a1_science_stackexchange_physics / trainer_state.json
ryanmarten's picture
End of training
4c431d2 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.988354430379747,
"eval_steps": 500,
"global_step": 1230,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004050632911392405,
"grad_norm": 7.357121025722606,
"learning_rate": 3.2520325203252037e-07,
"loss": 1.1877,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 7.424765382777217,
"learning_rate": 6.504065040650407e-07,
"loss": 1.2173,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 7.703759324247838,
"learning_rate": 9.75609756097561e-07,
"loss": 1.2125,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 7.3216789047170225,
"learning_rate": 1.3008130081300815e-06,
"loss": 1.1791,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 6.985487869542711,
"learning_rate": 1.6260162601626018e-06,
"loss": 1.1901,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 6.702135697027797,
"learning_rate": 1.951219512195122e-06,
"loss": 1.1481,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 5.424933225353772,
"learning_rate": 2.2764227642276426e-06,
"loss": 1.128,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 5.274712746336984,
"learning_rate": 2.601626016260163e-06,
"loss": 1.1381,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 2.7745991020686662,
"learning_rate": 2.926829268292683e-06,
"loss": 1.0773,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 2.5210679891631385,
"learning_rate": 3.2520325203252037e-06,
"loss": 1.0751,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 2.299882779507461,
"learning_rate": 3.577235772357724e-06,
"loss": 1.0389,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 4.955451099431155,
"learning_rate": 3.902439024390244e-06,
"loss": 1.0326,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 5.343951034877923,
"learning_rate": 4.227642276422765e-06,
"loss": 1.0537,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 5.220673387356522,
"learning_rate": 4.552845528455285e-06,
"loss": 1.0274,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 4.835025369077602,
"learning_rate": 4.8780487804878055e-06,
"loss": 1.0204,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 3.5129876249478285,
"learning_rate": 5.203252032520326e-06,
"loss": 1.0476,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 3.04161691010414,
"learning_rate": 5.528455284552846e-06,
"loss": 0.9732,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.7908189500856238,
"learning_rate": 5.853658536585366e-06,
"loss": 0.9433,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 2.387752254847135,
"learning_rate": 6.178861788617887e-06,
"loss": 0.9177,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 2.0445590766834325,
"learning_rate": 6.504065040650407e-06,
"loss": 0.9569,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.5863873040933747,
"learning_rate": 6.829268292682928e-06,
"loss": 0.9362,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.6426916187109923,
"learning_rate": 7.154471544715448e-06,
"loss": 0.9082,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.7624683329154551,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.908,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 1.3382521967342984,
"learning_rate": 7.804878048780489e-06,
"loss": 0.9154,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 1.0869325905959626,
"learning_rate": 8.130081300813009e-06,
"loss": 0.9087,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 1.1970531493322714,
"learning_rate": 8.45528455284553e-06,
"loss": 0.8849,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 1.081844072394682,
"learning_rate": 8.78048780487805e-06,
"loss": 0.8684,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 1.0005876500449296,
"learning_rate": 9.10569105691057e-06,
"loss": 0.882,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 1.0497129147682551,
"learning_rate": 9.43089430894309e-06,
"loss": 0.8547,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 0.9854204486582673,
"learning_rate": 9.756097560975611e-06,
"loss": 0.8944,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 0.9920899606976854,
"learning_rate": 1.008130081300813e-05,
"loss": 0.8874,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.9665649496236073,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.8582,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.9023081592867421,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.8366,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.8051412297572695,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.8344,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.7662034969436134,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.8342,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.7815915500456578,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.8419,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.6991095837514368,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.8578,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.8558149706554858,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.8234,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.8490650042423566,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.8594,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.7865199056786417,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.8373,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.8720073578958186,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.8247,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.8914677859554474,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.8113,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.9235957802194984,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.855,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.8505341280412683,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.827,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.736259098624615,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.8236,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.7028933268299694,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.8297,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.743262663669584,
"learning_rate": 1.528455284552846e-05,
"loss": 0.8545,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.8543328707031339,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.8287,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.8061238149284624,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.8117,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.7220903928212005,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.8086,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.9097763086100912,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.8273,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.8279023845714496,
"learning_rate": 1.691056910569106e-05,
"loss": 0.8262,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.8918953710491604,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.8064,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.7969136278451038,
"learning_rate": 1.75609756097561e-05,
"loss": 0.8389,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.8922149611649401,
"learning_rate": 1.788617886178862e-05,
"loss": 0.799,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.7582423434984019,
"learning_rate": 1.821138211382114e-05,
"loss": 0.8226,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.695730213021733,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.8328,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.9469935088508817,
"learning_rate": 1.886178861788618e-05,
"loss": 0.8303,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.6946570527886287,
"learning_rate": 1.91869918699187e-05,
"loss": 0.7976,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.9156548343356523,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.8119,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.7437626803537862,
"learning_rate": 1.983739837398374e-05,
"loss": 0.7991,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.8784365033590321,
"learning_rate": 2.016260162601626e-05,
"loss": 0.8152,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.843590639377991,
"learning_rate": 2.048780487804878e-05,
"loss": 0.8373,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.7588310224531974,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.8224,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 1.0544212035629266,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.7943,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.6992681795873665,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.7967,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.846009579000504,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.8156,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 0.7821499416641727,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.8038,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 0.7906404759576787,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.7994,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 0.8065452062682654,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.8161,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 0.8042715729579222,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.8102,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.8173881575083304,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.7889,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 0.8468038424783192,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.8252,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.8188543067438467,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.8284,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.8609571419715051,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.7814,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 0.7852462778576976,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.7697,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.8232443670192042,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.7913,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 0.7514776855104652,
"learning_rate": 2.536585365853659e-05,
"loss": 0.8002,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 0.782691943071837,
"learning_rate": 2.569105691056911e-05,
"loss": 0.8119,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 0.8652617712964507,
"learning_rate": 2.601626016260163e-05,
"loss": 0.8292,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 0.7989771416434857,
"learning_rate": 2.634146341463415e-05,
"loss": 0.8248,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 0.9022472770454814,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.8022,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 0.8938001863566319,
"learning_rate": 2.699186991869919e-05,
"loss": 0.7832,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 1.0563442446652955,
"learning_rate": 2.731707317073171e-05,
"loss": 0.8004,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 0.6721020717129841,
"learning_rate": 2.764227642276423e-05,
"loss": 0.7822,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 0.9302248147085038,
"learning_rate": 2.796747967479675e-05,
"loss": 0.8026,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 0.8130077054732896,
"learning_rate": 2.829268292682927e-05,
"loss": 0.7785,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 1.0667184898999946,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.8067,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 0.8893944862333222,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.8276,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 0.8379828070403038,
"learning_rate": 2.926829268292683e-05,
"loss": 0.793,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 1.024948393516918,
"learning_rate": 2.959349593495935e-05,
"loss": 0.7919,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 0.9955429548052227,
"learning_rate": 2.991869918699187e-05,
"loss": 0.8255,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 0.7741312596733728,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.7915,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 1.0017982551342928,
"learning_rate": 3.056910569105692e-05,
"loss": 0.8035,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 1.0866630871118075,
"learning_rate": 3.089430894308943e-05,
"loss": 0.7823,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 1.0270072423545096,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.8061,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 1.1424194476044773,
"learning_rate": 3.154471544715447e-05,
"loss": 0.8157,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 0.834227873992856,
"learning_rate": 3.186991869918699e-05,
"loss": 0.7856,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 0.9110526981365317,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.8399,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.8917117470507994,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.8156,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 0.9057422864285821,
"learning_rate": 3.284552845528456e-05,
"loss": 0.8078,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 0.9807364215258861,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.7843,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 0.9029018677011543,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.8031,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 0.8101907643613971,
"learning_rate": 3.382113821138212e-05,
"loss": 0.7973,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 1.0478103304245034,
"learning_rate": 3.414634146341463e-05,
"loss": 0.8061,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 0.9694809220377166,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.8251,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 0.8889038361709137,
"learning_rate": 3.479674796747968e-05,
"loss": 0.7895,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 0.7635788620787614,
"learning_rate": 3.51219512195122e-05,
"loss": 0.777,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 0.8474905538092927,
"learning_rate": 3.544715447154472e-05,
"loss": 0.776,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 0.9075769108705518,
"learning_rate": 3.577235772357724e-05,
"loss": 0.8025,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 1.0060895329232724,
"learning_rate": 3.609756097560976e-05,
"loss": 0.8061,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 0.939819311245968,
"learning_rate": 3.642276422764228e-05,
"loss": 0.8015,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 1.017262825821098,
"learning_rate": 3.67479674796748e-05,
"loss": 0.7837,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 0.8502262112819624,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.8067,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 0.9587306191789833,
"learning_rate": 3.739837398373984e-05,
"loss": 0.8394,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 0.8226568403225211,
"learning_rate": 3.772357723577236e-05,
"loss": 0.7879,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 0.9362277346908595,
"learning_rate": 3.804878048780488e-05,
"loss": 0.8083,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 1.0534578127476995,
"learning_rate": 3.83739837398374e-05,
"loss": 0.8292,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 1.1137744813863664,
"learning_rate": 3.869918699186992e-05,
"loss": 0.7927,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 1.1578440641893415,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.8028,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 0.9841913701077375,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.824,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 1.3266852979591688,
"learning_rate": 3.967479674796748e-05,
"loss": 0.8093,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 1.2844591044038312,
"learning_rate": 4e-05,
"loss": 0.7977,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 0.9300640620255223,
"learning_rate": 3.999991946137476e-05,
"loss": 0.8025,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 1.4638277883906021,
"learning_rate": 3.999967784614766e-05,
"loss": 0.8041,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 0.7785311572221093,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.797,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 1.2060884135106373,
"learning_rate": 3.999871139496895e-05,
"loss": 0.7946,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 0.8714851314065102,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.8065,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 0.9036950351806402,
"learning_rate": 3.999710067759846e-05,
"loss": 0.802,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 0.8183207948335134,
"learning_rate": 3.999605373449617e-05,
"loss": 0.8059,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.9961326425011064,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.8233,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 0.7540505486289353,
"learning_rate": 3.999347672161713e-05,
"loss": 0.8227,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 0.7965579423528135,
"learning_rate": 3.999194667259528e-05,
"loss": 0.7771,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 0.8408590281698622,
"learning_rate": 3.999025561118334e-05,
"loss": 0.8116,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 0.78011131014674,
"learning_rate": 3.998840355100086e-05,
"loss": 0.794,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 0.734049006276439,
"learning_rate": 3.998639050696409e-05,
"loss": 0.7877,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 0.761413239391788,
"learning_rate": 3.998421649528582e-05,
"loss": 0.7873,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 0.7547317014668593,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.7965,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.9106714067450988,
"learning_rate": 3.997938564033779e-05,
"loss": 0.8147,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 0.7565036747765088,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.7963,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.7145383294964522,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.7592,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 0.9662230930046013,
"learning_rate": 3.997093258045973e-05,
"loss": 0.7818,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.9111760994109757,
"learning_rate": 3.996779317598936e-05,
"loss": 0.7991,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 0.6238247882218976,
"learning_rate": 3.996449295365782e-05,
"loss": 0.7812,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.7876476359501811,
"learning_rate": 3.996103194004467e-05,
"loss": 0.792,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.8552701958092864,
"learning_rate": 3.995741016302441e-05,
"loss": 0.7982,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 0.8030236917310655,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.8171,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 0.7501304501965416,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.7949,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 0.8509240846665997,
"learning_rate": 3.994558054968643e-05,
"loss": 0.8136,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.791675665674136,
"learning_rate": 3.994131602367481e-05,
"loss": 0.7841,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 0.7974085470384731,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.7787,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.8010566652864607,
"learning_rate": 3.993230519343752e-05,
"loss": 0.8044,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 0.9024350326986876,
"learning_rate": 3.992755896178383e-05,
"loss": 0.815,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.8053881831130147,
"learning_rate": 3.992265223630981e-05,
"loss": 0.7934,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 0.8478443378133046,
"learning_rate": 3.991758505653355e-05,
"loss": 0.8187,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 1.0023560784762002,
"learning_rate": 3.991235746326543e-05,
"loss": 0.7623,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.7706689358761141,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.7936,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 0.8557884953655973,
"learning_rate": 3.990142120595444e-05,
"loss": 0.8003,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 1.337227635159225,
"learning_rate": 3.98957126299907e-05,
