ryanmarten's picture
End of training
9277add 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": 10.004693427295209,
"learning_rate": 3.2520325203252037e-07,
"loss": 1.3796,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 9.86824324084042,
"learning_rate": 6.504065040650407e-07,
"loss": 1.3736,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 9.923304216153873,
"learning_rate": 9.75609756097561e-07,
"loss": 1.3944,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 9.793642499054954,
"learning_rate": 1.3008130081300815e-06,
"loss": 1.3931,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 9.419092600462598,
"learning_rate": 1.6260162601626018e-06,
"loss": 1.3744,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 8.797831457394636,
"learning_rate": 1.951219512195122e-06,
"loss": 1.3404,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 7.22515226309723,
"learning_rate": 2.2764227642276426e-06,
"loss": 1.2931,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 6.577760144475995,
"learning_rate": 2.601626016260163e-06,
"loss": 1.275,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 4.220532361929984,
"learning_rate": 2.926829268292683e-06,
"loss": 1.207,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 3.6246036371581205,
"learning_rate": 3.2520325203252037e-06,
"loss": 1.1909,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 3.306170448800951,
"learning_rate": 3.577235772357724e-06,
"loss": 1.1986,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 3.9933908806496863,
"learning_rate": 3.902439024390244e-06,
"loss": 1.1309,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 4.326650044251223,
"learning_rate": 4.227642276422765e-06,
"loss": 1.1351,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 4.118753200999722,
"learning_rate": 4.552845528455285e-06,
"loss": 1.12,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 3.6636915643749215,
"learning_rate": 4.8780487804878055e-06,
"loss": 1.0632,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 2.807903023030099,
"learning_rate": 5.203252032520326e-06,
"loss": 1.0441,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 3.12167853471823,
"learning_rate": 5.528455284552846e-06,
"loss": 1.0396,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.6920502436055544,
"learning_rate": 5.853658536585366e-06,
"loss": 1.0281,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 2.110053954053194,
"learning_rate": 6.178861788617887e-06,
"loss": 0.9797,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.6200153266472799,
"learning_rate": 6.504065040650407e-06,
"loss": 0.9735,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.7193397639284709,
"learning_rate": 6.829268292682928e-06,
"loss": 0.9566,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.7935967402711563,
"learning_rate": 7.154471544715448e-06,
"loss": 0.9382,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.5974154146862103,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.9587,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 1.264887223753876,
"learning_rate": 7.804878048780489e-06,
"loss": 0.9386,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 1.213715774970272,
"learning_rate": 8.130081300813009e-06,
"loss": 0.9144,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 1.2179873080474801,
"learning_rate": 8.45528455284553e-06,
"loss": 0.9178,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 1.1273819110388876,
"learning_rate": 8.78048780487805e-06,
"loss": 0.9357,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 1.0011251177027147,
"learning_rate": 9.10569105691057e-06,
"loss": 0.9144,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 1.0648398326216078,
"learning_rate": 9.43089430894309e-06,
"loss": 0.9067,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 0.9733100783913939,
"learning_rate": 9.756097560975611e-06,
"loss": 0.9241,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 1.0250130673790656,
"learning_rate": 1.008130081300813e-05,
"loss": 0.9211,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.8982775376617136,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.9273,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.8931317570630534,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.9035,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.95573909870034,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.8999,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.9560376285108826,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.884,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 1.06431172169991,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.8765,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 1.0317381161517236,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.8633,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 1.0110471024135048,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.8742,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 1.0008996119617743,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.8753,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 1.0744072380891352,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.8689,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 1.1078001237810433,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.8885,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 1.0957664516912617,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.8543,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.9520531067941728,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.873,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 1.07101089527125,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.8816,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.9178644017979282,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.8664,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.9865079744374416,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.8786,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.9912767524862593,
"learning_rate": 1.528455284552846e-05,
"loss": 0.8823,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.9615116606526241,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.8605,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.9008688003437514,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.848,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.8943018769978403,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.8487,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.8851805530582184,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.8495,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 1.0317745142136727,
"learning_rate": 1.691056910569106e-05,
"loss": 0.841,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.8924396327835862,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.8409,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.9454194129395025,
"learning_rate": 1.75609756097561e-05,
"loss": 0.8493,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.987497952745904,
"learning_rate": 1.788617886178862e-05,
"loss": 0.8579,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.9240735794902795,
"learning_rate": 1.821138211382114e-05,
"loss": 0.8236,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.9569664446128245,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.843,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 1.0545201558454844,
"learning_rate": 1.886178861788618e-05,
"loss": 0.8418,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.8743989154316584,
"learning_rate": 1.91869918699187e-05,
"loss": 0.8323,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.997776780736207,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.8473,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.9948955301342011,
"learning_rate": 1.983739837398374e-05,
"loss": 0.8084,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 1.0511960850780724,
"learning_rate": 2.016260162601626e-05,
"loss": 0.8484,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 1.0055140025441067,
"learning_rate": 2.048780487804878e-05,
"loss": 0.8234,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 1.0846607892533107,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.8615,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 1.0602877540281315,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.8445,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 1.2810850023116775,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.8395,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 1.0440591376805315,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.8492,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 1.3476673600206697,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.8531,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 1.0713696734996134,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.8464,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 1.1166859434911962,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.8565,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 1.178244100636038,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.8426,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 1.0461893924553993,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.8117,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 1.0971810544396836,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.8577,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.9663012934030982,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.8229,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.9070333391696412,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.8253,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 0.9747760224443178,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.806,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.9403875047025927,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.8402,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 0.9656477954143978,
"learning_rate": 2.536585365853659e-05,
"loss": 0.8294,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 1.0644898460826773,
"learning_rate": 2.569105691056911e-05,
"loss": 0.8246,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 1.023804927948162,
"learning_rate": 2.601626016260163e-05,
"loss": 0.8427,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 0.951801098334677,
"learning_rate": 2.634146341463415e-05,
"loss": 0.8244,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 1.0922198455099041,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.8326,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 1.057307584255858,
"learning_rate": 2.699186991869919e-05,
"loss": 0.8409,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 1.218772739106358,
"learning_rate": 2.731707317073171e-05,
"loss": 0.8262,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 1.0573764114062014,
"learning_rate": 2.764227642276423e-05,
"loss": 0.8363,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 1.2297765568292516,
"learning_rate": 2.796747967479675e-05,
"loss": 0.8249,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 1.0169476841738558,
"learning_rate": 2.829268292682927e-05,
"loss": 0.8207,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 1.228946130265844,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.8513,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 1.0406426909364621,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.8546,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 1.0858628978176028,
"learning_rate": 2.926829268292683e-05,
"loss": 0.8696,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 0.9758170314323642,
"learning_rate": 2.959349593495935e-05,
"loss": 0.8286,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 1.2840059972631144,
"learning_rate": 2.991869918699187e-05,
"loss": 0.8202,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 1.0928866445494192,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.8202,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 1.0273319093519444,
"learning_rate": 3.056910569105692e-05,
"loss": 0.808,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 1.0320371933351296,
"learning_rate": 3.089430894308943e-05,
"loss": 0.8339,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 0.9462942084839352,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.8179,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 0.8605038683040473,
"learning_rate": 3.154471544715447e-05,
"loss": 0.8295,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 0.9202050989317262,
"learning_rate": 3.186991869918699e-05,
"loss": 0.839,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 0.8998595212539648,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.843,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 1.1445000353271642,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.8287,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 1.008172874275372,
"learning_rate": 3.284552845528456e-05,
"loss": 0.8153,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 1.0846423483007224,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.8136,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 1.0121895910099132,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.8127,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 1.0246418216873459,
"learning_rate": 3.382113821138212e-05,
"loss": 0.8246,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 0.9236919899810215,
"learning_rate": 3.414634146341463e-05,
"loss": 0.8226,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 0.9918508480962676,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.8312,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 1.1400152295400638,
"learning_rate": 3.479674796747968e-05,
"loss": 0.8488,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 1.1722118300851043,
"learning_rate": 3.51219512195122e-05,
"loss": 0.8168,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 0.8985154259591382,
"learning_rate": 3.544715447154472e-05,
"loss": 0.8201,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 1.0664196912961734,
"learning_rate": 3.577235772357724e-05,
"loss": 0.8757,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 1.0381379296341693,
"learning_rate": 3.609756097560976e-05,
"loss": 0.8437,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 0.9345875318249449,
"learning_rate": 3.642276422764228e-05,
"loss": 0.8228,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 1.0213376354003298,
"learning_rate": 3.67479674796748e-05,
"loss": 0.8147,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 1.1087168384350647,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.8179,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 0.9496561751873671,
"learning_rate": 3.739837398373984e-05,
"loss": 0.8351,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 1.0024799123513168,
"learning_rate": 3.772357723577236e-05,
"loss": 0.8396,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 1.0411434216465936,
"learning_rate": 3.804878048780488e-05,
"loss": 0.8461,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 0.878039419260935,
"learning_rate": 3.83739837398374e-05,
"loss": 0.8376,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 0.9198389909999117,
"learning_rate": 3.869918699186992e-05,
"loss": 0.8391,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 1.0477927314692563,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.8283,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 1.0913528792063034,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.8117,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 1.073592359174104,
"learning_rate": 3.967479674796748e-05,
"loss": 0.8045,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 1.0633782413208017,
"learning_rate": 4e-05,
"loss": 0.862,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 0.9525636240515236,
"learning_rate": 3.999991946137476e-05,
"loss": 0.8362,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 1.0000906688424975,
"learning_rate": 3.999967784614766e-05,
"loss": 0.8043,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 0.8812636827998283,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.8362,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 1.0336427379880995,
"learning_rate": 3.999871139496895e-05,
"loss": 0.8588,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 0.9507777279038656,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.8442,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 1.0580742375877734,
"learning_rate": 3.999710067759846e-05,
"loss": 0.8439,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 0.9309408119942485,
"learning_rate": 3.999605373449617e-05,
"loss": 0.8312,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.8695028915649585,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.8059,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 1.0454591169497167,
"learning_rate": 3.999347672161713e-05,
"loss": 0.817,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 0.8738752289390473,
"learning_rate": 3.999194667259528e-05,
"loss": 0.8233,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 1.3774634456048735,
"learning_rate": 3.999025561118334e-05,
"loss": 0.8533,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 0.9367594178384808,
"learning_rate": 3.998840355100086e-05,
"loss": 0.8518,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 1.2754268619523887,
"learning_rate": 3.998639050696409e-05,
"loss": 0.8474,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 0.9820371016660903,
"learning_rate": 3.998421649528582e-05,
"loss": 0.8383,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 0.9980286917964087,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.8376,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.9727881669912443,
"learning_rate": 3.997938564033779e-05,
"loss": 0.8484,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 0.9049208730417142,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.8471,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.9801805816519528,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.8581,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 0.7542688105415302,
"learning_rate": 3.997093258045973e-05,
"loss": 0.8465,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.929388952341545,
"learning_rate": 3.996779317598936e-05,
"loss": 0.8174,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 0.8753700050327151,
"learning_rate": 3.996449295365782e-05,
"loss": 0.7997,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.9525816481070986,
"learning_rate": 3.996103194004467e-05,
"loss": 0.8277,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.9500597814123136,
"learning_rate": 3.995741016302441e-05,
"loss": 0.8183,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 0.9562500255336047,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.844,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 0.9307653963935292,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.8414,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 0.9367235507428224,
"learning_rate": 3.994558054968643e-05,
"loss": 0.8141,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.7814277921171803,
"learning_rate": 3.994131602367481e-05,
"loss": 0.8185,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 0.7747222301879422,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.8252,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.8863825523478212,
"learning_rate": 3.993230519343752e-05,
"loss": 0.8423,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 0.8633778897983807,
"learning_rate": 3.992755896178383e-05,
"loss": 0.8542,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.8835435823966828,
"learning_rate": 3.992265223630981e-05,
"loss": 0.812,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 0.9164184152530932,
"learning_rate": 3.991758505653355e-05,
"loss": 0.8383,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 0.7373395253613425,
"learning_rate": 3.991235746326543e-05,
"loss": 0.8042,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.8562345432897505,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.8491,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 0.7748015706836923,
"learning_rate": 3.990142120595444e-05,
"loss": 0.8425,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 0.7655369612530196,
"learning_rate": 3.98957126299907e-05,
