a1_code_codefeedback / trainer_state.json
sedrickkeh's picture
End of training
e7f55ff verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.982278481012658,
"eval_steps": 500,
"global_step": 1230,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004050632911392405,
"grad_norm": 6.474354849322044,
"learning_rate": 3.2520325203252037e-07,
"loss": 1.0642,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 6.483142804178548,
"learning_rate": 6.504065040650407e-07,
"loss": 1.0614,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 6.4040533571345755,
"learning_rate": 9.75609756097561e-07,
"loss": 1.0551,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 6.229930543629941,
"learning_rate": 1.3008130081300815e-06,
"loss": 1.0453,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 6.238919524344026,
"learning_rate": 1.6260162601626018e-06,
"loss": 1.0873,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 5.840608448444309,
"learning_rate": 1.951219512195122e-06,
"loss": 1.0481,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 4.560861002800012,
"learning_rate": 2.2764227642276426e-06,
"loss": 1.0071,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 4.178773905436023,
"learning_rate": 2.601626016260163e-06,
"loss": 1.0075,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 2.801146892972886,
"learning_rate": 2.926829268292683e-06,
"loss": 0.9888,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 2.476410057222459,
"learning_rate": 3.2520325203252037e-06,
"loss": 0.9524,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 2.449142455106207,
"learning_rate": 3.577235772357724e-06,
"loss": 0.9701,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 3.534900089609012,
"learning_rate": 3.902439024390244e-06,
"loss": 0.9285,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 3.8541400719356678,
"learning_rate": 4.227642276422765e-06,
"loss": 0.956,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 3.629971327074421,
"learning_rate": 4.552845528455285e-06,
"loss": 0.9163,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 3.3363727555047364,
"learning_rate": 4.8780487804878055e-06,
"loss": 0.9385,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 2.185882063133234,
"learning_rate": 5.203252032520326e-06,
"loss": 0.8662,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 2.3740336636265806,
"learning_rate": 5.528455284552846e-06,
"loss": 0.9035,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.049495565141068,
"learning_rate": 5.853658536585366e-06,
"loss": 0.8475,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 1.5693509775507133,
"learning_rate": 6.178861788617887e-06,
"loss": 0.836,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.266271844817775,
"learning_rate": 6.504065040650407e-06,
"loss": 0.8242,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.2634518820867353,
"learning_rate": 6.829268292682928e-06,
"loss": 0.8115,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.4533876251677937,
"learning_rate": 7.154471544715448e-06,
"loss": 0.7938,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.3573014972281332,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.7938,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 1.0900136207360065,
"learning_rate": 7.804878048780489e-06,
"loss": 0.7833,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 1.1203907831118303,
"learning_rate": 8.130081300813009e-06,
"loss": 0.7704,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 1.0098324642421042,
"learning_rate": 8.45528455284553e-06,
"loss": 0.786,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 1.029083515922497,
"learning_rate": 8.78048780487805e-06,
"loss": 0.7578,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 0.9228556457092033,
"learning_rate": 9.10569105691057e-06,
"loss": 0.7555,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 0.8176507599190369,
"learning_rate": 9.43089430894309e-06,
"loss": 0.7616,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 0.8788881402940685,
"learning_rate": 9.756097560975611e-06,
"loss": 0.7609,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 0.8231765047646774,
"learning_rate": 1.008130081300813e-05,
"loss": 0.7581,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.7932701347763691,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.7582,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.7554316371605225,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.74,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.703788704638976,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.725,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.7012045597257317,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.7401,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.6569452157959436,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.7279,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.6607712613461495,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.7359,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.6840120948698405,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.7343,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.5946002774086647,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.7227,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.7146059086048339,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.7292,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.6342333375900306,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.7148,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.7264807684923233,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.7307,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.7212810689711346,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.7144,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.6190565664522967,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.7305,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.7497060536296275,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.7123,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.7231436612355596,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.7203,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.6138428436726511,
"learning_rate": 1.528455284552846e-05,
"loss": 0.6967,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.7711872963268629,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.6974,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.8757353527244967,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.707,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.6952551177963855,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.7062,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.8418505731145355,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.6963,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.7055964478301198,
"learning_rate": 1.691056910569106e-05,
"loss": 0.7087,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.6952386480346046,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.7037,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.8046317927675747,
"learning_rate": 1.75609756097561e-05,
"loss": 0.6945,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.7624824513619726,
"learning_rate": 1.788617886178862e-05,
"loss": 0.7033,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.6657829823213295,
"learning_rate": 1.821138211382114e-05,
"loss": 0.6928,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.716033873483883,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.6973,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.8443118344253713,
"learning_rate": 1.886178861788618e-05,
"loss": 0.6912,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.6756618308362743,
"learning_rate": 1.91869918699187e-05,
"loss": 0.697,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.7941992113823968,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.6805,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 1.1219390376045684,
"learning_rate": 1.983739837398374e-05,
"loss": 0.6893,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.6906577771342892,
"learning_rate": 2.016260162601626e-05,
"loss": 0.6785,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 1.0909615433309063,
"learning_rate": 2.048780487804878e-05,
"loss": 0.6961,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.8077887014494217,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.6938,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.731205908518557,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.691,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.7906785382946626,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.6751,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.8652076046732301,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.693,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 1.2409051524790382,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.6736,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 0.904249580907229,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.6666,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 1.136108207806525,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.7032,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 1.0901826233263694,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.6746,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 1.0805931715695567,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.6727,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 1.1811007798120938,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.6905,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 1.2423267623652317,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.6911,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.9343819701680494,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.6777,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 1.9475363898873441,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.6861,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 1.3773232651795633,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.6711,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 2.35994399458637,
"learning_rate": 2.536585365853659e-05,
"loss": 0.6762,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 2.3706251622704415,
"learning_rate": 2.569105691056911e-05,
"loss": 0.6815,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 1.269171051370134,
"learning_rate": 2.601626016260163e-05,
"loss": 0.6865,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 1.5133351391479744,
"learning_rate": 2.634146341463415e-05,
"loss": 0.6792,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 1.0403164085528793,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.6818,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 1.49222654221066,
"learning_rate": 2.699186991869919e-05,
"loss": 0.6827,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 0.9586930900005125,
"learning_rate": 2.731707317073171e-05,
"loss": 0.6697,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 1.2634115345843018,
"learning_rate": 2.764227642276423e-05,
"loss": 0.6846,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 0.8319049914462255,
"learning_rate": 2.796747967479675e-05,
"loss": 0.6832,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 1.4497568579451219,
"learning_rate": 2.829268292682927e-05,
"loss": 0.671,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 0.8108661842993733,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.6858,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 1.3138126740453366,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.6717,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 28.19901065148241,
"learning_rate": 2.926829268292683e-05,
"loss": 0.6736,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 1.8826888155296035,
"learning_rate": 2.959349593495935e-05,
"loss": 0.6823,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 1.2545676818025249,
"learning_rate": 2.991869918699187e-05,
"loss": 0.6726,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 1.2873831960452833,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.6782,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 1.6661304478750367,
"learning_rate": 3.056910569105692e-05,
"loss": 0.6734,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 1.359182708291026,
"learning_rate": 3.089430894308943e-05,
"loss": 0.6789,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 1.422837266819931,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.6815,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 1.8788708638449725,
"learning_rate": 3.154471544715447e-05,
"loss": 0.6715,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 1.3410335076860542,
"learning_rate": 3.186991869918699e-05,
"loss": 0.6499,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 1.9805585242842385,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.7031,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 1.8881561268873397,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.6697,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 1.5616900645014278,
"learning_rate": 3.284552845528456e-05,
"loss": 0.6741,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 1.5154078893557787,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.6834,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 1.7692791380204338,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.6945,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 1.212865554571536,
"learning_rate": 3.382113821138212e-05,
"loss": 0.6733,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 1.8613109866006743,
"learning_rate": 3.414634146341463e-05,
"loss": 0.6767,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 1.3608254739217487,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.6592,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 2.051510873632318,
"learning_rate": 3.479674796747968e-05,
"loss": 0.6853,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 1.902682771250894,
"learning_rate": 3.51219512195122e-05,
"loss": 0.6575,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 1.5749687217906596,
"learning_rate": 3.544715447154472e-05,
"loss": 0.6625,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 1.5760614111844107,
"learning_rate": 3.577235772357724e-05,
"loss": 0.6561,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 1.4580884862069405,
"learning_rate": 3.609756097560976e-05,
"loss": 0.6535,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 1.1886304886376549,
"learning_rate": 3.642276422764228e-05,
"loss": 0.6601,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 1.6371472489271286,
"learning_rate": 3.67479674796748e-05,
"loss": 0.6621,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 1.1803908507924363,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.6496,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 1.934873367327913,
"learning_rate": 3.739837398373984e-05,
"loss": 0.6547,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 1.590591782342671,
"learning_rate": 3.772357723577236e-05,
"loss": 0.6645,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 1.6637832687548741,
"learning_rate": 3.804878048780488e-05,
"loss": 0.6743,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 1.4524789117850756,
"learning_rate": 3.83739837398374e-05,
"loss": 0.664,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 1.7573391851951885,
"learning_rate": 3.869918699186992e-05,
"loss": 0.6475,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 1.3635796142915761,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.6496,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 1.3188508809693924,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.6408,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 1.3314659218270848,
"learning_rate": 3.967479674796748e-05,
"loss": 0.662,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 1.6350467067714645,
"learning_rate": 4e-05,
"loss": 0.6631,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 1.225234999092673,
"learning_rate": 3.999991946137476e-05,
"loss": 0.6544,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 1.7130116183372375,
"learning_rate": 3.999967784614766e-05,
"loss": 0.6717,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 1.3764781678216154,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.6581,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 1.2909818045148211,
"learning_rate": 3.999871139496895e-05,
"loss": 0.6589,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 1.381283970682795,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.6832,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 1.0905078114001976,
"learning_rate": 3.999710067759846e-05,
"loss": 0.6719,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 1.4547970382310564,
"learning_rate": 3.999605373449617e-05,
"loss": 0.6663,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 1.2283457904027586,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.674,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 0.873159893082871,
"learning_rate": 3.999347672161713e-05,
"loss": 0.6545,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 1.1240563519617293,
"learning_rate": 3.999194667259528e-05,
"loss": 0.6604,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 0.8731730517517008,
"learning_rate": 3.999025561118334e-05,
"loss": 0.6645,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 1.299660658749259,
"learning_rate": 3.998840355100086e-05,
"loss": 0.6627,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 0.9630728135987542,
"learning_rate": 3.998639050696409e-05,
"loss": 0.6639,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 1.3282004706507236,
"learning_rate": 3.998421649528582e-05,
"loss": 0.6666,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 1.0971215240855856,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.6689,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 1.2347340204599615,
"learning_rate": 3.997938564033779e-05,
"loss": 0.6575,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 1.3587369399456906,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.6602,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.944144868342619,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.6662,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 1.2867706924221498,
"learning_rate": 3.997093258045973e-05,
"loss": 0.6772,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 1.1858434865679053,
"learning_rate": 3.996779317598936e-05,
"loss": 0.6597,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 1.18694119211881,
"learning_rate": 3.996449295365782e-05,
"loss": 0.6662,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 1.015215756074195,
"learning_rate": 3.996103194004467e-05,
"loss": 0.666,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 1.2160600279287574,
"learning_rate": 3.995741016302441e-05,
"loss": 0.6583,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 1.0883431931119631,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.6768,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 1.7481726825130257,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.6528,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 1.3055458340463495,
"learning_rate": 3.994558054968643e-05,
"loss": 0.672,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 1.80660311918449,
"learning_rate": 3.994131602367481e-05,
"loss": 0.6542,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 1.2607606765124624,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.6537,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 1.2396534877956287,
"learning_rate": 3.993230519343752e-05,
"loss": 0.6768,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 1.5097319395045181,
"learning_rate": 3.992755896178383e-05,
"loss": 0.6654,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.9643849056350347,
"learning_rate": 3.992265223630981e-05,
"loss": 0.682,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 1.5870920370391375,
"learning_rate": 3.991758505653355e-05,
"loss": 0.652,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 1.0133399349604217,
"learning_rate": 3.991235746326543e-05,
"loss": 0.6722,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 1.8056803552492318,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.6682,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 1.3418578968989339,
"learning_rate": 3.990142120595444e-05,
"loss": 0.6549,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 1.7447074952189632,
"learning_rate": 3.98957126299907e-05,
