c1_math_0d_4s / trainer_state.json
neginr's picture
End of training
7b361a3 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": 6.223418672866073,
"learning_rate": 3.2520325203252037e-07,
"loss": 0.7952,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 6.128636745622828,
"learning_rate": 6.504065040650407e-07,
"loss": 0.805,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 6.381782181984575,
"learning_rate": 9.75609756097561e-07,
"loss": 0.8216,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 6.333793191403608,
"learning_rate": 1.3008130081300815e-06,
"loss": 0.8311,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 5.940199777698115,
"learning_rate": 1.6260162601626018e-06,
"loss": 0.8264,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 5.639947155869954,
"learning_rate": 1.951219512195122e-06,
"loss": 0.8055,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 4.355739652986022,
"learning_rate": 2.2764227642276426e-06,
"loss": 0.7669,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 4.055017409279571,
"learning_rate": 2.601626016260163e-06,
"loss": 0.769,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 2.379624699173893,
"learning_rate": 2.926829268292683e-06,
"loss": 0.7289,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 2.0295731475512255,
"learning_rate": 3.2520325203252037e-06,
"loss": 0.7334,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 1.8597919972176626,
"learning_rate": 3.577235772357724e-06,
"loss": 0.7132,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 3.826790123592512,
"learning_rate": 3.902439024390244e-06,
"loss": 0.7294,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 4.176795523142768,
"learning_rate": 4.227642276422765e-06,
"loss": 0.7289,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 4.198101537801199,
"learning_rate": 4.552845528455285e-06,
"loss": 0.7317,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 4.009055495528828,
"learning_rate": 4.8780487804878055e-06,
"loss": 0.7073,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 2.967502115047676,
"learning_rate": 5.203252032520326e-06,
"loss": 0.6516,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 2.7566856813310774,
"learning_rate": 5.528455284552846e-06,
"loss": 0.6507,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.13746557006516,
"learning_rate": 5.853658536585366e-06,
"loss": 0.645,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 1.4286721470996442,
"learning_rate": 6.178861788617887e-06,
"loss": 0.641,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.2316415606577473,
"learning_rate": 6.504065040650407e-06,
"loss": 0.6203,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.2622286877141706,
"learning_rate": 6.829268292682928e-06,
"loss": 0.598,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.2488255229842604,
"learning_rate": 7.154471544715448e-06,
"loss": 0.6087,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.0464706811796047,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.5993,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 0.8077209909463132,
"learning_rate": 7.804878048780489e-06,
"loss": 0.6044,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 0.9965421444681016,
"learning_rate": 8.130081300813009e-06,
"loss": 0.5962,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 0.8131751716977382,
"learning_rate": 8.45528455284553e-06,
"loss": 0.5652,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 0.6819837261276998,
"learning_rate": 8.78048780487805e-06,
"loss": 0.563,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 0.7579242315730137,
"learning_rate": 9.10569105691057e-06,
"loss": 0.5572,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 0.8131944386812923,
"learning_rate": 9.43089430894309e-06,
"loss": 0.5561,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 0.5317610763859923,
"learning_rate": 9.756097560975611e-06,
"loss": 0.5731,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 0.598354107538634,
"learning_rate": 1.008130081300813e-05,
"loss": 0.557,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.7971980189438191,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.5565,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.5570460979355879,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.5552,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.49706562730117265,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.529,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.6053606754109597,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.5409,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.593758847772035,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.541,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.5055112825984873,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.5373,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.5746444661051701,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.5373,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.5290277731948525,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.5343,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.44429740529524764,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.5271,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.6196118248375738,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.5219,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.48346762876889804,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.525,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.46843273361313625,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.5041,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.550301065116405,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.5217,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.5040285347290205,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.5167,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.5022787584873335,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.5086,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.47902680966491684,
"learning_rate": 1.528455284552846e-05,
"loss": 0.5211,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.47653621090876147,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.5103,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.6017267676680725,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.5126,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.4360417608698551,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.5165,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.5070252230087582,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.4983,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.5139126020676974,
"learning_rate": 1.691056910569106e-05,
"loss": 0.4916,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.478537568730014,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.5065,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.5247684437205558,
