"grad_norm": 0.3750653020053038,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.2809,
"step": 438
},
{
"epoch": 1.7797468354430381,
"grad_norm": 0.374364042985438,
"learning_rate": 3.248247071342966e-05,
"loss": 0.2785,
"step": 439
},
{
"epoch": 1.7837974683544304,
"grad_norm": 0.3673277288630421,
"learning_rate": 3.243807349689161e-05,
"loss": 0.2856,
"step": 440
},
{
"epoch": 1.7878481012658227,
"grad_norm": 0.4180813231036564,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.2935,
"step": 441
},
{
"epoch": 1.7918987341772152,
"grad_norm": 0.36858387017618277,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.2995,
"step": 442
},
{
"epoch": 1.7959493670886078,
"grad_norm": 0.4141844451463075,
"learning_rate": 3.230428223438075e-05,
"loss": 0.2792,
"step": 443
},
{
"epoch": 1.8,
"grad_norm": 0.37930381691973064,
"learning_rate": 3.225948647317459e-05,
"loss": 0.2995,
"step": 444
},
{
"epoch": 1.8040506329113923,
"grad_norm": 0.3461598713799288,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.2803,
"step": 445
},
{
"epoch": 1.8081012658227849,
"grad_norm": 0.37722272374216254,
"learning_rate": 3.216959910368034e-05,
"loss": 0.2969,
"step": 446
},
{
"epoch": 1.8121518987341774,
"grad_norm": 0.39383850256817515,
"learning_rate": 3.212450821933277e-05,
"loss": 0.2856,
"step": 447
},
{
"epoch": 1.8162025316455697,
"grad_norm": 0.3571712738004219,
"learning_rate": 3.207931968586281e-05,
"loss": 0.2895,
"step": 448
},
{
"epoch": 1.820253164556962,
"grad_norm": 0.4108712443892209,
"learning_rate": 3.203403386721272e-05,
"loss": 0.2895,
"step": 449
},
{
"epoch": 1.8243037974683545,
"grad_norm": 0.4014417624137479,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.2833,
"step": 450
},
{
"epoch": 1.828354430379747,
"grad_norm": 0.3961977349121418,
"learning_rate": 3.194317183405573e-05,
"loss": 0.2895,
"step": 451
},
{
"epoch": 1.8324050632911393,
"grad_norm": 0.4338148980931281,
"learning_rate": 3.189759635133914e-05,
"loss": 0.288,
"step": 452
},
{
"epoch": 1.8364556962025316,
"grad_norm": 0.3251403085246064,
"learning_rate": 3.185192504701718e-05,
"loss": 0.2855,
"step": 453
},
{
"epoch": 1.840506329113924,
"grad_norm": 0.3882080742360515,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.2853,
"step": 454
},
{
"epoch": 1.8445569620253166,
"grad_norm": 0.40177374717320075,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.2856,
"step": 455
},
{
"epoch": 1.848607594936709,
"grad_norm": 0.3951050255681593,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.2818,
"step": 456
},
{
"epoch": 1.8526582278481012,
"grad_norm": 0.3712475489278307,
"learning_rate": 3.166828898179731e-05,
"loss": 0.281,
"step": 457
},
{
"epoch": 1.8567088607594937,
"grad_norm": 0.4459106545113236,
"learning_rate": 3.162214410223536e-05,
"loss": 0.2692,
"step": 458
},
{
"epoch": 1.8607594936708862,
"grad_norm": 0.3750903471412591,
"learning_rate": 3.157590561952257e-05,
"loss": 0.2744,
"step": 459
},
{
"epoch": 1.8648101265822785,
"grad_norm": 0.47487493861251134,
"learning_rate": 3.152957390605732e-05,
"loss": 0.2892,
"step": 460
},
{
"epoch": 1.8688607594936708,
"grad_norm": 0.3825549302274178,
"learning_rate": 3.148314933498886e-05,
"loss": 0.2855,
"step": 461
},
{
"epoch": 1.8729113924050633,
"grad_norm": 0.4047035075174838,
"learning_rate": 3.143663228021431e-05,
"loss": 0.2788,
"step": 462
},
{
"epoch": 1.8769620253164558,
"grad_norm": 0.41474401171108916,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.2851,
"step": 463
},
{
"epoch": 1.8810126582278481,
"grad_norm": 0.3972443327353071,
"learning_rate": 3.134332221885661e-05,
"loss": 0.3008,
"step": 464
},
{
"epoch": 1.8850632911392404,
"grad_norm": 0.4095932073668544,
"learning_rate": 3.129652996377987e-05,
"loss": 0.2997,
"step": 465
},
{
"epoch": 1.889113924050633,
"grad_norm": 0.37696719071278284,
"learning_rate": 3.12496467280038e-05,
"loss": 0.2867,
"step": 466
},
{
"epoch": 1.8931645569620255,
"grad_norm": 0.3490692865271736,
"learning_rate": 3.120267288911952e-05,
"loss": 0.2748,
"step": 467
},
{
"epoch": 1.8972151898734178,
"grad_norm": 0.44184394643842034,
"learning_rate": 3.11556088254479e-05,
"loss": 0.2753,
"step": 468
},
{
"epoch": 1.90126582278481,
"grad_norm": 0.3772320730868035,
"learning_rate": 3.11084549160364e-05,
"loss": 0.2837,
"step": 469
},
{
"epoch": 1.9053164556962026,
"grad_norm": 0.4269113221169655,
"learning_rate": 3.106121154065615e-05,
"loss": 0.2892,
"step": 470
},
{
"epoch": 1.909367088607595,
"grad_norm": 0.3759232562192361,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.2924,
"step": 471
},
{
"epoch": 1.9134177215189874,
"grad_norm": 0.3829742083458787,
"learning_rate": 3.096645791467348e-05,
"loss": 0.2839,
"step": 472
},
{
"epoch": 1.9174683544303797,
"grad_norm": 0.36967957514343597,
"learning_rate": 3.091894842720373e-05,
"loss": 0.2993,
"step": 473
},
{
"epoch": 1.9215189873417722,
"grad_norm": 0.35539933132582974,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.2916,
"step": 474
},
{
"epoch": 1.9255696202531647,
"grad_norm": 0.40923983026678473,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.2949,
"step": 475
},
{
"epoch": 1.929620253164557,
"grad_norm": 0.3523266048455141,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.2951,
"step": 476
},
{
"epoch": 1.9336708860759493,
"grad_norm": 0.384232584476044,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.2873,
"step": 477
},
{
"epoch": 1.9377215189873418,
"grad_norm": 0.3700527865734961,
"learning_rate": 3.06800895716334e-05,
"loss": 0.2854,
"step": 478
},
{
"epoch": 1.9417721518987343,
"grad_norm": 0.3444383510223316,
"learning_rate": 3.063205821011029e-05,
"loss": 0.2844,
"step": 479
},
{
"epoch": 1.9458227848101266,
"grad_norm": 0.3466566782314073,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.2723,
"step": 480
},
{
"epoch": 1.949873417721519,
"grad_norm": 0.33068565421054796,
"learning_rate": 3.053573898718618e-05,
"loss": 0.2876,
"step": 481
},
{
"epoch": 1.9539240506329114,
"grad_norm": 0.34781193111013486,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.2698,
"step": 482
},
{
"epoch": 1.957974683544304,
"grad_norm": 0.34133872153504463,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.2799,
"step": 483
},
{
"epoch": 1.9620253164556962,
"grad_norm": 0.38229996975696096,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.2648,
"step": 484
},
{
"epoch": 1.9660759493670885,
"grad_norm": 0.3692374932974212,
"learning_rate": 3.034208541656255e-05,
"loss": 0.3118,
"step": 485
},
{
"epoch": 1.970126582278481,
"grad_norm": 0.40622539628510795,
"learning_rate": 3.029346281309218e-05,
"loss": 0.2984,
"step": 486
},
{
"epoch": 1.9741772151898735,
"grad_norm": 0.33221980559143033,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.286,
"step": 487
},
{
"epoch": 1.9782278481012658,
"grad_norm": 0.42867860255879847,
"learning_rate": 3.019596929201569e-05,
"loss": 0.2745,
"step": 488
},
{
"epoch": 1.9822784810126581,
"grad_norm": 0.40104187008540443,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.2987,
"step": 489
},
{
"epoch": 1.9863291139240506,
"grad_norm": 0.3848980881026057,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.2747,
"step": 490
},
{
"epoch": 1.9903797468354432,
"grad_norm": 0.3720312445974263,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.2798,
"step": 491
},
{
"epoch": 1.9944303797468355,
"grad_norm": 0.32988440274840736,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.2899,
"step": 492
},
{
"epoch": 1.9984810126582278,
"grad_norm": 0.35303426616194505,
"learning_rate": 2.995080534235264e-05,
"loss": 0.2764,
"step": 493
},
{
"epoch": 2.0040506329113925,
"grad_norm": 0.4900311155892411,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.1946,
"step": 494
},
{
"epoch": 2.008101265822785,
"grad_norm": 0.36630539575863963,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.1948,
"step": 495
},
{
"epoch": 2.012151898734177,
"grad_norm": 0.803065685325802,
"learning_rate": 2.980274210295326e-05,
"loss": 0.1984,
"step": 496
},
{
"epoch": 2.0162025316455696,
"grad_norm": 0.4107731759832203,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.1861,
"step": 497
},
{
"epoch": 2.020253164556962,
"grad_norm": 0.4233625404069576,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.201,
"step": 498
},
{
"epoch": 2.0243037974683546,
"grad_norm": 0.37211389721028065,
"learning_rate": 2.965396831793362e-05,
"loss": 0.1864,
"step": 499
},
{
"epoch": 2.0283544303797467,
"grad_norm": 0.36956964131509024,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.1846,
"step": 500
},
{
"epoch": 2.0324050632911392,
"grad_norm": 0.4192006843151016,
"learning_rate": 2.955439637024526e-05,
"loss": 0.182,
"step": 501
},
{
"epoch": 2.0364556962025318,
"grad_norm": 0.3885270340383296,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.1776,
"step": 502
},
{
"epoch": 2.0405063291139243,
"grad_norm": 0.4233909343953235,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.1866,
"step": 503
},
{
"epoch": 2.0445569620253163,
"grad_norm": 0.3803374111830457,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.1811,
"step": 504
},
{
"epoch": 2.048607594936709,
"grad_norm": 0.37531408985648984,
"learning_rate": 2.93543322968538e-05,
"loss": 0.2006,
"step": 505
},
{
"epoch": 2.0526582278481014,
"grad_norm": 0.39682310358243067,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.2043,
"step": 506
},
{
"epoch": 2.056708860759494,
"grad_norm": 0.373275817606965,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.1977,
"step": 507
},
{
"epoch": 2.060759493670886,
"grad_norm": 0.34654183822177714,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.1898,
"step": 508
},
{
"epoch": 2.0648101265822785,
"grad_norm": 0.3834090240605994,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.2039,
"step": 509
},
{
"epoch": 2.068860759493671,
"grad_norm": 0.33966486958016845,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.1967,
"step": 510
},
{
"epoch": 2.0729113924050635,
"grad_norm": 0.3669221485440549,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.1982,
"step": 511
},
{
"epoch": 2.0769620253164556,
"grad_norm": 0.3701930961086422,
"learning_rate": 2.900133526131588e-05,
"loss": 0.1891,
"step": 512
},
{
"epoch": 2.081012658227848,
"grad_norm": 0.40547705437775666,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.1952,
"step": 513
},
{
"epoch": 2.0850632911392406,
"grad_norm": 0.3540887148565302,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.1976,
"step": 514
},
{
"epoch": 2.089113924050633,
"grad_norm": 0.3682830709190753,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.1916,
"step": 515
},
{
"epoch": 2.093164556962025,
"grad_norm": 0.36558271913678136,
"learning_rate": 2.879801882631847e-05,
"loss": 0.1943,
"step": 516
},
{
"epoch": 2.0972151898734177,
"grad_norm": 0.36668553235561663,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.1889,
"step": 517
},
{
"epoch": 2.1012658227848102,
"grad_norm": 0.36339448510744066,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.2011,
"step": 518
},
{
"epoch": 2.1053164556962027,
"grad_norm": 0.3534553421325113,
"learning_rate": 2.864478605989494e-05,
"loss": 0.1809,
"step": 519
},
{
"epoch": 2.109367088607595,
"grad_norm": 0.36453651775334395,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.1876,
"step": 520
},
{
"epoch": 2.1134177215189873,
"grad_norm": 0.342861810160757,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.1871,
"step": 521
},
{
"epoch": 2.11746835443038,
"grad_norm": 0.3743990590026,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.1851,
"step": 522
},
{
"epoch": 2.1215189873417724,
"grad_norm": 0.3360490872445367,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.204,
"step": 523
},
{
"epoch": 2.1255696202531644,
"grad_norm": 0.37518974063589194,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.1994,
"step": 524
},
{
"epoch": 2.129620253164557,
"grad_norm": 0.34761477380389816,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.1853,
"step": 525
},
{
"epoch": 2.1336708860759495,
"grad_norm": 0.3390455388871827,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.1814,
"step": 526
},
{
"epoch": 2.137721518987342,
"grad_norm": 0.3506362500892247,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.1741,
"step": 527
},
{
"epoch": 2.141772151898734,
"grad_norm": 0.3448921829937808,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.1964,
"step": 528
},
{
"epoch": 2.1458227848101266,
"grad_norm": 0.404966901545389,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.1958,
"step": 529
},
{
"epoch": 2.149873417721519,
"grad_norm": 0.36601728703435005,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.1777,
"step": 530
},
{
"epoch": 2.1539240506329116,
"grad_norm": 0.3675245076433908,
"learning_rate": 2.802570062143278e-05,
"loss": 0.2083,
"step": 531
},
{
"epoch": 2.1579746835443037,
"grad_norm": 0.35314211446387506,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.1956,
"step": 532
},
{
"epoch": 2.162025316455696,
"grad_norm": 0.3845193182174272,
"learning_rate": 2.792159532078314e-05,
"loss": 0.1947,
"step": 533
},
{
"epoch": 2.1660759493670887,
"grad_norm": 0.3554117994129895,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.1949,
"step": 534
},
{
"epoch": 2.170126582278481,
"grad_norm": 0.4059965499645825,
"learning_rate": 2.781723482288857e-05,
"loss": 0.2052,
"step": 535
},
{
"epoch": 2.1741772151898733,
"grad_norm": 0.3301629668380151,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.1845,
"step": 536
},
{
"epoch": 2.178227848101266,
"grad_norm": 0.40320533143307496,
"learning_rate": 2.771262248976272e-05,