"loss": 0.7931,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 0.8195553360959965,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.8111,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 1.4739975745323137,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.7959,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 0.82086912611129,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.7366,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 1.388655451503468,
"learning_rate": 3.98712764319087e-05,
"loss": 0.8073,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.8093025927389879,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.7634,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 1.1709309662196867,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.7675,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.922270589951744,
"learning_rate": 3.985126869010129e-05,
"loss": 0.8063,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 1.227697417608106,
"learning_rate": 3.984427961114833e-05,
"loss": 0.7895,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 0.8764936060162462,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.7818,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 1.0860538054532902,
"learning_rate": 3.982982204151901e-05,
"loss": 0.8109,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.799188480169545,
"learning_rate": 3.982235366728193e-05,
"loss": 0.7931,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.68838330996928,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.7496,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 0.9658379962154666,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.7768,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.6962252659859032,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.8005,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.8145874480171655,
"learning_rate": 3.979088432599627e-05,
"loss": 0.795,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.723604651646915,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.7992,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 0.7887339805999639,
"learning_rate": 3.977419304032111e-05,
"loss": 0.7848,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 0.6651254918667732,
"learning_rate": 3.976560847560697e-05,
"loss": 0.8019,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 0.8464001758288876,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.7858,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 0.6233143186053487,
"learning_rate": 3.974796184811956e-05,
"loss": 0.7929,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.7696020640154256,
"learning_rate": 3.973889992746979e-05,
"loss": 0.7572,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.6368451305671496,
"learning_rate": 3.972967903243361e-05,
"loss": 0.7617,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.659919524911343,
"learning_rate": 3.972029923727486e-05,
"loss": 0.8022,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.7117791939755332,
"learning_rate": 3.971076061753709e-05,
"loss": 0.8102,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.7115543427465915,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.7852,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.7343953226179208,
"learning_rate": 3.969120721289402e-05,
"loss": 0.7653,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.7548585894892353,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.7902,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.8281594883467882,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.7728,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.6771452440238022,
"learning_rate": 3.966068788369422e-05,
"loss": 0.7943,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.7556959742760815,
"learning_rate": 3.965019797448622e-05,
"loss": 0.8048,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.9306962706834329,
"learning_rate": 3.963954980528515e-05,
"loss": 0.7937,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.8167957234762163,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.7688,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.696745286527455,
"learning_rate": 3.961777903121329e-05,
"loss": 0.7722,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.8054995569750728,
"learning_rate": 3.960665660168131e-05,
"loss": 0.7837,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.6736447400992428,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.7533,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.7524342486423747,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.7893,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 0.7234088718265472,
"learning_rate": 3.957234222185657e-05,
"loss": 0.783,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.7756984555016205,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.7544,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 0.6481773654868562,
"learning_rate": 3.954867765033605e-05,
"loss": 0.7864,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.778536493476031,
"learning_rate": 3.953660915306728e-05,
"loss": 0.7956,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.6441361668840987,
"learning_rate": 3.952438331063419e-05,
"loss": 0.7645,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.6451927857540642,
"learning_rate": 3.951200022150205e-05,
"loss": 0.7746,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.6385117785728711,
"learning_rate": 3.949945998540253e-05,
"loss": 0.7973,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.6902321648757327,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.7924,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.6289725021299989,
"learning_rate": 3.947390847755559e-05,
"loss": 0.7572,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.7373905189957886,
"learning_rate": 3.946089741159648e-05,
"loss": 0.7771,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.6157793357318533,
"learning_rate": 3.944772961024501e-05,
"loss": 0.7716,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.7298398286963342,
"learning_rate": 3.943440517955285e-05,
"loss": 0.7833,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.7988744258350667,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.7598,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.6854861224731701,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.7692,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.6419719999896315,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.7667,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.7055320314923652,
"learning_rate": 3.937954332854371e-05,
"loss": 0.7914,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.624855936984832,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.7713,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.7997610512466962,
"learning_rate": 3.935117547697387e-05,
"loss": 0.7544,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.6973924475484583,
"learning_rate": 3.933675771619675e-05,
"loss": 0.7757,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.7240542707060177,
"learning_rate": 3.932218421983131e-05,
"loss": 0.7778,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.8945028573289291,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.8088,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.6085267681107963,
"learning_rate": 3.929257049108054e-05,
"loss": 0.7593,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.7770175619103808,
"learning_rate": 3.927753049720011e-05,
"loss": 0.7721,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.7007855864772067,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.7921,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.6142061261021415,
"learning_rate": 3.92469848560784e-05,
"loss": 0.7717,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.6697106737363189,
"learning_rate": 3.923147945484751e-05,
"loss": 0.7977,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.5864461503867902,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.7769,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.6748267584297545,
"learning_rate": 3.920000411543654e-05,
"loss": 0.7632,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.6822573160031075,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.7588,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.6454979738042185,
"learning_rate": 3.916791024049648e-05,
"loss": 0.7485,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.730115367429912,
"learning_rate": 3.91516316745245e-05,
"loss": 0.7869,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.6016554860734091,
"learning_rate": 3.913519886394389e-05,
"loss": 0.7839,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.6466416112906492,
"learning_rate": 3.911861194110225e-05,
"loss": 0.7694,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.6144083454060468,
"learning_rate": 3.910187103958837e-05,
"loss": 0.7651,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.7027067241997592,
"learning_rate": 3.908497629423117e-05,
"loss": 0.7741,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.7190565088119352,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.7671,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.7416560970350694,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.7776,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.6126747908870802,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.7703,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.7885326837257709,
"learning_rate": 3.901586161429081e-05,
"loss": 0.7645,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.6356201862843519,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.7781,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.821065756484505,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.7851,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.706562266291596,
"learning_rate": 3.896241703466667e-05,
"loss": 0.7897,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.6618373795963769,
"learning_rate": 3.894429654085585e-05,
"loss": 0.7615,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.657014405511221,
"learning_rate": 3.892602347228505e-05,
"loss": 0.7684,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.6738902364171198,
"learning_rate": 3.890759797612307e-05,
"loss": 0.7665,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.7111591471607055,
"learning_rate": 3.888902020076632e-05,
"loss": 0.7895,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.6434016107045829,
"learning_rate": 3.887029029583764e-05,
"loss": 0.7849,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.6762561453064542,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.782,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.6822527373153331,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.7838,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.6037217367567999,
"learning_rate": 3.881318931821972e-05,
"loss": 0.7606,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.6043558270918458,
"learning_rate": 3.879385241571817e-05,
"loss": 0.7712,
"step": 246
},
{
"epoch": 1.0020253164556963,
"grad_norm": 0.6501754939836063,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.7041,
"step": 247
},
{
"epoch": 1.0060759493670886,
"grad_norm": 0.7738772941918096,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.6235,
"step": 248
},
{
"epoch": 1.010126582278481,
"grad_norm": 0.6733420315242706,
"learning_rate": 3.873493415863256e-05,
"loss": 0.6089,
"step": 249
},
{
"epoch": 1.0141772151898734,
"grad_norm": 0.8384559158285533,
"learning_rate": 3.871499275032111e-05,
"loss": 0.6044,
"step": 250
},
{
"epoch": 1.0182278481012659,
"grad_norm": 0.707134028514325,
"learning_rate": 3.869490061403091e-05,
"loss": 0.5887,
"step": 251
},
{
"epoch": 1.0222784810126582,
"grad_norm": 0.6697202186593402,
"learning_rate": 3.867465791158124e-05,
"loss": 0.5673,
"step": 252
},
{
"epoch": 1.0263291139240507,
"grad_norm": 0.6798218824595393,
"learning_rate": 3.865426480600407e-05,
"loss": 0.594,
"step": 253
},
{
"epoch": 1.030379746835443,
"grad_norm": 0.6750444225487667,
"learning_rate": 3.863372146154264e-05,
"loss": 0.6036,
"step": 254
},
{
"epoch": 1.0344303797468355,
"grad_norm": 0.6437099161237284,
"learning_rate": 3.861302804365024e-05,
"loss": 0.5939,
"step": 255
},
{
"epoch": 1.0384810126582278,
"grad_norm": 0.7567293994404617,
"learning_rate": 3.85921847189888e-05,
"loss": 0.5928,
"step": 256
},
{
"epoch": 1.0425316455696203,
"grad_norm": 0.6667958575065381,
"learning_rate": 3.85711916554276e-05,
"loss": 0.5757,
"step": 257
},
{
"epoch": 1.0465822784810126,
"grad_norm": 0.806308015946557,
"learning_rate": 3.85500490220419e-05,
"loss": 0.6131,
"step": 258
},
{
"epoch": 1.0506329113924051,
"grad_norm": 0.7412937747026481,
"learning_rate": 3.852875698911154e-05,
"loss": 0.608,
"step": 259
},
{
"epoch": 1.0546835443037974,
"grad_norm": 0.6546681540077655,
"learning_rate": 3.850731572811963e-05,
"loss": 0.6087,
"step": 260
},
{
"epoch": 1.05873417721519,
"grad_norm": 0.848864086632766,
"learning_rate": 3.848572541175116e-05,
"loss": 0.6086,
"step": 261
},
{
"epoch": 1.0627848101265822,
"grad_norm": 0.632306069712709,
"learning_rate": 3.846398621389154e-05,
"loss": 0.5992,
"step": 262
},
{
"epoch": 1.0668354430379747,
"grad_norm": 0.7442503955494355,
"learning_rate": 3.84420983096253e-05,
"loss": 0.6071,
"step": 263
},
{
"epoch": 1.070886075949367,
"grad_norm": 0.6408526091708983,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.5906,
"step": 264
},
{
"epoch": 1.0749367088607595,
"grad_norm": 0.8281678301460116,
"learning_rate": 3.839787708819787e-05,
"loss": 0.6026,
"step": 265
},
{
"epoch": 1.0789873417721518,
"grad_norm": 0.8105083791398079,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.5959,
"step": 266
},
{
"epoch": 1.0830379746835443,
"grad_norm": 0.7996285838867897,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.5791,
"step": 267
},
{
"epoch": 1.0870886075949366,
"grad_norm": 0.8738798837688567,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.5911,
"step": 268
},
{
"epoch": 1.0911392405063292,
"grad_norm": 0.6346735283928235,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.6063,
"step": 269
},
{
"epoch": 1.0951898734177214,
"grad_norm": 0.7865007667613132,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.5851,
"step": 270
},
{
"epoch": 1.099240506329114,
"grad_norm": 0.6208868282441605,
"learning_rate": 3.826166304956594e-05,
"loss": 0.5862,
"step": 271
},
{
"epoch": 1.1032911392405063,
"grad_norm": 0.63423802711674,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.6046,
"step": 272
},
{
"epoch": 1.1073417721518988,
"grad_norm": 0.6546539741767677,
"learning_rate": 3.821507978767479e-05,
"loss": 0.5885,
"step": 273
},
{
"epoch": 1.111392405063291,
"grad_norm": 0.6481519810845416,
"learning_rate": 3.819156801001693e-05,
"loss": 0.5947,
"step": 274
},
{
"epoch": 1.1154430379746836,
"grad_norm": 0.7295286332258266,
"learning_rate": 3.816790971997121e-05,
"loss": 0.5922,
"step": 275
},
{
"epoch": 1.1194936708860759,
"grad_norm": 0.7261917859363479,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.5896,
"step": 276
},
{
"epoch": 1.1235443037974684,
"grad_norm": 0.646196158853818,
"learning_rate": 3.81201543660571e-05,
"loss": 0.5981,
"step": 277
},
{
"epoch": 1.1275949367088607,
"grad_norm": 0.7669824756844837,
"learning_rate": 3.809605768680377e-05,
"loss": 0.5969,
"step": 278
},
{
"epoch": 1.1316455696202532,
"grad_norm": 0.5461003735703202,
"learning_rate": 3.807181526438958e-05,
"loss": 0.598,
"step": 279
},
{
"epoch": 1.1356962025316455,
"grad_norm": 0.7209000085590104,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.6097,
"step": 280
},
{
"epoch": 1.139746835443038,
"grad_norm": 0.6062853838790638,
"learning_rate": 3.802289397223145e-05,
"loss": 0.5927,
"step": 281
},
{
"epoch": 1.1437974683544303,
"grad_norm": 0.5868323179440023,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.6152,
"step": 282
},
{
"epoch": 1.1478481012658228,
"grad_norm": 0.7105400083757176,
"learning_rate": 3.797339206560096e-05,
"loss": 0.5958,
"step": 283
},
{
"epoch": 1.1518987341772151,
"grad_norm": 0.783324530265228,
"learning_rate": 3.794842387948027e-05,
"loss": 0.5804,
"step": 284
},
{
"epoch": 1.1559493670886076,
"grad_norm": 0.6361654259765143,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.5795,
"step": 285
},
{
"epoch": 1.16,
"grad_norm": 1.0408610272695302,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.6261,
"step": 286
},
{
"epoch": 1.1640506329113924,
"grad_norm": 0.7177491526473134,
"learning_rate": 3.787265280646825e-05,
"loss": 0.6051,
"step": 287
},
{
"epoch": 1.1681012658227847,
"grad_norm": 0.7461174114689153,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.6047,
"step": 288
},
{
"epoch": 1.1721518987341772,
"grad_norm": 0.8508971799941933,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.5778,
"step": 289
},
{
"epoch": 1.1762025316455695,
"grad_norm": 0.7391571718898076,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.6014,
"step": 290
},
{
"epoch": 1.180253164556962,
"grad_norm": 0.8062440891290986,
"learning_rate": 3.776961046830306e-05,
"loss": 0.6129,
"step": 291
},
{
"epoch": 1.1843037974683543,
"grad_norm": 0.6216866876062013,
"learning_rate": 3.774349157719215e-05,
"loss": 0.6068,
"step": 292
},
{
"epoch": 1.1883544303797469,
"grad_norm": 0.7470167485264986,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.6032,
"step": 293
},
{
"epoch": 1.1924050632911392,
"grad_norm": 0.7093847427248533,
"learning_rate": 3.769082529555359e-05,
"loss": 0.5944,
"step": 294
},
{
"epoch": 1.1964556962025317,
"grad_norm": 0.7468773509381677,
"learning_rate": 3.766427832919294e-05,
"loss": 0.601,
"step": 295
},
{
"epoch": 1.200506329113924,
"grad_norm": 0.5749797238451381,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.5759,
"step": 296
},
{
"epoch": 1.2045569620253165,
"grad_norm": 0.6745064413517762,
"learning_rate": 3.761075781441526e-05,
"loss": 0.5964,
"step": 297
},
{
"epoch": 1.2086075949367088,
"grad_norm": 0.7318016153125924,
"learning_rate": 3.75837846970451e-05,
"loss": 0.6043,
"step": 298
},
{
"epoch": 1.2126582278481013,
"grad_norm": 0.6202949298717851,
"learning_rate": 3.755666996229032e-05,
"loss": 0.5928,
"step": 299
},
{
"epoch": 1.2167088607594936,
"grad_norm": 0.7565209452005737,
"learning_rate": 3.752941382852927e-05,
"loss": 0.6148,
"step": 300
},
{
"epoch": 1.220759493670886,
"grad_norm": 0.6750931030577522,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.5962,
"step": 301
},
{
"epoch": 1.2248101265822784,
"grad_norm": 0.7322647131539731,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.6066,
"step": 302
},
{
"epoch": 1.228860759493671,
"grad_norm": 0.7141077043032026,
"learning_rate": 3.744679923406351e-05,
"loss": 0.5983,
"step": 303
},
{
"epoch": 1.2329113924050632,
"grad_norm": 0.7077037524104081,
"learning_rate": 3.741897971081043e-05,
"loss": 0.6012,
"step": 304
},
{
"epoch": 1.2369620253164557,
"grad_norm": 0.703114203251987,
"learning_rate": 3.739101989748946e-05,
"loss": 0.6206,
"step": 305
},
{
"epoch": 1.241012658227848,
"grad_norm": 0.779453533100084,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.6112,
"step": 306
},
{
"epoch": 1.2450632911392405,
"grad_norm": 0.639676887325618,
"learning_rate": 3.73346803025098e-05,
"loss": 0.6047,
"step": 307
},
{
"epoch": 1.2491139240506328,
"grad_norm": 0.8019385606801885,
"learning_rate": 3.730630097460247e-05,
"loss": 0.5779,
"step": 308
},
{
"epoch": 1.2531645569620253,
"grad_norm": 0.6939850611512531,
"learning_rate": 3.727778226412628e-05,
"loss": 0.6172,
"step": 309
},
{
"epoch": 1.2572151898734178,
"grad_norm": 0.899253534762381,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.6176,
"step": 310
},
{
"epoch": 1.2612658227848101,
"grad_norm": 0.5863773837280625,
"learning_rate": 3.722032761533114e-05,
"loss": 0.5958,
"step": 311
},
{
"epoch": 1.2653164556962024,
"grad_norm": 0.9407962840570052,
"learning_rate": 3.719139213974403e-05,
"loss": 0.6061,
"step": 312
},
{
"epoch": 1.269367088607595,
"grad_norm": 0.615350002240708,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.5805,
"step": 313
},
{
"epoch": 1.2734177215189875,
"grad_norm": 0.6948098310360096,
"learning_rate": 3.713310605140055e-05,
"loss": 0.6337,
"step": 314
},
{
"epoch": 1.2774683544303798,
"grad_norm": 0.7498586859517408,
"learning_rate": 3.710375590807233e-05,
"loss": 0.6072,
"step": 315
},
{
"epoch": 1.281518987341772,
"grad_norm": 0.5647974118074541,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.6093,
"step": 316
},
{
"epoch": 1.2855696202531646,
"grad_norm": 0.6742006365530202,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.5939,