"learning_rate": 1.75609756097561e-05,
"loss": 0.5073,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.5273474882875584,
"learning_rate": 1.788617886178862e-05,
"loss": 0.5012,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.4869053986495145,
"learning_rate": 1.821138211382114e-05,
"loss": 0.4825,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.3980873551543919,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.4965,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.4715580251717868,
"learning_rate": 1.886178861788618e-05,
"loss": 0.4972,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.5097877769978754,
"learning_rate": 1.91869918699187e-05,
"loss": 0.5102,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.6306398469646315,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.5039,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.5278687871041826,
"learning_rate": 1.983739837398374e-05,
"loss": 0.4793,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.7070842562118946,
"learning_rate": 2.016260162601626e-05,
"loss": 0.4987,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.7034200655946151,
"learning_rate": 2.048780487804878e-05,
"loss": 0.4897,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.7295506055079812,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.4669,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.5195942973788559,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.4801,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.553076619136969,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.4976,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.8655568473687507,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.4848,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 0.7832844448738286,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.4933,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 0.580870046535547,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.4813,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 0.8897341056819869,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.4897,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 0.7856024657116195,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.4911,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.7383529201792621,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.4779,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 1.1223565167218514,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.4865,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.7250532995301252,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.4918,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.9340337515701926,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.4872,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 1.0152456192362733,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.4707,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.7723216016232725,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.4961,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 0.9833880287779256,
"learning_rate": 2.536585365853659e-05,
"loss": 0.4775,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 0.9910684906520463,
"learning_rate": 2.569105691056911e-05,
"loss": 0.4693,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 1.2132851506029414,
"learning_rate": 2.601626016260163e-05,
"loss": 0.4898,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 0.7940255534328088,
"learning_rate": 2.634146341463415e-05,
"loss": 0.4785,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 0.7900763327446113,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.4862,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 1.3389349483981352,
"learning_rate": 2.699186991869919e-05,
"loss": 0.4741,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 0.6818819657287741,
"learning_rate": 2.731707317073171e-05,
"loss": 0.4785,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 1.1538661273427229,
"learning_rate": 2.764227642276423e-05,
"loss": 0.4702,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 0.8137393291098698,
"learning_rate": 2.796747967479675e-05,
"loss": 0.4768,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 0.9600936847791747,
"learning_rate": 2.829268292682927e-05,
"loss": 0.4781,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 0.7487657237892883,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.4754,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 0.9652780196560844,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.4757,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 0.8638725454452297,
"learning_rate": 2.926829268292683e-05,
"loss": 0.4714,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 0.858987028208688,
"learning_rate": 2.959349593495935e-05,
"loss": 0.4677,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 0.8558142977139351,
"learning_rate": 2.991869918699187e-05,
"loss": 0.4745,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 0.602311990201842,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.472,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 0.6488431164509836,
"learning_rate": 3.056910569105692e-05,
"loss": 0.4594,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 0.5685869578870032,
"learning_rate": 3.089430894308943e-05,
"loss": 0.467,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 0.5052213747926467,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.4746,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 0.5847962453443208,
"learning_rate": 3.154471544715447e-05,
"loss": 0.4794,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 0.6852063002673963,
"learning_rate": 3.186991869918699e-05,
"loss": 0.4879,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 0.683820650663953,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.4728,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.5354160876513833,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.4802,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 0.46746018518131666,
"learning_rate": 3.284552845528456e-05,
"loss": 0.4724,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 0.5586308343315202,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.4803,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 0.5277410957571124,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.475,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 0.5661426733285733,
"learning_rate": 3.382113821138212e-05,
"loss": 0.4799,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 0.483286877937077,
"learning_rate": 3.414634146341463e-05,
"loss": 0.4664,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 0.6084097469992477,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.4563,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 0.7192828404849808,
"learning_rate": 3.479674796747968e-05,
"loss": 0.4664,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 0.913193851529263,