"loss": 0.657,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 1.713826825942418,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.6749,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 1.6613880981416476,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.6642,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 1.5221071137203006,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.6577,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 1.5251116323592313,
"learning_rate": 3.98712764319087e-05,
"loss": 0.6614,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 1.6004369392102469,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.6581,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 1.0903412620817632,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.6636,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 1.220414445591047,
"learning_rate": 3.985126869010129e-05,
"loss": 0.6398,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 1.2559499825698726,
"learning_rate": 3.984427961114833e-05,
"loss": 0.6596,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 1.3592322387234117,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.6457,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 1.122469553781278,
"learning_rate": 3.982982204151901e-05,
"loss": 0.6469,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 1.7494671435574505,
"learning_rate": 3.982235366728193e-05,
"loss": 0.6506,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 1.117689919142851,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.6551,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 2.2764605694448066,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.656,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 2.16871073608932,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.6716,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 1.1875477614736685,
"learning_rate": 3.979088432599627e-05,
"loss": 0.6579,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 1.3797648478235371,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.6496,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 1.3288645021360268,
"learning_rate": 3.977419304032111e-05,
"loss": 0.6667,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 1.0290275716244577,
"learning_rate": 3.976560847560697e-05,
"loss": 0.6522,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 0.9178655871903216,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.6665,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 1.2781342541806548,
"learning_rate": 3.974796184811956e-05,
"loss": 0.6603,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.8850706385329972,
"learning_rate": 3.973889992746979e-05,
"loss": 0.6618,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.9887638513143595,
"learning_rate": 3.972967903243361e-05,
"loss": 0.6576,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 1.020439107687966,
"learning_rate": 3.972029923727486e-05,
"loss": 0.6775,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 1.085461252240094,
"learning_rate": 3.971076061753709e-05,
"loss": 0.6664,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.8072090331975253,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.6475,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.9693418075604752,
"learning_rate": 3.969120721289402e-05,
"loss": 0.6639,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 1.0770195275365195,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.642,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.8527918887692488,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.6499,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.7281399588524676,
"learning_rate": 3.966068788369422e-05,
"loss": 0.6453,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.8641307560603346,
"learning_rate": 3.965019797448622e-05,
"loss": 0.6551,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.8765565076761768,
"learning_rate": 3.963954980528515e-05,
"loss": 0.663,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.9234489341149446,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.6724,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.7159496484093195,
"learning_rate": 3.961777903121329e-05,
"loss": 0.661,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.9008152238479821,
"learning_rate": 3.960665660168131e-05,
"loss": 0.6623,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.8669739285332979,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.6772,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 1.2434727253939497,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.6489,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 1.0701037116680991,
"learning_rate": 3.957234222185657e-05,
"loss": 0.6712,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.9022608440929049,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.6621,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 1.0792772226359406,
"learning_rate": 3.954867765033605e-05,
"loss": 0.6432,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 1.0540315768393111,
"learning_rate": 3.953660915306728e-05,
"loss": 0.6577,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.8424606934896346,
"learning_rate": 3.952438331063419e-05,
"loss": 0.6666,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.9947991822597244,
"learning_rate": 3.951200022150205e-05,
"loss": 0.6645,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.8413393639014164,
"learning_rate": 3.949945998540253e-05,
"loss": 0.6569,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.7363481152354845,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.644,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 1.0324832819227012,
"learning_rate": 3.947390847755559e-05,
"loss": 0.6608,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.807038214568879,
"learning_rate": 3.946089741159648e-05,
"loss": 0.6499,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.7693033441814051,
"learning_rate": 3.944772961024501e-05,
"loss": 0.6661,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.7971669924836549,
"learning_rate": 3.943440517955285e-05,
"loss": 0.6538,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.7082464743924001,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.6399,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.6784500294100894,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.6333,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.8611115994259436,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.6652,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.8833486968838228,
"learning_rate": 3.937954332854371e-05,
"loss": 0.6523,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.9004706862535784,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.6344,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 1.2338663913392385,
"learning_rate": 3.935117547697387e-05,
"loss": 0.6471,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.7528653118851252,
"learning_rate": 3.933675771619675e-05,
"loss": 0.6478,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 1.3438899124936547,
"learning_rate": 3.932218421983131e-05,
"loss": 0.6531,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.8534205831552383,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.6322,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.9053390077788166,
"learning_rate": 3.929257049108054e-05,
"loss": 0.6409,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 1.1088040781097641,
"learning_rate": 3.927753049720011e-05,
"loss": 0.6503,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 1.2754237935299644,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.6474,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.8307950396284126,
"learning_rate": 3.92469848560784e-05,
"loss": 0.6421,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 1.5994331322111197,
"learning_rate": 3.923147945484751e-05,
"loss": 0.6454,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.8458287351757913,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.6596,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 1.463421169431786,
"learning_rate": 3.920000411543654e-05,
"loss": 0.6523,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.9028235527164244,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.6536,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 1.4850028057982634,
"learning_rate": 3.916791024049648e-05,
"loss": 0.6519,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 1.009366198567409,
"learning_rate": 3.91516316745245e-05,
"loss": 0.648,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 1.5151621914489402,
"learning_rate": 3.913519886394389e-05,
"loss": 0.6544,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.9767457448758948,
"learning_rate": 3.911861194110225e-05,
"loss": 0.6408,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 1.6308820884417192,
"learning_rate": 3.910187103958837e-05,
"loss": 0.644,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 1.371545680011608,
"learning_rate": 3.908497629423117e-05,
"loss": 0.6538,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 1.6673869701501398,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.6612,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 1.414756961400176,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.6392,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 1.2355353346447873,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.6469,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 1.33434938676518,
"learning_rate": 3.901586161429081e-05,
"loss": 0.6377,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.9492151531665349,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.6466,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 1.4152205939620925,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.6363,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.9939057666936282,
"learning_rate": 3.896241703466667e-05,
"loss": 0.6523,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 1.621840393911687,
"learning_rate": 3.894429654085585e-05,
"loss": 0.6526,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 1.1252635608677066,
"learning_rate": 3.892602347228505e-05,
"loss": 0.6452,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 1.813377917672578,
"learning_rate": 3.890759797612307e-05,
"loss": 0.6543,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 1.472783592112648,
"learning_rate": 3.888902020076632e-05,
"loss": 0.6535,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 1.2605238384719888,
"learning_rate": 3.887029029583764e-05,
"loss": 0.6135,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 1.316129614583567,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.6447,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 1.1800698267467966,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.6446,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.996074522071767,
"learning_rate": 3.881318931821972e-05,
"loss": 0.6442,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 1.2368495926690157,
"learning_rate": 3.879385241571817e-05,
"loss": 0.6313,
"step": 246
},
{
"epoch": 1.000506329113924,
"grad_norm": 1.1341367592721754,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.7025,
"step": 247
},
{
"epoch": 1.0045569620253165,
"grad_norm": 1.241856352963349,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.5962,
"step": 248
},
{
"epoch": 1.0086075949367088,
"grad_norm": 1.0204796856116367,
"learning_rate": 3.873493415863256e-05,
"loss": 0.5633,
"step": 249
},
{
"epoch": 1.0126582278481013,
"grad_norm": 1.12716740393278,
"learning_rate": 3.871499275032111e-05,
"loss": 0.5676,
"step": 250
},
{
"epoch": 1.0167088607594936,
"grad_norm": 0.7040796189310808,
"learning_rate": 3.869490061403091e-05,
"loss": 0.5405,
"step": 251
},
{
"epoch": 1.0207594936708861,
"grad_norm": 1.1674455203947274,
"learning_rate": 3.867465791158124e-05,
"loss": 0.5741,
"step": 252
},
{
"epoch": 1.0248101265822784,
"grad_norm": 0.831513810903045,
"learning_rate": 3.865426480600407e-05,
"loss": 0.539,
"step": 253
},
{
"epoch": 1.028860759493671,
"grad_norm": 1.1517376195068278,
"learning_rate": 3.863372146154264e-05,
"loss": 0.5967,
"step": 254
},
{
"epoch": 1.0329113924050632,
"grad_norm": 1.0620682738941556,
"learning_rate": 3.861302804365024e-05,
"loss": 0.5491,
"step": 255
},
{
"epoch": 1.0369620253164558,
"grad_norm": 0.9158660936293025,
"learning_rate": 3.85921847189888e-05,
"loss": 0.5585,
"step": 256
},
{
"epoch": 1.041012658227848,
"grad_norm": 0.7733834796810339,
"learning_rate": 3.85711916554276e-05,
"loss": 0.5836,
"step": 257
},
{
"epoch": 1.0450632911392406,
"grad_norm": 0.7694875431925313,
"learning_rate": 3.85500490220419e-05,
"loss": 0.5369,
"step": 258
},
{
"epoch": 1.0491139240506329,
"grad_norm": 0.8391341654527159,
"learning_rate": 3.852875698911154e-05,
"loss": 0.6098,
"step": 259
},
{
"epoch": 1.0531645569620254,
"grad_norm": 0.7638516088476978,
"learning_rate": 3.850731572811963e-05,
"loss": 0.5202,
"step": 260
},
{
"epoch": 1.0572151898734177,
"grad_norm": 0.7432034790279505,
"learning_rate": 3.848572541175116e-05,
"loss": 0.5808,
"step": 261
},
{
"epoch": 1.0612658227848102,
"grad_norm": 0.8724328264911398,
"learning_rate": 3.846398621389154e-05,
"loss": 0.5831,
"step": 262
},
{
"epoch": 1.0653164556962025,
"grad_norm": 0.7834651441391898,
"learning_rate": 3.84420983096253e-05,
"loss": 0.5435,
"step": 263
},
{
"epoch": 1.069367088607595,
"grad_norm": 0.8902100894377551,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.5744,
"step": 264
},
{
"epoch": 1.0734177215189873,
"grad_norm": 0.7608392682205541,
"learning_rate": 3.839787708819787e-05,
"loss": 0.5432,
"step": 265
},
{
"epoch": 1.0774683544303798,
"grad_norm": 0.7600758357449334,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.5768,
"step": 266
},
{
"epoch": 1.081518987341772,
"grad_norm": 0.8626030143839272,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.5708,
"step": 267
},
{
"epoch": 1.0855696202531646,
"grad_norm": 0.8331833110007257,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.5585,
"step": 268
},
{
"epoch": 1.089620253164557,
"grad_norm": 0.7485939969042157,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.5468,
"step": 269
},
{
"epoch": 1.0936708860759494,
"grad_norm": 0.7493631973987808,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.5743,
"step": 270
},
{
"epoch": 1.0977215189873417,
"grad_norm": 0.7637329447389053,
"learning_rate": 3.826166304956594e-05,
"loss": 0.5617,
"step": 271
},
{
"epoch": 1.1017721518987342,
"grad_norm": 0.6562565563159428,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.5314,
"step": 272
},
{
"epoch": 1.1058227848101265,
"grad_norm": 0.6369546334606763,
"learning_rate": 3.821507978767479e-05,
"loss": 0.5661,
"step": 273
},
{
"epoch": 1.109873417721519,
"grad_norm": 0.8219976300619848,
"learning_rate": 3.819156801001693e-05,
"loss": 0.5475,
"step": 274
},
{
"epoch": 1.1139240506329113,
"grad_norm": 0.724172424683593,
"learning_rate": 3.816790971997121e-05,
"loss": 0.5418,
"step": 275
},
{
"epoch": 1.1179746835443038,
"grad_norm": 0.8252557232202022,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.585,
"step": 276
},
{
"epoch": 1.1220253164556961,
"grad_norm": 0.6185775569273213,
"learning_rate": 3.81201543660571e-05,
"loss": 0.5071,
"step": 277
},
{
"epoch": 1.1260759493670887,
"grad_norm": 0.8414342922375957,
"learning_rate": 3.809605768680377e-05,
"loss": 0.571,
"step": 278
},
{
"epoch": 1.130126582278481,
"grad_norm": 0.7433954173400105,
"learning_rate": 3.807181526438958e-05,
"loss": 0.5843,
"step": 279
},
{
"epoch": 1.1341772151898735,
"grad_norm": 0.7576576549607834,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.5531,
"step": 280
},
{
"epoch": 1.1382278481012658,
"grad_norm": 1.0123627900862284,
"learning_rate": 3.802289397223145e-05,
"loss": 0.5306,
"step": 281
},
{
"epoch": 1.1422784810126583,
"grad_norm": 0.9136158847392416,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.5373,
"step": 282
},
{
"epoch": 1.1463291139240506,
"grad_norm": 0.6520697931160241,
"learning_rate": 3.797339206560096e-05,
"loss": 0.5816,
"step": 283
},
{
"epoch": 1.150379746835443,
"grad_norm": 0.6417762731448688,
"learning_rate": 3.794842387948027e-05,
"loss": 0.5509,
"step": 284
},
{
"epoch": 1.1544303797468354,
"grad_norm": 0.6609545966473199,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.5557,
"step": 285
},
{
"epoch": 1.158481012658228,
"grad_norm": 0.7356771238985378,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.5456,
"step": 286
},
{
"epoch": 1.1625316455696202,
"grad_norm": 0.6660101129518375,
"learning_rate": 3.787265280646825e-05,
"loss": 0.5817,
"step": 287
},
{
"epoch": 1.1665822784810127,
"grad_norm": 0.7346567064259215,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.5644,
"step": 288
},
{
"epoch": 1.170632911392405,
"grad_norm": 0.7019281380312725,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.563,
"step": 289
},
{
"epoch": 1.1746835443037975,
"grad_norm": 0.6127958300294609,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.5387,
"step": 290
},
{
"epoch": 1.1787341772151898,
"grad_norm": 0.6486876246748341,
"learning_rate": 3.776961046830306e-05,
"loss": 0.5516,
"step": 291
},
{
"epoch": 1.1827848101265823,
"grad_norm": 0.5946593340811529,
"learning_rate": 3.774349157719215e-05,
"loss": 0.5351,
"step": 292
},
{
"epoch": 1.1868354430379746,
"grad_norm": 0.7735724100776147,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.5898,
"step": 293
},
{
"epoch": 1.1908860759493671,
"grad_norm": 0.6136854664500483,
"learning_rate": 3.769082529555359e-05,
"loss": 0.5335,
"step": 294
},
{
"epoch": 1.1949367088607594,
"grad_norm": 0.7644865131658087,
"learning_rate": 3.766427832919294e-05,
"loss": 0.5909,
"step": 295
},
{
"epoch": 1.198987341772152,
"grad_norm": 0.6364199908700746,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.5301,
"step": 296
},
{
"epoch": 1.2030379746835442,
"grad_norm": 0.7500565343404131,
"learning_rate": 3.761075781441526e-05,
"loss": 0.5744,
"step": 297
},
{
"epoch": 1.2070886075949367,
"grad_norm": 0.691459993276526,
"learning_rate": 3.75837846970451e-05,
"loss": 0.5483,
"step": 298
},
{
"epoch": 1.211139240506329,
"grad_norm": 0.7476793904783704,
"learning_rate": 3.755666996229032e-05,
"loss": 0.6038,
"step": 299
},
{
"epoch": 1.2151898734177216,
"grad_norm": 0.6973882116925859,
"learning_rate": 3.752941382852927e-05,
"loss": 0.5548,
"step": 300
},
{
"epoch": 1.2192405063291138,
"grad_norm": 1.1086921562373009,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.5615,
"step": 301
},
{
"epoch": 1.2232911392405064,
"grad_norm": 0.8349235796602333,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.574,
"step": 302
},
{
"epoch": 1.2273417721518987,
"grad_norm": 1.059628009039191,
"learning_rate": 3.744679923406351e-05,
"loss": 0.5696,
"step": 303
},
{
"epoch": 1.2313924050632912,
"grad_norm": 1.3793332991328506,
"learning_rate": 3.741897971081043e-05,
"loss": 0.5518,
"step": 304
},
{
"epoch": 1.2354430379746835,
"grad_norm": 0.8153049100035671,
"learning_rate": 3.739101989748946e-05,
"loss": 0.5493,
"step": 305
},
{
"epoch": 1.239493670886076,
"grad_norm": 1.642175201106135,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.6059,
"step": 306
},
{
"epoch": 1.2435443037974683,
"grad_norm": 0.9976469533447115,
"learning_rate": 3.73346803025098e-05,
"loss": 0.552,
"step": 307
},
{
"epoch": 1.2475949367088608,
"grad_norm": 1.6772207730929416,
"learning_rate": 3.730630097460247e-05,
"loss": 0.5623,
"step": 308
},
{
"epoch": 1.251645569620253,
"grad_norm": 1.2647885344226129,
"learning_rate": 3.727778226412628e-05,
"loss": 0.5789,
"step": 309
},
{
"epoch": 1.2556962025316456,
"grad_norm": 1.5834160926349858,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.5595,
"step": 310
},
{
"epoch": 1.259746835443038,
"grad_norm": 1.3245099536453153,
"learning_rate": 3.722032761533114e-05,
"loss": 0.5557,
"step": 311
},
{
"epoch": 1.2637974683544304,
"grad_norm": 1.357640295851919,
"learning_rate": 3.719139213974403e-05,
"loss": 0.56,
"step": 312
},
{
"epoch": 1.2678481012658227,
"grad_norm": 1.3091269969390815,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.5856,
"step": 313
},
{
"epoch": 1.2718987341772152,
"grad_norm": 1.3546662451294271,
"learning_rate": 3.713310605140055e-05,
"loss": 0.5313,
"step": 314
},
{
"epoch": 1.2759493670886077,
"grad_norm": 1.198129789316893,
"learning_rate": 3.710375590807233e-05,
"loss": 0.5617,
"step": 315
},
{
"epoch": 1.28,
"grad_norm": 1.400022426714234,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.5432,
"step": 316
},
{
"epoch": 1.2840506329113923,
"grad_norm": 1.290245664357711,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.5753,
