e1_math_ms_phi / trainer_state.json
neginr's picture
End of training
cf92199 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.982278481012658,
"eval_steps": 500,
"global_step": 1230,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004050632911392405,
"grad_norm": 8.919178180597244,
"learning_rate": 3.2520325203252037e-07,
"loss": 1.632,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 8.673595264084508,
"learning_rate": 6.504065040650407e-07,
"loss": 1.6292,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 8.68368438140856,
"learning_rate": 9.75609756097561e-07,
"loss": 1.6525,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 8.651168105116529,
"learning_rate": 1.3008130081300815e-06,
"loss": 1.6468,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 8.328501743937613,
"learning_rate": 1.6260162601626018e-06,
"loss": 1.6195,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 8.107300757509313,
"learning_rate": 1.951219512195122e-06,
"loss": 1.6074,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 6.507053344602996,
"learning_rate": 2.2764227642276426e-06,
"loss": 1.5647,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 5.964016578087249,
"learning_rate": 2.601626016260163e-06,
"loss": 1.5367,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 3.6074650615287265,
"learning_rate": 2.926829268292683e-06,
"loss": 1.5002,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 3.000559035895805,
"learning_rate": 3.2520325203252037e-06,
"loss": 1.4673,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 3.0541132992399174,
"learning_rate": 3.577235772357724e-06,
"loss": 1.4768,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 5.888524493838664,
"learning_rate": 3.902439024390244e-06,
"loss": 1.4542,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 6.084399827162749,
"learning_rate": 4.227642276422765e-06,
"loss": 1.4238,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 5.667029158375839,
"learning_rate": 4.552845528455285e-06,
"loss": 1.471,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 5.481918568059667,
"learning_rate": 4.8780487804878055e-06,
"loss": 1.4239,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 3.7488655035030485,
"learning_rate": 5.203252032520326e-06,
"loss": 1.4199,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 3.1215126199052534,
"learning_rate": 5.528455284552846e-06,
"loss": 1.4089,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.3274774972232035,
"learning_rate": 5.853658536585366e-06,
"loss": 1.3313,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 1.9547513478680087,
"learning_rate": 6.178861788617887e-06,
"loss": 1.3512,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.6828900782979497,
"learning_rate": 6.504065040650407e-06,
"loss": 1.3183,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.7414608299683423,
"learning_rate": 6.829268292682928e-06,
"loss": 1.3517,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.5518174228917418,
"learning_rate": 7.154471544715448e-06,
"loss": 1.2882,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.7918950458456862,
"learning_rate": 7.4796747967479676e-06,
"loss": 1.3007,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 1.3757214515178946,
"learning_rate": 7.804878048780489e-06,
"loss": 1.2985,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 1.5010948828399315,
"learning_rate": 8.130081300813009e-06,
"loss": 1.2758,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 1.2741223244045816,
"learning_rate": 8.45528455284553e-06,
"loss": 1.2852,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 1.2566260486029976,
"learning_rate": 8.78048780487805e-06,
"loss": 1.3032,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 1.3516598052293818,
"learning_rate": 9.10569105691057e-06,
"loss": 1.3005,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 1.4319348259229154,
"learning_rate": 9.43089430894309e-06,
"loss": 1.2628,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 1.123797701572248,
"learning_rate": 9.756097560975611e-06,
"loss": 1.2436,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 1.3401947133251655,
"learning_rate": 1.008130081300813e-05,
"loss": 1.2282,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 1.4190119104838055,
"learning_rate": 1.0406504065040652e-05,
"loss": 1.2927,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 1.327790852389487,
"learning_rate": 1.0731707317073172e-05,
"loss": 1.2234,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 1.1962078252299655,
"learning_rate": 1.1056910569105692e-05,
"loss": 1.2336,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 1.0169161731611478,
"learning_rate": 1.1382113821138213e-05,
"loss": 1.2375,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.9105949020974995,
"learning_rate": 1.1707317073170731e-05,
"loss": 1.2546,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.8305659861062848,
"learning_rate": 1.2032520325203254e-05,
"loss": 1.2285,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.9077993000011223,
"learning_rate": 1.2357723577235774e-05,
"loss": 1.229,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.8733172385786414,
"learning_rate": 1.2682926829268294e-05,
"loss": 1.2013,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.8904872194202288,
"learning_rate": 1.3008130081300815e-05,
"loss": 1.2041,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 1.135052493962026,
"learning_rate": 1.3333333333333333e-05,
"loss": 1.2243,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.8914630591225475,
"learning_rate": 1.3658536585365855e-05,
"loss": 1.2044,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.9085805048682686,
"learning_rate": 1.3983739837398376e-05,
"loss": 1.2038,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.8994810502065087,
"learning_rate": 1.4308943089430896e-05,
"loss": 1.2143,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.831359430579628,
"learning_rate": 1.4634146341463415e-05,
"loss": 1.1941,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 1.0765056015603955,
"learning_rate": 1.4959349593495935e-05,
"loss": 1.2324,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.8200355290452236,
"learning_rate": 1.528455284552846e-05,
"loss": 1.1833,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 1.2559494180783668,
"learning_rate": 1.5609756097560978e-05,
"loss": 1.1832,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 1.1102906353564486,
"learning_rate": 1.5934959349593496e-05,
"loss": 1.2041,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.8687736935532478,
"learning_rate": 1.6260162601626018e-05,
"loss": 1.1877,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 1.2555261587271607,
"learning_rate": 1.6585365853658537e-05,
"loss": 1.1771,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 1.079416562961399,
"learning_rate": 1.691056910569106e-05,
"loss": 1.2076,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.9300367065867625,
"learning_rate": 1.7235772357723578e-05,
"loss": 1.1947,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.9856316575273348,
"learning_rate": 1.75609756097561e-05,
"loss": 1.1815,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.884399422848872,
"learning_rate": 1.788617886178862e-05,
"loss": 1.2023,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.8344193336125774,
"learning_rate": 1.821138211382114e-05,
"loss": 1.1613,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 1.0691639335166534,
"learning_rate": 1.8536585365853663e-05,
"loss": 1.1758,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.9310194463369935,
"learning_rate": 1.886178861788618e-05,
"loss": 1.1812,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.8733483566393783,
"learning_rate": 1.91869918699187e-05,
"loss": 1.1714,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.9180623561370465,
"learning_rate": 1.9512195121951222e-05,
"loss": 1.183,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.8931554128210901,
"learning_rate": 1.983739837398374e-05,
"loss": 1.1774,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.9832049531192,
"learning_rate": 2.016260162601626e-05,
"loss": 1.1886,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.943148995496054,
"learning_rate": 2.048780487804878e-05,
"loss": 1.1906,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 1.189623548781834,
"learning_rate": 2.0813008130081303e-05,
"loss": 1.1739,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.9929850225487156,
"learning_rate": 2.1138211382113822e-05,
"loss": 1.1779,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 1.0568944174098178,
"learning_rate": 2.1463414634146344e-05,
"loss": 1.1491,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 1.381830771590163,
"learning_rate": 2.1788617886178863e-05,
"loss": 1.1541,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 0.9583019821845234,
"learning_rate": 2.2113821138211385e-05,
"loss": 1.1594,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 1.2929329286802893,
"learning_rate": 2.2439024390243907e-05,
"loss": 1.1677,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 1.0007738510551807,
"learning_rate": 2.2764227642276426e-05,
"loss": 1.1655,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 1.0485647612686049,
"learning_rate": 2.3089430894308948e-05,
"loss": 1.1598,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.976662563153364,
"learning_rate": 2.3414634146341463e-05,
"loss": 1.1691,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 1.4815421452229116,
"learning_rate": 2.3739837398373985e-05,
"loss": 1.1641,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 1.024878886553717,
"learning_rate": 2.4065040650406507e-05,
"loss": 1.2109,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 1.5080638673137183,
"learning_rate": 2.4390243902439026e-05,
"loss": 1.1722,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 1.5231584680756194,
"learning_rate": 2.4715447154471548e-05,
"loss": 1.191,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.9846431663454013,
"learning_rate": 2.5040650406504066e-05,
"loss": 1.1696,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 1.3088947937437125,
"learning_rate": 2.536585365853659e-05,
"loss": 1.2019,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 0.9636869058642805,
"learning_rate": 2.569105691056911e-05,
"loss": 1.1885,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 1.4750147108541452,
"learning_rate": 2.601626016260163e-05,
"loss": 1.1763,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 1.2314295061912226,
"learning_rate": 2.634146341463415e-05,
"loss": 1.1641,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 1.4058202463294551,
"learning_rate": 2.6666666666666667e-05,
"loss": 1.1597,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 1.563017087764639,
"learning_rate": 2.699186991869919e-05,
"loss": 1.1513,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 1.2780323989693712,
"learning_rate": 2.731707317073171e-05,
"loss": 1.1872,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 1.2597160870985336,
"learning_rate": 2.764227642276423e-05,
"loss": 1.1758,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 1.412266960953602,
"learning_rate": 2.796747967479675e-05,
"loss": 1.1446,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 1.3820952304011276,
"learning_rate": 2.829268292682927e-05,
"loss": 1.1574,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 1.080871685122682,
"learning_rate": 2.8617886178861792e-05,
"loss": 1.1212,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 1.5958339730460864,
"learning_rate": 2.8943089430894314e-05,
"loss": 1.1656,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 1.3892444653514102,
"learning_rate": 2.926829268292683e-05,
"loss": 1.1455,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 0.9800145780527483,
"learning_rate": 2.959349593495935e-05,
"loss": 1.1507,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 1.49272112680718,
"learning_rate": 2.991869918699187e-05,
"loss": 1.1738,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 1.0382799720803453,
"learning_rate": 3.0243902439024392e-05,
"loss": 1.1795,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 1.251614689247021,
"learning_rate": 3.056910569105692e-05,
"loss": 1.1538,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 0.8911842168526203,
"learning_rate": 3.089430894308943e-05,
"loss": 1.155,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 1.2100035202633923,
"learning_rate": 3.1219512195121955e-05,
"loss": 1.1546,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 1.2740532101340547,
"learning_rate": 3.154471544715447e-05,
"loss": 1.1421,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 1.177709872335554,
"learning_rate": 3.186991869918699e-05,
"loss": 1.1285,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 1.2260003025651016,
"learning_rate": 3.2195121951219514e-05,
"loss": 1.1678,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 1.2472283041808074,
"learning_rate": 3.2520325203252037e-05,
"loss": 1.1523,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 1.3174212744835112,
"learning_rate": 3.284552845528456e-05,
"loss": 1.1642,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 1.1335960317139722,
"learning_rate": 3.3170731707317074e-05,
"loss": 1.1762,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 1.4390364231381467,
"learning_rate": 3.3495934959349596e-05,
"loss": 1.1492,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 1.0437016003027844,
"learning_rate": 3.382113821138212e-05,
"loss": 1.1495,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 1.747629182905173,
"learning_rate": 3.414634146341463e-05,
"loss": 1.149,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 1.4140935559814778,
"learning_rate": 3.4471544715447155e-05,
"loss": 1.1603,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 1.577482709722711,
"learning_rate": 3.479674796747968e-05,
"loss": 1.1554,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 1.532679643138614,
"learning_rate": 3.51219512195122e-05,
"loss": 1.1661,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 1.2569703595089612,
"learning_rate": 3.544715447154472e-05,
"loss": 1.161,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 1.3057796581946104,
"learning_rate": 3.577235772357724e-05,
"loss": 1.1904,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 1.2191766899760939,
"learning_rate": 3.609756097560976e-05,
"loss": 1.1595,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 1.1834847992519257,
"learning_rate": 3.642276422764228e-05,
"loss": 1.1859,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 1.2051230360019403,
"learning_rate": 3.67479674796748e-05,
"loss": 1.157,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 1.0324192504052867,
"learning_rate": 3.7073170731707325e-05,
"loss": 1.1289,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 1.4428641384835095,
"learning_rate": 3.739837398373984e-05,
"loss": 1.1485,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 1.1207943290102482,
"learning_rate": 3.772357723577236e-05,
"loss": 1.1422,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 1.388188280491624,
"learning_rate": 3.804878048780488e-05,
"loss": 1.1554,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 1.1874131662642784,
"learning_rate": 3.83739837398374e-05,
"loss": 1.1851,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 1.359634715788527,
"learning_rate": 3.869918699186992e-05,
"loss": 1.1452,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 1.2088795832181198,
"learning_rate": 3.9024390243902444e-05,
"loss": 1.1532,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 1.4543158199550668,
"learning_rate": 3.9349593495934966e-05,
"loss": 1.1713,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 1.1955275985715799,
"learning_rate": 3.967479674796748e-05,
"loss": 1.1434,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 1.2748059746551679,
"learning_rate": 4e-05,
"loss": 1.1365,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 1.3130509768061558,
"learning_rate": 3.999991946137476e-05,
"loss": 1.1773,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 1.1664086903361366,
"learning_rate": 3.999967784614766e-05,
"loss": 1.1448,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 1.0774434663031882,
"learning_rate": 3.9999275156264656e-05,
"loss": 1.1515,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 1.5966638873078685,
"learning_rate": 3.999871139496895e-05,
"loss": 1.1498,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 1.1380957018492472,
"learning_rate": 3.9997986566800995e-05,
"loss": 1.1743,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 1.404515303869956,
"learning_rate": 3.999710067759846e-05,
"loss": 1.1409,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 1.2629541372527973,
"learning_rate": 3.999605373449617e-05,
"loss": 1.1649,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 1.0753919258302234,
"learning_rate": 3.9994845745926075e-05,
"loss": 1.1642,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 1.2682038487391694,
"learning_rate": 3.999347672161713e-05,
"loss": 1.1672,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 1.1208716300372479,
"learning_rate": 3.999194667259528e-05,
"loss": 1.1702,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 1.5232901577764084,
"learning_rate": 3.999025561118334e-05,
"loss": 1.1391,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 1.2098522306940993,
"learning_rate": 3.998840355100086e-05,
"loss": 1.1871,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 1.1393017339205087,
"learning_rate": 3.998639050696409e-05,
"loss": 1.155,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 1.0502265680756275,
"learning_rate": 3.998421649528582e-05,
"loss": 1.149,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 1.6958547215171922,
"learning_rate": 3.9981881533475234e-05,
"loss": 1.1582,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.9547058199963063,
"learning_rate": 3.997938564033779e-05,
"loss": 1.1431,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 1.2940737865336196,
"learning_rate": 3.9976728835975064e-05,
"loss": 1.1477,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 1.2101100890023126,
"learning_rate": 3.9973911141784605e-05,
"loss": 1.1451,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 1.1003587720586252,
"learning_rate": 3.997093258045973e-05,
"loss": 1.1661,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 1.2368650678959234,
"learning_rate": 3.996779317598936e-05,
"loss": 1.1317,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 1.2491765721355004,
"learning_rate": 3.996449295365782e-05,
"loss": 1.1358,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.9408536245289014,
"learning_rate": 3.996103194004467e-05,
"loss": 1.1387,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 1.4441898744234942,
"learning_rate": 3.995741016302441e-05,
"loss": 1.1454,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 0.8238200163682206,
"learning_rate": 3.9953627651766364e-05,
"loss": 1.1395,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 1.3003555625058951,
"learning_rate": 3.9949684436734325e-05,
"loss": 1.1572,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 1.0299846808277295,
"learning_rate": 3.994558054968643e-05,
"loss": 1.129,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 1.102832515594259,
"learning_rate": 3.994131602367481e-05,
"loss": 1.1406,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 1.2432689651405364,
"learning_rate": 3.9936890893045376e-05,
"loss": 1.1456,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.8763290998629578,
"learning_rate": 3.993230519343752e-05,
"loss": 1.1498,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 1.240477078262571,
"learning_rate": 3.992755896178383e-05,
"loss": 1.166,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 1.0582376106511402,
"learning_rate": 3.992265223630981e-05,
"loss": 1.1366,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 1.146169264818053,
"learning_rate": 3.991758505653355e-05,
"loss": 1.1314,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 1.1067032878497,
"learning_rate": 3.991235746326543e-05,
"loss": 1.1725,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.9104306232192309,
"learning_rate": 3.9906969498607745e-05,
"loss": 1.1559,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 0.912581976169768,
"learning_rate": 3.990142120595444e-05,
"loss": 1.1585,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 0.9975882319575428,
"learning_rate": 3.98957126299907e-05,
"loss": 1.1405,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 1.0556489629966959,
"learning_rate": 3.9889843816692596e-05,