"loss": 0.1985,
"step": 537
},
{
"epoch": 2.1822784810126583,
"grad_norm": 0.3409931001549918,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.1969,
"step": 538
},
{
"epoch": 2.186329113924051,
"grad_norm": 0.35186740069640265,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.182,
"step": 539
},
{
"epoch": 2.190379746835443,
"grad_norm": 0.3516102308083508,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.1868,
"step": 540
},
{
"epoch": 2.1944303797468354,
"grad_norm": 0.3613538158312858,
"learning_rate": 2.7502655806328e-05,
"loss": 0.184,
"step": 541
},
{
"epoch": 2.198481012658228,
"grad_norm": 0.36870682694274065,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.1746,
"step": 542
},
{
"epoch": 2.2025316455696204,
"grad_norm": 0.37244910199475145,
"learning_rate": 2.739730822017673e-05,
"loss": 0.1842,
"step": 543
},
{
"epoch": 2.2065822784810125,
"grad_norm": 0.363234301126403,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.1905,
"step": 544
},
{
"epoch": 2.210632911392405,
"grad_norm": 0.36168523646172046,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.2,
"step": 545
},
{
"epoch": 2.2146835443037975,
"grad_norm": 0.3991254362313696,
"learning_rate": 2.723884107774236e-05,
"loss": 0.1858,
"step": 546
},
{
"epoch": 2.21873417721519,
"grad_norm": 0.33370877065022214,
"learning_rate": 2.718590152796239e-05,
"loss": 0.1925,
"step": 547
},
{
"epoch": 2.222784810126582,
"grad_norm": 0.3813115740100223,
"learning_rate": 2.71329041039194e-05,
"loss": 0.189,
"step": 548
},
{
"epoch": 2.2268354430379746,
"grad_norm": 0.3540831041862967,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.1884,
"step": 549
},
{
"epoch": 2.230886075949367,
"grad_norm": 0.35249265727662465,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.195,
"step": 550
},
{
"epoch": 2.2349367088607597,
"grad_norm": 0.31952974495208875,
"learning_rate": 2.697356885686189e-05,
"loss": 0.2007,
"step": 551
},
{
"epoch": 2.2389873417721518,
"grad_norm": 0.35450711764378295,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.2037,
"step": 552
},
{
"epoch": 2.2430379746835443,
"grad_norm": 0.3504830469965976,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.1796,
"step": 553
},
{
"epoch": 2.247088607594937,
"grad_norm": 0.3451126576762221,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.2078,
"step": 554
},
{
"epoch": 2.2511392405063293,
"grad_norm": 0.38240509438542036,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.1877,
"step": 555
},
{
"epoch": 2.2551898734177214,
"grad_norm": 0.37780959430280125,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.1917,
"step": 556
},
{
"epoch": 2.259240506329114,
"grad_norm": 0.3414749636945791,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.1977,
"step": 557
},
{
"epoch": 2.2632911392405064,
"grad_norm": 0.37548043746840437,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.1981,
"step": 558
},
{
"epoch": 2.267341772151899,
"grad_norm": 0.34465335197069497,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.1982,
"step": 559
},
{
"epoch": 2.271392405063291,
"grad_norm": 0.36771481115547733,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.2048,
"step": 560
},
{
"epoch": 2.2754430379746835,
"grad_norm": 0.3381879801304288,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.1797,
"step": 561
},
{
"epoch": 2.279493670886076,
"grad_norm": 0.37517891567819445,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.2011,
"step": 562
},
{
"epoch": 2.2835443037974685,
"grad_norm": 0.3589840194695012,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.1897,
"step": 563
},
{
"epoch": 2.2875949367088606,
"grad_norm": 0.36096138064839056,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.1964,
"step": 564
},
{
"epoch": 2.291645569620253,
"grad_norm": 0.33199164984950585,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.1952,
"step": 565
},
{
"epoch": 2.2956962025316456,
"grad_norm": 0.36132591967783084,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.1789,
"step": 566
},
{
"epoch": 2.299746835443038,
"grad_norm": 0.36749281844876563,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.211,
"step": 567
},
{
"epoch": 2.3037974683544302,
"grad_norm": 0.3440342153725153,
"learning_rate": 2.606146264569603e-05,
"loss": 0.2038,
"step": 568
},
{
"epoch": 2.3078481012658227,
"grad_norm": 0.3577339112700625,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.1921,
"step": 569
},
{
"epoch": 2.3118987341772153,
"grad_norm": 0.3504614290658007,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.1863,
"step": 570
},
{
"epoch": 2.3159493670886078,
"grad_norm": 0.38048091609289386,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.2109,
"step": 571
},
{
"epoch": 2.32,
"grad_norm": 0.320931057861772,
"learning_rate": 2.584472010606015e-05,
"loss": 0.1848,
"step": 572
},
{
"epoch": 2.3240506329113924,
"grad_norm": 0.34405087006073315,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.1839,
"step": 573
},
{
"epoch": 2.328101265822785,
"grad_norm": 0.3632345262185211,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.1915,
"step": 574
},
{
"epoch": 2.3321518987341774,
"grad_norm": 0.34098129467516525,
"learning_rate": 2.568166741322048e-05,
"loss": 0.1942,
"step": 575
},
{
"epoch": 2.3362025316455695,
"grad_norm": 0.3359704750944849,
"learning_rate": 2.56272244128509e-05,
"loss": 0.2058,
"step": 576
},
{
"epoch": 2.340253164556962,
"grad_norm": 0.3722691582578679,
"learning_rate": 2.55727360915895e-05,
"loss": 0.1973,
"step": 577
},
{
"epoch": 2.3443037974683545,
"grad_norm": 0.32076340331161673,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.1856,
"step": 578
},
{
"epoch": 2.348354430379747,
"grad_norm": 0.35236891506979506,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.1928,
"step": 579
},
{
"epoch": 2.352405063291139,
"grad_norm": 0.34603105148353813,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.1884,
"step": 580
},
{
"epoch": 2.3564556962025316,
"grad_norm": 0.34973370261092257,
"learning_rate": 2.535433837985559e-05,
"loss": 0.1908,
"step": 581
},
{
"epoch": 2.360506329113924,
"grad_norm": 0.3678727520329722,
"learning_rate": 2.529963004393324e-05,
"loss": 0.1954,
"step": 582
},
{
"epoch": 2.3645569620253166,
"grad_norm": 0.34412110818218755,
"learning_rate": 2.524487902551908e-05,
"loss": 0.1998,
"step": 583
},
{
"epoch": 2.3686075949367087,
"grad_norm": 0.349886616274705,
"learning_rate": 2.519008576557029e-05,
"loss": 0.2003,
"step": 584
},
{
"epoch": 2.372658227848101,
"grad_norm": 0.3500087852381463,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.1851,
"step": 585
},
{
"epoch": 2.3767088607594937,
"grad_norm": 0.35315339324463035,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.2117,
"step": 586
},
{
"epoch": 2.3807594936708862,
"grad_norm": 0.34884668471260877,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.2009,
"step": 587
},
{
"epoch": 2.3848101265822783,
"grad_norm": 0.37152309548871704,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.1873,
"step": 588
},
{
"epoch": 2.388860759493671,
"grad_norm": 0.3252745620983514,
"learning_rate": 2.491550129992345e-05,
"loss": 0.1884,
"step": 589
},
{
"epoch": 2.3929113924050633,
"grad_norm": 0.3527344561680257,
"learning_rate": 2.486046386967024e-05,
"loss": 0.2023,
"step": 590
},
{
"epoch": 2.396962025316456,
"grad_norm": 0.32344374569244916,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.1837,
"step": 591
},
{
"epoch": 2.401012658227848,
"grad_norm": 0.3802337858656375,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.2124,
"step": 592
},
{
"epoch": 2.4050632911392404,
"grad_norm": 0.3466088796240934,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.188,
"step": 593
},
{
"epoch": 2.409113924050633,
"grad_norm": 0.33873889009541863,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.1942,
"step": 594
},
{
"epoch": 2.4131645569620255,
"grad_norm": 0.36874802036296894,
"learning_rate": 2.458469841202444e-05,
"loss": 0.2113,
"step": 595
},
{
"epoch": 2.4172151898734175,
"grad_norm": 0.3745395239170412,
"learning_rate": 2.452943276859226e-05,
"loss": 0.1958,
"step": 596
},
{
"epoch": 2.42126582278481,
"grad_norm": 0.3453930816710518,
"learning_rate": 2.447413064573125e-05,
"loss": 0.204,
"step": 597
},
{
"epoch": 2.4253164556962026,
"grad_norm": 0.3667871453498349,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.2008,
"step": 598
},
{
"epoch": 2.429367088607595,
"grad_norm": 0.3583485953315999,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.2064,
"step": 599
},
{
"epoch": 2.433417721518987,
"grad_norm": 0.3354311184181883,
"learning_rate": 2.430800985597963e-05,
"loss": 0.1964,
"step": 600
},
{
"epoch": 2.4374683544303797,
"grad_norm": 0.35255180312459283,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.1959,
"step": 601
},
{
"epoch": 2.441518987341772,
"grad_norm": 0.3352825388713102,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.206,
"step": 602
},
{
"epoch": 2.4455696202531647,
"grad_norm": 0.34168139421451693,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.2065,
"step": 603
},
{
"epoch": 2.449620253164557,
"grad_norm": 0.3440366252332224,
"learning_rate": 2.408603181104957e-05,
"loss": 0.1894,
"step": 604
},
{
"epoch": 2.4536708860759493,
"grad_norm": 0.3417489858158557,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.1877,
"step": 605
},
{
"epoch": 2.457721518987342,
"grad_norm": 0.3235386041914362,
"learning_rate": 2.397484355008521e-05,
"loss": 0.188,
"step": 606
},
{
"epoch": 2.4617721518987343,
"grad_norm": 0.3315132273484202,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.2048,
"step": 607
},
{
"epoch": 2.4658227848101264,
"grad_norm": 0.3444826638226194,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.1884,
"step": 608
},
{
"epoch": 2.469873417721519,
"grad_norm": 0.3226502138801221,
"learning_rate": 2.380782218329337e-05,
"loss": 0.203,
"step": 609
},
{
"epoch": 2.4739240506329114,
"grad_norm": 0.32605592702074265,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.1848,
"step": 610
},
{
"epoch": 2.477974683544304,
"grad_norm": 0.3372131480668872,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.1986,
"step": 611
},
{
"epoch": 2.482025316455696,
"grad_norm": 0.3349401041523225,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.1878,
"step": 612
},
{
"epoch": 2.4860759493670885,
"grad_norm": 0.3256292880631055,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.1871,
"step": 613
},
{
"epoch": 2.490126582278481,
"grad_norm": 0.3391879065839507,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.2004,
"step": 614
},
{
"epoch": 2.4941772151898736,
"grad_norm": 0.3177767367975741,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.2114,
"step": 615
},
{
"epoch": 2.4982278481012656,
"grad_norm": 0.3410217323307578,
"learning_rate": 2.341705325992901e-05,
"loss": 0.1972,
"step": 616
},
{
"epoch": 2.502278481012658,
"grad_norm": 0.3305928895777982,
"learning_rate": 2.336111544604222e-05,
"loss": 0.1984,
"step": 617
},
{
"epoch": 2.5063291139240507,
"grad_norm": 0.3378937123145343,
"learning_rate": 2.33051505621937e-05,
"loss": 0.1966,
"step": 618
},
{
"epoch": 2.510379746835443,
"grad_norm": 0.3381323980242755,
"learning_rate": 2.324915905911693e-05,
"loss": 0.1874,
"step": 619
},
{
"epoch": 2.5144303797468357,
"grad_norm": 0.32886701561491505,
"learning_rate": 2.319314138775977e-05,
"loss": 0.1986,
"step": 620
},
{
"epoch": 2.5184810126582278,
"grad_norm": 0.32792487585698393,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.176,
"step": 621
},
{
"epoch": 2.5225316455696203,
"grad_norm": 0.32906244523029776,
"learning_rate": 2.308102934504593e-05,
"loss": 0.193,
"step": 622
},
{
"epoch": 2.526582278481013,
"grad_norm": 0.34035247879577246,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.1925,
"step": 623
},
{
"epoch": 2.530632911392405,
"grad_norm": 0.34518904964036434,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.1935,
"step": 624
},
{
"epoch": 2.5346835443037974,
"grad_norm": 0.35765903366809465,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.191,
"step": 625
},
{
"epoch": 2.53873417721519,
"grad_norm": 0.327239671315116,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.2064,
"step": 626
},
{
"epoch": 2.5427848101265824,
"grad_norm": 0.33103298806866754,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.1773,
"step": 627
},
{
"epoch": 2.546835443037975,
"grad_norm": 0.3381919652061972,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.1878,
"step": 628
},
{
"epoch": 2.550886075949367,
"grad_norm": 0.322907904052979,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.1953,
"step": 629
},
{
"epoch": 2.5549367088607595,
"grad_norm": 0.36948366793783366,
"learning_rate": 2.26316247733093e-05,
"loss": 0.2062,
"step": 630
},
{
"epoch": 2.558987341772152,
"grad_norm": 0.32634118377452853,
"learning_rate": 2.257534907110328e-05,
"loss": 0.1802,
"step": 631
},
{
"epoch": 2.563037974683544,
"grad_norm": 0.3379797117096483,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.1949,
"step": 632
},
{
"epoch": 2.5670886075949366,
"grad_norm": 0.31650953791139286,
"learning_rate": 2.246273589557294e-05,
"loss": 0.1914,
"step": 633
},
{
"epoch": 2.571139240506329,
"grad_norm": 0.34461031388487134,
"learning_rate": 2.240639932921966e-05,
"loss": 0.2016,
"step": 634
},
{
"epoch": 2.5751898734177217,
"grad_norm": 0.32776801477233547,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.2084,
"step": 635
},
{
"epoch": 2.579240506329114,
"grad_norm": 0.3190098856768315,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.1898,
"step": 636