"learning_rate": 3.51219512195122e-05,
"loss": 0.477,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 1.0055790130483067,
"learning_rate": 3.544715447154472e-05,
"loss": 0.4705,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 1.0850985920120886,
"learning_rate": 3.577235772357724e-05,
"loss": 0.4663,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 1.192250967266718,
"learning_rate": 3.609756097560976e-05,
"loss": 0.4723,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 0.7645574058531402,
"learning_rate": 3.642276422764228e-05,
"loss": 0.4635,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 1.2052252900402882,
"learning_rate": 3.67479674796748e-05,
"loss": 0.4611,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 0.7123763954364702,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.4754,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 1.1464139160405105,
"learning_rate": 3.739837398373984e-05,
"loss": 0.4666,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 0.895670612391412,
"learning_rate": 3.772357723577236e-05,
"loss": 0.4606,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 0.6994083815281071,
"learning_rate": 3.804878048780488e-05,
"loss": 0.4407,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 1.02624814321363,
"learning_rate": 3.83739837398374e-05,
"loss": 0.4681,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 0.8779114535786657,
"learning_rate": 3.869918699186992e-05,
"loss": 0.4626,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 0.7119656179965665,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.4766,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 0.9186894751185072,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.473,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 0.7526502750997665,
"learning_rate": 3.967479674796748e-05,
"loss": 0.4607,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 0.5832259667106181,
"learning_rate": 4e-05,
"loss": 0.4541,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 0.6649956412357945,
"learning_rate": 3.999991946137476e-05,
"loss": 0.4604,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.5868954810039616,
"learning_rate": 3.999967784614766e-05,
"loss": 0.4578,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 0.6211593960380289,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.4671,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 0.6691122368427926,
"learning_rate": 3.999871139496895e-05,
"loss": 0.4677,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 0.6293467372005882,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.466,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 0.8680649498438304,
"learning_rate": 3.999710067759846e-05,
"loss": 0.4775,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 0.7112518417684943,
"learning_rate": 3.999605373449617e-05,
"loss": 0.4597,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.6465317454100328,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.4544,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 0.5645672653363097,
"learning_rate": 3.999347672161713e-05,
"loss": 0.4619,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 0.6171316008620886,
"learning_rate": 3.999194667259528e-05,
"loss": 0.4565,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 0.6183255442424148,
"learning_rate": 3.999025561118334e-05,
"loss": 0.4613,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 0.5340577024708179,
"learning_rate": 3.998840355100086e-05,
"loss": 0.4524,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 0.5078308730998266,
"learning_rate": 3.998639050696409e-05,
"loss": 0.4489,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 0.5743154702674992,
"learning_rate": 3.998421649528582e-05,
"loss": 0.4682,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 0.51988824203462,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.4557,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.5288289887449817,
"learning_rate": 3.997938564033779e-05,
"loss": 0.4476,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 0.5802964287399642,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.4528,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.5058823421388833,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.4536,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 0.5926808630042381,
"learning_rate": 3.997093258045973e-05,
"loss": 0.4787,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.6021221817722917,
"learning_rate": 3.996779317598936e-05,
"loss": 0.437,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 0.5975643735288259,
"learning_rate": 3.996449295365782e-05,
"loss": 0.4731,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.5723587548312363,
"learning_rate": 3.996103194004467e-05,
"loss": 0.446,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.55211995117801,
"learning_rate": 3.995741016302441e-05,
"loss": 0.4547,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 0.5914758081757564,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.4538,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 0.6510435920651073,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.455,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 0.7677193548405306,
"learning_rate": 3.994558054968643e-05,
"loss": 0.4471,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.6917634470591633,
"learning_rate": 3.994131602367481e-05,
"loss": 0.4587,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 0.6619635460287344,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.4506,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.5933117726878424,
"learning_rate": 3.993230519343752e-05,
"loss": 0.4726,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 0.7453582812772214,
"learning_rate": 3.992755896178383e-05,
"loss": 0.4448,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.7928651238464575,
"learning_rate": 3.992265223630981e-05,
"loss": 0.463,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 0.5628030835263006,
"learning_rate": 3.991758505653355e-05,
"loss": 0.4519,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 0.6511234777146152,
"learning_rate": 3.991235746326543e-05,
"loss": 0.4493,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.8297515461260662,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.464,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 0.6858414932821196,
"learning_rate": 3.990142120595444e-05,
"loss": 0.4576,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 0.6593655275657745,
"learning_rate": 3.98957126299907e-05,
"loss": 0.4636,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 0.6988327954001843,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.4525,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 0.4973030573729136,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.4328,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 0.5660602969382748,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.4534,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 0.6659923397973627,