"loss": 1.1165,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 1.086349370633227,
"learning_rate": 3.9883814813326766e-05,
"loss": 1.1421,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 1.0814479914603263,
"learning_rate": 3.9877625668449956e-05,
"loss": 1.1532,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 1.171227238750386,
"learning_rate": 3.98712764319087e-05,
"loss": 1.1503,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.8995431231898411,
"learning_rate": 3.9864767154838864e-05,
"loss": 1.1169,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 0.9304114128129333,
"learning_rate": 3.9858097889665277e-05,
"loss": 1.1619,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.8398465005290133,
"learning_rate": 3.985126869010129e-05,
"loss": 1.1593,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 1.0094031732574285,
"learning_rate": 3.984427961114833e-05,
"loss": 1.1372,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 1.4146598273459479,
"learning_rate": 3.9837130709095475e-05,
"loss": 1.1554,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 0.9958808203316433,
"learning_rate": 3.982982204151901e-05,
"loss": 1.1451,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.8585793513169442,
"learning_rate": 3.982235366728193e-05,
"loss": 1.1447,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.9872104710648015,
"learning_rate": 3.9814725646533505e-05,
"loss": 1.1406,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 1.2988856260576516,
"learning_rate": 3.9806938040708746e-05,
"loss": 1.163,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.8080493998769621,
"learning_rate": 3.9798990912527976e-05,
"loss": 1.1436,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.8264723142407565,
"learning_rate": 3.979088432599627e-05,
"loss": 1.1285,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 1.1871651391217388,
"learning_rate": 3.9782618346402964e-05,
"loss": 1.1464,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 1.0376494869038189,
"learning_rate": 3.977419304032111e-05,
"loss": 1.1544,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 0.7541215608988976,
"learning_rate": 3.976560847560697e-05,
"loss": 1.1298,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 1.138373946123335,
"learning_rate": 3.9756864721399456e-05,
"loss": 1.1478,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 1.1117398406930232,
"learning_rate": 3.974796184811956e-05,
"loss": 1.1213,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.7479562389796871,
"learning_rate": 3.973889992746979e-05,
"loss": 1.1351,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.91376675675375,
"learning_rate": 3.972967903243361e-05,
"loss": 1.151,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 1.0238526991985826,
"learning_rate": 3.972029923727486e-05,
"loss": 1.1097,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.7763797158984429,
"learning_rate": 3.971076061753709e-05,
"loss": 1.1387,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.988853242799532,
"learning_rate": 3.9701063250043066e-05,
"loss": 1.1286,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 1.451579859611543,
"learning_rate": 3.969120721289402e-05,
"loss": 1.1628,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.810062644614527,
"learning_rate": 3.9681192585469146e-05,
"loss": 1.1082,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 1.5275557380709734,
"learning_rate": 3.9671019448424865e-05,
"loss": 1.1378,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 1.1781906854744864,
"learning_rate": 3.966068788369422e-05,
"loss": 1.1491,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.9940826626881009,
"learning_rate": 3.965019797448622e-05,
"loss": 1.115,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 1.183025127120996,
"learning_rate": 3.963954980528515e-05,
"loss": 1.1399,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.9981783888785299,
"learning_rate": 3.9628743461849905e-05,
"loss": 1.1607,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.7665658810938171,
"learning_rate": 3.961777903121329e-05,
"loss": 1.1457,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.9567554205434876,
"learning_rate": 3.960665660168131e-05,
"loss": 1.1442,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.6617907768181498,
"learning_rate": 3.9595376262832485e-05,
"loss": 1.1314,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.9449310639041723,
"learning_rate": 3.9583938105517127e-05,
"loss": 1.1228,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 1.0824988466266376,
"learning_rate": 3.957234222185657e-05,
"loss": 1.1329,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.8111268603737943,
"learning_rate": 3.9560588705242474e-05,
"loss": 1.1225,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 1.049747978390189,
"learning_rate": 3.954867765033605e-05,
"loss": 1.1354,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.9705871570919206,
"learning_rate": 3.953660915306728e-05,
"loss": 1.1196,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 1.0055274582209144,
"learning_rate": 3.952438331063419e-05,
"loss": 1.1388,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.9437335841669704,
"learning_rate": 3.951200022150205e-05,
"loss": 1.1425,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.9801163060403421,
"learning_rate": 3.949945998540253e-05,
"loss": 1.133,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.728230575141785,
"learning_rate": 3.9486762703332993e-05,
"loss": 1.1499,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.8922193689828343,
"learning_rate": 3.947390847755559e-05,
"loss": 1.1503,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.9253692993826793,
"learning_rate": 3.946089741159648e-05,
"loss": 1.151,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.92887391550479,
"learning_rate": 3.944772961024501e-05,
"loss": 1.1359,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.7871928607967363,
"learning_rate": 3.943440517955285e-05,
"loss": 1.0979,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.7500593937320247,
"learning_rate": 3.9420924226833126e-05,
"loss": 1.1465,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.8128036643195986,
"learning_rate": 3.9407286860659566e-05,
"loss": 1.1168,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.7892178709341505,
"learning_rate": 3.9393493190865657e-05,
"loss": 1.1244,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.7825033139320727,
"learning_rate": 3.937954332854371e-05,
"loss": 1.1213,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.8239781057171566,
"learning_rate": 3.9365437386044016e-05,
"loss": 1.1299,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.8439820267553576,
"learning_rate": 3.935117547697387e-05,
"loss": 1.1255,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.704992752375289,
"learning_rate": 3.933675771619675e-05,
"loss": 1.1601,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.8037897188235253,
"learning_rate": 3.932218421983131e-05,
"loss": 1.1398,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.652432358414381,
"learning_rate": 3.9307455105250484e-05,
"loss": 1.1165,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.6543779903303438,
"learning_rate": 3.929257049108054e-05,
"loss": 1.1572,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.726359569687531,
"learning_rate": 3.927753049720011e-05,
"loss": 1.1479,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.6465691342457687,
"learning_rate": 3.9262335244739234e-05,
"loss": 1.1494,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.6094877620835975,
"learning_rate": 3.92469848560784e-05,
"loss": 1.1091,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.7152689561266967,
"learning_rate": 3.923147945484751e-05,
"loss": 1.1147,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.6768904995469742,
"learning_rate": 3.9215819165924956e-05,
"loss": 1.1283,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.6760389207098225,
"learning_rate": 3.920000411543654e-05,
"loss": 1.1237,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.7165196045896339,
"learning_rate": 3.9184034430754495e-05,
"loss": 1.1431,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.7611829254465782,
"learning_rate": 3.916791024049648e-05,
"loss": 1.1464,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.564398484913344,
"learning_rate": 3.91516316745245e-05,
"loss": 1.158,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.6763665797979383,
"learning_rate": 3.913519886394389e-05,
"loss": 1.1321,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.7137030939866731,
"learning_rate": 3.911861194110225e-05,
"loss": 1.1168,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.7382912182162918,
"learning_rate": 3.910187103958837e-05,
"loss": 1.1671,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.7287200521008317,
"learning_rate": 3.908497629423117e-05,
"loss": 1.1609,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.845806623089827,
"learning_rate": 3.9067927841098614e-05,
"loss": 1.1115,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.8423035227230184,
"learning_rate": 3.9050725817496594e-05,
"loss": 1.127,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.7638047812579917,
"learning_rate": 3.9033370361967844e-05,
"loss": 1.1178,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.7623986737456457,
"learning_rate": 3.901586161429081e-05,
"loss": 1.1332,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.7482963215370838,
"learning_rate": 3.8998199715478545e-05,
"loss": 1.1432,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 1.0075390270349474,
"learning_rate": 3.8980384807777564e-05,
"loss": 1.1185,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.6946637274936335,
"learning_rate": 3.896241703466667e-05,
"loss": 1.1117,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.7611670722683926,
"learning_rate": 3.894429654085585e-05,
"loss": 1.1523,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.7680550024022222,
"learning_rate": 3.892602347228505e-05,
"loss": 1.1133,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.6623381404599441,
"learning_rate": 3.890759797612307e-05,
"loss": 1.1104,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.8887976468329106,
"learning_rate": 3.888902020076632e-05,
"loss": 1.1034,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.7818839026728887,
"learning_rate": 3.887029029583764e-05,
"loss": 1.1371,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.636880900657873,
"learning_rate": 3.8851408412185125e-05,
"loss": 1.1247,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.6515873778176838,
"learning_rate": 3.8832374701880855e-05,
"loss": 1.1022,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.7568888087073089,
"learning_rate": 3.881318931821972e-05,
"loss": 1.1347,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.6927841735034864,
"learning_rate": 3.879385241571817e-05,
"loss": 1.1248,
"step": 246
},
{
"epoch": 1.000506329113924,
"grad_norm": 0.851925020191978,
"learning_rate": 3.8774364150112955e-05,
"loss": 1.1166,
"step": 247
},
{
"epoch": 1.0045569620253165,
"grad_norm": 0.7302611160838869,
"learning_rate": 3.8754724678359884e-05,
"loss": 1.0658,
"step": 248
},
{
"epoch": 1.0086075949367088,
"grad_norm": 0.7428109735639431,
"learning_rate": 3.873493415863256e-05,
"loss": 1.0668,
"step": 249
},
{
"epoch": 1.0126582278481013,
"grad_norm": 1.1887128526261883,
"learning_rate": 3.871499275032111e-05,
"loss": 1.0699,
"step": 250
},
{
"epoch": 1.0167088607594936,
"grad_norm": 0.7604058542091272,
"learning_rate": 3.869490061403091e-05,
"loss": 1.0734,
"step": 251
},
{
"epoch": 1.0207594936708861,
"grad_norm": 0.8719444222113303,
"learning_rate": 3.867465791158124e-05,
"loss": 1.0639,
"step": 252
},
{
"epoch": 1.0248101265822784,
"grad_norm": 1.0327603805745695,
"learning_rate": 3.865426480600407e-05,
"loss": 1.0542,
"step": 253
},
{
"epoch": 1.028860759493671,
"grad_norm": 1.0792345357075703,
"learning_rate": 3.863372146154264e-05,
"loss": 1.0761,
"step": 254
},
{
"epoch": 1.0329113924050632,
"grad_norm": 0.9509351146577677,
"learning_rate": 3.861302804365024e-05,
"loss": 1.0675,
"step": 255
},
{
"epoch": 1.0369620253164558,
"grad_norm": 0.6290074755751839,
"learning_rate": 3.85921847189888e-05,
"loss": 1.059,
"step": 256
},
{
"epoch": 1.041012658227848,
"grad_norm": 0.8900017572410547,
"learning_rate": 3.85711916554276e-05,
"loss": 1.1002,
"step": 257
},
{
"epoch": 1.0450632911392406,
"grad_norm": 0.7853911945656405,
"learning_rate": 3.85500490220419e-05,
"loss": 1.0857,
"step": 258
},
{
"epoch": 1.0491139240506329,
"grad_norm": 1.0309399183360497,
"learning_rate": 3.852875698911154e-05,
"loss": 1.0644,
"step": 259
},
{
"epoch": 1.0531645569620254,
"grad_norm": 0.6388600141271606,
"learning_rate": 3.850731572811963e-05,
"loss": 1.0544,
"step": 260
},
{
"epoch": 1.0572151898734177,
"grad_norm": 0.8707863231627212,
"learning_rate": 3.848572541175116e-05,
"loss": 1.0587,
"step": 261
},
{
"epoch": 1.0612658227848102,
"grad_norm": 0.8281593640840309,
"learning_rate": 3.846398621389154e-05,
"loss": 1.0686,
"step": 262
},
{
"epoch": 1.0653164556962025,
"grad_norm": 0.7760911529684377,
"learning_rate": 3.84420983096253e-05,
"loss": 1.088,
"step": 263
},
{
"epoch": 1.069367088607595,
"grad_norm": 0.6681045683112204,
"learning_rate": 3.8420061875234606e-05,
"loss": 1.0605,
"step": 264
},
{
"epoch": 1.0734177215189873,
"grad_norm": 0.7509195777806481,
"learning_rate": 3.839787708819787e-05,
"loss": 1.069,
"step": 265
},
{
"epoch": 1.0774683544303798,
"grad_norm": 0.703463317290142,
"learning_rate": 3.8375544127188325e-05,
"loss": 1.0661,
"step": 266
},
{
"epoch": 1.081518987341772,
"grad_norm": 0.8302803602472276,
"learning_rate": 3.8353063172072564e-05,
"loss": 1.0722,
"step": 267
},
{
"epoch": 1.0855696202531646,
"grad_norm": 0.7408056703063259,
"learning_rate": 3.8330434403909105e-05,
"loss": 1.019,
"step": 268
},
{
"epoch": 1.089620253164557,
"grad_norm": 0.9200118082870449,
"learning_rate": 3.8307658004946934e-05,
"loss": 1.084,
"step": 269
},
{
"epoch": 1.0936708860759494,
"grad_norm": 0.6830782349878524,
"learning_rate": 3.8284734158624046e-05,
"loss": 1.0795,
"step": 270
},
{
"epoch": 1.0977215189873417,
"grad_norm": 0.6290171148020816,
"learning_rate": 3.826166304956594e-05,
"loss": 1.0777,
"step": 271
},
{
"epoch": 1.1017721518987342,
"grad_norm": 0.6298782435032536,
"learning_rate": 3.8238444863584164e-05,
"loss": 1.0826,
"step": 272
},
{
"epoch": 1.1058227848101265,
"grad_norm": 0.5848225109826446,
"learning_rate": 3.821507978767479e-05,
"loss": 1.0651,
"step": 273
},
{
"epoch": 1.109873417721519,
"grad_norm": 0.6586408215509544,
"learning_rate": 3.819156801001693e-05,
"loss": 1.0675,
"step": 274
},
{
"epoch": 1.1139240506329113,
"grad_norm": 0.5831527772707501,
"learning_rate": 3.816790971997121e-05,
"loss": 1.0769,
"step": 275
},
{
"epoch": 1.1179746835443038,
"grad_norm": 0.686575061941051,
"learning_rate": 3.8144105108078246e-05,
"loss": 1.0604,
"step": 276
},
{
"epoch": 1.1220253164556961,
"grad_norm": 0.5020364413954236,
"learning_rate": 3.81201543660571e-05,
"loss": 1.0347,
"step": 277
},
{
"epoch": 1.1260759493670887,
"grad_norm": 0.7035479813806363,
"learning_rate": 3.809605768680377e-05,
"loss": 1.0636,
"step": 278
},
{
"epoch": 1.130126582278481,
"grad_norm": 0.5565818581972175,
"learning_rate": 3.807181526438958e-05,
"loss": 1.0814,
"step": 279
},
{
"epoch": 1.1341772151898735,
"grad_norm": 0.5677306448031064,
"learning_rate": 3.8047427294059697e-05,
"loss": 1.0633,
"step": 280
},
{
"epoch": 1.1382278481012658,
"grad_norm": 0.6173961268923084,
"learning_rate": 3.802289397223145e-05,
"loss": 1.0691,
"step": 281
},
{
"epoch": 1.1422784810126583,
"grad_norm": 0.5301834334678659,
"learning_rate": 3.7998215496492854e-05,
"loss": 1.0833,
"step": 282
},
{
"epoch": 1.1463291139240506,
"grad_norm": 0.6011118987057419,
"learning_rate": 3.797339206560096e-05,
"loss": 1.0883,
"step": 283
},
{
"epoch": 1.150379746835443,
"grad_norm": 0.6650495403277895,
"learning_rate": 3.794842387948027e-05,
"loss": 1.0541,
"step": 284
},
{
"epoch": 1.1544303797468354,
"grad_norm": 0.6164482855959628,
"learning_rate": 3.7923311139221114e-05,
"loss": 1.0668,
"step": 285
},
{
"epoch": 1.158481012658228,
"grad_norm": 0.6772241753439625,
"learning_rate": 3.7898054047078054e-05,
"loss": 1.0516,
"step": 286
},
{
"epoch": 1.1625316455696202,
"grad_norm": 0.7133131528228839,
"learning_rate": 3.787265280646825e-05,
"loss": 1.0442,
"step": 287
},
{
"epoch": 1.1665822784810127,
"grad_norm": 0.7377511386446475,
"learning_rate": 3.7847107621969786e-05,
"loss": 1.0771,
"step": 288
},
{
"epoch": 1.170632911392405,
"grad_norm": 0.7287657017250347,
"learning_rate": 3.7821418699320064e-05,
"loss": 1.0777,
"step": 289
},
{
"epoch": 1.1746835443037975,
"grad_norm": 0.6323498752691613,
"learning_rate": 3.7795586245414145e-05,
"loss": 1.0728,
"step": 290
},
{
"epoch": 1.1787341772151898,
"grad_norm": 0.7400244654674555,
"learning_rate": 3.776961046830306e-05,
"loss": 1.0657,
"step": 291
},
{
"epoch": 1.1827848101265823,
"grad_norm": 0.9009653446257706,
"learning_rate": 3.774349157719215e-05,
"loss": 1.0922,
"step": 292
},
{
"epoch": 1.1868354430379746,
"grad_norm": 0.8967034096090364,
"learning_rate": 3.7717229782439365e-05,
"loss": 1.0859,
"step": 293
},
{
"epoch": 1.1908860759493671,
"grad_norm": 0.6800363016745378,
"learning_rate": 3.769082529555359e-05,
"loss": 1.0958,
"step": 294
},
{
"epoch": 1.1949367088607594,
"grad_norm": 0.7162778792815482,
"learning_rate": 3.766427832919294e-05,
"loss": 1.0635,
"step": 295
},
{
"epoch": 1.198987341772152,
"grad_norm": 0.7273180614195838,
"learning_rate": 3.7637589097163024e-05,
"loss": 1.0516,
"step": 296
},
{
"epoch": 1.2030379746835442,
"grad_norm": 0.721060170876544,
"learning_rate": 3.761075781441526e-05,
"loss": 1.0479,
"step": 297
},
{
"epoch": 1.2070886075949367,
"grad_norm": 0.6227924174080554,
"learning_rate": 3.75837846970451e-05,
"loss": 1.0493,
"step": 298
},
{
"epoch": 1.211139240506329,
"grad_norm": 0.7217088622855358,
"learning_rate": 3.755666996229032e-05,
"loss": 1.0852,
"step": 299
},
{
"epoch": 1.2151898734177216,
"grad_norm": 0.836686728518911,
"learning_rate": 3.752941382852927e-05,
"loss": 1.0743,
"step": 300
},
{
"epoch": 1.2192405063291138,
"grad_norm": 0.7277643740920103,
"learning_rate": 3.7502016515279115e-05,
"loss": 1.0897,
"step": 301
},
{
"epoch": 1.2232911392405064,
"grad_norm": 0.6850182653879813,
"learning_rate": 3.7474478243194043e-05,
"loss": 1.0568,
"step": 302
},
{
"epoch": 1.2273417721518987,
"grad_norm": 0.9017692704395514,
"learning_rate": 3.744679923406351e-05,
"loss": 1.0667,
"step": 303
},
{
"epoch": 1.2313924050632912,
"grad_norm": 0.8198373357548715,
"learning_rate": 3.741897971081043e-05,
"loss": 1.0589,
"step": 304
},
{
"epoch": 1.2354430379746835,
"grad_norm": 0.5883426801322128,
"learning_rate": 3.739101989748946e-05,
"loss": 1.069,
"step": 305
},
{
"epoch": 1.239493670886076,
"grad_norm": 0.6917554646370985,
"learning_rate": 3.7362920019285066e-05,
"loss": 1.0814,
"step": 306
},
{
"epoch": 1.2435443037974683,
"grad_norm": 0.6167703970582605,
"learning_rate": 3.73346803025098e-05,
"loss": 1.0755,
"step": 307
},
{
"epoch": 1.2475949367088608,
"grad_norm": 0.5263170992914957,
"learning_rate": 3.730630097460247e-05,
"loss": 1.0671,
"step": 308
},
{
"epoch": 1.251645569620253,
"grad_norm": 0.5359765553482525,
"learning_rate": 3.727778226412628e-05,
"loss": 1.0775,
"step": 309
},
{
"epoch": 1.2556962025316456,
"grad_norm": 0.5680587134045668,
"learning_rate": 3.7249124400767006e-05,
"loss": 1.0525,
"step": 310
},
{
"epoch": 1.259746835443038,
"grad_norm": 0.5631570642578028,
"learning_rate": 3.722032761533114e-05,
"loss": 1.0806,
"step": 311
},
{
"epoch": 1.2637974683544304,
"grad_norm": 0.5196473906165777,
"learning_rate": 3.719139213974403e-05,
"loss": 1.0692,
"step": 312
},
{
"epoch": 1.2678481012658227,
"grad_norm": 0.6044405981626504,
"learning_rate": 3.7162318207048006e-05,
"loss": 1.0651,
"step": 313
},
{
"epoch": 1.2718987341772152,
"grad_norm": 0.5880175975954993,
"learning_rate": 3.713310605140055e-05,
"loss": 1.0763,
"step": 314
},
{
"epoch": 1.2759493670886077,
"grad_norm": 0.6432550631333414,
"learning_rate": 3.710375590807233e-05,
"loss": 1.07,
"step": 315
},
{
"epoch": 1.28,
"grad_norm": 0.6439083646551924,
"learning_rate": 3.7074268013445365e-05,
"loss": 1.0855,
"step": 316
},
{
"epoch": 1.2840506329113923,
"grad_norm": 0.680451887198867,
"learning_rate": 3.7044642605011114e-05,
"loss": 1.0483,
"step": 317
},
{
"epoch": 1.2881012658227848,
"grad_norm": 0.613897568786428,
"learning_rate": 3.701487992136854e-05,
"loss": 1.0359,
"step": 318
},
{
"epoch": 1.2921518987341774,