"step": 317
},
{
"epoch": 1.2881012658227848,
"grad_norm": 1.52101035026535,
"learning_rate": 3.701487992136854e-05,
"loss": 0.5474,
"step": 318
},
{
"epoch": 1.2921518987341774,
"grad_norm": 1.5758313905735812,
"learning_rate": 3.69849802022222e-05,
"loss": 0.5902,
"step": 319
},
{
"epoch": 1.2962025316455696,
"grad_norm": 1.12169841135027,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.5516,
"step": 320
},
{
"epoch": 1.300253164556962,
"grad_norm": 1.357317999830611,
"learning_rate": 3.692477062175289e-05,
"loss": 0.5614,
"step": 321
},
{
"epoch": 1.3043037974683545,
"grad_norm": 0.971236012846228,
"learning_rate": 3.689446124534958e-05,
"loss": 0.5779,
"step": 322
},
{
"epoch": 1.308354430379747,
"grad_norm": 0.9966134944221307,
"learning_rate": 3.686401580327799e-05,
"loss": 0.5198,
"step": 323
},
{
"epoch": 1.3124050632911393,
"grad_norm": 0.8023107756832777,
"learning_rate": 3.683343454074149e-05,
"loss": 0.5693,
"step": 324
},
{
"epoch": 1.3164556962025316,
"grad_norm": 0.8942370556740934,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.5252,
"step": 325
},
{
"epoch": 1.320506329113924,
"grad_norm": 0.6742033823963365,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.5849,
"step": 326
},
{
"epoch": 1.3245569620253166,
"grad_norm": 0.9219001560591615,
"learning_rate": 3.674087829877297e-05,
"loss": 0.5281,
"step": 327
},
{
"epoch": 1.3286075949367089,
"grad_norm": 0.5821509994427708,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.5459,
"step": 328
},
{
"epoch": 1.3326582278481012,
"grad_norm": 1.048722184018075,
"learning_rate": 3.667849957966501e-05,
"loss": 0.5585,
"step": 329
},
{
"epoch": 1.3367088607594937,
"grad_norm": 0.6703959923967664,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.544,
"step": 330
},
{
"epoch": 1.3407594936708862,
"grad_norm": 1.047083910248625,
"learning_rate": 3.661558355626795e-05,
"loss": 0.5504,
"step": 331
},
{
"epoch": 1.3448101265822785,
"grad_norm": 0.7526474594919527,
"learning_rate": 3.658392468818163e-05,
"loss": 0.5631,
"step": 332
},
{
"epoch": 1.3488607594936708,
"grad_norm": 0.984302231720434,
"learning_rate": 3.655213225544574e-05,
"loss": 0.5551,
"step": 333
},
{
"epoch": 1.3529113924050633,
"grad_norm": 0.8839908721150211,
"learning_rate": 3.652020651411218e-05,
"loss": 0.5577,
"step": 334
},
{
"epoch": 1.3569620253164558,
"grad_norm": 0.5813241718314217,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.5353,
"step": 335
},
{
"epoch": 1.3610126582278481,
"grad_norm": 0.7739867701917744,
"learning_rate": 3.645595613522574e-05,
"loss": 0.5896,
"step": 336
},
{
"epoch": 1.3650632911392404,
"grad_norm": 0.6600280804930256,
"learning_rate": 3.642363201513657e-05,
"loss": 0.5212,
"step": 337
},
{
"epoch": 1.369113924050633,
"grad_norm": 0.7308829386910015,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.5637,
"step": 338
},
{
"epoch": 1.3731645569620254,
"grad_norm": 0.8223304950218445,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.5395,
"step": 339
},
{
"epoch": 1.3772151898734177,
"grad_norm": 0.7188954677361126,
"learning_rate": 3.632586705948318e-05,
"loss": 0.5727,
"step": 340
},
{
"epoch": 1.38126582278481,
"grad_norm": 0.7716658923129035,
"learning_rate": 3.629301541734311e-05,
"loss": 0.5242,
"step": 341
},
{
"epoch": 1.3853164556962025,
"grad_norm": 0.6668242060037731,
"learning_rate": 3.626003255349676e-05,
"loss": 0.601,
"step": 342
},
{
"epoch": 1.389367088607595,
"grad_norm": 0.6502275333476393,
"learning_rate": 3.622691873358357e-05,
"loss": 0.567,
"step": 343
},
{
"epoch": 1.3934177215189874,
"grad_norm": 0.6981204557715653,
"learning_rate": 3.61936742242977e-05,
"loss": 0.5404,
"step": 344
},
{
"epoch": 1.3974683544303796,
"grad_norm": 0.556182531018718,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.5405,
"step": 345
},
{
"epoch": 1.4015189873417722,
"grad_norm": 0.6236047572685106,
"learning_rate": 3.612679420964516e-05,
"loss": 0.5732,
"step": 346
},
{
"epoch": 1.4055696202531647,
"grad_norm": 0.6523213316932426,
"learning_rate": 3.609315924292092e-05,
"loss": 0.5305,
"step": 347
},
{
"epoch": 1.409620253164557,
"grad_norm": 0.7155197388667291,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.6212,
"step": 348
},
{
"epoch": 1.4136708860759493,
"grad_norm": 0.7049184476308888,
"learning_rate": 3.602550074513133e-05,
"loss": 0.5353,
"step": 349
},
{
"epoch": 1.4177215189873418,
"grad_norm": 0.6947156351841812,
"learning_rate": 3.599147775897822e-05,
"loss": 0.566,
"step": 350
},
{
"epoch": 1.4217721518987343,
"grad_norm": 0.8618774639954092,
"learning_rate": 3.595732597966167e-05,
"loss": 0.5743,
"step": 351
},
{
"epoch": 1.4258227848101266,
"grad_norm": 0.68346290974404,
"learning_rate": 3.592304568223542e-05,
"loss": 0.5502,
"step": 352
},
{
"epoch": 1.4298734177215189,
"grad_norm": 0.8396661468685334,
"learning_rate": 3.588863714278826e-05,
"loss": 0.5621,
"step": 353
},
{
"epoch": 1.4339240506329114,
"grad_norm": 0.6206325601447893,
"learning_rate": 3.585410063844186e-05,
"loss": 0.5578,
"step": 354
},
{
"epoch": 1.437974683544304,
"grad_norm": 0.7176132455804315,
"learning_rate": 3.581943644734846e-05,
"loss": 0.5698,
"step": 355
},
{
"epoch": 1.4420253164556962,
"grad_norm": 0.6088901006708863,
"learning_rate": 3.578464484868869e-05,
"loss": 0.5375,
"step": 356
},
{
"epoch": 1.4460759493670885,
"grad_norm": 0.7560729211819237,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.5956,
"step": 357
},
{
"epoch": 1.450126582278481,
"grad_norm": 0.6442426325581019,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.5407,
"step": 358
},
{
"epoch": 1.4541772151898735,
"grad_norm": 0.7568792256495813,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.5484,
"step": 359
},
{
"epoch": 1.4582278481012658,
"grad_norm": 0.8195169160910546,
"learning_rate": 3.564420999786543e-05,
"loss": 0.6079,
"step": 360
},
{
"epoch": 1.4622784810126581,
"grad_norm": 0.7078506190779682,
"learning_rate": 3.560878558491842e-05,
"loss": 0.5145,
"step": 361
},
{
"epoch": 1.4663291139240506,
"grad_norm": 0.6616459685906091,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.5968,
"step": 362
},
{
"epoch": 1.4703797468354431,
"grad_norm": 0.6113903313520046,
"learning_rate": 3.553755991230039e-05,
"loss": 0.5342,
"step": 363
},
{
"epoch": 1.4744303797468354,
"grad_norm": 0.6868722261173005,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.5707,
"step": 364
},
{
"epoch": 1.4784810126582277,
"grad_norm": 0.6480513566236376,
"learning_rate": 3.546583369120419e-05,
"loss": 0.5581,
"step": 365
},
{
"epoch": 1.4825316455696202,
"grad_norm": 0.696517466362994,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.576,
"step": 366
},
{
"epoch": 1.4865822784810128,
"grad_norm": 0.7059538737032156,
"learning_rate": 3.539360923231766e-05,
"loss": 0.5504,
"step": 367
},
{
"epoch": 1.490632911392405,
"grad_norm": 0.578751130906813,
"learning_rate": 3.535731089018394e-05,
"loss": 0.5545,
"step": 368
},
{
"epoch": 1.4946835443037974,
"grad_norm": 0.6516899902183392,
"learning_rate": 3.532088886237956e-05,
"loss": 0.5617,
"step": 369
},
{
"epoch": 1.4987341772151899,
"grad_norm": 0.6250681761366237,
"learning_rate": 3.528434344224253e-05,
"loss": 0.5534,
"step": 370
},
{
"epoch": 1.5027848101265824,
"grad_norm": 0.5607891969671928,
"learning_rate": 3.524767492410464e-05,
"loss": 0.5593,
"step": 371
},
{
"epoch": 1.5068354430379747,
"grad_norm": 0.6515421219492084,
"learning_rate": 3.521088360328908e-05,
"loss": 0.5876,
"step": 372
},
{
"epoch": 1.510886075949367,
"grad_norm": 0.6318528657893749,
"learning_rate": 3.517396977610811e-05,
"loss": 0.5512,
"step": 373
},
{
"epoch": 1.5149367088607595,
"grad_norm": 0.6274975899853215,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.5791,
"step": 374
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.8106203557574205,
"learning_rate": 3.509977579282971e-05,
"loss": 0.5679,
"step": 375
},
{
"epoch": 1.5230379746835443,
"grad_norm": 0.8243707606616362,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.5685,
"step": 376
},
{
"epoch": 1.5270886075949366,
"grad_norm": 0.7686781408906029,
"learning_rate": 3.502509536445719e-05,
"loss": 0.5493,
"step": 377
},
{
"epoch": 1.531139240506329,
"grad_norm": 1.3952793309776077,
"learning_rate": 3.498757348458147e-05,
"loss": 0.5831,
"step": 378
},
{
"epoch": 1.5351898734177216,
"grad_norm": 0.8290685354649189,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.5518,
"step": 379
},
{
"epoch": 1.539240506329114,
"grad_norm": 1.1492052901671208,
"learning_rate": 3.491216790442899e-05,
"loss": 0.597,
"step": 380
},
{
"epoch": 1.5432911392405062,
"grad_norm": 0.9542361311877778,
"learning_rate": 3.487428481145839e-05,
"loss": 0.5162,
"step": 381
},
{
"epoch": 1.5473417721518987,
"grad_norm": 0.8385183564166004,
"learning_rate": 3.483628192304278e-05,
"loss": 0.5604,
"step": 382
},
{
"epoch": 1.5513924050632912,
"grad_norm": 1.4277578761020004,
"learning_rate": 3.479815954525219e-05,
"loss": 0.5947,
"step": 383
},
{
"epoch": 1.5554430379746835,
"grad_norm": 0.8180287969849909,
"learning_rate": 3.475991798511899e-05,
"loss": 0.5606,
"step": 384
},
{
"epoch": 1.5594936708860758,
"grad_norm": 1.3537937635327895,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.5796,
"step": 385
},
{
"epoch": 1.5635443037974683,
"grad_norm": 0.8011727471018607,
"learning_rate": 3.468307855075128e-05,
"loss": 0.54,
"step": 386
},
{
"epoch": 1.5675949367088609,
"grad_norm": 0.9225500374433782,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.5674,
"step": 387
},
{
"epoch": 1.5716455696202531,
"grad_norm": 0.9805877010769618,
"learning_rate": 3.460576609535163e-05,
"loss": 0.5428,
"step": 388
},
{
"epoch": 1.5756962025316454,
"grad_norm": 0.8042507212213127,
"learning_rate": 3.456693326250006e-05,
"loss": 0.5438,
"step": 389
},
{
"epoch": 1.579746835443038,
"grad_norm": 0.7728186805443779,
"learning_rate": 3.452798310957058e-05,
"loss": 0.5885,
"step": 390
},
{
"epoch": 1.5837974683544305,
"grad_norm": 1.0382813143336926,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.5623,
"step": 391
},
{
"epoch": 1.5878481012658228,
"grad_norm": 0.638734079166629,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.581,
"step": 392
},
{
"epoch": 1.591898734177215,
"grad_norm": 0.7678569212031139,
"learning_rate": 3.441043187201574e-05,
"loss": 0.5589,
"step": 393
},
{
"epoch": 1.5959493670886076,
"grad_norm": 0.7161298256588402,
"learning_rate": 3.437101558517728e-05,
"loss": 0.5488,
"step": 394
},
{
"epoch": 1.6,
"grad_norm": 0.7000766586666707,
"learning_rate": 3.433148355615496e-05,
"loss": 0.5817,
"step": 395
},
{
"epoch": 1.6040506329113924,
"grad_norm": 0.699724612145923,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.5513,
"step": 396
},
{
"epoch": 1.6081012658227847,
"grad_norm": 0.651219759774385,
"learning_rate": 3.425207354603043e-05,
"loss": 0.558,
"step": 397
},
{
"epoch": 1.6121518987341772,
"grad_norm": 0.6690344552325893,
"learning_rate": 3.421219620448553e-05,
"loss": 0.5775,
"step": 398
},
{
"epoch": 1.6162025316455697,
"grad_norm": 0.7279401186293292,
"learning_rate": 3.417220439986623e-05,
"loss": 0.5392,
"step": 399
},
{
"epoch": 1.620253164556962,
"grad_norm": 0.6844449766700464,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.5581,
"step": 400
},
{
"epoch": 1.6243037974683543,
"grad_norm": 0.651420038695758,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.5648,
"step": 401
},
{
"epoch": 1.6283544303797468,
"grad_norm": 0.7539450482407491,
"learning_rate": 3.405154543304065e-05,
"loss": 0.5683,
"step": 402
},
{
"epoch": 1.6324050632911393,
"grad_norm": 0.5860230830052963,
"learning_rate": 3.401109900618843e-05,
"loss": 0.5412,
"step": 403
},
{
"epoch": 1.6364556962025316,
"grad_norm": 0.7736796699936704,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.6176,
"step": 404
},
{
"epoch": 1.640506329113924,
"grad_norm": 0.661685411424609,
"learning_rate": 3.392986794874714e-05,
"loss": 0.5324,
"step": 405
},
{
"epoch": 1.6445569620253164,
"grad_norm": 0.6661809430495648,
"learning_rate": 3.388908397238184e-05,
"loss": 0.5478,
"step": 406
},
{
"epoch": 1.648607594936709,
"grad_norm": 0.7117943093853628,
"learning_rate": 3.384818813524362e-05,
"loss": 0.5702,
"step": 407
},
{
"epoch": 1.6526582278481012,
"grad_norm": 0.5986573504064702,
"learning_rate": 3.380718076670195e-05,
"loss": 0.5673,
"step": 408
},
{
"epoch": 1.6567088607594935,
"grad_norm": 0.6683645685293332,
"learning_rate": 3.376606219702454e-05,
"loss": 0.5578,
"step": 409
},
{
"epoch": 1.660759493670886,
"grad_norm": 0.6563628358033483,
"learning_rate": 3.372483275737468e-05,
"loss": 0.5857,
"step": 410
},
{
"epoch": 1.6648101265822786,
"grad_norm": 0.580316442489648,
"learning_rate": 3.368349277980861e-05,
"loss": 0.5474,
"step": 411
},
{
"epoch": 1.6688607594936709,
"grad_norm": 0.597380790165879,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.548,
"step": 412
},
{
"epoch": 1.6729113924050631,
"grad_norm": 0.7525917184070671,
"learning_rate": 3.360048254360144e-05,
"loss": 0.5551,
"step": 413
},
{
"epoch": 1.6769620253164557,
"grad_norm": 0.7035454942935099,
"learning_rate": 3.355881295351336e-05,
"loss": 0.5667,
"step": 414
},
{
"epoch": 1.6810126582278482,
"grad_norm": 0.6339387166076031,
"learning_rate": 3.351703416260975e-05,
"loss": 0.5684,
"step": 415
},
{
"epoch": 1.6850632911392405,
"grad_norm": 0.7732048039807171,
"learning_rate": 3.347514650737126e-05,
"loss": 0.5453,
"step": 416
},
{
"epoch": 1.6891139240506328,
"grad_norm": 0.5815922747991128,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.5513,
"step": 417
},
{
"epoch": 1.6931645569620253,
"grad_norm": 0.8099523845470166,
"learning_rate": 3.339104595419334e-05,
"loss": 0.548,
"step": 418
},
{
"epoch": 1.6972151898734178,
"grad_norm": 0.7822048656514108,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.579,
"step": 419
},
{
"epoch": 1.70126582278481,
"grad_norm": 0.8121707112199703,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.5474,
"step": 420
},
{
"epoch": 1.7053164556962024,
"grad_norm": 0.7511401794701777,
"learning_rate": 3.326408710419996e-05,
"loss": 0.5703,
"step": 421
},
{
"epoch": 1.709367088607595,
"grad_norm": 0.730447551303335,
"learning_rate": 3.322155337795454e-05,
"loss": 0.5912,
"step": 422
},
{
"epoch": 1.7134177215189874,
"grad_norm": 0.7886804691907633,
"learning_rate": 3.317891316713587e-05,
"loss": 0.5435,
"step": 423
},
{
"epoch": 1.7174683544303797,
"grad_norm": 0.5477917642013045,
"learning_rate": 3.313616681516231e-05,
"loss": 0.5475,
"step": 424
},
{
"epoch": 1.721518987341772,
"grad_norm": 0.7035983690357113,
"learning_rate": 3.309331466630713e-05,
"loss": 0.5696,
"step": 425
},
{
"epoch": 1.7255696202531645,
"grad_norm": 0.6594560993518876,
"learning_rate": 3.305035706569563e-05,
"loss": 0.5755,
"step": 426
},
{
"epoch": 1.729620253164557,
"grad_norm": 0.6286503620962015,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.5546,
"step": 427
},
{
"epoch": 1.7336708860759493,
"grad_norm": 0.8410831351836002,
"learning_rate": 3.296412689394864e-05,
"loss": 0.5939,
"step": 428
},
{
"epoch": 1.7377215189873416,
"grad_norm": 0.6549887415566692,
"learning_rate": 3.292085501729909e-05,
"loss": 0.5448,
"step": 429
},
{
"epoch": 1.7417721518987341,
"grad_norm": 0.6066231494864504,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.5517,
"step": 430
},
{
"epoch": 1.7458227848101266,
"grad_norm": 0.6472740064832411,
"learning_rate": 3.283399942497381e-05,
"loss": 0.5576,
"step": 431
},
{
"epoch": 1.749873417721519,
"grad_norm": 0.6005940267890598,
"learning_rate": 3.279041640882108e-05,
"loss": 0.5487,
"step": 432
},
{
"epoch": 1.7539240506329112,
"grad_norm": 0.5937116410711792,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.5486,
"step": 433
},
{
"epoch": 1.7579746835443038,
"grad_norm": 0.5580024132729267,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.5779,
"step": 434
},
{
"epoch": 1.7620253164556963,
"grad_norm": 0.5024439753055502,
"learning_rate": 3.265905069502244e-05,
"loss": 0.5545,
"step": 435
},
{
"epoch": 1.7660759493670886,
"grad_norm": 0.5348206985014405,
"learning_rate": 3.261505774420016e-05,
"loss": 0.5654,
"step": 436
},
{
"epoch": 1.7701265822784809,
"grad_norm": 0.5611419233490575,
"learning_rate": 3.257096319343707e-05,
"loss": 0.5568,
"step": 437
},
{
"epoch": 1.7741772151898734,
"grad_norm": 0.5107048025350289,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.5735,
"step": 438
},
{
"epoch": 1.7782278481012659,
"grad_norm": 0.5543315313637863,
"learning_rate": 3.248247071342966e-05,
"loss": 0.5376,
"step": 439
},
{
"epoch": 1.7822784810126582,
"grad_norm": 0.6359235876140473,
"learning_rate": 3.243807349689161e-05,
"loss": 0.6047,
"step": 440
},
{
"epoch": 1.7863291139240505,
"grad_norm": 0.5969300182457072,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.5166,
"step": 441
},
{
"epoch": 1.790379746835443,
"grad_norm": 0.662333216083229,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.5852,
"step": 442
},
{
"epoch": 1.7944303797468355,
"grad_norm": 0.7459326588655024,
"learning_rate": 3.230428223438075e-05,
"loss": 0.5714,
"step": 443
},
{
"epoch": 1.7984810126582278,
"grad_norm": 0.6574458293260896,
"learning_rate": 3.225948647317459e-05,
"loss": 0.511,
"step": 444
},
{
"epoch": 1.80253164556962,
"grad_norm": 0.5730673611586229,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.58,
"step": 445
},
{
"epoch": 1.8065822784810126,
"grad_norm": 0.623433946118166,
"learning_rate": 3.216959910368034e-05,
"loss": 0.5636,
"step": 446
},
{
"epoch": 1.8106329113924051,
"grad_norm": 0.6183651483337774,
"learning_rate": 3.212450821933277e-05,
"loss": 0.5429,
"step": 447
},
{
"epoch": 1.8146835443037974,
"grad_norm": 0.7996075873612971,
"learning_rate": 3.207931968586281e-05,
"loss": 0.5827,
"step": 448
},
{
"epoch": 1.81873417721519,
"grad_norm": 0.6303483542653614,
"learning_rate": 3.203403386721272e-05,
"loss": 0.5478,
"step": 449
},
{
"epoch": 1.8227848101265822,
"grad_norm": 0.808710062696399,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.5565,
"step": 450
},
{
"epoch": 1.8268354430379747,
"grad_norm": 0.617472071084559,
"learning_rate": 3.194317183405573e-05,
"loss": 0.5335,
"step": 451
},
{
"epoch": 1.8308860759493673,
"grad_norm": 0.7243149467299994,
"learning_rate": 3.189759635133914e-05,
"loss": 0.5676,
"step": 452
},
{
"epoch": 1.8349367088607595,
"grad_norm": 0.7938735481333364,
"learning_rate": 3.185192504701718e-05,
"loss": 0.5462,
"step": 453
},
{
"epoch": 1.8389873417721518,
"grad_norm": 0.5962536537573995,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.5813,
"step": 454
},
{
"epoch": 1.8430379746835444,
"grad_norm": 0.6814731269420201,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.5625,
"step": 455
},
{
"epoch": 1.8470886075949369,
"grad_norm": 0.5745601727373038,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.5811,
"step": 456
},
{
"epoch": 1.8511392405063292,
"grad_norm": 0.7756592970459606,
"learning_rate": 3.166828898179731e-05,
"loss": 0.5763,
"step": 457
},
{
"epoch": 1.8551898734177215,
"grad_norm": 0.7188940630790319,
"learning_rate": 3.162214410223536e-05,
"loss": 0.5379,
"step": 458
},
{
"epoch": 1.859240506329114,
"grad_norm": 0.6727579786854035,
"learning_rate": 3.157590561952257e-05,
"loss": 0.5466,
"step": 459
},
{
"epoch": 1.8632911392405065,
"grad_norm": 0.7683592643920881,
"learning_rate": 3.152957390605732e-05,
"loss": 0.5558,
"step": 460
},
{
"epoch": 1.8673417721518988,
"grad_norm": 0.6163831011813435,
"learning_rate": 3.148314933498886e-05,
"loss": 0.5506,
"step": 461
},
{
"epoch": 1.871392405063291,
"grad_norm": 0.6580508920873542,
"learning_rate": 3.143663228021431e-05,
"loss": 0.5634,
"step": 462
},
{
"epoch": 1.8754430379746836,
"grad_norm": 0.6064752299076238,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.539,
"step": 463
},
{
"epoch": 1.879493670886076,
"grad_norm": 0.6362042711815233,
"learning_rate": 3.134332221885661e-05,
"loss": 0.5484,
"step": 464
},
{
"epoch": 1.8835443037974684,
"grad_norm": 0.7350137633088459,
"learning_rate": 3.129652996377987e-05,
"loss": 0.5836,
"step": 465
},
{
"epoch": 1.8875949367088607,
"grad_norm": 0.6868154412909067,
"learning_rate": 3.12496467280038e-05,
"loss": 0.5699,
"step": 466
},
{
"epoch": 1.8916455696202532,
"grad_norm": 0.6484623491982234,
"learning_rate": 3.120267288911952e-05,
"loss": 0.5455,
"step": 467
},
{
"epoch": 1.8956962025316457,
"grad_norm": 0.6590641791148368,
"learning_rate": 3.11556088254479e-05,
"loss": 0.5988,
"step": 468
},
{
"epoch": 1.899746835443038,
"grad_norm": 0.6145192632809904,
"learning_rate": 3.11084549160364e-05,
"loss": 0.5139,
"step": 469
},
{
"epoch": 1.9037974683544303,
"grad_norm": 0.6077611055139466,
"learning_rate": 3.106121154065615e-05,
"loss": 0.572,
"step": 470
},
{
"epoch": 1.9078481012658228,