"learning_rate": 3.98712764319087e-05,
"loss": 0.4501,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.593826797314028,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.4642,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 0.745853551334874,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.4618,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.7211412864891876,
"learning_rate": 3.985126869010129e-05,
"loss": 0.44,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 0.6038537926157606,
"learning_rate": 3.984427961114833e-05,
"loss": 0.4503,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 0.5997035708863198,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.4592,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 0.6613459559167731,
"learning_rate": 3.982982204151901e-05,
"loss": 0.4412,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.5734871452867881,
"learning_rate": 3.982235366728193e-05,
"loss": 0.4473,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.5031094530134188,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.4503,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 0.6184850840325999,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.4521,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.5625887325553036,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.4581,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.5494045626917026,
"learning_rate": 3.979088432599627e-05,
"loss": 0.4524,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.47461636080591574,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.4349,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 0.45781940431709806,
"learning_rate": 3.977419304032111e-05,
"loss": 0.4656,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 11.383195845614946,
"learning_rate": 3.976560847560697e-05,
"loss": 0.4506,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 1.1536372812844062,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.4607,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 1.1354471862767261,
"learning_rate": 3.974796184811956e-05,
"loss": 0.4561,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.7265601221730722,
"learning_rate": 3.973889992746979e-05,
"loss": 0.4499,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.890861552996204,
"learning_rate": 3.972967903243361e-05,
"loss": 0.4437,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.6582368767421704,
"learning_rate": 3.972029923727486e-05,
"loss": 0.4366,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.8369010829291688,
"learning_rate": 3.971076061753709e-05,
"loss": 0.4497,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.5868556020763777,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.4553,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.7511617614081706,
"learning_rate": 3.969120721289402e-05,
"loss": 0.4523,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.5882938205864616,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.4504,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.6313324847870809,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.4603,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.6767852523834864,
"learning_rate": 3.966068788369422e-05,
"loss": 0.447,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.6574716323911354,
"learning_rate": 3.965019797448622e-05,
"loss": 0.4531,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.5560857127473755,
"learning_rate": 3.963954980528515e-05,
"loss": 0.4522,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.6919488549440187,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.4566,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.5892162706154054,
"learning_rate": 3.961777903121329e-05,
"loss": 0.4411,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.6440384538480937,
"learning_rate": 3.960665660168131e-05,
"loss": 0.4422,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.4414711850794956,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.4322,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.6142825741737376,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.4426,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 0.5471663571697983,
"learning_rate": 3.957234222185657e-05,
"loss": 0.4491,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.5777294689543744,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.4502,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 0.5640075454445656,
"learning_rate": 3.954867765033605e-05,
"loss": 0.4453,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.5272674459704945,
"learning_rate": 3.953660915306728e-05,
"loss": 0.4472,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.48467440856991045,
"learning_rate": 3.952438331063419e-05,
"loss": 0.4408,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.455974860643311,
"learning_rate": 3.951200022150205e-05,
"loss": 0.4544,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.43711519447815833,
"learning_rate": 3.949945998540253e-05,
"loss": 0.4399,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.4356342057111041,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.4506,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.4757796014335013,
"learning_rate": 3.947390847755559e-05,
"loss": 0.4456,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.48248085360206466,
"learning_rate": 3.946089741159648e-05,
"loss": 0.4361,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.4268999109576842,
"learning_rate": 3.944772961024501e-05,
"loss": 0.4419,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.4787757148776982,
"learning_rate": 3.943440517955285e-05,
"loss": 0.456,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.4034235002390413,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.4347,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.48044416740866625,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.4524,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.4968481428921911,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.451,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.48131990454112633,
"learning_rate": 3.937954332854371e-05,
"loss": 0.4428,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.4108841607246741,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.4558,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.571110036938809,
"learning_rate": 3.935117547697387e-05,
"loss": 0.4411,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.5433482238143006,
"learning_rate": 3.933675771619675e-05,
"loss": 0.4428,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.42235180792642496,
"learning_rate": 3.932218421983131e-05,
"loss": 0.4415,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.44515209925758426,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.4407,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.43241212897213205,