"step": 317
},
{
"epoch": 1.289620253164557,
"grad_norm": 0.5992680451767054,
"learning_rate": 3.701487992136854e-05,
"loss": 0.6211,
"step": 318
},
{
"epoch": 1.2936708860759494,
"grad_norm": 0.6202567968838767,
"learning_rate": 3.69849802022222e-05,
"loss": 0.607,
"step": 319
},
{
"epoch": 1.2977215189873417,
"grad_norm": 0.67853178843719,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.6083,
"step": 320
},
{
"epoch": 1.3017721518987342,
"grad_norm": 0.672806240887955,
"learning_rate": 3.692477062175289e-05,
"loss": 0.5953,
"step": 321
},
{
"epoch": 1.3058227848101267,
"grad_norm": 0.6384489208057724,
"learning_rate": 3.689446124534958e-05,
"loss": 0.5901,
"step": 322
},
{
"epoch": 1.309873417721519,
"grad_norm": 0.6179764352717287,
"learning_rate": 3.686401580327799e-05,
"loss": 0.5812,
"step": 323
},
{
"epoch": 1.3139240506329113,
"grad_norm": 0.755630658535331,
"learning_rate": 3.683343454074149e-05,
"loss": 0.6072,
"step": 324
},
{
"epoch": 1.3179746835443038,
"grad_norm": 0.6826742177671765,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.6047,
"step": 325
},
{
"epoch": 1.3220253164556963,
"grad_norm": 0.6424057103719925,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.6234,
"step": 326
},
{
"epoch": 1.3260759493670886,
"grad_norm": 0.6772668274844256,
"learning_rate": 3.674087829877297e-05,
"loss": 0.5945,
"step": 327
},
{
"epoch": 1.330126582278481,
"grad_norm": 0.603795393237412,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.6009,
"step": 328
},
{
"epoch": 1.3341772151898734,
"grad_norm": 0.6156016904817295,
"learning_rate": 3.667849957966501e-05,
"loss": 0.5838,
"step": 329
},
{
"epoch": 1.338227848101266,
"grad_norm": 0.7202628350064344,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.6014,
"step": 330
},
{
"epoch": 1.3422784810126582,
"grad_norm": 0.5924384743393264,
"learning_rate": 3.661558355626795e-05,
"loss": 0.5888,
"step": 331
},
{
"epoch": 1.3463291139240505,
"grad_norm": 0.6082283911849152,
"learning_rate": 3.658392468818163e-05,
"loss": 0.6014,
"step": 332
},
{
"epoch": 1.350379746835443,
"grad_norm": 0.6341384967313501,
"learning_rate": 3.655213225544574e-05,
"loss": 0.6101,
"step": 333
},
{
"epoch": 1.3544303797468356,
"grad_norm": 0.5822634362224939,
"learning_rate": 3.652020651411218e-05,
"loss": 0.5858,
"step": 334
},
{
"epoch": 1.3584810126582278,
"grad_norm": 0.7335812393728912,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.5968,
"step": 335
},
{
"epoch": 1.3625316455696201,
"grad_norm": 0.7384155496884599,
"learning_rate": 3.645595613522574e-05,
"loss": 0.6116,
"step": 336
},
{
"epoch": 1.3665822784810127,
"grad_norm": 0.7193793720907526,
"learning_rate": 3.642363201513657e-05,
"loss": 0.6154,
"step": 337
},
{
"epoch": 1.3706329113924052,
"grad_norm": 0.5830254836180223,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.6217,
"step": 338
},
{
"epoch": 1.3746835443037975,
"grad_norm": 0.5993967847372769,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.5961,
"step": 339
},
{
"epoch": 1.3787341772151898,
"grad_norm": 0.5566326528633994,
"learning_rate": 3.632586705948318e-05,
"loss": 0.5761,
"step": 340
},
{
"epoch": 1.3827848101265823,
"grad_norm": 0.6108377546920231,
"learning_rate": 3.629301541734311e-05,
"loss": 0.584,
"step": 341
},
{
"epoch": 1.3868354430379748,
"grad_norm": 0.5679798764402259,
"learning_rate": 3.626003255349676e-05,
"loss": 0.6196,
"step": 342
},
{
"epoch": 1.390886075949367,
"grad_norm": 0.7120506110585241,
"learning_rate": 3.622691873358357e-05,
"loss": 0.6166,
"step": 343
},
{
"epoch": 1.3949367088607594,
"grad_norm": 0.6334484824238474,
"learning_rate": 3.61936742242977e-05,
"loss": 0.6081,
"step": 344
},
{
"epoch": 1.398987341772152,
"grad_norm": 0.7025775030322958,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.6157,
"step": 345
},
{
"epoch": 1.4030379746835444,
"grad_norm": 0.7610012846826253,
"learning_rate": 3.612679420964516e-05,
"loss": 0.6175,
"step": 346
},
{
"epoch": 1.4070886075949367,
"grad_norm": 0.5827284596453448,
"learning_rate": 3.609315924292092e-05,
"loss": 0.6101,
"step": 347
},
{
"epoch": 1.411139240506329,
"grad_norm": 0.7145869119403246,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.5899,
"step": 348
},
{
"epoch": 1.4151898734177215,
"grad_norm": 0.5590346294027106,
"learning_rate": 3.602550074513133e-05,
"loss": 0.6046,
"step": 349
},
{
"epoch": 1.419240506329114,
"grad_norm": 0.6863194364713018,
"learning_rate": 3.599147775897822e-05,
"loss": 0.5922,
"step": 350
},
{
"epoch": 1.4232911392405063,
"grad_norm": 0.5712314496604768,
"learning_rate": 3.595732597966167e-05,
"loss": 0.6229,
"step": 351
},
{
"epoch": 1.4273417721518986,
"grad_norm": 0.7704283069023754,
"learning_rate": 3.592304568223542e-05,
"loss": 0.6064,
"step": 352
},
{
"epoch": 1.4313924050632911,
"grad_norm": 0.5732702557863321,
"learning_rate": 3.588863714278826e-05,
"loss": 0.6073,
"step": 353
},
{
"epoch": 1.4354430379746836,
"grad_norm": 0.6123510785066215,
"learning_rate": 3.585410063844186e-05,
"loss": 0.5788,
"step": 354
},
{
"epoch": 1.439493670886076,
"grad_norm": 0.6272340983152361,
"learning_rate": 3.581943644734846e-05,
"loss": 0.5953,
"step": 355
},
{
"epoch": 1.4435443037974682,
"grad_norm": 0.6604503910103985,
"learning_rate": 3.578464484868869e-05,
"loss": 0.6028,
"step": 356
},
{
"epoch": 1.4475949367088607,
"grad_norm": 0.7012032823698879,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.5855,
"step": 357
},
{
"epoch": 1.4516455696202533,
"grad_norm": 0.7089679588740688,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.6349,
"step": 358
},
{
"epoch": 1.4556962025316456,
"grad_norm": 0.6906930148159617,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.5878,
"step": 359
},
{
"epoch": 1.4597468354430378,
"grad_norm": 0.7165738414970864,
"learning_rate": 3.564420999786543e-05,
"loss": 0.6145,
"step": 360
},
{
"epoch": 1.4637974683544304,
"grad_norm": 0.5824035174064334,
"learning_rate": 3.560878558491842e-05,
"loss": 0.5962,
"step": 361
},
{
"epoch": 1.4678481012658229,
"grad_norm": 0.7508065128348818,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.5886,
"step": 362
},
{
"epoch": 1.4718987341772152,
"grad_norm": 0.6657023686286115,
"learning_rate": 3.553755991230039e-05,
"loss": 0.6292,
"step": 363
},
{
"epoch": 1.4759493670886075,
"grad_norm": 0.7050378645904307,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.6049,
"step": 364
},
{
"epoch": 1.48,
"grad_norm": 0.7970903820948606,
"learning_rate": 3.546583369120419e-05,
"loss": 0.6145,
"step": 365
},
{
"epoch": 1.4840506329113925,
"grad_norm": 0.7236943657627367,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.6202,
"step": 366
},
{
"epoch": 1.4881012658227848,
"grad_norm": 0.7527595565691856,
"learning_rate": 3.539360923231766e-05,
"loss": 0.6031,
"step": 367
},
{
"epoch": 1.492151898734177,
"grad_norm": 0.7097147253514303,
"learning_rate": 3.535731089018394e-05,
"loss": 0.5931,
"step": 368
},
{
"epoch": 1.4962025316455696,
"grad_norm": 0.7011917981290551,
"learning_rate": 3.532088886237956e-05,
"loss": 0.6155,
"step": 369
},
{
"epoch": 1.5002531645569621,
"grad_norm": 0.7078927466857321,
"learning_rate": 3.528434344224253e-05,
"loss": 0.6006,
"step": 370
},
{
"epoch": 1.5043037974683544,
"grad_norm": 0.6440860225303101,
"learning_rate": 3.524767492410464e-05,
"loss": 0.6011,
"step": 371
},
{
"epoch": 1.5083544303797467,
"grad_norm": 0.694673837408342,
"learning_rate": 3.521088360328908e-05,
"loss": 0.6141,
"step": 372
},
{
"epoch": 1.5124050632911392,
"grad_norm": 0.670092564063503,
"learning_rate": 3.517396977610811e-05,
"loss": 0.5938,
"step": 373
},
{
"epoch": 1.5164556962025317,
"grad_norm": 0.5732120696509188,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.6128,
"step": 374
},
{
"epoch": 1.520506329113924,
"grad_norm": 0.6663479438733173,
"learning_rate": 3.509977579282971e-05,
"loss": 0.6236,
"step": 375
},
{
"epoch": 1.5245569620253163,
"grad_norm": 0.5433416576979395,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.6101,
"step": 376
},
{
"epoch": 1.5286075949367088,
"grad_norm": 0.613030099671511,
"learning_rate": 3.502509536445719e-05,
"loss": 0.6087,
"step": 377
},
{
"epoch": 1.5326582278481014,
"grad_norm": 0.5811948026972424,
"learning_rate": 3.498757348458147e-05,
"loss": 0.6184,
"step": 378
},
{
"epoch": 1.5367088607594936,
"grad_norm": 0.5378762697000645,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.5977,
"step": 379
},
{
"epoch": 1.540759493670886,
"grad_norm": 0.6637150959789496,
"learning_rate": 3.491216790442899e-05,
"loss": 0.6158,
"step": 380
},
{
"epoch": 1.5448101265822785,
"grad_norm": 0.5980611168312987,
"learning_rate": 3.487428481145839e-05,
"loss": 0.5917,
"step": 381
},
{
"epoch": 1.548860759493671,
"grad_norm": 0.5984168745248585,
"learning_rate": 3.483628192304278e-05,
"loss": 0.5761,
"step": 382
},
{
"epoch": 1.5529113924050633,
"grad_norm": 0.671227404618925,
"learning_rate": 3.479815954525219e-05,
"loss": 0.6141,
"step": 383
},
{
"epoch": 1.5569620253164556,
"grad_norm": 0.5794081819914237,
"learning_rate": 3.475991798511899e-05,
"loss": 0.6183,
"step": 384
},
{
"epoch": 1.561012658227848,
"grad_norm": 0.6778216481093903,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.6195,
"step": 385
},
{
"epoch": 1.5650632911392406,
"grad_norm": 0.6267287061600928,
"learning_rate": 3.468307855075128e-05,
"loss": 0.6086,
"step": 386
},
{
"epoch": 1.5691139240506329,
"grad_norm": 0.6093102541166201,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.614,
"step": 387
},
{
"epoch": 1.5731645569620252,
"grad_norm": 0.5959480796024088,
"learning_rate": 3.460576609535163e-05,
"loss": 0.6031,
"step": 388
},
{
"epoch": 1.5772151898734177,
"grad_norm": 0.549557027286607,
"learning_rate": 3.456693326250006e-05,
"loss": 0.605,
"step": 389
},
{
"epoch": 1.5812658227848102,
"grad_norm": 0.5497780312300823,
"learning_rate": 3.452798310957058e-05,
"loss": 0.6078,
"step": 390
},
{
"epoch": 1.5853164556962025,
"grad_norm": 0.5337392677809145,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.5937,
"step": 391
},
{
"epoch": 1.5893670886075948,
"grad_norm": 0.6374976312989193,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.6213,
"step": 392
},
{
"epoch": 1.5934177215189873,
"grad_norm": 0.5619752992255109,
"learning_rate": 3.441043187201574e-05,
"loss": 0.6279,
"step": 393
},
{
"epoch": 1.5974683544303798,
"grad_norm": 0.6113837439773606,
"learning_rate": 3.437101558517728e-05,
"loss": 0.5953,
"step": 394
},
{
"epoch": 1.6015189873417721,
"grad_norm": 0.5679820255144208,
"learning_rate": 3.433148355615496e-05,
"loss": 0.6146,
"step": 395
},
{
"epoch": 1.6055696202531644,
"grad_norm": 0.6059946337754479,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.6255,
"step": 396
},
{
"epoch": 1.609620253164557,
"grad_norm": 0.6160071668122475,
"learning_rate": 3.425207354603043e-05,
"loss": 0.6107,
"step": 397
},
{
"epoch": 1.6136708860759494,
"grad_norm": 0.5734656095164802,
"learning_rate": 3.421219620448553e-05,
"loss": 0.6078,
"step": 398
},
{
"epoch": 1.6177215189873417,
"grad_norm": 0.6839031084062255,
"learning_rate": 3.417220439986623e-05,
"loss": 0.5984,
"step": 399
},
{
"epoch": 1.621772151898734,
"grad_norm": 0.6193690770211241,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.6007,
"step": 400
},
{
"epoch": 1.6258227848101265,
"grad_norm": 0.6469268709591406,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.6128,
"step": 401
},
{
"epoch": 1.629873417721519,
"grad_norm": 0.7229082078490272,
"learning_rate": 3.405154543304065e-05,
"loss": 0.6126,
"step": 402
},
{
"epoch": 1.6339240506329114,
"grad_norm": 0.6278994803004404,
"learning_rate": 3.401109900618843e-05,
"loss": 0.6057,
"step": 403
},
{
"epoch": 1.6379746835443036,
"grad_norm": 0.5840175615971783,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.6232,
"step": 404
},
{
"epoch": 1.6420253164556962,
"grad_norm": 0.6068092425461267,
"learning_rate": 3.392986794874714e-05,
"loss": 0.6139,
"step": 405
},
{
"epoch": 1.6460759493670887,
"grad_norm": 0.5967104985226023,
"learning_rate": 3.388908397238184e-05,
"loss": 0.5867,
"step": 406
},
{
"epoch": 1.650126582278481,
"grad_norm": 0.5237752383651417,
"learning_rate": 3.384818813524362e-05,
"loss": 0.5921,
"step": 407
},
{
"epoch": 1.6541772151898733,
"grad_norm": 0.5977518362668297,
"learning_rate": 3.380718076670195e-05,
"loss": 0.617,
"step": 408
},
{
"epoch": 1.6582278481012658,
"grad_norm": 0.6065004761902109,
"learning_rate": 3.376606219702454e-05,
"loss": 0.6041,
"step": 409
},
{
"epoch": 1.6622784810126583,
"grad_norm": 0.581793080453738,
"learning_rate": 3.372483275737468e-05,
"loss": 0.5984,
"step": 410
},
{
"epoch": 1.6663291139240506,
"grad_norm": 0.5993771823144365,
"learning_rate": 3.368349277980861e-05,
"loss": 0.6108,
"step": 411
},
{
"epoch": 1.6703797468354429,
"grad_norm": 0.5922514963507861,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.6145,
"step": 412
},
{
"epoch": 1.6744303797468354,
"grad_norm": 0.5745436786789125,
"learning_rate": 3.360048254360144e-05,
"loss": 0.6067,
"step": 413
},
{
"epoch": 1.678481012658228,
"grad_norm": 0.5440056431273484,
"learning_rate": 3.355881295351336e-05,
"loss": 0.606,
"step": 414
},
{
"epoch": 1.6825316455696202,
"grad_norm": 0.5798944043172437,
"learning_rate": 3.351703416260975e-05,
"loss": 0.6242,
"step": 415
},
{
"epoch": 1.6865822784810125,
"grad_norm": 0.6164771165123981,
"learning_rate": 3.347514650737126e-05,
"loss": 0.6092,
"step": 416
},
{
"epoch": 1.690632911392405,
"grad_norm": 0.5282409694655134,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.5954,
"step": 417
},
{
"epoch": 1.6946835443037975,
"grad_norm": 0.5744124293809671,
"learning_rate": 3.339104595419334e-05,
"loss": 0.593,
"step": 418
},
{
"epoch": 1.6987341772151898,
"grad_norm": 0.5964589080097652,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.623,
"step": 419
},
{
"epoch": 1.7027848101265821,
"grad_norm": 0.51818399314006,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.5958,
"step": 420
},
{
"epoch": 1.7068354430379746,
"grad_norm": 0.5601441077298995,
"learning_rate": 3.326408710419996e-05,
"loss": 0.5914,
"step": 421
},
{
"epoch": 1.7108860759493671,
"grad_norm": 0.5356751801108312,
"learning_rate": 3.322155337795454e-05,
"loss": 0.6125,
"step": 422
},
{
"epoch": 1.7149367088607594,
"grad_norm": 0.5685429045130905,
"learning_rate": 3.317891316713587e-05,
"loss": 0.6035,
"step": 423
},
{
"epoch": 1.7189873417721517,
"grad_norm": 0.5254243260310507,
"learning_rate": 3.313616681516231e-05,
"loss": 0.6076,
"step": 424
},
{
"epoch": 1.7230379746835442,
"grad_norm": 0.5437745224631836,
"learning_rate": 3.309331466630713e-05,
"loss": 0.6279,
"step": 425
},
{
"epoch": 1.7270886075949368,
"grad_norm": 0.6081312945480227,
"learning_rate": 3.305035706569563e-05,
"loss": 0.6121,
"step": 426
},
{
"epoch": 1.7311392405063293,
"grad_norm": 0.5928595088794343,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.5723,
"step": 427
},
{
"epoch": 1.7351898734177216,
"grad_norm": 0.5782198984870364,
"learning_rate": 3.296412689394864e-05,
"loss": 0.6039,
"step": 428
},
{
"epoch": 1.7392405063291139,
"grad_norm": 0.546340240905146,
"learning_rate": 3.292085501729909e-05,
"loss": 0.6218,
"step": 429
},
{
"epoch": 1.7432911392405064,
"grad_norm": 0.5699410678063123,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.6049,
"step": 430
},
{
"epoch": 1.747341772151899,
"grad_norm": 0.5255145187051962,
"learning_rate": 3.283399942497381e-05,
"loss": 0.6074,
"step": 431
},
{
"epoch": 1.7513924050632912,
"grad_norm": 0.5642274035372045,
"learning_rate": 3.279041640882108e-05,
"loss": 0.6252,
"step": 432
},
{
"epoch": 1.7554430379746835,
"grad_norm": 0.5441428025148912,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.5908,
"step": 433
},
{
"epoch": 1.759493670886076,
"grad_norm": 0.6358066130919088,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.6153,
"step": 434
},
{
"epoch": 1.7635443037974685,
"grad_norm": 0.5476132870983478,
"learning_rate": 3.265905069502244e-05,
"loss": 0.6037,
"step": 435
},
{
"epoch": 1.7675949367088608,
"grad_norm": 0.6330577903543103,
"learning_rate": 3.261505774420016e-05,
"loss": 0.5951,
"step": 436
},
{
"epoch": 1.771645569620253,
"grad_norm": 0.6728543964899755,
"learning_rate": 3.257096319343707e-05,
"loss": 0.6124,
"step": 437
},
{
"epoch": 1.7756962025316456,
"grad_norm": 0.6118046869075859,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.6202,
"step": 438
},
{
"epoch": 1.7797468354430381,
"grad_norm": 0.5865845665437652,
"learning_rate": 3.248247071342966e-05,
"loss": 0.6077,
"step": 439
},
{
"epoch": 1.7837974683544304,
"grad_norm": 0.6339097657082721,
"learning_rate": 3.243807349689161e-05,
"loss": 0.5854,
"step": 440
},
{
"epoch": 1.7878481012658227,
"grad_norm": 0.5596694121427727,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.6191,
"step": 441
},
{
"epoch": 1.7918987341772152,
"grad_norm": 0.6307126241520362,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.6021,
"step": 442
},
{
"epoch": 1.7959493670886078,
"grad_norm": 0.527709292872088,
"learning_rate": 3.230428223438075e-05,
"loss": 0.6015,
"step": 443
},
{
"epoch": 1.8,
"grad_norm": 0.6480199316432507,
"learning_rate": 3.225948647317459e-05,
"loss": 0.6085,
"step": 444
},
{
"epoch": 1.8040506329113923,
"grad_norm": 0.5619710041426729,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.5987,
"step": 445
},
{
"epoch": 1.8081012658227849,
"grad_norm": 0.5428997827165996,
"learning_rate": 3.216959910368034e-05,
"loss": 0.6373,
"step": 446
},
{
"epoch": 1.8121518987341774,
"grad_norm": 0.5739659221926049,
"learning_rate": 3.212450821933277e-05,
"loss": 0.6081,
"step": 447
},
{
"epoch": 1.8162025316455697,
"grad_norm": 0.5614887614150708,
"learning_rate": 3.207931968586281e-05,
"loss": 0.6138,
"step": 448
},
{
"epoch": 1.820253164556962,
"grad_norm": 0.5998846838816598,
"learning_rate": 3.203403386721272e-05,
"loss": 0.5875,
"step": 449
},
{
"epoch": 1.8243037974683545,
"grad_norm": 0.5729510039142331,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.596,
"step": 450
},
{
"epoch": 1.828354430379747,
"grad_norm": 0.6655029598224216,
"learning_rate": 3.194317183405573e-05,
"loss": 0.6076,
"step": 451
},
{
"epoch": 1.8324050632911393,
"grad_norm": 0.576025892280621,
"learning_rate": 3.189759635133914e-05,
"loss": 0.6077,
"step": 452
},
{
"epoch": 1.8364556962025316,
"grad_norm": 0.6739308104205395,
"learning_rate": 3.185192504701718e-05,
"loss": 0.6002,
"step": 453
},
{
"epoch": 1.840506329113924,
"grad_norm": 0.6258087752016261,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.6459,
"step": 454
},
{
"epoch": 1.8445569620253166,
"grad_norm": 0.6871876733955403,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.6136,
"step": 455
},
{
"epoch": 1.848607594936709,
"grad_norm": 0.6252269010943755,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.6095,
"step": 456
},
{
"epoch": 1.8526582278481012,
"grad_norm": 0.5859013311198242,
"learning_rate": 3.166828898179731e-05,
"loss": 0.616,
"step": 457
},
{
"epoch": 1.8567088607594937,
"grad_norm": 0.6385298984307003,
"learning_rate": 3.162214410223536e-05,
"loss": 0.5975,
"step": 458
},
{
"epoch": 1.8607594936708862,
"grad_norm": 0.5784250372597951,
"learning_rate": 3.157590561952257e-05,
"loss": 0.5846,
"step": 459
},
{
"epoch": 1.8648101265822785,
"grad_norm": 0.5765898754594333,
"learning_rate": 3.152957390605732e-05,
"loss": 0.5819,
"step": 460
},
{
"epoch": 1.8688607594936708,
"grad_norm": 0.6217365708577794,
"learning_rate": 3.148314933498886e-05,
"loss": 0.6155,
"step": 461
},
{
"epoch": 1.8729113924050633,
"grad_norm": 0.6233388873651117,
"learning_rate": 3.143663228021431e-05,
"loss": 0.5852,
"step": 462
},
{
"epoch": 1.8769620253164558,
"grad_norm": 0.5846271475682301,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.6173,
"step": 463
},
{
"epoch": 1.8810126582278481,
"grad_norm": 0.5892144409607464,
"learning_rate": 3.134332221885661e-05,
"loss": 0.6063,
"step": 464
},
{
"epoch": 1.8850632911392404,
"grad_norm": 0.5779910922664048,
"learning_rate": 3.129652996377987e-05,
"loss": 0.6188,
"step": 465
},
{
"epoch": 1.889113924050633,
"grad_norm": 0.6358143401301061,
"learning_rate": 3.12496467280038e-05,
"loss": 0.608,
"step": 466
},
{
"epoch": 1.8931645569620255,
"grad_norm": 0.561644098298251,
"learning_rate": 3.120267288911952e-05,
"loss": 0.5925,
"step": 467
},
{
"epoch": 1.8972151898734178,
"grad_norm": 0.6331708238551711,
"learning_rate": 3.11556088254479e-05,
"loss": 0.5949,
"step": 468
},
{
"epoch": 1.90126582278481,
"grad_norm": 0.6066112001840233,
"learning_rate": 3.11084549160364e-05,
"loss": 0.6044,
"step": 469
},
{
"epoch": 1.9053164556962026,
"grad_norm": 0.6944750934044569,
"learning_rate": 3.106121154065615e-05,
"loss": 0.5936,
"step": 470
},
{
"epoch": 1.909367088607595,