"loss": 0.8278,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 0.7376327341793916,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.8364,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 0.7387778855436938,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.8086,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 0.7343401349359739,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.829,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 0.8111346763986166,
"learning_rate": 3.98712764319087e-05,
"loss": 0.8194,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.7923891191126099,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.8286,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 0.7923606130169327,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.8522,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.789116638153799,
"learning_rate": 3.985126869010129e-05,
"loss": 0.8405,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 0.7381158651357591,
"learning_rate": 3.984427961114833e-05,
"loss": 0.8604,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 0.7938302290702126,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.8145,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 0.7217186618693233,
"learning_rate": 3.982982204151901e-05,
"loss": 0.8206,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.7304648664602218,
"learning_rate": 3.982235366728193e-05,
"loss": 0.8357,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.8029957995833307,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.8471,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 0.6932127068462877,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.8017,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.7820119242139294,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.8096,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.7630496731026998,
"learning_rate": 3.979088432599627e-05,
"loss": 0.8433,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.6907671083094005,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.8181,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 0.7758556449074843,
"learning_rate": 3.977419304032111e-05,
"loss": 0.7923,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 0.7438790767793124,
"learning_rate": 3.976560847560697e-05,
"loss": 0.8281,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 0.7163640436412572,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.8468,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 0.7261546844970004,
"learning_rate": 3.974796184811956e-05,
"loss": 0.8151,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.7296005605648727,
"learning_rate": 3.973889992746979e-05,
"loss": 0.828,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.7567910518483075,
"learning_rate": 3.972967903243361e-05,
"loss": 0.8317,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.9072664457792458,
"learning_rate": 3.972029923727486e-05,
"loss": 0.8559,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.856729945749569,
"learning_rate": 3.971076061753709e-05,
"loss": 0.8299,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.7956859943269419,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.8056,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.7603365538848972,
"learning_rate": 3.969120721289402e-05,
"loss": 0.8051,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.8436200949320591,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.834,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.7535966759091176,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.8336,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.7803514001213459,
"learning_rate": 3.966068788369422e-05,
"loss": 0.8277,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.7713614912992535,
"learning_rate": 3.965019797448622e-05,
"loss": 0.8254,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.7240454979582007,
"learning_rate": 3.963954980528515e-05,
"loss": 0.8001,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.8310022620684312,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.8091,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.6865472592642535,
"learning_rate": 3.961777903121329e-05,
"loss": 0.8162,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.7943980739425294,
"learning_rate": 3.960665660168131e-05,
"loss": 0.8449,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.7385877733316012,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.8287,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.8301116839657243,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.8102,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 0.781816521151654,
"learning_rate": 3.957234222185657e-05,
"loss": 0.838,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.799113304596952,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.824,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 0.8358054059384119,
"learning_rate": 3.954867765033605e-05,
"loss": 0.8057,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.758600041731108,
"learning_rate": 3.953660915306728e-05,
"loss": 0.8439,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.8185222395427251,
"learning_rate": 3.952438331063419e-05,
"loss": 0.822,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.7259443736723704,
"learning_rate": 3.951200022150205e-05,
"loss": 0.852,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.7212625186184672,
"learning_rate": 3.949945998540253e-05,
"loss": 0.8159,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.7382098872116828,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.8267,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.6947159456099462,
"learning_rate": 3.947390847755559e-05,
"loss": 0.8047,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.7516092385006192,
"learning_rate": 3.946089741159648e-05,
"loss": 0.8522,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.7118728484797945,
"learning_rate": 3.944772961024501e-05,
"loss": 0.843,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.7138148744308567,
"learning_rate": 3.943440517955285e-05,
"loss": 0.8248,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.9174557152305183,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.8086,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.7679059017923763,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.837,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.9022670794280417,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.8135,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.8622370199714897,
"learning_rate": 3.937954332854371e-05,
"loss": 0.8171,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.8412918561321807,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.8013,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.8347572745510101,
"learning_rate": 3.935117547697387e-05,
"loss": 0.8177,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.7637293237114983,
"learning_rate": 3.933675771619675e-05,
"loss": 0.8258,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.8748314115895292,
"learning_rate": 3.932218421983131e-05,
"loss": 0.8084,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.7954292551947988,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.8243,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.823361213099621,
"learning_rate": 3.929257049108054e-05,
"loss": 0.8291,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.919840466866298,
"learning_rate": 3.927753049720011e-05,
"loss": 0.8238,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.7642537741803815,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.84,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.9157584533225639,
"learning_rate": 3.92469848560784e-05,
"loss": 0.8324,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.7278191860680464,
"learning_rate": 3.923147945484751e-05,
"loss": 0.8121,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.7809631912232056,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.8082,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.7280735124406051,
"learning_rate": 3.920000411543654e-05,
"loss": 0.7903,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.8155604658128605,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.8068,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.7368387215999596,
"learning_rate": 3.916791024049648e-05,
"loss": 0.8033,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.7518453314420176,
"learning_rate": 3.91516316745245e-05,
"loss": 0.8389,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.9255638194742798,
"learning_rate": 3.913519886394389e-05,
"loss": 0.8354,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.7032444460284674,
"learning_rate": 3.911861194110225e-05,
"loss": 0.818,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.7659805431913463,
"learning_rate": 3.910187103958837e-05,
"loss": 0.8239,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.7490572369409432,
"learning_rate": 3.908497629423117e-05,
"loss": 0.8244,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.7531690554236026,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.8327,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.7726496171701119,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.789,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.7597583686045692,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.8384,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.7042760913343034,
"learning_rate": 3.901586161429081e-05,
"loss": 0.8196,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.7551882798116235,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.8138,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.8186473423406924,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.8243,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.7935442205274802,
"learning_rate": 3.896241703466667e-05,
"loss": 0.8209,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.6928721820867808,
"learning_rate": 3.894429654085585e-05,
"loss": 0.8272,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.6949363260862331,
"learning_rate": 3.892602347228505e-05,
"loss": 0.8146,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.739292356054765,
"learning_rate": 3.890759797612307e-05,
"loss": 0.8056,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.661523091765273,
"learning_rate": 3.888902020076632e-05,
"loss": 0.8069,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.7562427817024111,
"learning_rate": 3.887029029583764e-05,
"loss": 0.8139,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.6813767219461144,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.792,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.763578901243012,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.8304,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.7759489436157322,
"learning_rate": 3.881318931821972e-05,
"loss": 0.8095,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.7839373925599333,
"learning_rate": 3.879385241571817e-05,
"loss": 0.8117,
"step": 246
},
{
"epoch": 1.0020253164556963,
"grad_norm": 0.8176234043899916,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.7216,
"step": 247
},
{
"epoch": 1.0060759493670886,
"grad_norm": 0.9044981905471841,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.5816,
"step": 248
},
{
"epoch": 1.010126582278481,
"grad_norm": 0.7570121259152276,
"learning_rate": 3.873493415863256e-05,
"loss": 0.6113,
"step": 249
},
{
"epoch": 1.0141772151898734,
"grad_norm": 0.9599669728794842,
"learning_rate": 3.871499275032111e-05,
"loss": 0.5945,
"step": 250
},
{
"epoch": 1.0182278481012659,
"grad_norm": 0.8412544562171668,
"learning_rate": 3.869490061403091e-05,
"loss": 0.5813,
"step": 251
},
{
"epoch": 1.0222784810126582,
"grad_norm": 0.8275565160425303,
"learning_rate": 3.867465791158124e-05,
"loss": 0.5723,
"step": 252
},
{
"epoch": 1.0263291139240507,
"grad_norm": 0.8797412659340897,
"learning_rate": 3.865426480600407e-05,
"loss": 0.5519,
"step": 253
},
{
"epoch": 1.030379746835443,
"grad_norm": 0.8526600302759778,
"learning_rate": 3.863372146154264e-05,
"loss": 0.6165,
"step": 254
},
{
"epoch": 1.0344303797468355,
"grad_norm": 0.8436302496525984,
"learning_rate": 3.861302804365024e-05,
"loss": 0.5862,
"step": 255
},
{
"epoch": 1.0384810126582278,
"grad_norm": 0.8001437845596726,
"learning_rate": 3.85921847189888e-05,
"loss": 0.5768,
"step": 256
},
{
"epoch": 1.0425316455696203,
"grad_norm": 0.8980399155466119,
"learning_rate": 3.85711916554276e-05,
"loss": 0.5865,
"step": 257
},
{
"epoch": 1.0465822784810126,
"grad_norm": 0.7128717563071048,
"learning_rate": 3.85500490220419e-05,
"loss": 0.5988,
"step": 258
},
{
"epoch": 1.0506329113924051,
"grad_norm": 0.787288678885512,
"learning_rate": 3.852875698911154e-05,
"loss": 0.5888,
"step": 259
},
{
"epoch": 1.0546835443037974,
"grad_norm": 0.7963701157066301,
"learning_rate": 3.850731572811963e-05,
"loss": 0.6168,
"step": 260
},
{
"epoch": 1.05873417721519,
"grad_norm": 0.7843384890666287,
"learning_rate": 3.848572541175116e-05,
"loss": 0.5885,
"step": 261
},
{
"epoch": 1.0627848101265822,
"grad_norm": 0.7822980661088962,
"learning_rate": 3.846398621389154e-05,
"loss": 0.5882,
"step": 262
},
{
"epoch": 1.0668354430379747,
"grad_norm": 0.7135571083340596,
"learning_rate": 3.84420983096253e-05,
"loss": 0.5967,
"step": 263
},
{
"epoch": 1.070886075949367,
"grad_norm": 0.7347893004619421,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.5553,
"step": 264
},
{
"epoch": 1.0749367088607595,
"grad_norm": 0.7459042343416314,
"learning_rate": 3.839787708819787e-05,
"loss": 0.5843,
"step": 265
},
{
"epoch": 1.0789873417721518,
"grad_norm": 0.756782468207476,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.5929,
"step": 266
},
{
"epoch": 1.0830379746835443,
"grad_norm": 0.6737767050029987,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.5864,
"step": 267
},
{
"epoch": 1.0870886075949366,
"grad_norm": 0.7626733024848522,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.5556,
"step": 268
},
{
"epoch": 1.0911392405063292,
"grad_norm": 0.8641104497912968,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.598,
"step": 269
},
{
"epoch": 1.0951898734177214,
"grad_norm": 0.7372518647288876,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.5775,
"step": 270
},
{
"epoch": 1.099240506329114,
"grad_norm": 0.7017359962190998,
"learning_rate": 3.826166304956594e-05,
"loss": 0.589,
"step": 271
},
{
"epoch": 1.1032911392405063,
"grad_norm": 0.6667557863327281,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.5843,
"step": 272
},
{
"epoch": 1.1073417721518988,
"grad_norm": 0.6830249557713303,
"learning_rate": 3.821507978767479e-05,
"loss": 0.5676,
"step": 273
},
{
"epoch": 1.111392405063291,
"grad_norm": 0.7412677565333684,
"learning_rate": 3.819156801001693e-05,
"loss": 0.5833,
"step": 274
},
{
"epoch": 1.1154430379746836,
"grad_norm": 0.7355135976097368,
"learning_rate": 3.816790971997121e-05,
"loss": 0.5827,
"step": 275
},
{
"epoch": 1.1194936708860759,
"grad_norm": 0.7207184833198015,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.5904,
"step": 276
},
{
"epoch": 1.1235443037974684,
"grad_norm": 0.7666674659573788,
"learning_rate": 3.81201543660571e-05,
"loss": 0.5755,
"step": 277
},
{
"epoch": 1.1275949367088607,
"grad_norm": 0.7717713370462717,
"learning_rate": 3.809605768680377e-05,
"loss": 0.5925,
"step": 278
},
{
"epoch": 1.1316455696202532,
"grad_norm": 0.6877617844768481,
"learning_rate": 3.807181526438958e-05,
"loss": 0.566,
"step": 279
},
{
"epoch": 1.1356962025316455,
"grad_norm": 0.7177417110728144,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.5941,
"step": 280
},
{
"epoch": 1.139746835443038,
"grad_norm": 0.7575426946756193,
"learning_rate": 3.802289397223145e-05,
"loss": 0.5864,
"step": 281
},
{
"epoch": 1.1437974683544303,
"grad_norm": 0.7012237173313887,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.5811,
"step": 282
},
{
"epoch": 1.1478481012658228,
"grad_norm": 0.754080082405236,
"learning_rate": 3.797339206560096e-05,
"loss": 0.6029,
"step": 283
},
{
"epoch": 1.1518987341772151,
"grad_norm": 0.6619301168799914,
"learning_rate": 3.794842387948027e-05,
"loss": 0.5639,
"step": 284
},
{
"epoch": 1.1559493670886076,
"grad_norm": 0.7389862543017989,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.5897,
"step": 285
},
{
"epoch": 1.16,
"grad_norm": 0.7123572562576377,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.5773,
"step": 286
},
{
"epoch": 1.1640506329113924,
"grad_norm": 0.7648099862525573,
"learning_rate": 3.787265280646825e-05,
"loss": 0.5936,
"step": 287
},
{
"epoch": 1.1681012658227847,
"grad_norm": 0.7158418868213963,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.5858,
"step": 288
},
{
"epoch": 1.1721518987341772,
"grad_norm": 0.7034443381522495,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.5685,
"step": 289
},
{
"epoch": 1.1762025316455695,
"grad_norm": 0.7275395107904723,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.5877,
"step": 290
},
{
"epoch": 1.180253164556962,
"grad_norm": 0.7235620207454525,
"learning_rate": 3.776961046830306e-05,
"loss": 0.5999,
"step": 291
},
{
"epoch": 1.1843037974683543,
"grad_norm": 0.7325277633619642,
"learning_rate": 3.774349157719215e-05,
"loss": 0.598,
"step": 292
},
{
"epoch": 1.1883544303797469,
"grad_norm": 0.7344871584641564,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.5993,
"step": 293
},
{
"epoch": 1.1924050632911392,
"grad_norm": 0.8225574621359677,
"learning_rate": 3.769082529555359e-05,
"loss": 0.5686,
"step": 294
},
{
"epoch": 1.1964556962025317,
"grad_norm": 0.7262416686397084,
"learning_rate": 3.766427832919294e-05,
"loss": 0.5722,
"step": 295
},
{
"epoch": 1.200506329113924,
"grad_norm": 0.7207747137991763,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.5922,
"step": 296
},
{
"epoch": 1.2045569620253165,
"grad_norm": 0.7531807232710078,
"learning_rate": 3.761075781441526e-05,
"loss": 0.5956,
"step": 297
},
{
"epoch": 1.2086075949367088,
"grad_norm": 0.8629687131615185,
"learning_rate": 3.75837846970451e-05,
"loss": 0.5654,
"step": 298
},
{
"epoch": 1.2126582278481013,
"grad_norm": 0.66657501663102,
"learning_rate": 3.755666996229032e-05,
"loss": 0.5896,
"step": 299
},
{
"epoch": 1.2167088607594936,
"grad_norm": 0.7914396275375452,
"learning_rate": 3.752941382852927e-05,
"loss": 0.6006,
"step": 300
},
{
"epoch": 1.220759493670886,
"grad_norm": 0.7719901077843927,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.5863,
"step": 301
},
{
"epoch": 1.2248101265822784,
"grad_norm": 0.7617878217211308,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.6008,
"step": 302
},
{
"epoch": 1.228860759493671,
"grad_norm": 0.7706350865901714,
"learning_rate": 3.744679923406351e-05,
"loss": 0.5755,
"step": 303
},
{
"epoch": 1.2329113924050632,
"grad_norm": 0.7826858699649715,
"learning_rate": 3.741897971081043e-05,
"loss": 0.598,
"step": 304
},
{
"epoch": 1.2369620253164557,
"grad_norm": 0.75575647575103,
"learning_rate": 3.739101989748946e-05,
"loss": 0.5948,
"step": 305
},
{
"epoch": 1.241012658227848,
"grad_norm": 0.7259998159033455,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.5816,
"step": 306
},
{
"epoch": 1.2450632911392405,
"grad_norm": 0.7516031614406877,
"learning_rate": 3.73346803025098e-05,
"loss": 0.5757,
"step": 307
},
{
"epoch": 1.2491139240506328,
"grad_norm": 0.6690044075848046,
"learning_rate": 3.730630097460247e-05,
"loss": 0.5864,
"step": 308
},
{
"epoch": 1.2531645569620253,
"grad_norm": 0.7054221126448951,
"learning_rate": 3.727778226412628e-05,
"loss": 0.5786,
"step": 309
},
{
"epoch": 1.2572151898734178,
"grad_norm": 0.7567823034855882,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.6132,
"step": 310
},
{
"epoch": 1.2612658227848101,
"grad_norm": 0.7093579180683525,
"learning_rate": 3.722032761533114e-05,
"loss": 0.5746,
"step": 311
},
{
"epoch": 1.2653164556962024,
"grad_norm": 0.7379485836741586,
"learning_rate": 3.719139213974403e-05,
"loss": 0.5849,
"step": 312
},
{
"epoch": 1.269367088607595,
"grad_norm": 0.7259764626312343,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.5762,
"step": 313
},
{
"epoch": 1.2734177215189875,
"grad_norm": 0.8456406145060527,
"learning_rate": 3.713310605140055e-05,
"loss": 0.5784,
"step": 314
},
{
"epoch": 1.2774683544303798,
"grad_norm": 0.7651297586846054,
"learning_rate": 3.710375590807233e-05,
"loss": 0.5749,
"step": 315
},
{
"epoch": 1.281518987341772,
"grad_norm": 0.7455107895582016,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.6179,
"step": 316
},
{
"epoch": 1.2855696202531646,
"grad_norm": 0.7434519079686347,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.6084,
"step": 317
},
{
"epoch": 1.289620253164557,
"grad_norm": 0.720162962354125,
"learning_rate": 3.701487992136854e-05,