"grad_norm": 0.6158987732106659,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.5487,
"step": 471
},
{
"epoch": 1.9118987341772153,
"grad_norm": 0.7080332139204996,
"learning_rate": 3.096645791467348e-05,
"loss": 0.5602,
"step": 472
},
{
"epoch": 1.9159493670886076,
"grad_norm": 0.6251782254855555,
"learning_rate": 3.091894842720373e-05,
"loss": 0.5347,
"step": 473
},
{
"epoch": 1.92,
"grad_norm": 0.9288847076853117,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.5851,
"step": 474
},
{
"epoch": 1.9240506329113924,
"grad_norm": 0.6760049454131851,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.5389,
"step": 475
},
{
"epoch": 1.928101265822785,
"grad_norm": 0.9353823728088811,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.5667,
"step": 476
},
{
"epoch": 1.9321518987341773,
"grad_norm": 0.8720691555949339,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.5603,
"step": 477
},
{
"epoch": 1.9362025316455695,
"grad_norm": 0.7345467845103417,
"learning_rate": 3.06800895716334e-05,
"loss": 0.5556,
"step": 478
},
{
"epoch": 1.940253164556962,
"grad_norm": 0.6725246261552398,
"learning_rate": 3.063205821011029e-05,
"loss": 0.5739,
"step": 479
},
{
"epoch": 1.9443037974683546,
"grad_norm": 0.7353771940730702,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.5859,
"step": 480
},
{
"epoch": 1.9483544303797469,
"grad_norm": 0.4996400348794062,
"learning_rate": 3.053573898718618e-05,
"loss": 0.5019,
"step": 481
},
{
"epoch": 1.9524050632911392,
"grad_norm": 0.5264947481684343,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.5484,
"step": 482
},
{
"epoch": 1.9564556962025317,
"grad_norm": 0.5285417560465397,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.579,
"step": 483
},
{
"epoch": 1.9605063291139242,
"grad_norm": 0.5121243869017996,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.5576,
"step": 484
},
{
"epoch": 1.9645569620253165,
"grad_norm": 0.5725155940727261,
"learning_rate": 3.034208541656255e-05,
"loss": 0.5591,
"step": 485
},
{
"epoch": 1.9686075949367088,
"grad_norm": 0.6044286460522778,
"learning_rate": 3.029346281309218e-05,
"loss": 0.584,
"step": 486
},
{
"epoch": 1.9726582278481013,
"grad_norm": 0.6634774974569839,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.5406,
"step": 487
},
{
"epoch": 1.9767088607594938,
"grad_norm": 0.6826452890058274,
"learning_rate": 3.019596929201569e-05,
"loss": 0.5667,
"step": 488
},
{
"epoch": 1.980759493670886,
"grad_norm": 0.5805690684417074,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.5675,
"step": 489
},
{
"epoch": 1.9848101265822784,
"grad_norm": 0.5395565535121861,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.5576,
"step": 490
},
{
"epoch": 1.988860759493671,
"grad_norm": 0.6277953535687352,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.5596,
"step": 491
},
{
"epoch": 1.9929113924050634,
"grad_norm": 0.5247686791178043,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.5722,
"step": 492
},
{
"epoch": 1.9969620253164557,
"grad_norm": 0.5362016021809037,
"learning_rate": 2.995080534235264e-05,
"loss": 0.561,
"step": 493
},
{
"epoch": 2.001012658227848,
"grad_norm": 0.6518400980695667,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.594,
"step": 494
},
{
"epoch": 2.0050632911392405,
"grad_norm": 0.7034153419324323,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.4431,
"step": 495
},
{
"epoch": 2.009113924050633,
"grad_norm": 0.7889661835280647,
"learning_rate": 2.980274210295326e-05,
"loss": 0.4288,
"step": 496
},
{
"epoch": 2.013164556962025,
"grad_norm": 0.7280438415060168,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.4147,
"step": 497
},
{
"epoch": 2.0172151898734176,
"grad_norm": 0.6755930425824148,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.4558,
"step": 498
},
{
"epoch": 2.02126582278481,
"grad_norm": 0.7103224311950204,
"learning_rate": 2.965396831793362e-05,
"loss": 0.3992,
"step": 499
},
{
"epoch": 2.0253164556962027,
"grad_norm": 0.6449286984204502,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.4241,
"step": 500
},
{
"epoch": 2.0293670886075947,
"grad_norm": 0.716786095464324,
"learning_rate": 2.955439637024526e-05,
"loss": 0.4214,
"step": 501
},
{
"epoch": 2.0334177215189873,
"grad_norm": 0.6992803163400471,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.438,
"step": 502
},
{
"epoch": 2.0374683544303798,
"grad_norm": 0.8807586408463017,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.4241,
"step": 503
},
{
"epoch": 2.0415189873417723,
"grad_norm": 0.6156321974129061,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.4411,
"step": 504
},
{
"epoch": 2.0455696202531644,
"grad_norm": 0.6969519821854392,
"learning_rate": 2.93543322968538e-05,
"loss": 0.4472,
"step": 505
},
{
"epoch": 2.049620253164557,
"grad_norm": 0.5483524274378506,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.4331,
"step": 506
},
{
"epoch": 2.0536708860759494,
"grad_norm": 0.5349287999642924,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.4418,
"step": 507
},
{
"epoch": 2.057721518987342,
"grad_norm": 0.5018433557279485,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.3885,
"step": 508
},
{
"epoch": 2.061772151898734,
"grad_norm": 0.5597826859526283,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.4443,
"step": 509
},
{
"epoch": 2.0658227848101265,
"grad_norm": 0.5761630515360596,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.4373,
"step": 510
},
{
"epoch": 2.069873417721519,
"grad_norm": 0.5002814614356598,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.4018,
"step": 511
},
{
"epoch": 2.0739240506329115,
"grad_norm": 0.6024301792244767,
"learning_rate": 2.900133526131588e-05,
"loss": 0.4577,
"step": 512
},
{
"epoch": 2.0779746835443036,
"grad_norm": 0.5540209766900261,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.4323,
"step": 513
},
{
"epoch": 2.082025316455696,
"grad_norm": 0.49131915996823927,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.3988,
"step": 514
},
{
"epoch": 2.0860759493670886,
"grad_norm": 0.6033663387466254,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.4369,
"step": 515
},
{
"epoch": 2.090126582278481,
"grad_norm": 0.5403056366325356,
"learning_rate": 2.879801882631847e-05,
"loss": 0.4262,
"step": 516
},
{
"epoch": 2.094177215189873,
"grad_norm": 0.5533684102227076,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.4344,
"step": 517
},
{
"epoch": 2.0982278481012657,
"grad_norm": 0.6033343884210709,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.4445,
"step": 518
},
{
"epoch": 2.1022784810126582,
"grad_norm": 0.4674949603146612,
"learning_rate": 2.864478605989494e-05,
"loss": 0.4022,
"step": 519
},
{
"epoch": 2.1063291139240508,
"grad_norm": 0.633536725957026,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.4273,
"step": 520
},
{
"epoch": 2.110379746835443,
"grad_norm": 0.5016574146787002,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.4277,
"step": 521
},
{
"epoch": 2.1144303797468353,
"grad_norm": 0.47352516896534125,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.3989,
"step": 522
},
{
"epoch": 2.118481012658228,
"grad_norm": 0.523367951428712,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.438,
"step": 523
},
{
"epoch": 2.1225316455696204,
"grad_norm": 0.5381136923768363,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.4339,
"step": 524
},
{
"epoch": 2.1265822784810124,
"grad_norm": 0.5837196380428405,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.4422,
"step": 525
},
{
"epoch": 2.130632911392405,
"grad_norm": 0.5221427964488214,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.3944,
"step": 526
},
{
"epoch": 2.1346835443037975,
"grad_norm": 0.7031894008188919,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.4412,
"step": 527
},
{
"epoch": 2.13873417721519,
"grad_norm": 0.6907462992777811,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.432,
"step": 528
},
{
"epoch": 2.1427848101265825,
"grad_norm": 0.6560803551333382,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.4079,
"step": 529
},
{
"epoch": 2.1468354430379746,
"grad_norm": 0.7362460456023995,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.4107,
"step": 530
},
{
"epoch": 2.150886075949367,
"grad_norm": 0.4833892393432303,
"learning_rate": 2.802570062143278e-05,
"loss": 0.4141,
"step": 531
},
{
"epoch": 2.1549367088607596,
"grad_norm": 0.6923965662738462,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.443,
"step": 532
},
{
"epoch": 2.1589873417721517,
"grad_norm": 0.4972699329513223,
"learning_rate": 2.792159532078314e-05,
"loss": 0.4197,
"step": 533
},
{
"epoch": 2.163037974683544,
"grad_norm": 0.5277132922525313,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.4336,
"step": 534
},
{
"epoch": 2.1670886075949367,
"grad_norm": 0.5123881481676859,
"learning_rate": 2.781723482288857e-05,
"loss": 0.425,
"step": 535
},
{
"epoch": 2.1711392405063292,
"grad_norm": 0.4923368925495476,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.4006,
"step": 536
},
{
"epoch": 2.1751898734177217,
"grad_norm": 0.5208210027460759,
"learning_rate": 2.771262248976272e-05,
"loss": 0.4202,
"step": 537
},
{
"epoch": 2.179240506329114,
"grad_norm": 0.5504193188151246,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.429,
"step": 538
},
{
"epoch": 2.1832911392405063,
"grad_norm": 0.5836624764828461,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.439,
"step": 539
},
{
"epoch": 2.187341772151899,
"grad_norm": 0.4330685240409073,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.4105,
"step": 540
},
{
"epoch": 2.191392405063291,
"grad_norm": 0.6193950618423172,
"learning_rate": 2.7502655806328e-05,
"loss": 0.4401,
"step": 541
},
{
"epoch": 2.1954430379746834,
"grad_norm": 0.4794592716452477,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.4232,
"step": 542
},
{
"epoch": 2.199493670886076,
"grad_norm": 0.5338584715853798,
"learning_rate": 2.739730822017673e-05,
"loss": 0.4208,
"step": 543
},
{
"epoch": 2.2035443037974685,
"grad_norm": 0.5307185191905863,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.4196,
"step": 544
},
{
"epoch": 2.207594936708861,
"grad_norm": 0.5255387534110728,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.4009,
"step": 545
},
{
"epoch": 2.211645569620253,
"grad_norm": 0.6206000544756747,
"learning_rate": 2.723884107774236e-05,
"loss": 0.4629,
"step": 546
},
{
"epoch": 2.2156962025316456,
"grad_norm": 0.5801630371060076,
"learning_rate": 2.718590152796239e-05,
"loss": 0.4157,
"step": 547
},
{
"epoch": 2.219746835443038,
"grad_norm": 0.5206972540951191,
"learning_rate": 2.71329041039194e-05,
"loss": 0.3946,
"step": 548
},
{
"epoch": 2.22379746835443,
"grad_norm": 0.6589070673762235,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.4598,
"step": 549
},
{
"epoch": 2.2278481012658227,
"grad_norm": 0.5138165967720153,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.4208,
"step": 550
},
{
"epoch": 2.231898734177215,
"grad_norm": 0.6738704278013986,
"learning_rate": 2.697356885686189e-05,
"loss": 0.438,
"step": 551
},
{
"epoch": 2.2359493670886077,
"grad_norm": 0.5287707362334351,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.431,
"step": 552
},
{
"epoch": 2.24,
"grad_norm": 0.7736928586883404,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.4935,
"step": 553
},
{
"epoch": 2.2440506329113923,
"grad_norm": 0.4822128753052239,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.385,
"step": 554
},
{
"epoch": 2.248101265822785,
"grad_norm": 0.6757729095651375,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.4402,
"step": 555
},
{
"epoch": 2.2521518987341773,
"grad_norm": 0.5060079874096033,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.4217,
"step": 556
},
{
"epoch": 2.2562025316455694,
"grad_norm": 0.5914273854993878,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.4209,
"step": 557
},
{
"epoch": 2.260253164556962,
"grad_norm": 0.6090547051302861,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.4247,
"step": 558
},
{
"epoch": 2.2643037974683544,
"grad_norm": 0.5584401499807955,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.4427,
"step": 559
},
{
"epoch": 2.268354430379747,
"grad_norm": 0.47046865058035214,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.4145,
"step": 560
},
{
"epoch": 2.2724050632911394,
"grad_norm": 0.707017122069457,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.4655,
"step": 561
},
{
"epoch": 2.2764556962025315,
"grad_norm": 0.5420059029022225,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.4819,
"step": 562
},
{
"epoch": 2.280506329113924,
"grad_norm": 0.5428988540710239,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.4141,
"step": 563
},
{
"epoch": 2.2845569620253166,
"grad_norm": 0.4918707094872838,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.4192,
"step": 564
},
{
"epoch": 2.2886075949367086,
"grad_norm": 0.5308063340988234,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.4431,
"step": 565
},
{
"epoch": 2.292658227848101,
"grad_norm": 0.48366941761840937,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.4483,
"step": 566
},
{
"epoch": 2.2967088607594937,
"grad_norm": 0.47966206541694595,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.4442,
"step": 567
},
{
"epoch": 2.300759493670886,
"grad_norm": 0.47592382613497336,
"learning_rate": 2.606146264569603e-05,
"loss": 0.4323,
"step": 568
},
{
"epoch": 2.3048101265822787,
"grad_norm": 0.5480611430435247,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.4887,
"step": 569
},
{
"epoch": 2.3088607594936708,
"grad_norm": 0.46151943151857633,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.4193,
"step": 570
},
{
"epoch": 2.3129113924050633,
"grad_norm": 0.5009361519814614,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.4006,
"step": 571
},
{
"epoch": 2.316962025316456,
"grad_norm": 0.5200540531837169,
"learning_rate": 2.584472010606015e-05,
"loss": 0.4424,
"step": 572
},
{
"epoch": 2.321012658227848,
"grad_norm": 0.46929704137598416,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.4443,
"step": 573
},
{
"epoch": 2.3250632911392404,
"grad_norm": 0.5124564066588312,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.4491,
"step": 574
},
{
"epoch": 2.329113924050633,
"grad_norm": 0.48536896172029026,
"learning_rate": 2.568166741322048e-05,
"loss": 0.4187,
"step": 575
},
{
"epoch": 2.3331645569620254,
"grad_norm": 0.5188394762409319,
"learning_rate": 2.56272244128509e-05,
"loss": 0.4436,
"step": 576
},
{
"epoch": 2.337215189873418,
"grad_norm": 0.4442186176333583,
"learning_rate": 2.55727360915895e-05,
"loss": 0.4291,
"step": 577
},
{
"epoch": 2.34126582278481,
"grad_norm": 0.5192763155319922,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.4418,
"step": 578
},
{
"epoch": 2.3453164556962025,
"grad_norm": 0.5159759804393628,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.4165,
"step": 579
},
{
"epoch": 2.349367088607595,
"grad_norm": 0.5541608452316855,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.4676,
"step": 580
},
{
"epoch": 2.353417721518987,
"grad_norm": 0.488269972511917,
"learning_rate": 2.535433837985559e-05,
"loss": 0.4124,
"step": 581
},
{
"epoch": 2.3574683544303796,
"grad_norm": 0.527394099021139,
"learning_rate": 2.529963004393324e-05,
"loss": 0.4597,
"step": 582
},
{
"epoch": 2.361518987341772,
"grad_norm": 0.4881047525097385,
"learning_rate": 2.524487902551908e-05,
"loss": 0.4185,
"step": 583
},
{
"epoch": 2.3655696202531646,
"grad_norm": 0.4952655639560917,
"learning_rate": 2.519008576557029e-05,
"loss": 0.4076,
"step": 584
},
{
"epoch": 2.369620253164557,
"grad_norm": 0.5057725765236856,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.4605,
"step": 585
},
{
"epoch": 2.3736708860759492,
"grad_norm": 0.5175378129790703,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.4203,
"step": 586
},
{
"epoch": 2.3777215189873417,
"grad_norm": 0.6088568810704776,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.4393,
"step": 587
},
{
"epoch": 2.3817721518987343,
"grad_norm": 0.46016195667854415,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.437,
"step": 588
},
{
"epoch": 2.3858227848101268,
"grad_norm": 0.5589738488491828,
"learning_rate": 2.491550129992345e-05,
"loss": 0.4323,
"step": 589
},
{
"epoch": 2.389873417721519,
"grad_norm": 0.4658014202108374,
"learning_rate": 2.486046386967024e-05,
"loss": 0.3988,
"step": 590
},
{
"epoch": 2.3939240506329114,
"grad_norm": 0.5951360668286324,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.4618,
"step": 591
},
{
"epoch": 2.397974683544304,
"grad_norm": 0.41416260106440755,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.4232,
"step": 592
},
{
"epoch": 2.4020253164556964,
"grad_norm": 0.5912686214173021,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.4508,
"step": 593
},
{
"epoch": 2.4060759493670885,
"grad_norm": 0.4794761570583838,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.4107,
"step": 594
},
{
"epoch": 2.410126582278481,
"grad_norm": 0.6175628878049441,
"learning_rate": 2.458469841202444e-05,
"loss": 0.4303,
"step": 595
},
{
"epoch": 2.4141772151898735,
"grad_norm": 0.5282690927849721,
"learning_rate": 2.452943276859226e-05,
"loss": 0.4336,
"step": 596
},
{
"epoch": 2.418227848101266,
"grad_norm": 0.5663850015397305,
"learning_rate": 2.447413064573125e-05,
"loss": 0.451,
"step": 597
},
{
"epoch": 2.422278481012658,
"grad_norm": 0.45498314516068644,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.4124,
"step": 598
},
{
"epoch": 2.4263291139240506,
"grad_norm": 0.5884680209789014,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.4406,
"step": 599
},
{
"epoch": 2.430379746835443,
"grad_norm": 0.5418796403528946,
"learning_rate": 2.430800985597963e-05,
"loss": 0.4439,
"step": 600
},
{
"epoch": 2.4344303797468356,
"grad_norm": 0.5442910124881788,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.4513,
"step": 601
},
{
"epoch": 2.4384810126582277,
"grad_norm": 0.49079075072390127,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.3837,
"step": 602
},
{
"epoch": 2.44253164556962,
"grad_norm": 0.474476810100094,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.4557,
"step": 603
},
{
"epoch": 2.4465822784810127,
"grad_norm": 0.48588804043415856,
"learning_rate": 2.408603181104957e-05,
"loss": 0.4344,
"step": 604
},
{
"epoch": 2.4506329113924052,
"grad_norm": 0.558754764459188,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.476,
"step": 605
},
{
"epoch": 2.4546835443037973,
"grad_norm": 0.4009640329089911,
"learning_rate": 2.397484355008521e-05,
"loss": 0.3783,
"step": 606
},
{
"epoch": 2.45873417721519,
"grad_norm": 0.522980201505538,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.426,
"step": 607
},
{
"epoch": 2.4627848101265823,
"grad_norm": 0.5720688915518074,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.4389,
"step": 608
},
{
"epoch": 2.466835443037975,
"grad_norm": 0.4742656638260161,
"learning_rate": 2.380782218329337e-05,
"loss": 0.4587,
"step": 609
},
{
"epoch": 2.470886075949367,
"grad_norm": 0.5999530960555551,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.4383,
"step": 610
},
{
"epoch": 2.4749367088607594,
"grad_norm": 0.434125771019456,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.406,
"step": 611
},
{
"epoch": 2.478987341772152,
"grad_norm": 0.5483112047413042,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.4548,
"step": 612
},
{
"epoch": 2.4830379746835445,
"grad_norm": 0.5078526170449391,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.444,
"step": 613
},
{
"epoch": 2.4870886075949366,
"grad_norm": 0.4438811439428039,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.4064,
"step": 614
},
{
"epoch": 2.491139240506329,
"grad_norm": 0.667185267246641,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.4207,
"step": 615
},
{
"epoch": 2.4951898734177216,
"grad_norm": 0.4835530988207502,
"learning_rate": 2.341705325992901e-05,
"loss": 0.4666,
"step": 616
},
{
"epoch": 2.499240506329114,
"grad_norm": 0.551503600142696,
"learning_rate": 2.336111544604222e-05,
"loss": 0.4128,
"step": 617
},
{
"epoch": 2.503291139240506,
"grad_norm": 0.5391298873267549,
"learning_rate": 2.33051505621937e-05,
"loss": 0.441,
"step": 618
},
{
"epoch": 2.5073417721518987,
"grad_norm": 0.5085895966219096,
"learning_rate": 2.324915905911693e-05,
"loss": 0.4349,
"step": 619
},
{
"epoch": 2.511392405063291,
"grad_norm": 0.5293475869196036,
"learning_rate": 2.319314138775977e-05,
"loss": 0.4382,
"step": 620
},
{
"epoch": 2.5154430379746833,
"grad_norm": 0.4829024733797844,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.3994,
"step": 621
},
{
"epoch": 2.519493670886076,
"grad_norm": 0.5127812112619523,
"learning_rate": 2.308102934504593e-05,
"loss": 0.4575,
"step": 622
},
{
"epoch": 2.5235443037974683,
"grad_norm": 0.48341921897993745,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.4144,
"step": 623
},
{