"grad_norm": 0.6379745612223656,
"learning_rate": 3.69849802022222e-05,
"loss": 1.0838,
"step": 319
},
{
"epoch": 1.2962025316455696,
"grad_norm": 0.5884206316242654,
"learning_rate": 3.6954943688380334e-05,
"loss": 1.0656,
"step": 320
},
{
"epoch": 1.300253164556962,
"grad_norm": 0.6970612946943731,
"learning_rate": 3.692477062175289e-05,
"loss": 1.0984,
"step": 321
},
{
"epoch": 1.3043037974683545,
"grad_norm": 0.6538909568911432,
"learning_rate": 3.689446124534958e-05,
"loss": 1.0483,
"step": 322
},
{
"epoch": 1.308354430379747,
"grad_norm": 0.6268620282723308,
"learning_rate": 3.686401580327799e-05,
"loss": 1.0637,
"step": 323
},
{
"epoch": 1.3124050632911393,
"grad_norm": 0.6425904174440024,
"learning_rate": 3.683343454074149e-05,
"loss": 1.0449,
"step": 324
},
{
"epoch": 1.3164556962025316,
"grad_norm": 0.5940749260803518,
"learning_rate": 3.6802717704037386e-05,
"loss": 1.0671,
"step": 325
},
{
"epoch": 1.320506329113924,
"grad_norm": 0.699255764735426,
"learning_rate": 3.6771865540554855e-05,
"loss": 1.0555,
"step": 326
},
{
"epoch": 1.3245569620253166,
"grad_norm": 0.5917509890942096,
"learning_rate": 3.674087829877297e-05,
"loss": 1.0541,
"step": 327
},
{
"epoch": 1.3286075949367089,
"grad_norm": 0.6137270127285811,
"learning_rate": 3.6709756228258735e-05,
"loss": 1.0698,
"step": 328
},
{
"epoch": 1.3326582278481012,
"grad_norm": 0.5755090651068986,
"learning_rate": 3.667849957966501e-05,
"loss": 1.0538,
"step": 329
},
{
"epoch": 1.3367088607594937,
"grad_norm": 0.6224987131983752,
"learning_rate": 3.6647108604728546e-05,
"loss": 1.0676,
"step": 330
},
{
"epoch": 1.3407594936708862,
"grad_norm": 0.6116251314444595,
"learning_rate": 3.661558355626795e-05,
"loss": 1.0946,
"step": 331
},
{
"epoch": 1.3448101265822785,
"grad_norm": 0.6781466562058358,
"learning_rate": 3.658392468818163e-05,
"loss": 1.0571,
"step": 332
},
{
"epoch": 1.3488607594936708,
"grad_norm": 0.6265405763510364,
"learning_rate": 3.655213225544574e-05,
"loss": 1.0687,
"step": 333
},
{
"epoch": 1.3529113924050633,
"grad_norm": 0.7071242727340072,
"learning_rate": 3.652020651411218e-05,
"loss": 1.0668,
"step": 334
},
{
"epoch": 1.3569620253164558,
"grad_norm": 0.6247363775477852,
"learning_rate": 3.6488147721306474e-05,
"loss": 1.0388,
"step": 335
},
{
"epoch": 1.3610126582278481,
"grad_norm": 0.6902335258085865,
"learning_rate": 3.645595613522574e-05,
"loss": 1.0933,
"step": 336
},
{
"epoch": 1.3650632911392404,
"grad_norm": 0.7099666299230967,
"learning_rate": 3.642363201513657e-05,
"loss": 1.0939,
"step": 337
},
{
"epoch": 1.369113924050633,
"grad_norm": 0.6853498903937094,
"learning_rate": 3.6391175621373006e-05,
"loss": 1.0441,
"step": 338
},
{
"epoch": 1.3731645569620254,
"grad_norm": 0.6253329167652545,
"learning_rate": 3.6358587215334355e-05,
"loss": 1.0303,
"step": 339
},
{
"epoch": 1.3772151898734177,
"grad_norm": 0.634332686548985,
"learning_rate": 3.632586705948318e-05,
"loss": 1.0892,
"step": 340
},
{
"epoch": 1.38126582278481,
"grad_norm": 0.6276856276234479,
"learning_rate": 3.629301541734311e-05,
"loss": 1.0893,
"step": 341
},
{
"epoch": 1.3853164556962025,
"grad_norm": 0.5851888393873211,
"learning_rate": 3.626003255349676e-05,
"loss": 1.0767,
"step": 342
},
{
"epoch": 1.389367088607595,
"grad_norm": 0.5523842753189205,
"learning_rate": 3.622691873358357e-05,
"loss": 1.0594,
"step": 343
},
{
"epoch": 1.3934177215189874,
"grad_norm": 0.6974949958407761,
"learning_rate": 3.61936742242977e-05,
"loss": 1.0714,
"step": 344
},
{
"epoch": 1.3974683544303796,
"grad_norm": 0.6282879258277368,
"learning_rate": 3.6160299293385864e-05,
"loss": 1.062,
"step": 345
},
{
"epoch": 1.4015189873417722,
"grad_norm": 0.5926713572554092,
"learning_rate": 3.612679420964516e-05,
"loss": 1.0924,
"step": 346
},
{
"epoch": 1.4055696202531647,
"grad_norm": 0.6846204542935677,
"learning_rate": 3.609315924292092e-05,
"loss": 1.0835,
"step": 347
},
{
"epoch": 1.409620253164557,
"grad_norm": 0.7897894760942937,
"learning_rate": 3.6059394664104554e-05,
"loss": 1.0745,
"step": 348
},
{
"epoch": 1.4136708860759493,
"grad_norm": 0.5358273925840039,
"learning_rate": 3.602550074513133e-05,
"loss": 1.0769,
"step": 349
},
{
"epoch": 1.4177215189873418,
"grad_norm": 0.7638590443177881,
"learning_rate": 3.599147775897822e-05,
"loss": 1.0693,
"step": 350
},
{
"epoch": 1.4217721518987343,
"grad_norm": 0.7206682857231989,
"learning_rate": 3.595732597966167e-05,
"loss": 1.0425,
"step": 351
},
{
"epoch": 1.4258227848101266,
"grad_norm": 0.6652489425605927,
"learning_rate": 3.592304568223542e-05,
"loss": 1.0573,
"step": 352
},
{
"epoch": 1.4298734177215189,
"grad_norm": 0.5997771105053521,
"learning_rate": 3.588863714278826e-05,
"loss": 1.0642,
"step": 353
},
{
"epoch": 1.4339240506329114,
"grad_norm": 0.7006602406091803,
"learning_rate": 3.585410063844186e-05,
"loss": 1.0797,
"step": 354
},
{
"epoch": 1.437974683544304,
"grad_norm": 0.6525063258068511,
"learning_rate": 3.581943644734846e-05,
"loss": 1.0876,
"step": 355
},
{
"epoch": 1.4420253164556962,
"grad_norm": 0.6544205795409309,
"learning_rate": 3.578464484868869e-05,
"loss": 1.0747,
"step": 356
},
{
"epoch": 1.4460759493670885,
"grad_norm": 0.6716548922121216,
"learning_rate": 3.5749726122669316e-05,
"loss": 1.0576,
"step": 357
},
{
"epoch": 1.450126582278481,
"grad_norm": 0.6360602548258462,
"learning_rate": 3.5714680550520943e-05,
"loss": 1.0947,
"step": 358
},
{
"epoch": 1.4541772151898735,
"grad_norm": 0.5688752779662498,
"learning_rate": 3.5679508414495794e-05,
"loss": 1.0599,
"step": 359
},
{
"epoch": 1.4582278481012658,
"grad_norm": 0.7600411916447805,
"learning_rate": 3.564420999786543e-05,
"loss": 1.0459,
"step": 360
},
{
"epoch": 1.4622784810126581,
"grad_norm": 0.5717931355122821,
"learning_rate": 3.560878558491842e-05,
"loss": 1.0689,
"step": 361
},
{
"epoch": 1.4663291139240506,
"grad_norm": 0.6897521782648611,
"learning_rate": 3.5573235460958145e-05,
"loss": 1.0764,
"step": 362
},
{
"epoch": 1.4703797468354431,
"grad_norm": 0.5987169183484596,
"learning_rate": 3.553755991230039e-05,
"loss": 1.0823,
"step": 363
},
{
"epoch": 1.4744303797468354,
"grad_norm": 0.6960298070899082,
"learning_rate": 3.5501759226271144e-05,
"loss": 1.0753,
"step": 364
},
{
"epoch": 1.4784810126582277,
"grad_norm": 0.5686492829795553,
"learning_rate": 3.546583369120419e-05,
"loss": 1.0655,
"step": 365
},
{
"epoch": 1.4825316455696202,
"grad_norm": 0.5729964978923757,
"learning_rate": 3.5429783596438864e-05,
"loss": 1.0632,
"step": 366
},
{
"epoch": 1.4865822784810128,
"grad_norm": 0.6957256175705179,
"learning_rate": 3.539360923231766e-05,
"loss": 1.0833,
"step": 367
},
{
"epoch": 1.490632911392405,
"grad_norm": 0.5313526646880745,
"learning_rate": 3.535731089018394e-05,
"loss": 1.0665,
"step": 368
},
{
"epoch": 1.4946835443037974,
"grad_norm": 0.7347693984835156,
"learning_rate": 3.532088886237956e-05,
"loss": 1.0567,
"step": 369
},
{
"epoch": 1.4987341772151899,
"grad_norm": 0.7709757856777937,
"learning_rate": 3.528434344224253e-05,
"loss": 1.084,
"step": 370
},
{
"epoch": 1.5027848101265824,
"grad_norm": 0.6404633469899204,
"learning_rate": 3.524767492410464e-05,
"loss": 1.0787,
"step": 371
},
{
"epoch": 1.5068354430379747,
"grad_norm": 0.7804844560791423,
"learning_rate": 3.521088360328908e-05,
"loss": 1.0586,
"step": 372
},
{
"epoch": 1.510886075949367,
"grad_norm": 0.6338499342758828,
"learning_rate": 3.517396977610811e-05,
"loss": 1.0648,
"step": 373
},
{
"epoch": 1.5149367088607595,
"grad_norm": 0.6337138070081827,
"learning_rate": 3.5136933739860595e-05,
"loss": 1.0425,
"step": 374
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.6864792072824519,
"learning_rate": 3.509977579282971e-05,
"loss": 1.1011,
"step": 375
},
{
"epoch": 1.5230379746835443,
"grad_norm": 0.6507818977505859,
"learning_rate": 3.5062496234280424e-05,
"loss": 1.0693,
"step": 376
},
{
"epoch": 1.5270886075949366,
"grad_norm": 0.7537152825223705,
"learning_rate": 3.502509536445719e-05,
"loss": 1.0491,
"step": 377
},
{
"epoch": 1.531139240506329,
"grad_norm": 0.6111283135720303,
"learning_rate": 3.498757348458147e-05,
"loss": 1.0936,
"step": 378
},
{
"epoch": 1.5351898734177216,
"grad_norm": 0.8147969541229043,
"learning_rate": 3.4949930896849324e-05,
"loss": 1.0711,
"step": 379
},
{
"epoch": 1.539240506329114,
"grad_norm": 0.5362473759660176,
"learning_rate": 3.491216790442899e-05,
"loss": 1.0961,
"step": 380
},
{
"epoch": 1.5432911392405062,
"grad_norm": 0.6198647916387064,
"learning_rate": 3.487428481145839e-05,
"loss": 1.0687,
"step": 381
},
{
"epoch": 1.5473417721518987,
"grad_norm": 0.620880320129933,
"learning_rate": 3.483628192304278e-05,
"loss": 1.0694,
"step": 382
},
{
"epoch": 1.5513924050632912,
"grad_norm": 0.5875929331133267,
"learning_rate": 3.479815954525219e-05,
"loss": 1.0671,
"step": 383
},
{
"epoch": 1.5554430379746835,
"grad_norm": 0.569549552740049,
"learning_rate": 3.475991798511899e-05,
"loss": 1.0597,
"step": 384
},
{
"epoch": 1.5594936708860758,
"grad_norm": 0.5798782723694165,
"learning_rate": 3.4721557550635464e-05,
"loss": 1.0533,
"step": 385
},
{
"epoch": 1.5635443037974683,
"grad_norm": 0.620673716876498,
"learning_rate": 3.468307855075128e-05,
"loss": 1.0774,
"step": 386
},
{
"epoch": 1.5675949367088609,
"grad_norm": 0.6485153368370037,
"learning_rate": 3.4644481295371005e-05,
"loss": 1.0501,
"step": 387
},
{
"epoch": 1.5716455696202531,
"grad_norm": 0.686513849791081,
"learning_rate": 3.460576609535163e-05,
"loss": 1.0803,
"step": 388
},
{
"epoch": 1.5756962025316454,
"grad_norm": 0.7342154527518032,
"learning_rate": 3.456693326250006e-05,
"loss": 1.0926,
"step": 389
},
{
"epoch": 1.579746835443038,
"grad_norm": 0.7236709903142894,
"learning_rate": 3.452798310957058e-05,
"loss": 1.0784,
"step": 390
},
{
"epoch": 1.5837974683544305,
"grad_norm": 0.8436438811509968,
"learning_rate": 3.4488915950262386e-05,
"loss": 1.0897,
"step": 391
},
{
"epoch": 1.5878481012658228,
"grad_norm": 0.7216457788532862,
"learning_rate": 3.4449732099216985e-05,
"loss": 1.0865,
"step": 392
},
{
"epoch": 1.591898734177215,
"grad_norm": 0.5894768929973824,
"learning_rate": 3.441043187201574e-05,
"loss": 1.0928,
"step": 393
},
{
"epoch": 1.5959493670886076,
"grad_norm": 0.6743427472533486,
"learning_rate": 3.437101558517728e-05,
"loss": 1.0677,
"step": 394
},
{
"epoch": 1.6,
"grad_norm": 0.661417610986794,
"learning_rate": 3.433148355615496e-05,
"loss": 1.072,
"step": 395
},
{
"epoch": 1.6040506329113924,
"grad_norm": 0.5980413818839474,
"learning_rate": 3.4291836103334294e-05,
"loss": 1.0707,
"step": 396
},
{
"epoch": 1.6081012658227847,
"grad_norm": 0.6301962016486362,
"learning_rate": 3.425207354603043e-05,
"loss": 1.0506,
"step": 397
},
{
"epoch": 1.6121518987341772,
"grad_norm": 0.6264889109263935,
"learning_rate": 3.421219620448553e-05,
"loss": 1.0374,
"step": 398
},
{
"epoch": 1.6162025316455697,
"grad_norm": 0.6080682030629735,
"learning_rate": 3.417220439986623e-05,
"loss": 1.0798,
"step": 399
},
{
"epoch": 1.620253164556962,
"grad_norm": 0.5374872563361605,
"learning_rate": 3.4132098454261024e-05,
"loss": 1.0691,
"step": 400
},
{
"epoch": 1.6243037974683543,
"grad_norm": 0.5632933057873966,
"learning_rate": 3.4091878690677676e-05,
"loss": 1.0559,
"step": 401
},
{
"epoch": 1.6283544303797468,
"grad_norm": 0.7171475758056569,
"learning_rate": 3.405154543304065e-05,
"loss": 1.0878,
"step": 402
},
{
"epoch": 1.6324050632911393,
"grad_norm": 0.6216014698453908,
"learning_rate": 3.401109900618843e-05,
"loss": 1.0877,
"step": 403
},
{
"epoch": 1.6364556962025316,
"grad_norm": 0.6475486715565288,
"learning_rate": 3.3970539735870996e-05,
"loss": 1.0644,
"step": 404
},
{
"epoch": 1.640506329113924,
"grad_norm": 0.6892698360121025,
"learning_rate": 3.392986794874714e-05,
"loss": 1.0702,
"step": 405
},
{
"epoch": 1.6445569620253164,
"grad_norm": 0.8235917319556474,
"learning_rate": 3.388908397238184e-05,
"loss": 1.093,
"step": 406
},
{
"epoch": 1.648607594936709,
"grad_norm": 0.5486611352546081,
"learning_rate": 3.384818813524362e-05,
"loss": 1.0917,
"step": 407
},
{
"epoch": 1.6526582278481012,
"grad_norm": 0.7461443278621569,
"learning_rate": 3.380718076670195e-05,
"loss": 1.054,
"step": 408
},
{
"epoch": 1.6567088607594935,
"grad_norm": 0.846969516183983,
"learning_rate": 3.376606219702454e-05,
"loss": 1.055,
"step": 409
},
{
"epoch": 1.660759493670886,
"grad_norm": 0.5014198455988566,
"learning_rate": 3.372483275737468e-05,
"loss": 1.0593,
"step": 410
},
{
"epoch": 1.6648101265822786,
"grad_norm": 0.6427045882811511,
"learning_rate": 3.368349277980861e-05,
"loss": 1.0763,
"step": 411
},
{
"epoch": 1.6688607594936709,
"grad_norm": 0.5173642011676177,
"learning_rate": 3.3642042597272844e-05,
"loss": 1.0767,
"step": 412
},
{
"epoch": 1.6729113924050631,
"grad_norm": 0.4970142290667448,
"learning_rate": 3.360048254360144e-05,
"loss": 1.0563,
"step": 413
},
{
"epoch": 1.6769620253164557,
"grad_norm": 0.6389208126874917,
"learning_rate": 3.355881295351336e-05,
"loss": 1.0466,
"step": 414
},
{
"epoch": 1.6810126582278482,
"grad_norm": 0.4907011809449259,
"learning_rate": 3.351703416260975e-05,
"loss": 1.0494,
"step": 415
},
{
"epoch": 1.6850632911392405,
"grad_norm": 0.5859539669080575,
"learning_rate": 3.347514650737126e-05,
"loss": 1.065,
"step": 416
},
{
"epoch": 1.6891139240506328,
"grad_norm": 0.5521156872316407,
"learning_rate": 3.3433150325155295e-05,
"loss": 1.0661,
"step": 417
},
{
"epoch": 1.6931645569620253,
"grad_norm": 0.5700032454898154,
"learning_rate": 3.339104595419334e-05,
"loss": 1.0705,
"step": 418
},
{
"epoch": 1.6972151898734178,
"grad_norm": 0.5324647864153427,
"learning_rate": 3.3348833733588204e-05,
"loss": 1.0592,
"step": 419
},
{
"epoch": 1.70126582278481,
"grad_norm": 0.5667554610447746,
"learning_rate": 3.3306514003311305e-05,
"loss": 1.0396,
"step": 420
},
{
"epoch": 1.7053164556962024,
"grad_norm": 0.5236086452588974,
"learning_rate": 3.326408710419996e-05,
"loss": 1.0473,
"step": 421
},
{
"epoch": 1.709367088607595,
"grad_norm": 0.5822604065227212,
"learning_rate": 3.322155337795454e-05,
"loss": 1.0321,
"step": 422
},
{
"epoch": 1.7134177215189874,
"grad_norm": 0.5408449259374103,
"learning_rate": 3.317891316713587e-05,
"loss": 1.085,
"step": 423
},
{
"epoch": 1.7174683544303797,
"grad_norm": 0.6090024830999022,
"learning_rate": 3.313616681516231e-05,
"loss": 1.0389,
"step": 424
},
{
"epoch": 1.721518987341772,
"grad_norm": 0.5695149904639301,
"learning_rate": 3.309331466630713e-05,
"loss": 1.0722,
"step": 425
},
{
"epoch": 1.7255696202531645,
"grad_norm": 0.5506639418630528,
"learning_rate": 3.305035706569563e-05,
"loss": 1.0773,
"step": 426
},
{
"epoch": 1.729620253164557,
"grad_norm": 0.5709849409928687,
"learning_rate": 3.3007294359302433e-05,
"loss": 1.0482,
"step": 427
},
{
"epoch": 1.7336708860759493,
"grad_norm": 0.5835088025137287,
"learning_rate": 3.296412689394864e-05,
"loss": 1.0387,
"step": 428
},
{
"epoch": 1.7377215189873416,
"grad_norm": 0.5305009034587226,
"learning_rate": 3.292085501729909e-05,
"loss": 1.0562,
"step": 429
},
{
"epoch": 1.7417721518987341,
"grad_norm": 0.6237100592554939,
"learning_rate": 3.2877479077859534e-05,
"loss": 1.0743,
"step": 430
},
{
"epoch": 1.7458227848101266,
"grad_norm": 0.5543336960760163,
"learning_rate": 3.283399942497381e-05,
"loss": 1.0759,
"step": 431
},
{
"epoch": 1.749873417721519,
"grad_norm": 0.5790570413736511,
"learning_rate": 3.279041640882108e-05,
"loss": 1.0792,
"step": 432
},
{
"epoch": 1.7539240506329112,
"grad_norm": 0.5120463004319638,
"learning_rate": 3.2746730380412964e-05,
"loss": 1.0809,
"step": 433
},
{
"epoch": 1.7579746835443038,
"grad_norm": 0.5273529420140992,
"learning_rate": 3.2702941691590726e-05,
"loss": 1.0389,
"step": 434
},
{
"epoch": 1.7620253164556963,
"grad_norm": 0.4618314417936204,
"learning_rate": 3.265905069502244e-05,
"loss": 1.0469,
"step": 435
},
{
"epoch": 1.7660759493670886,
"grad_norm": 0.4884854119265639,
"learning_rate": 3.261505774420016e-05,
"loss": 1.0779,
"step": 436
},
{
"epoch": 1.7701265822784809,
"grad_norm": 0.499570129013549,
"learning_rate": 3.257096319343707e-05,
"loss": 1.0439,
"step": 437
},
{
"epoch": 1.7741772151898734,
"grad_norm": 0.4574868623424848,
"learning_rate": 3.2526767397864614e-05,
"loss": 1.0701,
"step": 438
},
{
"epoch": 1.7782278481012659,
"grad_norm": 0.59387963230863,
"learning_rate": 3.248247071342966e-05,
"loss": 1.0648,
"step": 439
},
{
"epoch": 1.7822784810126582,
"grad_norm": 0.46976658981803177,
"learning_rate": 3.243807349689161e-05,
"loss": 1.0867,
"step": 440
},
{
"epoch": 1.7863291139240505,
"grad_norm": 0.6220536348554098,
"learning_rate": 3.2393576105819544e-05,
"loss": 1.0684,
"step": 441
},
{
"epoch": 1.790379746835443,
"grad_norm": 0.5247582785037236,
"learning_rate": 3.2348978898589333e-05,
"loss": 1.0901,
"step": 442
},
{
"epoch": 1.7944303797468355,
"grad_norm": 0.5388506427925662,
"learning_rate": 3.230428223438075e-05,
"loss": 1.0487,
"step": 443
},
{
"epoch": 1.7984810126582278,
"grad_norm": 0.5995545950935652,
"learning_rate": 3.225948647317459e-05,
"loss": 1.0567,
"step": 444
},
{
"epoch": 1.80253164556962,
"grad_norm": 0.6036606545319401,
"learning_rate": 3.2214591975749745e-05,
"loss": 1.0614,
"step": 445
},
{
"epoch": 1.8065822784810126,
"grad_norm": 0.592753131948971,
"learning_rate": 3.216959910368034e-05,
"loss": 1.0814,
"step": 446
},
{
"epoch": 1.8106329113924051,
"grad_norm": 0.5813218251277722,
"learning_rate": 3.212450821933277e-05,
"loss": 1.0771,
"step": 447
},
{
"epoch": 1.8146835443037974,
"grad_norm": 0.6008847053425626,
"learning_rate": 3.207931968586281e-05,
"loss": 1.084,
"step": 448
},
{
"epoch": 1.81873417721519,
"grad_norm": 0.5600702986975787,
"learning_rate": 3.203403386721272e-05,
"loss": 1.0624,
"step": 449
},
{
"epoch": 1.8227848101265822,
"grad_norm": 0.47896721831378614,
"learning_rate": 3.1988651128108245e-05,
"loss": 1.044,
"step": 450
},
{
"epoch": 1.8268354430379747,
"grad_norm": 0.5910449290022954,
"learning_rate": 3.194317183405573e-05,
"loss": 1.0781,
"step": 451
},
{
"epoch": 1.8308860759493673,
"grad_norm": 0.5889042728851661,
"learning_rate": 3.189759635133914e-05,
"loss": 1.0628,
"step": 452
},
{
"epoch": 1.8349367088607595,
"grad_norm": 0.540409036629324,
"learning_rate": 3.185192504701718e-05,
"loss": 1.1076,
"step": 453
},
{
"epoch": 1.8389873417721518,
"grad_norm": 0.661406619579928,
"learning_rate": 3.1806158288920234e-05,
"loss": 1.0854,
"step": 454
},
{
"epoch": 1.8430379746835444,
"grad_norm": 0.4816118350052103,
"learning_rate": 3.1760296445647477e-05,
"loss": 1.0432,
"step": 455
},
{
"epoch": 1.8470886075949369,
"grad_norm": 0.6834714473812729,
"learning_rate": 3.1714339886563896e-05,
"loss": 1.048,
"step": 456
},
{
"epoch": 1.8511392405063292,
"grad_norm": 0.5566223830689311,
"learning_rate": 3.166828898179731e-05,
"loss": 1.0722,
"step": 457
},
{
"epoch": 1.8551898734177215,
"grad_norm": 0.6215296245574825,
"learning_rate": 3.162214410223536e-05,
"loss": 1.0664,
"step": 458
},
{
"epoch": 1.859240506329114,
"grad_norm": 0.5501635355898049,
"learning_rate": 3.157590561952257e-05,
"loss": 1.0715,
"step": 459
},
{
"epoch": 1.8632911392405065,
"grad_norm": 0.5672975861268676,
"learning_rate": 3.152957390605732e-05,
"loss": 1.0635,
"step": 460
},
{
"epoch": 1.8673417721518988,
"grad_norm": 0.5816796198160982,
"learning_rate": 3.148314933498886e-05,
"loss": 1.0576,
"step": 461
},
{
"epoch": 1.871392405063291,
"grad_norm": 0.5577658083006578,
"learning_rate": 3.143663228021431e-05,
"loss": 1.0716,
"step": 462
},
{
"epoch": 1.8754430379746836,
"grad_norm": 0.49745229001439495,
"learning_rate": 3.1390023116375624e-05,
"loss": 1.0674,
"step": 463
},
{
"epoch": 1.879493670886076,
"grad_norm": 0.5197892843655256,
"learning_rate": 3.134332221885661e-05,
"loss": 1.0858,
"step": 464
},
{
"epoch": 1.8835443037974684,
"grad_norm": 0.5631218794195235,
"learning_rate": 3.129652996377987e-05,
"loss": 1.1043,
"step": 465
},
{
"epoch": 1.8875949367088607,
"grad_norm": 0.6009663443623484,
"learning_rate": 3.12496467280038e-05,
"loss": 1.0849,
"step": 466
},
{
"epoch": 1.8916455696202532,
"grad_norm": 0.5387077043467043,
"learning_rate": 3.120267288911952e-05,
"loss": 1.0683,
"step": 467
},
{
"epoch": 1.8956962025316457,
"grad_norm": 0.5609687470340434,
"learning_rate": 3.11556088254479e-05,
"loss": 1.0481,
"step": 468
},
{
"epoch": 1.899746835443038,
"grad_norm": 0.5631728400421996,
"learning_rate": 3.11084549160364e-05,
"loss": 1.0662,
"step": 469
},
{
"epoch": 1.9037974683544303,
"grad_norm": 0.5149489888685963,
"learning_rate": 3.106121154065615e-05,
"loss": 1.0609,
"step": 470
},
{
"epoch": 1.9078481012658228,
"grad_norm": 0.5480974227929971,
"learning_rate": 3.1013879079798805e-05,
"loss": 1.0369,
"step": 471
},
{
"epoch": 1.9118987341772153,
"grad_norm": 0.5250251379830478,
"learning_rate": 3.096645791467348e-05,
"loss": 1.0754,
"step": 472