"loss": 0.5977,
"step": 318
},
{
"epoch": 1.2936708860759494,
"grad_norm": 0.7562498540808521,
"learning_rate": 3.69849802022222e-05,
"loss": 0.5682,
"step": 319
},
{
"epoch": 1.2977215189873417,
"grad_norm": 0.7196103210091814,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.5936,
"step": 320
},
{
"epoch": 1.3017721518987342,
"grad_norm": 0.7434759260420499,
"learning_rate": 3.692477062175289e-05,
"loss": 0.5981,
"step": 321
},
{
"epoch": 1.3058227848101267,
"grad_norm": 0.7359614832568304,
"learning_rate": 3.689446124534958e-05,
"loss": 0.5976,
"step": 322
},
{
"epoch": 1.309873417721519,
"grad_norm": 0.7277692899960494,
"learning_rate": 3.686401580327799e-05,
"loss": 0.5753,
"step": 323
},
{
"epoch": 1.3139240506329113,
"grad_norm": 0.6742749468034184,
"learning_rate": 3.683343454074149e-05,
"loss": 0.5705,
"step": 324
},
{
"epoch": 1.3179746835443038,
"grad_norm": 0.7123588387464325,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.5662,
"step": 325
},
{
"epoch": 1.3220253164556963,
"grad_norm": 0.6895970617295654,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.5886,
"step": 326
},
{
"epoch": 1.3260759493670886,
"grad_norm": 0.6680128230256643,
"learning_rate": 3.674087829877297e-05,
"loss": 0.583,
"step": 327
},
{
"epoch": 1.330126582278481,
"grad_norm": 0.6913960228119983,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.5867,
"step": 328
},
{
"epoch": 1.3341772151898734,
"grad_norm": 0.6694118176815136,
"learning_rate": 3.667849957966501e-05,
"loss": 0.5715,
"step": 329
},
{
"epoch": 1.338227848101266,
"grad_norm": 0.7256325941382522,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.5745,
"step": 330
},
{
"epoch": 1.3422784810126582,
"grad_norm": 0.767092259583827,
"learning_rate": 3.661558355626795e-05,
"loss": 0.6062,
"step": 331
},
{
"epoch": 1.3463291139240505,
"grad_norm": 0.7679222005280212,
"learning_rate": 3.658392468818163e-05,
"loss": 0.5893,
"step": 332
},
{
"epoch": 1.350379746835443,
"grad_norm": 0.7409105613666013,
"learning_rate": 3.655213225544574e-05,
"loss": 0.6136,
"step": 333
},
{
"epoch": 1.3544303797468356,
"grad_norm": 0.7388589941969589,
"learning_rate": 3.652020651411218e-05,
"loss": 0.5943,
"step": 334
},
{
"epoch": 1.3584810126582278,
"grad_norm": 0.7511791874896535,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.5951,
"step": 335
},
{
"epoch": 1.3625316455696201,
"grad_norm": 0.7763223421172308,
"learning_rate": 3.645595613522574e-05,
"loss": 0.5864,
"step": 336
},
{
"epoch": 1.3665822784810127,
"grad_norm": 0.7037399273765952,
"learning_rate": 3.642363201513657e-05,
"loss": 0.6068,
"step": 337
},
{
"epoch": 1.3706329113924052,
"grad_norm": 0.8592795102688359,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.5872,
"step": 338
},
{
"epoch": 1.3746835443037975,
"grad_norm": 0.7764841434920271,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.5704,
"step": 339
},
{
"epoch": 1.3787341772151898,
"grad_norm": 0.6938143560491017,
"learning_rate": 3.632586705948318e-05,
"loss": 0.5905,
"step": 340
},
{
"epoch": 1.3827848101265823,
"grad_norm": 0.79212969444576,
"learning_rate": 3.629301541734311e-05,
"loss": 0.5768,
"step": 341
},
{
"epoch": 1.3868354430379748,
"grad_norm": 0.7350349035762532,
"learning_rate": 3.626003255349676e-05,
"loss": 0.6187,
"step": 342
},
{
"epoch": 1.390886075949367,
"grad_norm": 0.8232129418967443,
"learning_rate": 3.622691873358357e-05,
"loss": 0.6087,
"step": 343
},
{
"epoch": 1.3949367088607594,
"grad_norm": 0.7247331834594049,
"learning_rate": 3.61936742242977e-05,
"loss": 0.6114,
"step": 344
},
{
"epoch": 1.398987341772152,
"grad_norm": 0.6819023081950144,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.5869,
"step": 345
},
{
"epoch": 1.4030379746835444,
"grad_norm": 0.7544240666893031,
"learning_rate": 3.612679420964516e-05,
"loss": 0.605,
"step": 346
},
{
"epoch": 1.4070886075949367,
"grad_norm": 0.7698735240467657,
"learning_rate": 3.609315924292092e-05,
"loss": 0.6008,
"step": 347
},
{
"epoch": 1.411139240506329,
"grad_norm": 0.6526392017759234,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.593,
"step": 348
},
{
"epoch": 1.4151898734177215,
"grad_norm": 0.7547532963938162,
"learning_rate": 3.602550074513133e-05,
"loss": 0.6028,
"step": 349
},
{
"epoch": 1.419240506329114,
"grad_norm": 0.738540193371399,
"learning_rate": 3.599147775897822e-05,
"loss": 0.5997,
"step": 350
},
{
"epoch": 1.4232911392405063,
"grad_norm": 0.6676467158423623,
"learning_rate": 3.595732597966167e-05,
"loss": 0.5891,
"step": 351
},
{
"epoch": 1.4273417721518986,
"grad_norm": 0.7037988668997797,
"learning_rate": 3.592304568223542e-05,
"loss": 0.5834,
"step": 352
},
{
"epoch": 1.4313924050632911,
"grad_norm": 0.7763465768603007,
"learning_rate": 3.588863714278826e-05,
"loss": 0.5993,
"step": 353
},
{
"epoch": 1.4354430379746836,
"grad_norm": 0.6300320805647937,
"learning_rate": 3.585410063844186e-05,
"loss": 0.5758,
"step": 354
},
{
"epoch": 1.439493670886076,
"grad_norm": 0.7488178716628641,
"learning_rate": 3.581943644734846e-05,
"loss": 0.5969,
"step": 355
},
{
"epoch": 1.4435443037974682,
"grad_norm": 0.7291168988015532,
"learning_rate": 3.578464484868869e-05,
"loss": 0.5875,
"step": 356
},
{
"epoch": 1.4475949367088607,
"grad_norm": 0.7098163924267723,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.6099,
"step": 357
},
{
"epoch": 1.4516455696202533,
"grad_norm": 0.7653424783324582,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.584,
"step": 358
},
{
"epoch": 1.4556962025316456,
"grad_norm": 0.6873743812966822,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.5919,
"step": 359
},
{
"epoch": 1.4597468354430378,
"grad_norm": 0.659899312434721,
"learning_rate": 3.564420999786543e-05,
"loss": 0.5786,
"step": 360
},
{
"epoch": 1.4637974683544304,
"grad_norm": 0.7335940423254019,
"learning_rate": 3.560878558491842e-05,
"loss": 0.5673,
"step": 361
},
{
"epoch": 1.4678481012658229,
"grad_norm": 0.7179856316530925,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.605,
"step": 362
},
{
"epoch": 1.4718987341772152,
"grad_norm": 0.7141190327009512,
"learning_rate": 3.553755991230039e-05,
"loss": 0.5993,
"step": 363
},
{
"epoch": 1.4759493670886075,
"grad_norm": 0.733041647286087,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.5697,
"step": 364
},
{
"epoch": 1.48,
"grad_norm": 0.7123400233476338,
"learning_rate": 3.546583369120419e-05,
"loss": 0.6056,
"step": 365
},
{
"epoch": 1.4840506329113925,
"grad_norm": 0.7283749870531676,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.5904,
"step": 366
},
{
"epoch": 1.4881012658227848,
"grad_norm": 0.7220250245993246,
"learning_rate": 3.539360923231766e-05,
"loss": 0.5827,
"step": 367
},
{
"epoch": 1.492151898734177,
"grad_norm": 0.6704537456472008,
"learning_rate": 3.535731089018394e-05,
"loss": 0.5753,
"step": 368
},
{
"epoch": 1.4962025316455696,
"grad_norm": 0.7135887518714831,
"learning_rate": 3.532088886237956e-05,
"loss": 0.6047,
"step": 369
},
{
"epoch": 1.5002531645569621,
"grad_norm": 0.7157851825989271,
"learning_rate": 3.528434344224253e-05,
"loss": 0.583,
"step": 370
},
{
"epoch": 1.5043037974683544,
"grad_norm": 0.6983828987412893,
"learning_rate": 3.524767492410464e-05,
"loss": 0.5948,
"step": 371
},
{
"epoch": 1.5083544303797467,
"grad_norm": 0.699331242262123,
"learning_rate": 3.521088360328908e-05,
"loss": 0.583,
"step": 372
},
{
"epoch": 1.5124050632911392,
"grad_norm": 0.7063027539748569,
"learning_rate": 3.517396977610811e-05,
"loss": 0.5638,
"step": 373
},
{
"epoch": 1.5164556962025317,
"grad_norm": 0.6495421502435068,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.5875,
"step": 374
},
{
"epoch": 1.520506329113924,
"grad_norm": 0.7462797585212999,
"learning_rate": 3.509977579282971e-05,
"loss": 0.6194,
"step": 375
},
{
"epoch": 1.5245569620253163,
"grad_norm": 0.7069408248355523,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.5976,
"step": 376
},
{
"epoch": 1.5286075949367088,
"grad_norm": 0.6765056381618781,
"learning_rate": 3.502509536445719e-05,
"loss": 0.5877,
"step": 377
},
{
"epoch": 1.5326582278481014,
"grad_norm": 0.7219109875604066,
"learning_rate": 3.498757348458147e-05,
"loss": 0.5943,
"step": 378
},
{
"epoch": 1.5367088607594936,
"grad_norm": 0.7279361044022289,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.5987,
"step": 379
},
{
"epoch": 1.540759493670886,
"grad_norm": 0.7011001541760211,
"learning_rate": 3.491216790442899e-05,
"loss": 0.5913,
"step": 380
},
{
"epoch": 1.5448101265822785,
"grad_norm": 0.7904673823222896,
"learning_rate": 3.487428481145839e-05,
"loss": 0.5892,
"step": 381
},
{
"epoch": 1.548860759493671,
"grad_norm": 0.737937805772123,
"learning_rate": 3.483628192304278e-05,
"loss": 0.5937,
"step": 382
},
{
"epoch": 1.5529113924050633,
"grad_norm": 0.6589789370641471,
"learning_rate": 3.479815954525219e-05,
"loss": 0.5975,
"step": 383
},
{
"epoch": 1.5569620253164556,
"grad_norm": 0.7415702968102417,
"learning_rate": 3.475991798511899e-05,
"loss": 0.5866,
"step": 384
},
{
"epoch": 1.561012658227848,
"grad_norm": 0.683098226163998,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.6063,
"step": 385
},
{
"epoch": 1.5650632911392406,
"grad_norm": 0.6962332306244734,
"learning_rate": 3.468307855075128e-05,
"loss": 0.5803,
"step": 386
},
{
"epoch": 1.5691139240506329,
"grad_norm": 0.6706244028981881,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.5948,
"step": 387
},
{
"epoch": 1.5731645569620252,
"grad_norm": 0.7758756829539094,
"learning_rate": 3.460576609535163e-05,
"loss": 0.6113,
"step": 388
},
{
"epoch": 1.5772151898734177,
"grad_norm": 0.6424812972492893,
"learning_rate": 3.456693326250006e-05,
"loss": 0.6057,
"step": 389
},
{
"epoch": 1.5812658227848102,
"grad_norm": 0.7549805530627423,
"learning_rate": 3.452798310957058e-05,
"loss": 0.5863,
"step": 390
},
{
"epoch": 1.5853164556962025,
"grad_norm": 0.7252894071521755,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.604,
"step": 391
},
{
"epoch": 1.5893670886075948,
"grad_norm": 0.7174173185331172,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.6124,
"step": 392
},
{
"epoch": 1.5934177215189873,
"grad_norm": 0.6627418292844117,
"learning_rate": 3.441043187201574e-05,
"loss": 0.6122,
"step": 393
},
{
"epoch": 1.5974683544303798,
"grad_norm": 0.6923000055910502,
"learning_rate": 3.437101558517728e-05,
"loss": 0.5901,
"step": 394
},
{
"epoch": 1.6015189873417721,
"grad_norm": 0.6227371663300996,
"learning_rate": 3.433148355615496e-05,
"loss": 0.5993,
"step": 395
},
{
"epoch": 1.6055696202531644,
"grad_norm": 0.7264224781086559,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.6138,
"step": 396
},
{
"epoch": 1.609620253164557,
"grad_norm": 0.7096313664957014,
"learning_rate": 3.425207354603043e-05,
"loss": 0.617,
"step": 397
},
{
"epoch": 1.6136708860759494,
"grad_norm": 0.6874680332339222,
"learning_rate": 3.421219620448553e-05,
"loss": 0.5737,
"step": 398
},
{
"epoch": 1.6177215189873417,
"grad_norm": 0.7333109291304116,
"learning_rate": 3.417220439986623e-05,
"loss": 0.5939,
"step": 399
},
{
"epoch": 1.621772151898734,
"grad_norm": 0.6682301529832292,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.5593,
"step": 400
},
{
"epoch": 1.6258227848101265,
"grad_norm": 0.7474137114575522,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.6167,
"step": 401
},
{
"epoch": 1.629873417721519,
"grad_norm": 0.6706180262040364,
"learning_rate": 3.405154543304065e-05,
"loss": 0.5774,
"step": 402
},
{
"epoch": 1.6339240506329114,
"grad_norm": 0.641989379868484,
"learning_rate": 3.401109900618843e-05,
"loss": 0.5845,
"step": 403
},
{
"epoch": 1.6379746835443036,
"grad_norm": 0.6509215567952106,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.5883,
"step": 404
},
{
"epoch": 1.6420253164556962,
"grad_norm": 0.6753289661334198,
"learning_rate": 3.392986794874714e-05,
"loss": 0.5726,
"step": 405
},
{
"epoch": 1.6460759493670887,
"grad_norm": 0.6422151550148597,
"learning_rate": 3.388908397238184e-05,
"loss": 0.5959,
"step": 406
},
{
"epoch": 1.650126582278481,
"grad_norm": 0.7123735082358158,
"learning_rate": 3.384818813524362e-05,
"loss": 0.5801,
"step": 407
},
{
"epoch": 1.6541772151898733,
"grad_norm": 0.6544758913383721,
"learning_rate": 3.380718076670195e-05,
"loss": 0.6024,
"step": 408
},
{
"epoch": 1.6582278481012658,
"grad_norm": 0.6062604262962585,
"learning_rate": 3.376606219702454e-05,
"loss": 0.5836,
"step": 409
},
{
"epoch": 1.6622784810126583,
"grad_norm": 0.6397167913638488,
"learning_rate": 3.372483275737468e-05,
"loss": 0.6064,
"step": 410
},
{
"epoch": 1.6663291139240506,
"grad_norm": 0.6119240891155664,
"learning_rate": 3.368349277980861e-05,
"loss": 0.6097,
"step": 411
},
{
"epoch": 1.6703797468354429,
"grad_norm": 0.6664532778136305,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.5869,
"step": 412
},
{
"epoch": 1.6744303797468354,
"grad_norm": 0.6882413213864378,
"learning_rate": 3.360048254360144e-05,
"loss": 0.5949,
"step": 413
},
{
"epoch": 1.678481012658228,
"grad_norm": 0.6323612058317645,
"learning_rate": 3.355881295351336e-05,
"loss": 0.6124,
"step": 414
},
{
"epoch": 1.6825316455696202,
"grad_norm": 0.6313855659091396,
"learning_rate": 3.351703416260975e-05,
"loss": 0.602,
"step": 415
},
{
"epoch": 1.6865822784810125,
"grad_norm": 0.5981268565985506,
"learning_rate": 3.347514650737126e-05,
"loss": 0.5933,
"step": 416
},
{
"epoch": 1.690632911392405,
"grad_norm": 0.6850152012221604,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.6086,
"step": 417
},
{
"epoch": 1.6946835443037975,
"grad_norm": 0.6515506735705214,
"learning_rate": 3.339104595419334e-05,
"loss": 0.5938,
"step": 418
},
{
"epoch": 1.6987341772151898,
"grad_norm": 0.5975749021976261,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.6156,
"step": 419
},
{
"epoch": 1.7027848101265821,
"grad_norm": 0.6198974403734151,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.5692,
"step": 420
},
{
"epoch": 1.7068354430379746,
"grad_norm": 0.6368966815160166,
"learning_rate": 3.326408710419996e-05,
"loss": 0.5838,
"step": 421
},
{
"epoch": 1.7108860759493671,
"grad_norm": 0.635421575272419,
"learning_rate": 3.322155337795454e-05,
"loss": 0.5835,
"step": 422
},
{
"epoch": 1.7149367088607594,
"grad_norm": 0.6603850859826649,
"learning_rate": 3.317891316713587e-05,
"loss": 0.5886,
"step": 423
},
{
"epoch": 1.7189873417721517,
"grad_norm": 0.6594058380451611,
"learning_rate": 3.313616681516231e-05,
"loss": 0.5863,
"step": 424
},
{
"epoch": 1.7230379746835442,
"grad_norm": 0.6573085166838574,
"learning_rate": 3.309331466630713e-05,
"loss": 0.5976,
"step": 425
},
{
"epoch": 1.7270886075949368,
"grad_norm": 0.6453905226753844,
"learning_rate": 3.305035706569563e-05,
"loss": 0.6018,
"step": 426
},
{
"epoch": 1.7311392405063293,
"grad_norm": 0.6512973518541296,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.5958,
"step": 427
},
{
"epoch": 1.7351898734177216,
"grad_norm": 0.6933963059026946,
"learning_rate": 3.296412689394864e-05,
"loss": 0.5924,
"step": 428
},
{
"epoch": 1.7392405063291139,
"grad_norm": 0.6742479400812429,
"learning_rate": 3.292085501729909e-05,
"loss": 0.5919,
"step": 429
},
{
"epoch": 1.7432911392405064,
"grad_norm": 0.6277231719043137,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.5884,
"step": 430
},
{
"epoch": 1.747341772151899,
"grad_norm": 0.6707390510993932,
"learning_rate": 3.283399942497381e-05,
"loss": 0.5792,
"step": 431
},
{
"epoch": 1.7513924050632912,
"grad_norm": 0.6546101077387817,
"learning_rate": 3.279041640882108e-05,
"loss": 0.5948,
"step": 432
},
{
"epoch": 1.7554430379746835,
"grad_norm": 0.7223612995798812,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.5817,
"step": 433
},
{
"epoch": 1.759493670886076,
"grad_norm": 0.6815936882652763,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.5806,
"step": 434
},
{
"epoch": 1.7635443037974685,
"grad_norm": 0.6257131728010921,
"learning_rate": 3.265905069502244e-05,
"loss": 0.5954,
"step": 435
},
{
"epoch": 1.7675949367088608,
"grad_norm": 0.6688911125040209,
"learning_rate": 3.261505774420016e-05,
"loss": 0.59,
"step": 436
},
{
"epoch": 1.771645569620253,
"grad_norm": 0.6186754546829485,
"learning_rate": 3.257096319343707e-05,
"loss": 0.5971,
"step": 437
},
{
"epoch": 1.7756962025316456,
"grad_norm": 0.6600169311349564,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.5776,
"step": 438
},
{
"epoch": 1.7797468354430381,
"grad_norm": 0.6570214318321937,
"learning_rate": 3.248247071342966e-05,
"loss": 0.5995,
"step": 439
},
{
"epoch": 1.7837974683544304,
"grad_norm": 0.6524359076785037,
"learning_rate": 3.243807349689161e-05,
"loss": 0.5934,
"step": 440
},
{
"epoch": 1.7878481012658227,
"grad_norm": 0.6526063371282286,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.592,
"step": 441
},
{
"epoch": 1.7918987341772152,
"grad_norm": 0.6565839751404882,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.6019,
"step": 442
},
{
"epoch": 1.7959493670886078,
"grad_norm": 0.6357007283425619,
"learning_rate": 3.230428223438075e-05,
"loss": 0.5868,
"step": 443
},
{
"epoch": 1.8,
"grad_norm": 0.6627117509279796,
"learning_rate": 3.225948647317459e-05,
"loss": 0.6211,
"step": 444
},
{
"epoch": 1.8040506329113923,
"grad_norm": 0.6291554489202547,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.5926,
"step": 445
},
{
"epoch": 1.8081012658227849,
"grad_norm": 0.5983941525227289,
"learning_rate": 3.216959910368034e-05,
"loss": 0.6081,
"step": 446
},
{
"epoch": 1.8121518987341774,
"grad_norm": 0.6427530310819393,
"learning_rate": 3.212450821933277e-05,
"loss": 0.5825,
"step": 447
},
{
"epoch": 1.8162025316455697,
"grad_norm": 0.6477295233882678,
"learning_rate": 3.207931968586281e-05,
"loss": 0.5789,
"step": 448
},
{
"epoch": 1.820253164556962,
"grad_norm": 0.6487519690862404,
"learning_rate": 3.203403386721272e-05,
"loss": 0.5793,
"step": 449
},
{
"epoch": 1.8243037974683545,
"grad_norm": 0.6685093120449408,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.5718,
"step": 450
},
{
"epoch": 1.828354430379747,
"grad_norm": 0.6432854783051659,
"learning_rate": 3.194317183405573e-05,
"loss": 0.6143,
"step": 451
},
{
"epoch": 1.8324050632911393,
"grad_norm": 0.6038677772644971,
"learning_rate": 3.189759635133914e-05,
"loss": 0.5892,
"step": 452
},
{
"epoch": 1.8364556962025316,
"grad_norm": 0.7059138679734801,
"learning_rate": 3.185192504701718e-05,
"loss": 0.6024,
"step": 453
},
{
"epoch": 1.840506329113924,
"grad_norm": 0.6311685761848742,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.606,
"step": 454
},
{
"epoch": 1.8445569620253166,
"grad_norm": 0.6293762348945536,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.5866,
"step": 455
},
{
"epoch": 1.848607594936709,
"grad_norm": 0.6585820181481696,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.6017,
"step": 456
},
{
"epoch": 1.8526582278481012,
"grad_norm": 0.6180455552629438,
"learning_rate": 3.166828898179731e-05,
"loss": 0.5934,
"step": 457
},
{
"epoch": 1.8567088607594937,
"grad_norm": 0.6659627094817447,
"learning_rate": 3.162214410223536e-05,
"loss": 0.587,
"step": 458
},
{
"epoch": 1.8607594936708862,
"grad_norm": 0.6499824284740865,
"learning_rate": 3.157590561952257e-05,
"loss": 0.5752,
"step": 459
},
{
"epoch": 1.8648101265822785,
"grad_norm": 0.7317773125500089,
"learning_rate": 3.152957390605732e-05,
"loss": 0.5784,
"step": 460
},
{
"epoch": 1.8688607594936708,
"grad_norm": 0.6449844363110917,
"learning_rate": 3.148314933498886e-05,
"loss": 0.5907,
"step": 461
},
{
"epoch": 1.8729113924050633,
"grad_norm": 0.7087496869950337,
"learning_rate": 3.143663228021431e-05,
"loss": 0.6026,
"step": 462
},
{
"epoch": 1.8769620253164558,
"grad_norm": 0.6962095634061196,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.5916,
"step": 463
},
{
"epoch": 1.8810126582278481,
"grad_norm": 0.6090754214869452,
"learning_rate": 3.134332221885661e-05,
"loss": 0.6008,
"step": 464
},
{
"epoch": 1.8850632911392404,
"grad_norm": 0.6837431101261914,
"learning_rate": 3.129652996377987e-05,
"loss": 0.6186,
"step": 465
},
{
"epoch": 1.889113924050633,
"grad_norm": 0.6674864613008621,
"learning_rate": 3.12496467280038e-05,
"loss": 0.5903,
"step": 466
},
{
"epoch": 1.8931645569620255,
"grad_norm": 0.6027495123854183,
"learning_rate": 3.120267288911952e-05,
"loss": 0.6059,
"step": 467
},
{
"epoch": 1.8972151898734178,
"grad_norm": 0.6596200603038042,
"learning_rate": 3.11556088254479e-05,
"loss": 0.6145,
"step": 468
},
{
"epoch": 1.90126582278481,
"grad_norm": 0.6402764327521587,
"learning_rate": 3.11084549160364e-05,
"loss": 0.5918,
"step": 469
},
{
"epoch": 1.9053164556962026,
"grad_norm": 0.6413933438704263,
"learning_rate": 3.106121154065615e-05,
"loss": 0.5702,
"step": 470
},
{
"epoch": 1.909367088607595,
"grad_norm": 0.7190364411499097,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.589,
"step": 471