"learning_rate": 3.929257049108054e-05,
"loss": 0.4417,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.5077974716244278,
"learning_rate": 3.927753049720011e-05,
"loss": 0.4329,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.34642768605891094,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.4273,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.6054885884500858,
"learning_rate": 3.92469848560784e-05,
"loss": 0.4543,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.4505417953990085,
"learning_rate": 3.923147945484751e-05,
"loss": 0.437,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.6871849063272694,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.4528,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.4749973373307726,
"learning_rate": 3.920000411543654e-05,
"loss": 0.4336,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.4778425632851313,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.4476,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.46315932148500616,
"learning_rate": 3.916791024049648e-05,
"loss": 0.4357,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.35869815880517325,
"learning_rate": 3.91516316745245e-05,
"loss": 0.4387,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.41578015825204273,
"learning_rate": 3.913519886394389e-05,
"loss": 0.4581,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.4474776413057451,
"learning_rate": 3.911861194110225e-05,
"loss": 0.445,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.3488649743172224,
"learning_rate": 3.910187103958837e-05,
"loss": 0.4485,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.3782507732874626,
"learning_rate": 3.908497629423117e-05,
"loss": 0.4566,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.3677300163685344,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.4364,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.3161644151484949,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.4319,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.36538613312056734,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.4308,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.3120787343695632,
"learning_rate": 3.901586161429081e-05,
"loss": 0.4227,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.41218918595617626,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.4248,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.4003140656050267,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.4411,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.42505780715310404,
"learning_rate": 3.896241703466667e-05,
"loss": 0.4237,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.4133508889446291,
"learning_rate": 3.894429654085585e-05,
"loss": 0.4293,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.5282142547511397,
"learning_rate": 3.892602347228505e-05,
"loss": 0.4232,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.4284201634857587,
"learning_rate": 3.890759797612307e-05,
"loss": 0.4466,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.5163522009140149,
"learning_rate": 3.888902020076632e-05,
"loss": 0.423,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.3838548602678084,
"learning_rate": 3.887029029583764e-05,
"loss": 0.4438,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.4359269860259843,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.4481,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.5009651728491928,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.4411,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.4306600590853253,
"learning_rate": 3.881318931821972e-05,
"loss": 0.4449,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.35706285193160986,
"learning_rate": 3.879385241571817e-05,
"loss": 0.4346,
"step": 246
},
{
"epoch": 1.000506329113924,
"grad_norm": 0.4690723442288527,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.4394,
"step": 247
},
{
"epoch": 1.0045569620253165,
"grad_norm": 0.44126918559790757,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.4087,
"step": 248
},
{
"epoch": 1.0086075949367088,
"grad_norm": 0.3940311701091894,
"learning_rate": 3.873493415863256e-05,
"loss": 0.403,
"step": 249
},
{
"epoch": 1.0126582278481013,
"grad_norm": 0.40899083183722224,
"learning_rate": 3.871499275032111e-05,
"loss": 0.4101,
"step": 250
},
{
"epoch": 1.0167088607594936,
"grad_norm": 0.48428225669283315,
"learning_rate": 3.869490061403091e-05,
"loss": 0.4175,
"step": 251
},
{
"epoch": 1.0207594936708861,
"grad_norm": 0.6842109223776309,
"learning_rate": 3.867465791158124e-05,
"loss": 0.4113,
"step": 252
},
{
"epoch": 1.0248101265822784,
"grad_norm": 0.6092063900917932,
"learning_rate": 3.865426480600407e-05,
"loss": 0.3968,
"step": 253
},
{
"epoch": 1.028860759493671,
"grad_norm": 0.47048760222394076,
"learning_rate": 3.863372146154264e-05,
"loss": 0.4049,
"step": 254
},
{
"epoch": 1.0329113924050632,
"grad_norm": 0.44889573360405743,
"learning_rate": 3.861302804365024e-05,
"loss": 0.3997,
"step": 255
},
{
"epoch": 1.0369620253164558,
"grad_norm": 0.4572056927109752,
"learning_rate": 3.85921847189888e-05,
"loss": 0.4129,
"step": 256
},
{
"epoch": 1.041012658227848,
"grad_norm": 0.4804845139382862,
"learning_rate": 3.85711916554276e-05,
"loss": 0.4077,
"step": 257
},
{
"epoch": 1.0450632911392406,
"grad_norm": 0.4587714587346259,
"learning_rate": 3.85500490220419e-05,
"loss": 0.3922,
"step": 258
},
{
"epoch": 1.0491139240506329,
"grad_norm": 0.42720728757088633,
"learning_rate": 3.852875698911154e-05,
"loss": 0.3992,
"step": 259
},
{
"epoch": 1.0531645569620254,
"grad_norm": 0.5202526701075391,
"learning_rate": 3.850731572811963e-05,
"loss": 0.3949,
"step": 260
},
{
"epoch": 1.0572151898734177,
"grad_norm": 0.4645354651812409,
"learning_rate": 3.848572541175116e-05,
"loss": 0.3966,
"step": 261
},
{
"epoch": 1.0612658227848102,
"grad_norm": 0.4729185996830374,
"learning_rate": 3.846398621389154e-05,
"loss": 0.3993,
"step": 262
},
{
"epoch": 1.0653164556962025,
"grad_norm": 0.4332817672035301,
"learning_rate": 3.84420983096253e-05,
"loss": 0.3969,
"step": 263
},
{
"epoch": 1.069367088607595,
"grad_norm": 0.5010776989343766,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.3984,
"step": 264
},
{
"epoch": 1.0734177215189873,
"grad_norm": 0.4862888783744572,
"learning_rate": 3.839787708819787e-05,
"loss": 0.4045,
"step": 265
},
{
"epoch": 1.0774683544303798,
"grad_norm": 0.41840794412152604,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.3991,
"step": 266
},
{
"epoch": 1.081518987341772,
"grad_norm": 0.44955840610844217,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.3869,
"step": 267
},
{
"epoch": 1.0855696202531646,
"grad_norm": 0.39197041026975504,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.3919,
"step": 268
},
{
"epoch": 1.089620253164557,
"grad_norm": 0.4370353684706128,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.3985,