},
{
"epoch": 1.9159493670886076,
"grad_norm": 0.5194262265941454,
"learning_rate": 3.091894842720373e-05,
"loss": 1.0807,
"step": 473
},
{
"epoch": 1.92,
"grad_norm": 0.4804461383834654,
"learning_rate": 3.0871351000024425e-05,
"loss": 1.0749,
"step": 474
},
{
"epoch": 1.9240506329113924,
"grad_norm": 0.5218817783186183,
"learning_rate": 3.0823666016478716e-05,
"loss": 1.0763,
"step": 475
},
{
"epoch": 1.928101265822785,
"grad_norm": 0.588773474831664,
"learning_rate": 3.0775893860614896e-05,
"loss": 1.0606,
"step": 476
},
{
"epoch": 1.9321518987341773,
"grad_norm": 0.5871546877932584,
"learning_rate": 3.0728034917183336e-05,
"loss": 1.0623,
"step": 477
},
{
"epoch": 1.9362025316455695,
"grad_norm": 0.5845443512172266,
"learning_rate": 3.06800895716334e-05,
"loss": 1.0831,
"step": 478
},
{
"epoch": 1.940253164556962,
"grad_norm": 0.7213944517150224,
"learning_rate": 3.063205821011029e-05,
"loss": 1.064,
"step": 479
},
{
"epoch": 1.9443037974683546,
"grad_norm": 0.5628327495614887,
"learning_rate": 3.0583941219452016e-05,
"loss": 1.0869,
"step": 480
},
{
"epoch": 1.9483544303797469,
"grad_norm": 0.6790602677265252,
"learning_rate": 3.053573898718618e-05,
"loss": 1.0638,
"step": 481
},
{
"epoch": 1.9524050632911392,
"grad_norm": 0.5594070095718157,
"learning_rate": 3.0487451901526956e-05,
"loss": 1.0653,
"step": 482
},
{
"epoch": 1.9564556962025317,
"grad_norm": 0.47480766443377165,
"learning_rate": 3.0439080351371875e-05,
"loss": 1.0947,
"step": 483
},
{
"epoch": 1.9605063291139242,
"grad_norm": 0.5133839045515312,
"learning_rate": 3.0390624726298764e-05,
"loss": 1.0833,
"step": 484
},
{
"epoch": 1.9645569620253165,
"grad_norm": 0.48792081575503593,
"learning_rate": 3.034208541656255e-05,
"loss": 1.0645,
"step": 485
},
{
"epoch": 1.9686075949367088,
"grad_norm": 0.4836850884373659,
"learning_rate": 3.029346281309218e-05,
"loss": 1.0787,
"step": 486
},
{
"epoch": 1.9726582278481013,
"grad_norm": 0.4585614881504581,
"learning_rate": 3.0244757307487415e-05,
"loss": 1.0406,
"step": 487
},
{
"epoch": 1.9767088607594938,
"grad_norm": 0.48315112758596346,
"learning_rate": 3.019596929201569e-05,
"loss": 1.064,
"step": 488
},
{
"epoch": 1.980759493670886,
"grad_norm": 0.4655406971604461,
"learning_rate": 3.0147099159608985e-05,
"loss": 1.0789,
"step": 489
},
{
"epoch": 1.9848101265822784,
"grad_norm": 0.5195320947508908,
"learning_rate": 3.0098147303860616e-05,
"loss": 1.053,
"step": 490
},
{
"epoch": 1.988860759493671,
"grad_norm": 0.515988539815759,
"learning_rate": 3.0049114119022117e-05,
"loss": 1.0822,
"step": 491
},
{
"epoch": 1.9929113924050634,
"grad_norm": 0.5233495145417082,
"learning_rate": 3.0000000000000004e-05,
"loss": 1.0499,
"step": 492
},
{
"epoch": 1.9969620253164557,
"grad_norm": 0.44928310291284335,
"learning_rate": 2.995080534235264e-05,
"loss": 1.0898,
"step": 493
},
{
"epoch": 2.001012658227848,
"grad_norm": 0.5178001869086861,
"learning_rate": 2.9901530542287044e-05,
"loss": 1.0513,
"step": 494
},
{
"epoch": 2.0050632911392405,
"grad_norm": 0.6042785300637347,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.9753,
"step": 495
},
{
"epoch": 2.009113924050633,
"grad_norm": 0.639430561151299,
"learning_rate": 2.980274210295326e-05,
"loss": 0.9704,
"step": 496
},
{
"epoch": 2.013164556962025,
"grad_norm": 0.7533250152754637,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.955,
"step": 497
},
{
"epoch": 2.0172151898734176,
"grad_norm": 0.721913067520114,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.9744,
"step": 498
},
{
"epoch": 2.02126582278481,
"grad_norm": 0.7286790417904487,
"learning_rate": 2.965396831793362e-05,
"loss": 0.9884,
"step": 499
},
{
"epoch": 2.0253164556962027,
"grad_norm": 0.7361651107447638,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.9318,
"step": 500
},
{
"epoch": 2.0293670886075947,
"grad_norm": 0.5513097991884015,
"learning_rate": 2.955439637024526e-05,
"loss": 0.9706,
"step": 501
},
{
"epoch": 2.0334177215189873,
"grad_norm": 0.8278235560112607,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.98,
"step": 502
},
{
"epoch": 2.0374683544303798,
"grad_norm": 0.6902935098123835,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.9668,
"step": 503
},
{
"epoch": 2.0415189873417723,
"grad_norm": 0.5874058553909567,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.9531,
"step": 504
},
{
"epoch": 2.0455696202531644,
"grad_norm": 0.7731987038147143,
"learning_rate": 2.93543322968538e-05,
"loss": 0.9787,
"step": 505
},
{
"epoch": 2.049620253164557,
"grad_norm": 0.5896620095346067,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.978,
"step": 506
},
{
"epoch": 2.0536708860759494,
"grad_norm": 0.7773940926504807,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.9617,
"step": 507
},
{
"epoch": 2.057721518987342,
"grad_norm": 0.6773392228985918,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.9427,
"step": 508
},
{
"epoch": 2.061772151898734,
"grad_norm": 0.682225011210242,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.9846,
"step": 509
},
{
"epoch": 2.0658227848101265,
"grad_norm": 0.6692425782165297,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.9805,
"step": 510
},
{
"epoch": 2.069873417721519,
"grad_norm": 0.5345561189808844,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.943,
"step": 511
},
{
"epoch": 2.0739240506329115,
"grad_norm": 0.7278600577513941,
"learning_rate": 2.900133526131588e-05,
"loss": 0.9571,
"step": 512
},
{
"epoch": 2.0779746835443036,
"grad_norm": 0.5561855876599269,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.9722,
"step": 513
},
{
"epoch": 2.082025316455696,
"grad_norm": 0.5493456532962281,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.943,
"step": 514
},
{
"epoch": 2.0860759493670886,
"grad_norm": 0.5322616406267464,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.9789,
"step": 515
},
{
"epoch": 2.090126582278481,
"grad_norm": 0.5646509807591403,
"learning_rate": 2.879801882631847e-05,
"loss": 0.9669,
"step": 516
},
{
"epoch": 2.094177215189873,
"grad_norm": 0.6106521168104145,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.9922,
"step": 517
},
{
"epoch": 2.0982278481012657,
"grad_norm": 0.5621514441859692,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.9824,
"step": 518
},
{
"epoch": 2.1022784810126582,
"grad_norm": 0.5886510608880776,
"learning_rate": 2.864478605989494e-05,
"loss": 0.9583,
"step": 519
},
{
"epoch": 2.1063291139240508,
"grad_norm": 0.5485154301639841,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.9871,
"step": 520
},
{
"epoch": 2.110379746835443,
"grad_norm": 0.5644980565531262,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.9678,
"step": 521
},
{
"epoch": 2.1144303797468353,
"grad_norm": 0.6235037369250406,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.9635,
"step": 522
},
{
"epoch": 2.118481012658228,
"grad_norm": 0.48068003691756295,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.9749,
"step": 523
},
{
"epoch": 2.1225316455696204,
"grad_norm": 0.5756525785437145,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.9435,
"step": 524
},
{
"epoch": 2.1265822784810124,
"grad_norm": 0.5880472769941477,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.9477,
"step": 525
},
{
"epoch": 2.130632911392405,
"grad_norm": 0.6218332738802881,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.9614,
"step": 526
},
{
"epoch": 2.1346835443037975,
"grad_norm": 0.6680213561834931,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.9641,
"step": 527
},
{
"epoch": 2.13873417721519,
"grad_norm": 0.6140279575993761,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.9645,
"step": 528
},
{
"epoch": 2.1427848101265825,
"grad_norm": 0.5852382721494092,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.9532,
"step": 529
},
{
"epoch": 2.1468354430379746,
"grad_norm": 0.49536384505356523,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.9856,
"step": 530
},
{
"epoch": 2.150886075949367,
"grad_norm": 0.6165234796839585,
"learning_rate": 2.802570062143278e-05,
"loss": 0.9565,
"step": 531
},
{
"epoch": 2.1549367088607596,
"grad_norm": 0.5189581087557762,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.9987,
"step": 532
},
{
"epoch": 2.1589873417721517,
"grad_norm": 0.48846039793860613,
"learning_rate": 2.792159532078314e-05,
"loss": 0.9747,
"step": 533
},
{
"epoch": 2.163037974683544,
"grad_norm": 0.471165871379027,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.9604,
"step": 534
},
{
"epoch": 2.1670886075949367,
"grad_norm": 0.5435421708449117,
"learning_rate": 2.781723482288857e-05,
"loss": 0.985,
"step": 535
},
{
"epoch": 2.1711392405063292,
"grad_norm": 0.49541348005888486,
"learning_rate": 2.7764959925285517e-05,
"loss": 1.0017,
"step": 536
},
{
"epoch": 2.1751898734177217,
"grad_norm": 0.43892731650358213,
"learning_rate": 2.771262248976272e-05,
"loss": 0.9434,
"step": 537
},
{
"epoch": 2.179240506329114,
"grad_norm": 0.4898778810688344,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.9631,
"step": 538
},
{
"epoch": 2.1832911392405063,
"grad_norm": 0.4842671648791523,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.9441,
"step": 539
},
{
"epoch": 2.187341772151899,
"grad_norm": 0.45396300935991607,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.9704,
"step": 540
},
{
"epoch": 2.191392405063291,
"grad_norm": 0.5273849380628973,
"learning_rate": 2.7502655806328e-05,
"loss": 0.9784,
"step": 541
},
{
"epoch": 2.1954430379746834,
"grad_norm": 0.4450887493440926,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.9686,
"step": 542
},
{
"epoch": 2.199493670886076,
"grad_norm": 0.5336088746454483,
"learning_rate": 2.739730822017673e-05,
"loss": 0.9437,
"step": 543
},
{
"epoch": 2.2035443037974685,
"grad_norm": 0.5699792338548684,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.9663,
"step": 544
},
{
"epoch": 2.207594936708861,
"grad_norm": 0.5914745009340056,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.9757,
"step": 545
},
{
"epoch": 2.211645569620253,
"grad_norm": 0.5653780218506591,
"learning_rate": 2.723884107774236e-05,
"loss": 0.98,
"step": 546
},
{
"epoch": 2.2156962025316456,
"grad_norm": 0.5839928440378154,
"learning_rate": 2.718590152796239e-05,
"loss": 0.924,
"step": 547
},
{
"epoch": 2.219746835443038,
"grad_norm": 0.467662970779907,
"learning_rate": 2.71329041039194e-05,
"loss": 0.9812,
"step": 548
},
{
"epoch": 2.22379746835443,
"grad_norm": 0.5620360568059414,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.9566,
"step": 549
},
{
"epoch": 2.2278481012658227,
"grad_norm": 0.43987238524956085,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.9558,
"step": 550
},
{
"epoch": 2.231898734177215,
"grad_norm": 0.49282325997344917,
"learning_rate": 2.697356885686189e-05,
"loss": 0.9755,
"step": 551
},
{
"epoch": 2.2359493670886077,
"grad_norm": 0.5616416900333447,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.9672,
"step": 552
},
{
"epoch": 2.24,
"grad_norm": 0.49727203823842525,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.9619,
"step": 553
},
{
"epoch": 2.2440506329113923,
"grad_norm": 0.4943663805268212,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.9751,
"step": 554
},
{
"epoch": 2.248101265822785,
"grad_norm": 0.6141278952787332,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.9842,
"step": 555
},
{
"epoch": 2.2521518987341773,
"grad_norm": 0.5214583299363905,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.9823,
"step": 556
},
{
"epoch": 2.2562025316455694,
"grad_norm": 0.4805341424948711,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.9446,
"step": 557
},
{
"epoch": 2.260253164556962,
"grad_norm": 0.591950010002409,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.9753,
"step": 558
},
{
"epoch": 2.2643037974683544,
"grad_norm": 0.6080229569342978,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.997,
"step": 559
},
{
"epoch": 2.268354430379747,
"grad_norm": 0.42188942352105485,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.9667,
"step": 560
},
{
"epoch": 2.2724050632911394,
"grad_norm": 0.5570824834148913,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.974,
"step": 561
},
{
"epoch": 2.2764556962025315,
"grad_norm": 0.4552813012796245,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.9381,
"step": 562
},
{
"epoch": 2.280506329113924,
"grad_norm": 0.551508656676984,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.9591,
"step": 563
},
{
"epoch": 2.2845569620253166,
"grad_norm": 0.4911131005012389,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.9691,
"step": 564
},
{
"epoch": 2.2886075949367086,
"grad_norm": 0.481710501836991,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.9502,
"step": 565
},
{
"epoch": 2.292658227848101,
"grad_norm": 0.4883130603600267,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.9544,
"step": 566
},
{
"epoch": 2.2967088607594937,
"grad_norm": 0.5431071807550029,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.9675,
"step": 567
},
{
"epoch": 2.300759493670886,
"grad_norm": 0.43441489258742993,
"learning_rate": 2.606146264569603e-05,
"loss": 0.9613,
"step": 568
},
{
"epoch": 2.3048101265822787,
"grad_norm": 0.46937925897519844,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.9451,
"step": 569
},
{
"epoch": 2.3088607594936708,
"grad_norm": 0.5601786423507358,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.9741,
"step": 570
},
{
"epoch": 2.3129113924050633,
"grad_norm": 0.45278230577528467,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.9569,
"step": 571
},
{
"epoch": 2.316962025316456,
"grad_norm": 0.5087004484023451,
"learning_rate": 2.584472010606015e-05,
"loss": 0.976,
"step": 572
},
{
"epoch": 2.321012658227848,
"grad_norm": 0.4666921498009302,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.9715,
"step": 573
},
{
"epoch": 2.3250632911392404,
"grad_norm": 0.4391540004437758,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.966,
"step": 574
},
{
"epoch": 2.329113924050633,
"grad_norm": 0.4793013013585454,
"learning_rate": 2.568166741322048e-05,
"loss": 0.9614,
"step": 575
},
{
"epoch": 2.3331645569620254,
"grad_norm": 0.47415625620151874,
"learning_rate": 2.56272244128509e-05,
"loss": 0.9602,
"step": 576
},
{
"epoch": 2.337215189873418,
"grad_norm": 0.4700403662589336,
"learning_rate": 2.55727360915895e-05,
"loss": 0.9509,
"step": 577
},
{
"epoch": 2.34126582278481,
"grad_norm": 0.4864273574695621,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.9674,
"step": 578
},
{
"epoch": 2.3453164556962025,
"grad_norm": 0.5065676155807514,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.9619,
"step": 579
},
{
"epoch": 2.349367088607595,
"grad_norm": 0.5280827324275239,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.9723,
"step": 580
},
{
"epoch": 2.353417721518987,
"grad_norm": 0.4805925049352655,
"learning_rate": 2.535433837985559e-05,
"loss": 0.9858,
"step": 581
},
{
"epoch": 2.3574683544303796,
"grad_norm": 0.4882538501736841,
"learning_rate": 2.529963004393324e-05,
"loss": 0.9772,
"step": 582
},
{
"epoch": 2.361518987341772,
"grad_norm": 0.4478547925965307,
"learning_rate": 2.524487902551908e-05,
"loss": 0.9697,
"step": 583
},
{
"epoch": 2.3655696202531646,
"grad_norm": 0.41699643124655816,
"learning_rate": 2.519008576557029e-05,
"loss": 0.9658,
"step": 584
},
{
"epoch": 2.369620253164557,
"grad_norm": 0.4708912605930177,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.9556,
"step": 585
},
{
"epoch": 2.3736708860759492,
"grad_norm": 0.4410247746683741,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.9373,
"step": 586
},
{
"epoch": 2.3777215189873417,
"grad_norm": 0.4439179858344708,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.9407,
"step": 587
},
{
"epoch": 2.3817721518987343,
"grad_norm": 0.46611520161067166,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.9741,
"step": 588
},
{
"epoch": 2.3858227848101268,
"grad_norm": 0.4313974931834139,
"learning_rate": 2.491550129992345e-05,
"loss": 0.9462,
"step": 589
},
{
"epoch": 2.389873417721519,
"grad_norm": 0.4193079285763763,
"learning_rate": 2.486046386967024e-05,
"loss": 0.9479,
"step": 590
},
{
"epoch": 2.3939240506329114,
"grad_norm": 0.4492809315705656,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.9811,
"step": 591
},
{
"epoch": 2.397974683544304,
"grad_norm": 0.48279936856752276,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.9564,
"step": 592
},
{
"epoch": 2.4020253164556964,
"grad_norm": 0.4206463659705874,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.9884,
"step": 593
},
{
"epoch": 2.4060759493670885,
"grad_norm": 0.49316576415975527,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.9787,
"step": 594
},
{
"epoch": 2.410126582278481,
"grad_norm": 0.4847863852537041,
"learning_rate": 2.458469841202444e-05,
"loss": 0.9889,
"step": 595
},
{
"epoch": 2.4141772151898735,
"grad_norm": 0.4620681919757758,
"learning_rate": 2.452943276859226e-05,
"loss": 0.9736,
"step": 596
},
{
"epoch": 2.418227848101266,
"grad_norm": 0.4659993796137428,
"learning_rate": 2.447413064573125e-05,
"loss": 0.9917,
"step": 597
},
{
"epoch": 2.422278481012658,
"grad_norm": 0.49149106959666866,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.9675,
"step": 598
},
{
"epoch": 2.4263291139240506,
"grad_norm": 0.5034618512184392,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.9508,
"step": 599
},
{
"epoch": 2.430379746835443,
"grad_norm": 0.5003323200099419,
"learning_rate": 2.430800985597963e-05,
"loss": 0.9826,
"step": 600
},
{
"epoch": 2.4344303797468356,
"grad_norm": 0.4330246267096563,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.9932,
"step": 601
},
{
"epoch": 2.4384810126582277,
"grad_norm": 0.46432031610274777,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.9688,
"step": 602
},
{
"epoch": 2.44253164556962,
"grad_norm": 0.5491005608070328,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.9702,
"step": 603
},
{
"epoch": 2.4465822784810127,
"grad_norm": 0.5598427823296834,
"learning_rate": 2.408603181104957e-05,
"loss": 0.9536,
"step": 604
},
{
"epoch": 2.4506329113924052,
"grad_norm": 0.46856734833160557,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.9846,
"step": 605
},
{
"epoch": 2.4546835443037973,
"grad_norm": 0.48407382615175604,
"learning_rate": 2.397484355008521e-05,
"loss": 0.9477,
"step": 606
},
{
"epoch": 2.45873417721519,
"grad_norm": 0.4690032106299665,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.9921,
"step": 607
},
{
"epoch": 2.4627848101265823,
"grad_norm": 0.43095640490944653,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.9875,
"step": 608
},
{
"epoch": 2.466835443037975,
"grad_norm": 0.3997250915674939,
"learning_rate": 2.380782218329337e-05,
"loss": 0.9645,
"step": 609
},
{
"epoch": 2.470886075949367,
"grad_norm": 0.42339213871960174,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.9482,
"step": 610
},
{
"epoch": 2.4749367088607594,
"grad_norm": 0.46380282056151195,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.9651,
"step": 611
},
{
"epoch": 2.478987341772152,
"grad_norm": 0.4101196158161837,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.9723,
"step": 612
},
{
"epoch": 2.4830379746835445,
"grad_norm": 0.45676254232423374,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.9569,
"step": 613
},
{
"epoch": 2.4870886075949366,
"grad_norm": 0.455065960264916,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.9572,
"step": 614
},
{
"epoch": 2.491139240506329,
"grad_norm": 0.43863652931577707,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.9698,
"step": 615
},
{
"epoch": 2.4951898734177216,
"grad_norm": 0.41373573351358783,
"learning_rate": 2.341705325992901e-05,
"loss": 0.9788,
"step": 616
},
{
"epoch": 2.499240506329114,
"grad_norm": 0.4723943001668784,
"learning_rate": 2.336111544604222e-05,
"loss": 0.9849,
"step": 617
},
{
"epoch": 2.503291139240506,
"grad_norm": 0.43254739119292884,
"learning_rate": 2.33051505621937e-05,
"loss": 0.981,
"step": 618
},
{
"epoch": 2.5073417721518987,
"grad_norm": 0.4452355092716459,
"learning_rate": 2.324915905911693e-05,
"loss": 0.9838,
"step": 619
},
{
"epoch": 2.511392405063291,
"grad_norm": 0.43032330525349427,
"learning_rate": 2.319314138775977e-05,
"loss": 0.9413,
"step": 620
},
{
"epoch": 2.5154430379746833,
"grad_norm": 0.43547255653640415,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.9746,
"step": 621
},
{
"epoch": 2.519493670886076,
"grad_norm": 0.46906888229362587,
"learning_rate": 2.308102934504593e-05,
"loss": 0.972,
"step": 622
},
{
"epoch": 2.5235443037974683,
"grad_norm": 0.4414751661456316,
"learning_rate": 2.3024935876624222e-05,
"loss": 1.0093,
"step": 623
},
{
"epoch": 2.527594936708861,
"grad_norm": 0.4405911182132202,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.9546,
"step": 624
},
{
"epoch": 2.5316455696202533,