},
{
"epoch": 1.9134177215189874,
"grad_norm": 0.6143810409645176,
"learning_rate": 3.096645791467348e-05,
"loss": 0.5657,
"step": 472
},
{
"epoch": 1.9174683544303797,
"grad_norm": 0.6739518468545961,
"learning_rate": 3.091894842720373e-05,
"loss": 0.6158,
"step": 473
},
{
"epoch": 1.9215189873417722,
"grad_norm": 0.7110935490919414,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.6002,
"step": 474
},
{
"epoch": 1.9255696202531647,
"grad_norm": 0.5762719243262501,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.5838,
"step": 475
},
{
"epoch": 1.929620253164557,
"grad_norm": 0.6858171575864125,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.613,
"step": 476
},
{
"epoch": 1.9336708860759493,
"grad_norm": 0.6835710236846018,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.6035,
"step": 477
},
{
"epoch": 1.9377215189873418,
"grad_norm": 0.6203843322583785,
"learning_rate": 3.06800895716334e-05,
"loss": 0.5936,
"step": 478
},
{
"epoch": 1.9417721518987343,
"grad_norm": 0.6823832916522384,
"learning_rate": 3.063205821011029e-05,
"loss": 0.6207,
"step": 479
},
{
"epoch": 1.9458227848101266,
"grad_norm": 0.6246657149164356,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.596,
"step": 480
},
{
"epoch": 1.949873417721519,
"grad_norm": 0.6569879109633455,
"learning_rate": 3.053573898718618e-05,
"loss": 0.5959,
"step": 481
},
{
"epoch": 1.9539240506329114,
"grad_norm": 0.690342529343311,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.5928,
"step": 482
},
{
"epoch": 1.957974683544304,
"grad_norm": 0.6596866787085898,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.5933,
"step": 483
},
{
"epoch": 1.9620253164556962,
"grad_norm": 0.6091873817999601,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.5886,
"step": 484
},
{
"epoch": 1.9660759493670885,
"grad_norm": 0.6064179676016525,
"learning_rate": 3.034208541656255e-05,
"loss": 0.6097,
"step": 485
},
{
"epoch": 1.970126582278481,
"grad_norm": 0.6264089517234758,
"learning_rate": 3.029346281309218e-05,
"loss": 0.5861,
"step": 486
},
{
"epoch": 1.9741772151898735,
"grad_norm": 0.6213960662676646,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.5696,
"step": 487
},
{
"epoch": 1.9782278481012658,
"grad_norm": 0.6639089595570169,
"learning_rate": 3.019596929201569e-05,
"loss": 0.5946,
"step": 488
},
{
"epoch": 1.9822784810126581,
"grad_norm": 0.6418579350849266,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.6152,
"step": 489
},
{
"epoch": 1.9863291139240506,
"grad_norm": 0.6577437742306624,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.5827,
"step": 490
},
{
"epoch": 1.9903797468354432,
"grad_norm": 0.6817468876264824,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.5815,
"step": 491
},
{
"epoch": 1.9944303797468355,
"grad_norm": 0.6016072235836991,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.5732,
"step": 492
},
{
"epoch": 1.9984810126582278,
"grad_norm": 0.6217473773093328,
"learning_rate": 2.995080534235264e-05,
"loss": 0.5972,
"step": 493
},
{
"epoch": 2.0040506329113925,
"grad_norm": 0.9417454815520314,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.3667,
"step": 494
},
{
"epoch": 2.008101265822785,
"grad_norm": 0.5902846918628168,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.355,
"step": 495
},
{
"epoch": 2.012151898734177,
"grad_norm": 1.288083394466691,
"learning_rate": 2.980274210295326e-05,
"loss": 0.3553,
"step": 496
},
{
"epoch": 2.0162025316455696,
"grad_norm": 0.8479252451850584,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.3351,
"step": 497
},
{
"epoch": 2.020253164556962,
"grad_norm": 0.6641852263239901,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.3263,
"step": 498
},
{
"epoch": 2.0243037974683546,
"grad_norm": 0.7082572598527493,
"learning_rate": 2.965396831793362e-05,
"loss": 0.3554,
"step": 499
},
{
"epoch": 2.0283544303797467,
"grad_norm": 0.6904055550766527,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.3695,
"step": 500
},
{
"epoch": 2.0324050632911392,
"grad_norm": 0.6294130913688227,
"learning_rate": 2.955439637024526e-05,
"loss": 0.3328,
"step": 501
},
{
"epoch": 2.0364556962025318,
"grad_norm": 0.6530942776882078,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.3424,
"step": 502
},
{
"epoch": 2.0405063291139243,
"grad_norm": 0.6913592556629381,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.367,
"step": 503
},
{
"epoch": 2.0445569620253163,
"grad_norm": 0.6348472424913949,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.3578,
"step": 504
},
{
"epoch": 2.048607594936709,
"grad_norm": 0.6296474176256845,
"learning_rate": 2.93543322968538e-05,
"loss": 0.3278,
"step": 505
},
{
"epoch": 2.0526582278481014,
"grad_norm": 0.6154790948918614,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.3412,
"step": 506
},
{
"epoch": 2.056708860759494,
"grad_norm": 0.6669279049429048,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.3221,
"step": 507
},
{
"epoch": 2.060759493670886,
"grad_norm": 0.5878146244893304,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.3421,
"step": 508
},
{
"epoch": 2.0648101265822785,
"grad_norm": 0.6269164626126018,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.3408,
"step": 509
},
{
"epoch": 2.068860759493671,
"grad_norm": 0.5978857536864725,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.333,
"step": 510
},
{
"epoch": 2.0729113924050635,
"grad_norm": 0.5941210704552513,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.3391,
"step": 511
},
{
"epoch": 2.0769620253164556,
"grad_norm": 0.6254195368320388,
"learning_rate": 2.900133526131588e-05,
"loss": 0.3592,
"step": 512
},
{
"epoch": 2.081012658227848,
"grad_norm": 0.600293137498029,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.3484,
"step": 513
},
{
"epoch": 2.0850632911392406,
"grad_norm": 0.600846629223982,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.3414,
"step": 514
},
{
"epoch": 2.089113924050633,
"grad_norm": 0.6326060860624088,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.3349,
"step": 515
},
{
"epoch": 2.093164556962025,
"grad_norm": 0.6010367635219459,
"learning_rate": 2.879801882631847e-05,
"loss": 0.3537,
"step": 516
},
{
"epoch": 2.0972151898734177,
"grad_norm": 0.6075126148151161,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.3305,
"step": 517
},
{
"epoch": 2.1012658227848102,
"grad_norm": 0.6335981606441471,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.3331,
"step": 518
},
{
"epoch": 2.1053164556962027,
"grad_norm": 0.5923792111053628,
"learning_rate": 2.864478605989494e-05,
"loss": 0.3415,
"step": 519
},
{
"epoch": 2.109367088607595,
"grad_norm": 0.6056633753183017,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.3252,
"step": 520
},
{
"epoch": 2.1134177215189873,
"grad_norm": 0.6057724909385909,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.3504,
"step": 521
},
{
"epoch": 2.11746835443038,
"grad_norm": 0.5715124707002572,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.352,
"step": 522
},
{
"epoch": 2.1215189873417724,
"grad_norm": 0.6387210665710601,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.3321,
"step": 523
},
{
"epoch": 2.1255696202531644,
"grad_norm": 0.6309741320259192,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.3473,
"step": 524
},
{
"epoch": 2.129620253164557,
"grad_norm": 0.5870899060314128,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.3324,
"step": 525
},
{
"epoch": 2.1336708860759495,
"grad_norm": 0.6025008053388078,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.3356,
"step": 526
},
{
"epoch": 2.137721518987342,
"grad_norm": 0.5888085310102238,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.3091,
"step": 527
},
{
"epoch": 2.141772151898734,
"grad_norm": 0.5953485183202074,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.3584,
"step": 528
},
{
"epoch": 2.1458227848101266,
"grad_norm": 0.5626659502712237,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.3396,
"step": 529
},
{
"epoch": 2.149873417721519,
"grad_norm": 0.6103444898307342,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.3423,
"step": 530
},
{
"epoch": 2.1539240506329116,
"grad_norm": 0.5921095440118952,
"learning_rate": 2.802570062143278e-05,
"loss": 0.3585,
"step": 531
},
{
"epoch": 2.1579746835443037,
"grad_norm": 0.5795104571283244,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.3455,
"step": 532
},
{
"epoch": 2.162025316455696,
"grad_norm": 0.6093647135240207,
"learning_rate": 2.792159532078314e-05,
"loss": 0.3421,
"step": 533
},
{
"epoch": 2.1660759493670887,
"grad_norm": 0.5735468458628646,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.3586,
"step": 534
},
{
"epoch": 2.170126582278481,
"grad_norm": 0.5910149061234223,
"learning_rate": 2.781723482288857e-05,
"loss": 0.3269,
"step": 535
},
{
"epoch": 2.1741772151898733,
"grad_norm": 0.6066971999021511,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.3174,
"step": 536
},
{
"epoch": 2.178227848101266,
"grad_norm": 0.5902284852912734,
"learning_rate": 2.771262248976272e-05,
"loss": 0.3596,
"step": 537
},
{
"epoch": 2.1822784810126583,
"grad_norm": 0.6046251878394366,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.3269,
"step": 538
},
{
"epoch": 2.186329113924051,
"grad_norm": 0.5863097902757474,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.3259,
"step": 539
},
{
"epoch": 2.190379746835443,
"grad_norm": 0.6110487672585624,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.3311,
"step": 540
},
{
"epoch": 2.1944303797468354,
"grad_norm": 0.6059994184268671,
"learning_rate": 2.7502655806328e-05,
"loss": 0.3158,
"step": 541
},
{
"epoch": 2.198481012658228,
"grad_norm": 0.6006983548152651,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.3352,
"step": 542
},
{
"epoch": 2.2025316455696204,
"grad_norm": 0.609532448982531,
"learning_rate": 2.739730822017673e-05,
"loss": 0.3479,
"step": 543
},
{
"epoch": 2.2065822784810125,
"grad_norm": 0.6272242186402143,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.3459,
"step": 544
},
{
"epoch": 2.210632911392405,
"grad_norm": 0.6370016178947284,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.3334,
"step": 545
},
{
"epoch": 2.2146835443037975,
"grad_norm": 0.5890025192922338,
"learning_rate": 2.723884107774236e-05,
"loss": 0.3478,
"step": 546
},
{
"epoch": 2.21873417721519,
"grad_norm": 0.6293606296486266,
"learning_rate": 2.718590152796239e-05,
"loss": 0.3463,
"step": 547
},
{
"epoch": 2.222784810126582,
"grad_norm": 0.6103616793227594,
"learning_rate": 2.71329041039194e-05,
"loss": 0.3271,
"step": 548
},
{
"epoch": 2.2268354430379746,
"grad_norm": 0.6028796155298725,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.3299,
"step": 549
},
{
"epoch": 2.230886075949367,
"grad_norm": 0.6517848506947911,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.3404,
"step": 550
},
{
"epoch": 2.2349367088607597,
"grad_norm": 0.6058239372211752,
"learning_rate": 2.697356885686189e-05,
"loss": 0.3466,
"step": 551
},
{
"epoch": 2.2389873417721518,
"grad_norm": 0.6224806763109841,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.3448,
"step": 552
},
{
"epoch": 2.2430379746835443,
"grad_norm": 0.5782413272684929,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.3443,
"step": 553
},
{
"epoch": 2.247088607594937,
"grad_norm": 0.594736600946636,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.3357,
"step": 554
},
{
"epoch": 2.2511392405063293,
"grad_norm": 0.5876176879468104,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.3451,
"step": 555
},
{
"epoch": 2.2551898734177214,
"grad_norm": 0.5938950392895469,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.3423,
"step": 556
},
{
"epoch": 2.259240506329114,
"grad_norm": 0.6307780114221686,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.3187,
"step": 557
},
{
"epoch": 2.2632911392405064,
"grad_norm": 0.6247477383858354,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.3499,
"step": 558
},
{
"epoch": 2.267341772151899,
"grad_norm": 0.5979321786405076,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.3465,
"step": 559
},
{
"epoch": 2.271392405063291,
"grad_norm": 0.5856099206622496,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.3405,
"step": 560
},
{
"epoch": 2.2754430379746835,
"grad_norm": 0.6011578651918682,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.353,
"step": 561
},
{
"epoch": 2.279493670886076,
"grad_norm": 0.6016160823986206,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.3531,
"step": 562
},
{
"epoch": 2.2835443037974685,
"grad_norm": 0.5675829739949425,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.3375,
"step": 563
},
{
"epoch": 2.2875949367088606,
"grad_norm": 0.6009674079990768,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.3471,
"step": 564
},
{
"epoch": 2.291645569620253,
"grad_norm": 0.5957294606774565,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.3352,
"step": 565
},
{
"epoch": 2.2956962025316456,
"grad_norm": 0.5638498216484412,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.3464,
"step": 566
},
{
"epoch": 2.299746835443038,
"grad_norm": 0.5752115915053728,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.3416,
"step": 567
},
{
"epoch": 2.3037974683544302,
"grad_norm": 0.59356852555707,
"learning_rate": 2.606146264569603e-05,
"loss": 0.3587,
"step": 568
},
{
"epoch": 2.3078481012658227,
"grad_norm": 0.578577894256214,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.3382,
"step": 569
},
{
"epoch": 2.3118987341772153,
"grad_norm": 0.5761473709122804,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.3207,
"step": 570
},
{
"epoch": 2.3159493670886078,
"grad_norm": 0.6113076736744251,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.3533,
"step": 571
},
{
"epoch": 2.32,
"grad_norm": 0.5523552366690488,
"learning_rate": 2.584472010606015e-05,
"loss": 0.3249,
"step": 572
},
{
"epoch": 2.3240506329113924,
"grad_norm": 0.6041985223549319,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.3318,
"step": 573
},
{
"epoch": 2.328101265822785,
"grad_norm": 0.5779181325756721,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.3464,
"step": 574
},
{
"epoch": 2.3321518987341774,
"grad_norm": 0.5813882345910673,
"learning_rate": 2.568166741322048e-05,
"loss": 0.3355,
"step": 575
},
{
"epoch": 2.3362025316455695,
"grad_norm": 0.6055198154013148,
"learning_rate": 2.56272244128509e-05,
"loss": 0.355,
"step": 576
},
{
"epoch": 2.340253164556962,
"grad_norm": 0.596244546390521,
"learning_rate": 2.55727360915895e-05,
"loss": 0.3608,
"step": 577
},
{
"epoch": 2.3443037974683545,
"grad_norm": 0.5717699270566489,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.3255,
"step": 578
},
{
"epoch": 2.348354430379747,
"grad_norm": 0.6065369974469202,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.3564,
"step": 579
},
{
"epoch": 2.352405063291139,
"grad_norm": 0.6163316807013067,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.3491,
"step": 580
},
{
"epoch": 2.3564556962025316,
"grad_norm": 0.5956786292406672,
"learning_rate": 2.535433837985559e-05,
"loss": 0.3286,
"step": 581
},
{
"epoch": 2.360506329113924,
"grad_norm": 0.6270311015167285,
"learning_rate": 2.529963004393324e-05,
"loss": 0.3397,
"step": 582
},
{
"epoch": 2.3645569620253166,
"grad_norm": 0.5646343155115122,
"learning_rate": 2.524487902551908e-05,
"loss": 0.3576,
"step": 583
},
{
"epoch": 2.3686075949367087,
"grad_norm": 0.560222134940068,
"learning_rate": 2.519008576557029e-05,
"loss": 0.3516,
"step": 584
},
{
"epoch": 2.372658227848101,
"grad_norm": 0.5799558380461958,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.3468,
"step": 585
},
{
"epoch": 2.3767088607594937,
"grad_norm": 0.6220545867792966,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.3702,
"step": 586
},
{
"epoch": 2.3807594936708862,
"grad_norm": 0.5614137321770905,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.3345,
"step": 587
},
{
"epoch": 2.3848101265822783,
"grad_norm": 0.5994162800963297,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.3536,
"step": 588
},
{
"epoch": 2.388860759493671,
"grad_norm": 0.5960681788431547,
"learning_rate": 2.491550129992345e-05,
"loss": 0.344,
"step": 589
},
{
"epoch": 2.3929113924050633,
"grad_norm": 0.5949353313697402,
"learning_rate": 2.486046386967024e-05,
"loss": 0.3553,
"step": 590
},
{
"epoch": 2.396962025316456,
"grad_norm": 0.6401383919993436,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.3371,
"step": 591
},
{
"epoch": 2.401012658227848,
"grad_norm": 0.6103758464887339,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.3451,
"step": 592
},
{
"epoch": 2.4050632911392404,
"grad_norm": 0.6369428679246968,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.3537,
"step": 593
},
{
"epoch": 2.409113924050633,
"grad_norm": 0.5726873433266003,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.3349,
"step": 594
},
{
"epoch": 2.4131645569620255,
"grad_norm": 0.6250919030483638,
"learning_rate": 2.458469841202444e-05,
"loss": 0.348,
"step": 595
},
{
"epoch": 2.4172151898734175,
"grad_norm": 0.5634293415125493,
"learning_rate": 2.452943276859226e-05,
"loss": 0.3496,
"step": 596
},
{
"epoch": 2.42126582278481,
"grad_norm": 0.5426988922537426,
"learning_rate": 2.447413064573125e-05,
"loss": 0.3384,
"step": 597
},
{
"epoch": 2.4253164556962026,
"grad_norm": 0.6239199093145635,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.3431,
"step": 598
},
{
"epoch": 2.429367088607595,
"grad_norm": 0.564469068129197,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.3499,
"step": 599
},
{
"epoch": 2.433417721518987,
"grad_norm": 0.6107497042560383,
"learning_rate": 2.430800985597963e-05,
"loss": 0.3202,
"step": 600
},
{
"epoch": 2.4374683544303797,
"grad_norm": 0.6148004807003055,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.3373,
"step": 601
},
{
"epoch": 2.441518987341772,
"grad_norm": 0.585198764959575,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.3406,
"step": 602
},
{
"epoch": 2.4455696202531647,
"grad_norm": 0.6084815102500295,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.3678,
"step": 603
},
{
"epoch": 2.449620253164557,
"grad_norm": 0.6069918472994575,
"learning_rate": 2.408603181104957e-05,
"loss": 0.3684,
"step": 604
},
{
"epoch": 2.4536708860759493,
"grad_norm": 0.609890177530552,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.3399,
"step": 605
},
{
"epoch": 2.457721518987342,
"grad_norm": 0.577231317726262,
"learning_rate": 2.397484355008521e-05,
"loss": 0.3502,
"step": 606
},
{
"epoch": 2.4617721518987343,
"grad_norm": 0.5931697508220705,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.338,
"step": 607
},
{
"epoch": 2.4658227848101264,
"grad_norm": 0.6328256861178968,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.3554,
"step": 608
},
{
"epoch": 2.469873417721519,
"grad_norm": 0.5694621645693468,
"learning_rate": 2.380782218329337e-05,
"loss": 0.3408,
"step": 609
},
{
"epoch": 2.4739240506329114,
"grad_norm": 0.5914088597700327,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.3298,
"step": 610
},
{
"epoch": 2.477974683544304,
"grad_norm": 0.6372676755364345,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.3475,
"step": 611
},
{
"epoch": 2.482025316455696,
"grad_norm": 0.5694556874784791,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.3409,
"step": 612
},
{
"epoch": 2.4860759493670885,
"grad_norm": 0.6205649310886041,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.3741,
"step": 613
},
{
"epoch": 2.490126582278481,
"grad_norm": 0.5949620931696684,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.3488,
"step": 614
},
{
"epoch": 2.4941772151898736,
"grad_norm": 0.5881423817699204,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.3461,
"step": 615
},
{
"epoch": 2.4982278481012656,
"grad_norm": 0.6200166139111073,
"learning_rate": 2.341705325992901e-05,
"loss": 0.3399,
"step": 616
},
{
"epoch": 2.502278481012658,
"grad_norm": 0.6391711629778365,
"learning_rate": 2.336111544604222e-05,
"loss": 0.3378,
"step": 617
},
{
"epoch": 2.5063291139240507,
"grad_norm": 0.6150394316158574,
"learning_rate": 2.33051505621937e-05,
"loss": 0.3493,
"step": 618
},
{
"epoch": 2.510379746835443,
"grad_norm": 0.5824835944259482,
"learning_rate": 2.324915905911693e-05,
"loss": 0.3217,
"step": 619
},
{
"epoch": 2.5144303797468357,
"grad_norm": 0.5700805047384699,
"learning_rate": 2.319314138775977e-05,
"loss": 0.3471,
"step": 620
},
{
"epoch": 2.5184810126582278,
"grad_norm": 0.6032355912250699,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.3365,
"step": 621
},
{
"epoch": 2.5225316455696203,
"grad_norm": 0.6098068431073456,
"learning_rate": 2.308102934504593e-05,
"loss": 0.3211,
"step": 622
},
{
"epoch": 2.526582278481013,
"grad_norm": 0.5611823890412445,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.3506,
"step": 623
},
{