"step": 269
},
{
"epoch": 1.0936708860759494,
"grad_norm": 0.3500409582837721,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.3853,
"step": 270
},
{
"epoch": 1.0977215189873417,
"grad_norm": 0.39639554514445796,
"learning_rate": 3.826166304956594e-05,
"loss": 0.4044,
"step": 271
},
{
"epoch": 1.1017721518987342,
"grad_norm": 0.4474957742396197,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.387,
"step": 272
},
{
"epoch": 1.1058227848101265,
"grad_norm": 0.3668321220212158,
"learning_rate": 3.821507978767479e-05,
"loss": 0.3885,
"step": 273
},
{
"epoch": 1.109873417721519,
"grad_norm": 0.4142317797743725,
"learning_rate": 3.819156801001693e-05,
"loss": 0.3932,
"step": 274
},
{
"epoch": 1.1139240506329113,
"grad_norm": 0.408939103187847,
"learning_rate": 3.816790971997121e-05,
"loss": 0.3809,
"step": 275
},
{
"epoch": 1.1179746835443038,
"grad_norm": 0.35698923050469017,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.4027,
"step": 276
},
{
"epoch": 1.1220253164556961,
"grad_norm": 0.41013906621741014,
"learning_rate": 3.81201543660571e-05,
"loss": 0.403,
"step": 277
},
{
"epoch": 1.1260759493670887,
"grad_norm": 0.5114114192012433,
"learning_rate": 3.809605768680377e-05,
"loss": 0.4025,
"step": 278
},
{
"epoch": 1.130126582278481,
"grad_norm": 0.4821397383736097,
"learning_rate": 3.807181526438958e-05,
"loss": 0.3982,
"step": 279
},
{
"epoch": 1.1341772151898735,
"grad_norm": 0.45266673337663926,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.403,
"step": 280
},
{
"epoch": 1.1382278481012658,
"grad_norm": 0.423047134891627,
"learning_rate": 3.802289397223145e-05,
"loss": 0.395,
"step": 281
},
{
"epoch": 1.1422784810126583,
"grad_norm": 0.474817333075142,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.4051,
"step": 282
},
{
"epoch": 1.1463291139240506,
"grad_norm": 0.42288598123306287,
"learning_rate": 3.797339206560096e-05,
"loss": 0.4166,
"step": 283
},
{
"epoch": 1.150379746835443,
"grad_norm": 0.3952372169956468,
"learning_rate": 3.794842387948027e-05,
"loss": 0.3983,
"step": 284
},
{
"epoch": 1.1544303797468354,
"grad_norm": 0.35525325729457635,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.3965,
"step": 285
},
{
"epoch": 1.158481012658228,
"grad_norm": 0.3540408071135403,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.4161,
"step": 286
},
{
"epoch": 1.1625316455696202,
"grad_norm": 0.4090826525468005,
"learning_rate": 3.787265280646825e-05,
"loss": 0.4087,
"step": 287
},
{
"epoch": 1.1665822784810127,
"grad_norm": 0.367409046237202,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.3929,
"step": 288
},
{
"epoch": 1.170632911392405,
"grad_norm": 0.27785394171235134,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.3874,
"step": 289
},
{
"epoch": 1.1746835443037975,
"grad_norm": 0.42424688528308846,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.4032,
"step": 290
},
{
"epoch": 1.1787341772151898,
"grad_norm": 0.46843705175738687,
"learning_rate": 3.776961046830306e-05,
"loss": 0.3986,
"step": 291
},
{
"epoch": 1.1827848101265823,
"grad_norm": 0.36544237450531264,
"learning_rate": 3.774349157719215e-05,
"loss": 0.3936,
"step": 292
},
{
"epoch": 1.1868354430379746,
"grad_norm": 0.40516472419045385,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.3978,
"step": 293
},
{
"epoch": 1.1908860759493671,
"grad_norm": 0.39269882843917125,
"learning_rate": 3.769082529555359e-05,
"loss": 0.4091,
"step": 294
},
{
"epoch": 1.1949367088607594,
"grad_norm": 0.3807387835193761,
"learning_rate": 3.766427832919294e-05,
"loss": 0.4038,
"step": 295
},
{
"epoch": 1.198987341772152,
"grad_norm": 0.36391747163288907,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.4082,
"step": 296
},
{
"epoch": 1.2030379746835442,
"grad_norm": 0.449898145617493,
"learning_rate": 3.761075781441526e-05,
"loss": 0.3967,
"step": 297
},
{
"epoch": 1.2070886075949367,
"grad_norm": 0.40669918922631837,
"learning_rate": 3.75837846970451e-05,
"loss": 0.3956,
"step": 298
},
{
"epoch": 1.211139240506329,
"grad_norm": 0.3893996707353956,
"learning_rate": 3.755666996229032e-05,
"loss": 0.3998,
"step": 299
},
{
"epoch": 1.2151898734177216,
"grad_norm": 0.4221714487857021,
"learning_rate": 3.752941382852927e-05,
"loss": 0.413,
"step": 300
},
{
"epoch": 1.2192405063291138,
"grad_norm": 0.341880711834028,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.3993,
"step": 301
},
{
"epoch": 1.2232911392405064,
"grad_norm": 0.40483846913195337,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.3918,
"step": 302
},
{
"epoch": 1.2273417721518987,
"grad_norm": 0.5058342730403446,
"learning_rate": 3.744679923406351e-05,
"loss": 0.4015,
"step": 303
},
{
"epoch": 1.2313924050632912,
"grad_norm": 0.6224147333746857,
"learning_rate": 3.741897971081043e-05,
"loss": 0.4088,
"step": 304
},
{
"epoch": 1.2354430379746835,
"grad_norm": 0.4825808760537386,
"learning_rate": 3.739101989748946e-05,
"loss": 0.3943,
"step": 305
},
{
"epoch": 1.239493670886076,
"grad_norm": 0.3987798368654969,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.3935,
"step": 306
},
{
"epoch": 1.2435443037974683,
"grad_norm": 0.5050954892567086,
"learning_rate": 3.73346803025098e-05,
"loss": 0.3919,
"step": 307
},
{
"epoch": 1.2475949367088608,
"grad_norm": 0.49374073448497385,
"learning_rate": 3.730630097460247e-05,
"loss": 0.3941,
"step": 308
},
{
"epoch": 1.251645569620253,
"grad_norm": 0.42509728651382395,
"learning_rate": 3.727778226412628e-05,
"loss": 0.4049,
"step": 309
},
{
"epoch": 1.2556962025316456,
"grad_norm": 0.5076290627750819,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.4005,
"step": 310
},
{
"epoch": 1.259746835443038,
"grad_norm": 0.4406873745474715,
"learning_rate": 3.722032761533114e-05,
"loss": 0.4021,
"step": 311
},
{
"epoch": 1.2637974683544304,
"grad_norm": 0.4042205628259085,
"learning_rate": 3.719139213974403e-05,
"loss": 0.4045,
"step": 312
},
{
"epoch": 1.2678481012658227,
"grad_norm": 0.42353718430912113,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.3939,
"step": 313
},
{
"epoch": 1.2718987341772152,
"grad_norm": 0.5000898743100904,
"learning_rate": 3.713310605140055e-05,
"loss": 0.4167,
"step": 314
},
{
"epoch": 1.2759493670886077,
"grad_norm": 0.43338182857801955,
"learning_rate": 3.710375590807233e-05,
"loss": 0.3922,
"step": 315
},
{
"epoch": 1.28,
"grad_norm": 0.4047284872918609,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.4082,
"step": 316
},
{
"epoch": 1.2840506329113923,
"grad_norm": 0.4082444448347118,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.3955,
"step": 317
},
{
"epoch": 1.2881012658227848,
"grad_norm": 0.3880994107898263,
"learning_rate": 3.701487992136854e-05,
"loss": 0.4121,
"step": 318
},
{
"epoch": 1.2921518987341774,
"grad_norm": 0.4148633109610314,
"learning_rate": 3.69849802022222e-05,
"loss": 0.3999,
"step": 319
},
{
"epoch": 1.2962025316455696,
"grad_norm": 0.35790752853624797,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.394,
"step": 320
},
{
"epoch": 1.300253164556962,