"epoch": 2.527594936708861,
"grad_norm": 0.45640158100509026,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.4045,
"step": 624
},
{
"epoch": 2.5316455696202533,
"grad_norm": 0.47109763663678,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.4515,
"step": 625
},
{
"epoch": 2.5356962025316454,
"grad_norm": 0.5307771663438248,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.4193,
"step": 626
},
{
"epoch": 2.539746835443038,
"grad_norm": 0.5443365081761717,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.4482,
"step": 627
},
{
"epoch": 2.5437974683544304,
"grad_norm": 0.4400944953953194,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.4204,
"step": 628
},
{
"epoch": 2.547848101265823,
"grad_norm": 0.5305299055993307,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.4263,
"step": 629
},
{
"epoch": 2.5518987341772155,
"grad_norm": 0.4513514105398545,
"learning_rate": 2.26316247733093e-05,
"loss": 0.401,
"step": 630
},
{
"epoch": 2.5559493670886075,
"grad_norm": 0.5129236797632618,
"learning_rate": 2.257534907110328e-05,
"loss": 0.4545,
"step": 631
},
{
"epoch": 2.56,
"grad_norm": 0.4821314202953195,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.4258,
"step": 632
},
{
"epoch": 2.5640506329113926,
"grad_norm": 0.4385368285507809,
"learning_rate": 2.246273589557294e-05,
"loss": 0.4079,
"step": 633
},
{
"epoch": 2.5681012658227846,
"grad_norm": 0.5157927869151475,
"learning_rate": 2.240639932921966e-05,
"loss": 0.4254,
"step": 634
},
{
"epoch": 2.572151898734177,
"grad_norm": 0.49940365671022113,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.4669,
"step": 635
},
{
"epoch": 2.5762025316455697,
"grad_norm": 0.42842134618469535,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.4035,
"step": 636
},
{
"epoch": 2.580253164556962,
"grad_norm": 0.6324461147304068,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.451,
"step": 637
},
{
"epoch": 2.5843037974683547,
"grad_norm": 0.49089295215169726,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.4187,
"step": 638
},
{
"epoch": 2.5883544303797468,
"grad_norm": 0.5146267408277437,
"learning_rate": 2.212443486523819e-05,
"loss": 0.4403,
"step": 639
},
{
"epoch": 2.5924050632911393,
"grad_norm": 0.5672328575317099,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.4336,
"step": 640
},
{
"epoch": 2.596455696202532,
"grad_norm": 0.5692878253384939,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.4361,
"step": 641
},
{
"epoch": 2.600506329113924,
"grad_norm": 0.4747585938773478,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.4236,
"step": 642
},
{
"epoch": 2.6045569620253164,
"grad_norm": 0.6672119113825403,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.4217,
"step": 643
},
{
"epoch": 2.608607594936709,
"grad_norm": 0.4745346965497064,
"learning_rate": 2.184204266048803e-05,
"loss": 0.4234,
"step": 644
},
{
"epoch": 2.6126582278481014,
"grad_norm": 0.5156765050022633,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.425,
"step": 645
},
{
"epoch": 2.616708860759494,
"grad_norm": 0.5240145825633099,
"learning_rate": 2.17289787446841e-05,
"loss": 0.4422,
"step": 646
},
{
"epoch": 2.620759493670886,
"grad_norm": 0.5378982025598481,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.4469,
"step": 647
},
{
"epoch": 2.6248101265822785,
"grad_norm": 0.471599884665686,
"learning_rate": 2.161585912916385e-05,
"loss": 0.4373,
"step": 648
},
{
"epoch": 2.628860759493671,
"grad_norm": 0.42407584094114,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.4197,
"step": 649
},
{
"epoch": 2.632911392405063,
"grad_norm": 0.538685029838043,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.4282,
"step": 650
},
{
"epoch": 2.6369620253164556,
"grad_norm": 0.4295726221027334,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.4319,
"step": 651
},
{
"epoch": 2.641012658227848,
"grad_norm": 0.49458811149180903,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.4185,
"step": 652
},
{
"epoch": 2.6450632911392407,
"grad_norm": 0.48639494484920714,
"learning_rate": 2.133284032321232e-05,
"loss": 0.423,
"step": 653
},
{
"epoch": 2.649113924050633,
"grad_norm": 0.40716470589777654,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.4223,
"step": 654
},
{
"epoch": 2.6531645569620252,
"grad_norm": 0.546950418100323,
"learning_rate": 2.121955446741306e-05,
"loss": 0.4693,
"step": 655
},
{
"epoch": 2.6572151898734178,
"grad_norm": 0.4042072015300801,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.4011,
"step": 656
},
{
"epoch": 2.6612658227848103,
"grad_norm": 0.5332216591229726,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.4732,
"step": 657
},
{
"epoch": 2.6653164556962023,
"grad_norm": 0.41643474500746736,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.4252,
"step": 658
},
{
"epoch": 2.669367088607595,
"grad_norm": 0.4316372423217627,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.4115,
"step": 659
},
{
"epoch": 2.6734177215189874,
"grad_norm": 0.42655875291077666,
"learning_rate": 2.093617592756158e-05,
"loss": 0.4441,
"step": 660
},
{
"epoch": 2.67746835443038,
"grad_norm": 0.43398765618763396,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.4353,
"step": 661
},
{
"epoch": 2.6815189873417724,
"grad_norm": 0.4723712083568774,
"learning_rate": 2.082276853708978e-05,
"loss": 0.4398,
"step": 662
},
{
"epoch": 2.6855696202531645,
"grad_norm": 0.411051852338643,
"learning_rate": 2.076605467380071e-05,
"loss": 0.4169,
"step": 663
},
{
"epoch": 2.689620253164557,
"grad_norm": 0.4891397289482232,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.4557,
"step": 664
},
{
"epoch": 2.6936708860759495,
"grad_norm": 0.3947595824605611,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.3976,
"step": 665
},
{
"epoch": 2.6977215189873416,
"grad_norm": 0.47193024251570953,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.4239,
"step": 666
},
{
"epoch": 2.701772151898734,
"grad_norm": 0.47240218980490534,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.4383,
"step": 667
},
{
"epoch": 2.7058227848101266,
"grad_norm": 0.4774290812713164,
"learning_rate": 2.048240194884496e-05,
"loss": 0.4459,
"step": 668
},
{
"epoch": 2.709873417721519,
"grad_norm": 0.5087957237633663,
"learning_rate": 2.042565792045695e-05,
"loss": 0.4108,
"step": 669
},
{
"epoch": 2.7139240506329116,
"grad_norm": 0.4518653365817671,
"learning_rate": 2.036891046387857e-05,
"loss": 0.4372,
"step": 670
},
{
"epoch": 2.7179746835443037,
"grad_norm": 0.5451088547435853,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.4053,
"step": 671
},
{
"epoch": 2.7220253164556962,
"grad_norm": 0.5381111412773367,
"learning_rate": 2.025540709431948e-05,
"loss": 0.4653,
"step": 672
},
{
"epoch": 2.7260759493670887,
"grad_norm": 0.43334812115744137,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.4376,
"step": 673
},
{
"epoch": 2.730126582278481,
"grad_norm": 0.4669483278282843,
"learning_rate": 2.014189549672245e-05,
"loss": 0.4012,
"step": 674
},
{
"epoch": 2.7341772151898733,
"grad_norm": 0.46776215956908035,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.4462,
"step": 675
},
{
"epoch": 2.738227848101266,
"grad_norm": 0.4253683467269256,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.4213,
"step": 676
},
{
"epoch": 2.7422784810126584,
"grad_norm": 0.45486832423997725,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.4309,
"step": 677
},
{
"epoch": 2.746329113924051,
"grad_norm": 0.4954113381116748,
"learning_rate": 1.991486224484123e-05,
"loss": 0.4505,
"step": 678
},
{
"epoch": 2.750379746835443,
"grad_norm": 0.47085860643450944,
"learning_rate": 1.985810450327756e-05,
"loss": 0.4151,
"step": 679
},
{
"epoch": 2.7544303797468355,
"grad_norm": 0.47362951715049706,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.4212,
"step": 680
},
{
"epoch": 2.758481012658228,
"grad_norm": 0.49678660268146047,
"learning_rate": 1.974459290568053e-05,
"loss": 0.4078,
"step": 681
},
{
"epoch": 2.76253164556962,
"grad_norm": 0.5178195458889866,
"learning_rate": 1.968783996385397e-05,
"loss": 0.4543,
"step": 682
},
{
"epoch": 2.7665822784810126,
"grad_norm": 0.45041021272972254,
"learning_rate": 1.963108953612143e-05,
"loss": 0.4382,
"step": 683
},
{
"epoch": 2.770632911392405,
"grad_norm": 0.5114380637354445,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.4183,
"step": 684
},
{
"epoch": 2.7746835443037976,
"grad_norm": 0.4045023726693729,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.3992,
"step": 685
},
{
"epoch": 2.77873417721519,
"grad_norm": 0.48865889584443556,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.4352,
"step": 686
},
{
"epoch": 2.782784810126582,
"grad_norm": 0.4716107707510928,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.4491,
"step": 687
},
{
"epoch": 2.7868354430379747,
"grad_norm": 0.4450449972760815,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.4272,
"step": 688
},
{
"epoch": 2.790886075949367,
"grad_norm": 0.4272361539061772,
"learning_rate": 1.92906653591874e-05,
"loss": 0.4351,
"step": 689
},
{
"epoch": 2.7949367088607593,
"grad_norm": 0.5024055236229793,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.4086,
"step": 690
},
{
"epoch": 2.798987341772152,
"grad_norm": 0.41148318410091145,
"learning_rate": 1.917723146291022e-05,
"loss": 0.4186,
"step": 691
},
{
"epoch": 2.8030379746835443,
"grad_norm": 0.5721836783052416,
"learning_rate": 1.912052422608584e-05,
"loss": 0.4663,
"step": 692
},
{
"epoch": 2.807088607594937,
"grad_norm": 0.3924103752556536,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.4205,
"step": 693
},
{
"epoch": 2.8111392405063294,
"grad_norm": 0.4800916046383182,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.4236,
"step": 694
},
{
"epoch": 2.8151898734177214,
"grad_norm": 0.4459588790431608,
"learning_rate": 1.895044684123479e-05,
"loss": 0.4677,
"step": 695
},
{
"epoch": 2.819240506329114,
"grad_norm": 0.4612770078309109,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.3955,
"step": 696
},
{
"epoch": 2.8232911392405065,
"grad_norm": 0.44629134994907976,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.4278,
"step": 697
},
{
"epoch": 2.8273417721518985,
"grad_norm": 0.441174314848204,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.3994,
"step": 698
},
{
"epoch": 2.831392405063291,
"grad_norm": 0.42477672425784646,
"learning_rate": 1.872379746550743e-05,
"loss": 0.4159,
"step": 699
},
{
"epoch": 2.8354430379746836,
"grad_norm": 0.4253141755228025,
"learning_rate": 1.866715967678769e-05,
"loss": 0.4438,
"step": 700
},
{
"epoch": 2.839493670886076,
"grad_norm": 0.45035466025334736,
"learning_rate": 1.861053262258067e-05,
"loss": 0.4221,
"step": 701
},
{
"epoch": 2.8435443037974686,
"grad_norm": 0.45020078816594084,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.4606,
"step": 702
},
{
"epoch": 2.8475949367088607,
"grad_norm": 0.4568516087218622,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.4208,
"step": 703
},
{
"epoch": 2.851645569620253,
"grad_norm": 0.4281216046361796,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.4411,
"step": 704
},
{
"epoch": 2.8556962025316457,
"grad_norm": 0.4053335098635878,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.4098,
"step": 705
},
{
"epoch": 2.8597468354430378,
"grad_norm": 0.49098088519967176,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.4226,
"step": 706
},
{
"epoch": 2.8637974683544303,
"grad_norm": 0.43957674887997883,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.4195,
"step": 707
},
{
"epoch": 2.867848101265823,
"grad_norm": 0.42258822750151404,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.4372,
"step": 708
},
{
"epoch": 2.8718987341772153,
"grad_norm": 0.47675324729207913,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.4338,
"step": 709
},
{
"epoch": 2.875949367088608,
"grad_norm": 0.4472295975784575,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.4061,
"step": 710
},
{
"epoch": 2.88,
"grad_norm": 0.45353344348725394,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.4165,
"step": 711
},
{
"epoch": 2.8840506329113924,
"grad_norm": 0.4684507517774708,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.4489,
"step": 712
},
{
"epoch": 2.888101265822785,
"grad_norm": 0.5200130594211207,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.4415,
"step": 713
},
{
"epoch": 2.892151898734177,
"grad_norm": 0.4042733657814891,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.4156,
"step": 714
},
{
"epoch": 2.8962025316455695,
"grad_norm": 0.5048892049410512,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.4034,
"step": 715
},
{
"epoch": 2.900253164556962,
"grad_norm": 0.5965345829559612,
"learning_rate": 1.776272483901182e-05,
"loss": 0.4537,
"step": 716
},
{
"epoch": 2.9043037974683545,
"grad_norm": 0.566468798883805,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.4652,
"step": 717
},
{
"epoch": 2.908354430379747,
"grad_norm": 0.5979825939257544,
"learning_rate": 1.764995661794301e-05,
"loss": 0.3809,
"step": 718
},
{
"epoch": 2.912405063291139,
"grad_norm": 0.45231338821640965,
"learning_rate": 1.759360067078035e-05,
"loss": 0.4237,
"step": 719
},
{
"epoch": 2.9164556962025316,
"grad_norm": 0.6075760832177504,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.4474,
"step": 720
},
{
"epoch": 2.920506329113924,
"grad_norm": 0.45733826916227105,
"learning_rate": 1.748094737261012e-05,
"loss": 0.4105,
"step": 721
},
{
"epoch": 2.9245569620253162,
"grad_norm": 0.5952472456966734,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.466,
"step": 722
},
{
"epoch": 2.9286075949367087,
"grad_norm": 0.5680561047187549,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.4126,
"step": 723
},
{
"epoch": 2.9326582278481013,
"grad_norm": 0.5076590339869583,
"learning_rate": 1.731212071922883e-05,
"loss": 0.4379,
"step": 724
},
{
"epoch": 2.9367088607594938,
"grad_norm": 0.607631913011138,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.3979,
"step": 725
},
{
"epoch": 2.9407594936708863,
"grad_norm": 0.49010518803059205,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.4163,
"step": 726
},
{
"epoch": 2.9448101265822784,
"grad_norm": 0.5747989791850826,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.4361,
"step": 727
},
{
"epoch": 2.948860759493671,
"grad_norm": 0.576332761305199,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.4638,
"step": 728
},
{
"epoch": 2.9529113924050634,
"grad_norm": 0.5186447512299668,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.4464,
"step": 729
},
{
"epoch": 2.9569620253164555,
"grad_norm": 0.48061256873817865,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.4071,
"step": 730
},
{
"epoch": 2.961012658227848,
"grad_norm": 0.4575637715142271,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.4164,
"step": 731
},
{
"epoch": 2.9650632911392405,
"grad_norm": 0.42755515934116434,
"learning_rate": 1.686290200071915e-05,
"loss": 0.3964,
"step": 732
},
{
"epoch": 2.969113924050633,
"grad_norm": 0.4710409698586851,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.4653,
"step": 733
},
{
"epoch": 2.9731645569620255,
"grad_norm": 0.4249143459255096,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.4394,
"step": 734
},
{
"epoch": 2.9772151898734176,
"grad_norm": 0.47214077415695027,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.401,
"step": 735
},
{
"epoch": 2.98126582278481,
"grad_norm": 0.49890085388650035,
"learning_rate": 1.663888455395778e-05,
"loss": 0.4384,
"step": 736
},
{
"epoch": 2.9853164556962026,
"grad_norm": 0.4467486589048907,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.4326,
"step": 737
},
{
"epoch": 2.9893670886075947,
"grad_norm": 0.5210134735636267,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.4238,
"step": 738
},
{
"epoch": 2.993417721518987,
"grad_norm": 0.4307599174608029,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.4454,
"step": 739
},
{
"epoch": 2.9974683544303797,
"grad_norm": 0.4385270004238101,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.4412,
"step": 740
},
{
"epoch": 3.0015189873417722,
"grad_norm": 0.4425261980484347,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.3864,
"step": 741
},
{
"epoch": 3.0055696202531648,
"grad_norm": 0.6964724816686233,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.3459,
"step": 742
},
{
"epoch": 3.009620253164557,
"grad_norm": 0.5106266206382216,
"learning_rate": 1.624791353909428e-05,
"loss": 0.3135,
"step": 743
},
{
"epoch": 3.0136708860759494,
"grad_norm": 0.8500515385724966,
"learning_rate": 1.619217781670663e-05,
"loss": 0.3204,
"step": 744
},
{
"epoch": 3.017721518987342,
"grad_norm": 0.6344487742251707,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.2941,
"step": 745
},
{
"epoch": 3.0217721518987344,
"grad_norm": 0.7019941403289909,
"learning_rate": 1.608079882360134e-05,
"loss": 0.2939,
"step": 746
},
{
"epoch": 3.0258227848101265,
"grad_norm": 0.6166623696333943,
"learning_rate": 1.60251564499148e-05,
"loss": 0.3128,
"step": 747
},
{
"epoch": 3.029873417721519,
"grad_norm": 0.6163265192298683,
"learning_rate": 1.596954608907176e-05,
"loss": 0.3166,
"step": 748
},
{
"epoch": 3.0339240506329115,
"grad_norm": 0.5721047941129948,
"learning_rate": 1.591396818895043e-05,
"loss": 0.2998,
"step": 749
},
{
"epoch": 3.037974683544304,
"grad_norm": 0.5569636322857383,
"learning_rate": 1.585842319716759e-05,
"loss": 0.2966,
"step": 750
},
{
"epoch": 3.042025316455696,
"grad_norm": 0.4835436321091294,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.2701,
"step": 751
},
{
"epoch": 3.0460759493670886,
"grad_norm": 0.5797067171056159,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.3125,
"step": 752
},
{
"epoch": 3.050126582278481,
"grad_norm": 0.43361333546744457,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.2832,
"step": 753
},
{
"epoch": 3.0541772151898736,
"grad_norm": 0.5947330035458235,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.3307,
"step": 754
},
{
"epoch": 3.0582278481012657,
"grad_norm": 0.47325263689338504,
"learning_rate": 1.558120751116291e-05,
"loss": 0.3037,
"step": 755
},
{
"epoch": 3.062278481012658,
"grad_norm": 0.565561676283463,
"learning_rate": 1.552586935426876e-05,
"loss": 0.3526,
"step": 756
},
{
"epoch": 3.0663291139240507,
"grad_norm": 0.4394644926522546,
"learning_rate": 1.547056723140774e-05,
"loss": 0.2635,
"step": 757
},
{
"epoch": 3.0703797468354432,
"grad_norm": 0.4662275773392521,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.2856,
"step": 758
},
{
"epoch": 3.0744303797468353,
"grad_norm": 0.4354770829783537,
"learning_rate": 1.536007286907411e-05,
"loss": 0.2768,
"step": 759
},
{
"epoch": 3.078481012658228,
"grad_norm": 0.4290605501197191,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.3363,
"step": 760
},
{
"epoch": 3.0825316455696203,
"grad_norm": 0.5014933904246152,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.3065,
"step": 761
},
{
"epoch": 3.086582278481013,
"grad_norm": 0.42335550455232634,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.3071,
"step": 762
},
{
"epoch": 3.090632911392405,
"grad_norm": 0.41824932030316914,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.2822,
"step": 763
},
{
"epoch": 3.0946835443037974,
"grad_norm": 0.4567486449133481,
"learning_rate": 1.508449870007656e-05,
"loss": 0.309,
"step": 764
},
{
"epoch": 3.09873417721519,
"grad_norm": 0.3994493230888487,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.2798,
"step": 765
},
{
"epoch": 3.1027848101265825,
"grad_norm": 0.45359135927008826,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.3255,
"step": 766
},
{
"epoch": 3.1068354430379745,
"grad_norm": 0.37136799019266326,
"learning_rate": 1.4919625713405e-05,
"loss": 0.2814,
"step": 767
},
{
"epoch": 3.110886075949367,
"grad_norm": 0.3977883784084508,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.2916,
"step": 768
},
{
"epoch": 3.1149367088607596,
"grad_norm": 0.42794928792745707,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.3154,
"step": 769
},
{
"epoch": 3.118987341772152,
"grad_norm": 0.38621988663917217,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.295,
"step": 770
},
{
"epoch": 3.123037974683544,
"grad_norm": 0.41548003795858063,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.3149,
"step": 771
},
{
"epoch": 3.1270886075949367,
"grad_norm": 0.37505599066602163,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.2894,
"step": 772
},
{
"epoch": 3.131139240506329,
"grad_norm": 0.40209368537344375,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.3168,
"step": 773
},
{
"epoch": 3.1351898734177217,
"grad_norm": 0.3994414868532281,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.2866,
"step": 774
},
{
"epoch": 3.1392405063291138,
"grad_norm": 0.43985758658540103,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.2882,
"step": 775
},
{
"epoch": 3.1432911392405063,