"epoch": 2.530632911392405,
"grad_norm": 0.5816717685034394,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.3234,
"step": 624
},
{
"epoch": 2.5346835443037974,
"grad_norm": 0.6092433130836744,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.3338,
"step": 625
},
{
"epoch": 2.53873417721519,
"grad_norm": 0.5653203949256043,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.3356,
"step": 626
},
{
"epoch": 2.5427848101265824,
"grad_norm": 0.5824800337572055,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.3422,
"step": 627
},
{
"epoch": 2.546835443037975,
"grad_norm": 0.5849492566958202,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.3248,
"step": 628
},
{
"epoch": 2.550886075949367,
"grad_norm": 0.5480249567568154,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.3486,
"step": 629
},
{
"epoch": 2.5549367088607595,
"grad_norm": 0.5410522836614583,
"learning_rate": 2.26316247733093e-05,
"loss": 0.3282,
"step": 630
},
{
"epoch": 2.558987341772152,
"grad_norm": 0.5590794925864341,
"learning_rate": 2.257534907110328e-05,
"loss": 0.3521,
"step": 631
},
{
"epoch": 2.563037974683544,
"grad_norm": 0.5965579462121611,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.3571,
"step": 632
},
{
"epoch": 2.5670886075949366,
"grad_norm": 0.5845803589532097,
"learning_rate": 2.246273589557294e-05,
"loss": 0.3398,
"step": 633
},
{
"epoch": 2.571139240506329,
"grad_norm": 0.5738269095651082,
"learning_rate": 2.240639932921966e-05,
"loss": 0.3608,
"step": 634
},
{
"epoch": 2.5751898734177217,
"grad_norm": 0.5838185083716536,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.3271,
"step": 635
},
{
"epoch": 2.579240506329114,
"grad_norm": 0.5623328264743845,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.3521,
"step": 636
},
{
"epoch": 2.5832911392405062,
"grad_norm": 0.5811178086828238,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.3449,
"step": 637
},
{
"epoch": 2.5873417721518988,
"grad_norm": 0.5524979298220503,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.358,
"step": 638
},
{
"epoch": 2.5913924050632913,
"grad_norm": 0.5392055659567214,
"learning_rate": 2.212443486523819e-05,
"loss": 0.3379,
"step": 639
},
{
"epoch": 2.5954430379746833,
"grad_norm": 0.6034061749392852,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.3616,
"step": 640
},
{
"epoch": 2.599493670886076,
"grad_norm": 0.5478327330722319,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.3454,
"step": 641
},
{
"epoch": 2.6035443037974684,
"grad_norm": 0.5862961673005075,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.3439,
"step": 642
},
{
"epoch": 2.607594936708861,
"grad_norm": 0.5370123835677463,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.3301,
"step": 643
},
{
"epoch": 2.6116455696202534,
"grad_norm": 0.5797280811385412,
"learning_rate": 2.184204266048803e-05,
"loss": 0.3422,
"step": 644
},
{
"epoch": 2.6156962025316455,
"grad_norm": 0.5780640245238416,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.3306,
"step": 645
},
{
"epoch": 2.619746835443038,
"grad_norm": 0.5739702018293407,
"learning_rate": 2.17289787446841e-05,
"loss": 0.3355,
"step": 646
},
{
"epoch": 2.6237974683544305,
"grad_norm": 0.5542450516668087,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.3317,
"step": 647
},
{
"epoch": 2.6278481012658226,
"grad_norm": 0.5791759205044079,
"learning_rate": 2.161585912916385e-05,
"loss": 0.3253,
"step": 648
},
{
"epoch": 2.631898734177215,
"grad_norm": 0.5436588461490447,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.3287,
"step": 649
},
{
"epoch": 2.6359493670886076,
"grad_norm": 0.5729135281979345,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.3538,
"step": 650
},
{
"epoch": 2.64,
"grad_norm": 0.5493833480223291,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.356,
"step": 651
},
{
"epoch": 2.6440506329113926,
"grad_norm": 0.5824996660850993,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.3592,
"step": 652
},
{
"epoch": 2.6481012658227847,
"grad_norm": 0.5796961334241304,
"learning_rate": 2.133284032321232e-05,
"loss": 0.3388,
"step": 653
},
{
"epoch": 2.6521518987341772,
"grad_norm": 0.5854679006089802,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.3444,
"step": 654
},
{
"epoch": 2.6562025316455697,
"grad_norm": 0.5790567726146737,
"learning_rate": 2.121955446741306e-05,
"loss": 0.3259,
"step": 655
},
{
"epoch": 2.660253164556962,
"grad_norm": 0.5510804997349465,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.3455,
"step": 656
},
{
"epoch": 2.6643037974683543,
"grad_norm": 0.6158865425558006,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.3366,
"step": 657
},
{
"epoch": 2.668354430379747,
"grad_norm": 0.5558857935466524,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.3339,
"step": 658
},
{
"epoch": 2.6724050632911394,
"grad_norm": 0.5674989391400371,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.3488,
"step": 659
},
{
"epoch": 2.676455696202532,
"grad_norm": 0.5857708234658582,
"learning_rate": 2.093617592756158e-05,
"loss": 0.3441,
"step": 660
},
{
"epoch": 2.680506329113924,
"grad_norm": 0.618637711438647,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.3467,
"step": 661
},
{
"epoch": 2.6845569620253165,
"grad_norm": 0.5667633896796859,
"learning_rate": 2.082276853708978e-05,
"loss": 0.3312,
"step": 662
},
{
"epoch": 2.688607594936709,
"grad_norm": 0.5730583417997402,
"learning_rate": 2.076605467380071e-05,
"loss": 0.3329,
"step": 663
},
{
"epoch": 2.692658227848101,
"grad_norm": 0.5796156878566823,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.3325,
"step": 664
},
{
"epoch": 2.6967088607594936,
"grad_norm": 0.5615308163607495,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.3485,
"step": 665
},
{
"epoch": 2.700759493670886,
"grad_norm": 0.5672110977280976,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.3589,
"step": 666
},
{
"epoch": 2.7048101265822786,
"grad_norm": 0.5476120326009924,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.3381,
"step": 667
},
{
"epoch": 2.708860759493671,
"grad_norm": 0.5591260080713659,
"learning_rate": 2.048240194884496e-05,
"loss": 0.3475,
"step": 668
},
{
"epoch": 2.712911392405063,
"grad_norm": 0.5593272006907719,
"learning_rate": 2.042565792045695e-05,
"loss": 0.3497,
"step": 669
},
{
"epoch": 2.7169620253164557,
"grad_norm": 0.5404174641936051,
"learning_rate": 2.036891046387857e-05,
"loss": 0.3476,
"step": 670
},
{
"epoch": 2.721012658227848,
"grad_norm": 0.5662093780807766,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.356,
"step": 671
},
{
"epoch": 2.7250632911392403,
"grad_norm": 0.5394154801408607,
"learning_rate": 2.025540709431948e-05,
"loss": 0.3516,
"step": 672
},
{
"epoch": 2.729113924050633,
"grad_norm": 0.5700341146869652,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.3172,
"step": 673
},
{
"epoch": 2.7331645569620253,
"grad_norm": 0.5821191501614951,
"learning_rate": 2.014189549672245e-05,
"loss": 0.3605,
"step": 674
},
{
"epoch": 2.737215189873418,
"grad_norm": 0.564584282459589,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.347,
"step": 675
},
{
"epoch": 2.7412658227848103,
"grad_norm": 0.5481493100783352,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.3384,
"step": 676
},
{
"epoch": 2.7453164556962024,
"grad_norm": 0.567671721706152,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.3437,
"step": 677
},
{
"epoch": 2.749367088607595,
"grad_norm": 0.5577584003993251,
"learning_rate": 1.991486224484123e-05,
"loss": 0.332,
"step": 678
},
{
"epoch": 2.7534177215189874,
"grad_norm": 0.5739843290179172,
"learning_rate": 1.985810450327756e-05,
"loss": 0.3438,
"step": 679
},
{
"epoch": 2.7574683544303795,
"grad_norm": 0.5621491281541385,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.3493,
"step": 680
},
{
"epoch": 2.761518987341772,
"grad_norm": 0.5940702391610737,
"learning_rate": 1.974459290568053e-05,
"loss": 0.3574,
"step": 681
},
{
"epoch": 2.7655696202531646,
"grad_norm": 0.5369761231257867,
"learning_rate": 1.968783996385397e-05,
"loss": 0.3542,
"step": 682
},
{
"epoch": 2.769620253164557,
"grad_norm": 0.6053919321436417,
"learning_rate": 1.963108953612143e-05,
"loss": 0.3268,
"step": 683
},
{
"epoch": 2.7736708860759496,
"grad_norm": 0.5870997953422794,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.3352,
"step": 684
},
{
"epoch": 2.7777215189873417,
"grad_norm": 0.5629607233129359,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.3259,
"step": 685
},
{
"epoch": 2.781772151898734,
"grad_norm": 0.5694823013565996,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.3516,
"step": 686
},
{
"epoch": 2.7858227848101267,
"grad_norm": 0.5622445141275623,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.3334,
"step": 687
},
{
"epoch": 2.7898734177215188,
"grad_norm": 0.5750125258413183,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.3446,
"step": 688
},
{
"epoch": 2.7939240506329113,
"grad_norm": 0.5326427801363953,
"learning_rate": 1.92906653591874e-05,
"loss": 0.3434,
"step": 689
},
{
"epoch": 2.797974683544304,
"grad_norm": 0.5800928286953103,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.3545,
"step": 690
},
{
"epoch": 2.8020253164556963,
"grad_norm": 0.5487235751188719,
"learning_rate": 1.917723146291022e-05,
"loss": 0.3438,
"step": 691
},
{
"epoch": 2.806075949367089,
"grad_norm": 0.5522554298883271,
"learning_rate": 1.912052422608584e-05,
"loss": 0.3439,
"step": 692
},
{
"epoch": 2.810126582278481,
"grad_norm": 0.5890888538610736,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.3478,
"step": 693
},
{
"epoch": 2.8141772151898734,
"grad_norm": 0.5353201045508285,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.3696,
"step": 694
},
{
"epoch": 2.818227848101266,
"grad_norm": 0.5621832535652342,
"learning_rate": 1.895044684123479e-05,
"loss": 0.3536,
"step": 695
},
{
"epoch": 2.822278481012658,
"grad_norm": 0.5653643439437396,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.3401,
"step": 696
},
{
"epoch": 2.8263291139240505,
"grad_norm": 0.537870021449332,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.3579,
"step": 697
},
{
"epoch": 2.830379746835443,
"grad_norm": 0.5615328742345641,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.3483,
"step": 698
},
{
"epoch": 2.8344303797468355,
"grad_norm": 0.5527262359031931,
"learning_rate": 1.872379746550743e-05,
"loss": 0.3399,
"step": 699
},
{
"epoch": 2.838481012658228,
"grad_norm": 0.565508020258709,
"learning_rate": 1.866715967678769e-05,
"loss": 0.3299,
"step": 700
},
{
"epoch": 2.84253164556962,
"grad_norm": 0.5728689643335386,
"learning_rate": 1.861053262258067e-05,
"loss": 0.3391,
"step": 701
},
{
"epoch": 2.8465822784810126,
"grad_norm": 0.547934184191893,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.3304,
"step": 702
},
{
"epoch": 2.850632911392405,
"grad_norm": 0.583761546067117,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.3457,
"step": 703
},
{
"epoch": 2.8546835443037972,
"grad_norm": 0.5623606302104207,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.3239,
"step": 704
},
{
"epoch": 2.8587341772151897,
"grad_norm": 0.5595781003606272,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.347,
"step": 705
},
{
"epoch": 2.8627848101265823,
"grad_norm": 0.5649287428317293,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.3303,
"step": 706
},
{
"epoch": 2.8668354430379748,
"grad_norm": 0.5580092801691529,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.3386,
"step": 707
},
{
"epoch": 2.8708860759493673,
"grad_norm": 0.5657357082963987,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.3109,
"step": 708
},
{
"epoch": 2.8749367088607594,
"grad_norm": 0.5680244531133938,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.3275,
"step": 709
},
{
"epoch": 2.878987341772152,
"grad_norm": 0.5409620111843843,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.352,
"step": 710
},
{
"epoch": 2.8830379746835444,
"grad_norm": 0.5381547246725156,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.3411,
"step": 711
},
{
"epoch": 2.8870886075949365,
"grad_norm": 0.5636564596986513,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.3404,
"step": 712
},
{
"epoch": 2.891139240506329,
"grad_norm": 0.5503801995240662,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.3496,
"step": 713
},
{
"epoch": 2.8951898734177215,
"grad_norm": 0.5608612374957381,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.323,
"step": 714
},
{
"epoch": 2.899240506329114,
"grad_norm": 0.5534794305593992,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.3383,
"step": 715
},
{
"epoch": 2.9032911392405065,
"grad_norm": 0.5731128157496599,
"learning_rate": 1.776272483901182e-05,
"loss": 0.3348,
"step": 716
},
{
"epoch": 2.9073417721518986,
"grad_norm": 0.5804946441588634,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.336,
"step": 717
},
{
"epoch": 2.911392405063291,
"grad_norm": 0.565054665995655,
"learning_rate": 1.764995661794301e-05,
"loss": 0.3222,
"step": 718
},
{
"epoch": 2.9154430379746836,
"grad_norm": 0.5663180664633102,
"learning_rate": 1.759360067078035e-05,
"loss": 0.3258,
"step": 719
},
{
"epoch": 2.9194936708860757,
"grad_norm": 0.5849369724493076,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.348,
"step": 720
},
{
"epoch": 2.923544303797468,
"grad_norm": 0.5739907987009579,
"learning_rate": 1.748094737261012e-05,
"loss": 0.3213,
"step": 721
},
{
"epoch": 2.9275949367088607,
"grad_norm": 0.581605104732831,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.344,
"step": 722
},
{
"epoch": 2.9316455696202532,
"grad_norm": 0.614260698094826,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.3355,
"step": 723
},
{
"epoch": 2.9356962025316458,
"grad_norm": 0.5606801354136242,
"learning_rate": 1.731212071922883e-05,
"loss": 0.3546,
"step": 724
},
{
"epoch": 2.939746835443038,
"grad_norm": 0.562167519785487,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.3584,
"step": 725
},
{
"epoch": 2.9437974683544303,
"grad_norm": 0.5866310032691988,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.336,
"step": 726
},
{
"epoch": 2.947848101265823,
"grad_norm": 0.5512917706196985,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.3554,
"step": 727
},
{
"epoch": 2.951898734177215,
"grad_norm": 0.573082766585085,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.3314,
"step": 728
},
{
"epoch": 2.9559493670886074,
"grad_norm": 0.569588932690869,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.3443,
"step": 729
},
{
"epoch": 2.96,
"grad_norm": 0.5728505186270498,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.3534,
"step": 730
},
{
"epoch": 2.9640506329113925,
"grad_norm": 0.5656792822413913,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.3375,
"step": 731
},
{
"epoch": 2.968101265822785,
"grad_norm": 0.5413627457101798,
"learning_rate": 1.686290200071915e-05,
"loss": 0.3382,
"step": 732
},
{
"epoch": 2.972151898734177,
"grad_norm": 0.542289758840477,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.3516,
"step": 733
},
{
"epoch": 2.9762025316455696,
"grad_norm": 0.5483461762147299,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.3413,
"step": 734
},
{
"epoch": 2.980253164556962,
"grad_norm": 0.5624119206435044,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.3299,
"step": 735
},
{
"epoch": 2.984303797468354,
"grad_norm": 0.5453164568181075,
"learning_rate": 1.663888455395778e-05,
"loss": 0.3214,
"step": 736
},
{
"epoch": 2.9883544303797467,
"grad_norm": 0.5675962032184569,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.3219,
"step": 737
},
{
"epoch": 2.992405063291139,
"grad_norm": 0.5617116303959675,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.3324,
"step": 738
},
{
"epoch": 2.9964556962025317,
"grad_norm": 0.5535879777847904,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.3441,
"step": 739
},
{
"epoch": 3.002025316455696,
"grad_norm": 0.5986917124772161,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.2641,
"step": 740
},
{
"epoch": 3.0060759493670886,
"grad_norm": 0.6271191726518304,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.1718,
"step": 741
},
{
"epoch": 3.010126582278481,
"grad_norm": 0.5036777518389839,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.1836,
"step": 742
},
{
"epoch": 3.0141772151898736,
"grad_norm": 0.5009804001061989,
"learning_rate": 1.624791353909428e-05,
"loss": 0.1863,
"step": 743
},
{
"epoch": 3.0182278481012657,
"grad_norm": 0.7508569881457368,
"learning_rate": 1.619217781670663e-05,
"loss": 0.1994,
"step": 744
},
{
"epoch": 3.022278481012658,
"grad_norm": 0.6700795764353196,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.1782,
"step": 745
},
{
"epoch": 3.0263291139240507,
"grad_norm": 0.5966905163911352,
"learning_rate": 1.608079882360134e-05,
"loss": 0.1774,
"step": 746
},
{
"epoch": 3.030379746835443,
"grad_norm": 0.49708444793731255,
"learning_rate": 1.60251564499148e-05,
"loss": 0.1777,
"step": 747
},
{
"epoch": 3.0344303797468353,
"grad_norm": 0.5828411533217915,
"learning_rate": 1.596954608907176e-05,
"loss": 0.1942,
"step": 748
},
{
"epoch": 3.038481012658228,
"grad_norm": 0.6006215059216979,
"learning_rate": 1.591396818895043e-05,
"loss": 0.1804,
"step": 749
},
{
"epoch": 3.0425316455696203,
"grad_norm": 0.5237349129685631,
"learning_rate": 1.585842319716759e-05,
"loss": 0.1775,
"step": 750
},
{
"epoch": 3.046582278481013,
"grad_norm": 0.454200449830496,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.189,
"step": 751
},
{
"epoch": 3.050632911392405,
"grad_norm": 0.47905243290177635,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.192,
"step": 752
},
{
"epoch": 3.0546835443037974,
"grad_norm": 0.5348469377201885,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.1763,
"step": 753
},
{
"epoch": 3.05873417721519,
"grad_norm": 0.4983700499257158,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.1718,
"step": 754
},
{
"epoch": 3.0627848101265824,
"grad_norm": 0.4939992836440564,
"learning_rate": 1.558120751116291e-05,
"loss": 0.1637,
"step": 755
},
{
"epoch": 3.0668354430379745,
"grad_norm": 0.48621065436318955,
"learning_rate": 1.552586935426876e-05,
"loss": 0.1773,
"step": 756
},
{
"epoch": 3.070886075949367,
"grad_norm": 0.4464416074435812,
"learning_rate": 1.547056723140774e-05,
"loss": 0.152,
"step": 757
},
{
"epoch": 3.0749367088607595,
"grad_norm": 0.4853614956815793,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.1789,
"step": 758
},
{
"epoch": 3.078987341772152,
"grad_norm": 0.4740005353636371,
"learning_rate": 1.536007286907411e-05,
"loss": 0.1674,
"step": 759
},
{
"epoch": 3.083037974683544,
"grad_norm": 0.460100232813479,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.1687,
"step": 760
},
{
"epoch": 3.0870886075949366,
"grad_norm": 0.44915782106795243,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.1781,
"step": 761
},
{
"epoch": 3.091139240506329,
"grad_norm": 0.4520499342767741,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.1863,
"step": 762
},
{
"epoch": 3.0951898734177217,
"grad_norm": 0.4463831006384255,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.1699,
"step": 763
},
{
"epoch": 3.0992405063291137,
"grad_norm": 0.4583920395518712,
"learning_rate": 1.508449870007656e-05,
"loss": 0.1743,
"step": 764
},
{
"epoch": 3.1032911392405063,
"grad_norm": 0.4888046574592615,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.1697,
"step": 765
},
{
"epoch": 3.1073417721518988,
"grad_norm": 0.4532346430331465,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.1788,
"step": 766
},
{
"epoch": 3.1113924050632913,
"grad_norm": 0.46258284582077647,
"learning_rate": 1.4919625713405e-05,
"loss": 0.1688,
"step": 767
},
{
"epoch": 3.1154430379746834,
"grad_norm": 0.44452868159084585,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.1773,
"step": 768
},
{
"epoch": 3.119493670886076,
"grad_norm": 0.4588800230496731,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.1815,
"step": 769
},
{
"epoch": 3.1235443037974684,
"grad_norm": 0.45720925262901213,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.1931,
"step": 770
},
{
"epoch": 3.127594936708861,
"grad_norm": 0.44464675022073513,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.1815,
"step": 771
},
{
"epoch": 3.131645569620253,
"grad_norm": 0.46196903603487816,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.1588,
"step": 772
},
{
"epoch": 3.1356962025316455,
"grad_norm": 0.4593184173725093,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.1749,
"step": 773
},
{
"epoch": 3.139746835443038,
"grad_norm": 0.4722438338243619,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.185,
"step": 774
},
{
"epoch": 3.1437974683544305,
"grad_norm": 0.4693741916358073,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.1737,
"step": 775
},
{
"epoch": 3.1478481012658226,