"grad_norm": 0.5481340556487301,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.5805,
"step": 471
},
{
"epoch": 1.9134177215189874,
"grad_norm": 0.7988415063731097,
"learning_rate": 3.096645791467348e-05,
"loss": 0.6132,
"step": 472
},
{
"epoch": 1.9174683544303797,
"grad_norm": 0.55446212627412,
"learning_rate": 3.091894842720373e-05,
"loss": 0.5992,
"step": 473
},
{
"epoch": 1.9215189873417722,
"grad_norm": 0.6367604975145649,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.6149,
"step": 474
},
{
"epoch": 1.9255696202531647,
"grad_norm": 0.549691422624577,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.6042,
"step": 475
},
{
"epoch": 1.929620253164557,
"grad_norm": 0.645427836236271,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.6206,
"step": 476
},
{
"epoch": 1.9336708860759493,
"grad_norm": 0.5645109776638019,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.6038,
"step": 477
},
{
"epoch": 1.9377215189873418,
"grad_norm": 0.5639421228632894,
"learning_rate": 3.06800895716334e-05,
"loss": 0.5977,
"step": 478
},
{
"epoch": 1.9417721518987343,
"grad_norm": 0.5356946401817504,
"learning_rate": 3.063205821011029e-05,
"loss": 0.5905,
"step": 479
},
{
"epoch": 1.9458227848101266,
"grad_norm": 0.5863353464837168,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.6207,
"step": 480
},
{
"epoch": 1.949873417721519,
"grad_norm": 0.5755274822763732,
"learning_rate": 3.053573898718618e-05,
"loss": 0.5945,
"step": 481
},
{
"epoch": 1.9539240506329114,
"grad_norm": 0.5182659641346812,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.6109,
"step": 482
},
{
"epoch": 1.957974683544304,
"grad_norm": 0.5620409516789111,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.5985,
"step": 483
},
{
"epoch": 1.9620253164556962,
"grad_norm": 0.5159768742553567,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.6113,
"step": 484
},
{
"epoch": 1.9660759493670885,
"grad_norm": 0.5763983203599712,
"learning_rate": 3.034208541656255e-05,
"loss": 0.6172,
"step": 485
},
{
"epoch": 1.970126582278481,
"grad_norm": 0.6059907249978669,
"learning_rate": 3.029346281309218e-05,
"loss": 0.597,
"step": 486
},
{
"epoch": 1.9741772151898735,
"grad_norm": 0.5681463819972302,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.5813,
"step": 487
},
{
"epoch": 1.9782278481012658,
"grad_norm": 0.5903297200132464,
"learning_rate": 3.019596929201569e-05,
"loss": 0.5993,
"step": 488
},
{
"epoch": 1.9822784810126581,
"grad_norm": 0.5472165103625979,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.6107,
"step": 489
},
{
"epoch": 1.9863291139240506,
"grad_norm": 0.5409166524429191,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.6048,
"step": 490
},
{
"epoch": 1.9903797468354432,
"grad_norm": 0.5271960967613964,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.599,
"step": 491
},
{
"epoch": 1.9944303797468355,
"grad_norm": 0.5159857738019016,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.5834,
"step": 492
},
{
"epoch": 1.9984810126582278,
"grad_norm": 0.5249161556246615,
"learning_rate": 2.995080534235264e-05,
"loss": 0.5948,
"step": 493
},
{
"epoch": 2.0040506329113925,
"grad_norm": 0.805636671983596,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.4287,
"step": 494
},
{
"epoch": 2.008101265822785,
"grad_norm": 0.5333166774187614,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.3818,
"step": 495
},
{
"epoch": 2.012151898734177,
"grad_norm": 1.698956578740472,
"learning_rate": 2.980274210295326e-05,
"loss": 0.416,
"step": 496
},
{
"epoch": 2.0162025316455696,
"grad_norm": 0.6331047225603663,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.3838,
"step": 497
},
{
"epoch": 2.020253164556962,
"grad_norm": 0.6793672620020873,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.4027,
"step": 498
},
{
"epoch": 2.0243037974683546,
"grad_norm": 0.6500979691336092,
"learning_rate": 2.965396831793362e-05,
"loss": 0.3836,
"step": 499
},
{
"epoch": 2.0283544303797467,
"grad_norm": 0.5663320591813614,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.3877,
"step": 500
},
{
"epoch": 2.0324050632911392,
"grad_norm": 0.6275899957198964,
"learning_rate": 2.955439637024526e-05,
"loss": 0.4158,
"step": 501
},
{
"epoch": 2.0364556962025318,
"grad_norm": 0.6319866640117379,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.3779,
"step": 502
},
{
"epoch": 2.0405063291139243,
"grad_norm": 0.6073676919277431,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.4024,
"step": 503
},
{
"epoch": 2.0445569620253163,
"grad_norm": 0.5718587643321494,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.3876,
"step": 504
},
{
"epoch": 2.048607594936709,
"grad_norm": 0.6369923034618779,
"learning_rate": 2.93543322968538e-05,
"loss": 0.3639,
"step": 505
},
{
"epoch": 2.0526582278481014,
"grad_norm": 0.5578392286902798,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.3667,
"step": 506
},
{
"epoch": 2.056708860759494,
"grad_norm": 0.5778181474011912,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.387,
"step": 507
},
{
"epoch": 2.060759493670886,
"grad_norm": 0.5515657117839141,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.387,
"step": 508
},
{
"epoch": 2.0648101265822785,
"grad_norm": 0.5613774667754111,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.403,
"step": 509
},
{
"epoch": 2.068860759493671,
"grad_norm": 0.5243598309066884,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.3864,
"step": 510
},
{
"epoch": 2.0729113924050635,
"grad_norm": 0.5584777358701893,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.3946,
"step": 511
},
{
"epoch": 2.0769620253164556,
"grad_norm": 0.4960383328486839,
"learning_rate": 2.900133526131588e-05,
"loss": 0.395,
"step": 512
},
{
"epoch": 2.081012658227848,
"grad_norm": 0.5413589065830722,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.3882,
"step": 513
},
{
"epoch": 2.0850632911392406,
"grad_norm": 0.5182574559003844,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.3592,
"step": 514
},
{
"epoch": 2.089113924050633,
"grad_norm": 0.5408837758422048,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.3916,
"step": 515
},
{
"epoch": 2.093164556962025,
"grad_norm": 0.5100782310161501,
"learning_rate": 2.879801882631847e-05,
"loss": 0.3953,
"step": 516
},
{
"epoch": 2.0972151898734177,
"grad_norm": 0.5099483961641167,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.4053,
"step": 517
},
{
"epoch": 2.1012658227848102,
"grad_norm": 0.5248650501614557,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.368,
"step": 518
},
{
"epoch": 2.1053164556962027,
"grad_norm": 0.5072726224726579,
"learning_rate": 2.864478605989494e-05,
"loss": 0.3952,
"step": 519
},
{
"epoch": 2.109367088607595,
"grad_norm": 0.5365037330718331,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.387,
"step": 520
},
{
"epoch": 2.1134177215189873,
"grad_norm": 0.5590730810999428,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.3859,
"step": 521
},
{
"epoch": 2.11746835443038,
"grad_norm": 0.5381632429658239,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.3858,
"step": 522
},
{
"epoch": 2.1215189873417724,
"grad_norm": 0.5410005498949547,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.3923,
"step": 523
},
{
"epoch": 2.1255696202531644,
"grad_norm": 0.5078968146955498,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.4121,
"step": 524
},
{
"epoch": 2.129620253164557,
"grad_norm": 0.5296995844901422,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.3927,
"step": 525
},
{
"epoch": 2.1336708860759495,
"grad_norm": 0.4723476045334704,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.375,
"step": 526
},
{
"epoch": 2.137721518987342,
"grad_norm": 0.5252493203831382,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.3969,
"step": 527
},
{
"epoch": 2.141772151898734,
"grad_norm": 0.4905982526917173,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.4046,
"step": 528
},
{
"epoch": 2.1458227848101266,
"grad_norm": 0.519580299763162,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.3897,
"step": 529
},
{
"epoch": 2.149873417721519,
"grad_norm": 0.5664378522951868,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.3896,
"step": 530
},
{
"epoch": 2.1539240506329116,
"grad_norm": 0.5017290771805286,
"learning_rate": 2.802570062143278e-05,
"loss": 0.3776,
"step": 531
},
{
"epoch": 2.1579746835443037,
"grad_norm": 0.5008104300043716,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.3929,
"step": 532
},
{
"epoch": 2.162025316455696,
"grad_norm": 0.5071117741956293,
"learning_rate": 2.792159532078314e-05,
"loss": 0.3936,
"step": 533
},
{
"epoch": 2.1660759493670887,
"grad_norm": 0.489614235268161,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.4057,
"step": 534
},
{
"epoch": 2.170126582278481,
"grad_norm": 0.5674861553534223,
"learning_rate": 2.781723482288857e-05,
"loss": 0.3794,
"step": 535
},
{
"epoch": 2.1741772151898733,
"grad_norm": 0.5620502252188414,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.3778,
"step": 536
},
{
"epoch": 2.178227848101266,
"grad_norm": 0.5192049093013924,
"learning_rate": 2.771262248976272e-05,
"loss": 0.3817,
"step": 537
},
{
"epoch": 2.1822784810126583,
"grad_norm": 0.5334780620464697,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.3946,
"step": 538
},
{
"epoch": 2.186329113924051,
"grad_norm": 0.5368939094059946,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.419,
"step": 539
},
{
"epoch": 2.190379746835443,
"grad_norm": 0.6020815396403002,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.3768,
"step": 540
},
{
"epoch": 2.1944303797468354,
"grad_norm": 0.49624439715132995,
"learning_rate": 2.7502655806328e-05,
"loss": 0.3876,
"step": 541
},
{
"epoch": 2.198481012658228,
"grad_norm": 0.5930388215818916,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.3673,
"step": 542
},
{
"epoch": 2.2025316455696204,
"grad_norm": 0.5305118441568697,
"learning_rate": 2.739730822017673e-05,
"loss": 0.3729,
"step": 543
},
{
"epoch": 2.2065822784810125,
"grad_norm": 0.5512899905929632,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.388,
"step": 544
},
{
"epoch": 2.210632911392405,
"grad_norm": 0.4898761926377834,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.3924,
"step": 545
},
{
"epoch": 2.2146835443037975,
"grad_norm": 0.5585962183593443,
"learning_rate": 2.723884107774236e-05,
"loss": 0.3802,
"step": 546
},
{
"epoch": 2.21873417721519,
"grad_norm": 0.5173937763376121,
"learning_rate": 2.718590152796239e-05,
"loss": 0.3995,
"step": 547
},
{
"epoch": 2.222784810126582,
"grad_norm": 0.5358532035636683,
"learning_rate": 2.71329041039194e-05,
"loss": 0.3942,
"step": 548
},
{
"epoch": 2.2268354430379746,
"grad_norm": 0.5557098012935346,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.3802,
"step": 549
},
{
"epoch": 2.230886075949367,
"grad_norm": 0.5216631260505059,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.3976,
"step": 550
},
{
"epoch": 2.2349367088607597,
"grad_norm": 0.548484707249405,
"learning_rate": 2.697356885686189e-05,
"loss": 0.3934,
"step": 551
},
{
"epoch": 2.2389873417721518,
"grad_norm": 0.49170670797876764,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.425,
"step": 552
},
{
"epoch": 2.2430379746835443,
"grad_norm": 0.5557960129791512,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.4064,
"step": 553
},
{
"epoch": 2.247088607594937,
"grad_norm": 0.5183177342957853,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.4127,
"step": 554
},
{
"epoch": 2.2511392405063293,
"grad_norm": 0.4936928160467687,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.3917,
"step": 555
},
{
"epoch": 2.2551898734177214,
"grad_norm": 0.5911114413180744,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.3865,
"step": 556
},
{
"epoch": 2.259240506329114,
"grad_norm": 0.4963082774494974,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.3851,
"step": 557
},
{
"epoch": 2.2632911392405064,
"grad_norm": 0.5574085938696272,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.4024,
"step": 558
},
{
"epoch": 2.267341772151899,
"grad_norm": 0.5029247788977421,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.3929,
"step": 559
},
{
"epoch": 2.271392405063291,
"grad_norm": 0.5190940508562318,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.385,
"step": 560
},
{
"epoch": 2.2754430379746835,
"grad_norm": 0.4864394441466574,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.3795,
"step": 561
},
{
"epoch": 2.279493670886076,
"grad_norm": 0.5356769919034794,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.3777,
"step": 562
},
{
"epoch": 2.2835443037974685,
"grad_norm": 0.47845565870711027,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.394,
"step": 563
},
{
"epoch": 2.2875949367088606,
"grad_norm": 0.5335915947766228,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.4099,
"step": 564
},
{
"epoch": 2.291645569620253,
"grad_norm": 0.5278980614847779,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.3709,
"step": 565
},
{
"epoch": 2.2956962025316456,
"grad_norm": 0.5428989328848469,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.3918,
"step": 566
},
{
"epoch": 2.299746835443038,
"grad_norm": 0.5309585305281345,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.3947,
"step": 567
},
{
"epoch": 2.3037974683544302,
"grad_norm": 0.5090245896247894,
"learning_rate": 2.606146264569603e-05,
"loss": 0.3819,
"step": 568
},
{
"epoch": 2.3078481012658227,
"grad_norm": 0.5597994775994272,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.3904,
"step": 569
},
{
"epoch": 2.3118987341772153,
"grad_norm": 0.487671711103692,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.4015,
"step": 570
},
{
"epoch": 2.3159493670886078,
"grad_norm": 0.5547766281819857,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.4142,
"step": 571
},
{
"epoch": 2.32,
"grad_norm": 0.5309961358181071,
"learning_rate": 2.584472010606015e-05,
"loss": 0.3859,
"step": 572
},
{
"epoch": 2.3240506329113924,
"grad_norm": 0.5231718581784717,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.3834,
"step": 573
},
{
"epoch": 2.328101265822785,
"grad_norm": 0.526656356130501,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.3969,
"step": 574
},
{
"epoch": 2.3321518987341774,
"grad_norm": 0.5323174366467988,
"learning_rate": 2.568166741322048e-05,
"loss": 0.4249,
"step": 575
},
{
"epoch": 2.3362025316455695,
"grad_norm": 0.5485772189245224,
"learning_rate": 2.56272244128509e-05,
"loss": 0.3732,
"step": 576
},
{
"epoch": 2.340253164556962,
"grad_norm": 0.5487606185740096,
"learning_rate": 2.55727360915895e-05,
"loss": 0.3893,
"step": 577
},
{
"epoch": 2.3443037974683545,
"grad_norm": 0.45320450886860547,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.3966,
"step": 578
},
{
"epoch": 2.348354430379747,
"grad_norm": 0.5577531308781084,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.3875,
"step": 579
},
{
"epoch": 2.352405063291139,
"grad_norm": 0.5012656266906803,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.3839,
"step": 580
},
{
"epoch": 2.3564556962025316,
"grad_norm": 0.47954942907237397,
"learning_rate": 2.535433837985559e-05,
"loss": 0.367,
"step": 581
},
{
"epoch": 2.360506329113924,
"grad_norm": 0.5288365241695847,
"learning_rate": 2.529963004393324e-05,
"loss": 0.4013,
"step": 582
},
{
"epoch": 2.3645569620253166,
"grad_norm": 0.5259066368248149,
"learning_rate": 2.524487902551908e-05,
"loss": 0.3885,
"step": 583
},
{
"epoch": 2.3686075949367087,
"grad_norm": 0.5386614986214541,
"learning_rate": 2.519008576557029e-05,
"loss": 0.365,
"step": 584
},
{
"epoch": 2.372658227848101,
"grad_norm": 0.5169194781456984,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.3982,
"step": 585
},
{
"epoch": 2.3767088607594937,
"grad_norm": 0.4792641027685086,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.4106,
"step": 586
},
{
"epoch": 2.3807594936708862,
"grad_norm": 0.52993219487498,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.4044,
"step": 587
},
{
"epoch": 2.3848101265822783,
"grad_norm": 0.5057277906689147,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.3937,
"step": 588
},
{
"epoch": 2.388860759493671,
"grad_norm": 0.5045717257558978,
"learning_rate": 2.491550129992345e-05,
"loss": 0.3981,
"step": 589
},
{
"epoch": 2.3929113924050633,
"grad_norm": 0.49937295747499555,
"learning_rate": 2.486046386967024e-05,
"loss": 0.3822,
"step": 590
},
{
"epoch": 2.396962025316456,
"grad_norm": 0.5066308059524548,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.382,
"step": 591
},
{
"epoch": 2.401012658227848,
"grad_norm": 0.5171825504602564,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.4082,
"step": 592
},
{
"epoch": 2.4050632911392404,
"grad_norm": 0.5096968657982488,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.3876,
"step": 593
},
{
"epoch": 2.409113924050633,
"grad_norm": 0.4913228147767763,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.4078,
"step": 594
},
{
"epoch": 2.4131645569620255,
"grad_norm": 0.48374469005949505,
"learning_rate": 2.458469841202444e-05,
"loss": 0.3983,
"step": 595
},
{
"epoch": 2.4172151898734175,
"grad_norm": 0.5298518028899342,
"learning_rate": 2.452943276859226e-05,
"loss": 0.3801,
"step": 596
},
{
"epoch": 2.42126582278481,
"grad_norm": 0.508235386527028,
"learning_rate": 2.447413064573125e-05,
"loss": 0.3729,
"step": 597
},
{
"epoch": 2.4253164556962026,
"grad_norm": 0.49506312444253037,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.39,
"step": 598
},
{
"epoch": 2.429367088607595,
"grad_norm": 0.5082677869266008,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.3847,
"step": 599
},
{
"epoch": 2.433417721518987,
"grad_norm": 0.49629899094530816,
"learning_rate": 2.430800985597963e-05,
"loss": 0.3961,
"step": 600
},
{
"epoch": 2.4374683544303797,
"grad_norm": 0.5223900228587423,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.3884,
"step": 601
},
{
"epoch": 2.441518987341772,
"grad_norm": 0.5022780601498966,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.3937,
"step": 602
},
{
"epoch": 2.4455696202531647,
"grad_norm": 0.5027701401585879,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.4094,
"step": 603
},
{
"epoch": 2.449620253164557,
"grad_norm": 0.5138325086308025,
"learning_rate": 2.408603181104957e-05,
"loss": 0.4064,
"step": 604
},
{
"epoch": 2.4536708860759493,
"grad_norm": 0.5217868141883769,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.3903,
"step": 605
},
{
"epoch": 2.457721518987342,
"grad_norm": 0.5457417589365071,
"learning_rate": 2.397484355008521e-05,
"loss": 0.4138,
"step": 606
},
{
"epoch": 2.4617721518987343,
"grad_norm": 0.4961941282406955,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.4088,
"step": 607
},
{
"epoch": 2.4658227848101264,
"grad_norm": 0.5673651505143675,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.4065,
"step": 608
},
{
"epoch": 2.469873417721519,
"grad_norm": 0.496795027734954,
"learning_rate": 2.380782218329337e-05,
"loss": 0.3837,
"step": 609
},
{
"epoch": 2.4739240506329114,
"grad_norm": 0.5663638519427953,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.4153,
"step": 610
},
{
"epoch": 2.477974683544304,
"grad_norm": 0.5263510587265858,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.394,
"step": 611
},
{
"epoch": 2.482025316455696,
"grad_norm": 0.5473837432130367,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.3729,
"step": 612
},
{
"epoch": 2.4860759493670885,
"grad_norm": 0.596745223518634,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.3783,
"step": 613
},
{
"epoch": 2.490126582278481,
"grad_norm": 0.5177646328185753,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.3898,
"step": 614
},
{
"epoch": 2.4941772151898736,
"grad_norm": 0.6058198021976957,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.4066,
"step": 615
},
{
"epoch": 2.4982278481012656,
"grad_norm": 0.5299405783953162,
"learning_rate": 2.341705325992901e-05,
"loss": 0.3869,
"step": 616
},
{
"epoch": 2.502278481012658,
"grad_norm": 0.507854044505029,
"learning_rate": 2.336111544604222e-05,
"loss": 0.3831,
"step": 617
},
{
"epoch": 2.5063291139240507,
"grad_norm": 0.5638658148945871,
"learning_rate": 2.33051505621937e-05,
"loss": 0.3923,
"step": 618
},
{
"epoch": 2.510379746835443,
"grad_norm": 0.46130581449073943,
"learning_rate": 2.324915905911693e-05,
"loss": 0.3885,
"step": 619
},
{
"epoch": 2.5144303797468357,
"grad_norm": 0.5718106276597306,
"learning_rate": 2.319314138775977e-05,
"loss": 0.3789,
"step": 620
},
{
"epoch": 2.5184810126582278,
"grad_norm": 0.5120071019440388,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.3815,
"step": 621
},
{
"epoch": 2.5225316455696203,
"grad_norm": 0.5393623709351485,
"learning_rate": 2.308102934504593e-05,
"loss": 0.3834,
"step": 622
},
{
"epoch": 2.526582278481013,