"grad_norm": 0.37568643832570947,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.2988,
"step": 776
},
{
"epoch": 3.147341772151899,
"grad_norm": 0.3831991388051374,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.2997,
"step": 777
},
{
"epoch": 3.1513924050632913,
"grad_norm": 0.3976703169309294,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.2972,
"step": 778
},
{
"epoch": 3.1554430379746834,
"grad_norm": 0.43336609394655534,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.3252,
"step": 779
},
{
"epoch": 3.159493670886076,
"grad_norm": 0.34927675408266406,
"learning_rate": 1.420958430225303e-05,
"loss": 0.3074,
"step": 780
},
{
"epoch": 3.1635443037974684,
"grad_norm": 0.392273315368155,
"learning_rate": 1.415527989393985e-05,
"loss": 0.3348,
"step": 781
},
{
"epoch": 3.167594936708861,
"grad_norm": 0.36768140166133517,
"learning_rate": 1.410102255819891e-05,
"loss": 0.2983,
"step": 782
},
{
"epoch": 3.171645569620253,
"grad_norm": 0.35008506422205893,
"learning_rate": 1.404681273201131e-05,
"loss": 0.2919,
"step": 783
},
{
"epoch": 3.1756962025316455,
"grad_norm": 0.35809259005953026,
"learning_rate": 1.399265085197556e-05,
"loss": 0.2993,
"step": 784
},
{
"epoch": 3.179746835443038,
"grad_norm": 0.40084930335540353,
"learning_rate": 1.393853735430398e-05,
"loss": 0.3296,
"step": 785
},
{
"epoch": 3.1837974683544306,
"grad_norm": 0.3742475351506065,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.3138,
"step": 786
},
{
"epoch": 3.1878481012658226,
"grad_norm": 0.3673929789319437,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.2953,
"step": 787
},
{
"epoch": 3.191898734177215,
"grad_norm": 0.41347573309554,
"learning_rate": 1.377649151173163e-05,
"loss": 0.3197,
"step": 788
},
{
"epoch": 3.1959493670886077,
"grad_norm": 0.3634082815684886,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.2972,
"step": 789
},
{
"epoch": 3.2,
"grad_norm": 0.3702230096712765,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.3082,
"step": 790
},
{
"epoch": 3.2040506329113922,
"grad_norm": 0.3618051654692856,
"learning_rate": 1.361489677627324e-05,
"loss": 0.2799,
"step": 791
},
{
"epoch": 3.2081012658227848,
"grad_norm": 0.41142643099707266,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.3376,
"step": 792
},
{
"epoch": 3.2121518987341773,
"grad_norm": 0.3681809504151678,
"learning_rate": 1.350742335331241e-05,
"loss": 0.2948,
"step": 793
},
{
"epoch": 3.21620253164557,
"grad_norm": 0.3511567885903376,
"learning_rate": 1.345376486102198e-05,
"loss": 0.2973,
"step": 794
},
{
"epoch": 3.220253164556962,
"grad_norm": 0.3880125947938426,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.3377,
"step": 795
},
{
"epoch": 3.2243037974683544,
"grad_norm": 0.37052002844166615,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.29,
"step": 796
},
{
"epoch": 3.228354430379747,
"grad_norm": 0.43764516282859717,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.3214,
"step": 797
},
{
"epoch": 3.2324050632911394,
"grad_norm": 0.37926031055795484,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.2941,
"step": 798
},
{
"epoch": 3.2364556962025315,
"grad_norm": 0.39719924281711944,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.3098,
"step": 799
},
{
"epoch": 3.240506329113924,
"grad_norm": 0.4307812710017444,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.3076,
"step": 800
},
{
"epoch": 3.2445569620253165,
"grad_norm": 0.3999177602865204,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.332,
"step": 801
},
{
"epoch": 3.248607594936709,
"grad_norm": 0.42805750850233215,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.2902,
"step": 802
},
{
"epoch": 3.252658227848101,
"grad_norm": 0.37861739800426364,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.284,
"step": 803
},
{
"epoch": 3.2567088607594936,
"grad_norm": 0.3846915604600669,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.3163,
"step": 804
},
{
"epoch": 3.260759493670886,
"grad_norm": 0.4021531737212712,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.3163,
"step": 805
},
{
"epoch": 3.2648101265822786,
"grad_norm": 0.3813809411070368,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.2888,
"step": 806
},
{
"epoch": 3.2688607594936707,
"grad_norm": 0.4489208523998237,
"learning_rate": 1.276115892225764e-05,
"loss": 0.3185,
"step": 807
},
{
"epoch": 3.2729113924050632,
"grad_norm": 0.3801925937239439,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.3138,
"step": 808
},
{
"epoch": 3.2769620253164558,
"grad_norm": 0.40227074424736187,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.3172,
"step": 809
},
{
"epoch": 3.2810126582278483,
"grad_norm": 0.38080568782425844,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.2943,
"step": 810
},
{
"epoch": 3.2850632911392403,
"grad_norm": 0.41316432960193294,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.3272,
"step": 811
},
{
"epoch": 3.289113924050633,
"grad_norm": 0.36005921729105744,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.2843,
"step": 812
},
{
"epoch": 3.2931645569620254,
"grad_norm": 0.48826411270331044,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.3387,
"step": 813
},
{
"epoch": 3.297215189873418,
"grad_norm": 0.35086564222780364,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.266,
"step": 814
},
{
"epoch": 3.30126582278481,
"grad_norm": 0.40707238624095876,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.3045,
"step": 815
},
{
"epoch": 3.3053164556962025,
"grad_norm": 0.35681503306659135,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.2872,
"step": 816
},
{
"epoch": 3.309367088607595,
"grad_norm": 0.4074747490780596,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.3218,
"step": 817
},
{
"epoch": 3.3134177215189875,
"grad_norm": 0.4198992597497969,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.3114,
"step": 818
},
{
"epoch": 3.3174683544303796,
"grad_norm": 0.4073309074849308,
"learning_rate": 1.213055323844297e-05,
"loss": 0.3074,
"step": 819
},
{
"epoch": 3.321518987341772,
"grad_norm": 0.3698496295274397,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.2563,
"step": 820
},
{
"epoch": 3.3255696202531646,
"grad_norm": 0.39951952528586576,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.2878,
"step": 821
},
{
"epoch": 3.329620253164557,
"grad_norm": 0.41903848690708695,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.3052,
"step": 822
},
{
"epoch": 3.333670886075949,
"grad_norm": 0.391243874387018,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.2812,
"step": 823
},
{
"epoch": 3.3377215189873417,
"grad_norm": 0.42555498929901747,
"learning_rate": 1.187045262895488e-05,
"loss": 0.2964,
"step": 824
},
{
"epoch": 3.3417721518987342,
"grad_norm": 0.39005504884235653,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.3137,
"step": 825
},
{
"epoch": 3.3458227848101267,
"grad_norm": 0.4270029085583203,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.2738,
"step": 826
},
{
"epoch": 3.349873417721519,
"grad_norm": 0.4172576299112741,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.2939,
"step": 827
},
{
"epoch": 3.3539240506329113,
"grad_norm": 0.39245741236832576,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.3236,
"step": 828
},
{
"epoch": 3.357974683544304,
"grad_norm": 0.45894087044759924,
"learning_rate": 1.161198884952377e-05,
"loss": 0.2902,
"step": 829
},
{
"epoch": 3.3620253164556964,
"grad_norm": 0.39911005312462,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.2971,
"step": 830
},
{
"epoch": 3.3660759493670884,
"grad_norm": 0.41228323247114806,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.3213,
"step": 831
},
{
"epoch": 3.370126582278481,
"grad_norm": 0.40852528830491014,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.2994,
"step": 832
},
{
"epoch": 3.3741772151898735,
"grad_norm": 0.3694136517298872,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.2972,
"step": 833
},
{
"epoch": 3.378227848101266,
"grad_norm": 0.34507835502125744,
"learning_rate": 1.135521394010506e-05,
"loss": 0.2637,
"step": 834
},
{
"epoch": 3.382278481012658,
"grad_norm": 0.42218408327161827,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.3409,
"step": 835
},
{
"epoch": 3.3863291139240506,
"grad_norm": 0.3693595332794577,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.2738,
"step": 836
},
{
"epoch": 3.390379746835443,
"grad_norm": 0.3702982785746903,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.2884,
"step": 837
},
{
"epoch": 3.3944303797468356,
"grad_norm": 0.34553998784604795,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.2802,
"step": 838
},
{
"epoch": 3.3984810126582277,
"grad_norm": 0.39186623775000656,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.3329,
"step": 839
},
{
"epoch": 3.40253164556962,
"grad_norm": 0.35258257056341014,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.2952,
"step": 840
},
{
"epoch": 3.4065822784810127,
"grad_norm": 0.3716669935771999,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.287,
"step": 841
},
{
"epoch": 3.410632911392405,
"grad_norm": 0.3654106687091707,
"learning_rate": 1.0948015846744e-05,
"loss": 0.2999,
"step": 842
},
{
"epoch": 3.4146835443037973,
"grad_norm": 0.36948902063844596,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.301,
"step": 843
},
{
"epoch": 3.41873417721519,
"grad_norm": 0.35342349945682405,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.3008,
"step": 844
},
{
"epoch": 3.4227848101265823,
"grad_norm": 0.3744041221796353,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.3159,
"step": 845
},
{
"epoch": 3.426835443037975,
"grad_norm": 0.3495548477628517,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.2998,
"step": 846
},
{
"epoch": 3.430886075949367,
"grad_norm": 0.3747724375884927,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.3243,
"step": 847
},
{
"epoch": 3.4349367088607594,
"grad_norm": 0.3671134199726299,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.2977,
"step": 848
},
{
"epoch": 3.438987341772152,
"grad_norm": 0.37504534750877344,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.2887,
"step": 849
},
{
"epoch": 3.4430379746835444,
"grad_norm": 0.3697542530515682,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.3066,
"step": 850
},
{
"epoch": 3.4470886075949365,
"grad_norm": 0.417881625188286,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.3326,
"step": 851
},
{
"epoch": 3.451139240506329,
"grad_norm": 0.39616180219184016,
"learning_rate": 1.044560362975474e-05,
"loss": 0.3073,
"step": 852
},
{
"epoch": 3.4551898734177215,
"grad_norm": 0.33351823074230347,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.2618,
"step": 853
},
{
"epoch": 3.459240506329114,
"grad_norm": 0.40027552855063353,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.3246,
"step": 854
},
{
"epoch": 3.463291139240506,
"grad_norm": 0.37216841273226675,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.3097,
"step": 855
},
{
"epoch": 3.4673417721518986,
"grad_norm": 0.3899216382269836,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.2856,
"step": 856
},
{
"epoch": 3.471392405063291,
"grad_norm": 0.34837488603528605,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.2888,
"step": 857
},
{
"epoch": 3.4754430379746837,
"grad_norm": 0.3790046141609114,
"learning_rate": 1.014782400334433e-05,
"loss": 0.3341,
"step": 858
},
{
"epoch": 3.479493670886076,
"grad_norm": 0.371072588893698,
"learning_rate": 1.009846945771296e-05,
"loss": 0.2846,
"step": 859
},
{
"epoch": 3.4835443037974683,
"grad_norm": 0.42692878117588945,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.3141,
"step": 860
},
{
"epoch": 3.487594936708861,
"grad_norm": 0.3382694089906159,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.2824,
"step": 861
},
{
"epoch": 3.4916455696202533,
"grad_norm": 0.3556746716313721,
"learning_rate": 9.950885880977891e-06,
"loss": 0.2957,
"step": 862
},
{
"epoch": 3.4956962025316454,
"grad_norm": 0.39281123083110014,
"learning_rate": 9.901852696139382e-06,
"loss": 0.3062,
"step": 863
},
{
"epoch": 3.499746835443038,
"grad_norm": 0.39984050049480363,
"learning_rate": 9.852900840391027e-06,
"loss": 0.315,
"step": 864
},
{
"epoch": 3.5037974683544304,
"grad_norm": 0.3445442168313929,
"learning_rate": 9.804030707984313e-06,
"loss": 0.2767,
"step": 865
},
{
"epoch": 3.507848101265823,
"grad_norm": 0.40523080900827513,
"learning_rate": 9.755242692512599e-06,
"loss": 0.3463,
"step": 866
},
{
"epoch": 3.5118987341772154,
"grad_norm": 0.34612725602958255,
"learning_rate": 9.70653718690782e-06,
"loss": 0.28,
"step": 867
},
{
"epoch": 3.5159493670886075,
"grad_norm": 0.38785340600860285,
"learning_rate": 9.657914583437454e-06,
"loss": 0.3077,
"step": 868
},
{
"epoch": 3.52,
"grad_norm": 0.3748766247445659,
"learning_rate": 9.609375273701246e-06,
"loss": 0.3117,
"step": 869
},
{
"epoch": 3.5240506329113925,
"grad_norm": 0.3730965978354007,
"learning_rate": 9.560919648628133e-06,
"loss": 0.3057,
"step": 870
},
{
"epoch": 3.5281012658227846,
"grad_norm": 0.37317958189860084,
"learning_rate": 9.512548098473047e-06,
"loss": 0.3176,
"step": 871
},
{
"epoch": 3.532151898734177,
"grad_norm": 0.370982283940031,
"learning_rate": 9.464261012813825e-06,
"loss": 0.2997,
"step": 872
},
{
"epoch": 3.5362025316455696,
"grad_norm": 0.36437958148649646,
"learning_rate": 9.416058780547987e-06,
"loss": 0.2849,
"step": 873
},
{
"epoch": 3.540253164556962,
"grad_norm": 0.3544364775541987,
"learning_rate": 9.367941789889714e-06,
"loss": 0.2822,
"step": 874
},
{
"epoch": 3.5443037974683547,
"grad_norm": 0.38563538742283315,
"learning_rate": 9.319910428366607e-06,
"loss": 0.321,
"step": 875
},
{
"epoch": 3.5483544303797467,
"grad_norm": 0.371470472163226,
"learning_rate": 9.271965082816667e-06,
"loss": 0.3176,
"step": 876
},
{
"epoch": 3.5524050632911393,
"grad_norm": 0.3622408849980764,
"learning_rate": 9.224106139385111e-06,
"loss": 0.2655,
"step": 877
},
{
"epoch": 3.5564556962025318,
"grad_norm": 0.383753863794064,
"learning_rate": 9.176333983521291e-06,
"loss": 0.3095,
"step": 878
},
{
"epoch": 3.560506329113924,
"grad_norm": 0.35460261841071244,
"learning_rate": 9.12864899997558e-06,
"loss": 0.3167,
"step": 879
},
{
"epoch": 3.5645569620253164,
"grad_norm": 0.3622615253384486,
"learning_rate": 9.08105157279628e-06,
"loss": 0.3088,
"step": 880
},
{
"epoch": 3.568607594936709,
"grad_norm": 0.3720999571807848,
"learning_rate": 9.03354208532653e-06,
"loss": 0.2966,
"step": 881
},
{
"epoch": 3.5726582278481014,
"grad_norm": 0.36735811913100624,
"learning_rate": 8.986120920201205e-06,
"loss": 0.3051,
"step": 882
},
{
"epoch": 3.576708860759494,
"grad_norm": 0.35991041410923047,
"learning_rate": 8.938788459343852e-06,
"loss": 0.2856,
"step": 883
},
{
"epoch": 3.580759493670886,
"grad_norm": 0.34722893253671905,
"learning_rate": 8.8915450839636e-06,
"loss": 0.3073,
"step": 884
},
{
"epoch": 3.5848101265822785,
"grad_norm": 0.3953034274988757,
"learning_rate": 8.844391174552116e-06,
"loss": 0.3065,
"step": 885
},
{
"epoch": 3.588860759493671,
"grad_norm": 0.3869644722902605,
"learning_rate": 8.797327110880479e-06,
"loss": 0.3188,
"step": 886
},
{
"epoch": 3.592911392405063,
"grad_norm": 0.3775324956840317,
"learning_rate": 8.750353271996206e-06,
"loss": 0.3058,
"step": 887
},
{
"epoch": 3.5969620253164556,
"grad_norm": 0.3710994185908825,
"learning_rate": 8.703470036220132e-06,
"loss": 0.2801,
"step": 888
},
{
"epoch": 3.601012658227848,
"grad_norm": 0.384011714600533,
"learning_rate": 8.656677781143394e-06,
"loss": 0.2947,
"step": 889
},
{
"epoch": 3.6050632911392406,
"grad_norm": 0.3454870828748119,
"learning_rate": 8.609976883624377e-06,
"loss": 0.3163,
"step": 890
},
{
"epoch": 3.609113924050633,
"grad_norm": 0.3701291880237732,
"learning_rate": 8.563367719785698e-06,
"loss": 0.2892,
"step": 891
},
{
"epoch": 3.613164556962025,
"grad_norm": 0.33566399134922265,
"learning_rate": 8.516850665011138e-06,
"loss": 0.2484,
"step": 892
},
{
"epoch": 3.6172151898734177,
"grad_norm": 0.3877717810668131,
"learning_rate": 8.47042609394269e-06,
"loss": 0.3253,
"step": 893
},
{
"epoch": 3.6212658227848102,
"grad_norm": 0.3584054588522732,
"learning_rate": 8.424094380477432e-06,
"loss": 0.2969,
"step": 894
},
{
"epoch": 3.6253164556962023,
"grad_norm": 0.35530087031816704,
"learning_rate": 8.37785589776465e-06,
"loss": 0.2875,
"step": 895
},
{
"epoch": 3.629367088607595,
"grad_norm": 0.35600437020393927,
"learning_rate": 8.331711018202694e-06,
"loss": 0.3045,
"step": 896
},
{
"epoch": 3.6334177215189873,
"grad_norm": 0.3617294479525603,
"learning_rate": 8.285660113436104e-06,
"loss": 0.311,
"step": 897
},
{
"epoch": 3.63746835443038,
"grad_norm": 0.37839507972934644,
"learning_rate": 8.239703554352527e-06,
"loss": 0.2904,
"step": 898
},
{
"epoch": 3.6415189873417724,
"grad_norm": 0.42765289876184265,
"learning_rate": 8.193841711079775e-06,
"loss": 0.3315,
"step": 899
},
{
"epoch": 3.6455696202531644,
"grad_norm": 0.3543153045877358,
"learning_rate": 8.148074952982828e-06,
"loss": 0.2779,
"step": 900
},
{
"epoch": 3.649620253164557,
"grad_norm": 0.3839773726461832,
"learning_rate": 8.102403648660859e-06,
"loss": 0.3102,
"step": 901
},
{
"epoch": 3.6536708860759495,
"grad_norm": 0.3942872960865684,
"learning_rate": 8.056828165944282e-06,
"loss": 0.3021,
"step": 902
},
{
"epoch": 3.6577215189873415,
"grad_norm": 0.38157557801147945,
"learning_rate": 8.011348871891762e-06,
"loss": 0.3141,
"step": 903
},
{
"epoch": 3.661772151898734,
"grad_norm": 0.38232817883951253,
"learning_rate": 7.965966132787287e-06,
"loss": 0.3124,
"step": 904
},
{
"epoch": 3.6658227848101266,
"grad_norm": 0.3908361626411096,
"learning_rate": 7.920680314137189e-06,
"loss": 0.2871,
"step": 905
},
{
"epoch": 3.669873417721519,
"grad_norm": 0.43103191947884445,
"learning_rate": 7.875491780667246e-06,
"loss": 0.2951,
"step": 906
},
{
"epoch": 3.6739240506329116,
"grad_norm": 0.4513857385549617,
"learning_rate": 7.830400896319667e-06,
"loss": 0.2892,
"step": 907
},
{
"epoch": 3.6779746835443037,
"grad_norm": 0.3641998301436886,
"learning_rate": 7.785408024250259e-06,
"loss": 0.3376,
"step": 908
},
{
"epoch": 3.682025316455696,
"grad_norm": 0.398935155391295,
"learning_rate": 7.74051352682542e-06,
"loss": 0.2787,
"step": 909
},
{
"epoch": 3.6860759493670887,
"grad_norm": 0.4351635489581466,
"learning_rate": 7.695717765619257e-06,
"loss": 0.3318,
"step": 910
},
{
"epoch": 3.690126582278481,
"grad_norm": 0.34978419363240926,
"learning_rate": 7.651021101410673e-06,
"loss": 0.2648,
"step": 911
},
{
"epoch": 3.6941772151898733,
"grad_norm": 0.4470278269350219,
"learning_rate": 7.606423894180464e-06,
"loss": 0.3068,
"step": 912
},
{
"epoch": 3.698227848101266,
"grad_norm": 0.41035660701916593,
"learning_rate": 7.56192650310839e-06,
"loss": 0.2819,
"step": 913
},
{
"epoch": 3.7022784810126583,
"grad_norm": 0.3777246122964969,
"learning_rate": 7.517529286570349e-06,
"loss": 0.3183,
"step": 914
},
{
"epoch": 3.706329113924051,
"grad_norm": 0.34493893256324604,
"learning_rate": 7.473232602135387e-06,
"loss": 0.2999,
"step": 915
},
{
"epoch": 3.710379746835443,
"grad_norm": 0.34907284778716535,
"learning_rate": 7.429036806562935e-06,
"loss": 0.2566,
"step": 916
},
{
"epoch": 3.7144303797468354,
"grad_norm": 0.3790837911234484,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.3151,
"step": 917
},
{
"epoch": 3.718481012658228,
"grad_norm": 0.3759314026284005,
"learning_rate": 7.340949304977567e-06,
"loss": 0.3388,
"step": 918
},
{
"epoch": 3.72253164556962,
"grad_norm": 0.3627651386533068,
"learning_rate": 7.297058308409282e-06,
"loss": 0.3037,
"step": 919
},
{
"epoch": 3.7265822784810125,
"grad_norm": 0.39576854194375305,
"learning_rate": 7.25326961958704e-06,
"loss": 0.2806,
"step": 920
},
{
"epoch": 3.730632911392405,
"grad_norm": 0.38498345922116656,
"learning_rate": 7.209583591178921e-06,
"loss": 0.3274,
"step": 921
},
{
"epoch": 3.7346835443037976,
"grad_norm": 0.3368175018870942,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.2792,
"step": 922
},
{
"epoch": 3.73873417721519,
"grad_norm": 0.37672843541515516,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.2989,
"step": 923
},
{
"epoch": 3.742784810126582,
"grad_norm": 0.3508586154992745,
"learning_rate": 7.079144982700909e-06,
"loss": 0.331,
"step": 924
},
{
"epoch": 3.7468354430379747,
"grad_norm": 0.34131251451055616,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.3262,
"step": 925
},
{
"epoch": 3.750886075949367,
"grad_norm": 0.3427597525997347,
"learning_rate": 6.99270564069757e-06,
"loss": 0.2751,
"step": 926
},
{
"epoch": 3.7549367088607593,
"grad_norm": 0.3627547067665001,
"learning_rate": 6.949642934304375e-06,
"loss": 0.2986,
"step": 927
},
{
"epoch": 3.7589873417721518,
"grad_norm": 0.37526434528767233,
"learning_rate": 6.906685333692871e-06,
"loss": 0.3057,
"step": 928
},
{
"epoch": 3.7630379746835443,