},
{
"epoch": 2.5832911392405062,
"grad_norm": 0.33571846999637334,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.1862,
"step": 637
},
{
"epoch": 2.5873417721518988,
"grad_norm": 0.3387095331100848,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.1949,
"step": 638
},
{
"epoch": 2.5913924050632913,
"grad_norm": 0.3398466930168304,
"learning_rate": 2.212443486523819e-05,
"loss": 0.1774,
"step": 639
},
{
"epoch": 2.5954430379746833,
"grad_norm": 0.3649576267661619,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.1974,
"step": 640
},
{
"epoch": 2.599493670886076,
"grad_norm": 0.36591328165952924,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.183,
"step": 641
},
{
"epoch": 2.6035443037974684,
"grad_norm": 0.3456786859231735,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.2012,
"step": 642
},
{
"epoch": 2.607594936708861,
"grad_norm": 0.38056766195489444,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.1916,
"step": 643
},
{
"epoch": 2.6116455696202534,
"grad_norm": 0.3519555099000642,
"learning_rate": 2.184204266048803e-05,
"loss": 0.2007,
"step": 644
},
{
"epoch": 2.6156962025316455,
"grad_norm": 0.38378289470701965,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.1979,
"step": 645
},
{
"epoch": 2.619746835443038,
"grad_norm": 0.37002752705320235,
"learning_rate": 2.17289787446841e-05,
"loss": 0.2003,
"step": 646
},
{
"epoch": 2.6237974683544305,
"grad_norm": 0.3645045759379658,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.2084,
"step": 647
},
{
"epoch": 2.6278481012658226,
"grad_norm": 0.35570655140462715,
"learning_rate": 2.161585912916385e-05,
"loss": 0.1863,
"step": 648
},
{
"epoch": 2.631898734177215,
"grad_norm": 0.3142617312950977,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.191,
"step": 649
},
{
"epoch": 2.6359493670886076,
"grad_norm": 0.34568594889979154,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.1992,
"step": 650
},
{
"epoch": 2.64,
"grad_norm": 0.34806692119962146,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.1882,
"step": 651
},
{
"epoch": 2.6440506329113926,
"grad_norm": 0.30950012296672225,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.1874,
"step": 652
},
{
"epoch": 2.6481012658227847,
"grad_norm": 0.35777025582582944,
"learning_rate": 2.133284032321232e-05,
"loss": 0.1915,
"step": 653
},
{
"epoch": 2.6521518987341772,
"grad_norm": 0.33084611676247583,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.1851,
"step": 654
},
{
"epoch": 2.6562025316455697,
"grad_norm": 0.3312829961773294,
"learning_rate": 2.121955446741306e-05,
"loss": 0.1936,
"step": 655
},
{
"epoch": 2.660253164556962,
"grad_norm": 0.34641050480722985,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.1926,
"step": 656
},
{
"epoch": 2.6643037974683543,
"grad_norm": 0.3247759609776688,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.1837,
"step": 657
},
{
"epoch": 2.668354430379747,
"grad_norm": 0.32416187943746455,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.2008,
"step": 658
},
{
"epoch": 2.6724050632911394,
"grad_norm": 0.3213151404875882,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.2114,
"step": 659
},
{
"epoch": 2.676455696202532,
"grad_norm": 0.34397134772928023,
"learning_rate": 2.093617592756158e-05,
"loss": 0.1998,
"step": 660
},
{
"epoch": 2.680506329113924,
"grad_norm": 0.32066688241486285,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.2016,
"step": 661
},
{
"epoch": 2.6845569620253165,
"grad_norm": 0.32988109461571996,
"learning_rate": 2.082276853708978e-05,
"loss": 0.1744,
"step": 662
},
{
"epoch": 2.688607594936709,
"grad_norm": 0.31336619173248925,
"learning_rate": 2.076605467380071e-05,
"loss": 0.194,
"step": 663
},
{
"epoch": 2.692658227848101,
"grad_norm": 0.3323932436532686,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.1953,
"step": 664
},
{
"epoch": 2.6967088607594936,
"grad_norm": 0.3067991419131723,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.204,
"step": 665
},
{
"epoch": 2.700759493670886,
"grad_norm": 0.3406908815022445,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.1918,
"step": 666
},
{
"epoch": 2.7048101265822786,
"grad_norm": 0.3131445726179939,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.2083,
"step": 667
},
{
"epoch": 2.708860759493671,
"grad_norm": 0.32393260820827613,
"learning_rate": 2.048240194884496e-05,
"loss": 0.1929,
"step": 668
},
{
"epoch": 2.712911392405063,
"grad_norm": 0.33567704699551937,
"learning_rate": 2.042565792045695e-05,
"loss": 0.1994,
"step": 669
},
{
"epoch": 2.7169620253164557,
"grad_norm": 0.3264860191511337,
"learning_rate": 2.036891046387857e-05,
"loss": 0.1979,
"step": 670
},
{
"epoch": 2.721012658227848,
"grad_norm": 0.32948295996039456,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.1881,
"step": 671
},
{
"epoch": 2.7250632911392403,
"grad_norm": 0.33414032183820763,
"learning_rate": 2.025540709431948e-05,
"loss": 0.1963,
"step": 672
},
{
"epoch": 2.729113924050633,
"grad_norm": 0.3146575947262753,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.1896,
"step": 673
},
{
"epoch": 2.7331645569620253,
"grad_norm": 0.3557913839686474,
"learning_rate": 2.014189549672245e-05,
"loss": 0.1927,
"step": 674
},
{
"epoch": 2.737215189873418,
"grad_norm": 0.35392166253141916,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.1944,
"step": 675
},
{
"epoch": 2.7412658227848103,
"grad_norm": 0.3363416249830106,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.1958,
"step": 676
},
{
"epoch": 2.7453164556962024,
"grad_norm": 0.38167241513711114,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.205,
"step": 677
},
{
"epoch": 2.749367088607595,
"grad_norm": 0.3429011676778506,
"learning_rate": 1.991486224484123e-05,
"loss": 0.1939,
"step": 678
},
{
"epoch": 2.7534177215189874,
"grad_norm": 0.37163148253681577,
"learning_rate": 1.985810450327756e-05,
"loss": 0.1933,
"step": 679
},
{
"epoch": 2.7574683544303795,
"grad_norm": 0.3370788593441941,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.2001,
"step": 680
},
{
"epoch": 2.761518987341772,
"grad_norm": 0.37130116677512953,
"learning_rate": 1.974459290568053e-05,
"loss": 0.2039,
"step": 681
},
{
"epoch": 2.7655696202531646,
"grad_norm": 0.3185245226328395,
"learning_rate": 1.968783996385397e-05,
"loss": 0.1897,
"step": 682
},
{
"epoch": 2.769620253164557,
"grad_norm": 0.32564953218710885,
"learning_rate": 1.963108953612143e-05,
"loss": 0.1933,
"step": 683
},
{
"epoch": 2.7736708860759496,
"grad_norm": 0.3427014908232673,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.1825,
"step": 684
},
{
"epoch": 2.7777215189873417,
"grad_norm": 0.3191018842698593,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.1976,
"step": 685
},
{
"epoch": 2.781772151898734,
"grad_norm": 0.34825145846912864,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.2036,
"step": 686
},
{
"epoch": 2.7858227848101267,
"grad_norm": 0.3244018701651075,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.1904,
"step": 687
},
{
"epoch": 2.7898734177215188,
"grad_norm": 0.35895075994615866,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.1908,
"step": 688
},
{
"epoch": 2.7939240506329113,
"grad_norm": 0.3274689284889573,
"learning_rate": 1.92906653591874e-05,
"loss": 0.192,
"step": 689
},
{
"epoch": 2.797974683544304,
"grad_norm": 0.32881998020854425,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.2071,
"step": 690
},
{
"epoch": 2.8020253164556963,
"grad_norm": 0.3436804640881603,
"learning_rate": 1.917723146291022e-05,
"loss": 0.1929,
"step": 691
},
{
"epoch": 2.806075949367089,
"grad_norm": 0.3217314173985633,
"learning_rate": 1.912052422608584e-05,
"loss": 0.1842,
"step": 692
},
{
"epoch": 2.810126582278481,
"grad_norm": 0.3087309487820781,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.1957,
"step": 693
},
{
"epoch": 2.8141772151898734,
"grad_norm": 0.3471991433107877,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.1859,
"step": 694
},
{
"epoch": 2.818227848101266,
"grad_norm": 0.32959812983581516,
"learning_rate": 1.895044684123479e-05,
"loss": 0.192,
"step": 695
},
{
"epoch": 2.822278481012658,
"grad_norm": 0.3379641091745217,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.2067,
"step": 696
},
{
"epoch": 2.8263291139240505,
"grad_norm": 0.3655216983274233,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.21,
"step": 697
},
{
"epoch": 2.830379746835443,
"grad_norm": 0.33144079644422086,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.189,
"step": 698
},
{
"epoch": 2.8344303797468355,
"grad_norm": 0.32162406179840985,
"learning_rate": 1.872379746550743e-05,
"loss": 0.2039,
"step": 699
},
{
"epoch": 2.838481012658228,
"grad_norm": 0.3295989971035275,
"learning_rate": 1.866715967678769e-05,
"loss": 0.2096,
"step": 700
},
{
"epoch": 2.84253164556962,
"grad_norm": 0.32421014787513414,
"learning_rate": 1.861053262258067e-05,
"loss": 0.1996,
"step": 701
},
{
"epoch": 2.8465822784810126,
"grad_norm": 0.31728218222990895,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.2041,
"step": 702
},
{
"epoch": 2.850632911392405,
"grad_norm": 0.3316501238913074,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.1917,
"step": 703
},
{
"epoch": 2.8546835443037972,
"grad_norm": 0.34599416544707035,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.1854,
"step": 704
},
{
"epoch": 2.8587341772151897,
"grad_norm": 0.3150316788609285,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.1968,
"step": 705
},
{
"epoch": 2.8627848101265823,
"grad_norm": 0.32866505050107153,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.1922,
"step": 706
},
{
"epoch": 2.8668354430379748,
"grad_norm": 0.32944026266358134,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.1905,
"step": 707
},
{
"epoch": 2.8708860759493673,
"grad_norm": 0.31272694189304484,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.1867,
"step": 708
},
{
"epoch": 2.8749367088607594,
"grad_norm": 0.3226659546361858,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.2004,
"step": 709
},
{
"epoch": 2.878987341772152,
"grad_norm": 0.3212389789623033,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.204,
"step": 710
},
{
"epoch": 2.8830379746835444,
"grad_norm": 0.3083981456243947,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.1959,
"step": 711
},
{
"epoch": 2.8870886075949365,
"grad_norm": 0.31420858959334863,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.1838,
"step": 712
},
{
"epoch": 2.891139240506329,
"grad_norm": 0.3168318290000783,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.1931,
"step": 713
},
{
"epoch": 2.8951898734177215,
"grad_norm": 0.3192229159045572,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.185,
"step": 714
},
{
"epoch": 2.899240506329114,
"grad_norm": 0.33106585843795666,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.1807,
"step": 715
},
{
"epoch": 2.9032911392405065,
"grad_norm": 0.34050004340428125,
"learning_rate": 1.776272483901182e-05,
"loss": 0.1907,
"step": 716
},
{
"epoch": 2.9073417721518986,
"grad_norm": 0.3203765098266979,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.2015,
"step": 717
},
{
"epoch": 2.911392405063291,
"grad_norm": 0.32308168095848055,
"learning_rate": 1.764995661794301e-05,
"loss": 0.1896,
"step": 718
},
{
"epoch": 2.9154430379746836,
"grad_norm": 0.33152650515777093,
"learning_rate": 1.759360067078035e-05,
"loss": 0.2036,
"step": 719
},
{
"epoch": 2.9194936708860757,
"grad_norm": 0.31527557479419277,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.1906,
"step": 720
},
{
"epoch": 2.923544303797468,
"grad_norm": 0.3287943587469149,
"learning_rate": 1.748094737261012e-05,
"loss": 0.1859,
"step": 721
},
{
"epoch": 2.9275949367088607,
"grad_norm": 0.33022440919045126,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.1923,
"step": 722
},
{
"epoch": 2.9316455696202532,
"grad_norm": 0.3273818274741229,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.2035,
"step": 723
},
{
"epoch": 2.9356962025316458,
"grad_norm": 0.3361008073479113,
"learning_rate": 1.731212071922883e-05,
"loss": 0.1833,
"step": 724
},
{
"epoch": 2.939746835443038,
"grad_norm": 0.33746080201484985,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.1943,
"step": 725
},
{
"epoch": 2.9437974683544303,
"grad_norm": 0.30832310940666446,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.2012,
"step": 726
},
{
"epoch": 2.947848101265823,
"grad_norm": 0.34414394549308663,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.1939,
"step": 727
},
{
"epoch": 2.951898734177215,
"grad_norm": 0.33716860814166105,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.2007,
"step": 728
},
{
"epoch": 2.9559493670886074,
"grad_norm": 0.32748196991888767,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.1848,
"step": 729
},
{
"epoch": 2.96,
"grad_norm": 0.3587025003448406,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.1992,
"step": 730
},
{
"epoch": 2.9640506329113925,
"grad_norm": 0.32505854564382064,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.1993,
"step": 731
},
{
"epoch": 2.968101265822785,
"grad_norm": 0.3411239142959039,
"learning_rate": 1.686290200071915e-05,
"loss": 0.1922,
"step": 732
},
{
"epoch": 2.972151898734177,
"grad_norm": 0.36152885863378026,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.1985,
"step": 733
},
{
"epoch": 2.9762025316455696,
"grad_norm": 0.3318041624005512,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.1985,
"step": 734
},
{
"epoch": 2.980253164556962,
"grad_norm": 0.32960047909158846,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.208,
"step": 735
},
{
"epoch": 2.984303797468354,