"grad_norm": 0.5290851308208688,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.3911,
"step": 623
},
{
"epoch": 2.530632911392405,
"grad_norm": 0.4968584402663297,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.3764,
"step": 624
},
{
"epoch": 2.5346835443037974,
"grad_norm": 0.5198578052378253,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.3877,
"step": 625
},
{
"epoch": 2.53873417721519,
"grad_norm": 0.5101396496897631,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.3899,
"step": 626
},
{
"epoch": 2.5427848101265824,
"grad_norm": 0.47740453714546216,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.3745,
"step": 627
},
{
"epoch": 2.546835443037975,
"grad_norm": 0.5033874008481569,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.399,
"step": 628
},
{
"epoch": 2.550886075949367,
"grad_norm": 0.5196901383210085,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.3856,
"step": 629
},
{
"epoch": 2.5549367088607595,
"grad_norm": 0.47529308123504466,
"learning_rate": 2.26316247733093e-05,
"loss": 0.3729,
"step": 630
},
{
"epoch": 2.558987341772152,
"grad_norm": 0.5521330908490354,
"learning_rate": 2.257534907110328e-05,
"loss": 0.3957,
"step": 631
},
{
"epoch": 2.563037974683544,
"grad_norm": 0.46475835242995406,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.4017,
"step": 632
},
{
"epoch": 2.5670886075949366,
"grad_norm": 0.540284425762592,
"learning_rate": 2.246273589557294e-05,
"loss": 0.393,
"step": 633
},
{
"epoch": 2.571139240506329,
"grad_norm": 0.49783412101338714,
"learning_rate": 2.240639932921966e-05,
"loss": 0.3806,
"step": 634
},
{
"epoch": 2.5751898734177217,
"grad_norm": 0.47373932613815994,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.3929,
"step": 635
},
{
"epoch": 2.579240506329114,
"grad_norm": 0.48466026237194804,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.3758,
"step": 636
},
{
"epoch": 2.5832911392405062,
"grad_norm": 0.5240457596125045,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.3876,
"step": 637
},
{
"epoch": 2.5873417721518988,
"grad_norm": 0.4803016406530167,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.3828,
"step": 638
},
{
"epoch": 2.5913924050632913,
"grad_norm": 0.49903951735877516,
"learning_rate": 2.212443486523819e-05,
"loss": 0.3766,
"step": 639
},
{
"epoch": 2.5954430379746833,
"grad_norm": 0.5050405448628907,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.4157,
"step": 640
},
{
"epoch": 2.599493670886076,
"grad_norm": 0.4893982919162751,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.3637,
"step": 641
},
{
"epoch": 2.6035443037974684,
"grad_norm": 0.5284231915310953,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.3995,
"step": 642
},
{
"epoch": 2.607594936708861,
"grad_norm": 0.540301264074371,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.381,
"step": 643
},
{
"epoch": 2.6116455696202534,
"grad_norm": 0.48751632469706613,
"learning_rate": 2.184204266048803e-05,
"loss": 0.3762,
"step": 644
},
{
"epoch": 2.6156962025316455,
"grad_norm": 0.504192029291679,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.405,
"step": 645
},
{
"epoch": 2.619746835443038,
"grad_norm": 0.4900022896485269,
"learning_rate": 2.17289787446841e-05,
"loss": 0.3915,
"step": 646
},
{
"epoch": 2.6237974683544305,
"grad_norm": 0.49564818126953947,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.3997,
"step": 647
},
{
"epoch": 2.6278481012658226,
"grad_norm": 0.5074261598645381,
"learning_rate": 2.161585912916385e-05,
"loss": 0.3812,
"step": 648
},
{
"epoch": 2.631898734177215,
"grad_norm": 0.5304611806811149,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.3796,
"step": 649
},
{
"epoch": 2.6359493670886076,
"grad_norm": 0.46985338630871776,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.3927,
"step": 650
},
{
"epoch": 2.64,
"grad_norm": 0.5274816934489053,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.4029,
"step": 651
},
{
"epoch": 2.6440506329113926,
"grad_norm": 0.5155970016221268,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.3802,
"step": 652
},
{
"epoch": 2.6481012658227847,
"grad_norm": 0.48700566139002127,
"learning_rate": 2.133284032321232e-05,
"loss": 0.3947,
"step": 653
},
{
"epoch": 2.6521518987341772,
"grad_norm": 0.4822866147458846,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.4028,
"step": 654
},
{
"epoch": 2.6562025316455697,
"grad_norm": 0.48882533152827023,
"learning_rate": 2.121955446741306e-05,
"loss": 0.3876,
"step": 655
},
{
"epoch": 2.660253164556962,
"grad_norm": 0.4808330047990712,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.3851,
"step": 656
},
{
"epoch": 2.6643037974683543,
"grad_norm": 0.5198127247641465,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.3807,
"step": 657
},
{
"epoch": 2.668354430379747,
"grad_norm": 0.5043231992548044,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.3836,
"step": 658
},
{
"epoch": 2.6724050632911394,
"grad_norm": 0.471732950849583,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.403,
"step": 659
},
{
"epoch": 2.676455696202532,
"grad_norm": 0.48768568716875477,
"learning_rate": 2.093617592756158e-05,
"loss": 0.396,
"step": 660
},
{
"epoch": 2.680506329113924,
"grad_norm": 0.5164824074171982,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.3945,
"step": 661
},
{
"epoch": 2.6845569620253165,
"grad_norm": 0.44362133760095174,
"learning_rate": 2.082276853708978e-05,
"loss": 0.4012,
"step": 662
},
{
"epoch": 2.688607594936709,
"grad_norm": 0.5424636133054441,
"learning_rate": 2.076605467380071e-05,
"loss": 0.3921,
"step": 663
},
{
"epoch": 2.692658227848101,
"grad_norm": 0.4879320485597456,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.3839,
"step": 664
},
{
"epoch": 2.6967088607594936,
"grad_norm": 0.4690178569671435,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.3937,
"step": 665
},
{
"epoch": 2.700759493670886,
"grad_norm": 0.5166539612222131,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.3781,
"step": 666
},
{
"epoch": 2.7048101265822786,
"grad_norm": 0.49678908987531584,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.4005,
"step": 667
},
{
"epoch": 2.708860759493671,
"grad_norm": 0.4756486183991966,
"learning_rate": 2.048240194884496e-05,
"loss": 0.3969,
"step": 668
},
{
"epoch": 2.712911392405063,
"grad_norm": 0.5116247364806675,
"learning_rate": 2.042565792045695e-05,
"loss": 0.3908,
"step": 669
},
{
"epoch": 2.7169620253164557,
"grad_norm": 0.46826860800038805,
"learning_rate": 2.036891046387857e-05,
"loss": 0.3935,
"step": 670
},
{
"epoch": 2.721012658227848,
"grad_norm": 0.5162937174364828,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.402,
"step": 671
},
{
"epoch": 2.7250632911392403,
"grad_norm": 0.49510821734878696,
"learning_rate": 2.025540709431948e-05,
"loss": 0.3995,
"step": 672
},
{
"epoch": 2.729113924050633,
"grad_norm": 0.5005030848170915,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.3691,
"step": 673
},
{
"epoch": 2.7331645569620253,
"grad_norm": 0.458481590020088,
"learning_rate": 2.014189549672245e-05,
"loss": 0.4105,
"step": 674
},
{
"epoch": 2.737215189873418,
"grad_norm": 0.49321668364280713,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.3771,
"step": 675
},
{
"epoch": 2.7412658227848103,
"grad_norm": 0.472527857837478,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.4009,
"step": 676
},
{
"epoch": 2.7453164556962024,
"grad_norm": 0.5022015726648594,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.4072,
"step": 677
},
{
"epoch": 2.749367088607595,
"grad_norm": 0.48760586373927134,
"learning_rate": 1.991486224484123e-05,
"loss": 0.3937,
"step": 678
},
{
"epoch": 2.7534177215189874,
"grad_norm": 0.5323909096454066,
"learning_rate": 1.985810450327756e-05,
"loss": 0.4003,
"step": 679
},
{
"epoch": 2.7574683544303795,
"grad_norm": 0.47254780463862117,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.4225,
"step": 680
},
{
"epoch": 2.761518987341772,
"grad_norm": 0.47120478959433,
"learning_rate": 1.974459290568053e-05,
"loss": 0.4038,
"step": 681
},
{
"epoch": 2.7655696202531646,
"grad_norm": 0.4594869104773301,
"learning_rate": 1.968783996385397e-05,
"loss": 0.4358,
"step": 682
},
{
"epoch": 2.769620253164557,
"grad_norm": 0.5008529955760294,
"learning_rate": 1.963108953612143e-05,
"loss": 0.3907,
"step": 683
},
{
"epoch": 2.7736708860759496,
"grad_norm": 0.4900989070728069,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.3879,
"step": 684
},
{
"epoch": 2.7777215189873417,
"grad_norm": 0.5333902803944582,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.3834,
"step": 685
},
{
"epoch": 2.781772151898734,
"grad_norm": 0.4882997792399622,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.3745,
"step": 686
},
{
"epoch": 2.7858227848101267,
"grad_norm": 0.5330437284709124,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.4042,
"step": 687
},
{
"epoch": 2.7898734177215188,
"grad_norm": 0.4938235903639061,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.3798,
"step": 688
},
{
"epoch": 2.7939240506329113,
"grad_norm": 0.5015297647268909,
"learning_rate": 1.92906653591874e-05,
"loss": 0.3775,
"step": 689
},
{
"epoch": 2.797974683544304,
"grad_norm": 0.46983888389104816,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.3978,
"step": 690
},
{
"epoch": 2.8020253164556963,
"grad_norm": 0.47126749503389154,
"learning_rate": 1.917723146291022e-05,
"loss": 0.3857,
"step": 691
},
{
"epoch": 2.806075949367089,
"grad_norm": 0.44525189100086054,
"learning_rate": 1.912052422608584e-05,
"loss": 0.3978,
"step": 692
},
{
"epoch": 2.810126582278481,
"grad_norm": 0.4818002078286936,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.3775,
"step": 693
},
{
"epoch": 2.8141772151898734,
"grad_norm": 0.4886079247174068,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.3588,
"step": 694
},
{
"epoch": 2.818227848101266,
"grad_norm": 0.5009523669051549,
"learning_rate": 1.895044684123479e-05,
"loss": 0.3966,
"step": 695
},
{
"epoch": 2.822278481012658,
"grad_norm": 0.48656292422114505,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.4002,
"step": 696
},
{
"epoch": 2.8263291139240505,
"grad_norm": 0.46348396160794086,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.3912,
"step": 697
},
{
"epoch": 2.830379746835443,
"grad_norm": 0.48119412321376664,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.3773,
"step": 698
},
{
"epoch": 2.8344303797468355,
"grad_norm": 0.4526427618334827,
"learning_rate": 1.872379746550743e-05,
"loss": 0.4124,
"step": 699
},
{
"epoch": 2.838481012658228,
"grad_norm": 0.4707631219887712,
"learning_rate": 1.866715967678769e-05,
"loss": 0.392,
"step": 700
},
{
"epoch": 2.84253164556962,
"grad_norm": 0.4819353367354442,
"learning_rate": 1.861053262258067e-05,
"loss": 0.3753,
"step": 701
},
{
"epoch": 2.8465822784810126,
"grad_norm": 0.5141132825023353,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.3877,
"step": 702
},
{
"epoch": 2.850632911392405,
"grad_norm": 0.47046045515133966,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.3772,
"step": 703
},
{
"epoch": 2.8546835443037972,
"grad_norm": 0.46309990954293057,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.3961,
"step": 704
},
{
"epoch": 2.8587341772151897,
"grad_norm": 0.4542058709155521,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.3934,
"step": 705
},
{
"epoch": 2.8627848101265823,
"grad_norm": 0.4777927167513398,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.3912,
"step": 706
},
{
"epoch": 2.8668354430379748,
"grad_norm": 0.4708627213989777,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.3825,
"step": 707
},
{
"epoch": 2.8708860759493673,
"grad_norm": 0.47195116943965465,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.3857,
"step": 708
},
{
"epoch": 2.8749367088607594,
"grad_norm": 0.4571950853402927,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.3747,
"step": 709
},
{
"epoch": 2.878987341772152,
"grad_norm": 0.47398785812233857,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.4,
"step": 710
},
{
"epoch": 2.8830379746835444,
"grad_norm": 0.49696567158611393,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.3816,
"step": 711
},
{
"epoch": 2.8870886075949365,
"grad_norm": 0.485728134335276,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.3919,
"step": 712
},
{
"epoch": 2.891139240506329,
"grad_norm": 0.46453533577832523,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.395,
"step": 713
},
{
"epoch": 2.8951898734177215,
"grad_norm": 0.48905687989305474,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.3728,
"step": 714
},
{
"epoch": 2.899240506329114,
"grad_norm": 0.4782460710457942,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.3989,
"step": 715
},
{
"epoch": 2.9032911392405065,
"grad_norm": 0.4733029187187664,
"learning_rate": 1.776272483901182e-05,
"loss": 0.3893,
"step": 716
},
{
"epoch": 2.9073417721518986,
"grad_norm": 0.45941519144760606,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.3821,
"step": 717
},
{
"epoch": 2.911392405063291,
"grad_norm": 0.47470767582155743,
"learning_rate": 1.764995661794301e-05,
"loss": 0.3813,
"step": 718
},
{
"epoch": 2.9154430379746836,
"grad_norm": 0.49465117426237754,
"learning_rate": 1.759360067078035e-05,
"loss": 0.3492,
"step": 719
},
{
"epoch": 2.9194936708860757,
"grad_norm": 0.4647171580572261,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.3764,
"step": 720
},
{
"epoch": 2.923544303797468,
"grad_norm": 0.4703552685131904,
"learning_rate": 1.748094737261012e-05,
"loss": 0.4148,
"step": 721
},
{
"epoch": 2.9275949367088607,
"grad_norm": 0.48667391909708396,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.3816,
"step": 722
},
{
"epoch": 2.9316455696202532,
"grad_norm": 0.4801612636753923,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.3869,
"step": 723
},
{
"epoch": 2.9356962025316458,
"grad_norm": 0.4616111425065936,
"learning_rate": 1.731212071922883e-05,
"loss": 0.381,
"step": 724
},
{
"epoch": 2.939746835443038,
"grad_norm": 0.48068063412326295,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.3844,
"step": 725
},
{
"epoch": 2.9437974683544303,
"grad_norm": 0.4923558569596147,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.3813,
"step": 726
},
{
"epoch": 2.947848101265823,
"grad_norm": 0.458721613865435,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.3986,
"step": 727
},
{
"epoch": 2.951898734177215,
"grad_norm": 0.5168530585192836,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.3867,
"step": 728
},
{
"epoch": 2.9559493670886074,
"grad_norm": 0.49200606733279345,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.3852,
"step": 729
},
{
"epoch": 2.96,
"grad_norm": 0.48940810430331494,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.381,
"step": 730
},
{
"epoch": 2.9640506329113925,
"grad_norm": 0.497344781289111,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.3881,
"step": 731
},
{
"epoch": 2.968101265822785,
"grad_norm": 0.4806141719673312,
"learning_rate": 1.686290200071915e-05,
"loss": 0.4043,
"step": 732
},
{
"epoch": 2.972151898734177,
"grad_norm": 0.42736977716895663,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.3796,
"step": 733
},
{
"epoch": 2.9762025316455696,
"grad_norm": 0.4820867951133501,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.3722,
"step": 734
},
{
"epoch": 2.980253164556962,
"grad_norm": 0.4790149682398204,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.3951,
"step": 735
},
{
"epoch": 2.984303797468354,
"grad_norm": 0.47592717648275684,
"learning_rate": 1.663888455395778e-05,
"loss": 0.3699,
"step": 736
},
{
"epoch": 2.9883544303797467,
"grad_norm": 0.4951585917745847,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.3918,
"step": 737
},
{
"epoch": 2.992405063291139,
"grad_norm": 0.4825041855196529,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.3873,
"step": 738
},
{
"epoch": 2.9964556962025317,
"grad_norm": 0.4703350132419328,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.3725,
"step": 739
},
{
"epoch": 3.002025316455696,
"grad_norm": 0.5442538215574824,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.3094,
"step": 740
},
{
"epoch": 3.0060759493670886,
"grad_norm": 0.6675486529161807,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.2046,
"step": 741
},
{
"epoch": 3.010126582278481,
"grad_norm": 0.4425286391417739,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.2439,
"step": 742
},
{
"epoch": 3.0141772151898736,
"grad_norm": 0.8545725198842813,
"learning_rate": 1.624791353909428e-05,
"loss": 0.2537,
"step": 743
},
{
"epoch": 3.0182278481012657,
"grad_norm": 0.7390613791865969,
"learning_rate": 1.619217781670663e-05,
"loss": 0.2527,
"step": 744
},
{
"epoch": 3.022278481012658,
"grad_norm": 0.5367804736315809,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.2723,
"step": 745
},
{
"epoch": 3.0263291139240507,
"grad_norm": 0.5947465667060565,
"learning_rate": 1.608079882360134e-05,
"loss": 0.2494,
"step": 746
},
{
"epoch": 3.030379746835443,
"grad_norm": 0.609967938143923,
"learning_rate": 1.60251564499148e-05,
"loss": 0.2408,
"step": 747
},
{
"epoch": 3.0344303797468353,
"grad_norm": 0.555382104445193,
"learning_rate": 1.596954608907176e-05,
"loss": 0.2412,
"step": 748
},
{
"epoch": 3.038481012658228,
"grad_norm": 0.4490653449075135,
"learning_rate": 1.591396818895043e-05,
"loss": 0.226,
"step": 749
},
{
"epoch": 3.0425316455696203,
"grad_norm": 0.5085219220832761,
"learning_rate": 1.585842319716759e-05,
"loss": 0.2465,
"step": 750
},
{
"epoch": 3.046582278481013,
"grad_norm": 0.5480758433456606,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.247,
"step": 751
},
{
"epoch": 3.050632911392405,
"grad_norm": 0.4779662689216839,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.2242,
"step": 752
},
{
"epoch": 3.0546835443037974,
"grad_norm": 0.4504995430383242,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.244,
"step": 753
},
{
"epoch": 3.05873417721519,
"grad_norm": 0.44997225053045015,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.2174,
"step": 754
},
{
"epoch": 3.0627848101265824,
"grad_norm": 0.4658077677352596,
"learning_rate": 1.558120751116291e-05,
"loss": 0.2441,
"step": 755
},
{
"epoch": 3.0668354430379745,
"grad_norm": 0.46819554366413696,
"learning_rate": 1.552586935426876e-05,
"loss": 0.2447,
"step": 756
},
{
"epoch": 3.070886075949367,
"grad_norm": 0.4415734728422274,
"learning_rate": 1.547056723140774e-05,
"loss": 0.2113,
"step": 757
},
{
"epoch": 3.0749367088607595,
"grad_norm": 0.43690753417571837,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.2478,
"step": 758
},
{
"epoch": 3.078987341772152,
"grad_norm": 0.4677806728892679,
"learning_rate": 1.536007286907411e-05,
"loss": 0.2326,
"step": 759
},
{
"epoch": 3.083037974683544,
"grad_norm": 0.41548068387607523,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.2147,
"step": 760
},
{
"epoch": 3.0870886075949366,
"grad_norm": 0.41704279432571983,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.2065,
"step": 761
},
{
"epoch": 3.091139240506329,
"grad_norm": 0.4085544471222903,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.2554,
"step": 762
},
{
"epoch": 3.0951898734177217,
"grad_norm": 0.4365148992974762,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.2423,
"step": 763
},
{
"epoch": 3.0992405063291137,
"grad_norm": 0.41440934718474426,
"learning_rate": 1.508449870007656e-05,
"loss": 0.2066,
"step": 764
},
{
"epoch": 3.1032911392405063,
"grad_norm": 0.4027667798956743,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.2488,
"step": 765
},
{
"epoch": 3.1073417721518988,
"grad_norm": 0.4276553047623516,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.2281,
"step": 766
},
{
"epoch": 3.1113924050632913,
"grad_norm": 0.40369046417369064,
"learning_rate": 1.4919625713405e-05,
"loss": 0.2298,
"step": 767
},
{
"epoch": 3.1154430379746834,
"grad_norm": 0.41499201478145287,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.249,
"step": 768
},
{
"epoch": 3.119493670886076,
"grad_norm": 0.40975169568035386,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.2329,
"step": 769
},
{
"epoch": 3.1235443037974684,
"grad_norm": 0.42228638837468313,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.2388,
"step": 770
},
{
"epoch": 3.127594936708861,
"grad_norm": 0.4143010903359201,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.2419,
"step": 771
},
{
"epoch": 3.131645569620253,
"grad_norm": 0.40237544201898445,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.2575,
"step": 772
},
{
"epoch": 3.1356962025316455,
"grad_norm": 0.41141276560940654,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.2363,
"step": 773
},
{
"epoch": 3.139746835443038,
"grad_norm": 0.40609312298724376,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.2317,
"step": 774
},
{
"epoch": 3.1437974683544305,
"grad_norm": 0.4172908693524053,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.227,