"grad_norm": 0.3245700166288051,
"learning_rate": 6.86383318483769e-06,
"loss": 0.293,
"step": 929
},
{
"epoch": 3.767088607594937,
"grad_norm": 0.36173847070612225,
"learning_rate": 6.821086832864139e-06,
"loss": 0.3137,
"step": 930
},
{
"epoch": 3.7711392405063293,
"grad_norm": 0.36741790509762956,
"learning_rate": 6.77844662204546e-06,
"loss": 0.296,
"step": 931
},
{
"epoch": 3.7751898734177214,
"grad_norm": 0.3333882704967745,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.2851,
"step": 932
},
{
"epoch": 3.779240506329114,
"grad_norm": 0.36808299808997974,
"learning_rate": 6.693485996688695e-06,
"loss": 0.3154,
"step": 933
},
{
"epoch": 3.7832911392405064,
"grad_norm": 0.3597595928641474,
"learning_rate": 6.651166266411801e-06,
"loss": 0.265,
"step": 934
},
{
"epoch": 3.7873417721518985,
"grad_norm": 0.3549118546678562,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.3107,
"step": 935
},
{
"epoch": 3.791392405063291,
"grad_norm": 0.36085601787218335,
"learning_rate": 6.566849674844711e-06,
"loss": 0.3261,
"step": 936
},
{
"epoch": 3.7954430379746835,
"grad_norm": 0.3469607814639242,
"learning_rate": 6.524853492628747e-06,
"loss": 0.2873,
"step": 937
},
{
"epoch": 3.799493670886076,
"grad_norm": 0.36188569798880776,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.2866,
"step": 938
},
{
"epoch": 3.8035443037974686,
"grad_norm": 0.38514171772305855,
"learning_rate": 6.441187046486648e-06,
"loss": 0.3147,
"step": 939
},
{
"epoch": 3.8075949367088606,
"grad_norm": 0.3520224529754323,
"learning_rate": 6.399517456398567e-06,
"loss": 0.2675,
"step": 940
},
{
"epoch": 3.811645569620253,
"grad_norm": 0.35398377955706917,
"learning_rate": 6.357957402727164e-06,
"loss": 0.3154,
"step": 941
},
{
"epoch": 3.8156962025316457,
"grad_norm": 0.35533536744490285,
"learning_rate": 6.316507220191395e-06,
"loss": 0.2989,
"step": 942
},
{
"epoch": 3.8197468354430377,
"grad_norm": 0.37256149899552304,
"learning_rate": 6.275167242625331e-06,
"loss": 0.3189,
"step": 943
},
{
"epoch": 3.8237974683544302,
"grad_norm": 0.35235175616549785,
"learning_rate": 6.233937802975471e-06,
"loss": 0.281,
"step": 944
},
{
"epoch": 3.8278481012658228,
"grad_norm": 0.8433084589548899,
"learning_rate": 6.192819233298046e-06,
"loss": 0.3132,
"step": 945
},
{
"epoch": 3.8318987341772153,
"grad_norm": 0.32662418968682577,
"learning_rate": 6.151811864756383e-06,
"loss": 0.2898,
"step": 946
},
{
"epoch": 3.835949367088608,
"grad_norm": 0.3886882731178603,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.3233,
"step": 947
},
{
"epoch": 3.84,
"grad_norm": 0.4111242372036189,
"learning_rate": 6.070132051252868e-06,
"loss": 0.3226,
"step": 948
},
{
"epoch": 3.8440506329113924,
"grad_norm": 0.3252339917278776,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.2785,
"step": 949
},
{
"epoch": 3.848101265822785,
"grad_norm": 0.3537853625792128,
"learning_rate": 5.988900993811575e-06,
"loss": 0.3315,
"step": 950
},
{
"epoch": 3.852151898734177,
"grad_norm": 0.33822341651981946,
"learning_rate": 5.948454566959363e-06,
"loss": 0.3054,
"step": 951
},
{
"epoch": 3.8562025316455695,
"grad_norm": 0.3838845201912748,
"learning_rate": 5.908121309322328e-06,
"loss": 0.3139,
"step": 952
},
{
"epoch": 3.860253164556962,
"grad_norm": 0.3249520850543603,
"learning_rate": 5.867901545738976e-06,
"loss": 0.3086,
"step": 953
},
{
"epoch": 3.8643037974683545,
"grad_norm": 0.3492442850413824,
"learning_rate": 5.827795600133774e-06,
"loss": 0.3048,
"step": 954
},
{
"epoch": 3.868354430379747,
"grad_norm": 0.3439399323503794,
"learning_rate": 5.787803795514466e-06,
"loss": 0.3061,
"step": 955
},
{
"epoch": 3.872405063291139,
"grad_norm": 0.33726952648324027,
"learning_rate": 5.747926453969576e-06,
"loss": 0.2873,
"step": 956
},
{
"epoch": 3.8764556962025316,
"grad_norm": 0.32427364454956353,
"learning_rate": 5.708163896665708e-06,
"loss": 0.2793,
"step": 957
},
{
"epoch": 3.880506329113924,
"grad_norm": 0.4048428473400336,
"learning_rate": 5.668516443845047e-06,
"loss": 0.3413,
"step": 958
},
{
"epoch": 3.884556962025316,
"grad_norm": 0.3472544693198482,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.2655,
"step": 959
},
{
"epoch": 3.8886075949367087,
"grad_norm": 0.3618594545943795,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.3408,
"step": 960
},
{
"epoch": 3.8926582278481012,
"grad_norm": 0.3390687508114334,
"learning_rate": 5.550267900783019e-06,
"loss": 0.3062,
"step": 961
},
{
"epoch": 3.8967088607594937,
"grad_norm": 0.3363997320494114,
"learning_rate": 5.511084049737623e-06,
"loss": 0.3003,
"step": 962
},
{
"epoch": 3.9007594936708863,
"grad_norm": 0.3267511148571096,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.2908,
"step": 963
},
{
"epoch": 3.9048101265822783,
"grad_norm": 0.3593974518568918,
"learning_rate": 5.433066737499948e-06,
"loss": 0.3155,
"step": 964
},
{
"epoch": 3.908860759493671,
"grad_norm": 0.34186939706036096,
"learning_rate": 5.394233904648376e-06,
"loss": 0.2794,
"step": 965
},
{
"epoch": 3.9129113924050634,
"grad_norm": 0.33167288147483714,
"learning_rate": 5.355518704628997e-06,
"loss": 0.3169,
"step": 966
},
{
"epoch": 3.9169620253164554,
"grad_norm": 0.36262620051900063,
"learning_rate": 5.316921449248731e-06,
"loss": 0.3088,
"step": 967
},
{
"epoch": 3.921012658227848,
"grad_norm": 0.3603649778077229,
"learning_rate": 5.278442449364538e-06,
"loss": 0.2866,
"step": 968
},
{
"epoch": 3.9250632911392405,
"grad_norm": 0.36968822626259906,
"learning_rate": 5.240082014881016e-06,
"loss": 0.3245,
"step": 969
},
{
"epoch": 3.929113924050633,
"grad_norm": 0.3525810353713833,
"learning_rate": 5.201840454747822e-06,
"loss": 0.2861,
"step": 970
},
{
"epoch": 3.9331645569620255,
"grad_norm": 0.3457696224454169,
"learning_rate": 5.163718076957223e-06,
"loss": 0.2707,
"step": 971
},
{
"epoch": 3.9372151898734176,
"grad_norm": 0.37807236029413,
"learning_rate": 5.125715188541609e-06,
"loss": 0.3103,
"step": 972
},
{
"epoch": 3.94126582278481,
"grad_norm": 0.3569154733193089,
"learning_rate": 5.087832095571021e-06,
"loss": 0.3092,
"step": 973
},
{
"epoch": 3.9453164556962026,
"grad_norm": 0.36438576658192273,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.283,
"step": 974
},
{
"epoch": 3.9493670886075947,
"grad_norm": 0.3333897398791141,
"learning_rate": 5.01242651541854e-06,
"loss": 0.2824,
"step": 975
},
{
"epoch": 3.953417721518987,
"grad_norm": 0.3688035922439317,
"learning_rate": 4.974904635542815e-06,
"loss": 0.3089,
"step": 976
},
{
"epoch": 3.9574683544303797,
"grad_norm": 0.36809176619987594,
"learning_rate": 4.937503765719582e-06,
"loss": 0.3031,
"step": 977
},
{
"epoch": 3.961518987341772,
"grad_norm": 0.39709632219556823,
"learning_rate": 4.900224207170299e-06,
"loss": 0.3135,
"step": 978
},
{
"epoch": 3.9655696202531647,
"grad_norm": 0.3258976768532089,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.2788,
"step": 979
},
{
"epoch": 3.969620253164557,
"grad_norm": 0.3523052248031429,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.3075,
"step": 980
},
{
"epoch": 3.9736708860759493,
"grad_norm": 0.34542277222435747,
"learning_rate": 4.789116396710924e-06,
"loss": 0.2886,
"step": 981
},
{
"epoch": 3.977721518987342,
"grad_norm": 0.3609849333299127,
"learning_rate": 4.752325075895368e-06,
"loss": 0.3057,
"step": 982
},
{
"epoch": 3.981772151898734,
"grad_norm": 0.34405851499868984,
"learning_rate": 4.715656557757473e-06,
"loss": 0.2893,
"step": 983
},
{
"epoch": 3.9858227848101264,
"grad_norm": 0.33753023052317693,
"learning_rate": 4.679111137620442e-06,
"loss": 0.3259,
"step": 984
},
{
"epoch": 3.989873417721519,
"grad_norm": 0.3566095749823653,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.2908,
"step": 985
},
{
"epoch": 3.9939240506329114,
"grad_norm": 0.3567551993872493,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.2972,
"step": 986
},
{
"epoch": 3.997974683544304,
"grad_norm": 0.41687990087576654,
"learning_rate": 4.570216403561141e-06,
"loss": 0.3404,
"step": 987
},
{
"epoch": 4.002025316455696,
"grad_norm": 0.4143891730974457,
"learning_rate": 4.534166308795815e-06,
"loss": 0.2711,
"step": 988
},
{
"epoch": 4.006075949367089,
"grad_norm": 0.5902991419886133,
"learning_rate": 4.498240773728859e-06,
"loss": 0.2368,
"step": 989
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.4731033665895566,
"learning_rate": 4.462440087699609e-06,
"loss": 0.2093,
"step": 990
},
{
"epoch": 4.014177215189873,
"grad_norm": 0.3565991896665581,
"learning_rate": 4.426764539041861e-06,
"loss": 0.2317,
"step": 991
},
{
"epoch": 4.018227848101266,
"grad_norm": 0.3886162282837213,
"learning_rate": 4.391214415081582e-06,
"loss": 0.1915,
"step": 992
},
{
"epoch": 4.022278481012658,
"grad_norm": 0.6092458526648769,
"learning_rate": 4.355790002134579e-06,
"loss": 0.2302,
"step": 993
},
{
"epoch": 4.02632911392405,
"grad_norm": 0.6393153267739573,
"learning_rate": 4.320491585504207e-06,
"loss": 0.2219,
"step": 994
},
{
"epoch": 4.030379746835443,
"grad_norm": 0.5113856800090277,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.2163,
"step": 995
},
{
"epoch": 4.034430379746835,
"grad_norm": 0.4188829721570954,
"learning_rate": 4.250273877330691e-06,
"loss": 0.2269,
"step": 996
},
{
"epoch": 4.038481012658228,
"grad_norm": 0.4452716956578497,
"learning_rate": 4.215355151311313e-06,
"loss": 0.2241,
"step": 997
},
{
"epoch": 4.04253164556962,
"grad_norm": 0.4694881902120784,
"learning_rate": 4.180563552651542e-06,
"loss": 0.2169,
"step": 998
},
{
"epoch": 4.046582278481012,
"grad_norm": 0.5013480753451433,
"learning_rate": 4.145899361558147e-06,
"loss": 0.2332,
"step": 999
},
{
"epoch": 4.050632911392405,
"grad_norm": 0.4425336888334855,
"learning_rate": 4.111362857211738e-06,
"loss": 0.2099,
"step": 1000
},
{
"epoch": 4.054683544303797,
"grad_norm": 0.41965026016057605,
"learning_rate": 4.076954317764592e-06,
"loss": 0.2073,
"step": 1001
},
{
"epoch": 4.0587341772151895,
"grad_norm": 0.38785531598635287,
"learning_rate": 4.042674020338335e-06,
"loss": 0.2042,
"step": 1002
},
{
"epoch": 4.062784810126582,
"grad_norm": 0.3914508532456278,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.2092,
"step": 1003
},
{
"epoch": 4.0668354430379745,
"grad_norm": 0.40146630515890763,
"learning_rate": 3.974499254868674e-06,
"loss": 0.209,
"step": 1004
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.4222301735324834,
"learning_rate": 3.940605335895451e-06,
"loss": 0.2259,
"step": 1005
},
{
"epoch": 4.0749367088607595,
"grad_norm": 0.3983025216668947,
"learning_rate": 3.90684075707908e-06,
"loss": 0.2223,
"step": 1006
},
{
"epoch": 4.078987341772152,
"grad_norm": 0.3612679577850778,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.2102,
"step": 1007
},
{
"epoch": 4.083037974683545,
"grad_norm": 0.37025290807271927,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.2394,
"step": 1008
},
{
"epoch": 4.087088607594937,
"grad_norm": 0.3702764180469184,
"learning_rate": 3.806325775702304e-06,
"loss": 0.2267,
"step": 1009
},
{
"epoch": 4.091139240506329,
"grad_norm": 0.35022196826318547,
"learning_rate": 3.773081266416434e-06,
"loss": 0.2055,
"step": 1010
},
{
"epoch": 4.095189873417722,
"grad_norm": 0.374198942089115,
"learning_rate": 3.739967446503245e-06,
"loss": 0.2348,
"step": 1011
},
{
"epoch": 4.099240506329114,
"grad_norm": 0.36087509096487813,
"learning_rate": 3.706984582656894e-06,
"loss": 0.2139,
"step": 1012
},
{
"epoch": 4.103291139240507,
"grad_norm": 0.33566307551514984,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.2135,
"step": 1013
},
{
"epoch": 4.107341772151899,
"grad_norm": 0.3535954148001838,
"learning_rate": 3.641412784665648e-06,
"loss": 0.2212,
"step": 1014
},
{
"epoch": 4.111392405063291,
"grad_norm": 0.34356101909949577,
"learning_rate": 3.608824378627005e-06,
"loss": 0.2134,
"step": 1015
},
{
"epoch": 4.115443037974684,
"grad_norm": 0.33395941471939034,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.1974,
"step": 1016
},
{
"epoch": 4.119493670886076,
"grad_norm": 0.3548444345258014,
"learning_rate": 3.544043864774269e-06,
"loss": 0.2374,
"step": 1017
},
{
"epoch": 4.123544303797468,
"grad_norm": 0.34696290948619435,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.2158,
"step": 1018
},
{
"epoch": 4.127594936708861,
"grad_norm": 0.38586030477702626,
"learning_rate": 3.479793485887819e-06,
"loss": 0.2471,
"step": 1019
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.33689424521836514,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.2093,
"step": 1020
},
{
"epoch": 4.135696202531646,
"grad_norm": 0.32814488214600773,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.1966,
"step": 1021
},
{
"epoch": 4.139746835443038,
"grad_norm": 0.32337311156790455,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.2144,
"step": 1022
},
{
"epoch": 4.14379746835443,
"grad_norm": 0.34250323992813353,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.221,
"step": 1023
},
{
"epoch": 4.147848101265823,
"grad_norm": 0.3140436196145639,
"learning_rate": 3.321500420335e-06,
"loss": 0.2057,
"step": 1024
},
{
"epoch": 4.151898734177215,
"grad_norm": 0.3459730964293516,
"learning_rate": 3.290243771741275e-06,
"loss": 0.2357,
"step": 1025
},
{
"epoch": 4.155949367088607,
"grad_norm": 0.3323810607974555,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.2335,
"step": 1026
},
{
"epoch": 4.16,
"grad_norm": 0.306519379314799,
"learning_rate": 3.228134459445149e-06,
"loss": 0.1926,
"step": 1027
},
{
"epoch": 4.164050632911392,
"grad_norm": 0.3359898069848761,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.2221,
"step": 1028
},
{
"epoch": 4.168101265822785,
"grad_norm": 0.3286344791954443,
"learning_rate": 3.166565459258513e-06,
"loss": 0.2053,
"step": 1029
},
{
"epoch": 4.172151898734177,
"grad_norm": 0.3503871474561193,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.2373,
"step": 1030
},
{
"epoch": 4.176202531645569,
"grad_norm": 0.3137003608739278,
"learning_rate": 3.105538754650419e-06,
"loss": 0.2184,
"step": 1031
},
{
"epoch": 4.180253164556962,
"grad_norm": 0.3028382411541864,
"learning_rate": 3.07522937824712e-06,
"loss": 0.2239,
"step": 1032
},
{
"epoch": 4.184303797468354,
"grad_norm": 0.3084612826334476,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.2405,
"step": 1033
},
{
"epoch": 4.188354430379746,
"grad_norm": 0.3089808970363591,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.2171,
"step": 1034
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.31866549686748347,
"learning_rate": 2.985120078631465e-06,
"loss": 0.2201,
"step": 1035
},
{
"epoch": 4.1964556962025314,
"grad_norm": 0.30363159891377184,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.1887,
"step": 1036
},
{
"epoch": 4.200506329113924,
"grad_norm": 0.3173815818246447,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.2192,
"step": 1037
},
{
"epoch": 4.2045569620253165,
"grad_norm": 0.30746028031258016,
"learning_rate": 2.896244091927678e-06,
"loss": 0.2192,
"step": 1038
},
{
"epoch": 4.2086075949367086,
"grad_norm": 0.3309922156058177,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.223,
"step": 1039
},
{
"epoch": 4.2126582278481015,
"grad_norm": 0.3212385077635499,
"learning_rate": 2.837681792951994e-06,
"loss": 0.2371,
"step": 1040
},
{
"epoch": 4.216708860759494,
"grad_norm": 0.30771937444280173,
"learning_rate": 2.808607860255981e-06,
"loss": 0.2125,
"step": 1041
},
{
"epoch": 4.220759493670886,
"grad_norm": 0.3323988953574119,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.218,
"step": 1042
},
{
"epoch": 4.224810126582279,
"grad_norm": 0.32082732165340955,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.2304,
"step": 1043
},
{
"epoch": 4.228860759493671,
"grad_norm": 0.30012000961982543,
"learning_rate": 2.722217735873718e-06,
"loss": 0.1907,
"step": 1044
},
{
"epoch": 4.232911392405064,
"grad_norm": 0.31394266705706664,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.2044,
"step": 1045
},
{
"epoch": 4.236962025316456,
"grad_norm": 0.3126909963611483,
"learning_rate": 2.665319697490205e-06,
"loss": 0.2254,
"step": 1046
},
{
"epoch": 4.241012658227848,
"grad_norm": 0.3207207529962264,
"learning_rate": 2.637079980714945e-06,
"loss": 0.2205,
"step": 1047
},
{
"epoch": 4.245063291139241,
"grad_norm": 0.3159371301637313,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.2127,
"step": 1048
},
{
"epoch": 4.249113924050633,
"grad_norm": 0.30407349065050027,
"learning_rate": 2.581020289189571e-06,
"loss": 0.2029,
"step": 1049
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.3219726957925647,
"learning_rate": 2.553200765936501e-06,
"loss": 0.2277,
"step": 1050
},
{
"epoch": 4.257215189873418,
"grad_norm": 0.31728083525539963,
"learning_rate": 2.525521756805962e-06,
"loss": 0.2167,
"step": 1051
},
{
"epoch": 4.26126582278481,
"grad_norm": 0.3425233124240346,
"learning_rate": 2.497983484720885e-06,
"loss": 0.2366,
"step": 1052
},
{
"epoch": 4.265316455696203,
"grad_norm": 0.3628748923567657,
"learning_rate": 2.470586171470728e-06,
"loss": 0.2591,
"step": 1053
},
{
"epoch": 4.269367088607595,
"grad_norm": 0.30263919647383186,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.1988,
"step": 1054
},
{
"epoch": 4.273417721518987,
"grad_norm": 0.32351678517798854,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.233,
"step": 1055
},
{
"epoch": 4.27746835443038,
"grad_norm": 0.3237002403505897,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.2214,
"step": 1056
},
{
"epoch": 4.281518987341772,
"grad_norm": 0.30006606961949156,
"learning_rate": 2.362410902836978e-06,
"loss": 0.1916,
"step": 1057
},
{
"epoch": 4.285569620253165,
"grad_norm": 0.3088418036304727,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.2133,
"step": 1058
},
{
"epoch": 4.289620253164557,
"grad_norm": 0.3308846565110285,
"learning_rate": 2.309174704446411e-06,
"loss": 0.2389,
"step": 1059
},
{
"epoch": 4.293670886075949,
"grad_norm": 0.3244210795838209,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.2183,
"step": 1060
},
{
"epoch": 4.297721518987342,
"grad_norm": 0.31995623598482936,
"learning_rate": 2.256508422807855e-06,
"loss": 0.2102,
"step": 1061
},
{
"epoch": 4.301772151898734,
"grad_norm": 0.3219230896605183,
"learning_rate": 2.230389531696946e-06,
"loss": 0.2085,
"step": 1062
},
{
"epoch": 4.305822784810126,
"grad_norm": 0.3369552371810801,
"learning_rate": 2.204413754585857e-06,
"loss": 0.2329,
"step": 1063
},
{
"epoch": 4.309873417721519,
"grad_norm": 0.3296892908524687,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.2103,
"step": 1064
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.3028617385998846,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.1888,
"step": 1065
},
{
"epoch": 4.317974683544303,
"grad_norm": 0.35387731921276566,
"learning_rate": 2.127347193531757e-06,
"loss": 0.2429,
"step": 1066
},
{
"epoch": 4.322025316455696,
"grad_norm": 0.3193120632382296,
"learning_rate": 2.101945952921942e-06,
"loss": 0.2228,
"step": 1067
},
{
"epoch": 4.326075949367088,
"grad_norm": 0.32795635231208087,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.2054,
"step": 1068
},
{
"epoch": 4.330126582278481,
"grad_norm": 0.30203024962281105,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.2153,
"step": 1069
},
{
"epoch": 4.334177215189873,
"grad_norm": 0.30178030670508565,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.1938,
"step": 1070
},
{
"epoch": 4.3382278481012655,
"grad_norm": 0.3443680372346513,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.235,
"step": 1071
},
{
"epoch": 4.3422784810126585,
"grad_norm": 0.3077384778099141,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.1964,
"step": 1072
},
{
"epoch": 4.3463291139240505,
"grad_norm": 0.3222627169313154,
"learning_rate": 1.95257270594031e-06,
"loss": 0.2212,
"step": 1073
},
{
"epoch": 4.3503797468354435,
"grad_norm": 0.32254864837177,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.2272,
"step": 1074
},
{
"epoch": 4.3544303797468356,
"grad_norm": 0.30962384023319867,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.2084,
"step": 1075
},
{
"epoch": 4.358481012658228,
"grad_norm": 0.31162790933702844,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.2107,
"step": 1076
},
{
"epoch": 4.362531645569621,
"grad_norm": 0.31249022436898166,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.2145,
"step": 1077
},
{
"epoch": 4.366582278481013,
"grad_norm": 0.32810524961546544,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.2394,
"step": 1078
},
{
"epoch": 4.370632911392405,
"grad_norm": 0.3060243840177739,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.1993,
"step": 1079
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.3187363072899373,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.2296,
"step": 1080
},
{
"epoch": 4.37873417721519,
"grad_norm": 0.30048385927476623,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.22,