"grad_norm": 0.4538633357836371,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.1947,
"step": 776
},
{
"epoch": 3.151898734177215,
"grad_norm": 0.4665174872041237,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.1883,
"step": 777
},
{
"epoch": 3.1559493670886076,
"grad_norm": 0.4444915698424835,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.1826,
"step": 778
},
{
"epoch": 3.16,
"grad_norm": 0.44791728673686704,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.1719,
"step": 779
},
{
"epoch": 3.164050632911392,
"grad_norm": 0.46878343393553196,
"learning_rate": 1.420958430225303e-05,
"loss": 0.1892,
"step": 780
},
{
"epoch": 3.1681012658227847,
"grad_norm": 0.4398459497717241,
"learning_rate": 1.415527989393985e-05,
"loss": 0.1838,
"step": 781
},
{
"epoch": 3.1721518987341772,
"grad_norm": 0.4667262279413063,
"learning_rate": 1.410102255819891e-05,
"loss": 0.1983,
"step": 782
},
{
"epoch": 3.1762025316455698,
"grad_norm": 0.4526078614833072,
"learning_rate": 1.404681273201131e-05,
"loss": 0.1604,
"step": 783
},
{
"epoch": 3.180253164556962,
"grad_norm": 0.4358681379793921,
"learning_rate": 1.399265085197556e-05,
"loss": 0.1806,
"step": 784
},
{
"epoch": 3.1843037974683543,
"grad_norm": 0.4268875000853456,
"learning_rate": 1.393853735430398e-05,
"loss": 0.1704,
"step": 785
},
{
"epoch": 3.188354430379747,
"grad_norm": 0.4519634011071898,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.1684,
"step": 786
},
{
"epoch": 3.1924050632911394,
"grad_norm": 0.4583551991392413,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.1697,
"step": 787
},
{
"epoch": 3.1964556962025314,
"grad_norm": 0.45136440561365443,
"learning_rate": 1.377649151173163e-05,
"loss": 0.1587,
"step": 788
},
{
"epoch": 3.200506329113924,
"grad_norm": 0.4632298649173296,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.1622,
"step": 789
},
{
"epoch": 3.2045569620253165,
"grad_norm": 0.4642617403309669,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.1849,
"step": 790
},
{
"epoch": 3.208607594936709,
"grad_norm": 0.44607969688754784,
"learning_rate": 1.361489677627324e-05,
"loss": 0.1757,
"step": 791
},
{
"epoch": 3.212658227848101,
"grad_norm": 0.4633938638131549,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.168,
"step": 792
},
{
"epoch": 3.2167088607594936,
"grad_norm": 0.4448598284847409,
"learning_rate": 1.350742335331241e-05,
"loss": 0.1684,
"step": 793
},
{
"epoch": 3.220759493670886,
"grad_norm": 0.44930650022508356,
"learning_rate": 1.345376486102198e-05,
"loss": 0.1619,
"step": 794
},
{
"epoch": 3.2248101265822786,
"grad_norm": 0.4714730257780953,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.1744,
"step": 795
},
{
"epoch": 3.2288607594936707,
"grad_norm": 0.44262153352822087,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.1623,
"step": 796
},
{
"epoch": 3.232911392405063,
"grad_norm": 0.4546661657077998,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.1755,
"step": 797
},
{
"epoch": 3.2369620253164557,
"grad_norm": 0.4348423642995229,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.1585,
"step": 798
},
{
"epoch": 3.2410126582278482,
"grad_norm": 0.4350796650152279,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.1628,
"step": 799
},
{
"epoch": 3.2450632911392403,
"grad_norm": 0.4514691712060649,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.1486,
"step": 800
},
{
"epoch": 3.249113924050633,
"grad_norm": 0.4599594803131515,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.1832,
"step": 801
},
{
"epoch": 3.2531645569620253,
"grad_norm": 0.4571549457099645,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.1664,
"step": 802
},
{
"epoch": 3.257215189873418,
"grad_norm": 0.4646023850361749,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.1616,
"step": 803
},
{
"epoch": 3.26126582278481,
"grad_norm": 0.4522131643226597,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.1693,
"step": 804
},
{
"epoch": 3.2653164556962024,
"grad_norm": 0.4536930973390179,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.1875,
"step": 805
},
{
"epoch": 3.269367088607595,
"grad_norm": 0.4519634090723797,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.1686,
"step": 806
},
{
"epoch": 3.2734177215189875,
"grad_norm": 0.4439121736080914,
"learning_rate": 1.276115892225764e-05,
"loss": 0.188,
"step": 807
},
{
"epoch": 3.27746835443038,
"grad_norm": 0.4367917441776648,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.1545,
"step": 808
},
{
"epoch": 3.281518987341772,
"grad_norm": 0.44085674148686754,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.1869,
"step": 809
},
{
"epoch": 3.2855696202531646,
"grad_norm": 0.44573967516571,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.1827,
"step": 810
},
{
"epoch": 3.289620253164557,
"grad_norm": 0.4594843427104667,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.1661,
"step": 811
},
{
"epoch": 3.293670886075949,
"grad_norm": 0.45656857668771283,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.1603,
"step": 812
},
{
"epoch": 3.2977215189873417,
"grad_norm": 0.4354571432815786,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.1818,
"step": 813
},
{
"epoch": 3.301772151898734,
"grad_norm": 0.4180241619207137,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.1557,
"step": 814
},
{
"epoch": 3.3058227848101267,
"grad_norm": 0.44420165183386406,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.1789,
"step": 815
},
{
"epoch": 3.309873417721519,
"grad_norm": 0.4435542975933394,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.1743,
"step": 816
},
{
"epoch": 3.3139240506329113,
"grad_norm": 0.4614680163130065,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.1657,
"step": 817
},
{
"epoch": 3.317974683544304,
"grad_norm": 0.4708751935785985,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.1574,
"step": 818
},
{
"epoch": 3.3220253164556963,
"grad_norm": 0.4454506935567742,
"learning_rate": 1.213055323844297e-05,
"loss": 0.1853,
"step": 819
},
{
"epoch": 3.3260759493670884,
"grad_norm": 0.4676979211035502,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.1558,
"step": 820
},
{
"epoch": 3.330126582278481,
"grad_norm": 0.4402967705890882,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.1757,
"step": 821
},
{
"epoch": 3.3341772151898734,
"grad_norm": 0.4426061492048064,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.1752,
"step": 822
},
{
"epoch": 3.338227848101266,
"grad_norm": 0.4384235178706347,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.1805,
"step": 823
},
{
"epoch": 3.3422784810126585,
"grad_norm": 0.45915818056217406,
"learning_rate": 1.187045262895488e-05,
"loss": 0.1596,
"step": 824
},
{
"epoch": 3.3463291139240505,
"grad_norm": 0.4501886348783903,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.1725,
"step": 825
},
{
"epoch": 3.350379746835443,
"grad_norm": 0.47601875071843747,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.1936,
"step": 826
},
{
"epoch": 3.3544303797468356,
"grad_norm": 0.45516863027957344,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.1644,
"step": 827
},
{
"epoch": 3.3584810126582276,
"grad_norm": 0.44024137363294996,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.1796,
"step": 828
},
{
"epoch": 3.36253164556962,
"grad_norm": 0.452233563916536,
"learning_rate": 1.161198884952377e-05,
"loss": 0.1456,
"step": 829
},
{
"epoch": 3.3665822784810127,
"grad_norm": 0.46220424892590106,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.1826,
"step": 830
},
{
"epoch": 3.370632911392405,
"grad_norm": 0.4664512542777458,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.1715,
"step": 831
},
{
"epoch": 3.3746835443037977,
"grad_norm": 0.4450733113528357,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.1773,
"step": 832
},
{
"epoch": 3.3787341772151898,
"grad_norm": 0.4420395155884992,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.1618,
"step": 833
},
{
"epoch": 3.3827848101265823,
"grad_norm": 0.4510280076396199,
"learning_rate": 1.135521394010506e-05,
"loss": 0.1724,
"step": 834
},
{
"epoch": 3.386835443037975,
"grad_norm": 0.42992375818372164,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.1709,
"step": 835
},
{
"epoch": 3.390886075949367,
"grad_norm": 0.4522989540067541,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.1741,
"step": 836
},
{
"epoch": 3.3949367088607594,
"grad_norm": 0.4455973676809599,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.1866,
"step": 837
},
{
"epoch": 3.398987341772152,
"grad_norm": 0.4413274852959181,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.1854,
"step": 838
},
{
"epoch": 3.4030379746835444,
"grad_norm": 0.4360545659175171,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.1707,
"step": 839
},
{
"epoch": 3.407088607594937,
"grad_norm": 0.4376441215633842,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.1644,
"step": 840
},
{
"epoch": 3.411139240506329,
"grad_norm": 0.4540430421141661,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.1856,
"step": 841
},
{
"epoch": 3.4151898734177215,
"grad_norm": 0.444554822892427,
"learning_rate": 1.0948015846744e-05,
"loss": 0.1828,
"step": 842
},
{
"epoch": 3.419240506329114,
"grad_norm": 0.4403218602923545,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.1926,
"step": 843
},
{
"epoch": 3.423291139240506,
"grad_norm": 0.45424320139631924,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.1756,
"step": 844
},
{
"epoch": 3.4273417721518986,
"grad_norm": 0.44740629361834194,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.1654,
"step": 845
},
{
"epoch": 3.431392405063291,
"grad_norm": 0.45405786412695387,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.1551,
"step": 846
},
{
"epoch": 3.4354430379746836,
"grad_norm": 0.4482189117634364,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.1742,
"step": 847
},
{
"epoch": 3.439493670886076,
"grad_norm": 0.44654619531671186,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.1732,
"step": 848
},
{
"epoch": 3.4435443037974682,
"grad_norm": 0.4616688959327128,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.172,
"step": 849
},
{
"epoch": 3.4475949367088607,
"grad_norm": 0.4332001173997673,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.175,
"step": 850
},
{
"epoch": 3.4516455696202533,
"grad_norm": 0.4467430833894322,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.1692,
"step": 851
},
{
"epoch": 3.4556962025316453,
"grad_norm": 0.459338993906443,
"learning_rate": 1.044560362975474e-05,
"loss": 0.1744,
"step": 852
},
{
"epoch": 3.459746835443038,
"grad_norm": 0.4448733455898962,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.1778,
"step": 853
},
{
"epoch": 3.4637974683544304,
"grad_norm": 0.43110006718167804,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.1819,
"step": 854
},
{
"epoch": 3.467848101265823,
"grad_norm": 0.43600919470348704,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.1586,
"step": 855
},
{
"epoch": 3.4718987341772154,
"grad_norm": 0.43118740837180214,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.1545,
"step": 856
},
{
"epoch": 3.4759493670886075,
"grad_norm": 0.4385212757797293,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.1783,
"step": 857
},
{
"epoch": 3.48,
"grad_norm": 0.4707033430408071,
"learning_rate": 1.014782400334433e-05,
"loss": 0.1791,
"step": 858
},
{
"epoch": 3.4840506329113925,
"grad_norm": 0.4417786114258212,
"learning_rate": 1.009846945771296e-05,
"loss": 0.1445,
"step": 859
},
{
"epoch": 3.4881012658227846,
"grad_norm": 0.4403840945965605,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.1682,
"step": 860
},
{
"epoch": 3.492151898734177,
"grad_norm": 0.4416852479383303,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.1848,
"step": 861
},
{
"epoch": 3.4962025316455696,
"grad_norm": 0.43681583618613956,
"learning_rate": 9.950885880977891e-06,
"loss": 0.1633,
"step": 862
},
{
"epoch": 3.500253164556962,
"grad_norm": 0.4270739165482594,
"learning_rate": 9.901852696139382e-06,
"loss": 0.1645,
"step": 863
},
{
"epoch": 3.5043037974683546,
"grad_norm": 0.41966162382285743,
"learning_rate": 9.852900840391027e-06,
"loss": 0.1753,
"step": 864
},
{
"epoch": 3.5083544303797467,
"grad_norm": 0.4296390522779729,
"learning_rate": 9.804030707984313e-06,
"loss": 0.1727,
"step": 865
},
{
"epoch": 3.512405063291139,
"grad_norm": 0.43184144089142135,
"learning_rate": 9.755242692512599e-06,
"loss": 0.1647,
"step": 866
},
{
"epoch": 3.5164556962025317,
"grad_norm": 0.44172567848727323,
"learning_rate": 9.70653718690782e-06,
"loss": 0.1672,
"step": 867
},
{
"epoch": 3.520506329113924,
"grad_norm": 0.4428046627853052,
"learning_rate": 9.657914583437454e-06,
"loss": 0.1641,
"step": 868
},
{
"epoch": 3.5245569620253163,
"grad_norm": 0.43844081698607307,
"learning_rate": 9.609375273701246e-06,
"loss": 0.1634,
"step": 869
},
{
"epoch": 3.528607594936709,
"grad_norm": 0.4551562931618646,
"learning_rate": 9.560919648628133e-06,
"loss": 0.1728,
"step": 870
},
{
"epoch": 3.5326582278481014,
"grad_norm": 0.43698343160728315,
"learning_rate": 9.512548098473047e-06,
"loss": 0.1739,
"step": 871
},
{
"epoch": 3.536708860759494,
"grad_norm": 0.42728206720642886,
"learning_rate": 9.464261012813825e-06,
"loss": 0.167,
"step": 872
},
{
"epoch": 3.540759493670886,
"grad_norm": 0.43966658607606734,
"learning_rate": 9.416058780547987e-06,
"loss": 0.1823,
"step": 873
},
{
"epoch": 3.5448101265822785,
"grad_norm": 0.42810758940556265,
"learning_rate": 9.367941789889714e-06,
"loss": 0.1627,
"step": 874
},
{
"epoch": 3.548860759493671,
"grad_norm": 0.4400715191434691,
"learning_rate": 9.319910428366607e-06,
"loss": 0.1912,
"step": 875
},
{
"epoch": 3.552911392405063,
"grad_norm": 0.44496976826010265,
"learning_rate": 9.271965082816667e-06,
"loss": 0.1713,
"step": 876
},
{
"epoch": 3.5569620253164556,
"grad_norm": 0.4272606183129963,
"learning_rate": 9.224106139385111e-06,
"loss": 0.1664,
"step": 877
},
{
"epoch": 3.561012658227848,
"grad_norm": 0.4258899170671759,
"learning_rate": 9.176333983521291e-06,
"loss": 0.1711,
"step": 878
},
{
"epoch": 3.5650632911392406,
"grad_norm": 0.4234072144757895,
"learning_rate": 9.12864899997558e-06,
"loss": 0.1732,
"step": 879
},
{
"epoch": 3.569113924050633,
"grad_norm": 0.43510207127452494,
"learning_rate": 9.08105157279628e-06,
"loss": 0.1592,
"step": 880
},
{
"epoch": 3.573164556962025,
"grad_norm": 0.42752214101486974,
"learning_rate": 9.03354208532653e-06,
"loss": 0.1906,
"step": 881
},
{
"epoch": 3.5772151898734177,
"grad_norm": 0.41102171797595527,
"learning_rate": 8.986120920201205e-06,
"loss": 0.1609,
"step": 882
},
{
"epoch": 3.58126582278481,
"grad_norm": 0.451267295890508,
"learning_rate": 8.938788459343852e-06,
"loss": 0.1736,
"step": 883
},
{
"epoch": 3.5853164556962023,
"grad_norm": 0.44264588432606167,
"learning_rate": 8.8915450839636e-06,
"loss": 0.1472,
"step": 884
},
{
"epoch": 3.589367088607595,
"grad_norm": 0.4363782520831449,
"learning_rate": 8.844391174552116e-06,
"loss": 0.1568,
"step": 885
},
{
"epoch": 3.5934177215189873,
"grad_norm": 0.4358970283235929,
"learning_rate": 8.797327110880479e-06,
"loss": 0.155,
"step": 886
},
{
"epoch": 3.59746835443038,
"grad_norm": 0.4326810691203048,
"learning_rate": 8.750353271996206e-06,
"loss": 0.1684,
"step": 887
},
{
"epoch": 3.6015189873417723,
"grad_norm": 0.4330449266373825,
"learning_rate": 8.703470036220132e-06,
"loss": 0.1814,
"step": 888
},
{
"epoch": 3.6055696202531644,
"grad_norm": 0.4510312612424354,
"learning_rate": 8.656677781143394e-06,
"loss": 0.1711,
"step": 889
},
{
"epoch": 3.609620253164557,
"grad_norm": 0.4514071266646793,
"learning_rate": 8.609976883624377e-06,
"loss": 0.1459,
"step": 890
},
{
"epoch": 3.6136708860759494,
"grad_norm": 0.4373635903187739,
"learning_rate": 8.563367719785698e-06,
"loss": 0.1742,
"step": 891
},
{
"epoch": 3.6177215189873415,
"grad_norm": 0.4404394494473405,
"learning_rate": 8.516850665011138e-06,
"loss": 0.1777,
"step": 892
},
{
"epoch": 3.621772151898734,
"grad_norm": 0.42746455589595544,
"learning_rate": 8.47042609394269e-06,
"loss": 0.18,
"step": 893
},
{
"epoch": 3.6258227848101265,
"grad_norm": 0.4472837152292449,
"learning_rate": 8.424094380477432e-06,
"loss": 0.1641,
"step": 894
},
{
"epoch": 3.629873417721519,
"grad_norm": 0.4263661584419103,
"learning_rate": 8.37785589776465e-06,
"loss": 0.1575,
"step": 895
},
{
"epoch": 3.6339240506329116,
"grad_norm": 0.4445021263484194,
"learning_rate": 8.331711018202694e-06,
"loss": 0.1634,
"step": 896
},
{
"epoch": 3.6379746835443036,
"grad_norm": 0.42827015314636085,
"learning_rate": 8.285660113436104e-06,
"loss": 0.1566,
"step": 897
},
{
"epoch": 3.642025316455696,
"grad_norm": 0.4553733880362178,
"learning_rate": 8.239703554352527e-06,
"loss": 0.1492,
"step": 898
},
{
"epoch": 3.6460759493670887,
"grad_norm": 0.43742658238706217,
"learning_rate": 8.193841711079775e-06,
"loss": 0.1774,
"step": 899
},
{
"epoch": 3.6501265822784807,
"grad_norm": 0.43532034624662413,
"learning_rate": 8.148074952982828e-06,
"loss": 0.1596,
"step": 900
},
{
"epoch": 3.6541772151898733,
"grad_norm": 0.4327289753450771,
"learning_rate": 8.102403648660859e-06,
"loss": 0.1658,
"step": 901
},
{
"epoch": 3.6582278481012658,
"grad_norm": 0.4305170584748057,
"learning_rate": 8.056828165944282e-06,
"loss": 0.157,
"step": 902
},
{
"epoch": 3.6622784810126583,
"grad_norm": 0.4418309015442352,
"learning_rate": 8.011348871891762e-06,
"loss": 0.1597,
"step": 903
},
{
"epoch": 3.666329113924051,
"grad_norm": 0.43013007123832203,
"learning_rate": 7.965966132787287e-06,
"loss": 0.1604,
"step": 904
},
{
"epoch": 3.670379746835443,
"grad_norm": 0.43623781863733624,
"learning_rate": 7.920680314137189e-06,
"loss": 0.1745,
"step": 905
},
{
"epoch": 3.6744303797468354,
"grad_norm": 0.4219677325134748,
"learning_rate": 7.875491780667246e-06,
"loss": 0.162,
"step": 906
},
{
"epoch": 3.678481012658228,
"grad_norm": 0.44047750387424917,
"learning_rate": 7.830400896319667e-06,
"loss": 0.1964,
"step": 907
},
{
"epoch": 3.68253164556962,
"grad_norm": 0.4493272059656011,
"learning_rate": 7.785408024250259e-06,
"loss": 0.1822,
"step": 908
},
{
"epoch": 3.6865822784810125,
"grad_norm": 0.4328106709229672,
"learning_rate": 7.74051352682542e-06,
"loss": 0.1514,
"step": 909
},
{
"epoch": 3.690632911392405,
"grad_norm": 0.4357964654605219,
"learning_rate": 7.695717765619257e-06,
"loss": 0.162,
"step": 910
},
{
"epoch": 3.6946835443037975,
"grad_norm": 0.4409615901117508,
"learning_rate": 7.651021101410673e-06,
"loss": 0.1567,
"step": 911
},
{
"epoch": 3.69873417721519,
"grad_norm": 0.4379276744215732,
"learning_rate": 7.606423894180464e-06,
"loss": 0.1608,
"step": 912
},
{
"epoch": 3.702784810126582,
"grad_norm": 0.4301916004812862,
"learning_rate": 7.56192650310839e-06,
"loss": 0.1703,
"step": 913
},
{
"epoch": 3.7068354430379746,
"grad_norm": 0.4405538776753518,
"learning_rate": 7.517529286570349e-06,
"loss": 0.1665,
"step": 914
},
{
"epoch": 3.710886075949367,
"grad_norm": 0.44168161536217654,
"learning_rate": 7.473232602135387e-06,
"loss": 0.1676,
"step": 915
},
{
"epoch": 3.714936708860759,
"grad_norm": 0.4358360230636989,
"learning_rate": 7.429036806562935e-06,
"loss": 0.1745,
"step": 916
},
{
"epoch": 3.7189873417721517,
"grad_norm": 0.4597843389845769,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.1593,
"step": 917
},
{
"epoch": 3.7230379746835442,
"grad_norm": 0.44371190545690353,
"learning_rate": 7.340949304977567e-06,
"loss": 0.1696,
"step": 918
},
{
"epoch": 3.7270886075949368,
"grad_norm": 0.4357409095517997,
"learning_rate": 7.297058308409282e-06,
"loss": 0.1613,
"step": 919
},
{
"epoch": 3.7311392405063293,
"grad_norm": 0.4323323184522045,
"learning_rate": 7.25326961958704e-06,
"loss": 0.1673,
"step": 920
},
{
"epoch": 3.7351898734177214,
"grad_norm": 0.41922303166900643,
"learning_rate": 7.209583591178921e-06,
"loss": 0.1698,
"step": 921
},
{
"epoch": 3.739240506329114,
"grad_norm": 0.4294866227219242,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.1669,
"step": 922
},
{
"epoch": 3.7432911392405064,
"grad_norm": 0.4313925986144899,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.1686,
"step": 923
},
{
"epoch": 3.747341772151899,
"grad_norm": 0.4387418351703761,
"learning_rate": 7.079144982700909e-06,
"loss": 0.1601,
"step": 924
},
{
"epoch": 3.7513924050632914,
"grad_norm": 0.44492725574447156,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.1588,
"step": 925
},
{
"epoch": 3.7554430379746835,
"grad_norm": 0.4346465761551741,
"learning_rate": 6.99270564069757e-06,
"loss": 0.1527,
"step": 926
},
{
"epoch": 3.759493670886076,
"grad_norm": 0.4270236511868173,
"learning_rate": 6.949642934304375e-06,
"loss": 0.1666,
"step": 927
},
{
"epoch": 3.7635443037974685,
"grad_norm": 0.428614638180669,
"learning_rate": 6.906685333692871e-06,
"loss": 0.1608,