"grad_norm": 0.4257045930215373,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.9844,
"step": 625
},
{
"epoch": 2.5356962025316454,
"grad_norm": 0.42751225488727,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.9895,
"step": 626
},
{
"epoch": 2.539746835443038,
"grad_norm": 0.4374512809675193,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.9566,
"step": 627
},
{
"epoch": 2.5437974683544304,
"grad_norm": 0.4718982449461261,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.9625,
"step": 628
},
{
"epoch": 2.547848101265823,
"grad_norm": 0.4822321400078465,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.9795,
"step": 629
},
{
"epoch": 2.5518987341772155,
"grad_norm": 0.43107742997922993,
"learning_rate": 2.26316247733093e-05,
"loss": 0.9845,
"step": 630
},
{
"epoch": 2.5559493670886075,
"grad_norm": 0.43969119019373465,
"learning_rate": 2.257534907110328e-05,
"loss": 0.9771,
"step": 631
},
{
"epoch": 2.56,
"grad_norm": 0.44447179600934333,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.9871,
"step": 632
},
{
"epoch": 2.5640506329113926,
"grad_norm": 0.5106945240479446,
"learning_rate": 2.246273589557294e-05,
"loss": 0.9999,
"step": 633
},
{
"epoch": 2.5681012658227846,
"grad_norm": 0.5010415542480293,
"learning_rate": 2.240639932921966e-05,
"loss": 0.9699,
"step": 634
},
{
"epoch": 2.572151898734177,
"grad_norm": 0.49737208496861973,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.9725,
"step": 635
},
{
"epoch": 2.5762025316455697,
"grad_norm": 0.43659996936024736,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.9777,
"step": 636
},
{
"epoch": 2.580253164556962,
"grad_norm": 0.4855875995435425,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.9496,
"step": 637
},
{
"epoch": 2.5843037974683547,
"grad_norm": 0.5073343460176186,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.9793,
"step": 638
},
{
"epoch": 2.5883544303797468,
"grad_norm": 0.4905700280704824,
"learning_rate": 2.212443486523819e-05,
"loss": 0.9543,
"step": 639
},
{
"epoch": 2.5924050632911393,
"grad_norm": 0.5390036535821043,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.9765,
"step": 640
},
{
"epoch": 2.596455696202532,
"grad_norm": 0.44197690639406556,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.9955,
"step": 641
},
{
"epoch": 2.600506329113924,
"grad_norm": 0.5797476832067858,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.969,
"step": 642
},
{
"epoch": 2.6045569620253164,
"grad_norm": 0.4052698395628387,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.9657,
"step": 643
},
{
"epoch": 2.608607594936709,
"grad_norm": 0.4962216304195398,
"learning_rate": 2.184204266048803e-05,
"loss": 0.9613,
"step": 644
},
{
"epoch": 2.6126582278481014,
"grad_norm": 0.44575037255679884,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.9435,
"step": 645
},
{
"epoch": 2.616708860759494,
"grad_norm": 0.4857387580972144,
"learning_rate": 2.17289787446841e-05,
"loss": 0.9805,
"step": 646
},
{
"epoch": 2.620759493670886,
"grad_norm": 0.4279985289102459,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.9784,
"step": 647
},
{
"epoch": 2.6248101265822785,
"grad_norm": 0.4699298848122388,
"learning_rate": 2.161585912916385e-05,
"loss": 0.9694,
"step": 648
},
{
"epoch": 2.628860759493671,
"grad_norm": 0.5447894028003447,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.9844,
"step": 649
},
{
"epoch": 2.632911392405063,
"grad_norm": 0.4771158617167532,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.9764,
"step": 650
},
{
"epoch": 2.6369620253164556,
"grad_norm": 0.45931696137625866,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.9969,
"step": 651
},
{
"epoch": 2.641012658227848,
"grad_norm": 0.48909664100803973,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.9637,
"step": 652
},
{
"epoch": 2.6450632911392407,
"grad_norm": 0.4383503846675232,
"learning_rate": 2.133284032321232e-05,
"loss": 0.9605,
"step": 653
},
{
"epoch": 2.649113924050633,
"grad_norm": 0.4530984721108198,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.989,
"step": 654
},
{
"epoch": 2.6531645569620252,
"grad_norm": 0.4449365909431596,
"learning_rate": 2.121955446741306e-05,
"loss": 0.9644,
"step": 655
},
{
"epoch": 2.6572151898734178,
"grad_norm": 0.49548010841853063,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.939,
"step": 656
},
{
"epoch": 2.6612658227848103,
"grad_norm": 0.5570861097420376,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.9559,
"step": 657
},
{
"epoch": 2.6653164556962023,
"grad_norm": 0.48030733724396746,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.9846,
"step": 658
},
{
"epoch": 2.669367088607595,
"grad_norm": 0.5676156820144774,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.9579,
"step": 659
},
{
"epoch": 2.6734177215189874,
"grad_norm": 0.5286652820022774,
"learning_rate": 2.093617592756158e-05,
"loss": 0.9594,
"step": 660
},
{
"epoch": 2.67746835443038,
"grad_norm": 0.5057804250046244,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.9843,
"step": 661
},
{
"epoch": 2.6815189873417724,
"grad_norm": 0.5324628236253326,
"learning_rate": 2.082276853708978e-05,
"loss": 0.9534,
"step": 662
},
{
"epoch": 2.6855696202531645,
"grad_norm": 0.5290574374065112,
"learning_rate": 2.076605467380071e-05,
"loss": 0.9624,
"step": 663
},
{
"epoch": 2.689620253164557,
"grad_norm": 0.4425680912032712,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.9422,
"step": 664
},
{
"epoch": 2.6936708860759495,
"grad_norm": 0.485284813823776,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.9554,
"step": 665
},
{
"epoch": 2.6977215189873416,
"grad_norm": 0.41640336802861394,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.9721,
"step": 666
},
{
"epoch": 2.701772151898734,
"grad_norm": 0.3983737630432093,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.9741,
"step": 667
},
{
"epoch": 2.7058227848101266,
"grad_norm": 0.4405031200298459,
"learning_rate": 2.048240194884496e-05,
"loss": 0.9656,
"step": 668
},
{
"epoch": 2.709873417721519,
"grad_norm": 0.46571236074915373,
"learning_rate": 2.042565792045695e-05,
"loss": 0.9623,
"step": 669
},
{
"epoch": 2.7139240506329116,
"grad_norm": 0.4228238244472744,
"learning_rate": 2.036891046387857e-05,
"loss": 0.958,
"step": 670
},
{
"epoch": 2.7179746835443037,
"grad_norm": 0.421813537263323,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.9975,
"step": 671
},
{
"epoch": 2.7220253164556962,
"grad_norm": 0.4167730274354173,
"learning_rate": 2.025540709431948e-05,
"loss": 0.9826,
"step": 672
},
{
"epoch": 2.7260759493670887,
"grad_norm": 0.46553122087706356,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.978,
"step": 673
},
{
"epoch": 2.730126582278481,
"grad_norm": 0.4217883852826991,
"learning_rate": 2.014189549672245e-05,
"loss": 0.9706,
"step": 674
},
{
"epoch": 2.7341772151898733,
"grad_norm": 0.45172608489339805,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.9492,
"step": 675
},
{
"epoch": 2.738227848101266,
"grad_norm": 0.4930990637539112,
"learning_rate": 2.0028379327907327e-05,
"loss": 1.001,
"step": 676
},
{
"epoch": 2.7422784810126584,
"grad_norm": 0.48830318290343944,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.9875,
"step": 677
},
{
"epoch": 2.746329113924051,
"grad_norm": 0.44610136099263814,
"learning_rate": 1.991486224484123e-05,
"loss": 0.9953,
"step": 678
},
{
"epoch": 2.750379746835443,
"grad_norm": 0.4307476315241404,
"learning_rate": 1.985810450327756e-05,
"loss": 0.9564,
"step": 679
},
{
"epoch": 2.7544303797468355,
"grad_norm": 0.4925846530742079,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.9934,
"step": 680
},
{
"epoch": 2.758481012658228,
"grad_norm": 0.4323635778577378,
"learning_rate": 1.974459290568053e-05,
"loss": 0.9801,
"step": 681
},
{
"epoch": 2.76253164556962,
"grad_norm": 0.4065914161003242,
"learning_rate": 1.968783996385397e-05,
"loss": 0.9892,
"step": 682
},
{
"epoch": 2.7665822784810126,
"grad_norm": 0.42365103132273824,
"learning_rate": 1.963108953612143e-05,
"loss": 0.9699,
"step": 683
},
{
"epoch": 2.770632911392405,
"grad_norm": 0.39862697523303786,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.9772,
"step": 684
},
{
"epoch": 2.7746835443037976,
"grad_norm": 0.39001797964684043,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.9733,
"step": 685
},
{
"epoch": 2.77873417721519,
"grad_norm": 0.42681017097181867,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.9564,
"step": 686
},
{
"epoch": 2.782784810126582,
"grad_norm": 0.4094159570199675,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.9504,
"step": 687
},
{
"epoch": 2.7868354430379747,
"grad_norm": 0.45327410080266933,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.9634,
"step": 688
},
{
"epoch": 2.790886075949367,
"grad_norm": 0.40036775628923693,
"learning_rate": 1.92906653591874e-05,
"loss": 0.979,
"step": 689
},
{
"epoch": 2.7949367088607593,
"grad_norm": 0.4353720600175797,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.9802,
"step": 690
},
{
"epoch": 2.798987341772152,
"grad_norm": 0.442540874522496,
"learning_rate": 1.917723146291022e-05,
"loss": 0.9762,
"step": 691
},
{
"epoch": 2.8030379746835443,
"grad_norm": 0.4185556429992911,
"learning_rate": 1.912052422608584e-05,
"loss": 0.9739,
"step": 692
},
{
"epoch": 2.807088607594937,
"grad_norm": 0.41941806579486485,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.968,
"step": 693
},
{
"epoch": 2.8111392405063294,
"grad_norm": 0.4183562995265651,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.9653,
"step": 694
},
{
"epoch": 2.8151898734177214,
"grad_norm": 0.453683909874934,
"learning_rate": 1.895044684123479e-05,
"loss": 0.9945,
"step": 695
},
{
"epoch": 2.819240506329114,
"grad_norm": 0.4622198594941008,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.9758,
"step": 696
},
{
"epoch": 2.8232911392405065,
"grad_norm": 0.4711540795249115,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.9811,
"step": 697
},
{
"epoch": 2.8273417721518985,
"grad_norm": 0.4306777800800549,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.9656,
"step": 698
},
{
"epoch": 2.831392405063291,
"grad_norm": 0.44793532807776654,
"learning_rate": 1.872379746550743e-05,
"loss": 0.9831,
"step": 699
},
{
"epoch": 2.8354430379746836,
"grad_norm": 0.4413889719173009,
"learning_rate": 1.866715967678769e-05,
"loss": 0.9853,
"step": 700
},
{
"epoch": 2.839493670886076,
"grad_norm": 0.41815242386096624,
"learning_rate": 1.861053262258067e-05,
"loss": 0.9771,
"step": 701
},
{
"epoch": 2.8435443037974686,
"grad_norm": 0.38310637673668196,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.9681,
"step": 702
},
{
"epoch": 2.8475949367088607,
"grad_norm": 0.4079229705668711,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.9866,
"step": 703
},
{
"epoch": 2.851645569620253,
"grad_norm": 0.42535560742701073,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.982,
"step": 704
},
{
"epoch": 2.8556962025316457,
"grad_norm": 0.39511618846488955,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.9577,
"step": 705
},
{
"epoch": 2.8597468354430378,
"grad_norm": 0.4115797622672065,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.9753,
"step": 706
},
{
"epoch": 2.8637974683544303,
"grad_norm": 0.37212764259055475,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.9476,
"step": 707
},
{
"epoch": 2.867848101265823,
"grad_norm": 0.40650234160298204,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.9689,
"step": 708
},
{
"epoch": 2.8718987341772153,
"grad_norm": 0.41213176697386483,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.9397,
"step": 709
},
{
"epoch": 2.875949367088608,
"grad_norm": 0.4128979442259312,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.9622,
"step": 710
},
{
"epoch": 2.88,
"grad_norm": 0.41639244869972536,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.9704,
"step": 711
},
{
"epoch": 2.8840506329113924,
"grad_norm": 0.3820336745790507,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.9479,
"step": 712
},
{
"epoch": 2.888101265822785,
"grad_norm": 0.41981457225400665,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.9667,
"step": 713
},
{
"epoch": 2.892151898734177,
"grad_norm": 0.5491019992774512,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.9659,
"step": 714
},
{
"epoch": 2.8962025316455695,
"grad_norm": 0.44527491203776365,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.9479,
"step": 715
},
{
"epoch": 2.900253164556962,
"grad_norm": 0.4180535052090943,
"learning_rate": 1.776272483901182e-05,
"loss": 0.9572,
"step": 716
},
{
"epoch": 2.9043037974683545,
"grad_norm": 0.5322380411097211,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.9713,
"step": 717
},
{
"epoch": 2.908354430379747,
"grad_norm": 0.41525584176338204,
"learning_rate": 1.764995661794301e-05,
"loss": 0.9798,
"step": 718
},
{
"epoch": 2.912405063291139,
"grad_norm": 0.5039715332954958,
"learning_rate": 1.759360067078035e-05,
"loss": 0.933,
"step": 719
},
{
"epoch": 2.9164556962025316,
"grad_norm": 0.42647554194380194,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.9976,
"step": 720
},
{
"epoch": 2.920506329113924,
"grad_norm": 0.5557188393500095,
"learning_rate": 1.748094737261012e-05,
"loss": 0.9782,
"step": 721
},
{
"epoch": 2.9245569620253162,
"grad_norm": 0.3715756040132988,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.9925,
"step": 722
},
{
"epoch": 2.9286075949367087,
"grad_norm": 0.4618861575972924,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.9715,
"step": 723
},
{
"epoch": 2.9326582278481013,
"grad_norm": 0.49268611878425317,
"learning_rate": 1.731212071922883e-05,
"loss": 0.9719,
"step": 724
},
{
"epoch": 2.9367088607594938,
"grad_norm": 0.42077248647798987,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.9761,
"step": 725
},
{
"epoch": 2.9407594936708863,
"grad_norm": 0.4201617518218664,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.978,
"step": 726
},
{
"epoch": 2.9448101265822784,
"grad_norm": 0.4062606036297453,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.9605,
"step": 727
},
{
"epoch": 2.948860759493671,
"grad_norm": 0.4250105482134818,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.9382,
"step": 728
},
{
"epoch": 2.9529113924050634,
"grad_norm": 0.4113279357224978,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.957,
"step": 729
},
{
"epoch": 2.9569620253164555,
"grad_norm": 0.460541611684496,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.9699,
"step": 730
},
{
"epoch": 2.961012658227848,
"grad_norm": 0.39807798821960616,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.9715,
"step": 731
},
{
"epoch": 2.9650632911392405,
"grad_norm": 0.432689379939096,
"learning_rate": 1.686290200071915e-05,
"loss": 0.9728,
"step": 732
},
{
"epoch": 2.969113924050633,
"grad_norm": 0.47962303258696043,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.9562,
"step": 733
},
{
"epoch": 2.9731645569620255,
"grad_norm": 0.3965598910596417,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.9655,
"step": 734
},
{
"epoch": 2.9772151898734176,
"grad_norm": 0.4443400767453623,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.968,
"step": 735
},
{
"epoch": 2.98126582278481,
"grad_norm": 0.43023973631057927,
"learning_rate": 1.663888455395778e-05,
"loss": 0.9606,
"step": 736
},
{
"epoch": 2.9853164556962026,
"grad_norm": 0.42914204372699066,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.9399,
"step": 737
},
{
"epoch": 2.9893670886075947,
"grad_norm": 0.39770931735694787,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.9676,
"step": 738
},
{
"epoch": 2.993417721518987,
"grad_norm": 0.5011571875361633,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.9538,
"step": 739
},
{
"epoch": 2.9974683544303797,
"grad_norm": 0.41605003237183164,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.9604,
"step": 740
},
{
"epoch": 3.0015189873417722,
"grad_norm": 0.4852173543047961,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.9543,
"step": 741
},
{
"epoch": 3.0055696202531648,
"grad_norm": 0.6098647735753382,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.8547,
"step": 742
},
{
"epoch": 3.009620253164557,
"grad_norm": 0.9998362378893455,
"learning_rate": 1.624791353909428e-05,
"loss": 0.8415,
"step": 743
},
{
"epoch": 3.0136708860759494,
"grad_norm": 0.5133083118210368,
"learning_rate": 1.619217781670663e-05,
"loss": 0.8524,
"step": 744
},
{
"epoch": 3.017721518987342,
"grad_norm": 0.7257540523313422,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.8384,
"step": 745
},
{
"epoch": 3.0217721518987344,
"grad_norm": 0.6699594958007665,
"learning_rate": 1.608079882360134e-05,
"loss": 0.8299,
"step": 746
},
{
"epoch": 3.0258227848101265,
"grad_norm": 0.6293353512438571,
"learning_rate": 1.60251564499148e-05,
"loss": 0.8454,
"step": 747
},
{
"epoch": 3.029873417721519,
"grad_norm": 0.5576558817940911,
"learning_rate": 1.596954608907176e-05,
"loss": 0.8115,
"step": 748
},
{
"epoch": 3.0339240506329115,
"grad_norm": 0.6926279444498484,
"learning_rate": 1.591396818895043e-05,
"loss": 0.8372,
"step": 749
},
{
"epoch": 3.037974683544304,
"grad_norm": 0.6296537753653931,
"learning_rate": 1.585842319716759e-05,
"loss": 0.823,
"step": 750
},
{
"epoch": 3.042025316455696,
"grad_norm": 0.6976665069698015,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.8182,
"step": 751
},
{
"epoch": 3.0460759493670886,
"grad_norm": 0.6630740566068996,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.8427,
"step": 752
},
{
"epoch": 3.050126582278481,
"grad_norm": 0.6600274739463632,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.823,
"step": 753
},
{
"epoch": 3.0541772151898736,
"grad_norm": 0.5367312446114819,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.8114,
"step": 754
},
{
"epoch": 3.0582278481012657,
"grad_norm": 0.6707391893847942,
"learning_rate": 1.558120751116291e-05,
"loss": 0.8113,
"step": 755
},
{
"epoch": 3.062278481012658,
"grad_norm": 0.5180606137237209,
"learning_rate": 1.552586935426876e-05,
"loss": 0.827,
"step": 756
},
{
"epoch": 3.0663291139240507,
"grad_norm": 0.610667876086823,
"learning_rate": 1.547056723140774e-05,
"loss": 0.8509,
"step": 757
},
{
"epoch": 3.0703797468354432,
"grad_norm": 0.5797436630926927,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.8404,
"step": 758
},
{
"epoch": 3.0744303797468353,
"grad_norm": 0.5628979593060021,
"learning_rate": 1.536007286907411e-05,
"loss": 0.8395,
"step": 759
},
{
"epoch": 3.078481012658228,
"grad_norm": 0.6130635794708162,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.8173,
"step": 760
},
{
"epoch": 3.0825316455696203,
"grad_norm": 0.5094031201175856,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.8392,
"step": 761
},
{
"epoch": 3.086582278481013,
"grad_norm": 0.6162708717571006,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.8462,
"step": 762
},
{
"epoch": 3.090632911392405,
"grad_norm": 0.5005468776707757,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.8249,
"step": 763
},
{
"epoch": 3.0946835443037974,
"grad_norm": 0.5577164143097275,
"learning_rate": 1.508449870007656e-05,
"loss": 0.8338,
"step": 764
},
{
"epoch": 3.09873417721519,
"grad_norm": 0.4379287359141932,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.8283,
"step": 765
},
{
"epoch": 3.1027848101265825,
"grad_norm": 0.554934297976691,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.8443,
"step": 766
},
{
"epoch": 3.1068354430379745,
"grad_norm": 0.45392551407834164,
"learning_rate": 1.4919625713405e-05,
"loss": 0.8121,
"step": 767
},
{
"epoch": 3.110886075949367,
"grad_norm": 0.47500051055523,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.8223,
"step": 768
},
{
"epoch": 3.1149367088607596,
"grad_norm": 0.44998481311665706,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.8523,
"step": 769
},
{
"epoch": 3.118987341772152,
"grad_norm": 0.4454977076171069,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.8471,
"step": 770
},
{
"epoch": 3.123037974683544,
"grad_norm": 0.48583545815345136,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.8427,
"step": 771
},
{
"epoch": 3.1270886075949367,
"grad_norm": 0.43971815582837814,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.8358,
"step": 772
},
{
"epoch": 3.131139240506329,
"grad_norm": 0.4704299149127121,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.8328,
"step": 773
},
{
"epoch": 3.1351898734177217,
"grad_norm": 0.4102783997465394,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.8069,
"step": 774
},
{
"epoch": 3.1392405063291138,
"grad_norm": 0.42694989987152215,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.8163,
"step": 775
},
{
"epoch": 3.1432911392405063,
"grad_norm": 0.45284828227279356,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.8409,
"step": 776
},
{
"epoch": 3.147341772151899,