"step": 1081
},
{
"epoch": 4.382784810126582,
"grad_norm": 0.3026519729663568,
"learning_rate": 1.738336950434061e-06,
"loss": 0.2083,
"step": 1082
},
{
"epoch": 4.386835443037975,
"grad_norm": 0.331725460289037,
"learning_rate": 1.715265841375957e-06,
"loss": 0.2306,
"step": 1083
},
{
"epoch": 4.390886075949367,
"grad_norm": 0.31205640195726114,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.2203,
"step": 1084
},
{
"epoch": 4.39493670886076,
"grad_norm": 0.29863195855711644,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.2096,
"step": 1085
},
{
"epoch": 4.398987341772152,
"grad_norm": 0.3317463688508735,
"learning_rate": 1.646936827927441e-06,
"loss": 0.2193,
"step": 1086
},
{
"epoch": 4.403037974683544,
"grad_norm": 0.3050257604690793,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.1918,
"step": 1087
},
{
"epoch": 4.407088607594937,
"grad_norm": 0.30642437878011497,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.2059,
"step": 1088
},
{
"epoch": 4.411139240506329,
"grad_norm": 0.3181419266915969,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.218,
"step": 1089
},
{
"epoch": 4.415189873417722,
"grad_norm": 0.30290227992623614,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.2164,
"step": 1090
},
{
"epoch": 4.419240506329114,
"grad_norm": 0.31714419397241994,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.2276,
"step": 1091
},
{
"epoch": 4.423291139240506,
"grad_norm": 0.3349460662528124,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.2075,
"step": 1092
},
{
"epoch": 4.427341772151899,
"grad_norm": 0.30674704958260246,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.2109,
"step": 1093
},
{
"epoch": 4.431392405063291,
"grad_norm": 0.3089049173880383,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.2303,
"step": 1094
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.3164443701985928,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.2269,
"step": 1095
},
{
"epoch": 4.439493670886076,
"grad_norm": 0.3005849712627327,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.2142,
"step": 1096
},
{
"epoch": 4.443544303797468,
"grad_norm": 0.312731700989552,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.2365,
"step": 1097
},
{
"epoch": 4.44759493670886,
"grad_norm": 0.3198442551141024,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.216,
"step": 1098
},
{
"epoch": 4.451645569620253,
"grad_norm": 0.31560444483918965,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.2049,
"step": 1099
},
{
"epoch": 4.455696202531645,
"grad_norm": 0.3512434907195414,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.203,
"step": 1100
},
{
"epoch": 4.459746835443038,
"grad_norm": 0.315955348664699,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.2239,
"step": 1101
},
{
"epoch": 4.46379746835443,
"grad_norm": 0.3046231200350712,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.2091,
"step": 1102
},
{
"epoch": 4.467848101265822,
"grad_norm": 0.30415588018832124,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.2321,
"step": 1103
},
{
"epoch": 4.471898734177215,
"grad_norm": 0.30740800068979585,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.2432,
"step": 1104
},
{
"epoch": 4.4759493670886075,
"grad_norm": 0.31269423509548094,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.2161,
"step": 1105
},
{
"epoch": 4.48,
"grad_norm": 0.3093188846304979,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.215,
"step": 1106
},
{
"epoch": 4.4840506329113925,
"grad_norm": 0.30454410307329155,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.203,
"step": 1107
},
{
"epoch": 4.488101265822785,
"grad_norm": 0.32369208400125316,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.2088,
"step": 1108
},
{
"epoch": 4.4921518987341775,
"grad_norm": 0.2985374916940252,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.2145,
"step": 1109
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.31807806952674195,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.2062,
"step": 1110
},
{
"epoch": 4.500253164556962,
"grad_norm": 0.29824733697863914,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.1987,
"step": 1111
},
{
"epoch": 4.504303797468355,
"grad_norm": 0.3176874975342112,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.2122,
"step": 1112
},
{
"epoch": 4.508354430379747,
"grad_norm": 0.32379508692477976,
"learning_rate": 1.092402023876933e-06,
"loss": 0.2479,
"step": 1113
},
{
"epoch": 4.512405063291139,
"grad_norm": 0.296239202254087,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.2146,
"step": 1114
},
{
"epoch": 4.516455696202532,
"grad_norm": 0.2905227165487182,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.195,
"step": 1115
},
{
"epoch": 4.520506329113924,
"grad_norm": 0.4095301732725045,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.2479,
"step": 1116
},
{
"epoch": 4.524556962025317,
"grad_norm": 0.29758579699294296,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.2153,
"step": 1117
},
{
"epoch": 4.528607594936709,
"grad_norm": 0.29973197542107,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.2183,
"step": 1118
},
{
"epoch": 4.532658227848101,
"grad_norm": 0.30840705471903457,
"learning_rate": 9.841383857091947e-07,
"loss": 0.2064,
"step": 1119
},
{
"epoch": 4.536708860759494,
"grad_norm": 0.32281940210431515,
"learning_rate": 9.666296380321616e-07,
"loss": 0.2044,
"step": 1120
},
{
"epoch": 4.540759493670886,
"grad_norm": 0.31452295139869824,
"learning_rate": 9.492741825034124e-07,
"loss": 0.2272,
"step": 1121
},
{
"epoch": 4.544810126582279,
"grad_norm": 0.3003535449353701,
"learning_rate": 9.320721589013892e-07,
"loss": 0.2054,
"step": 1122
},
{
"epoch": 4.548860759493671,
"grad_norm": 0.32502179538325676,
"learning_rate": 9.150237057688339e-07,
"loss": 0.2374,
"step": 1123
},
{
"epoch": 4.552911392405063,
"grad_norm": 0.30136989779659173,
"learning_rate": 8.981289604116328e-07,
"loss": 0.1959,
"step": 1124
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.3236332860226761,
"learning_rate": 8.813880588977542e-07,
"loss": 0.2255,
"step": 1125
},
{
"epoch": 4.561012658227848,
"grad_norm": 0.3218316206546295,
"learning_rate": 8.648011360561126e-07,
"loss": 0.2191,
"step": 1126
},
{
"epoch": 4.56506329113924,
"grad_norm": 0.32429863258352737,
"learning_rate": 8.483683254755037e-07,
"loss": 0.2029,
"step": 1127
},
{
"epoch": 4.569113924050633,
"grad_norm": 0.3106175502727386,
"learning_rate": 8.320897595035227e-07,
"loss": 0.2361,
"step": 1128
},
{
"epoch": 4.573164556962025,
"grad_norm": 0.3186027629811239,
"learning_rate": 8.159655692455093e-07,
"loss": 0.2254,
"step": 1129
},
{
"epoch": 4.577215189873417,
"grad_norm": 0.32481563999989715,
"learning_rate": 7.999958845634648e-07,
"loss": 0.2181,
"step": 1130
},
{
"epoch": 4.58126582278481,
"grad_norm": 0.2925321676248862,
"learning_rate": 7.841808340750478e-07,
"loss": 0.2089,
"step": 1131
},
{
"epoch": 4.585316455696202,
"grad_norm": 0.32475810732346816,
"learning_rate": 7.685205451524869e-07,
"loss": 0.2227,
"step": 1132
},
{
"epoch": 4.589367088607595,
"grad_norm": 0.3156591936809026,
"learning_rate": 7.530151439216027e-07,
"loss": 0.1962,
"step": 1133
},
{
"epoch": 4.593417721518987,
"grad_norm": 0.32105306407476586,
"learning_rate": 7.376647552607675e-07,
"loss": 0.2099,
"step": 1134
},
{
"epoch": 4.597468354430379,
"grad_norm": 0.3309146069428135,
"learning_rate": 7.224695027998963e-07,
"loss": 0.2411,
"step": 1135
},
{
"epoch": 4.601518987341772,
"grad_norm": 0.3072509458091158,
"learning_rate": 7.07429508919466e-07,
"loss": 0.2032,
"step": 1136
},
{
"epoch": 4.605569620253164,
"grad_norm": 0.31552890758978247,
"learning_rate": 6.925448947495206e-07,
"loss": 0.2394,
"step": 1137
},
{
"epoch": 4.609620253164557,
"grad_norm": 0.29671384092404024,
"learning_rate": 6.778157801686936e-07,
"loss": 0.1948,
"step": 1138
},
{
"epoch": 4.613670886075949,
"grad_norm": 0.3255893911955539,
"learning_rate": 6.632422838032515e-07,
"loss": 0.2399,
"step": 1139
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.2868157637679356,
"learning_rate": 6.488245230261281e-07,
"loss": 0.2005,
"step": 1140
},
{
"epoch": 4.6217721518987345,
"grad_norm": 0.3036668777388273,
"learning_rate": 6.345626139559868e-07,
"loss": 0.2105,
"step": 1141
},
{
"epoch": 4.6258227848101265,
"grad_norm": 0.3243942809571357,
"learning_rate": 6.204566714562866e-07,
"loss": 0.2252,
"step": 1142
},
{
"epoch": 4.629873417721519,
"grad_norm": 0.3120845854030891,
"learning_rate": 6.06506809134344e-07,
"loss": 0.2195,
"step": 1143
},
{
"epoch": 4.633924050632912,
"grad_norm": 0.30606307521743187,
"learning_rate": 5.927131393404373e-07,
"loss": 0.2217,
"step": 1144
},
{
"epoch": 4.637974683544304,
"grad_norm": 0.29222076561484794,
"learning_rate": 5.790757731668817e-07,
"loss": 0.195,
"step": 1145
},
{
"epoch": 4.642025316455696,
"grad_norm": 0.33301589951345634,
"learning_rate": 5.655948204471507e-07,
"loss": 0.2194,
"step": 1146
},
{
"epoch": 4.646075949367089,
"grad_norm": 0.29640181762757284,
"learning_rate": 5.522703897549875e-07,
"loss": 0.2017,
"step": 1147
},
{
"epoch": 4.650126582278481,
"grad_norm": 0.3272501877231962,
"learning_rate": 5.391025884035239e-07,
"loss": 0.2304,
"step": 1148
},
{
"epoch": 4.654177215189874,
"grad_norm": 0.287867625051333,
"learning_rate": 5.260915224444207e-07,
"loss": 0.1846,
"step": 1149
},
{
"epoch": 4.658227848101266,
"grad_norm": 0.30128606129046487,
"learning_rate": 5.132372966670129e-07,
"loss": 0.2211,
"step": 1150
},
{
"epoch": 4.662278481012658,
"grad_norm": 0.3038631467942001,
"learning_rate": 5.005400145974704e-07,
"loss": 0.2245,
"step": 1151
},
{
"epoch": 4.666329113924051,
"grad_norm": 0.30601006519280977,
"learning_rate": 4.879997784979562e-07,
"loss": 0.2089,
"step": 1152
},
{
"epoch": 4.670379746835443,
"grad_norm": 0.30594903634918186,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.2363,
"step": 1153
},
{
"epoch": 4.674430379746836,
"grad_norm": 0.2896457780094536,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.203,
"step": 1154
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.3280412048099317,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.2512,
"step": 1155
},
{
"epoch": 4.68253164556962,
"grad_norm": 0.30982969574402247,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.2034,
"step": 1156
},
{
"epoch": 4.686582278481013,
"grad_norm": 0.3076623889801228,
"learning_rate": 4.27657778143431e-07,
"loss": 0.2008,
"step": 1157
},
{
"epoch": 4.690632911392405,
"grad_norm": 0.316917266155974,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.2205,
"step": 1158
},
{
"epoch": 4.694683544303797,
"grad_norm": 0.32644506772444876,
"learning_rate": 4.046237371675177e-07,
"loss": 0.2284,
"step": 1159
},
{
"epoch": 4.69873417721519,
"grad_norm": 0.2938487243902423,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.1949,
"step": 1160
},
{
"epoch": 4.702784810126582,
"grad_norm": 0.31920997945792506,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.2144,
"step": 1161
},
{
"epoch": 4.706835443037974,
"grad_norm": 0.3375476875171742,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.2179,
"step": 1162
},
{
"epoch": 4.710886075949367,
"grad_norm": 0.3262444136550121,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.2352,
"step": 1163
},
{
"epoch": 4.714936708860759,
"grad_norm": 0.28781497351495416,
"learning_rate": 3.498020255137813e-07,
"loss": 0.1888,
"step": 1164
},
{
"epoch": 4.718987341772152,
"grad_norm": 0.3242893618790839,
"learning_rate": 3.393121163057811e-07,
"loss": 0.2419,
"step": 1165
},
{
"epoch": 4.723037974683544,
"grad_norm": 0.5211513553265888,
"learning_rate": 3.289805515751399e-07,
"loss": 0.1987,
"step": 1166
},
{
"epoch": 4.727088607594936,
"grad_norm": 0.3377954795743951,
"learning_rate": 3.188074145308573e-07,
"loss": 0.2586,
"step": 1167
},
{
"epoch": 4.731139240506329,
"grad_norm": 0.29789393215434595,
"learning_rate": 3.087927871059804e-07,
"loss": 0.2104,
"step": 1168
},
{
"epoch": 4.735189873417721,
"grad_norm": 0.29154901616064743,
"learning_rate": 2.989367499569418e-07,
"loss": 0.2081,
"step": 1169
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.3114070419147311,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.2244,
"step": 1170
},
{
"epoch": 4.743291139240506,
"grad_norm": 0.2991746482799073,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.1961,
"step": 1171
},
{
"epoch": 4.7473417721518985,
"grad_norm": 0.3141941838225669,
"learning_rate": 2.703209675663887e-07,
"loss": 0.2145,
"step": 1172
},
{
"epoch": 4.751392405063291,
"grad_norm": 0.3024219325831472,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.1899,
"step": 1173
},
{
"epoch": 4.7554430379746835,
"grad_norm": 0.3406650305944353,
"learning_rate": 2.520381518804471e-07,
"loss": 0.2395,
"step": 1174
},
{
"epoch": 4.759493670886076,
"grad_norm": 0.3038691441802032,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.2002,
"step": 1175
},
{
"epoch": 4.7635443037974685,
"grad_norm": 0.3058184399815397,
"learning_rate": 2.343915243930317e-07,
"loss": 0.2226,
"step": 1176
},
{
"epoch": 4.767594936708861,
"grad_norm": 0.3101814703219188,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.1911,
"step": 1177
},
{
"epoch": 4.7716455696202535,
"grad_norm": 0.3077928143945994,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.2234,
"step": 1178
},
{
"epoch": 4.775696202531646,
"grad_norm": 0.3163320529669286,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.2238,
"step": 1179
},
{
"epoch": 4.779746835443038,
"grad_norm": 0.30279334176332856,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.2104,
"step": 1180
},
{
"epoch": 4.783797468354431,
"grad_norm": 0.294738232068403,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.1954,
"step": 1181
},
{
"epoch": 4.787848101265823,
"grad_norm": 0.3090606776424862,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.2128,
"step": 1182
},
{
"epoch": 4.791898734177215,
"grad_norm": 0.3103337343021391,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.238,
"step": 1183
},
{
"epoch": 4.795949367088608,
"grad_norm": 0.30509335938451254,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.2133,
"step": 1184
},
{
"epoch": 4.8,
"grad_norm": 0.3235935561365773,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.2542,
"step": 1185
},
{
"epoch": 4.804050632911393,
"grad_norm": 0.2792618306182332,
"learning_rate": 1.557203888516745e-07,
"loss": 0.1827,
"step": 1186
},
{
"epoch": 4.808101265822785,
"grad_norm": 0.32201861209839505,
"learning_rate": 1.487313098987131e-07,
"loss": 0.2292,
"step": 1187
},
{
"epoch": 4.812151898734177,
"grad_norm": 0.30043510397809614,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.2072,
"step": 1188
},
{
"epoch": 4.81620253164557,
"grad_norm": 0.29971474287087124,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.2199,
"step": 1189
},
{
"epoch": 4.820253164556962,
"grad_norm": 0.30208453573799016,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.2277,
"step": 1190
},
{
"epoch": 4.824303797468355,
"grad_norm": 0.30962539570564174,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.2107,
"step": 1191
},
{
"epoch": 4.828354430379747,
"grad_norm": 0.30161518902131956,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.214,
"step": 1192
},
{
"epoch": 4.832405063291139,
"grad_norm": 0.29843461518662595,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.2222,
"step": 1193
},
{
"epoch": 4.836455696202532,
"grad_norm": 0.3062160456949203,
"learning_rate": 1.042873700093061e-07,
"loss": 0.2126,
"step": 1194
},
{
"epoch": 4.840506329113924,
"grad_norm": 0.2971756055328466,
"learning_rate": 9.857879404556291e-08,
"loss": 0.2107,
"step": 1195
},
{
"epoch": 4.844556962025316,
"grad_norm": 0.31421646599002073,
"learning_rate": 9.303050139225722e-08,
"loss": 0.2384,
"step": 1196
},
{
"epoch": 4.848607594936709,
"grad_norm": 0.3186125501861975,
"learning_rate": 8.76425367345779e-08,
"loss": 0.2317,
"step": 1197
},
{
"epoch": 4.852658227848101,
"grad_norm": 0.28128393793340084,
"learning_rate": 8.241494346644897e-08,
"loss": 0.1815,
"step": 1198
},
{
"epoch": 4.856708860759493,
"grad_norm": 0.315363607817519,
"learning_rate": 7.734776369019204e-08,
"loss": 0.2146,
"step": 1199
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.3083340161255706,
"learning_rate": 7.244103821617332e-08,
"loss": 0.2246,
"step": 1200
},
{
"epoch": 4.864810126582278,
"grad_norm": 0.29392823845765126,
"learning_rate": 6.769480656248606e-08,
"loss": 0.2041,
"step": 1201
},
{
"epoch": 4.868860759493671,
"grad_norm": 0.31662653839770716,
"learning_rate": 6.310910695462635e-08,
"loss": 0.2231,
"step": 1202
},
{
"epoch": 4.872911392405063,
"grad_norm": 0.3087775434998742,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.2187,
"step": 1203
},
{
"epoch": 4.876962025316455,
"grad_norm": 0.2945980880638434,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.1982,
"step": 1204
},
{
"epoch": 4.881012658227848,
"grad_norm": 0.2959681951214001,
"learning_rate": 5.031556326567488e-08,
"loss": 0.1986,
"step": 1205
},
{
"epoch": 4.88506329113924,
"grad_norm": 0.3170146436786069,
"learning_rate": 4.637234823364312e-08,
"loss": 0.2482,
"step": 1206
},
{
"epoch": 4.889113924050633,
"grad_norm": 0.28166722125923926,
"learning_rate": 4.258983697558838e-08,
"loss": 0.1988,
"step": 1207
},
{
"epoch": 4.8931645569620255,
"grad_norm": 0.30695117737718314,
"learning_rate": 3.896805995533548e-08,
"loss": 0.2209,
"step": 1208
},
{
"epoch": 4.8972151898734175,
"grad_norm": 0.31342315422120115,
"learning_rate": 3.550704634218028e-08,
"loss": 0.2274,
"step": 1209
},
{
"epoch": 4.9012658227848105,
"grad_norm": 0.2928587174542329,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.1793,
"step": 1210
},
{
"epoch": 4.905316455696203,
"grad_norm": 0.3268737142218603,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.2369,
"step": 1211
},
{
"epoch": 4.909367088607595,
"grad_norm": 0.31439857742358757,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.221,
"step": 1212
},
{
"epoch": 4.913417721518988,
"grad_norm": 0.3080328682917125,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.193,
"step": 1213
},
{
"epoch": 4.91746835443038,
"grad_norm": 0.3157766800786009,
"learning_rate": 2.061435966221881e-08,
"loss": 0.2244,
"step": 1214
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.30590614374014174,
"learning_rate": 1.811846652477245e-08,
"loss": 0.2124,
"step": 1215
},
{
"epoch": 4.925569620253165,
"grad_norm": 0.3080481329692172,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.214,
"step": 1216
},
{
"epoch": 4.929620253164557,
"grad_norm": 0.32547855192374375,
"learning_rate": 1.360949303591097e-08,
"loss": 0.2225,
"step": 1217
},
{
"epoch": 4.93367088607595,
"grad_norm": 0.29144313855891457,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.1839,
"step": 1218
},
{
"epoch": 4.937721518987342,
"grad_norm": 0.3241413356352155,
"learning_rate": 9.744388816668172e-09,
"loss": 0.2387,
"step": 1219
},
{
"epoch": 4.941772151898734,
"grad_norm": 0.3047636386656546,
"learning_rate": 8.05332740472009e-09,
"loss": 0.1904,
"step": 1220
},
{
"epoch": 4.945822784810127,
"grad_norm": 0.33615040374648375,
"learning_rate": 6.523278382872811e-09,
"loss": 0.2355,
"step": 1221
},
{
"epoch": 4.949873417721519,
"grad_norm": 0.30400854583970144,
"learning_rate": 5.15425407393133e-09,
"loss": 0.2075,
"step": 1222
},
{
"epoch": 4.953924050632912,
"grad_norm": 0.30204213074638747,
"learning_rate": 3.94626550383137e-09,
"loss": 0.2048,
"step": 1223
},
{
"epoch": 4.957974683544304,
"grad_norm": 0.30741717110770833,
"learning_rate": 2.899322401546112e-09,
"loss": 0.2171,
"step": 1224
},
{
"epoch": 4.962025316455696,
"grad_norm": 0.31684363098239965,
"learning_rate": 2.013433199010706e-09,
"loss": 0.2192,
"step": 1225
},
{
"epoch": 4.966075949367089,
"grad_norm": 0.3047424284042036,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.2051,
"step": 1226
},
{
"epoch": 4.970126582278481,
"grad_norm": 0.3059548313399919,
"learning_rate": 7.248437353468695e-10,
"loss": 0.2047,
"step": 1227
},
{
"epoch": 4.974177215189873,
"grad_norm": 0.3207763195448412,
"learning_rate": 3.221538523412449e-10,
"loss": 0.2156,
"step": 1228
},
{
"epoch": 4.978227848101266,
"grad_norm": 0.3117165570556346,
"learning_rate": 8.053862524670663e-11,
"loss": 0.2291,
"step": 1229
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.30431813193316853,
"learning_rate": 0.0,
"loss": 0.2066,
"step": 1230
},
{
"epoch": 4.982278481012658,
"step": 1230,
"total_flos": 2.334498549389066e+18,
"train_loss": 0.4423948734998703,
"train_runtime": 115891.4291,
"train_samples_per_second": 1.363,
"train_steps_per_second": 0.011
}
],
"logging_steps": 1.0,
"max_steps": 1230,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.334498549389066e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}