"step": 775
},
{
"epoch": 3.1478481012658226,
"grad_norm": 0.4038231656491561,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.2323,
"step": 776
},
{
"epoch": 3.151898734177215,
"grad_norm": 0.4252719684086989,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.2336,
"step": 777
},
{
"epoch": 3.1559493670886076,
"grad_norm": 0.41181604837216595,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.2507,
"step": 778
},
{
"epoch": 3.16,
"grad_norm": 0.39386332260414786,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.2564,
"step": 779
},
{
"epoch": 3.164050632911392,
"grad_norm": 0.40352885554151896,
"learning_rate": 1.420958430225303e-05,
"loss": 0.2501,
"step": 780
},
{
"epoch": 3.1681012658227847,
"grad_norm": 0.42453876970743726,
"learning_rate": 1.415527989393985e-05,
"loss": 0.2306,
"step": 781
},
{
"epoch": 3.1721518987341772,
"grad_norm": 0.4041376848353027,
"learning_rate": 1.410102255819891e-05,
"loss": 0.223,
"step": 782
},
{
"epoch": 3.1762025316455698,
"grad_norm": 0.402206366472267,
"learning_rate": 1.404681273201131e-05,
"loss": 0.2364,
"step": 783
},
{
"epoch": 3.180253164556962,
"grad_norm": 0.4087234041670038,
"learning_rate": 1.399265085197556e-05,
"loss": 0.2419,
"step": 784
},
{
"epoch": 3.1843037974683543,
"grad_norm": 0.4023376665436527,
"learning_rate": 1.393853735430398e-05,
"loss": 0.233,
"step": 785
},
{
"epoch": 3.188354430379747,
"grad_norm": 0.41204160185850075,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.2276,
"step": 786
},
{
"epoch": 3.1924050632911394,
"grad_norm": 0.40557815628107813,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.2367,
"step": 787
},
{
"epoch": 3.1964556962025314,
"grad_norm": 0.4115320340468774,
"learning_rate": 1.377649151173163e-05,
"loss": 0.2127,
"step": 788
},
{
"epoch": 3.200506329113924,
"grad_norm": 0.4029022858425113,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.2094,
"step": 789
},
{
"epoch": 3.2045569620253165,
"grad_norm": 0.39623060413983185,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.2146,
"step": 790
},
{
"epoch": 3.208607594936709,
"grad_norm": 0.45560900466083865,
"learning_rate": 1.361489677627324e-05,
"loss": 0.2184,
"step": 791
},
{
"epoch": 3.212658227848101,
"grad_norm": 0.38732393556518296,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.2273,
"step": 792
},
{
"epoch": 3.2167088607594936,
"grad_norm": 0.38826499650745216,
"learning_rate": 1.350742335331241e-05,
"loss": 0.2391,
"step": 793
},
{
"epoch": 3.220759493670886,
"grad_norm": 0.3852553298021364,
"learning_rate": 1.345376486102198e-05,
"loss": 0.2387,
"step": 794
},
{
"epoch": 3.2248101265822786,
"grad_norm": 0.4002667646493048,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.2692,
"step": 795
},
{
"epoch": 3.2288607594936707,
"grad_norm": 0.3990098961738234,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.2475,
"step": 796
},
{
"epoch": 3.232911392405063,
"grad_norm": 0.39724765319452565,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.2219,
"step": 797
},
{
"epoch": 3.2369620253164557,
"grad_norm": 0.3890925170890548,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.2309,
"step": 798
},
{
"epoch": 3.2410126582278482,
"grad_norm": 0.3955481364379494,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.2301,
"step": 799
},
{
"epoch": 3.2450632911392403,
"grad_norm": 0.3982155356530945,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.2645,
"step": 800
},
{
"epoch": 3.249113924050633,
"grad_norm": 0.3800151804440712,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.2343,
"step": 801
},
{
"epoch": 3.2531645569620253,
"grad_norm": 0.3908192243746257,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.2435,
"step": 802
},
{
"epoch": 3.257215189873418,
"grad_norm": 0.3940568493308798,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.2371,
"step": 803
},
{
"epoch": 3.26126582278481,
"grad_norm": 0.40911145914886404,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.246,
"step": 804
},
{
"epoch": 3.2653164556962024,
"grad_norm": 0.3975208353567077,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.2363,
"step": 805
},
{
"epoch": 3.269367088607595,
"grad_norm": 0.4055879209850727,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.2184,
"step": 806
},
{
"epoch": 3.2734177215189875,
"grad_norm": 0.4156228720286413,
"learning_rate": 1.276115892225764e-05,
"loss": 0.2302,
"step": 807
},
{
"epoch": 3.27746835443038,
"grad_norm": 0.37675872488945183,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.2169,
"step": 808
},
{
"epoch": 3.281518987341772,
"grad_norm": 0.3974830240414087,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.2226,
"step": 809
},
{
"epoch": 3.2855696202531646,
"grad_norm": 0.4118052213658784,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.2254,
"step": 810
},
{
"epoch": 3.289620253164557,
"grad_norm": 0.3852218094988751,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.2498,
"step": 811
},
{
"epoch": 3.293670886075949,
"grad_norm": 0.3992479955881525,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.2189,
"step": 812
},
{
"epoch": 3.2977215189873417,
"grad_norm": 0.38036158541814064,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.233,
"step": 813
},
{
"epoch": 3.301772151898734,
"grad_norm": 0.3943198140787574,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.2439,
"step": 814
},
{
"epoch": 3.3058227848101267,
"grad_norm": 0.3781012284511662,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.2329,
"step": 815
},
{
"epoch": 3.309873417721519,
"grad_norm": 0.3969789848229199,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.234,
"step": 816
},
{
"epoch": 3.3139240506329113,
"grad_norm": 0.3995536631634163,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.2244,
"step": 817
},
{
"epoch": 3.317974683544304,
"grad_norm": 0.39401648913353204,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.2276,
"step": 818
},
{
"epoch": 3.3220253164556963,
"grad_norm": 0.3926517192896875,
"learning_rate": 1.213055323844297e-05,
"loss": 0.239,
"step": 819
},
{
"epoch": 3.3260759493670884,
"grad_norm": 0.4055764839333673,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.2277,
"step": 820
},
{
"epoch": 3.330126582278481,
"grad_norm": 0.39861478973018244,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.2331,
"step": 821
},
{
"epoch": 3.3341772151898734,
"grad_norm": 0.38647487319496077,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.211,
"step": 822
},
{
"epoch": 3.338227848101266,
"grad_norm": 0.38459027809467,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.2255,
"step": 823
},
{
"epoch": 3.3422784810126585,
"grad_norm": 0.39907752788995254,
"learning_rate": 1.187045262895488e-05,
"loss": 0.2359,
"step": 824
},
{
"epoch": 3.3463291139240505,
"grad_norm": 0.40948467266745636,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.2456,
"step": 825
},
{
"epoch": 3.350379746835443,
"grad_norm": 0.40418699461095,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.2357,
"step": 826
},
{
"epoch": 3.3544303797468356,
"grad_norm": 0.402352851135571,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.2092,
"step": 827
},
{
"epoch": 3.3584810126582276,
"grad_norm": 0.383755149007403,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.2196,
"step": 828
},
{
"epoch": 3.36253164556962,
"grad_norm": 0.38753094184408166,
"learning_rate": 1.161198884952377e-05,
"loss": 0.2452,
"step": 829
},
{
"epoch": 3.3665822784810127,
"grad_norm": 0.3952975505409735,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.2548,
"step": 830
},
{
"epoch": 3.370632911392405,
"grad_norm": 0.3929863633784028,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.2219,
"step": 831
},
{
"epoch": 3.3746835443037977,
"grad_norm": 0.4205522687446161,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.2163,
"step": 832
},
{
"epoch": 3.3787341772151898,
"grad_norm": 0.3893170795742087,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.2298,
"step": 833
},
{
"epoch": 3.3827848101265823,
"grad_norm": 0.40373976911322274,
"learning_rate": 1.135521394010506e-05,
"loss": 0.2486,
"step": 834
},
{
"epoch": 3.386835443037975,
"grad_norm": 0.42114699734360417,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.2181,
"step": 835
},
{
"epoch": 3.390886075949367,
"grad_norm": 0.4032361218102273,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.2397,
"step": 836
},
{
"epoch": 3.3949367088607594,
"grad_norm": 0.39711284127217666,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.2387,
"step": 837
},
{
"epoch": 3.398987341772152,
"grad_norm": 0.3995225042045124,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.2342,
"step": 838
},
{
"epoch": 3.4030379746835444,
"grad_norm": 0.3799424726635692,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.22,
"step": 839
},
{
"epoch": 3.407088607594937,
"grad_norm": 0.38454787546651936,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.2299,
"step": 840
},
{
"epoch": 3.411139240506329,
"grad_norm": 0.39845063429760796,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.2527,
"step": 841
},
{
"epoch": 3.4151898734177215,
"grad_norm": 0.39628953408786555,
"learning_rate": 1.0948015846744e-05,
"loss": 0.2251,
"step": 842
},
{
"epoch": 3.419240506329114,
"grad_norm": 0.39846535057160876,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.2213,
"step": 843
},
{
"epoch": 3.423291139240506,
"grad_norm": 0.3772191154561103,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.228,
"step": 844
},
{
"epoch": 3.4273417721518986,
"grad_norm": 0.3930434926777099,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.2532,
"step": 845
},
{
"epoch": 3.431392405063291,
"grad_norm": 0.39831285371166963,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.2356,
"step": 846
},
{
"epoch": 3.4354430379746836,
"grad_norm": 0.40546347880657735,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.2164,
"step": 847
},
{
"epoch": 3.439493670886076,
"grad_norm": 0.3892566176818296,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.2266,
"step": 848
},
{
"epoch": 3.4435443037974682,
"grad_norm": 0.3880704231690668,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.2371,
"step": 849
},
{
"epoch": 3.4475949367088607,
"grad_norm": 0.378129870610881,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.2508,
"step": 850
},
{
"epoch": 3.4516455696202533,
"grad_norm": 0.3975785397094189,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.2249,
"step": 851
},
{
"epoch": 3.4556962025316453,
"grad_norm": 0.39100968949636383,
"learning_rate": 1.044560362975474e-05,
"loss": 0.2768,
"step": 852
},
{
"epoch": 3.459746835443038,
"grad_norm": 0.3836851850774309,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.228,
"step": 853
},
{
"epoch": 3.4637974683544304,
"grad_norm": 0.3937856943770143,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.2071,
"step": 854
},
{
"epoch": 3.467848101265823,
"grad_norm": 0.3990758987322364,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.2369,
"step": 855
},
{
"epoch": 3.4718987341772154,
"grad_norm": 0.38867271875931453,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.2514,
"step": 856
},
{
"epoch": 3.4759493670886075,
"grad_norm": 0.40343771996942407,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.2169,
"step": 857
},
{
"epoch": 3.48,
"grad_norm": 0.3927806934105635,
"learning_rate": 1.014782400334433e-05,
"loss": 0.2456,
"step": 858
},
{
"epoch": 3.4840506329113925,
"grad_norm": 0.4048767292526437,
"learning_rate": 1.009846945771296e-05,
"loss": 0.2531,
"step": 859
},
{
"epoch": 3.4881012658227846,
"grad_norm": 0.384811836945055,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.2187,
"step": 860
},
{
"epoch": 3.492151898734177,
"grad_norm": 0.390470780832265,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.2134,
"step": 861
},
{
"epoch": 3.4962025316455696,
"grad_norm": 0.3926324392948825,
"learning_rate": 9.950885880977891e-06,
"loss": 0.2288,
"step": 862
},
{
"epoch": 3.500253164556962,
"grad_norm": 0.3997341312361244,
"learning_rate": 9.901852696139382e-06,
"loss": 0.219,
"step": 863
},
{
"epoch": 3.5043037974683546,
"grad_norm": 0.38949164489727783,
"learning_rate": 9.852900840391027e-06,
"loss": 0.2486,
"step": 864
},
{
"epoch": 3.5083544303797467,
"grad_norm": 0.3881043586916374,
"learning_rate": 9.804030707984313e-06,
"loss": 0.2452,
"step": 865
},
{
"epoch": 3.512405063291139,
"grad_norm": 0.3980374823415403,
"learning_rate": 9.755242692512599e-06,
"loss": 0.2315,
"step": 866
},
{
"epoch": 3.5164556962025317,
"grad_norm": 0.3868787786339173,
"learning_rate": 9.70653718690782e-06,
"loss": 0.2367,
"step": 867
},
{
"epoch": 3.520506329113924,
"grad_norm": 0.3862530365147463,
"learning_rate": 9.657914583437454e-06,
"loss": 0.2139,
"step": 868
},
{
"epoch": 3.5245569620253163,
"grad_norm": 0.3851058684135963,
"learning_rate": 9.609375273701246e-06,
"loss": 0.2238,
"step": 869
},
{
"epoch": 3.528607594936709,
"grad_norm": 0.39491796310113825,
"learning_rate": 9.560919648628133e-06,
"loss": 0.2225,
"step": 870
},
{
"epoch": 3.5326582278481014,
"grad_norm": 0.3788249890233726,
"learning_rate": 9.512548098473047e-06,
"loss": 0.2317,
"step": 871
},
{
"epoch": 3.536708860759494,
"grad_norm": 0.3947871373059924,
"learning_rate": 9.464261012813825e-06,
"loss": 0.2379,
"step": 872
},
{
"epoch": 3.540759493670886,
"grad_norm": 0.38378107791780747,
"learning_rate": 9.416058780547987e-06,
"loss": 0.2136,
"step": 873
},
{
"epoch": 3.5448101265822785,
"grad_norm": 0.3936538858483406,
"learning_rate": 9.367941789889714e-06,
"loss": 0.2329,
"step": 874
},
{
"epoch": 3.548860759493671,
"grad_norm": 0.3775841670901123,
"learning_rate": 9.319910428366607e-06,
"loss": 0.2554,
"step": 875
},
{
"epoch": 3.552911392405063,
"grad_norm": 0.3910314976121111,
"learning_rate": 9.271965082816667e-06,
"loss": 0.2327,
"step": 876
},
{
"epoch": 3.5569620253164556,
"grad_norm": 0.3877935809448209,
"learning_rate": 9.224106139385111e-06,
"loss": 0.2196,
"step": 877
},
{
"epoch": 3.561012658227848,
"grad_norm": 0.41093097188535,
"learning_rate": 9.176333983521291e-06,
"loss": 0.2396,
"step": 878
},
{
"epoch": 3.5650632911392406,
"grad_norm": 0.37966959915360354,
"learning_rate": 9.12864899997558e-06,
"loss": 0.2314,
"step": 879
},
{
"epoch": 3.569113924050633,
"grad_norm": 0.39108280896948133,
"learning_rate": 9.08105157279628e-06,
"loss": 0.2089,
"step": 880
},
{
"epoch": 3.573164556962025,
"grad_norm": 0.39042467313014595,
"learning_rate": 9.03354208532653e-06,
"loss": 0.2119,
"step": 881
},
{
"epoch": 3.5772151898734177,
"grad_norm": 0.3804572371050465,
"learning_rate": 8.986120920201205e-06,
"loss": 0.191,
"step": 882
},
{
"epoch": 3.58126582278481,
"grad_norm": 0.38996151119055356,
"learning_rate": 8.938788459343852e-06,
"loss": 0.2247,
"step": 883
},
{
"epoch": 3.5853164556962023,
"grad_norm": 0.4023878814903596,
"learning_rate": 8.8915450839636e-06,
"loss": 0.2298,
"step": 884
},
{
"epoch": 3.589367088607595,
"grad_norm": 0.4081808251818312,
"learning_rate": 8.844391174552116e-06,
"loss": 0.2292,
"step": 885
},
{
"epoch": 3.5934177215189873,
"grad_norm": 0.3808897320997891,
"learning_rate": 8.797327110880479e-06,
"loss": 0.2663,
"step": 886
},
{
"epoch": 3.59746835443038,
"grad_norm": 0.38934229823427086,
"learning_rate": 8.750353271996206e-06,
"loss": 0.2174,
"step": 887
},
{
"epoch": 3.6015189873417723,
"grad_norm": 0.4160036426058114,
"learning_rate": 8.703470036220132e-06,
"loss": 0.2065,
"step": 888
},
{
"epoch": 3.6055696202531644,
"grad_norm": 0.38807923391093013,
"learning_rate": 8.656677781143394e-06,
"loss": 0.2245,
"step": 889
},
{
"epoch": 3.609620253164557,
"grad_norm": 0.40978758460290404,
"learning_rate": 8.609976883624377e-06,
"loss": 0.2193,
"step": 890
},
{
"epoch": 3.6136708860759494,
"grad_norm": 0.4037176068639973,
"learning_rate": 8.563367719785698e-06,
"loss": 0.1987,
"step": 891
},
{
"epoch": 3.6177215189873415,
"grad_norm": 0.41100402697638816,
"learning_rate": 8.516850665011138e-06,
"loss": 0.2087,
"step": 892
},
{
"epoch": 3.621772151898734,
"grad_norm": 0.3944338127804413,
"learning_rate": 8.47042609394269e-06,
"loss": 0.233,
"step": 893
},
{
"epoch": 3.6258227848101265,
"grad_norm": 0.39765271109609535,
"learning_rate": 8.424094380477432e-06,
"loss": 0.2573,
"step": 894
},
{
"epoch": 3.629873417721519,
"grad_norm": 0.397707748895049,
"learning_rate": 8.37785589776465e-06,
"loss": 0.2329,
"step": 895
},
{
"epoch": 3.6339240506329116,
"grad_norm": 0.38544630516570594,
"learning_rate": 8.331711018202694e-06,
"loss": 0.2455,
"step": 896
},
{
"epoch": 3.6379746835443036,
"grad_norm": 0.3886198840383641,
"learning_rate": 8.285660113436104e-06,
"loss": 0.2274,
"step": 897
},
{
"epoch": 3.642025316455696,
"grad_norm": 0.3758536120942117,
"learning_rate": 8.239703554352527e-06,
"loss": 0.2501,
"step": 898
},
{
"epoch": 3.6460759493670887,
"grad_norm": 0.39818314451738396,
"learning_rate": 8.193841711079775e-06,
"loss": 0.2041,
"step": 899
},
{
"epoch": 3.6501265822784807,
"grad_norm": 0.382156848693061,
"learning_rate": 8.148074952982828e-06,
"loss": 0.2275,
"step": 900
},
{
"epoch": 3.6541772151898733,
"grad_norm": 0.3948091496828988,
"learning_rate": 8.102403648660859e-06,
"loss": 0.2127,
"step": 901
},
{
"epoch": 3.6582278481012658,
"grad_norm": 0.41239503743838796,
"learning_rate": 8.056828165944282e-06,
"loss": 0.2375,
"step": 902
},
{
"epoch": 3.6622784810126583,
"grad_norm": 0.38504448382820095,
"learning_rate": 8.011348871891762e-06,
"loss": 0.2395,
"step": 903
},
{
"epoch": 3.666329113924051,
"grad_norm": 0.3902250039224954,
"learning_rate": 7.965966132787287e-06,
"loss": 0.2112,
"step": 904
},
{
"epoch": 3.670379746835443,
"grad_norm": 0.3987610934829964,
"learning_rate": 7.920680314137189e-06,
"loss": 0.2416,
"step": 905
},
{
"epoch": 3.6744303797468354,
"grad_norm": 0.40566905976791195,
"learning_rate": 7.875491780667246e-06,
"loss": 0.2466,
"step": 906
},
{
"epoch": 3.678481012658228,
"grad_norm": 0.39221900651128494,
"learning_rate": 7.830400896319667e-06,
"loss": 0.2355,
"step": 907
},
{
"epoch": 3.68253164556962,
"grad_norm": 0.3969321149951802,
"learning_rate": 7.785408024250259e-06,
"loss": 0.252,
"step": 908
},
{
"epoch": 3.6865822784810125,
"grad_norm": 0.3909003280842836,
"learning_rate": 7.74051352682542e-06,
"loss": 0.2201,
"step": 909
},
{
"epoch": 3.690632911392405,
"grad_norm": 0.39726604361818973,
"learning_rate": 7.695717765619257e-06,
"loss": 0.223,
"step": 910
},
{
"epoch": 3.6946835443037975,
"grad_norm": 0.3798866420030086,
"learning_rate": 7.651021101410673e-06,
"loss": 0.2688,
"step": 911
},
{
"epoch": 3.69873417721519,
"grad_norm": 0.38795227556669615,
"learning_rate": 7.606423894180464e-06,
"loss": 0.2255,
"step": 912
},
{
"epoch": 3.702784810126582,
"grad_norm": 0.37182142442504806,
"learning_rate": 7.56192650310839e-06,
"loss": 0.269,
"step": 913
},
{
"epoch": 3.7068354430379746,
"grad_norm": 0.3890517271643863,
"learning_rate": 7.517529286570349e-06,
"loss": 0.2445,
"step": 914
},
{
"epoch": 3.710886075949367,
"grad_norm": 0.37452214354357927,
"learning_rate": 7.473232602135387e-06,
"loss": 0.2265,
"step": 915
},
{
"epoch": 3.714936708860759,
"grad_norm": 0.3840924460825667,
"learning_rate": 7.429036806562935e-06,
"loss": 0.2264,
"step": 916
},
{
"epoch": 3.7189873417721517,
"grad_norm": 0.3926440882036039,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.2086,
"step": 917
},
{
"epoch": 3.7230379746835442,
"grad_norm": 0.4000645483636327,
"learning_rate": 7.340949304977567e-06,
"loss": 0.2193,
"step": 918
},
{
"epoch": 3.7270886075949368,
"grad_norm": 0.3891329029243332,
"learning_rate": 7.297058308409282e-06,
"loss": 0.2361,
"step": 919
},
{
"epoch": 3.7311392405063293,
"grad_norm": 0.39114018349335905,
"learning_rate": 7.25326961958704e-06,
"loss": 0.2343,
"step": 920
},
{
"epoch": 3.7351898734177214,
"grad_norm": 0.39904335136281527,
"learning_rate": 7.209583591178921e-06,
"loss": 0.2187,
"step": 921
},
{
"epoch": 3.739240506329114,
"grad_norm": 0.38767487446815,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.2205,
"step": 922
},
{
"epoch": 3.7432911392405064,
"grad_norm": 0.3986192837336703,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.2214,
"step": 923
},
{
"epoch": 3.747341772151899,
"grad_norm": 0.3868921190158506,
"learning_rate": 7.079144982700909e-06,
"loss": 0.2341,
"step": 924
},
{
"epoch": 3.7513924050632914,
"grad_norm": 0.3879385467955555,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.2526,
"step": 925
},
{
"epoch": 3.7554430379746835,
"grad_norm": 0.3773516187785109,
"learning_rate": 6.99270564069757e-06,
"loss": 0.234,
"step": 926
},
{
"epoch": 3.759493670886076,
"grad_norm": 0.3756641250687707,
"learning_rate": 6.949642934304375e-06,
"loss": 0.2207,
"step": 927
},
{
"epoch": 3.7635443037974685,
"grad_norm": 0.3962871666717583,