"step": 928
},
{
"epoch": 3.7675949367088606,
"grad_norm": 0.4373107373948653,
"learning_rate": 6.86383318483769e-06,
"loss": 0.1508,
"step": 929
},
{
"epoch": 3.771645569620253,
"grad_norm": 0.43942591963337796,
"learning_rate": 6.821086832864139e-06,
"loss": 0.1718,
"step": 930
},
{
"epoch": 3.7756962025316456,
"grad_norm": 0.4330969382784416,
"learning_rate": 6.77844662204546e-06,
"loss": 0.156,
"step": 931
},
{
"epoch": 3.779746835443038,
"grad_norm": 0.4261270869400158,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.167,
"step": 932
},
{
"epoch": 3.7837974683544306,
"grad_norm": 0.43358906429866867,
"learning_rate": 6.693485996688695e-06,
"loss": 0.1719,
"step": 933
},
{
"epoch": 3.7878481012658227,
"grad_norm": 0.4432895591222381,
"learning_rate": 6.651166266411801e-06,
"loss": 0.1536,
"step": 934
},
{
"epoch": 3.7918987341772152,
"grad_norm": 0.42609514205560683,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.1591,
"step": 935
},
{
"epoch": 3.7959493670886078,
"grad_norm": 0.4190789190322287,
"learning_rate": 6.566849674844711e-06,
"loss": 0.1628,
"step": 936
},
{
"epoch": 3.8,
"grad_norm": 0.4315756325964381,
"learning_rate": 6.524853492628747e-06,
"loss": 0.1857,
"step": 937
},
{
"epoch": 3.8040506329113923,
"grad_norm": 0.42618503219596127,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.1657,
"step": 938
},
{
"epoch": 3.808101265822785,
"grad_norm": 0.45714482365477366,
"learning_rate": 6.441187046486648e-06,
"loss": 0.1708,
"step": 939
},
{
"epoch": 3.8121518987341774,
"grad_norm": 0.45016089122942105,
"learning_rate": 6.399517456398567e-06,
"loss": 0.1823,
"step": 940
},
{
"epoch": 3.81620253164557,
"grad_norm": 0.42661721763119387,
"learning_rate": 6.357957402727164e-06,
"loss": 0.1762,
"step": 941
},
{
"epoch": 3.820253164556962,
"grad_norm": 0.4465157325244347,
"learning_rate": 6.316507220191395e-06,
"loss": 0.1573,
"step": 942
},
{
"epoch": 3.8243037974683545,
"grad_norm": 0.42429927215361857,
"learning_rate": 6.275167242625331e-06,
"loss": 0.1474,
"step": 943
},
{
"epoch": 3.828354430379747,
"grad_norm": 0.4218731908492854,
"learning_rate": 6.233937802975471e-06,
"loss": 0.1675,
"step": 944
},
{
"epoch": 3.832405063291139,
"grad_norm": 0.42729607976864475,
"learning_rate": 6.192819233298046e-06,
"loss": 0.1554,
"step": 945
},
{
"epoch": 3.8364556962025316,
"grad_norm": 0.40481571421789014,
"learning_rate": 6.151811864756383e-06,
"loss": 0.1716,
"step": 946
},
{
"epoch": 3.840506329113924,
"grad_norm": 0.42826077201934715,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.1618,
"step": 947
},
{
"epoch": 3.8445569620253166,
"grad_norm": 0.42654755012927636,
"learning_rate": 6.070132051252868e-06,
"loss": 0.1536,
"step": 948
},
{
"epoch": 3.848607594936709,
"grad_norm": 0.43735328994788186,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.1574,
"step": 949
},
{
"epoch": 3.852658227848101,
"grad_norm": 0.4480604554400636,
"learning_rate": 5.988900993811575e-06,
"loss": 0.1602,
"step": 950
},
{
"epoch": 3.8567088607594937,
"grad_norm": 0.4330704473253033,
"learning_rate": 5.948454566959363e-06,
"loss": 0.1695,
"step": 951
},
{
"epoch": 3.8607594936708862,
"grad_norm": 0.4265474930123265,
"learning_rate": 5.908121309322328e-06,
"loss": 0.1576,
"step": 952
},
{
"epoch": 3.8648101265822783,
"grad_norm": 0.4480696700630478,
"learning_rate": 5.867901545738976e-06,
"loss": 0.1585,
"step": 953
},
{
"epoch": 3.868860759493671,
"grad_norm": 0.4310354785817411,
"learning_rate": 5.827795600133774e-06,
"loss": 0.1989,
"step": 954
},
{
"epoch": 3.8729113924050633,
"grad_norm": 0.4190455744108073,
"learning_rate": 5.787803795514466e-06,
"loss": 0.1607,
"step": 955
},
{
"epoch": 3.876962025316456,
"grad_norm": 0.4393420675361396,
"learning_rate": 5.747926453969576e-06,
"loss": 0.1719,
"step": 956
},
{
"epoch": 3.8810126582278484,
"grad_norm": 0.4117239625162937,
"learning_rate": 5.708163896665708e-06,
"loss": 0.1791,
"step": 957
},
{
"epoch": 3.8850632911392404,
"grad_norm": 0.41544768596970383,
"learning_rate": 5.668516443845047e-06,
"loss": 0.1737,
"step": 958
},
{
"epoch": 3.889113924050633,
"grad_norm": 0.42041999315520656,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.1842,
"step": 959
},
{
"epoch": 3.8931645569620255,
"grad_norm": 0.42119062773207405,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.1633,
"step": 960
},
{
"epoch": 3.8972151898734175,
"grad_norm": 0.4200661365918226,
"learning_rate": 5.550267900783019e-06,
"loss": 0.1543,
"step": 961
},
{
"epoch": 3.90126582278481,
"grad_norm": 0.44672969210598545,
"learning_rate": 5.511084049737623e-06,
"loss": 0.1567,
"step": 962
},
{
"epoch": 3.9053164556962026,
"grad_norm": 0.4220585265905678,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.1463,
"step": 963
},
{
"epoch": 3.909367088607595,
"grad_norm": 0.419147612538862,
"learning_rate": 5.433066737499948e-06,
"loss": 0.1608,
"step": 964
},
{
"epoch": 3.9134177215189876,
"grad_norm": 0.42645691469575625,
"learning_rate": 5.394233904648376e-06,
"loss": 0.1398,
"step": 965
},
{
"epoch": 3.9174683544303797,
"grad_norm": 0.42621049241628417,
"learning_rate": 5.355518704628997e-06,
"loss": 0.1843,
"step": 966
},
{
"epoch": 3.921518987341772,
"grad_norm": 0.41937861499477913,
"learning_rate": 5.316921449248731e-06,
"loss": 0.1637,
"step": 967
},
{
"epoch": 3.9255696202531647,
"grad_norm": 0.4338643033554104,
"learning_rate": 5.278442449364538e-06,
"loss": 0.1532,
"step": 968
},
{
"epoch": 3.9296202531645568,
"grad_norm": 0.42513154520265195,
"learning_rate": 5.240082014881016e-06,
"loss": 0.1651,
"step": 969
},
{
"epoch": 3.9336708860759493,
"grad_norm": 0.4201804389106929,
"learning_rate": 5.201840454747822e-06,
"loss": 0.1706,
"step": 970
},
{
"epoch": 3.937721518987342,
"grad_norm": 0.41006242135703364,
"learning_rate": 5.163718076957223e-06,
"loss": 0.1769,
"step": 971
},
{
"epoch": 3.9417721518987343,
"grad_norm": 0.4264452813356826,
"learning_rate": 5.125715188541609e-06,
"loss": 0.1499,
"step": 972
},
{
"epoch": 3.945822784810127,
"grad_norm": 0.4460921094882639,
"learning_rate": 5.087832095571021e-06,
"loss": 0.1398,
"step": 973
},
{
"epoch": 3.949873417721519,
"grad_norm": 0.42665720870962215,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.1615,
"step": 974
},
{
"epoch": 3.9539240506329114,
"grad_norm": 0.4206376488090742,
"learning_rate": 5.01242651541854e-06,
"loss": 0.1579,
"step": 975
},
{
"epoch": 3.957974683544304,
"grad_norm": 0.43507254777936377,
"learning_rate": 4.974904635542815e-06,
"loss": 0.1758,
"step": 976
},
{
"epoch": 3.962025316455696,
"grad_norm": 0.41551635068296183,
"learning_rate": 4.937503765719582e-06,
"loss": 0.16,
"step": 977
},
{
"epoch": 3.9660759493670885,
"grad_norm": 0.43418625049624593,
"learning_rate": 4.900224207170299e-06,
"loss": 0.1576,
"step": 978
},
{
"epoch": 3.970126582278481,
"grad_norm": 0.43733438253289236,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.1795,
"step": 979
},
{
"epoch": 3.9741772151898735,
"grad_norm": 0.4228019041772109,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.1689,
"step": 980
},
{
"epoch": 3.978227848101266,
"grad_norm": 0.43195303074678204,
"learning_rate": 4.789116396710924e-06,
"loss": 0.1617,
"step": 981
},
{
"epoch": 3.982278481012658,
"grad_norm": 0.40243214393401755,
"learning_rate": 4.752325075895368e-06,
"loss": 0.1793,
"step": 982
},
{
"epoch": 3.9863291139240506,
"grad_norm": 0.4243286077711075,
"learning_rate": 4.715656557757473e-06,
"loss": 0.1756,
"step": 983
},
{
"epoch": 3.990379746835443,
"grad_norm": 0.4314560329715484,
"learning_rate": 4.679111137620442e-06,
"loss": 0.1778,
"step": 984
},
{
"epoch": 3.9944303797468352,
"grad_norm": 0.4296028807156098,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.1469,
"step": 985
},
{
"epoch": 3.9984810126582278,
"grad_norm": 0.41330728438849157,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.1746,
"step": 986
},
{
"epoch": 4.004050632911392,
"grad_norm": 0.4440563253385535,
"learning_rate": 4.570216403561141e-06,
"loss": 0.1017,
"step": 987
},
{
"epoch": 4.008101265822785,
"grad_norm": 0.45177772735511174,
"learning_rate": 4.534166308795815e-06,
"loss": 0.1079,
"step": 988
},
{
"epoch": 4.012151898734177,
"grad_norm": 0.3825127916593969,
"learning_rate": 4.498240773728859e-06,
"loss": 0.0938,
"step": 989
},
{
"epoch": 4.01620253164557,
"grad_norm": 0.32000971929905686,
"learning_rate": 4.462440087699609e-06,
"loss": 0.095,
"step": 990
},
{
"epoch": 4.020253164556962,
"grad_norm": 0.30559527558307126,
"learning_rate": 4.426764539041861e-06,
"loss": 0.0971,
"step": 991
},
{
"epoch": 4.024303797468354,
"grad_norm": 0.3542957336264319,
"learning_rate": 4.391214415081582e-06,
"loss": 0.1225,
"step": 992
},
{
"epoch": 4.028354430379747,
"grad_norm": 0.3723754664066784,
"learning_rate": 4.355790002134579e-06,
"loss": 0.1056,
"step": 993
},
{
"epoch": 4.032405063291139,
"grad_norm": 0.39485736256955395,
"learning_rate": 4.320491585504207e-06,
"loss": 0.1062,
"step": 994
},
{
"epoch": 4.036455696202531,
"grad_norm": 0.36219826610572065,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.0808,
"step": 995
},
{
"epoch": 4.040506329113924,
"grad_norm": 0.389205280834126,
"learning_rate": 4.250273877330691e-06,
"loss": 0.099,
"step": 996
},
{
"epoch": 4.044556962025316,
"grad_norm": 0.3976269694591632,
"learning_rate": 4.215355151311313e-06,
"loss": 0.0934,
"step": 997
},
{
"epoch": 4.048607594936709,
"grad_norm": 0.36763698916350807,
"learning_rate": 4.180563552651542e-06,
"loss": 0.0902,
"step": 998
},
{
"epoch": 4.052658227848101,
"grad_norm": 0.3304920222968978,
"learning_rate": 4.145899361558147e-06,
"loss": 0.0912,
"step": 999
},
{
"epoch": 4.056708860759493,
"grad_norm": 0.32154578612526913,
"learning_rate": 4.111362857211738e-06,
"loss": 0.1063,
"step": 1000
},
{
"epoch": 4.060759493670886,
"grad_norm": 0.34878945741226974,
"learning_rate": 4.076954317764592e-06,
"loss": 0.1203,
"step": 1001
},
{
"epoch": 4.0648101265822785,
"grad_norm": 0.3399144010157978,
"learning_rate": 4.042674020338335e-06,
"loss": 0.0895,
"step": 1002
},
{
"epoch": 4.0688607594936705,
"grad_norm": 0.31403003180006117,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.0918,
"step": 1003
},
{
"epoch": 4.0729113924050635,
"grad_norm": 0.33255437818870065,
"learning_rate": 3.974499254868674e-06,
"loss": 0.1024,
"step": 1004
},
{
"epoch": 4.076962025316456,
"grad_norm": 0.31389998505327293,
"learning_rate": 3.940605335895451e-06,
"loss": 0.1049,
"step": 1005
},
{
"epoch": 4.0810126582278485,
"grad_norm": 0.31069366324968545,
"learning_rate": 3.90684075707908e-06,
"loss": 0.0909,
"step": 1006
},
{
"epoch": 4.085063291139241,
"grad_norm": 0.30697556396786724,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.0914,
"step": 1007
},
{
"epoch": 4.089113924050633,
"grad_norm": 0.29676939987422246,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.0821,
"step": 1008
},
{
"epoch": 4.093164556962026,
"grad_norm": 0.3081864765753793,
"learning_rate": 3.806325775702304e-06,
"loss": 0.089,
"step": 1009
},
{
"epoch": 4.097215189873418,
"grad_norm": 0.2978807282268539,
"learning_rate": 3.773081266416434e-06,
"loss": 0.0727,
"step": 1010
},
{
"epoch": 4.10126582278481,
"grad_norm": 0.3123403295200488,
"learning_rate": 3.739967446503245e-06,
"loss": 0.0912,
"step": 1011
},
{
"epoch": 4.105316455696203,
"grad_norm": 0.31945998721623287,
"learning_rate": 3.706984582656894e-06,
"loss": 0.0947,
"step": 1012
},
{
"epoch": 4.109367088607595,
"grad_norm": 0.31033582894685796,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.0863,
"step": 1013
},
{
"epoch": 4.113417721518988,
"grad_norm": 0.329378670210074,
"learning_rate": 3.641412784665648e-06,
"loss": 0.1127,
"step": 1014
},
{
"epoch": 4.11746835443038,
"grad_norm": 0.33691441007204453,
"learning_rate": 3.608824378627005e-06,
"loss": 0.1062,
"step": 1015
},
{
"epoch": 4.121518987341772,
"grad_norm": 0.3169919373696831,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.0824,
"step": 1016
},
{
"epoch": 4.125569620253165,
"grad_norm": 0.30709059497144114,
"learning_rate": 3.544043864774269e-06,
"loss": 0.0853,
"step": 1017
},
{
"epoch": 4.129620253164557,
"grad_norm": 0.31515073934073856,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.1078,
"step": 1018
},
{
"epoch": 4.133670886075949,
"grad_norm": 0.3287696342138486,
"learning_rate": 3.479793485887819e-06,
"loss": 0.1008,
"step": 1019
},
{
"epoch": 4.137721518987342,
"grad_norm": 0.3024455770514863,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.101,
"step": 1020
},
{
"epoch": 4.141772151898734,
"grad_norm": 0.3090514432864323,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.0946,
"step": 1021
},
{
"epoch": 4.145822784810127,
"grad_norm": 0.31856850391904995,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.1025,
"step": 1022
},
{
"epoch": 4.149873417721519,
"grad_norm": 0.28868573165297756,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.0799,
"step": 1023
},
{
"epoch": 4.153924050632911,
"grad_norm": 0.3071466658453563,
"learning_rate": 3.321500420335e-06,
"loss": 0.0815,
"step": 1024
},
{
"epoch": 4.157974683544304,
"grad_norm": 0.30401976540851194,
"learning_rate": 3.290243771741275e-06,
"loss": 0.1,
"step": 1025
},
{
"epoch": 4.162025316455696,
"grad_norm": 0.3153422630236962,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.1096,
"step": 1026
},
{
"epoch": 4.166075949367088,
"grad_norm": 0.291002194572653,
"learning_rate": 3.228134459445149e-06,
"loss": 0.0765,
"step": 1027
},
{
"epoch": 4.170126582278481,
"grad_norm": 0.3074381746143392,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.0861,
"step": 1028
},
{
"epoch": 4.174177215189873,
"grad_norm": 0.3009824107319316,
"learning_rate": 3.166565459258513e-06,
"loss": 0.0996,
"step": 1029
},
{
"epoch": 4.178227848101266,
"grad_norm": 0.3053844845196557,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.0954,
"step": 1030
},
{
"epoch": 4.182278481012658,
"grad_norm": 0.31862898532209416,
"learning_rate": 3.105538754650419e-06,
"loss": 0.0979,
"step": 1031
},
{
"epoch": 4.18632911392405,
"grad_norm": 0.3117445782707549,
"learning_rate": 3.07522937824712e-06,
"loss": 0.0954,
"step": 1032
},
{
"epoch": 4.190379746835443,
"grad_norm": 0.34365957915043155,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.1008,
"step": 1033
},
{
"epoch": 4.194430379746835,
"grad_norm": 0.31217280539301406,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.098,
"step": 1034
},
{
"epoch": 4.1984810126582275,
"grad_norm": 0.3156205541211237,
"learning_rate": 2.985120078631465e-06,
"loss": 0.0944,
"step": 1035
},
{
"epoch": 4.2025316455696204,
"grad_norm": 0.29095122624987413,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.0841,
"step": 1036
},
{
"epoch": 4.2065822784810125,
"grad_norm": 0.28614965250640506,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.0875,
"step": 1037
},
{
"epoch": 4.2106329113924055,
"grad_norm": 0.30067497171497953,
"learning_rate": 2.896244091927678e-06,
"loss": 0.0832,
"step": 1038
},
{
"epoch": 4.2146835443037975,
"grad_norm": 0.29719698042409526,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.1037,
"step": 1039
},
{
"epoch": 4.21873417721519,
"grad_norm": 0.29804111462684324,
"learning_rate": 2.837681792951994e-06,
"loss": 0.0898,
"step": 1040
},
{
"epoch": 4.222784810126583,
"grad_norm": 0.3078453119860937,
"learning_rate": 2.808607860255981e-06,
"loss": 0.0971,
"step": 1041
},
{
"epoch": 4.226835443037975,
"grad_norm": 0.30010751839135674,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.0929,
"step": 1042
},
{
"epoch": 4.230886075949367,
"grad_norm": 0.2918260032795571,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.0903,
"step": 1043
},
{
"epoch": 4.23493670886076,
"grad_norm": 0.30303866682239083,
"learning_rate": 2.722217735873718e-06,
"loss": 0.1037,
"step": 1044
},
{
"epoch": 4.238987341772152,
"grad_norm": 0.3058359346217749,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.0899,
"step": 1045
},
{
"epoch": 4.243037974683545,
"grad_norm": 0.29567958198236083,
"learning_rate": 2.665319697490205e-06,
"loss": 0.1042,
"step": 1046
},
{
"epoch": 4.247088607594937,
"grad_norm": 0.3101067098144972,
"learning_rate": 2.637079980714945e-06,
"loss": 0.0969,
"step": 1047
},
{
"epoch": 4.251139240506329,
"grad_norm": 0.30494416584082606,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.0902,
"step": 1048
},
{
"epoch": 4.255189873417722,
"grad_norm": 0.318918318750103,
"learning_rate": 2.581020289189571e-06,
"loss": 0.1068,
"step": 1049
},
{
"epoch": 4.259240506329114,
"grad_norm": 0.3003154577781419,
"learning_rate": 2.553200765936501e-06,
"loss": 0.092,
"step": 1050
},
{
"epoch": 4.263291139240506,
"grad_norm": 0.3019986178171186,
"learning_rate": 2.525521756805962e-06,
"loss": 0.0839,
"step": 1051
},
{
"epoch": 4.267341772151899,
"grad_norm": 0.30465662090237866,
"learning_rate": 2.497983484720885e-06,
"loss": 0.0984,
"step": 1052
},
{
"epoch": 4.271392405063291,
"grad_norm": 0.3087208324044594,
"learning_rate": 2.470586171470728e-06,
"loss": 0.0869,
"step": 1053
},
{
"epoch": 4.275443037974684,
"grad_norm": 0.29298656869915435,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.0968,
"step": 1054
},
{
"epoch": 4.279493670886076,
"grad_norm": 0.2983774699259279,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.1165,
"step": 1055
},
{
"epoch": 4.283544303797468,
"grad_norm": 0.2831723249923194,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.0673,
"step": 1056
},
{
"epoch": 4.287594936708861,
"grad_norm": 0.2977770560461507,
"learning_rate": 2.362410902836978e-06,
"loss": 0.0964,
"step": 1057
},
{
"epoch": 4.291645569620253,
"grad_norm": 0.2958524883643253,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.0879,
"step": 1058
},
{
"epoch": 4.295696202531645,
"grad_norm": 0.2909220853656131,
"learning_rate": 2.309174704446411e-06,
"loss": 0.0976,
"step": 1059
},
{
"epoch": 4.299746835443038,
"grad_norm": 0.305680917605246,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.0789,
"step": 1060
},
{
"epoch": 4.30379746835443,
"grad_norm": 0.30612684812370766,
"learning_rate": 2.256508422807855e-06,
"loss": 0.0824,
"step": 1061
},
{
"epoch": 4.307848101265823,
"grad_norm": 0.3040616530614763,
"learning_rate": 2.230389531696946e-06,
"loss": 0.0884,
"step": 1062
},
{
"epoch": 4.311898734177215,
"grad_norm": 0.30344168649310316,
"learning_rate": 2.204413754585857e-06,
"loss": 0.0961,
"step": 1063
},
{
"epoch": 4.315949367088607,
"grad_norm": 0.2934371059992982,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.0973,
"step": 1064
},
{
"epoch": 4.32,
"grad_norm": 0.314452855391954,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.101,
"step": 1065
},
{
"epoch": 4.324050632911392,
"grad_norm": 0.301006889722246,
"learning_rate": 2.127347193531757e-06,
"loss": 0.0902,
"step": 1066
},
{
"epoch": 4.328101265822784,
"grad_norm": 0.2815342362145772,
"learning_rate": 2.101945952921942e-06,
"loss": 0.0847,
"step": 1067
},
{
"epoch": 4.332151898734177,
"grad_norm": 0.31702000799855784,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.115,
"step": 1068
},
{
"epoch": 4.3362025316455695,
"grad_norm": 0.27375117324125053,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.0811,
"step": 1069
},
{
"epoch": 4.340253164556962,
"grad_norm": 0.31479293858676843,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.0875,
"step": 1070
},
{
"epoch": 4.3443037974683545,
"grad_norm": 0.2769524395140214,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.0838,
"step": 1071
},
{
"epoch": 4.348354430379747,
"grad_norm": 0.297374513872857,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.0806,
"step": 1072
},
{
"epoch": 4.3524050632911395,
"grad_norm": 0.29590341799788245,
"learning_rate": 1.95257270594031e-06,
"loss": 0.09,
"step": 1073
},
{
"epoch": 4.356455696202532,
"grad_norm": 0.3032775291519343,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.0772,
"step": 1074
},
{
"epoch": 4.360506329113924,
"grad_norm": 0.31885552987691135,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.0969,
"step": 1075
},
{
"epoch": 4.364556962025317,
"grad_norm": 0.33246765479990104,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.1066,
"step": 1076
},
{
"epoch": 4.368607594936709,
"grad_norm": 0.30236679765907876,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.0985,
"step": 1077
},
{
"epoch": 4.372658227848102,
"grad_norm": 0.29396454152386714,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.075,
"step": 1078
},
{
"epoch": 4.376708860759494,
"grad_norm": 0.3121606797325783,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.0971,
"step": 1079
},
{
"epoch": 4.380759493670886,
"grad_norm": 0.3094725308387631,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.1111,
"step": 1080
},
{
"epoch": 4.384810126582279,