"grad_norm": 0.3736138081689745,
"learning_rate": 3.692477062175289e-05,
"loss": 0.3826,
"step": 321
},
{
"epoch": 1.3043037974683545,
"grad_norm": 0.4075415126130377,
"learning_rate": 3.689446124534958e-05,
"loss": 0.3898,
"step": 322
},
{
"epoch": 1.308354430379747,
"grad_norm": 0.4059318631092302,
"learning_rate": 3.686401580327799e-05,
"loss": 0.3991,
"step": 323
},
{
"epoch": 1.3124050632911393,
"grad_norm": 0.3270349973671135,
"learning_rate": 3.683343454074149e-05,
"loss": 0.3843,
"step": 324
},
{
"epoch": 1.3164556962025316,
"grad_norm": 0.38506152653419684,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.3919,
"step": 325
},
{
"epoch": 1.320506329113924,
"grad_norm": 0.4299200525362855,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.3988,
"step": 326
},
{
"epoch": 1.3245569620253166,
"grad_norm": 0.37557633243638,
"learning_rate": 3.674087829877297e-05,
"loss": 0.3913,
"step": 327
},
{
"epoch": 1.3286075949367089,
"grad_norm": 0.45840162019032515,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.3915,
"step": 328
},
{
"epoch": 1.3326582278481012,
"grad_norm": 0.3206537273734398,
"learning_rate": 3.667849957966501e-05,
"loss": 0.4014,
"step": 329
},
{
"epoch": 1.3367088607594937,
"grad_norm": 0.48113240745524316,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.39,
"step": 330
},
{
"epoch": 1.3407594936708862,
"grad_norm": 0.36820319355321296,
"learning_rate": 3.661558355626795e-05,
"loss": 0.4049,
"step": 331
},
{
"epoch": 1.3448101265822785,
"grad_norm": 0.3626937896050084,
"learning_rate": 3.658392468818163e-05,
"loss": 0.3929,
"step": 332
},
{
"epoch": 1.3488607594936708,
"grad_norm": 0.3783831202108202,
"learning_rate": 3.655213225544574e-05,
"loss": 0.4075,
"step": 333
},
{
"epoch": 1.3529113924050633,
"grad_norm": 0.4782607716243734,
"learning_rate": 3.652020651411218e-05,
"loss": 0.3873,
"step": 334
},
{
"epoch": 1.3569620253164558,
"grad_norm": 0.39108156116414317,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.3964,
"step": 335
},
{
"epoch": 1.3610126582278481,
"grad_norm": 0.3957956529232561,
"learning_rate": 3.645595613522574e-05,
"loss": 0.4077,
"step": 336
},
{
"epoch": 1.3650632911392404,
"grad_norm": 0.3939619041447444,
"learning_rate": 3.642363201513657e-05,
"loss": 0.3858,
"step": 337
},
{
"epoch": 1.369113924050633,
"grad_norm": 0.4111246025161492,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.3951,
"step": 338
},
{
"epoch": 1.3731645569620254,
"grad_norm": 0.3399241380661189,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.3951,
"step": 339
},
{
"epoch": 1.3772151898734177,
"grad_norm": 0.5009344015913203,
"learning_rate": 3.632586705948318e-05,
"loss": 0.3971,
"step": 340
},
{
"epoch": 1.38126582278481,
"grad_norm": 0.35235285122175286,
"learning_rate": 3.629301541734311e-05,
"loss": 0.4004,
"step": 341
},
{
"epoch": 1.3853164556962025,
"grad_norm": 0.3183860324277687,
"learning_rate": 3.626003255349676e-05,
"loss": 0.3881,
"step": 342
},
{
"epoch": 1.389367088607595,
"grad_norm": 0.3522435241537998,
"learning_rate": 3.622691873358357e-05,
"loss": 0.4029,
"step": 343
},
{
"epoch": 1.3934177215189874,
"grad_norm": 0.420113351815351,
"learning_rate": 3.61936742242977e-05,
"loss": 0.3892,
"step": 344
},
{
"epoch": 1.3974683544303796,
"grad_norm": 0.3505813707026184,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.4,
"step": 345
},
{
"epoch": 1.4015189873417722,
"grad_norm": 0.3147016299867735,
"learning_rate": 3.612679420964516e-05,
"loss": 0.3988,
"step": 346
},
{
"epoch": 1.4055696202531647,
"grad_norm": 0.3525611197431202,
"learning_rate": 3.609315924292092e-05,
"loss": 0.397,
"step": 347
},
{
"epoch": 1.409620253164557,
"grad_norm": 0.35733274189149417,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.3962,
"step": 348
},
{
"epoch": 1.4136708860759493,
"grad_norm": 0.39030307089129024,
"learning_rate": 3.602550074513133e-05,
"loss": 0.4026,
"step": 349
},
{
"epoch": 1.4177215189873418,
"grad_norm": 0.37139738113564735,
"learning_rate": 3.599147775897822e-05,
"loss": 0.4022,
"step": 350
},
{
"epoch": 1.4217721518987343,
"grad_norm": 0.44853245867955754,
"learning_rate": 3.595732597966167e-05,
"loss": 0.4104,
"step": 351
},
{
"epoch": 1.4258227848101266,
"grad_norm": 0.385173925772124,
"learning_rate": 3.592304568223542e-05,
"loss": 0.3974,
"step": 352
},
{
"epoch": 1.4298734177215189,
"grad_norm": 0.36880719961449954,
"learning_rate": 3.588863714278826e-05,
"loss": 0.4016,
"step": 353
},
{
"epoch": 1.4339240506329114,
"grad_norm": 0.3539239708850062,
"learning_rate": 3.585410063844186e-05,
"loss": 0.3963,
"step": 354
},
{
"epoch": 1.437974683544304,
"grad_norm": 0.42572937874718275,
"learning_rate": 3.581943644734846e-05,
"loss": 0.401,
"step": 355
},
{
"epoch": 1.4420253164556962,
"grad_norm": 0.3430904214895455,
"learning_rate": 3.578464484868869e-05,
"loss": 0.3968,
"step": 356
},
{
"epoch": 1.4460759493670885,
"grad_norm": 0.302950517509909,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.374,
"step": 357
},
{
"epoch": 1.450126582278481,
"grad_norm": 0.3263890601215173,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.4022,
"step": 358
},
{
"epoch": 1.4541772151898735,
"grad_norm": 0.36119122777006485,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.3924,
"step": 359
},
{
"epoch": 1.4582278481012658,
"grad_norm": 0.35587353661474114,
"learning_rate": 3.564420999786543e-05,
"loss": 0.4014,
"step": 360
},
{
"epoch": 1.4622784810126581,
"grad_norm": 0.3733776881723431,
"learning_rate": 3.560878558491842e-05,
"loss": 0.3964,
"step": 361
},
{
"epoch": 1.4663291139240506,
"grad_norm": 0.36115068804069855,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.3973,
"step": 362
},
{
"epoch": 1.4703797468354431,
"grad_norm": 0.40737936288584264,
"learning_rate": 3.553755991230039e-05,
"loss": 0.3885,
"step": 363
},
{
"epoch": 1.4744303797468354,
"grad_norm": 0.38821161427224454,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.4009,
"step": 364
},
{
"epoch": 1.4784810126582277,
"grad_norm": 0.41994013513896367,
"learning_rate": 3.546583369120419e-05,
"loss": 0.4072,
"step": 365
},
{
"epoch": 1.4825316455696202,
"grad_norm": 0.3826301811216342,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.4128,
"step": 366
},
{
"epoch": 1.4865822784810128,
"grad_norm": 0.34176454377609133,
"learning_rate": 3.539360923231766e-05,
"loss": 0.3929,
"step": 367
},
{
"epoch": 1.490632911392405,
"grad_norm": 0.39118666328841023,
"learning_rate": 3.535731089018394e-05,
"loss": 0.3983,
"step": 368
},
{
"epoch": 1.4946835443037974,
"grad_norm": 0.33998879804233373,
"learning_rate": 3.532088886237956e-05,
"loss": 0.3999,
"step": 369
},
{
"epoch": 1.4987341772151899,
"grad_norm": 0.3929502544515907,
"learning_rate": 3.528434344224253e-05,
"loss": 0.3911,
"step": 370
},
{
"epoch": 1.5027848101265824,
"grad_norm": 0.36990530239638786,
"learning_rate": 3.524767492410464e-05,
"loss": 0.3967,
"step": 371
},
{
"epoch": 1.5068354430379747,
"grad_norm": 0.386243805815637,
"learning_rate": 3.521088360328908e-05,