"grad_norm": 0.45140687224397874,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.8324,
"step": 777
},
{
"epoch": 3.1513924050632913,
"grad_norm": 0.4426227025822779,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.8448,
"step": 778
},
{
"epoch": 3.1554430379746834,
"grad_norm": 0.4637602364733529,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.8396,
"step": 779
},
{
"epoch": 3.159493670886076,
"grad_norm": 0.4362058027102316,
"learning_rate": 1.420958430225303e-05,
"loss": 0.8446,
"step": 780
},
{
"epoch": 3.1635443037974684,
"grad_norm": 0.4027702760652104,
"learning_rate": 1.415527989393985e-05,
"loss": 0.843,
"step": 781
},
{
"epoch": 3.167594936708861,
"grad_norm": 0.4431697238511214,
"learning_rate": 1.410102255819891e-05,
"loss": 0.8274,
"step": 782
},
{
"epoch": 3.171645569620253,
"grad_norm": 0.3972875255567799,
"learning_rate": 1.404681273201131e-05,
"loss": 0.8197,
"step": 783
},
{
"epoch": 3.1756962025316455,
"grad_norm": 0.4114042486953834,
"learning_rate": 1.399265085197556e-05,
"loss": 0.8367,
"step": 784
},
{
"epoch": 3.179746835443038,
"grad_norm": 0.41675480795124686,
"learning_rate": 1.393853735430398e-05,
"loss": 0.8178,
"step": 785
},
{
"epoch": 3.1837974683544306,
"grad_norm": 0.43763219561068745,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.8309,
"step": 786
},
{
"epoch": 3.1878481012658226,
"grad_norm": 0.4497905515686786,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.8511,
"step": 787
},
{
"epoch": 3.191898734177215,
"grad_norm": 0.44142128320248647,
"learning_rate": 1.377649151173163e-05,
"loss": 0.8273,
"step": 788
},
{
"epoch": 3.1959493670886077,
"grad_norm": 0.44215952802647784,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.8421,
"step": 789
},
{
"epoch": 3.2,
"grad_norm": 0.4153554942008414,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.8024,
"step": 790
},
{
"epoch": 3.2040506329113922,
"grad_norm": 0.4471829527911052,
"learning_rate": 1.361489677627324e-05,
"loss": 0.8352,
"step": 791
},
{
"epoch": 3.2081012658227848,
"grad_norm": 0.4487238457445614,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.8246,
"step": 792
},
{
"epoch": 3.2121518987341773,
"grad_norm": 0.4183645587880581,
"learning_rate": 1.350742335331241e-05,
"loss": 0.8442,
"step": 793
},
{
"epoch": 3.21620253164557,
"grad_norm": 0.43365504212840883,
"learning_rate": 1.345376486102198e-05,
"loss": 0.8428,
"step": 794
},
{
"epoch": 3.220253164556962,
"grad_norm": 0.41481790494575743,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.8062,
"step": 795
},
{
"epoch": 3.2243037974683544,
"grad_norm": 0.4721083284435898,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.8329,
"step": 796
},
{
"epoch": 3.228354430379747,
"grad_norm": 0.4491498337044083,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.86,
"step": 797
},
{
"epoch": 3.2324050632911394,
"grad_norm": 0.43913787857823344,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.8375,
"step": 798
},
{
"epoch": 3.2364556962025315,
"grad_norm": 0.4992056541216991,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.829,
"step": 799
},
{
"epoch": 3.240506329113924,
"grad_norm": 0.4340975768970369,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.8393,
"step": 800
},
{
"epoch": 3.2445569620253165,
"grad_norm": 0.46429244126501495,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.8409,
"step": 801
},
{
"epoch": 3.248607594936709,
"grad_norm": 0.4452295783549923,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.8376,
"step": 802
},
{
"epoch": 3.252658227848101,
"grad_norm": 0.4888091549528938,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.8595,
"step": 803
},
{
"epoch": 3.2567088607594936,
"grad_norm": 0.4271324296438308,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.8455,
"step": 804
},
{
"epoch": 3.260759493670886,
"grad_norm": 0.4545673019833688,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.8474,
"step": 805
},
{
"epoch": 3.2648101265822786,
"grad_norm": 0.4505089485906897,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.8212,
"step": 806
},
{
"epoch": 3.2688607594936707,
"grad_norm": 0.44668366949066,
"learning_rate": 1.276115892225764e-05,
"loss": 0.8362,
"step": 807
},
{
"epoch": 3.2729113924050632,
"grad_norm": 0.42825903310209396,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.8391,
"step": 808
},
{
"epoch": 3.2769620253164558,
"grad_norm": 0.4301623096798999,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.8251,
"step": 809
},
{
"epoch": 3.2810126582278483,
"grad_norm": 0.5098550186405754,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.8437,
"step": 810
},
{
"epoch": 3.2850632911392403,
"grad_norm": 0.43302094345102304,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.8331,
"step": 811
},
{
"epoch": 3.289113924050633,
"grad_norm": 0.4343791576132872,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.8269,
"step": 812
},
{
"epoch": 3.2931645569620254,
"grad_norm": 0.46766154889267636,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.8471,
"step": 813
},
{
"epoch": 3.297215189873418,
"grad_norm": 0.42778279165023864,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.8476,
"step": 814
},
{
"epoch": 3.30126582278481,
"grad_norm": 0.49107378448750116,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.8209,
"step": 815
},
{
"epoch": 3.3053164556962025,
"grad_norm": 0.4479261571742715,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.8333,
"step": 816
},
{
"epoch": 3.309367088607595,
"grad_norm": 0.4789105875597741,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.8571,
"step": 817
},
{
"epoch": 3.3134177215189875,
"grad_norm": 0.5329721286986038,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.8523,
"step": 818
},
{
"epoch": 3.3174683544303796,
"grad_norm": 0.46259433414846185,
"learning_rate": 1.213055323844297e-05,
"loss": 0.8095,
"step": 819
},
{
"epoch": 3.321518987341772,
"grad_norm": 0.5259114584572695,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.8397,
"step": 820
},
{
"epoch": 3.3255696202531646,
"grad_norm": 0.4836172265148426,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.8383,
"step": 821
},
{
"epoch": 3.329620253164557,
"grad_norm": 0.46757896114685804,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.8454,
"step": 822
},
{
"epoch": 3.333670886075949,
"grad_norm": 0.5043648345740435,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.8208,
"step": 823
},
{
"epoch": 3.3377215189873417,
"grad_norm": 0.4119046145138724,
"learning_rate": 1.187045262895488e-05,
"loss": 0.8424,
"step": 824
},
{
"epoch": 3.3417721518987342,
"grad_norm": 0.44817024666942895,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.8431,
"step": 825
},
{
"epoch": 3.3458227848101267,
"grad_norm": 0.4855028783722228,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.8495,
"step": 826
},
{
"epoch": 3.349873417721519,
"grad_norm": 0.45892791391414056,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.8552,
"step": 827
},
{
"epoch": 3.3539240506329113,
"grad_norm": 0.4963438792477058,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.8759,
"step": 828
},
{
"epoch": 3.357974683544304,
"grad_norm": 0.46419441894807184,
"learning_rate": 1.161198884952377e-05,
"loss": 0.8325,
"step": 829
},
{
"epoch": 3.3620253164556964,
"grad_norm": 0.4935981750891952,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.8347,
"step": 830
},
{
"epoch": 3.3660759493670884,
"grad_norm": 0.4176566292455059,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.8269,
"step": 831
},
{
"epoch": 3.370126582278481,
"grad_norm": 0.4915632996320811,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.8398,
"step": 832
},
{
"epoch": 3.3741772151898735,
"grad_norm": 0.4101370811182063,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.8144,
"step": 833
},
{
"epoch": 3.378227848101266,
"grad_norm": 0.42701773988821434,
"learning_rate": 1.135521394010506e-05,
"loss": 0.8262,
"step": 834
},
{
"epoch": 3.382278481012658,
"grad_norm": 0.4448767211298008,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.8301,
"step": 835
},
{
"epoch": 3.3863291139240506,
"grad_norm": 0.3980739925778004,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.8297,
"step": 836
},
{
"epoch": 3.390379746835443,
"grad_norm": 0.40616790288021926,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.8304,
"step": 837
},
{
"epoch": 3.3944303797468356,
"grad_norm": 0.4042366157996179,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.8438,
"step": 838
},
{
"epoch": 3.3984810126582277,
"grad_norm": 0.4233769395879206,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.8415,
"step": 839
},
{
"epoch": 3.40253164556962,
"grad_norm": 0.4120086598375207,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.8356,
"step": 840
},
{
"epoch": 3.4065822784810127,
"grad_norm": 0.43267875276824536,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.8309,
"step": 841
},
{
"epoch": 3.410632911392405,
"grad_norm": 0.4257364512482632,
"learning_rate": 1.0948015846744e-05,
"loss": 0.832,
"step": 842
},
{
"epoch": 3.4146835443037973,
"grad_norm": 0.4179075053173085,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.8324,
"step": 843
},
{
"epoch": 3.41873417721519,
"grad_norm": 0.43931632107595153,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.8297,
"step": 844
},
{
"epoch": 3.4227848101265823,
"grad_norm": 0.42591232992295625,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.8521,
"step": 845
},
{
"epoch": 3.426835443037975,
"grad_norm": 0.4343870298142052,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.8176,
"step": 846
},
{
"epoch": 3.430886075949367,
"grad_norm": 0.4062255379942965,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.8266,
"step": 847
},
{
"epoch": 3.4349367088607594,
"grad_norm": 0.43821300082805736,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.8263,
"step": 848
},
{
"epoch": 3.438987341772152,
"grad_norm": 0.4236140464142908,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.8201,
"step": 849
},
{
"epoch": 3.4430379746835444,
"grad_norm": 0.45064888957194515,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.8504,
"step": 850
},
{
"epoch": 3.4470886075949365,
"grad_norm": 0.40791242302710695,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.8291,
"step": 851
},
{
"epoch": 3.451139240506329,
"grad_norm": 0.42526583244574795,
"learning_rate": 1.044560362975474e-05,
"loss": 0.8439,
"step": 852
},
{
"epoch": 3.4551898734177215,
"grad_norm": 0.4527216557279818,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.8273,
"step": 853
},
{
"epoch": 3.459240506329114,
"grad_norm": 0.40144077096837627,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.8209,
"step": 854
},
{
"epoch": 3.463291139240506,
"grad_norm": 0.43806488944231187,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.8717,
"step": 855
},
{
"epoch": 3.4673417721518986,
"grad_norm": 0.4219689255809355,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.8585,
"step": 856
},
{
"epoch": 3.471392405063291,
"grad_norm": 0.4136218588562535,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.8182,
"step": 857
},
{
"epoch": 3.4754430379746837,
"grad_norm": 0.42825412649947314,
"learning_rate": 1.014782400334433e-05,
"loss": 0.8268,
"step": 858
},
{
"epoch": 3.479493670886076,
"grad_norm": 0.4223623375130303,
"learning_rate": 1.009846945771296e-05,
"loss": 0.8578,
"step": 859
},
{
"epoch": 3.4835443037974683,
"grad_norm": 0.42328305058220883,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.8369,
"step": 860
},
{
"epoch": 3.487594936708861,
"grad_norm": 0.41878451826527646,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.8249,
"step": 861
},
{
"epoch": 3.4916455696202533,
"grad_norm": 0.4384021428747495,
"learning_rate": 9.950885880977891e-06,
"loss": 0.8337,
"step": 862
},
{
"epoch": 3.4956962025316454,
"grad_norm": 0.39665779589894257,
"learning_rate": 9.901852696139382e-06,
"loss": 0.8303,
"step": 863
},
{
"epoch": 3.499746835443038,
"grad_norm": 0.44276905707293535,
"learning_rate": 9.852900840391027e-06,
"loss": 0.8347,
"step": 864
},
{
"epoch": 3.5037974683544304,
"grad_norm": 0.4085380943892777,
"learning_rate": 9.804030707984313e-06,
"loss": 0.8134,
"step": 865
},
{
"epoch": 3.507848101265823,
"grad_norm": 0.4461163065929184,
"learning_rate": 9.755242692512599e-06,
"loss": 0.834,
"step": 866
},
{
"epoch": 3.5118987341772154,
"grad_norm": 0.41989574217104286,
"learning_rate": 9.70653718690782e-06,
"loss": 0.8356,
"step": 867
},
{
"epoch": 3.5159493670886075,
"grad_norm": 0.454558505269666,
"learning_rate": 9.657914583437454e-06,
"loss": 0.8503,
"step": 868
},
{
"epoch": 3.52,
"grad_norm": 0.4080071982516211,
"learning_rate": 9.609375273701246e-06,
"loss": 0.8452,
"step": 869
},
{
"epoch": 3.5240506329113925,
"grad_norm": 0.444880547512302,
"learning_rate": 9.560919648628133e-06,
"loss": 0.8505,
"step": 870
},
{
"epoch": 3.5281012658227846,
"grad_norm": 0.4308896460657528,
"learning_rate": 9.512548098473047e-06,
"loss": 0.8355,
"step": 871
},
{
"epoch": 3.532151898734177,
"grad_norm": 0.4277339989458913,
"learning_rate": 9.464261012813825e-06,
"loss": 0.8335,
"step": 872
},
{
"epoch": 3.5362025316455696,
"grad_norm": 0.44753054164552986,
"learning_rate": 9.416058780547987e-06,
"loss": 0.8388,
"step": 873
},
{
"epoch": 3.540253164556962,
"grad_norm": 0.41451432124843707,
"learning_rate": 9.367941789889714e-06,
"loss": 0.8391,
"step": 874
},
{
"epoch": 3.5443037974683547,
"grad_norm": 0.42818914869295777,
"learning_rate": 9.319910428366607e-06,
"loss": 0.8269,
"step": 875
},
{
"epoch": 3.5483544303797467,
"grad_norm": 0.4246002921520773,
"learning_rate": 9.271965082816667e-06,
"loss": 0.8256,
"step": 876
},
{
"epoch": 3.5524050632911393,
"grad_norm": 0.42892013867414236,
"learning_rate": 9.224106139385111e-06,
"loss": 0.8481,
"step": 877
},
{
"epoch": 3.5564556962025318,
"grad_norm": 0.437575894394203,
"learning_rate": 9.176333983521291e-06,
"loss": 0.8532,
"step": 878
},
{
"epoch": 3.560506329113924,
"grad_norm": 0.42343656551437714,
"learning_rate": 9.12864899997558e-06,
"loss": 0.8265,
"step": 879
},
{
"epoch": 3.5645569620253164,
"grad_norm": 0.4441463962830876,
"learning_rate": 9.08105157279628e-06,
"loss": 0.8403,
"step": 880
},
{
"epoch": 3.568607594936709,
"grad_norm": 0.41190349438077656,
"learning_rate": 9.03354208532653e-06,
"loss": 0.8345,
"step": 881
},
{
"epoch": 3.5726582278481014,
"grad_norm": 0.46968193732785457,
"learning_rate": 8.986120920201205e-06,
"loss": 0.8523,
"step": 882
},
{
"epoch": 3.576708860759494,
"grad_norm": 0.41339432264752046,
"learning_rate": 8.938788459343852e-06,
"loss": 0.8227,
"step": 883
},
{
"epoch": 3.580759493670886,
"grad_norm": 0.4097513290419891,
"learning_rate": 8.8915450839636e-06,
"loss": 0.8512,
"step": 884
},
{
"epoch": 3.5848101265822785,
"grad_norm": 0.41284430563960584,
"learning_rate": 8.844391174552116e-06,
"loss": 0.8241,
"step": 885
},
{
"epoch": 3.588860759493671,
"grad_norm": 0.4183879351127537,
"learning_rate": 8.797327110880479e-06,
"loss": 0.8386,
"step": 886
},
{
"epoch": 3.592911392405063,
"grad_norm": 0.4459224547882989,
"learning_rate": 8.750353271996206e-06,
"loss": 0.8307,
"step": 887
},
{
"epoch": 3.5969620253164556,
"grad_norm": 0.42210959838507844,
"learning_rate": 8.703470036220132e-06,
"loss": 0.8403,
"step": 888
},
{
"epoch": 3.601012658227848,
"grad_norm": 0.44935874890227134,
"learning_rate": 8.656677781143394e-06,
"loss": 0.8556,
"step": 889
},
{
"epoch": 3.6050632911392406,
"grad_norm": 0.3944484656931401,
"learning_rate": 8.609976883624377e-06,
"loss": 0.822,
"step": 890
},
{
"epoch": 3.609113924050633,
"grad_norm": 0.4523550989635727,
"learning_rate": 8.563367719785698e-06,
"loss": 0.8423,
"step": 891
},
{
"epoch": 3.613164556962025,
"grad_norm": 0.41739339318890945,
"learning_rate": 8.516850665011138e-06,
"loss": 0.8129,
"step": 892
},
{
"epoch": 3.6172151898734177,
"grad_norm": 0.42338577406063665,
"learning_rate": 8.47042609394269e-06,
"loss": 0.8337,
"step": 893
},
{
"epoch": 3.6212658227848102,
"grad_norm": 0.41369817234007034,
"learning_rate": 8.424094380477432e-06,
"loss": 0.8311,
"step": 894
},
{
"epoch": 3.6253164556962023,
"grad_norm": 0.41559226753446804,
"learning_rate": 8.37785589776465e-06,
"loss": 0.8173,
"step": 895
},
{
"epoch": 3.629367088607595,
"grad_norm": 0.4103339797445647,
"learning_rate": 8.331711018202694e-06,
"loss": 0.8493,
"step": 896
},
{
"epoch": 3.6334177215189873,
"grad_norm": 0.4440492017820311,
"learning_rate": 8.285660113436104e-06,
"loss": 0.8288,
"step": 897
},
{
"epoch": 3.63746835443038,
"grad_norm": 0.4160848215360993,
"learning_rate": 8.239703554352527e-06,
"loss": 0.813,
"step": 898
},
{
"epoch": 3.6415189873417724,
"grad_norm": 0.39268872315510756,
"learning_rate": 8.193841711079775e-06,
"loss": 0.841,
"step": 899
},
{
"epoch": 3.6455696202531644,
"grad_norm": 0.41167854859638964,
"learning_rate": 8.148074952982828e-06,
"loss": 0.8386,
"step": 900
},
{
"epoch": 3.649620253164557,
"grad_norm": 0.43302960056043327,
"learning_rate": 8.102403648660859e-06,
"loss": 0.8323,
"step": 901
},
{
"epoch": 3.6536708860759495,
"grad_norm": 0.44322743578062757,
"learning_rate": 8.056828165944282e-06,
"loss": 0.8355,
"step": 902
},
{
"epoch": 3.6577215189873415,
"grad_norm": 0.436011960041533,
"learning_rate": 8.011348871891762e-06,
"loss": 0.8498,
"step": 903
},
{
"epoch": 3.661772151898734,
"grad_norm": 0.4465107899027855,
"learning_rate": 7.965966132787287e-06,
"loss": 0.8446,
"step": 904
},
{
"epoch": 3.6658227848101266,
"grad_norm": 0.43281950322991647,
"learning_rate": 7.920680314137189e-06,
"loss": 0.8314,
"step": 905
},
{
"epoch": 3.669873417721519,
"grad_norm": 0.3894591854584584,
"learning_rate": 7.875491780667246e-06,
"loss": 0.8198,
"step": 906
},
{
"epoch": 3.6739240506329116,
"grad_norm": 0.4514088951817415,
"learning_rate": 7.830400896319667e-06,
"loss": 0.8242,
"step": 907
},
{
"epoch": 3.6779746835443037,
"grad_norm": 0.3979724144737252,
"learning_rate": 7.785408024250259e-06,
"loss": 0.8437,
"step": 908
},
{
"epoch": 3.682025316455696,
"grad_norm": 0.403644523467473,
"learning_rate": 7.74051352682542e-06,
"loss": 0.8549,
"step": 909
},
{
"epoch": 3.6860759493670887,
"grad_norm": 0.41193809308334384,
"learning_rate": 7.695717765619257e-06,
"loss": 0.8274,
"step": 910
},
{
"epoch": 3.690126582278481,
"grad_norm": 0.44426955035236126,
"learning_rate": 7.651021101410673e-06,
"loss": 0.8375,
"step": 911
},
{
"epoch": 3.6941772151898733,
"grad_norm": 0.40738068207001127,
"learning_rate": 7.606423894180464e-06,
"loss": 0.8325,
"step": 912
},
{
"epoch": 3.698227848101266,
"grad_norm": 0.3932634817161196,
"learning_rate": 7.56192650310839e-06,
"loss": 0.8475,
"step": 913
},
{
"epoch": 3.7022784810126583,
"grad_norm": 0.3955871073925122,
"learning_rate": 7.517529286570349e-06,
"loss": 0.8426,
"step": 914
},
{
"epoch": 3.706329113924051,
"grad_norm": 0.4263092260778542,
"learning_rate": 7.473232602135387e-06,
"loss": 0.831,
"step": 915
},
{
"epoch": 3.710379746835443,
"grad_norm": 0.41738494887604755,
"learning_rate": 7.429036806562935e-06,
"loss": 0.8174,
"step": 916
},
{
"epoch": 3.7144303797468354,
"grad_norm": 0.41568736380258414,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.8276,
"step": 917
},
{
"epoch": 3.718481012658228,
"grad_norm": 0.4035693712349663,
"learning_rate": 7.340949304977567e-06,
"loss": 0.8216,
"step": 918
},
{
"epoch": 3.72253164556962,
"grad_norm": 0.39807525416204215,
"learning_rate": 7.297058308409282e-06,
"loss": 0.8347,
"step": 919
},
{
"epoch": 3.7265822784810125,
"grad_norm": 0.3865436222347132,
"learning_rate": 7.25326961958704e-06,
"loss": 0.8468,
"step": 920
},
{
"epoch": 3.730632911392405,
"grad_norm": 0.41381961120363125,
"learning_rate": 7.209583591178921e-06,
"loss": 0.8335,
"step": 921
},
{
"epoch": 3.7346835443037976,
"grad_norm": 0.4215413327950253,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.8347,
"step": 922
},
{
"epoch": 3.73873417721519,
"grad_norm": 0.4003481025223345,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.8418,
"step": 923
},
{
"epoch": 3.742784810126582,
"grad_norm": 0.4152373357353929,
"learning_rate": 7.079144982700909e-06,
"loss": 0.861,
"step": 924
},
{
"epoch": 3.7468354430379747,
"grad_norm": 0.39581908919188236,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.8392,
"step": 925
},
{
"epoch": 3.750886075949367,
"grad_norm": 0.430673747797509,
"learning_rate": 6.99270564069757e-06,
"loss": 0.8295,
"step": 926
},
{
"epoch": 3.7549367088607593,
"grad_norm": 0.43301732646989083,
"learning_rate": 6.949642934304375e-06,
"loss": 0.8109,
"step": 927
},
{
"epoch": 3.7589873417721518,
"grad_norm": 0.4009990437579464,
"learning_rate": 6.906685333692871e-06,
"loss": 0.8302,
"step": 928
},
{
"epoch": 3.7630379746835443,
"grad_norm": 0.3966171099603616,
"learning_rate": 6.86383318483769e-06,