"grad_norm": 0.3626133272561222,
"learning_rate": 1.663888455395778e-05,
"loss": 0.2019,
"step": 736
},
{
"epoch": 2.9883544303797467,
"grad_norm": 0.34359236034951335,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.19,
"step": 737
},
{
"epoch": 2.992405063291139,
"grad_norm": 0.33723582204964425,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.1954,
"step": 738
},
{
"epoch": 2.9964556962025317,
"grad_norm": 0.34317665192618396,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.189,
"step": 739
},
{
"epoch": 3.002025316455696,
"grad_norm": 0.364578582111282,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.1719,
"step": 740
},
{
"epoch": 3.0060759493670886,
"grad_norm": 0.37826608199136275,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.1273,
"step": 741
},
{
"epoch": 3.010126582278481,
"grad_norm": 0.30822314160056047,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.1197,
"step": 742
},
{
"epoch": 3.0141772151898736,
"grad_norm": 0.3725278132797627,
"learning_rate": 1.624791353909428e-05,
"loss": 0.1138,
"step": 743
},
{
"epoch": 3.0182278481012657,
"grad_norm": 0.46752905150960067,
"learning_rate": 1.619217781670663e-05,
"loss": 0.1205,
"step": 744
},
{
"epoch": 3.022278481012658,
"grad_norm": 0.4003654871007492,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.1136,
"step": 745
},
{
"epoch": 3.0263291139240507,
"grad_norm": 0.3230174428650925,
"learning_rate": 1.608079882360134e-05,
"loss": 0.1249,
"step": 746
},
{
"epoch": 3.030379746835443,
"grad_norm": 0.3835947056533741,
"learning_rate": 1.60251564499148e-05,
"loss": 0.125,
"step": 747
},
{
"epoch": 3.0344303797468353,
"grad_norm": 0.3613309670695289,
"learning_rate": 1.596954608907176e-05,
"loss": 0.1116,
"step": 748
},
{
"epoch": 3.038481012658228,
"grad_norm": 0.3003499965767677,
"learning_rate": 1.591396818895043e-05,
"loss": 0.1161,
"step": 749
},
{
"epoch": 3.0425316455696203,
"grad_norm": 0.3556765120604414,
"learning_rate": 1.585842319716759e-05,
"loss": 0.1078,
"step": 750
},
{
"epoch": 3.046582278481013,
"grad_norm": 0.3363863368383935,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.1118,
"step": 751
},
{
"epoch": 3.050632911392405,
"grad_norm": 0.3478524026827633,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.1176,
"step": 752
},
{
"epoch": 3.0546835443037974,
"grad_norm": 0.35543632630556166,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.1198,
"step": 753
},
{
"epoch": 3.05873417721519,
"grad_norm": 0.31534705402765273,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.1184,
"step": 754
},
{
"epoch": 3.0627848101265824,
"grad_norm": 0.3167751988616509,
"learning_rate": 1.558120751116291e-05,
"loss": 0.1227,
"step": 755
},
{
"epoch": 3.0668354430379745,
"grad_norm": 0.31834387006968107,
"learning_rate": 1.552586935426876e-05,
"loss": 0.1354,
"step": 756
},
{
"epoch": 3.070886075949367,
"grad_norm": 0.2910889292133272,
"learning_rate": 1.547056723140774e-05,
"loss": 0.1306,
"step": 757
},
{
"epoch": 3.0749367088607595,
"grad_norm": 0.3092544156386163,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.1053,
"step": 758
},
{
"epoch": 3.078987341772152,
"grad_norm": 0.32778213549561175,
"learning_rate": 1.536007286907411e-05,
"loss": 0.114,
"step": 759
},
{
"epoch": 3.083037974683544,
"grad_norm": 0.3207925796679365,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.1066,
"step": 760
},
{
"epoch": 3.0870886075949366,
"grad_norm": 0.34970117464242667,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.1144,
"step": 761
},
{
"epoch": 3.091139240506329,
"grad_norm": 0.32246002171269456,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.1211,
"step": 762
},
{
"epoch": 3.0951898734177217,
"grad_norm": 0.29210485704365124,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.1201,
"step": 763
},
{
"epoch": 3.0992405063291137,
"grad_norm": 0.3194507641344166,
"learning_rate": 1.508449870007656e-05,
"loss": 0.1338,
"step": 764
},
{
"epoch": 3.1032911392405063,
"grad_norm": 0.31306179989476807,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.1173,
"step": 765
},
{
"epoch": 3.1073417721518988,
"grad_norm": 0.29558818450208585,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.1242,
"step": 766
},
{
"epoch": 3.1113924050632913,
"grad_norm": 0.3169177003412419,
"learning_rate": 1.4919625713405e-05,
"loss": 0.1178,
"step": 767
},
{
"epoch": 3.1154430379746834,
"grad_norm": 0.32377596299554906,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.1262,
"step": 768
},
{
"epoch": 3.119493670886076,
"grad_norm": 0.31031456170577193,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.1204,
"step": 769
},
{
"epoch": 3.1235443037974684,
"grad_norm": 0.30215555108417513,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.1129,
"step": 770
},
{
"epoch": 3.127594936708861,
"grad_norm": 0.330827172105525,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.1219,
"step": 771
},
{
"epoch": 3.131645569620253,
"grad_norm": 0.30795383699352696,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.1058,
"step": 772
},
{
"epoch": 3.1356962025316455,
"grad_norm": 0.2840893697046477,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.1183,
"step": 773
},
{
"epoch": 3.139746835443038,
"grad_norm": 0.3206800657342683,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.1209,
"step": 774
},
{
"epoch": 3.1437974683544305,
"grad_norm": 0.3000758229909626,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.1134,
"step": 775
},
{
"epoch": 3.1478481012658226,
"grad_norm": 0.29715802932407587,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.1173,
"step": 776
},
{
"epoch": 3.151898734177215,
"grad_norm": 0.31332083684135187,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.1043,
"step": 777
},
{
"epoch": 3.1559493670886076,
"grad_norm": 0.3029244327910965,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.1167,
"step": 778
},
{
"epoch": 3.16,
"grad_norm": 0.3103950972646009,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.1168,
"step": 779
},
{
"epoch": 3.164050632911392,
"grad_norm": 0.31392595777153637,
"learning_rate": 1.420958430225303e-05,
"loss": 0.1192,
"step": 780
},
{
"epoch": 3.1681012658227847,
"grad_norm": 0.30521169788641356,
"learning_rate": 1.415527989393985e-05,
"loss": 0.1188,
"step": 781
},
{
"epoch": 3.1721518987341772,
"grad_norm": 0.31567462760621207,
"learning_rate": 1.410102255819891e-05,
"loss": 0.12,
"step": 782
},
{
"epoch": 3.1762025316455698,
"grad_norm": 0.29663468640100793,
"learning_rate": 1.404681273201131e-05,
"loss": 0.1288,
"step": 783
},
{
"epoch": 3.180253164556962,
"grad_norm": 0.2976388038799971,
"learning_rate": 1.399265085197556e-05,
"loss": 0.113,
"step": 784
},
{
"epoch": 3.1843037974683543,
"grad_norm": 0.29999609696795,
"learning_rate": 1.393853735430398e-05,
"loss": 0.1128,
"step": 785
},
{
"epoch": 3.188354430379747,
"grad_norm": 0.3068094916282679,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.1143,
"step": 786
},
{
"epoch": 3.1924050632911394,
"grad_norm": 0.289904648113877,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.1232,
"step": 787
},
{
"epoch": 3.1964556962025314,
"grad_norm": 0.32283901439917967,
"learning_rate": 1.377649151173163e-05,
"loss": 0.1158,
"step": 788
},
{
"epoch": 3.200506329113924,
"grad_norm": 0.31374782288394926,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.112,
"step": 789
},
{
"epoch": 3.2045569620253165,
"grad_norm": 0.29900826899863403,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.1086,
"step": 790
},
{
"epoch": 3.208607594936709,
"grad_norm": 0.2974269360968285,
"learning_rate": 1.361489677627324e-05,
"loss": 0.1347,
"step": 791
},
{
"epoch": 3.212658227848101,
"grad_norm": 0.28474619726713096,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.1209,
"step": 792
},
{
"epoch": 3.2167088607594936,
"grad_norm": 0.2919737240666102,
"learning_rate": 1.350742335331241e-05,
"loss": 0.1182,
"step": 793
},
{
"epoch": 3.220759493670886,
"grad_norm": 0.2905617878375585,
"learning_rate": 1.345376486102198e-05,
"loss": 0.115,
"step": 794
},
{
"epoch": 3.2248101265822786,
"grad_norm": 0.3108842916610003,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.1178,
"step": 795
},
{
"epoch": 3.2288607594936707,
"grad_norm": 0.3001851465038203,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.1138,
"step": 796
},
{
"epoch": 3.232911392405063,
"grad_norm": 0.31620198793886867,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.1176,
"step": 797
},
{
"epoch": 3.2369620253164557,
"grad_norm": 0.32849571174919223,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.1117,
"step": 798
},
{
"epoch": 3.2410126582278482,
"grad_norm": 0.30353856920539,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.1287,
"step": 799
},
{
"epoch": 3.2450632911392403,
"grad_norm": 0.29533877741132547,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.123,
"step": 800
},
{
"epoch": 3.249113924050633,
"grad_norm": 0.31435852561444766,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.1227,
"step": 801
},
{
"epoch": 3.2531645569620253,
"grad_norm": 0.30817309726742037,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.1211,
"step": 802
},
{
"epoch": 3.257215189873418,
"grad_norm": 0.30011663314500947,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.1171,
"step": 803
},
{
"epoch": 3.26126582278481,
"grad_norm": 0.31263422565731264,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.1244,
"step": 804
},
{
"epoch": 3.2653164556962024,
"grad_norm": 0.32682338486264423,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.1426,
"step": 805
},
{
"epoch": 3.269367088607595,
"grad_norm": 0.2944123677743002,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.1157,
"step": 806
},
{
"epoch": 3.2734177215189875,
"grad_norm": 0.3229895952331933,
"learning_rate": 1.276115892225764e-05,
"loss": 0.1163,
"step": 807
},
{
"epoch": 3.27746835443038,
"grad_norm": 0.28803907761265946,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.1108,
"step": 808
},
{
"epoch": 3.281518987341772,
"grad_norm": 0.2965487906405644,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.125,
"step": 809
},
{
"epoch": 3.2855696202531646,
"grad_norm": 0.31368992985816596,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.1148,
"step": 810
},
{
"epoch": 3.289620253164557,
"grad_norm": 0.29967969797115807,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.11,
"step": 811
},
{
"epoch": 3.293670886075949,
"grad_norm": 0.2927962464995236,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.1038,
"step": 812
},
{
"epoch": 3.2977215189873417,
"grad_norm": 0.3418917447650004,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.1211,
"step": 813
},
{
"epoch": 3.301772151898734,
"grad_norm": 0.3014212270452523,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.1221,
"step": 814
},
{
"epoch": 3.3058227848101267,
"grad_norm": 0.29949316057249176,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.1236,
"step": 815
},
{
"epoch": 3.309873417721519,
"grad_norm": 0.278575783641677,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.1049,
"step": 816
},
{
"epoch": 3.3139240506329113,
"grad_norm": 0.29101724113481986,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.1125,
"step": 817
},
{
"epoch": 3.317974683544304,
"grad_norm": 0.29192968045939566,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.1173,
"step": 818
},
{
"epoch": 3.3220253164556963,
"grad_norm": 0.2929521770215952,
"learning_rate": 1.213055323844297e-05,
"loss": 0.115,
"step": 819
},
{
"epoch": 3.3260759493670884,
"grad_norm": 0.28593047455470544,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.104,
"step": 820
},
{
"epoch": 3.330126582278481,
"grad_norm": 0.2839007876071549,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.119,
"step": 821
},
{
"epoch": 3.3341772151898734,
"grad_norm": 0.29370421666274243,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.1032,
"step": 822
},
{
"epoch": 3.338227848101266,
"grad_norm": 0.29274058276346115,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.1124,
"step": 823
},
{
"epoch": 3.3422784810126585,
"grad_norm": 0.29506814649229696,
"learning_rate": 1.187045262895488e-05,
"loss": 0.1162,
"step": 824
},
{
"epoch": 3.3463291139240505,
"grad_norm": 0.31387271117087273,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.1067,
"step": 825
},
{
"epoch": 3.350379746835443,
"grad_norm": 0.28698434389816685,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.1192,
"step": 826
},
{
"epoch": 3.3544303797468356,
"grad_norm": 0.29902520477862293,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.1259,
"step": 827
},
{
"epoch": 3.3584810126582276,
"grad_norm": 0.30021783406690017,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.1127,
"step": 828
},
{
"epoch": 3.36253164556962,
"grad_norm": 0.2781540251318242,
"learning_rate": 1.161198884952377e-05,
"loss": 0.11,
"step": 829
},
{
"epoch": 3.3665822784810127,
"grad_norm": 0.2996870416515317,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.1206,
"step": 830
},
{
"epoch": 3.370632911392405,
"grad_norm": 0.3009221411205192,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.1208,
"step": 831
},
{
"epoch": 3.3746835443037977,
"grad_norm": 0.2894928720265309,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.1096,
"step": 832
},
{
"epoch": 3.3787341772151898,
"grad_norm": 0.2980242206108984,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.1124,
"step": 833
},
{
"epoch": 3.3827848101265823,
"grad_norm": 0.29841438649573593,
"learning_rate": 1.135521394010506e-05,
"loss": 0.1112,
"step": 834
},
{
"epoch": 3.386835443037975,