"learning_rate": 6.906685333692871e-06,
"loss": 0.2221,
"step": 928
},
{
"epoch": 3.7675949367088606,
"grad_norm": 0.3780626843029462,
"learning_rate": 6.86383318483769e-06,
"loss": 0.2264,
"step": 929
},
{
"epoch": 3.771645569620253,
"grad_norm": 0.3869694250686247,
"learning_rate": 6.821086832864139e-06,
"loss": 0.2154,
"step": 930
},
{
"epoch": 3.7756962025316456,
"grad_norm": 0.3780994664824515,
"learning_rate": 6.77844662204546e-06,
"loss": 0.2227,
"step": 931
},
{
"epoch": 3.779746835443038,
"grad_norm": 0.37975217461627075,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.2211,
"step": 932
},
{
"epoch": 3.7837974683544306,
"grad_norm": 0.3952458664487907,
"learning_rate": 6.693485996688695e-06,
"loss": 0.2447,
"step": 933
},
{
"epoch": 3.7878481012658227,
"grad_norm": 0.39835040200284305,
"learning_rate": 6.651166266411801e-06,
"loss": 0.2014,
"step": 934
},
{
"epoch": 3.7918987341772152,
"grad_norm": 0.3786372048353804,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.2269,
"step": 935
},
{
"epoch": 3.7959493670886078,
"grad_norm": 0.377662680058512,
"learning_rate": 6.566849674844711e-06,
"loss": 0.2072,
"step": 936
},
{
"epoch": 3.8,
"grad_norm": 0.3709874130217897,
"learning_rate": 6.524853492628747e-06,
"loss": 0.2418,
"step": 937
},
{
"epoch": 3.8040506329113923,
"grad_norm": 0.392466917790635,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.2238,
"step": 938
},
{
"epoch": 3.808101265822785,
"grad_norm": 0.3767432548986641,
"learning_rate": 6.441187046486648e-06,
"loss": 0.2539,
"step": 939
},
{
"epoch": 3.8121518987341774,
"grad_norm": 0.378914983316852,
"learning_rate": 6.399517456398567e-06,
"loss": 0.2102,
"step": 940
},
{
"epoch": 3.81620253164557,
"grad_norm": 0.379321202220843,
"learning_rate": 6.357957402727164e-06,
"loss": 0.2265,
"step": 941
},
{
"epoch": 3.820253164556962,
"grad_norm": 0.364585839098531,
"learning_rate": 6.316507220191395e-06,
"loss": 0.2638,
"step": 942
},
{
"epoch": 3.8243037974683545,
"grad_norm": 0.3953768332765741,
"learning_rate": 6.275167242625331e-06,
"loss": 0.2401,
"step": 943
},
{
"epoch": 3.828354430379747,
"grad_norm": 0.38491918706128453,
"learning_rate": 6.233937802975471e-06,
"loss": 0.2173,
"step": 944
},
{
"epoch": 3.832405063291139,
"grad_norm": 0.3685605226354334,
"learning_rate": 6.192819233298046e-06,
"loss": 0.2474,
"step": 945
},
{
"epoch": 3.8364556962025316,
"grad_norm": 0.3944622901807385,
"learning_rate": 6.151811864756383e-06,
"loss": 0.2312,
"step": 946
},
{
"epoch": 3.840506329113924,
"grad_norm": 0.3830358023560218,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.2055,
"step": 947
},
{
"epoch": 3.8445569620253166,
"grad_norm": 0.38523945259104403,
"learning_rate": 6.070132051252868e-06,
"loss": 0.2104,
"step": 948
},
{
"epoch": 3.848607594936709,
"grad_norm": 0.3827076639906433,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.2197,
"step": 949
},
{
"epoch": 3.852658227848101,
"grad_norm": 0.38779525210024773,
"learning_rate": 5.988900993811575e-06,
"loss": 0.2142,
"step": 950
},
{
"epoch": 3.8567088607594937,
"grad_norm": 0.38520507669338977,
"learning_rate": 5.948454566959363e-06,
"loss": 0.2077,
"step": 951
},
{
"epoch": 3.8607594936708862,
"grad_norm": 0.3892798183765175,
"learning_rate": 5.908121309322328e-06,
"loss": 0.2393,
"step": 952
},
{
"epoch": 3.8648101265822783,
"grad_norm": 0.3777200911762426,
"learning_rate": 5.867901545738976e-06,
"loss": 0.2293,
"step": 953
},
{
"epoch": 3.868860759493671,
"grad_norm": 0.36929529861832805,
"learning_rate": 5.827795600133774e-06,
"loss": 0.242,
"step": 954
},
{
"epoch": 3.8729113924050633,
"grad_norm": 0.3864029115208743,
"learning_rate": 5.787803795514466e-06,
"loss": 0.2233,
"step": 955
},
{
"epoch": 3.876962025316456,
"grad_norm": 0.3857913747045394,
"learning_rate": 5.747926453969576e-06,
"loss": 0.2384,
"step": 956
},
{
"epoch": 3.8810126582278484,
"grad_norm": 0.39597704143620727,
"learning_rate": 5.708163896665708e-06,
"loss": 0.2451,
"step": 957
},
{
"epoch": 3.8850632911392404,
"grad_norm": 0.391634186693407,
"learning_rate": 5.668516443845047e-06,
"loss": 0.2289,
"step": 958
},
{
"epoch": 3.889113924050633,
"grad_norm": 0.3751277749603061,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.2267,
"step": 959
},
{
"epoch": 3.8931645569620255,
"grad_norm": 0.38495780023699977,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.2486,
"step": 960
},
{
"epoch": 3.8972151898734175,
"grad_norm": 0.3589726709526898,
"learning_rate": 5.550267900783019e-06,
"loss": 0.2261,
"step": 961
},
{
"epoch": 3.90126582278481,
"grad_norm": 0.3786705106088617,
"learning_rate": 5.511084049737623e-06,
"loss": 0.2171,
"step": 962
},
{
"epoch": 3.9053164556962026,
"grad_norm": 0.3781244393261993,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.2526,
"step": 963
},
{
"epoch": 3.909367088607595,
"grad_norm": 0.37592579259641146,
"learning_rate": 5.433066737499948e-06,
"loss": 0.2255,
"step": 964
},
{
"epoch": 3.9134177215189876,
"grad_norm": 0.3764036018398716,
"learning_rate": 5.394233904648376e-06,
"loss": 0.2366,
"step": 965
},
{
"epoch": 3.9174683544303797,
"grad_norm": 0.366522638738241,
"learning_rate": 5.355518704628997e-06,
"loss": 0.2329,
"step": 966
},
{
"epoch": 3.921518987341772,
"grad_norm": 0.3665831296715056,
"learning_rate": 5.316921449248731e-06,
"loss": 0.224,
"step": 967
},
{
"epoch": 3.9255696202531647,
"grad_norm": 0.38970700047091067,
"learning_rate": 5.278442449364538e-06,
"loss": 0.2163,
"step": 968
},
{
"epoch": 3.9296202531645568,
"grad_norm": 0.380062211660772,
"learning_rate": 5.240082014881016e-06,
"loss": 0.2101,
"step": 969
},
{
"epoch": 3.9336708860759493,
"grad_norm": 0.3740984226726544,
"learning_rate": 5.201840454747822e-06,
"loss": 0.2199,
"step": 970
},
{
"epoch": 3.937721518987342,
"grad_norm": 0.3726076910877676,
"learning_rate": 5.163718076957223e-06,
"loss": 0.2349,
"step": 971
},
{
"epoch": 3.9417721518987343,
"grad_norm": 0.38941582604738745,
"learning_rate": 5.125715188541609e-06,
"loss": 0.2143,
"step": 972
},
{
"epoch": 3.945822784810127,
"grad_norm": 0.37735484567834443,
"learning_rate": 5.087832095571021e-06,
"loss": 0.236,
"step": 973
},
{
"epoch": 3.949873417721519,
"grad_norm": 0.378366807902618,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.2268,
"step": 974
},
{
"epoch": 3.9539240506329114,
"grad_norm": 0.3730855416801937,
"learning_rate": 5.01242651541854e-06,
"loss": 0.2387,
"step": 975
},
{
"epoch": 3.957974683544304,
"grad_norm": 0.37022151462069175,
"learning_rate": 4.974904635542815e-06,
"loss": 0.2418,
"step": 976
},
{
"epoch": 3.962025316455696,
"grad_norm": 0.3807584801459076,
"learning_rate": 4.937503765719582e-06,
"loss": 0.2161,
"step": 977
},
{
"epoch": 3.9660759493670885,
"grad_norm": 0.3781300521209357,
"learning_rate": 4.900224207170299e-06,
"loss": 0.2315,
"step": 978
},
{
"epoch": 3.970126582278481,
"grad_norm": 0.37958003695832954,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.2168,
"step": 979
},
{
"epoch": 3.9741772151898735,
"grad_norm": 0.3801041061386129,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.2431,
"step": 980
},
{
"epoch": 3.978227848101266,
"grad_norm": 0.4037274909667036,
"learning_rate": 4.789116396710924e-06,
"loss": 0.2262,
"step": 981
},
{
"epoch": 3.982278481012658,
"grad_norm": 0.36415055926933526,
"learning_rate": 4.752325075895368e-06,
"loss": 0.2284,
"step": 982
},
{
"epoch": 3.9863291139240506,
"grad_norm": 0.3649804395416501,
"learning_rate": 4.715656557757473e-06,
"loss": 0.2508,
"step": 983
},
{
"epoch": 3.990379746835443,
"grad_norm": 0.3827470258038263,
"learning_rate": 4.679111137620442e-06,
"loss": 0.243,
"step": 984
},
{
"epoch": 3.9944303797468352,
"grad_norm": 0.3848835105020911,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.2081,
"step": 985
},
{
"epoch": 3.9984810126582278,
"grad_norm": 0.3644754910521408,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.2268,
"step": 986
},
{
"epoch": 4.004050632911392,
"grad_norm": 0.5527494409585632,
"learning_rate": 4.570216403561141e-06,
"loss": 0.1716,
"step": 987
},
{
"epoch": 4.008101265822785,
"grad_norm": 0.41813052579791465,
"learning_rate": 4.534166308795815e-06,
"loss": 0.1659,
"step": 988
},
{
"epoch": 4.012151898734177,
"grad_norm": 0.3877086500521018,
"learning_rate": 4.498240773728859e-06,
"loss": 0.1484,
"step": 989
},
{
"epoch": 4.01620253164557,
"grad_norm": 0.2905131423722041,
"learning_rate": 4.462440087699609e-06,
"loss": 0.1601,
"step": 990
},
{
"epoch": 4.020253164556962,
"grad_norm": 0.33246750276478687,
"learning_rate": 4.426764539041861e-06,
"loss": 0.1766,
"step": 991
},
{
"epoch": 4.024303797468354,
"grad_norm": 0.40267761266313223,
"learning_rate": 4.391214415081582e-06,
"loss": 0.1577,
"step": 992
},
{
"epoch": 4.028354430379747,
"grad_norm": 0.43329069453025687,
"learning_rate": 4.355790002134579e-06,
"loss": 0.1446,
"step": 993
},
{
"epoch": 4.032405063291139,
"grad_norm": 0.5069942502799991,
"learning_rate": 4.320491585504207e-06,
"loss": 0.1605,
"step": 994
},
{
"epoch": 4.036455696202531,
"grad_norm": 0.40096062073442795,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.1437,
"step": 995
},
{
"epoch": 4.040506329113924,
"grad_norm": 0.3311614346316088,
"learning_rate": 4.250273877330691e-06,
"loss": 0.1514,
"step": 996
},
{
"epoch": 4.044556962025316,
"grad_norm": 0.34649580708404887,
"learning_rate": 4.215355151311313e-06,
"loss": 0.1624,
"step": 997
},
{
"epoch": 4.048607594936709,
"grad_norm": 0.3408948932282546,
"learning_rate": 4.180563552651542e-06,
"loss": 0.1679,
"step": 998
},
{
"epoch": 4.052658227848101,
"grad_norm": 0.35789498713690243,
"learning_rate": 4.145899361558147e-06,
"loss": 0.1412,
"step": 999
},
{
"epoch": 4.056708860759493,
"grad_norm": 0.37062589544160834,
"learning_rate": 4.111362857211738e-06,
"loss": 0.1613,
"step": 1000
},
{
"epoch": 4.060759493670886,
"grad_norm": 0.3608656867159992,
"learning_rate": 4.076954317764592e-06,
"loss": 0.2,
"step": 1001
},
{
"epoch": 4.0648101265822785,
"grad_norm": 0.32202401115940527,
"learning_rate": 4.042674020338335e-06,
"loss": 0.1587,
"step": 1002
},
{
"epoch": 4.0688607594936705,
"grad_norm": 0.3124942406733075,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.134,
"step": 1003
},
{
"epoch": 4.0729113924050635,
"grad_norm": 0.3126545912998488,
"learning_rate": 3.974499254868674e-06,
"loss": 0.1562,
"step": 1004
},
{
"epoch": 4.076962025316456,
"grad_norm": 0.3079557191921071,
"learning_rate": 3.940605335895451e-06,
"loss": 0.1333,
"step": 1005
},
{
"epoch": 4.0810126582278485,
"grad_norm": 0.33662440397979726,
"learning_rate": 3.90684075707908e-06,
"loss": 0.1474,
"step": 1006
},
{
"epoch": 4.085063291139241,
"grad_norm": 0.32997062850508746,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.1502,
"step": 1007
},
{
"epoch": 4.089113924050633,
"grad_norm": 0.32323562568778486,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.163,
"step": 1008
},
{
"epoch": 4.093164556962026,
"grad_norm": 0.29373924697100634,
"learning_rate": 3.806325775702304e-06,
"loss": 0.1828,
"step": 1009
},
{
"epoch": 4.097215189873418,
"grad_norm": 0.30658409879649906,
"learning_rate": 3.773081266416434e-06,
"loss": 0.1483,
"step": 1010
},
{
"epoch": 4.10126582278481,
"grad_norm": 0.2945243542355097,
"learning_rate": 3.739967446503245e-06,
"loss": 0.1603,
"step": 1011
},
{
"epoch": 4.105316455696203,
"grad_norm": 0.30640232364166575,
"learning_rate": 3.706984582656894e-06,
"loss": 0.1644,
"step": 1012
},
{
"epoch": 4.109367088607595,
"grad_norm": 0.31296119326413147,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.1619,
"step": 1013
},
{
"epoch": 4.113417721518988,
"grad_norm": 0.2830771391540428,
"learning_rate": 3.641412784665648e-06,
"loss": 0.1295,
"step": 1014
},
{
"epoch": 4.11746835443038,
"grad_norm": 0.28077985927821847,
"learning_rate": 3.608824378627005e-06,
"loss": 0.1392,
"step": 1015
},
{
"epoch": 4.121518987341772,
"grad_norm": 0.290039333786592,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.1675,
"step": 1016
},
{
"epoch": 4.125569620253165,
"grad_norm": 0.2979223763522679,
"learning_rate": 3.544043864774269e-06,
"loss": 0.1537,
"step": 1017
},
{
"epoch": 4.129620253164557,
"grad_norm": 0.302660816853464,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.1368,
"step": 1018
},
{
"epoch": 4.133670886075949,
"grad_norm": 0.30695920580823616,
"learning_rate": 3.479793485887819e-06,
"loss": 0.1706,
"step": 1019
},
{
"epoch": 4.137721518987342,
"grad_norm": 0.30005566362658836,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.1364,
"step": 1020
},
{
"epoch": 4.141772151898734,
"grad_norm": 0.2981351600521079,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.1639,
"step": 1021
},
{
"epoch": 4.145822784810127,
"grad_norm": 0.28963205970324835,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.1564,
"step": 1022
},
{
"epoch": 4.149873417721519,
"grad_norm": 0.28785722791035423,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.153,
"step": 1023
},
{
"epoch": 4.153924050632911,
"grad_norm": 0.3002935089121761,
"learning_rate": 3.321500420335e-06,
"loss": 0.1605,
"step": 1024
},
{
"epoch": 4.157974683544304,
"grad_norm": 0.29187155001216936,
"learning_rate": 3.290243771741275e-06,
"loss": 0.1537,
"step": 1025
},
{
"epoch": 4.162025316455696,
"grad_norm": 0.2853312598238615,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.1415,
"step": 1026
},
{
"epoch": 4.166075949367088,
"grad_norm": 0.29276451346913546,
"learning_rate": 3.228134459445149e-06,
"loss": 0.1742,
"step": 1027
},
{
"epoch": 4.170126582278481,
"grad_norm": 0.29819168095218285,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.1498,
"step": 1028
},
{
"epoch": 4.174177215189873,
"grad_norm": 0.28373352319403333,
"learning_rate": 3.166565459258513e-06,
"loss": 0.159,
"step": 1029
},
{
"epoch": 4.178227848101266,
"grad_norm": 0.2847132504238253,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.1459,
"step": 1030
},
{
"epoch": 4.182278481012658,
"grad_norm": 0.298467378912953,
"learning_rate": 3.105538754650419e-06,
"loss": 0.1577,
"step": 1031
},
{
"epoch": 4.18632911392405,
"grad_norm": 0.2947156447095453,
"learning_rate": 3.07522937824712e-06,
"loss": 0.1698,
"step": 1032
},
{
"epoch": 4.190379746835443,
"grad_norm": 0.28694127783793194,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.1618,
"step": 1033
},
{
"epoch": 4.194430379746835,
"grad_norm": 0.298041551315183,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.1683,
"step": 1034
},
{
"epoch": 4.1984810126582275,
"grad_norm": 0.2777737543983037,
"learning_rate": 2.985120078631465e-06,
"loss": 0.1608,
"step": 1035
},
{
"epoch": 4.2025316455696204,
"grad_norm": 0.2907880948164746,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.1779,
"step": 1036
},
{
"epoch": 4.2065822784810125,
"grad_norm": 0.29770654981409644,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.1467,
"step": 1037
},
{
"epoch": 4.2106329113924055,
"grad_norm": 0.2941449500421362,
"learning_rate": 2.896244091927678e-06,
"loss": 0.1629,
"step": 1038
},
{
"epoch": 4.2146835443037975,
"grad_norm": 0.29228887759195193,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.1736,
"step": 1039
},
{
"epoch": 4.21873417721519,
"grad_norm": 0.30005069984935206,
"learning_rate": 2.837681792951994e-06,
"loss": 0.1507,
"step": 1040
},
{
"epoch": 4.222784810126583,
"grad_norm": 0.28529202119965014,
"learning_rate": 2.808607860255981e-06,
"loss": 0.1623,
"step": 1041
},
{
"epoch": 4.226835443037975,
"grad_norm": 0.2809873225182775,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.1472,
"step": 1042
},
{
"epoch": 4.230886075949367,
"grad_norm": 0.31289884351845604,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.1331,
"step": 1043
},
{
"epoch": 4.23493670886076,
"grad_norm": 0.28790977532792733,
"learning_rate": 2.722217735873718e-06,
"loss": 0.1297,
"step": 1044
},
{
"epoch": 4.238987341772152,
"grad_norm": 0.30631962668686896,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.1419,
"step": 1045
},
{
"epoch": 4.243037974683545,
"grad_norm": 0.289061235706911,
"learning_rate": 2.665319697490205e-06,
"loss": 0.1661,
"step": 1046
},
{
"epoch": 4.247088607594937,
"grad_norm": 0.2886962330676168,
"learning_rate": 2.637079980714945e-06,
"loss": 0.1592,
"step": 1047
},
{
"epoch": 4.251139240506329,
"grad_norm": 0.2926096923178884,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.1797,
"step": 1048
},
{
"epoch": 4.255189873417722,
"grad_norm": 0.2749906647894509,
"learning_rate": 2.581020289189571e-06,
"loss": 0.1683,
"step": 1049
},
{
"epoch": 4.259240506329114,
"grad_norm": 0.29404053938939223,
"learning_rate": 2.553200765936501e-06,
"loss": 0.1668,
"step": 1050
},
{
"epoch": 4.263291139240506,
"grad_norm": 0.2928447509790546,
"learning_rate": 2.525521756805962e-06,
"loss": 0.1571,
"step": 1051
},
{
"epoch": 4.267341772151899,
"grad_norm": 0.3040988180388799,
"learning_rate": 2.497983484720885e-06,
"loss": 0.1687,
"step": 1052
},
{
"epoch": 4.271392405063291,
"grad_norm": 0.29630636322023013,
"learning_rate": 2.470586171470728e-06,
"loss": 0.1527,
"step": 1053
},
{
"epoch": 4.275443037974684,
"grad_norm": 0.295127497777346,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.183,
"step": 1054
},
{
"epoch": 4.279493670886076,
"grad_norm": 0.2988533187529217,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.1639,
"step": 1055
},
{
"epoch": 4.283544303797468,
"grad_norm": 0.2902437655556626,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.1597,
"step": 1056
},
{
"epoch": 4.287594936708861,
"grad_norm": 0.30579516900524245,
"learning_rate": 2.362410902836978e-06,
"loss": 0.143,
"step": 1057
},
{
"epoch": 4.291645569620253,
"grad_norm": 0.27857601602119975,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.1501,
"step": 1058
},
{
"epoch": 4.295696202531645,
"grad_norm": 0.29257406932122687,
"learning_rate": 2.309174704446411e-06,
"loss": 0.158,
"step": 1059
},
{
"epoch": 4.299746835443038,
"grad_norm": 0.28170625901041746,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.1729,
"step": 1060
},
{
"epoch": 4.30379746835443,
"grad_norm": 0.2869064768457476,
"learning_rate": 2.256508422807855e-06,
"loss": 0.1389,
"step": 1061
},
{
"epoch": 4.307848101265823,
"grad_norm": 0.30496847320224735,
"learning_rate": 2.230389531696946e-06,
"loss": 0.1607,
"step": 1062
},
{
"epoch": 4.311898734177215,
"grad_norm": 0.2945064775903355,
"learning_rate": 2.204413754585857e-06,
"loss": 0.1591,
"step": 1063
},
{
"epoch": 4.315949367088607,
"grad_norm": 0.286674293933279,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.1856,
"step": 1064
},
{
"epoch": 4.32,
"grad_norm": 0.29176918297568094,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.1398,
"step": 1065
},
{
"epoch": 4.324050632911392,
"grad_norm": 0.27944373546931955,
"learning_rate": 2.127347193531757e-06,
"loss": 0.1534,
"step": 1066
},
{
"epoch": 4.328101265822784,
"grad_norm": 0.29833139856074137,
"learning_rate": 2.101945952921942e-06,
"loss": 0.1473,
"step": 1067
},
{
"epoch": 4.332151898734177,
"grad_norm": 0.2811281589642247,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.1464,
"step": 1068
},
{
"epoch": 4.3362025316455695,
"grad_norm": 0.2839203580582184,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.1542,
"step": 1069
},
{
"epoch": 4.340253164556962,
"grad_norm": 0.2875836349536334,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.1475,
"step": 1070
},
{
"epoch": 4.3443037974683545,
"grad_norm": 0.2759365902686032,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.1461,
"step": 1071
},
{
"epoch": 4.348354430379747,
"grad_norm": 0.2885671092674566,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.1714,
"step": 1072
},
{
"epoch": 4.3524050632911395,
"grad_norm": 0.2763970557001938,
"learning_rate": 1.95257270594031e-06,
"loss": 0.1459,
"step": 1073
},
{
"epoch": 4.356455696202532,
"grad_norm": 0.28881178030119203,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.1645,
"step": 1074
},
{
"epoch": 4.360506329113924,
"grad_norm": 0.29202777818846853,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.1325,
"step": 1075
},
{
"epoch": 4.364556962025317,
"grad_norm": 0.28374453540434547,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.1568,
"step": 1076
},
{
"epoch": 4.368607594936709,
"grad_norm": 0.29988681818947865,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.153,
"step": 1077
},
{
"epoch": 4.372658227848102,
"grad_norm": 0.2983167060996779,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.1725,
"step": 1078
},
{
"epoch": 4.376708860759494,
"grad_norm": 0.2923604657172818,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.1618,
"step": 1079
},
{
"epoch": 4.380759493670886,
"grad_norm": 0.28842811913941824,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.1338,
"step": 1080
},
{
"epoch": 4.384810126582279,