"loss": 0.8428,
"step": 929
},
{
"epoch": 3.767088607594937,
"grad_norm": 0.39220498811098414,
"learning_rate": 6.821086832864139e-06,
"loss": 0.8556,
"step": 930
},
{
"epoch": 3.7711392405063293,
"grad_norm": 0.39564256177691987,
"learning_rate": 6.77844662204546e-06,
"loss": 0.8414,
"step": 931
},
{
"epoch": 3.7751898734177214,
"grad_norm": 0.37886228257575677,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.8303,
"step": 932
},
{
"epoch": 3.779240506329114,
"grad_norm": 0.40620617718274743,
"learning_rate": 6.693485996688695e-06,
"loss": 0.8362,
"step": 933
},
{
"epoch": 3.7832911392405064,
"grad_norm": 0.40902102542521,
"learning_rate": 6.651166266411801e-06,
"loss": 0.8295,
"step": 934
},
{
"epoch": 3.7873417721518985,
"grad_norm": 0.3867055900729841,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.8195,
"step": 935
},
{
"epoch": 3.791392405063291,
"grad_norm": 0.40306616137174706,
"learning_rate": 6.566849674844711e-06,
"loss": 0.838,
"step": 936
},
{
"epoch": 3.7954430379746835,
"grad_norm": 0.41574247116543633,
"learning_rate": 6.524853492628747e-06,
"loss": 0.8188,
"step": 937
},
{
"epoch": 3.799493670886076,
"grad_norm": 0.4172623250923879,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.8508,
"step": 938
},
{
"epoch": 3.8035443037974686,
"grad_norm": 0.4076163774092128,
"learning_rate": 6.441187046486648e-06,
"loss": 0.8523,
"step": 939
},
{
"epoch": 3.8075949367088606,
"grad_norm": 0.4275838965251661,
"learning_rate": 6.399517456398567e-06,
"loss": 0.8028,
"step": 940
},
{
"epoch": 3.811645569620253,
"grad_norm": 0.4042242099902195,
"learning_rate": 6.357957402727164e-06,
"loss": 0.816,
"step": 941
},
{
"epoch": 3.8156962025316457,
"grad_norm": 0.4163901804931075,
"learning_rate": 6.316507220191395e-06,
"loss": 0.8433,
"step": 942
},
{
"epoch": 3.8197468354430377,
"grad_norm": 0.3946602042096483,
"learning_rate": 6.275167242625331e-06,
"loss": 0.8203,
"step": 943
},
{
"epoch": 3.8237974683544302,
"grad_norm": 0.39235265139732123,
"learning_rate": 6.233937802975471e-06,
"loss": 0.833,
"step": 944
},
{
"epoch": 3.8278481012658228,
"grad_norm": 0.3955478923936077,
"learning_rate": 6.192819233298046e-06,
"loss": 0.8087,
"step": 945
},
{
"epoch": 3.8318987341772153,
"grad_norm": 0.39023703855306485,
"learning_rate": 6.151811864756383e-06,
"loss": 0.8464,
"step": 946
},
{
"epoch": 3.835949367088608,
"grad_norm": 0.41311382326731605,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.8284,
"step": 947
},
{
"epoch": 3.84,
"grad_norm": 0.4064028361002765,
"learning_rate": 6.070132051252868e-06,
"loss": 0.8533,
"step": 948
},
{
"epoch": 3.8440506329113924,
"grad_norm": 0.43245748679363444,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.8401,
"step": 949
},
{
"epoch": 3.848101265822785,
"grad_norm": 0.4051409932335795,
"learning_rate": 5.988900993811575e-06,
"loss": 0.8126,
"step": 950
},
{
"epoch": 3.852151898734177,
"grad_norm": 0.39054258538714604,
"learning_rate": 5.948454566959363e-06,
"loss": 0.8362,
"step": 951
},
{
"epoch": 3.8562025316455695,
"grad_norm": 0.41157827109350537,
"learning_rate": 5.908121309322328e-06,
"loss": 0.8289,
"step": 952
},
{
"epoch": 3.860253164556962,
"grad_norm": 0.41042416418447203,
"learning_rate": 5.867901545738976e-06,
"loss": 0.8258,
"step": 953
},
{
"epoch": 3.8643037974683545,
"grad_norm": 0.4185835870249711,
"learning_rate": 5.827795600133774e-06,
"loss": 0.8352,
"step": 954
},
{
"epoch": 3.868354430379747,
"grad_norm": 0.38944447293323714,
"learning_rate": 5.787803795514466e-06,
"loss": 0.8355,
"step": 955
},
{
"epoch": 3.872405063291139,
"grad_norm": 0.39749515401315505,
"learning_rate": 5.747926453969576e-06,
"loss": 0.8217,
"step": 956
},
{
"epoch": 3.8764556962025316,
"grad_norm": 0.416371870868455,
"learning_rate": 5.708163896665708e-06,
"loss": 0.8272,
"step": 957
},
{
"epoch": 3.880506329113924,
"grad_norm": 0.4120567749071727,
"learning_rate": 5.668516443845047e-06,
"loss": 0.851,
"step": 958
},
{
"epoch": 3.884556962025316,
"grad_norm": 0.39785317798575115,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.8393,
"step": 959
},
{
"epoch": 3.8886075949367087,
"grad_norm": 0.39437313042594835,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.8387,
"step": 960
},
{
"epoch": 3.8926582278481012,
"grad_norm": 0.40588681510386776,
"learning_rate": 5.550267900783019e-06,
"loss": 0.8467,
"step": 961
},
{
"epoch": 3.8967088607594937,
"grad_norm": 0.4149361206667543,
"learning_rate": 5.511084049737623e-06,
"loss": 0.832,
"step": 962
},
{
"epoch": 3.9007594936708863,
"grad_norm": 0.38555150643145236,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.8325,
"step": 963
},
{
"epoch": 3.9048101265822783,
"grad_norm": 0.4256681914296973,
"learning_rate": 5.433066737499948e-06,
"loss": 0.8368,
"step": 964
},
{
"epoch": 3.908860759493671,
"grad_norm": 0.4126984496612008,
"learning_rate": 5.394233904648376e-06,
"loss": 0.8511,
"step": 965
},
{
"epoch": 3.9129113924050634,
"grad_norm": 0.39167960011806974,
"learning_rate": 5.355518704628997e-06,
"loss": 0.8235,
"step": 966
},
{
"epoch": 3.9169620253164554,
"grad_norm": 0.3954610514932393,
"learning_rate": 5.316921449248731e-06,
"loss": 0.854,
"step": 967
},
{
"epoch": 3.921012658227848,
"grad_norm": 0.3976267580755455,
"learning_rate": 5.278442449364538e-06,
"loss": 0.807,
"step": 968
},
{
"epoch": 3.9250632911392405,
"grad_norm": 0.39205527081598573,
"learning_rate": 5.240082014881016e-06,
"loss": 0.839,
"step": 969
},
{
"epoch": 3.929113924050633,
"grad_norm": 0.39554454987437326,
"learning_rate": 5.201840454747822e-06,
"loss": 0.8455,
"step": 970
},
{
"epoch": 3.9331645569620255,
"grad_norm": 0.3986555170166855,
"learning_rate": 5.163718076957223e-06,
"loss": 0.8178,
"step": 971
},
{
"epoch": 3.9372151898734176,
"grad_norm": 0.40313909947321747,
"learning_rate": 5.125715188541609e-06,
"loss": 0.8255,
"step": 972
},
{
"epoch": 3.94126582278481,
"grad_norm": 0.39942155904316234,
"learning_rate": 5.087832095571021e-06,
"loss": 0.8522,
"step": 973
},
{
"epoch": 3.9453164556962026,
"grad_norm": 0.3933706459075555,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.8432,
"step": 974
},
{
"epoch": 3.9493670886075947,
"grad_norm": 0.4227135857241554,
"learning_rate": 5.01242651541854e-06,
"loss": 0.831,
"step": 975
},
{
"epoch": 3.953417721518987,
"grad_norm": 0.39252672427987256,
"learning_rate": 4.974904635542815e-06,
"loss": 0.8357,
"step": 976
},
{
"epoch": 3.9574683544303797,
"grad_norm": 0.3815454271801741,
"learning_rate": 4.937503765719582e-06,
"loss": 0.8233,
"step": 977
},
{
"epoch": 3.961518987341772,
"grad_norm": 0.3863057303643772,
"learning_rate": 4.900224207170299e-06,
"loss": 0.8273,
"step": 978
},
{
"epoch": 3.9655696202531647,
"grad_norm": 0.4300746771797314,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.8123,
"step": 979
},
{
"epoch": 3.969620253164557,
"grad_norm": 0.40969765653693874,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.8326,
"step": 980
},
{
"epoch": 3.9736708860759493,
"grad_norm": 0.40059742734954723,
"learning_rate": 4.789116396710924e-06,
"loss": 0.8358,
"step": 981
},
{
"epoch": 3.977721518987342,
"grad_norm": 0.38865510216234733,
"learning_rate": 4.752325075895368e-06,
"loss": 0.8247,
"step": 982
},
{
"epoch": 3.981772151898734,
"grad_norm": 0.3902921088805331,
"learning_rate": 4.715656557757473e-06,
"loss": 0.8175,
"step": 983
},
{
"epoch": 3.9858227848101264,
"grad_norm": 0.3829147501465742,
"learning_rate": 4.679111137620442e-06,
"loss": 0.805,
"step": 984
},
{
"epoch": 3.989873417721519,
"grad_norm": 0.37522749608161143,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.8279,
"step": 985
},
{
"epoch": 3.9939240506329114,
"grad_norm": 0.3960158175213234,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.8631,
"step": 986
},
{
"epoch": 3.997974683544304,
"grad_norm": 0.3866156413065647,
"learning_rate": 4.570216403561141e-06,
"loss": 0.8334,
"step": 987
},
{
"epoch": 4.002025316455696,
"grad_norm": 0.5378038630394847,
"learning_rate": 4.534166308795815e-06,
"loss": 0.7804,
"step": 988
},
{
"epoch": 4.006075949367089,
"grad_norm": 0.8205915182763533,
"learning_rate": 4.498240773728859e-06,
"loss": 0.7315,
"step": 989
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.6135293482309975,
"learning_rate": 4.462440087699609e-06,
"loss": 0.7112,
"step": 990
},
{
"epoch": 4.014177215189873,
"grad_norm": 0.8006562859565675,
"learning_rate": 4.426764539041861e-06,
"loss": 0.7173,
"step": 991
},
{
"epoch": 4.018227848101266,
"grad_norm": 1.0646613127381366,
"learning_rate": 4.391214415081582e-06,
"loss": 0.7063,
"step": 992
},
{
"epoch": 4.022278481012658,
"grad_norm": 0.8119572475692994,
"learning_rate": 4.355790002134579e-06,
"loss": 0.7207,
"step": 993
},
{
"epoch": 4.02632911392405,
"grad_norm": 0.7533676225386695,
"learning_rate": 4.320491585504207e-06,
"loss": 0.711,
"step": 994
},
{
"epoch": 4.030379746835443,
"grad_norm": 0.8604012651689938,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.7045,
"step": 995
},
{
"epoch": 4.034430379746835,
"grad_norm": 0.8209680197100339,
"learning_rate": 4.250273877330691e-06,
"loss": 0.7123,
"step": 996
},
{
"epoch": 4.038481012658228,
"grad_norm": 0.7055162187160072,
"learning_rate": 4.215355151311313e-06,
"loss": 0.7163,
"step": 997
},
{
"epoch": 4.04253164556962,
"grad_norm": 0.5913994237869326,
"learning_rate": 4.180563552651542e-06,
"loss": 0.7131,
"step": 998
},
{
"epoch": 4.046582278481012,
"grad_norm": 0.6447616624932161,
"learning_rate": 4.145899361558147e-06,
"loss": 0.6941,
"step": 999
},
{
"epoch": 4.050632911392405,
"grad_norm": 0.6965692252303175,
"learning_rate": 4.111362857211738e-06,
"loss": 0.7132,
"step": 1000
},
{
"epoch": 4.054683544303797,
"grad_norm": 0.6822825928854558,
"learning_rate": 4.076954317764592e-06,
"loss": 0.7062,
"step": 1001
},
{
"epoch": 4.0587341772151895,
"grad_norm": 0.5238535747929336,
"learning_rate": 4.042674020338335e-06,
"loss": 0.6885,
"step": 1002
},
{
"epoch": 4.062784810126582,
"grad_norm": 0.5708851326447116,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.7137,
"step": 1003
},
{
"epoch": 4.0668354430379745,
"grad_norm": 0.5693583929473824,
"learning_rate": 3.974499254868674e-06,
"loss": 0.72,
"step": 1004
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.5675276223470386,
"learning_rate": 3.940605335895451e-06,
"loss": 0.7091,
"step": 1005
},
{
"epoch": 4.0749367088607595,
"grad_norm": 0.543920653459443,
"learning_rate": 3.90684075707908e-06,
"loss": 0.714,
"step": 1006
},
{
"epoch": 4.078987341772152,
"grad_norm": 0.5227736598683538,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.703,
"step": 1007
},
{
"epoch": 4.083037974683545,
"grad_norm": 0.5725441454531157,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.7131,
"step": 1008
},
{
"epoch": 4.087088607594937,
"grad_norm": 0.5678358605083701,
"learning_rate": 3.806325775702304e-06,
"loss": 0.6959,
"step": 1009
},
{
"epoch": 4.091139240506329,
"grad_norm": 0.5437771224014913,
"learning_rate": 3.773081266416434e-06,
"loss": 0.7088,
"step": 1010
},
{
"epoch": 4.095189873417722,
"grad_norm": 0.4667145160048971,
"learning_rate": 3.739967446503245e-06,
"loss": 0.69,
"step": 1011
},
{
"epoch": 4.099240506329114,
"grad_norm": 0.4787965621677959,
"learning_rate": 3.706984582656894e-06,
"loss": 0.6975,
"step": 1012
},
{
"epoch": 4.103291139240507,
"grad_norm": 0.5338419534730515,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.7026,
"step": 1013
},
{
"epoch": 4.107341772151899,
"grad_norm": 0.5628245086843368,
"learning_rate": 3.641412784665648e-06,
"loss": 0.7055,
"step": 1014
},
{
"epoch": 4.111392405063291,
"grad_norm": 0.47677137986699647,
"learning_rate": 3.608824378627005e-06,
"loss": 0.718,
"step": 1015
},
{
"epoch": 4.115443037974684,
"grad_norm": 0.4549812008595826,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.7088,
"step": 1016
},
{
"epoch": 4.119493670886076,
"grad_norm": 0.4774496600085126,
"learning_rate": 3.544043864774269e-06,
"loss": 0.7037,
"step": 1017
},
{
"epoch": 4.123544303797468,
"grad_norm": 0.5278184498717162,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.712,
"step": 1018
},
{
"epoch": 4.127594936708861,
"grad_norm": 0.5317644035176917,
"learning_rate": 3.479793485887819e-06,
"loss": 0.7075,
"step": 1019
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.4650853550070621,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.7175,
"step": 1020
},
{
"epoch": 4.135696202531646,
"grad_norm": 0.46774090050836875,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.6992,
"step": 1021
},
{
"epoch": 4.139746835443038,
"grad_norm": 0.46773316023076744,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.6936,
"step": 1022
},
{
"epoch": 4.14379746835443,
"grad_norm": 0.4996065944305862,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.7001,
"step": 1023
},
{
"epoch": 4.147848101265823,
"grad_norm": 0.47992872085592647,
"learning_rate": 3.321500420335e-06,
"loss": 0.701,
"step": 1024
},
{
"epoch": 4.151898734177215,
"grad_norm": 0.4510648557523204,
"learning_rate": 3.290243771741275e-06,
"loss": 0.7068,
"step": 1025
},
{
"epoch": 4.155949367088607,
"grad_norm": 0.47752197550029546,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.7062,
"step": 1026
},
{
"epoch": 4.16,
"grad_norm": 0.4864820860028587,
"learning_rate": 3.228134459445149e-06,
"loss": 0.7084,
"step": 1027
},
{
"epoch": 4.164050632911392,
"grad_norm": 0.4463532550040719,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.7024,
"step": 1028
},
{
"epoch": 4.168101265822785,
"grad_norm": 0.428362823367227,
"learning_rate": 3.166565459258513e-06,
"loss": 0.6795,
"step": 1029
},
{
"epoch": 4.172151898734177,
"grad_norm": 0.45370198175030574,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.6985,
"step": 1030
},
{
"epoch": 4.176202531645569,
"grad_norm": 0.4519282011494011,
"learning_rate": 3.105538754650419e-06,
"loss": 0.704,
"step": 1031
},
{
"epoch": 4.180253164556962,
"grad_norm": 0.46207927929558285,
"learning_rate": 3.07522937824712e-06,
"loss": 0.6913,
"step": 1032
},
{
"epoch": 4.184303797468354,
"grad_norm": 0.4325764611816221,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.7055,
"step": 1033
},
{
"epoch": 4.188354430379746,
"grad_norm": 0.44617626630149526,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.7068,
"step": 1034
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.4338305339507494,
"learning_rate": 2.985120078631465e-06,
"loss": 0.6916,
"step": 1035
},
{
"epoch": 4.1964556962025314,
"grad_norm": 0.47531486101474213,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.6973,
"step": 1036
},
{
"epoch": 4.200506329113924,
"grad_norm": 0.471508716430008,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.6933,
"step": 1037
},
{
"epoch": 4.2045569620253165,
"grad_norm": 0.46599185251674896,
"learning_rate": 2.896244091927678e-06,
"loss": 0.7165,
"step": 1038
},
{
"epoch": 4.2086075949367086,
"grad_norm": 0.4368257375197407,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.7118,
"step": 1039
},
{
"epoch": 4.2126582278481015,
"grad_norm": 0.43949007868307566,
"learning_rate": 2.837681792951994e-06,
"loss": 0.6886,
"step": 1040
},
{
"epoch": 4.216708860759494,
"grad_norm": 0.45277385369953055,
"learning_rate": 2.808607860255981e-06,
"loss": 0.7161,
"step": 1041
},
{
"epoch": 4.220759493670886,
"grad_norm": 0.45040089964043806,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.6955,
"step": 1042
},
{
"epoch": 4.224810126582279,
"grad_norm": 0.4340483760543899,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.7084,
"step": 1043
},
{
"epoch": 4.228860759493671,
"grad_norm": 0.4424639543086904,
"learning_rate": 2.722217735873718e-06,
"loss": 0.6924,
"step": 1044
},
{
"epoch": 4.232911392405064,
"grad_norm": 0.4387694699561883,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.7152,
"step": 1045
},
{
"epoch": 4.236962025316456,
"grad_norm": 0.43449038365174913,
"learning_rate": 2.665319697490205e-06,
"loss": 0.7119,
"step": 1046
},
{
"epoch": 4.241012658227848,
"grad_norm": 0.4467820144639237,
"learning_rate": 2.637079980714945e-06,
"loss": 0.6958,
"step": 1047
},
{
"epoch": 4.245063291139241,
"grad_norm": 0.4402037305552043,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.7034,
"step": 1048
},
{
"epoch": 4.249113924050633,
"grad_norm": 0.43360271607976036,
"learning_rate": 2.581020289189571e-06,
"loss": 0.6967,
"step": 1049
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.4424166813166147,
"learning_rate": 2.553200765936501e-06,
"loss": 0.6947,
"step": 1050
},
{
"epoch": 4.257215189873418,
"grad_norm": 0.44038405397035735,
"learning_rate": 2.525521756805962e-06,
"loss": 0.7233,
"step": 1051
},
{
"epoch": 4.26126582278481,
"grad_norm": 0.43114991438659656,
"learning_rate": 2.497983484720885e-06,
"loss": 0.7079,
"step": 1052
},
{
"epoch": 4.265316455696203,
"grad_norm": 0.43657640839455686,
"learning_rate": 2.470586171470728e-06,
"loss": 0.7037,
"step": 1053
},
{
"epoch": 4.269367088607595,
"grad_norm": 0.4452881518531483,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.7018,
"step": 1054
},
{
"epoch": 4.273417721518987,
"grad_norm": 0.44614435605597713,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.7073,
"step": 1055
},
{
"epoch": 4.27746835443038,
"grad_norm": 0.43264531143401874,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.6887,
"step": 1056
},
{
"epoch": 4.281518987341772,
"grad_norm": 0.4325822900012594,
"learning_rate": 2.362410902836978e-06,
"loss": 0.7029,
"step": 1057
},
{
"epoch": 4.285569620253165,
"grad_norm": 0.42906508714780717,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.7012,
"step": 1058
},
{
"epoch": 4.289620253164557,
"grad_norm": 0.43459514889367945,
"learning_rate": 2.309174704446411e-06,
"loss": 0.7004,
"step": 1059
},
{
"epoch": 4.293670886075949,
"grad_norm": 0.4313407467191593,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.6897,
"step": 1060
},
{
"epoch": 4.297721518987342,
"grad_norm": 0.4336113322248844,
"learning_rate": 2.256508422807855e-06,
"loss": 0.7011,
"step": 1061
},
{
"epoch": 4.301772151898734,
"grad_norm": 0.42930081261946523,
"learning_rate": 2.230389531696946e-06,
"loss": 0.6827,
"step": 1062
},
{
"epoch": 4.305822784810126,
"grad_norm": 0.42232829089535134,
"learning_rate": 2.204413754585857e-06,
"loss": 0.6902,
"step": 1063
},
{
"epoch": 4.309873417721519,
"grad_norm": 0.4477614298577946,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.6917,
"step": 1064
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.44082820495276326,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.7031,
"step": 1065
},
{
"epoch": 4.317974683544303,
"grad_norm": 0.43766541928310015,
"learning_rate": 2.127347193531757e-06,
"loss": 0.7034,
"step": 1066
},
{
"epoch": 4.322025316455696,
"grad_norm": 0.45224124799832566,
"learning_rate": 2.101945952921942e-06,
"loss": 0.7233,
"step": 1067
},
{
"epoch": 4.326075949367088,
"grad_norm": 0.4415428961255853,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.7155,
"step": 1068
},
{
"epoch": 4.330126582278481,
"grad_norm": 0.42666659472461177,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.6946,
"step": 1069
},
{
"epoch": 4.334177215189873,
"grad_norm": 0.4406641473532928,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.6856,
"step": 1070
},
{
"epoch": 4.3382278481012655,
"grad_norm": 0.42147847491007207,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.6897,
"step": 1071
},
{
"epoch": 4.3422784810126585,
"grad_norm": 0.43587382908313144,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.6845,
"step": 1072
},
{
"epoch": 4.3463291139240505,
"grad_norm": 0.4373089087552151,
"learning_rate": 1.95257270594031e-06,
"loss": 0.7038,
"step": 1073
},
{
"epoch": 4.3503797468354435,
"grad_norm": 0.44171118822519995,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.7114,
"step": 1074
},
{
"epoch": 4.3544303797468356,
"grad_norm": 0.43069528441433674,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.6932,
"step": 1075
},
{
"epoch": 4.358481012658228,
"grad_norm": 0.43760084877196975,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.7046,
"step": 1076
},
{
"epoch": 4.362531645569621,
"grad_norm": 0.44029447016066847,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.7009,
"step": 1077
},
{
"epoch": 4.366582278481013,
"grad_norm": 0.4358764475689891,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.6927,
"step": 1078
},
{
"epoch": 4.370632911392405,
"grad_norm": 0.4181062530899562,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.687,
"step": 1079
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.43271052354967543,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.7173,
"step": 1080
},
{
"epoch": 4.37873417721519,
"grad_norm": 0.4415207079818095,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.7136,
"step": 1081
},
{