"grad_norm": 0.3116170320059298,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.0827,
"step": 1081
},
{
"epoch": 4.388860759493671,
"grad_norm": 0.3029599096585966,
"learning_rate": 1.738336950434061e-06,
"loss": 0.0997,
"step": 1082
},
{
"epoch": 4.392911392405063,
"grad_norm": 0.3210592947055813,
"learning_rate": 1.715265841375957e-06,
"loss": 0.1108,
"step": 1083
},
{
"epoch": 4.396962025316456,
"grad_norm": 0.3107450689451434,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.104,
"step": 1084
},
{
"epoch": 4.401012658227848,
"grad_norm": 0.2916076483466408,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.0787,
"step": 1085
},
{
"epoch": 4.405063291139241,
"grad_norm": 0.3006314712648664,
"learning_rate": 1.646936827927441e-06,
"loss": 0.0991,
"step": 1086
},
{
"epoch": 4.409113924050633,
"grad_norm": 0.2971940216171281,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.0833,
"step": 1087
},
{
"epoch": 4.413164556962025,
"grad_norm": 0.28610372551913565,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.1085,
"step": 1088
},
{
"epoch": 4.417215189873418,
"grad_norm": 0.2913365891127378,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.0969,
"step": 1089
},
{
"epoch": 4.42126582278481,
"grad_norm": 0.3125441322722971,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.0968,
"step": 1090
},
{
"epoch": 4.425316455696202,
"grad_norm": 0.3122092576036691,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.1036,
"step": 1091
},
{
"epoch": 4.429367088607595,
"grad_norm": 0.307313358805454,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.1019,
"step": 1092
},
{
"epoch": 4.433417721518987,
"grad_norm": 0.30415657602587226,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.0826,
"step": 1093
},
{
"epoch": 4.43746835443038,
"grad_norm": 0.3071303974312258,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.1276,
"step": 1094
},
{
"epoch": 4.441518987341772,
"grad_norm": 0.2979407994438442,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.0908,
"step": 1095
},
{
"epoch": 4.445569620253164,
"grad_norm": 0.2984477897644635,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.0986,
"step": 1096
},
{
"epoch": 4.449620253164557,
"grad_norm": 0.3107161330405733,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.0958,
"step": 1097
},
{
"epoch": 4.453670886075949,
"grad_norm": 0.2980002507618588,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.1068,
"step": 1098
},
{
"epoch": 4.457721518987341,
"grad_norm": 0.31506086087537066,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.0929,
"step": 1099
},
{
"epoch": 4.461772151898734,
"grad_norm": 0.2943797105027466,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.0953,
"step": 1100
},
{
"epoch": 4.465822784810126,
"grad_norm": 0.28686491855379703,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.0879,
"step": 1101
},
{
"epoch": 4.469873417721519,
"grad_norm": 0.3013246380284793,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.0893,
"step": 1102
},
{
"epoch": 4.473924050632911,
"grad_norm": 0.28981940414040963,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.0892,
"step": 1103
},
{
"epoch": 4.4779746835443035,
"grad_norm": 0.29084626044459816,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.0877,
"step": 1104
},
{
"epoch": 4.4820253164556965,
"grad_norm": 0.2912966893202847,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.0837,
"step": 1105
},
{
"epoch": 4.4860759493670885,
"grad_norm": 0.31468821869379116,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.1115,
"step": 1106
},
{
"epoch": 4.490126582278481,
"grad_norm": 0.28698810414757425,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.0815,
"step": 1107
},
{
"epoch": 4.494177215189874,
"grad_norm": 0.29914549858397066,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.0875,
"step": 1108
},
{
"epoch": 4.498227848101266,
"grad_norm": 0.2944132904866547,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.1041,
"step": 1109
},
{
"epoch": 4.502278481012659,
"grad_norm": 0.2974561460242331,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.086,
"step": 1110
},
{
"epoch": 4.506329113924051,
"grad_norm": 0.28494690700995845,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.0978,
"step": 1111
},
{
"epoch": 4.510379746835443,
"grad_norm": 0.2982966168925565,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.0921,
"step": 1112
},
{
"epoch": 4.514430379746836,
"grad_norm": 0.3015747075450197,
"learning_rate": 1.092402023876933e-06,
"loss": 0.1062,
"step": 1113
},
{
"epoch": 4.518481012658228,
"grad_norm": 0.3171353072158858,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.1022,
"step": 1114
},
{
"epoch": 4.52253164556962,
"grad_norm": 0.29343366837001,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.1006,
"step": 1115
},
{
"epoch": 4.526582278481013,
"grad_norm": 0.2892876268919545,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.092,
"step": 1116
},
{
"epoch": 4.530632911392405,
"grad_norm": 0.29163024802476983,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.1129,
"step": 1117
},
{
"epoch": 4.534683544303798,
"grad_norm": 0.3057052646916121,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.1111,
"step": 1118
},
{
"epoch": 4.53873417721519,
"grad_norm": 0.288408984061695,
"learning_rate": 9.841383857091947e-07,
"loss": 0.0829,
"step": 1119
},
{
"epoch": 4.542784810126582,
"grad_norm": 0.28690986579997435,
"learning_rate": 9.666296380321616e-07,
"loss": 0.0828,
"step": 1120
},
{
"epoch": 4.546835443037975,
"grad_norm": 0.2939245326182489,
"learning_rate": 9.492741825034124e-07,
"loss": 0.0942,
"step": 1121
},
{
"epoch": 4.550886075949367,
"grad_norm": 0.29560847468880636,
"learning_rate": 9.320721589013892e-07,
"loss": 0.0911,
"step": 1122
},
{
"epoch": 4.55493670886076,
"grad_norm": 0.3033672509451601,
"learning_rate": 9.150237057688339e-07,
"loss": 0.1019,
"step": 1123
},
{
"epoch": 4.558987341772152,
"grad_norm": 0.29624953360602213,
"learning_rate": 8.981289604116328e-07,
"loss": 0.0778,
"step": 1124
},
{
"epoch": 4.563037974683544,
"grad_norm": 0.2968275888020755,
"learning_rate": 8.813880588977542e-07,
"loss": 0.1029,
"step": 1125
},
{
"epoch": 4.567088607594937,
"grad_norm": 0.3016354476119144,
"learning_rate": 8.648011360561126e-07,
"loss": 0.096,
"step": 1126
},
{
"epoch": 4.571139240506329,
"grad_norm": 0.29636099786046827,
"learning_rate": 8.483683254755037e-07,
"loss": 0.1023,
"step": 1127
},
{
"epoch": 4.575189873417721,
"grad_norm": 0.29612622649551296,
"learning_rate": 8.320897595035227e-07,
"loss": 0.0831,
"step": 1128
},
{
"epoch": 4.579240506329114,
"grad_norm": 0.29508044222814916,
"learning_rate": 8.159655692455093e-07,
"loss": 0.0831,
"step": 1129
},
{
"epoch": 4.583291139240506,
"grad_norm": 0.32094546195441465,
"learning_rate": 7.999958845634648e-07,
"loss": 0.1045,
"step": 1130
},
{
"epoch": 4.587341772151898,
"grad_norm": 0.30939597533219465,
"learning_rate": 7.841808340750478e-07,
"loss": 0.0911,
"step": 1131
},
{
"epoch": 4.591392405063291,
"grad_norm": 0.3075398125736892,
"learning_rate": 7.685205451524869e-07,
"loss": 0.0937,
"step": 1132
},
{
"epoch": 4.595443037974683,
"grad_norm": 0.2860963816241099,
"learning_rate": 7.530151439216027e-07,
"loss": 0.0877,
"step": 1133
},
{
"epoch": 4.599493670886076,
"grad_norm": 0.30058622509578664,
"learning_rate": 7.376647552607675e-07,
"loss": 0.1043,
"step": 1134
},
{
"epoch": 4.603544303797468,
"grad_norm": 0.31406474683455204,
"learning_rate": 7.224695027998963e-07,
"loss": 0.0844,
"step": 1135
},
{
"epoch": 4.6075949367088604,
"grad_norm": 0.2996123289528497,
"learning_rate": 7.07429508919466e-07,
"loss": 0.0898,
"step": 1136
},
{
"epoch": 4.611645569620253,
"grad_norm": 0.3208455092106972,
"learning_rate": 6.925448947495206e-07,
"loss": 0.1159,
"step": 1137
},
{
"epoch": 4.6156962025316455,
"grad_norm": 0.2950033262476518,
"learning_rate": 6.778157801686936e-07,
"loss": 0.0913,
"step": 1138
},
{
"epoch": 4.619746835443038,
"grad_norm": 0.2877751018506998,
"learning_rate": 6.632422838032515e-07,
"loss": 0.1031,
"step": 1139
},
{
"epoch": 4.6237974683544305,
"grad_norm": 0.2940597324310048,
"learning_rate": 6.488245230261281e-07,
"loss": 0.0844,
"step": 1140
},
{
"epoch": 4.627848101265823,
"grad_norm": 0.2873610663099083,
"learning_rate": 6.345626139559868e-07,
"loss": 0.0853,
"step": 1141
},
{
"epoch": 4.6318987341772155,
"grad_norm": 0.30213054423055963,
"learning_rate": 6.204566714562866e-07,
"loss": 0.0917,
"step": 1142
},
{
"epoch": 4.635949367088608,
"grad_norm": 0.2788872366995321,
"learning_rate": 6.06506809134344e-07,
"loss": 0.085,
"step": 1143
},
{
"epoch": 4.64,
"grad_norm": 0.3050743111425735,
"learning_rate": 5.927131393404373e-07,
"loss": 0.0836,
"step": 1144
},
{
"epoch": 4.644050632911393,
"grad_norm": 0.30571740655872903,
"learning_rate": 5.790757731668817e-07,
"loss": 0.0948,
"step": 1145
},
{
"epoch": 4.648101265822785,
"grad_norm": 0.2926632476669542,
"learning_rate": 5.655948204471507e-07,
"loss": 0.085,
"step": 1146
},
{
"epoch": 4.652151898734177,
"grad_norm": 0.30446215144827077,
"learning_rate": 5.522703897549875e-07,
"loss": 0.0988,
"step": 1147
},
{
"epoch": 4.65620253164557,
"grad_norm": 0.30972649884046305,
"learning_rate": 5.391025884035239e-07,
"loss": 0.1215,
"step": 1148
},
{
"epoch": 4.660253164556962,
"grad_norm": 0.2945928643416498,
"learning_rate": 5.260915224444207e-07,
"loss": 0.0987,
"step": 1149
},
{
"epoch": 4.664303797468355,
"grad_norm": 0.28730738524927424,
"learning_rate": 5.132372966670129e-07,
"loss": 0.0786,
"step": 1150
},
{
"epoch": 4.668354430379747,
"grad_norm": 0.2960024786173416,
"learning_rate": 5.005400145974704e-07,
"loss": 0.0992,
"step": 1151
},
{
"epoch": 4.672405063291139,
"grad_norm": 0.3027810636156665,
"learning_rate": 4.879997784979562e-07,
"loss": 0.1006,
"step": 1152
},
{
"epoch": 4.676455696202532,
"grad_norm": 0.2950173711164706,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.0775,
"step": 1153
},
{
"epoch": 4.680506329113924,
"grad_norm": 0.2979879996808229,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.096,
"step": 1154
},
{
"epoch": 4.684556962025317,
"grad_norm": 0.2979060654024272,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.1158,
"step": 1155
},
{
"epoch": 4.688607594936709,
"grad_norm": 0.29921048170731324,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.0904,
"step": 1156
},
{
"epoch": 4.692658227848101,
"grad_norm": 0.29255177593348813,
"learning_rate": 4.27657778143431e-07,
"loss": 0.0902,
"step": 1157
},
{
"epoch": 4.696708860759494,
"grad_norm": 0.3083807168112228,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.1002,
"step": 1158
},
{
"epoch": 4.700759493670886,
"grad_norm": 0.3024756003537126,
"learning_rate": 4.046237371675177e-07,
"loss": 0.0926,
"step": 1159
},
{
"epoch": 4.704810126582278,
"grad_norm": 0.3116145569122567,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.1028,
"step": 1160
},
{
"epoch": 4.708860759493671,
"grad_norm": 0.3105853082352048,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.0874,
"step": 1161
},
{
"epoch": 4.712911392405063,
"grad_norm": 0.30335948870782303,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.1052,
"step": 1162
},
{
"epoch": 4.716962025316455,
"grad_norm": 0.3011937024508463,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.0887,
"step": 1163
},
{
"epoch": 4.721012658227848,
"grad_norm": 0.29506846413770343,
"learning_rate": 3.498020255137813e-07,
"loss": 0.1012,
"step": 1164
},
{
"epoch": 4.72506329113924,
"grad_norm": 0.2905087549966518,
"learning_rate": 3.393121163057811e-07,
"loss": 0.0806,
"step": 1165
},
{
"epoch": 4.729113924050633,
"grad_norm": 0.2885840488124097,
"learning_rate": 3.289805515751399e-07,
"loss": 0.0807,
"step": 1166
},
{
"epoch": 4.733164556962025,
"grad_norm": 0.30498582688294323,
"learning_rate": 3.188074145308573e-07,
"loss": 0.0856,
"step": 1167
},
{
"epoch": 4.737215189873417,
"grad_norm": 0.2960930657339419,
"learning_rate": 3.087927871059804e-07,
"loss": 0.1035,
"step": 1168
},
{
"epoch": 4.74126582278481,
"grad_norm": 0.3182151065576106,
"learning_rate": 2.989367499569418e-07,
"loss": 0.1126,
"step": 1169
},
{
"epoch": 4.745316455696202,
"grad_norm": 0.3117426239095743,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.0936,
"step": 1170
},
{
"epoch": 4.749367088607595,
"grad_norm": 0.279503445984087,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.0901,
"step": 1171
},
{
"epoch": 4.7534177215189874,
"grad_norm": 0.30910389618235407,
"learning_rate": 2.703209675663887e-07,
"loss": 0.0967,
"step": 1172
},
{
"epoch": 4.7574683544303795,
"grad_norm": 0.2857685637546387,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.1059,
"step": 1173
},
{
"epoch": 4.7615189873417725,
"grad_norm": 0.2900048265751233,
"learning_rate": 2.520381518804471e-07,
"loss": 0.081,
"step": 1174
},
{
"epoch": 4.7655696202531646,
"grad_norm": 0.28693514608782067,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.0908,
"step": 1175
},
{
"epoch": 4.769620253164557,
"grad_norm": 0.3125268292818404,
"learning_rate": 2.343915243930317e-07,
"loss": 0.0932,
"step": 1176
},
{
"epoch": 4.77367088607595,
"grad_norm": 0.2916945356351481,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.0845,
"step": 1177
},
{
"epoch": 4.777721518987342,
"grad_norm": 0.30939771326341237,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.1,
"step": 1178
},
{
"epoch": 4.781772151898734,
"grad_norm": 0.3064125902956753,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.0932,
"step": 1179
},
{
"epoch": 4.785822784810127,
"grad_norm": 0.29659060901993783,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.1154,
"step": 1180
},
{
"epoch": 4.789873417721519,
"grad_norm": 0.3028751670994611,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.0997,
"step": 1181
},
{
"epoch": 4.793924050632912,
"grad_norm": 0.3052572336501952,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.0886,
"step": 1182
},
{
"epoch": 4.797974683544304,
"grad_norm": 0.29659111006945316,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.1013,
"step": 1183
},
{
"epoch": 4.802025316455696,
"grad_norm": 0.31209403993035195,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.0831,
"step": 1184
},
{
"epoch": 4.806075949367089,
"grad_norm": 0.292874502988922,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.1066,
"step": 1185
},
{
"epoch": 4.810126582278481,
"grad_norm": 0.2994921464411499,
"learning_rate": 1.557203888516745e-07,
"loss": 0.104,
"step": 1186
},
{
"epoch": 4.814177215189874,
"grad_norm": 0.2988422955462081,
"learning_rate": 1.487313098987131e-07,
"loss": 0.0899,
"step": 1187
},
{
"epoch": 4.818227848101266,
"grad_norm": 0.3066398719691197,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.0848,
"step": 1188
},
{
"epoch": 4.822278481012658,
"grad_norm": 0.3150544053883717,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.1034,
"step": 1189
},
{
"epoch": 4.826329113924051,
"grad_norm": 0.2913086367032518,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.0938,
"step": 1190
},
{
"epoch": 4.830379746835443,
"grad_norm": 0.28966099928948347,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.0787,
"step": 1191
},
{
"epoch": 4.834430379746835,
"grad_norm": 0.34044958017067345,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.0836,
"step": 1192
},
{
"epoch": 4.838481012658228,
"grad_norm": 0.30702740257105593,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.0902,
"step": 1193
},
{
"epoch": 4.84253164556962,
"grad_norm": 0.2983477519052036,
"learning_rate": 1.042873700093061e-07,
"loss": 0.0938,
"step": 1194
},
{
"epoch": 4.846582278481012,
"grad_norm": 0.29346655323688653,
"learning_rate": 9.857879404556291e-08,
"loss": 0.0912,
"step": 1195
},
{
"epoch": 4.850632911392405,
"grad_norm": 0.29317079451639033,
"learning_rate": 9.303050139225722e-08,
"loss": 0.0983,
"step": 1196
},
{
"epoch": 4.854683544303797,
"grad_norm": 0.2971524390539934,
"learning_rate": 8.76425367345779e-08,
"loss": 0.0916,
"step": 1197
},
{
"epoch": 4.85873417721519,
"grad_norm": 0.28103166582887923,
"learning_rate": 8.241494346644897e-08,
"loss": 0.093,
"step": 1198
},
{
"epoch": 4.862784810126582,
"grad_norm": 0.29456672107520904,
"learning_rate": 7.734776369019204e-08,
"loss": 0.0784,
"step": 1199
},
{
"epoch": 4.866835443037974,
"grad_norm": 0.2828991563365941,
"learning_rate": 7.244103821617332e-08,
"loss": 0.0957,
"step": 1200
},
{
"epoch": 4.870886075949367,
"grad_norm": 0.29599806353324526,
"learning_rate": 6.769480656248606e-08,
"loss": 0.0908,
"step": 1201
},
{
"epoch": 4.874936708860759,
"grad_norm": 0.2884110254970776,
"learning_rate": 6.310910695462635e-08,
"loss": 0.0775,
"step": 1202
},
{
"epoch": 4.878987341772152,
"grad_norm": 0.3219334882919305,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.0857,
"step": 1203
},
{
"epoch": 4.883037974683544,
"grad_norm": 0.30941151223641655,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.0985,
"step": 1204
},
{
"epoch": 4.8870886075949365,
"grad_norm": 0.28688359345552655,
"learning_rate": 5.031556326567488e-08,
"loss": 0.0853,
"step": 1205
},
{
"epoch": 4.891139240506329,
"grad_norm": 0.3110976024311402,
"learning_rate": 4.637234823364312e-08,
"loss": 0.0964,
"step": 1206
},
{
"epoch": 4.8951898734177215,
"grad_norm": 0.30199858778609917,
"learning_rate": 4.258983697558838e-08,
"loss": 0.0905,
"step": 1207
},
{
"epoch": 4.899240506329114,
"grad_norm": 0.2882898158869149,
"learning_rate": 3.896805995533548e-08,
"loss": 0.09,
"step": 1208
},
{
"epoch": 4.9032911392405065,
"grad_norm": 0.30358757179604157,
"learning_rate": 3.550704634218028e-08,
"loss": 0.084,
"step": 1209
},
{
"epoch": 4.907341772151899,
"grad_norm": 0.28740663403193084,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.1085,
"step": 1210
},
{
"epoch": 4.911392405063291,
"grad_norm": 0.29376625933844686,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.1098,
"step": 1211
},
{
"epoch": 4.915443037974684,
"grad_norm": 0.3024180374886568,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.0955,
"step": 1212
},
{
"epoch": 4.919493670886076,
"grad_norm": 0.3073253767733607,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.11,
"step": 1213
},
{
"epoch": 4.923544303797469,
"grad_norm": 0.29731383682835083,
"learning_rate": 2.061435966221881e-08,
"loss": 0.0813,
"step": 1214
},
{
"epoch": 4.927594936708861,
"grad_norm": 0.2960770237917369,
"learning_rate": 1.811846652477245e-08,
"loss": 0.0834,
"step": 1215
},
{
"epoch": 4.931645569620253,
"grad_norm": 0.2923922255127805,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.0889,
"step": 1216
},
{
"epoch": 4.935696202531646,
"grad_norm": 0.30386138951475666,
"learning_rate": 1.360949303591097e-08,
"loss": 0.103,
"step": 1217
},
{
"epoch": 4.939746835443038,
"grad_norm": 0.30020018996103565,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.0897,
"step": 1218
},
{
"epoch": 4.943797468354431,
"grad_norm": 0.30153613853444866,
"learning_rate": 9.744388816668172e-09,
"loss": 0.0973,
"step": 1219
},
{
"epoch": 4.947848101265823,
"grad_norm": 0.29660613383511797,
"learning_rate": 8.05332740472009e-09,
"loss": 0.0825,
"step": 1220
},
{
"epoch": 4.951898734177215,
"grad_norm": 0.2872333000028156,
"learning_rate": 6.523278382872811e-09,
"loss": 0.1013,
"step": 1221
},
{
"epoch": 4.955949367088608,
"grad_norm": 0.3085325321593508,
"learning_rate": 5.15425407393133e-09,
"loss": 0.106,
"step": 1222
},
{
"epoch": 4.96,
"grad_norm": 0.3124183552366456,
"learning_rate": 3.94626550383137e-09,
"loss": 0.1051,
"step": 1223
},
{
"epoch": 4.964050632911392,
"grad_norm": 0.29819734436777734,
"learning_rate": 2.899322401546112e-09,
"loss": 0.1056,
"step": 1224
},
{
"epoch": 4.968101265822785,
"grad_norm": 0.30148710906440934,
"learning_rate": 2.013433199010706e-09,
"loss": 0.0867,
"step": 1225
},
{
"epoch": 4.972151898734177,
"grad_norm": 0.30056115510510584,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.0997,
"step": 1226
},
{
"epoch": 4.976202531645569,
"grad_norm": 0.2955477626741755,
"learning_rate": 7.248437353468695e-10,
"loss": 0.112,
"step": 1227
},
{
"epoch": 4.980253164556962,
"grad_norm": 0.2972089187897695,
"learning_rate": 3.221538523412449e-10,
"loss": 0.0847,
"step": 1228
},
{
"epoch": 4.984303797468354,
"grad_norm": 0.3006315083275406,
"learning_rate": 8.053862524670663e-11,
"loss": 0.1011,
"step": 1229
},
{
"epoch": 4.988354430379747,
"grad_norm": 0.3065057089609727,
"learning_rate": 0.0,
"loss": 0.0872,
"step": 1230
},
{
"epoch": 4.988354430379747,
"step": 1230,
"total_flos": 9.255890833993892e+17,
"train_loss": 0.41349471477958244,
"train_runtime": 30600.327,
"train_samples_per_second": 5.163,
"train_steps_per_second": 0.04
}
],
"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": 9.255890833993892e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}