"grad_norm": 0.2717893733171465,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.1408,
"step": 1081
},
{
"epoch": 4.388860759493671,
"grad_norm": 0.2972710909396453,
"learning_rate": 1.738336950434061e-06,
"loss": 0.1332,
"step": 1082
},
{
"epoch": 4.392911392405063,
"grad_norm": 0.2793711115179343,
"learning_rate": 1.715265841375957e-06,
"loss": 0.1471,
"step": 1083
},
{
"epoch": 4.396962025316456,
"grad_norm": 0.2944025127177289,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.1626,
"step": 1084
},
{
"epoch": 4.401012658227848,
"grad_norm": 0.29034317336464543,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.1806,
"step": 1085
},
{
"epoch": 4.405063291139241,
"grad_norm": 0.2701946895033376,
"learning_rate": 1.646936827927441e-06,
"loss": 0.1326,
"step": 1086
},
{
"epoch": 4.409113924050633,
"grad_norm": 0.2847153233279307,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.1431,
"step": 1087
},
{
"epoch": 4.413164556962025,
"grad_norm": 0.28181505358628134,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.134,
"step": 1088
},
{
"epoch": 4.417215189873418,
"grad_norm": 0.28182517959049674,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.1292,
"step": 1089
},
{
"epoch": 4.42126582278481,
"grad_norm": 0.29303251656212004,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.163,
"step": 1090
},
{
"epoch": 4.425316455696202,
"grad_norm": 0.29150398348102763,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.1705,
"step": 1091
},
{
"epoch": 4.429367088607595,
"grad_norm": 0.273184484534952,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.1627,
"step": 1092
},
{
"epoch": 4.433417721518987,
"grad_norm": 0.29179331118706514,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.1404,
"step": 1093
},
{
"epoch": 4.43746835443038,
"grad_norm": 0.2861990371975788,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.1671,
"step": 1094
},
{
"epoch": 4.441518987341772,
"grad_norm": 0.29326120283495216,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.1408,
"step": 1095
},
{
"epoch": 4.445569620253164,
"grad_norm": 0.2905193638545743,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.1583,
"step": 1096
},
{
"epoch": 4.449620253164557,
"grad_norm": 0.27997159421066387,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.145,
"step": 1097
},
{
"epoch": 4.453670886075949,
"grad_norm": 0.2758013445765164,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.1601,
"step": 1098
},
{
"epoch": 4.457721518987341,
"grad_norm": 0.2868934854194082,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.1489,
"step": 1099
},
{
"epoch": 4.461772151898734,
"grad_norm": 0.3045157159294296,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.1692,
"step": 1100
},
{
"epoch": 4.465822784810126,
"grad_norm": 0.288657076845399,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.1459,
"step": 1101
},
{
"epoch": 4.469873417721519,
"grad_norm": 0.2911012299453321,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.156,
"step": 1102
},
{
"epoch": 4.473924050632911,
"grad_norm": 0.28618526147705137,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.1652,
"step": 1103
},
{
"epoch": 4.4779746835443035,
"grad_norm": 0.28134439156658525,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.1717,
"step": 1104
},
{
"epoch": 4.4820253164556965,
"grad_norm": 0.2768632599098236,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.1412,
"step": 1105
},
{
"epoch": 4.4860759493670885,
"grad_norm": 0.2789054806789415,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.176,
"step": 1106
},
{
"epoch": 4.490126582278481,
"grad_norm": 0.2892656404621324,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.1546,
"step": 1107
},
{
"epoch": 4.494177215189874,
"grad_norm": 0.2853781719149108,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.1347,
"step": 1108
},
{
"epoch": 4.498227848101266,
"grad_norm": 0.2707598004637536,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.1601,
"step": 1109
},
{
"epoch": 4.502278481012659,
"grad_norm": 0.2823200204642309,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.1419,
"step": 1110
},
{
"epoch": 4.506329113924051,
"grad_norm": 0.27686368266178835,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.1545,
"step": 1111
},
{
"epoch": 4.510379746835443,
"grad_norm": 0.2964470689913743,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.1708,
"step": 1112
},
{
"epoch": 4.514430379746836,
"grad_norm": 0.28672018417255435,
"learning_rate": 1.092402023876933e-06,
"loss": 0.164,
"step": 1113
},
{
"epoch": 4.518481012658228,
"grad_norm": 0.2841877670650692,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.1272,
"step": 1114
},
{
"epoch": 4.52253164556962,
"grad_norm": 0.2885033649662011,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.1633,
"step": 1115
},
{
"epoch": 4.526582278481013,
"grad_norm": 0.29437490593635096,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.1858,
"step": 1116
},
{
"epoch": 4.530632911392405,
"grad_norm": 0.292146084931673,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.1628,
"step": 1117
},
{
"epoch": 4.534683544303798,
"grad_norm": 0.299862777066963,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.1524,
"step": 1118
},
{
"epoch": 4.53873417721519,
"grad_norm": 0.28007139179235707,
"learning_rate": 9.841383857091947e-07,
"loss": 0.1483,
"step": 1119
},
{
"epoch": 4.542784810126582,
"grad_norm": 0.28209568013810443,
"learning_rate": 9.666296380321616e-07,
"loss": 0.1745,
"step": 1120
},
{
"epoch": 4.546835443037975,
"grad_norm": 0.2974288945107882,
"learning_rate": 9.492741825034124e-07,
"loss": 0.1479,
"step": 1121
},
{
"epoch": 4.550886075949367,
"grad_norm": 0.2819302562340476,
"learning_rate": 9.320721589013892e-07,
"loss": 0.1431,
"step": 1122
},
{
"epoch": 4.55493670886076,
"grad_norm": 0.3061155682640011,
"learning_rate": 9.150237057688339e-07,
"loss": 0.1707,
"step": 1123
},
{
"epoch": 4.558987341772152,
"grad_norm": 0.27038121612387567,
"learning_rate": 8.981289604116328e-07,
"loss": 0.1521,
"step": 1124
},
{
"epoch": 4.563037974683544,
"grad_norm": 0.2862976948810326,
"learning_rate": 8.813880588977542e-07,
"loss": 0.1423,
"step": 1125
},
{
"epoch": 4.567088607594937,
"grad_norm": 0.2843942870717705,
"learning_rate": 8.648011360561126e-07,
"loss": 0.1537,
"step": 1126
},
{
"epoch": 4.571139240506329,
"grad_norm": 0.287198770378647,
"learning_rate": 8.483683254755037e-07,
"loss": 0.1765,
"step": 1127
},
{
"epoch": 4.575189873417721,
"grad_norm": 0.28179383762947907,
"learning_rate": 8.320897595035227e-07,
"loss": 0.1676,
"step": 1128
},
{
"epoch": 4.579240506329114,
"grad_norm": 0.292343471980248,
"learning_rate": 8.159655692455093e-07,
"loss": 0.1467,
"step": 1129
},
{
"epoch": 4.583291139240506,
"grad_norm": 0.29104465246319233,
"learning_rate": 7.999958845634648e-07,
"loss": 0.144,
"step": 1130
},
{
"epoch": 4.587341772151898,
"grad_norm": 0.28793788288489675,
"learning_rate": 7.841808340750478e-07,
"loss": 0.1736,
"step": 1131
},
{
"epoch": 4.591392405063291,
"grad_norm": 0.28659001775076265,
"learning_rate": 7.685205451524869e-07,
"loss": 0.1218,
"step": 1132
},
{
"epoch": 4.595443037974683,
"grad_norm": 0.299309583924964,
"learning_rate": 7.530151439216027e-07,
"loss": 0.1443,
"step": 1133
},
{
"epoch": 4.599493670886076,
"grad_norm": 0.28496614566439904,
"learning_rate": 7.376647552607675e-07,
"loss": 0.1514,
"step": 1134
},
{
"epoch": 4.603544303797468,
"grad_norm": 0.2796887231795183,
"learning_rate": 7.224695027998963e-07,
"loss": 0.1398,
"step": 1135
},
{
"epoch": 4.6075949367088604,
"grad_norm": 0.2835165542254995,
"learning_rate": 7.07429508919466e-07,
"loss": 0.1636,
"step": 1136
},
{
"epoch": 4.611645569620253,
"grad_norm": 0.2846742506841411,
"learning_rate": 6.925448947495206e-07,
"loss": 0.1355,
"step": 1137
},
{
"epoch": 4.6156962025316455,
"grad_norm": 0.28944585590794775,
"learning_rate": 6.778157801686936e-07,
"loss": 0.158,
"step": 1138
},
{
"epoch": 4.619746835443038,
"grad_norm": 0.2914439792581438,
"learning_rate": 6.632422838032515e-07,
"loss": 0.1762,
"step": 1139
},
{
"epoch": 4.6237974683544305,
"grad_norm": 0.2815752606975582,
"learning_rate": 6.488245230261281e-07,
"loss": 0.1588,
"step": 1140
},
{
"epoch": 4.627848101265823,
"grad_norm": 0.29606342288775744,
"learning_rate": 6.345626139559868e-07,
"loss": 0.1537,
"step": 1141
},
{
"epoch": 4.6318987341772155,
"grad_norm": 0.29854319554232295,
"learning_rate": 6.204566714562866e-07,
"loss": 0.1601,
"step": 1142
},
{
"epoch": 4.635949367088608,
"grad_norm": 0.27713232173540997,
"learning_rate": 6.06506809134344e-07,
"loss": 0.1456,
"step": 1143
},
{
"epoch": 4.64,
"grad_norm": 0.2926661737095211,
"learning_rate": 5.927131393404373e-07,
"loss": 0.1518,
"step": 1144
},
{
"epoch": 4.644050632911393,
"grad_norm": 0.273017900287255,
"learning_rate": 5.790757731668817e-07,
"loss": 0.1447,
"step": 1145
},
{
"epoch": 4.648101265822785,
"grad_norm": 0.2915327509694073,
"learning_rate": 5.655948204471507e-07,
"loss": 0.1586,
"step": 1146
},
{
"epoch": 4.652151898734177,
"grad_norm": 0.2901061796939774,
"learning_rate": 5.522703897549875e-07,
"loss": 0.1633,
"step": 1147
},
{
"epoch": 4.65620253164557,
"grad_norm": 0.2836373975346331,
"learning_rate": 5.391025884035239e-07,
"loss": 0.1682,
"step": 1148
},
{
"epoch": 4.660253164556962,
"grad_norm": 0.2831534775320347,
"learning_rate": 5.260915224444207e-07,
"loss": 0.1609,
"step": 1149
},
{
"epoch": 4.664303797468355,
"grad_norm": 0.2850112171538381,
"learning_rate": 5.132372966670129e-07,
"loss": 0.1431,
"step": 1150
},
{
"epoch": 4.668354430379747,
"grad_norm": 0.2862564051040808,
"learning_rate": 5.005400145974704e-07,
"loss": 0.1561,
"step": 1151
},
{
"epoch": 4.672405063291139,
"grad_norm": 0.27681116490826035,
"learning_rate": 4.879997784979562e-07,
"loss": 0.1653,
"step": 1152
},
{
"epoch": 4.676455696202532,
"grad_norm": 0.2800348408376165,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.122,
"step": 1153
},
{
"epoch": 4.680506329113924,
"grad_norm": 0.2882993360946391,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.1453,
"step": 1154
},
{
"epoch": 4.684556962025317,
"grad_norm": 0.28114881649328505,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.1673,
"step": 1155
},
{
"epoch": 4.688607594936709,
"grad_norm": 0.28589876163433364,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.1565,
"step": 1156
},
{
"epoch": 4.692658227848101,
"grad_norm": 0.29765074357551813,
"learning_rate": 4.27657778143431e-07,
"loss": 0.1619,
"step": 1157
},
{
"epoch": 4.696708860759494,
"grad_norm": 0.2958938962101301,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.1634,
"step": 1158
},
{
"epoch": 4.700759493670886,
"grad_norm": 0.2790461226685561,
"learning_rate": 4.046237371675177e-07,
"loss": 0.1738,
"step": 1159
},
{
"epoch": 4.704810126582278,
"grad_norm": 0.30236883465763703,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.1507,
"step": 1160
},
{
"epoch": 4.708860759493671,
"grad_norm": 0.2778402519264542,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.1508,
"step": 1161
},
{
"epoch": 4.712911392405063,
"grad_norm": 0.3007403511994216,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.177,
"step": 1162
},
{
"epoch": 4.716962025316455,
"grad_norm": 0.2819742505635935,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.1624,
"step": 1163
},
{
"epoch": 4.721012658227848,
"grad_norm": 0.29972720891546584,
"learning_rate": 3.498020255137813e-07,
"loss": 0.1631,
"step": 1164
},
{
"epoch": 4.72506329113924,
"grad_norm": 0.2756885650724342,
"learning_rate": 3.393121163057811e-07,
"loss": 0.1448,
"step": 1165
},
{
"epoch": 4.729113924050633,
"grad_norm": 0.28368336890423806,
"learning_rate": 3.289805515751399e-07,
"loss": 0.1601,
"step": 1166
},
{
"epoch": 4.733164556962025,
"grad_norm": 0.293688027305225,
"learning_rate": 3.188074145308573e-07,
"loss": 0.1719,
"step": 1167
},
{
"epoch": 4.737215189873417,
"grad_norm": 0.2783150307534161,
"learning_rate": 3.087927871059804e-07,
"loss": 0.1387,
"step": 1168
},
{
"epoch": 4.74126582278481,
"grad_norm": 0.2758062234984419,
"learning_rate": 2.989367499569418e-07,
"loss": 0.1627,
"step": 1169
},
{
"epoch": 4.745316455696202,
"grad_norm": 0.28267700382174876,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.1441,
"step": 1170
},
{
"epoch": 4.749367088607595,
"grad_norm": 0.28806016653346117,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.1473,
"step": 1171
},
{
"epoch": 4.7534177215189874,
"grad_norm": 0.2869768642989331,
"learning_rate": 2.703209675663887e-07,
"loss": 0.1374,
"step": 1172
},
{
"epoch": 4.7574683544303795,
"grad_norm": 0.29202254739842376,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.1487,
"step": 1173
},
{
"epoch": 4.7615189873417725,
"grad_norm": 0.2834492357319221,
"learning_rate": 2.520381518804471e-07,
"loss": 0.1635,
"step": 1174
},
{
"epoch": 4.7655696202531646,
"grad_norm": 0.2962832181041542,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.1502,
"step": 1175
},
{
"epoch": 4.769620253164557,
"grad_norm": 0.293571073706303,
"learning_rate": 2.343915243930317e-07,
"loss": 0.14,
"step": 1176
},
{
"epoch": 4.77367088607595,
"grad_norm": 0.28295526903994733,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.167,
"step": 1177
},
{
"epoch": 4.777721518987342,
"grad_norm": 0.27755105000882213,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.1351,
"step": 1178
},
{
"epoch": 4.781772151898734,
"grad_norm": 0.2988881623957229,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.1524,
"step": 1179
},
{
"epoch": 4.785822784810127,
"grad_norm": 0.2856418602014717,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.1662,
"step": 1180
},
{
"epoch": 4.789873417721519,
"grad_norm": 0.287297059478593,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.161,
"step": 1181
},
{
"epoch": 4.793924050632912,
"grad_norm": 0.27990392414030335,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.1439,
"step": 1182
},
{
"epoch": 4.797974683544304,
"grad_norm": 0.2904711492679225,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.1591,
"step": 1183
},
{
"epoch": 4.802025316455696,
"grad_norm": 0.2984372257549442,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.1507,
"step": 1184
},
{
"epoch": 4.806075949367089,
"grad_norm": 0.27609628027450384,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.1535,
"step": 1185
},
{
"epoch": 4.810126582278481,
"grad_norm": 0.2826284626565299,
"learning_rate": 1.557203888516745e-07,
"loss": 0.1487,
"step": 1186
},
{
"epoch": 4.814177215189874,
"grad_norm": 0.28777133326345183,
"learning_rate": 1.487313098987131e-07,
"loss": 0.1317,
"step": 1187
},
{
"epoch": 4.818227848101266,
"grad_norm": 0.28793581432963844,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.1628,
"step": 1188
},
{
"epoch": 4.822278481012658,
"grad_norm": 0.27599682206629744,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.1413,
"step": 1189
},
{
"epoch": 4.826329113924051,
"grad_norm": 0.27849053777888905,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.1725,
"step": 1190
},
{
"epoch": 4.830379746835443,
"grad_norm": 0.27678393734677265,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.1458,
"step": 1191
},
{
"epoch": 4.834430379746835,
"grad_norm": 0.2819189374799565,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.1695,
"step": 1192
},
{
"epoch": 4.838481012658228,
"grad_norm": 0.27586434178814123,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.1542,
"step": 1193
},
{
"epoch": 4.84253164556962,
"grad_norm": 0.2870365100246339,
"learning_rate": 1.042873700093061e-07,
"loss": 0.1631,
"step": 1194
},
{
"epoch": 4.846582278481012,
"grad_norm": 0.29109142328002935,
"learning_rate": 9.857879404556291e-08,
"loss": 0.14,
"step": 1195
},
{
"epoch": 4.850632911392405,
"grad_norm": 0.29011398421584494,
"learning_rate": 9.303050139225722e-08,
"loss": 0.1848,
"step": 1196
},
{
"epoch": 4.854683544303797,
"grad_norm": 0.27846061639593556,
"learning_rate": 8.76425367345779e-08,
"loss": 0.1404,
"step": 1197
},
{
"epoch": 4.85873417721519,
"grad_norm": 0.2827540523945022,
"learning_rate": 8.241494346644897e-08,
"loss": 0.1549,
"step": 1198
},
{
"epoch": 4.862784810126582,
"grad_norm": 0.2948910259623969,
"learning_rate": 7.734776369019204e-08,
"loss": 0.1776,
"step": 1199
},
{
"epoch": 4.866835443037974,
"grad_norm": 0.2910376118132819,
"learning_rate": 7.244103821617332e-08,
"loss": 0.1756,
"step": 1200
},
{
"epoch": 4.870886075949367,
"grad_norm": 0.2904238988608664,
"learning_rate": 6.769480656248606e-08,
"loss": 0.158,
"step": 1201
},
{
"epoch": 4.874936708860759,
"grad_norm": 0.2918769717356335,
"learning_rate": 6.310910695462635e-08,
"loss": 0.18,
"step": 1202
},
{
"epoch": 4.878987341772152,
"grad_norm": 0.2841246538090745,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.1608,
"step": 1203
},
{
"epoch": 4.883037974683544,
"grad_norm": 0.30842301009059453,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.1787,
"step": 1204
},
{
"epoch": 4.8870886075949365,
"grad_norm": 0.2810029955371153,
"learning_rate": 5.031556326567488e-08,
"loss": 0.1512,
"step": 1205
},
{
"epoch": 4.891139240506329,
"grad_norm": 0.28926849308599495,
"learning_rate": 4.637234823364312e-08,
"loss": 0.1543,
"step": 1206
},
{
"epoch": 4.8951898734177215,
"grad_norm": 0.28424126059301436,
"learning_rate": 4.258983697558838e-08,
"loss": 0.1565,
"step": 1207
},
{
"epoch": 4.899240506329114,
"grad_norm": 0.27909137404130857,
"learning_rate": 3.896805995533548e-08,
"loss": 0.1563,
"step": 1208
},
{
"epoch": 4.9032911392405065,
"grad_norm": 0.2722501233026398,
"learning_rate": 3.550704634218028e-08,
"loss": 0.1604,
"step": 1209
},
{
"epoch": 4.907341772151899,
"grad_norm": 0.2742844436081079,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.1745,
"step": 1210
},
{
"epoch": 4.911392405063291,
"grad_norm": 0.28256346151262574,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.139,
"step": 1211
},
{
"epoch": 4.915443037974684,
"grad_norm": 0.2756664011929319,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.1408,
"step": 1212
},
{
"epoch": 4.919493670886076,
"grad_norm": 0.2809131856768422,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.1477,
"step": 1213
},
{
"epoch": 4.923544303797469,
"grad_norm": 0.28622820100493773,
"learning_rate": 2.061435966221881e-08,
"loss": 0.1505,
"step": 1214
},
{
"epoch": 4.927594936708861,
"grad_norm": 0.27583113179883956,
"learning_rate": 1.811846652477245e-08,
"loss": 0.139,
"step": 1215
},
{
"epoch": 4.931645569620253,
"grad_norm": 0.27675828813430026,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.152,
"step": 1216
},
{
"epoch": 4.935696202531646,
"grad_norm": 0.27198593136436955,
"learning_rate": 1.360949303591097e-08,
"loss": 0.1447,
"step": 1217
},
{
"epoch": 4.939746835443038,
"grad_norm": 0.2840451023441452,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.1749,
"step": 1218
},
{
"epoch": 4.943797468354431,
"grad_norm": 0.297347633089949,
"learning_rate": 9.744388816668172e-09,
"loss": 0.1527,
"step": 1219
},
{
"epoch": 4.947848101265823,
"grad_norm": 0.29114783297218294,
"learning_rate": 8.05332740472009e-09,
"loss": 0.1557,
"step": 1220
},
{
"epoch": 4.951898734177215,
"grad_norm": 0.3021636025825127,
"learning_rate": 6.523278382872811e-09,
"loss": 0.1462,
"step": 1221
},
{
"epoch": 4.955949367088608,
"grad_norm": 0.28354873552696547,
"learning_rate": 5.15425407393133e-09,
"loss": 0.1504,
"step": 1222
},
{
"epoch": 4.96,
"grad_norm": 0.2820059043679998,
"learning_rate": 3.94626550383137e-09,
"loss": 0.1299,
"step": 1223
},
{
"epoch": 4.964050632911392,
"grad_norm": 0.2841510276178814,
"learning_rate": 2.899322401546112e-09,
"loss": 0.1686,
"step": 1224
},
{
"epoch": 4.968101265822785,
"grad_norm": 0.28540499413174014,
"learning_rate": 2.013433199010706e-09,
"loss": 0.1485,
"step": 1225
},
{
"epoch": 4.972151898734177,
"grad_norm": 0.28823391899080425,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.1545,
"step": 1226
},
{
"epoch": 4.976202531645569,
"grad_norm": 0.2829342431870061,
"learning_rate": 7.248437353468695e-10,
"loss": 0.1488,
"step": 1227
},
{
"epoch": 4.980253164556962,
"grad_norm": 0.2858718468861255,
"learning_rate": 3.221538523412449e-10,
"loss": 0.1595,
"step": 1228
},
{
"epoch": 4.984303797468354,
"grad_norm": 0.2900893574067172,
"learning_rate": 8.053862524670663e-11,
"loss": 0.1541,
"step": 1229
},
{
"epoch": 4.988354430379747,
"grad_norm": 0.28981501555026234,
"learning_rate": 0.0,
"loss": 0.1428,
"step": 1230
},
{
"epoch": 4.988354430379747,
"step": 1230,
"total_flos": 1.9258602225475256e+18,
"train_loss": 0.441497222437122,
"train_runtime": 69675.7889,
"train_samples_per_second": 2.268,
"train_steps_per_second": 0.018
}
],
"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": 1.9258602225475256e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}