"grad_norm": 0.2972349207512905,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.113,
"step": 835
},
{
"epoch": 3.390886075949367,
"grad_norm": 0.29736346311799033,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.1103,
"step": 836
},
{
"epoch": 3.3949367088607594,
"grad_norm": 0.29611111901719833,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.1209,
"step": 837
},
{
"epoch": 3.398987341772152,
"grad_norm": 0.2984199049105023,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.1063,
"step": 838
},
{
"epoch": 3.4030379746835444,
"grad_norm": 0.2887168694264156,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.1196,
"step": 839
},
{
"epoch": 3.407088607594937,
"grad_norm": 0.28581960070077445,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.1246,
"step": 840
},
{
"epoch": 3.411139240506329,
"grad_norm": 0.29644359284133603,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.1242,
"step": 841
},
{
"epoch": 3.4151898734177215,
"grad_norm": 0.30782699629593624,
"learning_rate": 1.0948015846744e-05,
"loss": 0.1096,
"step": 842
},
{
"epoch": 3.419240506329114,
"grad_norm": 0.29032059157903173,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.1145,
"step": 843
},
{
"epoch": 3.423291139240506,
"grad_norm": 0.2949972814596448,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.1,
"step": 844
},
{
"epoch": 3.4273417721518986,
"grad_norm": 0.29603752358936813,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.1049,
"step": 845
},
{
"epoch": 3.431392405063291,
"grad_norm": 0.2942749724686279,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.118,
"step": 846
},
{
"epoch": 3.4354430379746836,
"grad_norm": 0.29080419900008303,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.1088,
"step": 847
},
{
"epoch": 3.439493670886076,
"grad_norm": 0.29060774805525463,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.1083,
"step": 848
},
{
"epoch": 3.4435443037974682,
"grad_norm": 0.3609320898133826,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.1345,
"step": 849
},
{
"epoch": 3.4475949367088607,
"grad_norm": 0.2970619866193548,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.1144,
"step": 850
},
{
"epoch": 3.4516455696202533,
"grad_norm": 0.27608885267672933,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.11,
"step": 851
},
{
"epoch": 3.4556962025316453,
"grad_norm": 0.29456123465814166,
"learning_rate": 1.044560362975474e-05,
"loss": 0.1211,
"step": 852
},
{
"epoch": 3.459746835443038,
"grad_norm": 0.29686818828764044,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.1184,
"step": 853
},
{
"epoch": 3.4637974683544304,
"grad_norm": 0.2855555425288743,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.1295,
"step": 854
},
{
"epoch": 3.467848101265823,
"grad_norm": 0.29104570866116114,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.1214,
"step": 855
},
{
"epoch": 3.4718987341772154,
"grad_norm": 0.3054556737737249,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.1191,
"step": 856
},
{
"epoch": 3.4759493670886075,
"grad_norm": 0.2955018672640258,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.1153,
"step": 857
},
{
"epoch": 3.48,
"grad_norm": 0.2935044345072772,
"learning_rate": 1.014782400334433e-05,
"loss": 0.1221,
"step": 858
},
{
"epoch": 3.4840506329113925,
"grad_norm": 0.3024750144957813,
"learning_rate": 1.009846945771296e-05,
"loss": 0.1002,
"step": 859
},
{
"epoch": 3.4881012658227846,
"grad_norm": 0.2992649345266789,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.1206,
"step": 860
},
{
"epoch": 3.492151898734177,
"grad_norm": 0.2823950614106017,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.1119,
"step": 861
},
{
"epoch": 3.4962025316455696,
"grad_norm": 0.2822723849764097,
"learning_rate": 9.950885880977891e-06,
"loss": 0.1074,
"step": 862
},
{
"epoch": 3.500253164556962,
"grad_norm": 0.29707658533886105,
"learning_rate": 9.901852696139382e-06,
"loss": 0.1154,
"step": 863
},
{
"epoch": 3.5043037974683546,
"grad_norm": 0.2874098016793272,
"learning_rate": 9.852900840391027e-06,
"loss": 0.1336,
"step": 864
},
{
"epoch": 3.5083544303797467,
"grad_norm": 0.2822443220052716,
"learning_rate": 9.804030707984313e-06,
"loss": 0.1284,
"step": 865
},
{
"epoch": 3.512405063291139,
"grad_norm": 0.2854730404059136,
"learning_rate": 9.755242692512599e-06,
"loss": 0.1054,
"step": 866
},
{
"epoch": 3.5164556962025317,
"grad_norm": 0.2941136176272844,
"learning_rate": 9.70653718690782e-06,
"loss": 0.1059,
"step": 867
},
{
"epoch": 3.520506329113924,
"grad_norm": 0.28811481380355825,
"learning_rate": 9.657914583437454e-06,
"loss": 0.1223,
"step": 868
},
{
"epoch": 3.5245569620253163,
"grad_norm": 0.2850793381733345,
"learning_rate": 9.609375273701246e-06,
"loss": 0.1158,
"step": 869
},
{
"epoch": 3.528607594936709,
"grad_norm": 0.28195442209776866,
"learning_rate": 9.560919648628133e-06,
"loss": 0.132,
"step": 870
},
{
"epoch": 3.5326582278481014,
"grad_norm": 0.2851756200935532,
"learning_rate": 9.512548098473047e-06,
"loss": 0.1182,
"step": 871
},
{
"epoch": 3.536708860759494,
"grad_norm": 0.2936312603823942,
"learning_rate": 9.464261012813825e-06,
"loss": 0.1234,
"step": 872
},
{
"epoch": 3.540759493670886,
"grad_norm": 0.2721008655569761,
"learning_rate": 9.416058780547987e-06,
"loss": 0.1114,
"step": 873
},
{
"epoch": 3.5448101265822785,
"grad_norm": 0.28527657958057967,
"learning_rate": 9.367941789889714e-06,
"loss": 0.1109,
"step": 874
},
{
"epoch": 3.548860759493671,
"grad_norm": 0.28485659194417706,
"learning_rate": 9.319910428366607e-06,
"loss": 0.1083,
"step": 875
},
{
"epoch": 3.552911392405063,
"grad_norm": 0.2811459430691036,
"learning_rate": 9.271965082816667e-06,
"loss": 0.1246,
"step": 876
},
{
"epoch": 3.5569620253164556,
"grad_norm": 0.2789460778345869,
"learning_rate": 9.224106139385111e-06,
"loss": 0.101,
"step": 877
},
{
"epoch": 3.561012658227848,
"grad_norm": 0.2808262543758036,
"learning_rate": 9.176333983521291e-06,
"loss": 0.1139,
"step": 878
},
{
"epoch": 3.5650632911392406,
"grad_norm": 0.28787450648034835,
"learning_rate": 9.12864899997558e-06,
"loss": 0.1184,
"step": 879
},
{
"epoch": 3.569113924050633,
"grad_norm": 0.2882524428426226,
"learning_rate": 9.08105157279628e-06,
"loss": 0.1394,
"step": 880
},
{
"epoch": 3.573164556962025,
"grad_norm": 0.2777225549826923,
"learning_rate": 9.03354208532653e-06,
"loss": 0.1158,
"step": 881
},
{
"epoch": 3.5772151898734177,
"grad_norm": 0.2808150600879713,
"learning_rate": 8.986120920201205e-06,
"loss": 0.1011,
"step": 882
},
{
"epoch": 3.58126582278481,
"grad_norm": 0.29284832871329725,
"learning_rate": 8.938788459343852e-06,
"loss": 0.1041,
"step": 883
},
{
"epoch": 3.5853164556962023,
"grad_norm": 0.2913642072399025,
"learning_rate": 8.8915450839636e-06,
"loss": 0.1037,
"step": 884
},
{
"epoch": 3.589367088607595,
"grad_norm": 0.2790945913499418,
"learning_rate": 8.844391174552116e-06,
"loss": 0.1202,
"step": 885
},
{
"epoch": 3.5934177215189873,
"grad_norm": 0.28209588005128783,
"learning_rate": 8.797327110880479e-06,
"loss": 0.1181,
"step": 886
},
{
"epoch": 3.59746835443038,
"grad_norm": 0.28880045830437306,
"learning_rate": 8.750353271996206e-06,
"loss": 0.1003,
"step": 887
},
{
"epoch": 3.6015189873417723,
"grad_norm": 0.291740961240262,
"learning_rate": 8.703470036220132e-06,
"loss": 0.1054,
"step": 888
},
{
"epoch": 3.6055696202531644,
"grad_norm": 0.2910608569476958,
"learning_rate": 8.656677781143394e-06,
"loss": 0.1122,
"step": 889
},
{
"epoch": 3.609620253164557,
"grad_norm": 0.2797094463878048,
"learning_rate": 8.609976883624377e-06,
"loss": 0.1168,
"step": 890
},
{
"epoch": 3.6136708860759494,
"grad_norm": 0.2750134035007396,
"learning_rate": 8.563367719785698e-06,
"loss": 0.1168,
"step": 891
},
{
"epoch": 3.6177215189873415,
"grad_norm": 0.2845184449890979,
"learning_rate": 8.516850665011138e-06,
"loss": 0.1152,
"step": 892
},
{
"epoch": 3.621772151898734,
"grad_norm": 0.30736510606279804,
"learning_rate": 8.47042609394269e-06,
"loss": 0.1069,
"step": 893
},
{
"epoch": 3.6258227848101265,
"grad_norm": 0.2956653619099626,
"learning_rate": 8.424094380477432e-06,
"loss": 0.123,
"step": 894
},
{
"epoch": 3.629873417721519,
"grad_norm": 0.2817085232476072,
"learning_rate": 8.37785589776465e-06,
"loss": 0.1229,
"step": 895
},
{
"epoch": 3.6339240506329116,
"grad_norm": 0.29374406758159133,
"learning_rate": 8.331711018202694e-06,
"loss": 0.1105,
"step": 896
},
{
"epoch": 3.6379746835443036,
"grad_norm": 0.29081678859519644,
"learning_rate": 8.285660113436104e-06,
"loss": 0.1239,
"step": 897
},
{
"epoch": 3.642025316455696,
"grad_norm": 0.2892956481426453,
"learning_rate": 8.239703554352527e-06,
"loss": 0.1055,
"step": 898
},
{
"epoch": 3.6460759493670887,
"grad_norm": 0.2926170044057536,
"learning_rate": 8.193841711079775e-06,
"loss": 0.1146,
"step": 899
},
{
"epoch": 3.6501265822784807,
"grad_norm": 0.29136800051755124,
"learning_rate": 8.148074952982828e-06,
"loss": 0.1272,
"step": 900
},
{
"epoch": 3.6541772151898733,
"grad_norm": 0.28394100435169095,
"learning_rate": 8.102403648660859e-06,
"loss": 0.1108,
"step": 901
},
{
"epoch": 3.6582278481012658,
"grad_norm": 0.2833327165496927,
"learning_rate": 8.056828165944282e-06,
"loss": 0.1108,
"step": 902
},
{
"epoch": 3.6622784810126583,
"grad_norm": 0.2777915600001154,
"learning_rate": 8.011348871891762e-06,
"loss": 0.1202,
"step": 903
},
{
"epoch": 3.666329113924051,
"grad_norm": 0.2887793931242193,
"learning_rate": 7.965966132787287e-06,
"loss": 0.1049,
"step": 904
},
{
"epoch": 3.670379746835443,
"grad_norm": 0.28334247753935304,
"learning_rate": 7.920680314137189e-06,
"loss": 0.1073,
"step": 905
},
{
"epoch": 3.6744303797468354,
"grad_norm": 0.2826696185121108,
"learning_rate": 7.875491780667246e-06,
"loss": 0.1199,
"step": 906
},
{
"epoch": 3.678481012658228,
"grad_norm": 0.2752739716153221,
"learning_rate": 7.830400896319667e-06,
"loss": 0.1024,
"step": 907
},
{
"epoch": 3.68253164556962,
"grad_norm": 0.29539194344239633,
"learning_rate": 7.785408024250259e-06,
"loss": 0.1047,
"step": 908
},
{
"epoch": 3.6865822784810125,
"grad_norm": 0.2919291045895852,
"learning_rate": 7.74051352682542e-06,
"loss": 0.1163,
"step": 909
},
{
"epoch": 3.690632911392405,
"grad_norm": 0.2835801403586845,
"learning_rate": 7.695717765619257e-06,
"loss": 0.1123,
"step": 910
},
{
"epoch": 3.6946835443037975,
"grad_norm": 0.27747890275866977,
"learning_rate": 7.651021101410673e-06,
"loss": 0.1057,
"step": 911
},
{
"epoch": 3.69873417721519,
"grad_norm": 0.29135276065070065,
"learning_rate": 7.606423894180464e-06,
"loss": 0.1192,
"step": 912
},
{
"epoch": 3.702784810126582,
"grad_norm": 0.2816246486949913,
"learning_rate": 7.56192650310839e-06,
"loss": 0.1102,
"step": 913
},
{
"epoch": 3.7068354430379746,
"grad_norm": 0.29483157770742774,
"learning_rate": 7.517529286570349e-06,
"loss": 0.0997,
"step": 914
},
{
"epoch": 3.710886075949367,
"grad_norm": 0.2776730618428603,
"learning_rate": 7.473232602135387e-06,
"loss": 0.1113,
"step": 915
},
{
"epoch": 3.714936708860759,
"grad_norm": 0.27057237213929247,
"learning_rate": 7.429036806562935e-06,
"loss": 0.1092,
"step": 916
},
{
"epoch": 3.7189873417721517,
"grad_norm": 0.2930564569325932,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.1152,
"step": 917
},
{
"epoch": 3.7230379746835442,
"grad_norm": 0.3105195502883366,
"learning_rate": 7.340949304977567e-06,
"loss": 0.1257,
"step": 918
},
{
"epoch": 3.7270886075949368,
"grad_norm": 0.27243009496989407,
"learning_rate": 7.297058308409282e-06,
"loss": 0.1167,
"step": 919
},
{
"epoch": 3.7311392405063293,
"grad_norm": 0.2813165633954234,
"learning_rate": 7.25326961958704e-06,
"loss": 0.0984,
"step": 920
},
{
"epoch": 3.7351898734177214,
"grad_norm": 0.2744245992642221,
"learning_rate": 7.209583591178921e-06,
"loss": 0.1185,
"step": 921
},
{
"epoch": 3.739240506329114,
"grad_norm": 0.2864598138898537,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.1112,
"step": 922
},
{
"epoch": 3.7432911392405064,
"grad_norm": 0.2885379852457433,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.1024,
"step": 923
},
{
"epoch": 3.747341772151899,
"grad_norm": 0.2850654013882221,
"learning_rate": 7.079144982700909e-06,
"loss": 0.1082,
"step": 924
},
{
"epoch": 3.7513924050632914,
"grad_norm": 0.2830850781622,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.1107,
"step": 925
},
{
"epoch": 3.7554430379746835,
"grad_norm": 0.28456585435750326,
"learning_rate": 6.99270564069757e-06,
"loss": 0.1062,
"step": 926
},
{
"epoch": 3.759493670886076,
"grad_norm": 0.28276786428610473,
"learning_rate": 6.949642934304375e-06,
"loss": 0.1069,
"step": 927
},
{
"epoch": 3.7635443037974685,
"grad_norm": 0.28710546299514306,
"learning_rate": 6.906685333692871e-06,
"loss": 0.1059,
"step": 928
},
{
"epoch": 3.7675949367088606,
"grad_norm": 0.28953735062182506,
"learning_rate": 6.86383318483769e-06,
"loss": 0.1302,
"step": 929
},
{
"epoch": 3.771645569620253,
"grad_norm": 0.2790070383396659,
"learning_rate": 6.821086832864139e-06,
"loss": 0.1182,
"step": 930
},
{
"epoch": 3.7756962025316456,
"grad_norm": 0.28210857542390005,
"learning_rate": 6.77844662204546e-06,
"loss": 0.1133,
"step": 931
},
{
"epoch": 3.779746835443038,
"grad_norm": 0.2825722257043948,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.1199,
"step": 932
},
{
"epoch": 3.7837974683544306,
"grad_norm": 0.27380024318604806,
"learning_rate": 6.693485996688695e-06,
"loss": 0.1122,
"step": 933
},
{
"epoch": 3.7878481012658227,
"grad_norm": 0.2722114338024708,
"learning_rate": 6.651166266411801e-06,
"loss": 0.1226,
"step": 934
},
{
"epoch": 3.7918987341772152,