"epoch": 4.382784810126582,
"grad_norm": 0.424539304106672,
"learning_rate": 1.738336950434061e-06,
"loss": 0.6931,
"step": 1082
},
{
"epoch": 4.386835443037975,
"grad_norm": 0.44286192703360766,
"learning_rate": 1.715265841375957e-06,
"loss": 0.697,
"step": 1083
},
{
"epoch": 4.390886075949367,
"grad_norm": 0.42460163264470935,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.7025,
"step": 1084
},
{
"epoch": 4.39493670886076,
"grad_norm": 0.43167908901655394,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.701,
"step": 1085
},
{
"epoch": 4.398987341772152,
"grad_norm": 0.4531023079303449,
"learning_rate": 1.646936827927441e-06,
"loss": 0.6954,
"step": 1086
},
{
"epoch": 4.403037974683544,
"grad_norm": 0.4265473665693666,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.6877,
"step": 1087
},
{
"epoch": 4.407088607594937,
"grad_norm": 0.4303785713344227,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.7077,
"step": 1088
},
{
"epoch": 4.411139240506329,
"grad_norm": 0.42670505361334066,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.7132,
"step": 1089
},
{
"epoch": 4.415189873417722,
"grad_norm": 0.42200178241407776,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.7093,
"step": 1090
},
{
"epoch": 4.419240506329114,
"grad_norm": 0.4277797326792725,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.6967,
"step": 1091
},
{
"epoch": 4.423291139240506,
"grad_norm": 0.42841933870028404,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.694,
"step": 1092
},
{
"epoch": 4.427341772151899,
"grad_norm": 0.43424143800177945,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.6908,
"step": 1093
},
{
"epoch": 4.431392405063291,
"grad_norm": 0.43028358229741964,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.7,
"step": 1094
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.4427078992013297,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.7098,
"step": 1095
},
{
"epoch": 4.439493670886076,
"grad_norm": 0.4300736661093881,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.7236,
"step": 1096
},
{
"epoch": 4.443544303797468,
"grad_norm": 0.42971014790272455,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.7082,
"step": 1097
},
{
"epoch": 4.44759493670886,
"grad_norm": 0.4177102301297652,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.7012,
"step": 1098
},
{
"epoch": 4.451645569620253,
"grad_norm": 0.42582422073612103,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.6889,
"step": 1099
},
{
"epoch": 4.455696202531645,
"grad_norm": 0.4377930387005071,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.7124,
"step": 1100
},
{
"epoch": 4.459746835443038,
"grad_norm": 0.43616023957820743,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.7147,
"step": 1101
},
{
"epoch": 4.46379746835443,
"grad_norm": 0.43093825797089075,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.6898,
"step": 1102
},
{
"epoch": 4.467848101265822,
"grad_norm": 0.43233796435099264,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.7044,
"step": 1103
},
{
"epoch": 4.471898734177215,
"grad_norm": 0.4298789639342173,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.7084,
"step": 1104
},
{
"epoch": 4.4759493670886075,
"grad_norm": 0.4211420678424715,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.6905,
"step": 1105
},
{
"epoch": 4.48,
"grad_norm": 0.4307918493362568,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.7037,
"step": 1106
},
{
"epoch": 4.4840506329113925,
"grad_norm": 0.43666254631061896,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.6904,
"step": 1107
},
{
"epoch": 4.488101265822785,
"grad_norm": 0.42669554806094023,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.6909,
"step": 1108
},
{
"epoch": 4.4921518987341775,
"grad_norm": 0.41675733567632745,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.6838,
"step": 1109
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.426756795729151,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.6966,
"step": 1110
},
{
"epoch": 4.500253164556962,
"grad_norm": 0.4300140907706569,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.6894,
"step": 1111
},
{
"epoch": 4.504303797468355,
"grad_norm": 0.4249921194674582,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.6806,
"step": 1112
},
{
"epoch": 4.508354430379747,
"grad_norm": 0.42316922563918463,
"learning_rate": 1.092402023876933e-06,
"loss": 0.7051,
"step": 1113
},
{
"epoch": 4.512405063291139,
"grad_norm": 0.44468834218545794,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.7163,
"step": 1114
},
{
"epoch": 4.516455696202532,
"grad_norm": 0.42840054230743446,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.7197,
"step": 1115
},
{
"epoch": 4.520506329113924,
"grad_norm": 0.42723742207331095,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.7053,
"step": 1116
},
{
"epoch": 4.524556962025317,
"grad_norm": 0.42087115798807706,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.6968,
"step": 1117
},
{
"epoch": 4.528607594936709,
"grad_norm": 0.423158956967241,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.6946,
"step": 1118
},
{
"epoch": 4.532658227848101,
"grad_norm": 0.42604412452990226,
"learning_rate": 9.841383857091947e-07,
"loss": 0.7105,
"step": 1119
},
{
"epoch": 4.536708860759494,
"grad_norm": 0.4358607636549177,
"learning_rate": 9.666296380321616e-07,
"loss": 0.6916,
"step": 1120
},
{
"epoch": 4.540759493670886,
"grad_norm": 0.41856368390634746,
"learning_rate": 9.492741825034124e-07,
"loss": 0.7033,
"step": 1121
},
{
"epoch": 4.544810126582279,
"grad_norm": 0.43025358164523614,
"learning_rate": 9.320721589013892e-07,
"loss": 0.6954,
"step": 1122
},
{
"epoch": 4.548860759493671,
"grad_norm": 0.43173513492169824,
"learning_rate": 9.150237057688339e-07,
"loss": 0.7004,
"step": 1123
},
{
"epoch": 4.552911392405063,
"grad_norm": 0.4269420297935245,
"learning_rate": 8.981289604116328e-07,
"loss": 0.6926,
"step": 1124
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.4409688940914616,
"learning_rate": 8.813880588977542e-07,
"loss": 0.6947,
"step": 1125
},
{
"epoch": 4.561012658227848,
"grad_norm": 0.4462289573231369,
"learning_rate": 8.648011360561126e-07,
"loss": 0.7082,
"step": 1126
},
{
"epoch": 4.56506329113924,
"grad_norm": 0.42135363850510565,
"learning_rate": 8.483683254755037e-07,
"loss": 0.6973,
"step": 1127
},
{
"epoch": 4.569113924050633,
"grad_norm": 0.42468306548865903,
"learning_rate": 8.320897595035227e-07,
"loss": 0.7102,
"step": 1128
},
{
"epoch": 4.573164556962025,
"grad_norm": 0.4214054771545712,
"learning_rate": 8.159655692455093e-07,
"loss": 0.6915,
"step": 1129
},
{
"epoch": 4.577215189873417,
"grad_norm": 0.42348458690831375,
"learning_rate": 7.999958845634648e-07,
"loss": 0.7028,
"step": 1130
},
{
"epoch": 4.58126582278481,
"grad_norm": 0.42838827037885974,
"learning_rate": 7.841808340750478e-07,
"loss": 0.7159,
"step": 1131
},
{
"epoch": 4.585316455696202,
"grad_norm": 0.4484354255970924,
"learning_rate": 7.685205451524869e-07,
"loss": 0.6906,
"step": 1132
},
{
"epoch": 4.589367088607595,
"grad_norm": 0.4328604678010873,
"learning_rate": 7.530151439216027e-07,
"loss": 0.7046,
"step": 1133
},
{
"epoch": 4.593417721518987,
"grad_norm": 0.4278567622318977,
"learning_rate": 7.376647552607675e-07,
"loss": 0.6955,
"step": 1134
},
{
"epoch": 4.597468354430379,
"grad_norm": 0.42957228086909116,
"learning_rate": 7.224695027998963e-07,
"loss": 0.705,
"step": 1135
},
{
"epoch": 4.601518987341772,
"grad_norm": 0.4230681608104427,
"learning_rate": 7.07429508919466e-07,
"loss": 0.6861,
"step": 1136
},
{
"epoch": 4.605569620253164,
"grad_norm": 0.426882441780937,
"learning_rate": 6.925448947495206e-07,
"loss": 0.7076,
"step": 1137
},
{
"epoch": 4.609620253164557,
"grad_norm": 0.41572051055999243,
"learning_rate": 6.778157801686936e-07,
"loss": 0.6984,
"step": 1138
},
{
"epoch": 4.613670886075949,
"grad_norm": 0.43699526791036136,
"learning_rate": 6.632422838032515e-07,
"loss": 0.7049,
"step": 1139
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.4317982991406899,
"learning_rate": 6.488245230261281e-07,
"loss": 0.6914,
"step": 1140
},
{
"epoch": 4.6217721518987345,
"grad_norm": 0.4173677011178122,
"learning_rate": 6.345626139559868e-07,
"loss": 0.7,
"step": 1141
},
{
"epoch": 4.6258227848101265,
"grad_norm": 0.4237009874620114,
"learning_rate": 6.204566714562866e-07,
"loss": 0.6983,
"step": 1142
},
{
"epoch": 4.629873417721519,
"grad_norm": 0.4444386531039824,
"learning_rate": 6.06506809134344e-07,
"loss": 0.7164,
"step": 1143
},
{
"epoch": 4.633924050632912,
"grad_norm": 0.4371923372147083,
"learning_rate": 5.927131393404373e-07,
"loss": 0.7211,
"step": 1144
},
{
"epoch": 4.637974683544304,
"grad_norm": 0.43428852607877233,
"learning_rate": 5.790757731668817e-07,
"loss": 0.7002,
"step": 1145
},
{
"epoch": 4.642025316455696,
"grad_norm": 0.4254789709558514,
"learning_rate": 5.655948204471507e-07,
"loss": 0.7019,
"step": 1146
},
{
"epoch": 4.646075949367089,
"grad_norm": 0.42653339197867096,
"learning_rate": 5.522703897549875e-07,
"loss": 0.7066,
"step": 1147
},
{
"epoch": 4.650126582278481,
"grad_norm": 0.4321802200093975,
"learning_rate": 5.391025884035239e-07,
"loss": 0.7165,
"step": 1148
},
{
"epoch": 4.654177215189874,
"grad_norm": 0.4309588019853538,
"learning_rate": 5.260915224444207e-07,
"loss": 0.7277,
"step": 1149
},
{
"epoch": 4.658227848101266,
"grad_norm": 0.43655709032723294,
"learning_rate": 5.132372966670129e-07,
"loss": 0.7153,
"step": 1150
},
{
"epoch": 4.662278481012658,
"grad_norm": 0.4345257519761863,
"learning_rate": 5.005400145974704e-07,
"loss": 0.7135,
"step": 1151
},
{
"epoch": 4.666329113924051,
"grad_norm": 0.4314025736985042,
"learning_rate": 4.879997784979562e-07,
"loss": 0.7013,
"step": 1152
},
{
"epoch": 4.670379746835443,
"grad_norm": 0.43127719530758246,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.6937,
"step": 1153
},
{
"epoch": 4.674430379746836,
"grad_norm": 0.4301789284255141,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.6949,
"step": 1154
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.4303863973674103,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.6983,
"step": 1155
},
{
"epoch": 4.68253164556962,
"grad_norm": 0.42119442554469505,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.6972,
"step": 1156
},
{
"epoch": 4.686582278481013,
"grad_norm": 0.43279395777890595,
"learning_rate": 4.27657778143431e-07,
"loss": 0.7077,
"step": 1157
},
{
"epoch": 4.690632911392405,
"grad_norm": 0.43191350802261236,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.7059,
"step": 1158
},
{
"epoch": 4.694683544303797,
"grad_norm": 0.4382341612461369,
"learning_rate": 4.046237371675177e-07,
"loss": 0.7093,
"step": 1159
},
{
"epoch": 4.69873417721519,
"grad_norm": 0.4112902294649588,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.7009,
"step": 1160
},
{
"epoch": 4.702784810126582,
"grad_norm": 0.426954004562,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.6964,
"step": 1161
},
{
"epoch": 4.706835443037974,
"grad_norm": 0.4350761147932092,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.7143,
"step": 1162
},
{
"epoch": 4.710886075949367,
"grad_norm": 0.41566990676267745,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.6914,
"step": 1163
},
{
"epoch": 4.714936708860759,
"grad_norm": 0.4276763642597431,
"learning_rate": 3.498020255137813e-07,
"loss": 0.6855,
"step": 1164
},
{
"epoch": 4.718987341772152,
"grad_norm": 0.43888894212358315,
"learning_rate": 3.393121163057811e-07,
"loss": 0.7048,
"step": 1165
},
{
"epoch": 4.723037974683544,
"grad_norm": 0.420992748351888,
"learning_rate": 3.289805515751399e-07,
"loss": 0.7143,
"step": 1166
},
{
"epoch": 4.727088607594936,
"grad_norm": 0.41981595908076746,
"learning_rate": 3.188074145308573e-07,
"loss": 0.7145,
"step": 1167
},
{
"epoch": 4.731139240506329,
"grad_norm": 0.41750896397713394,
"learning_rate": 3.087927871059804e-07,
"loss": 0.7105,
"step": 1168
},
{
"epoch": 4.735189873417721,
"grad_norm": 0.41505946134915606,
"learning_rate": 2.989367499569418e-07,
"loss": 0.6993,
"step": 1169
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.4292289445388557,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.6854,
"step": 1170
},
{
"epoch": 4.743291139240506,
"grad_norm": 0.4280533533293539,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.7042,
"step": 1171
},
{
"epoch": 4.7473417721518985,
"grad_norm": 0.41745230150446416,
"learning_rate": 2.703209675663887e-07,
"loss": 0.7024,
"step": 1172
},
{
"epoch": 4.751392405063291,
"grad_norm": 0.4181751119668438,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.6923,
"step": 1173
},
{
"epoch": 4.7554430379746835,
"grad_norm": 0.4223973261740004,
"learning_rate": 2.520381518804471e-07,
"loss": 0.6941,
"step": 1174
},
{
"epoch": 4.759493670886076,
"grad_norm": 0.41305704752336037,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.6834,
"step": 1175
},
{
"epoch": 4.7635443037974685,
"grad_norm": 0.42122236650404093,
"learning_rate": 2.343915243930317e-07,
"loss": 0.705,
"step": 1176
},
{
"epoch": 4.767594936708861,
"grad_norm": 0.41835562333990123,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.7022,
"step": 1177
},
{
"epoch": 4.7716455696202535,
"grad_norm": 0.4214276152057772,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.6827,
"step": 1178
},
{
"epoch": 4.775696202531646,
"grad_norm": 0.4252474705951066,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.707,
"step": 1179
},
{
"epoch": 4.779746835443038,
"grad_norm": 0.42229611977473613,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.7011,
"step": 1180
},
{
"epoch": 4.783797468354431,
"grad_norm": 0.4333175559313292,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.7125,
"step": 1181
},
{
"epoch": 4.787848101265823,
"grad_norm": 0.42509733627258217,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.695,
"step": 1182
},
{
"epoch": 4.791898734177215,
"grad_norm": 0.4309391636569834,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.7083,
"step": 1183
},
{
"epoch": 4.795949367088608,
"grad_norm": 0.4169489957906063,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.6968,
"step": 1184
},
{
"epoch": 4.8,
"grad_norm": 0.4236981064232541,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.6974,
"step": 1185
},
{
"epoch": 4.804050632911393,
"grad_norm": 0.4294041749513422,
"learning_rate": 1.557203888516745e-07,
"loss": 0.6909,
"step": 1186
},
{
"epoch": 4.808101265822785,
"grad_norm": 0.42694921239735645,
"learning_rate": 1.487313098987131e-07,
"loss": 0.7018,
"step": 1187
},
{
"epoch": 4.812151898734177,
"grad_norm": 0.4169753335711223,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.6905,
"step": 1188
},
{
"epoch": 4.81620253164557,
"grad_norm": 0.4097038137874668,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.6918,
"step": 1189
},
{
"epoch": 4.820253164556962,
"grad_norm": 0.4311362617809195,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.6969,
"step": 1190
},
{
"epoch": 4.824303797468355,
"grad_norm": 0.4179908082192621,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.6897,
"step": 1191
},
{
"epoch": 4.828354430379747,
"grad_norm": 0.4259540802649157,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.6955,
"step": 1192
},
{
"epoch": 4.832405063291139,
"grad_norm": 0.4255668831459623,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.7218,
"step": 1193
},
{
"epoch": 4.836455696202532,
"grad_norm": 0.4165406665604323,
"learning_rate": 1.042873700093061e-07,
"loss": 0.7105,
"step": 1194
},
{
"epoch": 4.840506329113924,
"grad_norm": 0.4308755263035076,
"learning_rate": 9.857879404556291e-08,
"loss": 0.7012,
"step": 1195
},
{
"epoch": 4.844556962025316,
"grad_norm": 0.424739930880005,
"learning_rate": 9.303050139225722e-08,
"loss": 0.6965,
"step": 1196
},
{
"epoch": 4.848607594936709,
"grad_norm": 0.42557455253594223,
"learning_rate": 8.76425367345779e-08,
"loss": 0.7105,
"step": 1197
},
{
"epoch": 4.852658227848101,
"grad_norm": 0.41076568091150983,
"learning_rate": 8.241494346644897e-08,
"loss": 0.6811,
"step": 1198
},
{
"epoch": 4.856708860759493,
"grad_norm": 0.42236356203521314,
"learning_rate": 7.734776369019204e-08,
"loss": 0.7098,
"step": 1199
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.425147414178042,
"learning_rate": 7.244103821617332e-08,
"loss": 0.717,
"step": 1200
},
{
"epoch": 4.864810126582278,
"grad_norm": 0.4174541137251538,
"learning_rate": 6.769480656248606e-08,
"loss": 0.7097,
"step": 1201
},
{
"epoch": 4.868860759493671,
"grad_norm": 0.42906219364095366,
"learning_rate": 6.310910695462635e-08,
"loss": 0.703,
"step": 1202
},
{
"epoch": 4.872911392405063,
"grad_norm": 0.4196365032277408,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.7033,
"step": 1203
},
{
"epoch": 4.876962025316455,
"grad_norm": 0.4223705529781405,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.6723,
"step": 1204
},
{
"epoch": 4.881012658227848,
"grad_norm": 0.4271231418754079,
"learning_rate": 5.031556326567488e-08,
"loss": 0.7065,
"step": 1205
},
{
"epoch": 4.88506329113924,
"grad_norm": 0.4129608355523833,
"learning_rate": 4.637234823364312e-08,
"loss": 0.675,
"step": 1206
},
{
"epoch": 4.889113924050633,
"grad_norm": 0.42865488187966044,
"learning_rate": 4.258983697558838e-08,
"loss": 0.6804,
"step": 1207
},
{
"epoch": 4.8931645569620255,
"grad_norm": 0.4237075464351991,
"learning_rate": 3.896805995533548e-08,
"loss": 0.7013,
"step": 1208
},
{
"epoch": 4.8972151898734175,
"grad_norm": 0.4167080650661818,
"learning_rate": 3.550704634218028e-08,
"loss": 0.6806,
"step": 1209
},
{
"epoch": 4.9012658227848105,
"grad_norm": 0.4119022855644748,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.6907,
"step": 1210
},
{
"epoch": 4.905316455696203,
"grad_norm": 0.42246083478281926,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.7077,
"step": 1211
},
{
"epoch": 4.909367088607595,
"grad_norm": 0.41828190078966726,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.6831,
"step": 1212
},
{
"epoch": 4.913417721518988,
"grad_norm": 0.42919164306894236,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.699,
"step": 1213
},
{
"epoch": 4.91746835443038,
"grad_norm": 0.4428261549377888,
"learning_rate": 2.061435966221881e-08,
"loss": 0.7138,
"step": 1214
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.42240026346444,
"learning_rate": 1.811846652477245e-08,
"loss": 0.7013,
"step": 1215
},
{
"epoch": 4.925569620253165,
"grad_norm": 0.4214154663109981,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.6878,
"step": 1216
},
{
"epoch": 4.929620253164557,
"grad_norm": 0.4247459993141148,
"learning_rate": 1.360949303591097e-08,
"loss": 0.7048,
"step": 1217
},
{
"epoch": 4.93367088607595,
"grad_norm": 0.4272667051835984,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.6942,
"step": 1218
},
{
"epoch": 4.937721518987342,
"grad_norm": 0.4379076009312832,
"learning_rate": 9.744388816668172e-09,
"loss": 0.7007,
"step": 1219
},
{
"epoch": 4.941772151898734,
"grad_norm": 0.4196922127042964,
"learning_rate": 8.05332740472009e-09,
"loss": 0.6763,
"step": 1220
},
{
"epoch": 4.945822784810127,
"grad_norm": 0.43062812830396013,
"learning_rate": 6.523278382872811e-09,
"loss": 0.7062,
"step": 1221
},
{
"epoch": 4.949873417721519,
"grad_norm": 0.4253989024454168,
"learning_rate": 5.15425407393133e-09,
"loss": 0.7031,
"step": 1222
},
{
"epoch": 4.953924050632912,
"grad_norm": 0.4230988072285643,
"learning_rate": 3.94626550383137e-09,
"loss": 0.6915,
"step": 1223
},
{
"epoch": 4.957974683544304,
"grad_norm": 0.4263341423618711,
"learning_rate": 2.899322401546112e-09,
"loss": 0.6793,
"step": 1224
},
{
"epoch": 4.962025316455696,
"grad_norm": 0.4144393537817264,
"learning_rate": 2.013433199010706e-09,
"loss": 0.6898,
"step": 1225
},
{
"epoch": 4.966075949367089,
"grad_norm": 0.4287271981991976,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.7084,
"step": 1226
},
{
"epoch": 4.970126582278481,
"grad_norm": 0.421592849120306,
"learning_rate": 7.248437353468695e-10,
"loss": 0.6985,
"step": 1227
},
{
"epoch": 4.974177215189873,
"grad_norm": 0.41761133067426787,
"learning_rate": 3.221538523412449e-10,
"loss": 0.6869,
"step": 1228
},
{
"epoch": 4.978227848101266,
"grad_norm": 0.43636965209543643,
"learning_rate": 8.053862524670663e-11,
"loss": 0.7214,
"step": 1229
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.4129270843682327,
"learning_rate": 0.0,
"loss": 0.7077,
"step": 1230
},
{
"epoch": 4.982278481012658,
"step": 1230,
"total_flos": 1.9490496700092252e+18,
"train_loss": 0.9533841308054886,
"train_runtime": 13422.2897,
"train_samples_per_second": 11.771,
"train_steps_per_second": 0.092
}
],
"logging_steps": 1,
"max_steps": 1230,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.9490496700092252e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}