"grad_norm": 0.27947466659797454,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.1156,
"step": 935
},
{
"epoch": 3.7959493670886078,
"grad_norm": 0.27383720965878605,
"learning_rate": 6.566849674844711e-06,
"loss": 0.0931,
"step": 936
},
{
"epoch": 3.8,
"grad_norm": 0.28168978881931267,
"learning_rate": 6.524853492628747e-06,
"loss": 0.114,
"step": 937
},
{
"epoch": 3.8040506329113923,
"grad_norm": 0.2827814420453267,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.1149,
"step": 938
},
{
"epoch": 3.808101265822785,
"grad_norm": 0.28731904376225775,
"learning_rate": 6.441187046486648e-06,
"loss": 0.1005,
"step": 939
},
{
"epoch": 3.8121518987341774,
"grad_norm": 0.287261226748459,
"learning_rate": 6.399517456398567e-06,
"loss": 0.1005,
"step": 940
},
{
"epoch": 3.81620253164557,
"grad_norm": 0.2779512409425681,
"learning_rate": 6.357957402727164e-06,
"loss": 0.1171,
"step": 941
},
{
"epoch": 3.820253164556962,
"grad_norm": 0.2739006192095547,
"learning_rate": 6.316507220191395e-06,
"loss": 0.1206,
"step": 942
},
{
"epoch": 3.8243037974683545,
"grad_norm": 0.2767020851612418,
"learning_rate": 6.275167242625331e-06,
"loss": 0.1059,
"step": 943
},
{
"epoch": 3.828354430379747,
"grad_norm": 0.2719326273928105,
"learning_rate": 6.233937802975471e-06,
"loss": 0.1058,
"step": 944
},
{
"epoch": 3.832405063291139,
"grad_norm": 0.28052376226304815,
"learning_rate": 6.192819233298046e-06,
"loss": 0.0954,
"step": 945
},
{
"epoch": 3.8364556962025316,
"grad_norm": 0.28808496944059486,
"learning_rate": 6.151811864756383e-06,
"loss": 0.1038,
"step": 946
},
{
"epoch": 3.840506329113924,
"grad_norm": 0.2861729036595112,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.1125,
"step": 947
},
{
"epoch": 3.8445569620253166,
"grad_norm": 0.2776188155858491,
"learning_rate": 6.070132051252868e-06,
"loss": 0.1092,
"step": 948
},
{
"epoch": 3.848607594936709,
"grad_norm": 0.2879730149602658,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.1145,
"step": 949
},
{
"epoch": 3.852658227848101,
"grad_norm": 0.28512955166500326,
"learning_rate": 5.988900993811575e-06,
"loss": 0.1119,
"step": 950
},
{
"epoch": 3.8567088607594937,
"grad_norm": 0.2953927138289508,
"learning_rate": 5.948454566959363e-06,
"loss": 0.1211,
"step": 951
},
{
"epoch": 3.8607594936708862,
"grad_norm": 0.28968020262672184,
"learning_rate": 5.908121309322328e-06,
"loss": 0.107,
"step": 952
},
{
"epoch": 3.8648101265822783,
"grad_norm": 0.2802654938506691,
"learning_rate": 5.867901545738976e-06,
"loss": 0.105,
"step": 953
},
{
"epoch": 3.868860759493671,
"grad_norm": 0.2725221577477711,
"learning_rate": 5.827795600133774e-06,
"loss": 0.1075,
"step": 954
},
{
"epoch": 3.8729113924050633,
"grad_norm": 0.2764061223199379,
"learning_rate": 5.787803795514466e-06,
"loss": 0.0992,
"step": 955
},
{
"epoch": 3.876962025316456,
"grad_norm": 0.2924432016328262,
"learning_rate": 5.747926453969576e-06,
"loss": 0.1375,
"step": 956
},
{
"epoch": 3.8810126582278484,
"grad_norm": 0.28291824662374604,
"learning_rate": 5.708163896665708e-06,
"loss": 0.1239,
"step": 957
},
{
"epoch": 3.8850632911392404,
"grad_norm": 0.27194888891101704,
"learning_rate": 5.668516443845047e-06,
"loss": 0.1061,
"step": 958
},
{
"epoch": 3.889113924050633,
"grad_norm": 0.2759857319927484,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.1153,
"step": 959
},
{
"epoch": 3.8931645569620255,
"grad_norm": 0.28043211717224925,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.1174,
"step": 960
},
{
"epoch": 3.8972151898734175,
"grad_norm": 0.27303412566183066,
"learning_rate": 5.550267900783019e-06,
"loss": 0.1189,
"step": 961
},
{
"epoch": 3.90126582278481,
"grad_norm": 0.26316075582734066,
"learning_rate": 5.511084049737623e-06,
"loss": 0.1068,
"step": 962
},
{
"epoch": 3.9053164556962026,
"grad_norm": 0.2703517162147365,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.1297,
"step": 963
},
{
"epoch": 3.909367088607595,
"grad_norm": 0.2704267960355312,
"learning_rate": 5.433066737499948e-06,
"loss": 0.1146,
"step": 964
},
{
"epoch": 3.9134177215189876,
"grad_norm": 0.27739441253475783,
"learning_rate": 5.394233904648376e-06,
"loss": 0.1145,
"step": 965
},
{
"epoch": 3.9174683544303797,
"grad_norm": 0.26494956879576836,
"learning_rate": 5.355518704628997e-06,
"loss": 0.1306,
"step": 966
},
{
"epoch": 3.921518987341772,
"grad_norm": 0.28124528179286457,
"learning_rate": 5.316921449248731e-06,
"loss": 0.1224,
"step": 967
},
{
"epoch": 3.9255696202531647,
"grad_norm": 0.27793573030945606,
"learning_rate": 5.278442449364538e-06,
"loss": 0.1112,
"step": 968
},
{
"epoch": 3.9296202531645568,
"grad_norm": 0.2830689644972569,
"learning_rate": 5.240082014881016e-06,
"loss": 0.1224,
"step": 969
},
{
"epoch": 3.9336708860759493,
"grad_norm": 0.2824832692501287,
"learning_rate": 5.201840454747822e-06,
"loss": 0.1052,
"step": 970
},
{
"epoch": 3.937721518987342,
"grad_norm": 0.28332461722904734,
"learning_rate": 5.163718076957223e-06,
"loss": 0.1159,
"step": 971
},
{
"epoch": 3.9417721518987343,
"grad_norm": 0.28515903155733885,
"learning_rate": 5.125715188541609e-06,
"loss": 0.1082,
"step": 972
},
{
"epoch": 3.945822784810127,
"grad_norm": 0.288173586694882,
"learning_rate": 5.087832095571021e-06,
"loss": 0.1127,
"step": 973
},
{
"epoch": 3.949873417721519,
"grad_norm": 0.2910606018925096,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.1126,
"step": 974
},
{
"epoch": 3.9539240506329114,
"grad_norm": 0.27716649463348314,
"learning_rate": 5.01242651541854e-06,
"loss": 0.1144,
"step": 975
},
{
"epoch": 3.957974683544304,
"grad_norm": 0.27826425494083573,
"learning_rate": 4.974904635542815e-06,
"loss": 0.125,
"step": 976
},
{
"epoch": 3.962025316455696,
"grad_norm": 0.2710265848270513,
"learning_rate": 4.937503765719582e-06,
"loss": 0.1218,
"step": 977
},
{
"epoch": 3.9660759493670885,
"grad_norm": 0.26186742256793655,
"learning_rate": 4.900224207170299e-06,
"loss": 0.1192,
"step": 978
},
{
"epoch": 3.970126582278481,
"grad_norm": 0.2766540034847084,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.1001,
"step": 979
},
{
"epoch": 3.9741772151898735,
"grad_norm": 0.2728571750612267,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.112,
"step": 980
},
{
"epoch": 3.978227848101266,
"grad_norm": 0.27764707785842396,
"learning_rate": 4.789116396710924e-06,
"loss": 0.099,
"step": 981
},
{
"epoch": 3.982278481012658,
"grad_norm": 0.2855112980480514,
"learning_rate": 4.752325075895368e-06,
"loss": 0.1149,
"step": 982
},
{
"epoch": 3.9863291139240506,
"grad_norm": 0.2671140281138487,
"learning_rate": 4.715656557757473e-06,
"loss": 0.1085,
"step": 983
},
{
"epoch": 3.990379746835443,
"grad_norm": 0.2758292257220625,
"learning_rate": 4.679111137620442e-06,
"loss": 0.1273,
"step": 984
},
{
"epoch": 3.9944303797468352,
"grad_norm": 0.28663797705879096,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.1291,
"step": 985
},
{
"epoch": 3.9984810126582278,
"grad_norm": 0.28107296940321524,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.1089,
"step": 986
},
{
"epoch": 4.004050632911392,
"grad_norm": 0.35453803247281257,
"learning_rate": 4.570216403561141e-06,
"loss": 0.0825,
"step": 987
},
{
"epoch": 4.008101265822785,
"grad_norm": 0.297305125924895,
"learning_rate": 4.534166308795815e-06,
"loss": 0.0688,
"step": 988
},
{
"epoch": 4.012151898734177,
"grad_norm": 0.2670004493129719,
"learning_rate": 4.498240773728859e-06,
"loss": 0.0762,
"step": 989
},
{
"epoch": 4.01620253164557,
"grad_norm": 0.2157636993733985,
"learning_rate": 4.462440087699609e-06,
"loss": 0.0681,
"step": 990
},
{
"epoch": 4.020253164556962,
"grad_norm": 0.2371742888348774,
"learning_rate": 4.426764539041861e-06,
"loss": 0.0825,
"step": 991
},
{
"epoch": 4.024303797468354,
"grad_norm": 0.2589369063616861,
"learning_rate": 4.391214415081582e-06,
"loss": 0.0618,
"step": 992
},
{
"epoch": 4.028354430379747,
"grad_norm": 0.27467263624183147,
"learning_rate": 4.355790002134579e-06,
"loss": 0.0584,
"step": 993
},
{
"epoch": 4.032405063291139,
"grad_norm": 0.33723902878299294,
"learning_rate": 4.320491585504207e-06,
"loss": 0.0753,
"step": 994
},
{
"epoch": 4.036455696202531,
"grad_norm": 0.30023421363157843,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.0629,
"step": 995
},
{
"epoch": 4.040506329113924,
"grad_norm": 0.29543998163432816,
"learning_rate": 4.250273877330691e-06,
"loss": 0.0695,
"step": 996
},
{
"epoch": 4.044556962025316,
"grad_norm": 0.25494124902159226,
"learning_rate": 4.215355151311313e-06,
"loss": 0.0636,
"step": 997
},
{
"epoch": 4.048607594936709,
"grad_norm": 0.24578268745351972,
"learning_rate": 4.180563552651542e-06,
"loss": 0.0861,
"step": 998
},
{
"epoch": 4.052658227848101,
"grad_norm": 0.23892627072330516,
"learning_rate": 4.145899361558147e-06,
"loss": 0.0644,
"step": 999
},
{
"epoch": 4.056708860759493,
"grad_norm": 0.24743854420960723,
"learning_rate": 4.111362857211738e-06,
"loss": 0.0689,
"step": 1000
},
{
"epoch": 4.060759493670886,
"grad_norm": 0.2736209702392604,
"learning_rate": 4.076954317764592e-06,
"loss": 0.0667,
"step": 1001
},
{
"epoch": 4.0648101265822785,
"grad_norm": 0.24584365685321766,
"learning_rate": 4.042674020338335e-06,
"loss": 0.0622,
"step": 1002
},
{
"epoch": 4.0688607594936705,
"grad_norm": 0.227365519561464,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.0634,
"step": 1003
},
{
"epoch": 4.0729113924050635,
"grad_norm": 0.23186454764411193,
"learning_rate": 3.974499254868674e-06,
"loss": 0.0732,
"step": 1004
},
{
"epoch": 4.076962025316456,
"grad_norm": 0.22634319036843883,
"learning_rate": 3.940605335895451e-06,
"loss": 0.0779,
"step": 1005
},
{
"epoch": 4.0810126582278485,
"grad_norm": 0.22633343312846185,
"learning_rate": 3.90684075707908e-06,
"loss": 0.0716,
"step": 1006
},
{
"epoch": 4.085063291139241,
"grad_norm": 0.24384130215392502,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.0593,
"step": 1007
},
{
"epoch": 4.089113924050633,
"grad_norm": 0.24489522590434198,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.0633,
"step": 1008
},
{
"epoch": 4.093164556962026,
"grad_norm": 0.2577435175329827,
"learning_rate": 3.806325775702304e-06,
"loss": 0.0733,
"step": 1009
},
{
"epoch": 4.097215189873418,
"grad_norm": 0.2372001716708643,
"learning_rate": 3.773081266416434e-06,
"loss": 0.0634,
"step": 1010
},
{
"epoch": 4.10126582278481,
"grad_norm": 0.23019804947936193,
"learning_rate": 3.739967446503245e-06,
"loss": 0.0781,
"step": 1011
},
{
"epoch": 4.105316455696203,
"grad_norm": 0.23660108768771373,
"learning_rate": 3.706984582656894e-06,
"loss": 0.0717,
"step": 1012
},
{
"epoch": 4.109367088607595,
"grad_norm": 0.23017040640059652,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.0739,
"step": 1013
},
{
"epoch": 4.113417721518988,
"grad_norm": 0.22083756041574834,
"learning_rate": 3.641412784665648e-06,
"loss": 0.0688,
"step": 1014
},
{
"epoch": 4.11746835443038,
"grad_norm": 0.22332141758145926,
"learning_rate": 3.608824378627005e-06,
"loss": 0.0654,
"step": 1015
},
{
"epoch": 4.121518987341772,
"grad_norm": 0.2348397906243097,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.0821,
"step": 1016
},
{
"epoch": 4.125569620253165,
"grad_norm": 0.23204566625512701,
"learning_rate": 3.544043864774269e-06,
"loss": 0.0763,
"step": 1017
},
{
"epoch": 4.129620253164557,
"grad_norm": 0.23105275405244888,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.0604,
"step": 1018
},
{
"epoch": 4.133670886075949,
"grad_norm": 0.23251224724928327,
"learning_rate": 3.479793485887819e-06,
"loss": 0.0777,
"step": 1019
},
{
"epoch": 4.137721518987342,
"grad_norm": 0.2310162697728697,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.0642,
"step": 1020
},
{
"epoch": 4.141772151898734,
"grad_norm": 0.22083049452993553,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.0615,
"step": 1021
},
{
"epoch": 4.145822784810127,
"grad_norm": 0.23083121343291746,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.0808,
"step": 1022
},
{
"epoch": 4.149873417721519,
"grad_norm": 0.2214292747547024,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.0727,
"step": 1023
},
{
"epoch": 4.153924050632911,
"grad_norm": 0.23957664049744068,
"learning_rate": 3.321500420335e-06,
"loss": 0.0681,
"step": 1024
},
{
"epoch": 4.157974683544304,
"grad_norm": 0.23890830215074985,
"learning_rate": 3.290243771741275e-06,
"loss": 0.0695,
"step": 1025
},
{
"epoch": 4.162025316455696,
"grad_norm": 0.2288492622336535,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.0775,
"step": 1026
},
{
"epoch": 4.166075949367088,
"grad_norm": 0.2312917509579206,
"learning_rate": 3.228134459445149e-06,
"loss": 0.0626,
"step": 1027
},
{
"epoch": 4.170126582278481,
"grad_norm": 0.22201051919621134,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.0712,
"step": 1028
},
{
"epoch": 4.174177215189873,
"grad_norm": 0.22517246125925233,
"learning_rate": 3.166565459258513e-06,
"loss": 0.0782,
"step": 1029
},
{
"epoch": 4.178227848101266,
"grad_norm": 0.21723839773720163,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.0594,
"step": 1030
},
{
"epoch": 4.182278481012658,
"grad_norm": 0.2213051750722828,
"learning_rate": 3.105538754650419e-06,
"loss": 0.0616,
"step": 1031
},
{
"epoch": 4.18632911392405,
"grad_norm": 0.22660115957080967,
"learning_rate": 3.07522937824712e-06,
"loss": 0.0683,
"step": 1032
},
{
"epoch": 4.190379746835443,
"grad_norm": 0.21027958851467948,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.0755,
"step": 1033
},
{
"epoch": 4.194430379746835,
"grad_norm": 0.22078829040653,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.0766,
"step": 1034
},
{
"epoch": 4.1984810126582275,
"grad_norm": 0.22011928841234685,
"learning_rate": 2.985120078631465e-06,
"loss": 0.0835,
"step": 1035
},
{
"epoch": 4.2025316455696204,
"grad_norm": 0.23647956122704525,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.0695,
"step": 1036
},
{
"epoch": 4.2065822784810125,
"grad_norm": 0.23079352560202857,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.0645,
"step": 1037
},
{
"epoch": 4.2106329113924055,
"grad_norm": 0.22574726890238767,
"learning_rate": 2.896244091927678e-06,
"loss": 0.0752,
"step": 1038
},
{
"epoch": 4.2146835443037975,
"grad_norm": 0.22508123949066267,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.0699,
"step": 1039
},
{
"epoch": 4.21873417721519,
"grad_norm": 0.2211866203847552,
"learning_rate": 2.837681792951994e-06,
"loss": 0.0623,
"step": 1040
},
{
"epoch": 4.222784810126583,
"grad_norm": 0.22681992863035805,
"learning_rate": 2.808607860255981e-06,
"loss": 0.0784,
"step": 1041
},
{
"epoch": 4.226835443037975,
"grad_norm": 0.2169954587883081,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.0712,
"step": 1042
},
{
"epoch": 4.230886075949367,
"grad_norm": 0.2281788657061911,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.0693,
"step": 1043
},
{
"epoch": 4.23493670886076,
"grad_norm": 0.21514525479089228,
"learning_rate": 2.722217735873718e-06,
"loss": 0.0691,
"step": 1044
},
{
"epoch": 4.238987341772152,
"grad_norm": 0.22171235209436463,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.07,
"step": 1045
},
{
"epoch": 4.243037974683545,
"grad_norm": 0.23137375067661,
"learning_rate": 2.665319697490205e-06,
"loss": 0.0668,
"step": 1046
},
{
"epoch": 4.247088607594937,
"grad_norm": 0.22380885676215465,
"learning_rate": 2.637079980714945e-06,
"loss": 0.0669,
"step": 1047
},
{
"epoch": 4.251139240506329,
"grad_norm": 0.22851844810375663,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.0893,
"step": 1048
},
{
"epoch": 4.255189873417722,
"grad_norm": 0.23084320859221744,
"learning_rate": 2.581020289189571e-06,
"loss": 0.0732,
"step": 1049
},
{
"epoch": 4.259240506329114,
"grad_norm": 0.2208553049095809,
"learning_rate": 2.553200765936501e-06,
"loss": 0.0691,
"step": 1050
},
{
"epoch": 4.263291139240506,
"grad_norm": 0.22312535574754863,
"learning_rate": 2.525521756805962e-06,
"loss": 0.0559,
"step": 1051
},
{
"epoch": 4.267341772151899,
"grad_norm": 0.22739694898268092,
"learning_rate": 2.497983484720885e-06,
"loss": 0.0674,
"step": 1052
},
{
"epoch": 4.271392405063291,
"grad_norm": 0.23094253465567793,
"learning_rate": 2.470586171470728e-06,
"loss": 0.0776,
"step": 1053
},
{
"epoch": 4.275443037974684,
"grad_norm": 0.2038388529017639,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.0694,
"step": 1054
},
{
"epoch": 4.279493670886076,
"grad_norm": 0.22367699798336105,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.0735,
"step": 1055
},
{
"epoch": 4.283544303797468,
"grad_norm": 0.2290918253484516,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.0719,
"step": 1056
},
{
"epoch": 4.287594936708861,
"grad_norm": 0.22607971013611983,
"learning_rate": 2.362410902836978e-06,
"loss": 0.0652,
"step": 1057
},
{
"epoch": 4.291645569620253,
"grad_norm": 0.2201545204462739,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.0673,
"step": 1058
},
{
"epoch": 4.295696202531645,
"grad_norm": 0.22133204994201486,
"learning_rate": 2.309174704446411e-06,
"loss": 0.0681,
"step": 1059
},
{
"epoch": 4.299746835443038,
"grad_norm": 0.2230398030580475,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.076,
"step": 1060
},
{
"epoch": 4.30379746835443,
"grad_norm": 0.22314893473501218,
"learning_rate": 2.256508422807855e-06,
"loss": 0.0647,
"step": 1061
},
{
"epoch": 4.307848101265823,
"grad_norm": 0.23000347679775593,
"learning_rate": 2.230389531696946e-06,
"loss": 0.071,
"step": 1062
},
{
"epoch": 4.311898734177215,
"grad_norm": 0.23549940168556438,
"learning_rate": 2.204413754585857e-06,
"loss": 0.0813,
"step": 1063
},
{
"epoch": 4.315949367088607,
"grad_norm": 0.22687330949012066,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.0755,
"step": 1064
},
{
"epoch": 4.32,
"grad_norm": 0.22320754110999286,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.0655,
"step": 1065
},
{
"epoch": 4.324050632911392,
"grad_norm": 0.22532406504739158,
"learning_rate": 2.127347193531757e-06,
"loss": 0.065,
"step": 1066
},
{
"epoch": 4.328101265822784,
"grad_norm": 0.21387575521573446,
"learning_rate": 2.101945952921942e-06,
"loss": 0.0627,
"step": 1067
},
{
"epoch": 4.332151898734177,
"grad_norm": 0.2273201866439436,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.0762,
"step": 1068
},
{
"epoch": 4.3362025316455695,
"grad_norm": 0.22981792733973141,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.0639,
"step": 1069
},
{
"epoch": 4.340253164556962,
"grad_norm": 0.22732119069291692,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.0673,
"step": 1070
},
{
"epoch": 4.3443037974683545,
"grad_norm": 0.23295360893335987,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.0694,
"step": 1071
},
{
"epoch": 4.348354430379747,
"grad_norm": 0.22187883326277802,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.0785,
"step": 1072
},
{
"epoch": 4.3524050632911395,
"grad_norm": 0.2262325944787212,
"learning_rate": 1.95257270594031e-06,
"loss": 0.0546,
"step": 1073
},
{
"epoch": 4.356455696202532,
"grad_norm": 0.2295249722527472,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.0836,
"step": 1074
},
{
"epoch": 4.360506329113924,
"grad_norm": 0.23213513674247307,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.0619,
"step": 1075
},
{
"epoch": 4.364556962025317,
"grad_norm": 0.23392979980101164,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.0835,
"step": 1076
},
{
"epoch": 4.368607594936709,
"grad_norm": 0.22976227537922542,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.0831,
"step": 1077
},
{
"epoch": 4.372658227848102,
"grad_norm": 0.21569748059178515,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.0639,
"step": 1078
},
{
"epoch": 4.376708860759494,
"grad_norm": 0.22215398705180678,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.0728,
"step": 1079
},
{
"epoch": 4.380759493670886,
"grad_norm": 0.2205101548609658,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.064,
"step": 1080
},
{
"epoch": 4.384810126582279,
"grad_norm": 0.23035764622239405,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.0754,
"step": 1081
},
{
"epoch": 4.388860759493671,
"grad_norm": 0.22409563236857713,
"learning_rate": 1.738336950434061e-06,
"loss": 0.0697,
"step": 1082
},
{
"epoch": 4.392911392405063,
"grad_norm": 0.22065240545318782,
"learning_rate": 1.715265841375957e-06,
"loss": 0.0705,
"step": 1083
},
{
"epoch": 4.396962025316456,
"grad_norm": 0.22192015627730155,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.0553,
"step": 1084
},
{
"epoch": 4.401012658227848,
"grad_norm": 0.23101971847479424,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.0829,
"step": 1085
},
{
"epoch": 4.405063291139241,
"grad_norm": 0.2174627499536193,
"learning_rate": 1.646936827927441e-06,
"loss": 0.0738,
"step": 1086
},
{
"epoch": 4.409113924050633,
"grad_norm": 0.22483488931426932,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.0655,
"step": 1087
},
{
"epoch": 4.413164556962025,
"grad_norm": 0.22093412662372797,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.0641,
"step": 1088
},
{
"epoch": 4.417215189873418,
"grad_norm": 0.21729766717914653,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.0734,
"step": 1089
},
{
"epoch": 4.42126582278481,
"grad_norm": 0.21502914502326315,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.0605,
"step": 1090
},
{
"epoch": 4.425316455696202,
"grad_norm": 0.21468131790108883,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.0726,
"step": 1091
},
{
"epoch": 4.429367088607595,
"grad_norm": 0.22297383149422212,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.0603,
"step": 1092
},
{
"epoch": 4.433417721518987,
"grad_norm": 0.2396622121082945,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.0678,
"step": 1093
},
{
"epoch": 4.43746835443038,
"grad_norm": 0.22805990772700016,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.0674,
"step": 1094
},
{
"epoch": 4.441518987341772,
"grad_norm": 0.2250228365743954,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.062,
"step": 1095
},
{
"epoch": 4.445569620253164,
"grad_norm": 0.2098167712133337,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.0633,
"step": 1096
},
{
"epoch": 4.449620253164557,
"grad_norm": 0.22457029647154209,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.0671,
"step": 1097
},
{
"epoch": 4.453670886075949,
"grad_norm": 0.2134105245583983,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.0679,
"step": 1098
},
{
"epoch": 4.457721518987341,
"grad_norm": 0.2174327926805258,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.0681,
"step": 1099
},
{
"epoch": 4.461772151898734,
"grad_norm": 0.2258014530329191,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.0695,
"step": 1100
},
{
"epoch": 4.465822784810126,
"grad_norm": 0.22493584852627704,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.0658,
"step": 1101
},
{
"epoch": 4.469873417721519,
"grad_norm": 0.21992393634159466,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.0664,
"step": 1102
},
{
"epoch": 4.473924050632911,
"grad_norm": 0.24319356403345083,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.0809,
"step": 1103
},
{
"epoch": 4.4779746835443035,
"grad_norm": 0.21992115359359493,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.0634,
"step": 1104
},
{
"epoch": 4.4820253164556965,
"grad_norm": 0.22284761964339828,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.0748,
"step": 1105
},
{
"epoch": 4.4860759493670885,
"grad_norm": 0.22009442763892392,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.0599,
"step": 1106
},
{
"epoch": 4.490126582278481,
"grad_norm": 0.22201666994065775,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.0661,
"step": 1107
},
{
"epoch": 4.494177215189874,
"grad_norm": 0.2479230825724556,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.0701,
"step": 1108
},
{
"epoch": 4.498227848101266,
"grad_norm": 0.21882910288283533,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.0562,
"step": 1109
},
{
"epoch": 4.502278481012659,
"grad_norm": 0.2223727027832003,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.0782,
"step": 1110
},
{
"epoch": 4.506329113924051,
"grad_norm": 0.21156476362882204,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.0708,
"step": 1111
},
{
"epoch": 4.510379746835443,
"grad_norm": 0.21834246046996528,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.0714,
"step": 1112
},
{
"epoch": 4.514430379746836,
"grad_norm": 0.2235422018794454,
"learning_rate": 1.092402023876933e-06,
"loss": 0.0725,
"step": 1113
},
{
"epoch": 4.518481012658228,
"grad_norm": 0.21855730310026103,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.0832,
"step": 1114
},
{
"epoch": 4.52253164556962,
"grad_norm": 0.2151072422844024,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.0713,
"step": 1115
},
{
"epoch": 4.526582278481013,
"grad_norm": 0.22142033585585572,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.0775,
"step": 1116
},
{
"epoch": 4.530632911392405,
"grad_norm": 0.2138907227233124,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.0609,
"step": 1117
},
{
"epoch": 4.534683544303798,
"grad_norm": 0.22513179825895346,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.0687,
"step": 1118
},
{
"epoch": 4.53873417721519,
"grad_norm": 0.2275003048324263,
"learning_rate": 9.841383857091947e-07,
"loss": 0.0634,
"step": 1119
},
{
"epoch": 4.542784810126582,
"grad_norm": 0.22147887715364564,
"learning_rate": 9.666296380321616e-07,
"loss": 0.0792,
"step": 1120
},
{
"epoch": 4.546835443037975,
"grad_norm": 0.22874096674339675,
"learning_rate": 9.492741825034124e-07,
"loss": 0.0702,
"step": 1121
},
{
"epoch": 4.550886075949367,
"grad_norm": 0.22255392908976826,
"learning_rate": 9.320721589013892e-07,
"loss": 0.0668,
"step": 1122
},
{
"epoch": 4.55493670886076,
"grad_norm": 0.22273968462179783,
"learning_rate": 9.150237057688339e-07,
"loss": 0.0711,
"step": 1123
},
{
"epoch": 4.558987341772152,
"grad_norm": 0.2293852392692537,
"learning_rate": 8.981289604116328e-07,
"loss": 0.0733,
"step": 1124
},
{
"epoch": 4.563037974683544,
"grad_norm": 0.22576290810509123,
"learning_rate": 8.813880588977542e-07,
"loss": 0.0682,
"step": 1125
},
{
"epoch": 4.567088607594937,
"grad_norm": 0.22680098136710333,
"learning_rate": 8.648011360561126e-07,
"loss": 0.0677,
"step": 1126
},
{
"epoch": 4.571139240506329,
"grad_norm": 0.22501637252735035,
"learning_rate": 8.483683254755037e-07,
"loss": 0.0685,
"step": 1127
},
{
"epoch": 4.575189873417721,
"grad_norm": 0.20969936296885303,
"learning_rate": 8.320897595035227e-07,
"loss": 0.0651,
"step": 1128
},
{
"epoch": 4.579240506329114,
"grad_norm": 0.21520714164569826,
"learning_rate": 8.159655692455093e-07,
"loss": 0.0629,
"step": 1129
},
{
"epoch": 4.583291139240506,
"grad_norm": 0.23317437231085383,
"learning_rate": 7.999958845634648e-07,
"loss": 0.0664,
"step": 1130
},
{
"epoch": 4.587341772151898,
"grad_norm": 0.22431874246484673,
"learning_rate": 7.841808340750478e-07,
"loss": 0.066,
"step": 1131
},
{
"epoch": 4.591392405063291,
"grad_norm": 0.22172964213861804,
"learning_rate": 7.685205451524869e-07,
"loss": 0.0703,
"step": 1132
},
{
"epoch": 4.595443037974683,
"grad_norm": 0.21472912875827047,
"learning_rate": 7.530151439216027e-07,
"loss": 0.0687,
"step": 1133
},
{
"epoch": 4.599493670886076,
"grad_norm": 0.22914041898241294,
"learning_rate": 7.376647552607675e-07,
"loss": 0.0674,
"step": 1134
},
{
"epoch": 4.603544303797468,
"grad_norm": 0.21959068989367253,
"learning_rate": 7.224695027998963e-07,
"loss": 0.066,
"step": 1135
},
{
"epoch": 4.6075949367088604,
"grad_norm": 0.2200968025566797,
"learning_rate": 7.07429508919466e-07,
"loss": 0.0816,
"step": 1136
},
{
"epoch": 4.611645569620253,
"grad_norm": 0.2172309966519211,
"learning_rate": 6.925448947495206e-07,
"loss": 0.0563,
"step": 1137
},
{
"epoch": 4.6156962025316455,
"grad_norm": 0.21861730289322728,
"learning_rate": 6.778157801686936e-07,
"loss": 0.0617,
"step": 1138
},
{
"epoch": 4.619746835443038,
"grad_norm": 0.22685540119067424,
"learning_rate": 6.632422838032515e-07,
"loss": 0.0654,
"step": 1139
},
{
"epoch": 4.6237974683544305,
"grad_norm": 0.2150737634790382,
"learning_rate": 6.488245230261281e-07,
"loss": 0.0559,
"step": 1140
},
{
"epoch": 4.627848101265823,
"grad_norm": 0.236467310871022,
"learning_rate": 6.345626139559868e-07,
"loss": 0.0762,
"step": 1141
},
{
"epoch": 4.6318987341772155,
"grad_norm": 0.22990572174015875,
"learning_rate": 6.204566714562866e-07,
"loss": 0.0741,
"step": 1142
},
{
"epoch": 4.635949367088608,
"grad_norm": 0.22072410971702014,
"learning_rate": 6.06506809134344e-07,
"loss": 0.0872,
"step": 1143
},
{
"epoch": 4.64,
"grad_norm": 0.22107206017868603,
"learning_rate": 5.927131393404373e-07,
"loss": 0.067,
"step": 1144
},
{
"epoch": 4.644050632911393,
"grad_norm": 0.2179876525659186,
"learning_rate": 5.790757731668817e-07,
"loss": 0.0737,
"step": 1145
},
{
"epoch": 4.648101265822785,
"grad_norm": 0.21792594191725986,
"learning_rate": 5.655948204471507e-07,
"loss": 0.0777,
"step": 1146
},
{
"epoch": 4.652151898734177,
"grad_norm": 0.2164586166212343,
"learning_rate": 5.522703897549875e-07,
"loss": 0.0676,
"step": 1147
},
{
"epoch": 4.65620253164557,
"grad_norm": 0.21312145545227754,
"learning_rate": 5.391025884035239e-07,
"loss": 0.0611,
"step": 1148
},
{
"epoch": 4.660253164556962,
"grad_norm": 0.21776710454916795,
"learning_rate": 5.260915224444207e-07,
"loss": 0.0792,
"step": 1149
},
{
"epoch": 4.664303797468355,
"grad_norm": 0.21956783551660028,
"learning_rate": 5.132372966670129e-07,
"loss": 0.0731,
"step": 1150
},
{
"epoch": 4.668354430379747,
"grad_norm": 0.21581514987758,
"learning_rate": 5.005400145974704e-07,
"loss": 0.0579,
"step": 1151
},
{
"epoch": 4.672405063291139,
"grad_norm": 0.22481974607600783,
"learning_rate": 4.879997784979562e-07,
"loss": 0.0667,
"step": 1152
},
{
"epoch": 4.676455696202532,
"grad_norm": 0.2166558131071334,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.068,
"step": 1153
},
{
"epoch": 4.680506329113924,
"grad_norm": 0.21959159460837505,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.0664,
"step": 1154
},
{
"epoch": 4.684556962025317,
"grad_norm": 0.2247905538405723,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.075,
"step": 1155
},
{
"epoch": 4.688607594936709,
"grad_norm": 0.228424353442305,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.076,
"step": 1156
},
{
"epoch": 4.692658227848101,
"grad_norm": 0.22648515316086487,
"learning_rate": 4.27657778143431e-07,
"loss": 0.0735,
"step": 1157
},
{
"epoch": 4.696708860759494,
"grad_norm": 0.21546096999349432,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.0555,
"step": 1158
},
{
"epoch": 4.700759493670886,
"grad_norm": 0.22165425151791968,
"learning_rate": 4.046237371675177e-07,
"loss": 0.074,
"step": 1159
},
{
"epoch": 4.704810126582278,
"grad_norm": 0.22201876429282694,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.067,
"step": 1160
},
{
"epoch": 4.708860759493671,
"grad_norm": 0.22999015174743445,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.066,
"step": 1161
},
{
"epoch": 4.712911392405063,
"grad_norm": 0.21322675534833252,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.0836,
"step": 1162
},
{
"epoch": 4.716962025316455,
"grad_norm": 0.22107707869809107,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.0674,
"step": 1163
},
{
"epoch": 4.721012658227848,
"grad_norm": 0.215387404359144,
"learning_rate": 3.498020255137813e-07,
"loss": 0.0736,
"step": 1164
},
{
"epoch": 4.72506329113924,
"grad_norm": 0.2044784672478571,
"learning_rate": 3.393121163057811e-07,
"loss": 0.066,
"step": 1165
},
{
"epoch": 4.729113924050633,
"grad_norm": 0.21795895672251336,
"learning_rate": 3.289805515751399e-07,
"loss": 0.0705,
"step": 1166
},
{
"epoch": 4.733164556962025,
"grad_norm": 0.22168152246849313,
"learning_rate": 3.188074145308573e-07,
"loss": 0.0616,
"step": 1167
},
{
"epoch": 4.737215189873417,
"grad_norm": 0.21389687169150703,
"learning_rate": 3.087927871059804e-07,
"loss": 0.0585,
"step": 1168
},
{
"epoch": 4.74126582278481,
"grad_norm": 0.21822658442176945,
"learning_rate": 2.989367499569418e-07,
"loss": 0.0645,
"step": 1169
},
{
"epoch": 4.745316455696202,
"grad_norm": 0.21821258933708348,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.0795,
"step": 1170
},
{
"epoch": 4.749367088607595,
"grad_norm": 0.22240850931945622,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.0617,
"step": 1171
},
{
"epoch": 4.7534177215189874,
"grad_norm": 0.22643885550193701,
"learning_rate": 2.703209675663887e-07,
"loss": 0.0718,
"step": 1172
},
{
"epoch": 4.7574683544303795,
"grad_norm": 0.22128939123902958,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.0589,
"step": 1173
},
{
"epoch": 4.7615189873417725,
"grad_norm": 0.2231361564835323,
"learning_rate": 2.520381518804471e-07,
"loss": 0.0729,
"step": 1174
},
{
"epoch": 4.7655696202531646,
"grad_norm": 0.2448456130779054,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.0865,
"step": 1175
},
{
"epoch": 4.769620253164557,
"grad_norm": 0.22512591493539014,
"learning_rate": 2.343915243930317e-07,
"loss": 0.0668,
"step": 1176
},
{
"epoch": 4.77367088607595,
"grad_norm": 0.22714262024133994,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.0626,
"step": 1177
},
{
"epoch": 4.777721518987342,
"grad_norm": 0.21710827773923477,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.0607,
"step": 1178
},
{
"epoch": 4.781772151898734,
"grad_norm": 0.21864842681082122,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.0727,
"step": 1179
},
{
"epoch": 4.785822784810127,
"grad_norm": 0.2177485715820106,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.0696,
"step": 1180
},
{
"epoch": 4.789873417721519,
"grad_norm": 0.2252110193461747,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.0605,
"step": 1181
},
{
"epoch": 4.793924050632912,
"grad_norm": 0.2257767423549085,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.0637,
"step": 1182
},
{
"epoch": 4.797974683544304,
"grad_norm": 0.22799143185751888,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.0793,
"step": 1183
},
{
"epoch": 4.802025316455696,
"grad_norm": 0.22542225309498892,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.0562,
"step": 1184
},
{
"epoch": 4.806075949367089,
"grad_norm": 0.21464786271682817,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.0699,
"step": 1185
},
{
"epoch": 4.810126582278481,
"grad_norm": 0.21716081074593244,
"learning_rate": 1.557203888516745e-07,
"loss": 0.0575,
"step": 1186
},
{
"epoch": 4.814177215189874,
"grad_norm": 0.24218922000996887,
"learning_rate": 1.487313098987131e-07,
"loss": 0.073,
"step": 1187
},
{
"epoch": 4.818227848101266,
"grad_norm": 0.22234201494121972,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.0643,
"step": 1188
},
{
"epoch": 4.822278481012658,
"grad_norm": 0.21473646261434187,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.0652,
"step": 1189
},
{
"epoch": 4.826329113924051,
"grad_norm": 0.22408460900082225,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.0578,
"step": 1190
},
{
"epoch": 4.830379746835443,
"grad_norm": 0.22822585788636587,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.0691,
"step": 1191
},
{
"epoch": 4.834430379746835,
"grad_norm": 0.22119255001331226,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.0737,
"step": 1192
},
{
"epoch": 4.838481012658228,
"grad_norm": 0.22040394568132232,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.0646,
"step": 1193
},
{
"epoch": 4.84253164556962,
"grad_norm": 0.2815173858818427,
"learning_rate": 1.042873700093061e-07,
"loss": 0.0673,
"step": 1194
},
{
"epoch": 4.846582278481012,
"grad_norm": 0.22316523476926156,
"learning_rate": 9.857879404556291e-08,
"loss": 0.068,
"step": 1195
},
{
"epoch": 4.850632911392405,
"grad_norm": 0.2200014266641066,
"learning_rate": 9.303050139225722e-08,
"loss": 0.0779,
"step": 1196
},
{
"epoch": 4.854683544303797,
"grad_norm": 0.21672538729383267,
"learning_rate": 8.76425367345779e-08,
"loss": 0.0623,
"step": 1197
},
{
"epoch": 4.85873417721519,
"grad_norm": 0.21403546879517266,
"learning_rate": 8.241494346644897e-08,
"loss": 0.0728,
"step": 1198
},
{
"epoch": 4.862784810126582,
"grad_norm": 0.21746236323509188,
"learning_rate": 7.734776369019204e-08,
"loss": 0.0635,
"step": 1199
},
{
"epoch": 4.866835443037974,
"grad_norm": 0.22423047889385916,
"learning_rate": 7.244103821617332e-08,
"loss": 0.0622,
"step": 1200
},
{
"epoch": 4.870886075949367,
"grad_norm": 0.23156649246691058,
"learning_rate": 6.769480656248606e-08,
"loss": 0.0695,
"step": 1201
},
{
"epoch": 4.874936708860759,
"grad_norm": 0.21280074025863238,
"learning_rate": 6.310910695462635e-08,
"loss": 0.0685,
"step": 1202
},
{
"epoch": 4.878987341772152,
"grad_norm": 0.22488973354953173,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.0703,
"step": 1203
},
{
"epoch": 4.883037974683544,
"grad_norm": 0.23054766590438863,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.0606,
"step": 1204
},
{
"epoch": 4.8870886075949365,
"grad_norm": 0.21489785395549324,
"learning_rate": 5.031556326567488e-08,
"loss": 0.0609,
"step": 1205
},
{
"epoch": 4.891139240506329,
"grad_norm": 0.23157148835153135,
"learning_rate": 4.637234823364312e-08,
"loss": 0.0791,
"step": 1206
},
{
"epoch": 4.8951898734177215,
"grad_norm": 0.22887058893465576,
"learning_rate": 4.258983697558838e-08,
"loss": 0.065,
"step": 1207
},
{
"epoch": 4.899240506329114,
"grad_norm": 0.21916283088278193,
"learning_rate": 3.896805995533548e-08,
"loss": 0.0793,
"step": 1208
},
{
"epoch": 4.9032911392405065,
"grad_norm": 0.22302660417651649,
"learning_rate": 3.550704634218028e-08,
"loss": 0.0811,
"step": 1209
},
{
"epoch": 4.907341772151899,
"grad_norm": 0.21472149195392512,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.0688,
"step": 1210
},
{
"epoch": 4.911392405063291,
"grad_norm": 0.21688349585665231,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.0748,
"step": 1211
},
{
"epoch": 4.915443037974684,
"grad_norm": 0.22954061164424927,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.0787,
"step": 1212
},
{
"epoch": 4.919493670886076,
"grad_norm": 0.2272912322110633,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.0778,
"step": 1213
},
{
"epoch": 4.923544303797469,
"grad_norm": 0.2304091671284524,
"learning_rate": 2.061435966221881e-08,
"loss": 0.0772,
"step": 1214
},
{
"epoch": 4.927594936708861,
"grad_norm": 0.21562930558016008,
"learning_rate": 1.811846652477245e-08,
"loss": 0.0689,
"step": 1215
},
{
"epoch": 4.931645569620253,
"grad_norm": 0.21876113778415987,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.0703,
"step": 1216
},
{
"epoch": 4.935696202531646,
"grad_norm": 0.2256191168146549,
"learning_rate": 1.360949303591097e-08,
"loss": 0.067,
"step": 1217
},
{
"epoch": 4.939746835443038,
"grad_norm": 0.22336613343565107,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.0766,
"step": 1218
},
{
"epoch": 4.943797468354431,
"grad_norm": 0.2213389940950149,
"learning_rate": 9.744388816668172e-09,
"loss": 0.0571,
"step": 1219
},
{
"epoch": 4.947848101265823,
"grad_norm": 0.21345550737442745,
"learning_rate": 8.05332740472009e-09,
"loss": 0.0673,
"step": 1220
},
{
"epoch": 4.951898734177215,
"grad_norm": 0.21831460107482004,
"learning_rate": 6.523278382872811e-09,
"loss": 0.0542,
"step": 1221
},
{
"epoch": 4.955949367088608,
"grad_norm": 0.21328142433745323,
"learning_rate": 5.15425407393133e-09,
"loss": 0.0746,
"step": 1222
},
{
"epoch": 4.96,
"grad_norm": 0.22228387368199812,
"learning_rate": 3.94626550383137e-09,
"loss": 0.0645,
"step": 1223
},
{
"epoch": 4.964050632911392,
"grad_norm": 0.2267888057226826,
"learning_rate": 2.899322401546112e-09,
"loss": 0.0719,
"step": 1224
},
{
"epoch": 4.968101265822785,
"grad_norm": 0.22577682121152526,
"learning_rate": 2.013433199010706e-09,
"loss": 0.0834,
"step": 1225
},
{
"epoch": 4.972151898734177,
"grad_norm": 0.21611049857598746,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.084,
"step": 1226
},
{
"epoch": 4.976202531645569,
"grad_norm": 0.21666378732130187,
"learning_rate": 7.248437353468695e-10,
"loss": 0.0553,
"step": 1227
},
{
"epoch": 4.980253164556962,
"grad_norm": 0.21439060373592603,
"learning_rate": 3.221538523412449e-10,
"loss": 0.0752,
"step": 1228
},
{
"epoch": 4.984303797468354,
"grad_norm": 0.22239764971109358,
"learning_rate": 8.053862524670663e-11,
"loss": 0.0674,
"step": 1229
},
{
"epoch": 4.988354430379747,
"grad_norm": 0.23001369655179305,
"learning_rate": 0.0,
"loss": 0.0667,
"step": 1230
},
{
"epoch": 4.988354430379747,
"step": 1230,
"total_flos": 1.3796467879622738e+18,
"train_loss": 0.21418862139609288,
"train_runtime": 44832.7347,
"train_samples_per_second": 3.524,
"train_steps_per_second": 0.027
}
],
"logging_steps": 1,
"max_steps": 1230,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.3796467879622738e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}