"loss": 0.3946,
"step": 372
},
{
"epoch": 1.510886075949367,
"grad_norm": 0.3133994755402812,
"learning_rate": 3.517396977610811e-05,
"loss": 0.3926,
"step": 373
},
{
"epoch": 1.5149367088607595,
"grad_norm": 0.4615213244111576,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.4035,
"step": 374
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.37164253229337435,
"learning_rate": 3.509977579282971e-05,
"loss": 0.3905,
"step": 375
},
{
"epoch": 1.5230379746835443,
"grad_norm": 0.3677837755689629,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.3882,
"step": 376
},
{
"epoch": 1.5270886075949366,
"grad_norm": 0.3585919059475591,
"learning_rate": 3.502509536445719e-05,
"loss": 0.3961,
"step": 377
},
{
"epoch": 1.531139240506329,
"grad_norm": 0.4569551631001496,
"learning_rate": 3.498757348458147e-05,
"loss": 0.4045,
"step": 378
},
{
"epoch": 1.5351898734177216,
"grad_norm": 0.34297461776799715,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.3977,
"step": 379
},
{
"epoch": 1.539240506329114,
"grad_norm": 0.4276781828252824,
"learning_rate": 3.491216790442899e-05,
"loss": 0.3974,
"step": 380
},
{
"epoch": 1.5432911392405062,
"grad_norm": 0.5515614002416039,
"learning_rate": 3.487428481145839e-05,
"loss": 0.4015,
"step": 381
},
{
"epoch": 1.5473417721518987,
"grad_norm": 0.46411547937644504,
"learning_rate": 3.483628192304278e-05,
"loss": 0.3804,
"step": 382
},
{
"epoch": 1.5513924050632912,
"grad_norm": 0.43222963524669444,
"learning_rate": 3.479815954525219e-05,
"loss": 0.3967,
"step": 383
},
{
"epoch": 1.5554430379746835,
"grad_norm": 0.45098711640054956,
"learning_rate": 3.475991798511899e-05,
"loss": 0.3917,
"step": 384
},
{
"epoch": 1.5594936708860758,
"grad_norm": 0.3394555093125839,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.3892,
"step": 385
},
{
"epoch": 1.5635443037974683,
"grad_norm": 0.40180555453527045,
"learning_rate": 3.468307855075128e-05,
"loss": 0.4085,
"step": 386
},
{
"epoch": 1.5675949367088609,
"grad_norm": 0.4291941894711293,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.4004,
"step": 387
},
{
"epoch": 1.5716455696202531,
"grad_norm": 0.5080462676067932,
"learning_rate": 3.460576609535163e-05,
"loss": 0.3928,
"step": 388
},
{
"epoch": 1.5756962025316454,
"grad_norm": 0.8851305343356615,
"learning_rate": 3.456693326250006e-05,
"loss": 0.4092,
"step": 389
},
{
"epoch": 1.579746835443038,
"grad_norm": 0.5304585885603884,
"learning_rate": 3.452798310957058e-05,
"loss": 0.3961,
"step": 390
},
{
"epoch": 1.5837974683544305,
"grad_norm": 0.6348631076538275,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.3999,
"step": 391
},
{
"epoch": 1.5878481012658228,
"grad_norm": 0.4192302282630994,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.3892,
"step": 392
},
{
"epoch": 1.591898734177215,
"grad_norm": 0.4664307236352895,
"learning_rate": 3.441043187201574e-05,
"loss": 0.4018,
"step": 393
},
{
"epoch": 1.5959493670886076,
"grad_norm": 0.4841307674257871,
"learning_rate": 3.437101558517728e-05,
"loss": 0.3969,
"step": 394
},
{
"epoch": 1.6,
"grad_norm": 0.38434696702314114,
"learning_rate": 3.433148355615496e-05,
"loss": 0.3896,
"step": 395
},
{
"epoch": 1.6040506329113924,
"grad_norm": 0.4524185499388156,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.4094,
"step": 396
},
{
"epoch": 1.6081012658227847,
"grad_norm": 0.37249973077288273,
"learning_rate": 3.425207354603043e-05,
"loss": 0.3853,
"step": 397
},
{
"epoch": 1.6121518987341772,
"grad_norm": 0.3567148904162671,
"learning_rate": 3.421219620448553e-05,
"loss": 0.4023,
"step": 398
},
{
"epoch": 1.6162025316455697,
"grad_norm": 0.3678691124280393,
"learning_rate": 3.417220439986623e-05,
"loss": 0.3985,
"step": 399
},
{
"epoch": 1.620253164556962,
"grad_norm": 0.3948142979052231,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.3827,
"step": 400
},
{
"epoch": 1.6243037974683543,
"grad_norm": 0.42362396157684196,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.398,
"step": 401
},
{
"epoch": 1.6283544303797468,
"grad_norm": 0.5492507013846221,
"learning_rate": 3.405154543304065e-05,
"loss": 0.4019,
"step": 402
},
{
"epoch": 1.6324050632911393,
"grad_norm": 0.3950920278618248,
"learning_rate": 3.401109900618843e-05,
"loss": 0.4009,
"step": 403
},
{
"epoch": 1.6364556962025316,
"grad_norm": 0.49345254611454914,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.3768,
"step": 404
},
{
"epoch": 1.640506329113924,
"grad_norm": 0.4167815313257816,
"learning_rate": 3.392986794874714e-05,
"loss": 0.3947,
"step": 405
},
{
"epoch": 1.6445569620253164,
"grad_norm": 0.5280952382248076,
"learning_rate": 3.388908397238184e-05,
"loss": 0.3956,
"step": 406
},
{
"epoch": 1.648607594936709,
"grad_norm": 0.4631304610294181,
"learning_rate": 3.384818813524362e-05,
"loss": 0.3902,
"step": 407
},
{
"epoch": 1.6526582278481012,
"grad_norm": 0.33122549030695325,
"learning_rate": 3.380718076670195e-05,
"loss": 0.3818,
"step": 408
},
{
"epoch": 1.6567088607594935,
"grad_norm": 0.4466878218994967,
"learning_rate": 3.376606219702454e-05,
"loss": 0.3726,
"step": 409
},
{
"epoch": 1.660759493670886,
"grad_norm": 0.32992444810603827,
"learning_rate": 3.372483275737468e-05,
"loss": 0.3933,
"step": 410
},
{
"epoch": 1.6648101265822786,
"grad_norm": 0.39571132697673117,
"learning_rate": 3.368349277980861e-05,
"loss": 0.3991,
"step": 411
},
{
"epoch": 1.6688607594936709,
"grad_norm": 0.35730910593326826,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.4051,
"step": 412
},
{
"epoch": 1.6729113924050631,
"grad_norm": 0.3522667236872628,
"learning_rate": 3.360048254360144e-05,
"loss": 0.3815,
"step": 413
},
{
"epoch": 1.6769620253164557,
"grad_norm": 0.3575888284314355,
"learning_rate": 3.355881295351336e-05,
"loss": 0.3886,
"step": 414
},
{
"epoch": 1.6810126582278482,
"grad_norm": 0.3728842841498144,
"learning_rate": 3.351703416260975e-05,
"loss": 0.3784,
"step": 415
},
{
"epoch": 1.6850632911392405,
"grad_norm": 0.3920987097561368,
"learning_rate": 3.347514650737126e-05,
"loss": 0.3941,
"step": 416
},
{
"epoch": 1.6891139240506328,
"grad_norm": 0.3984758181770451,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.3861,
"step": 417
},
{
"epoch": 1.6931645569620253,
"grad_norm": 0.312431082020951,
"learning_rate": 3.339104595419334e-05,
"loss": 0.3893,
"step": 418
},
{
"epoch": 1.6972151898734178,
"grad_norm": 0.4457695996389306,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.4054,
"step": 419
},
{
"epoch": 1.70126582278481,
"grad_norm": 0.4033525040406623,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.3933,
"step": 420
},
{
"epoch": 1.7053164556962024,
"grad_norm": 0.3030446840732993,
"learning_rate": 3.326408710419996e-05,
"loss": 0.3832,
"step": 421
},
{
"epoch": 1.709367088607595,
"grad_norm": 0.3447926237046455,
"learning_rate": 3.322155337795454e-05,
"loss": 0.4071,
"step": 422
},
{
"epoch": 1.7134177215189874,
"grad_norm": 0.3235628175837872,
"learning_rate": 3.317891316713587e-05,
"loss": 0.3948,
"step": 423
},
{
"epoch": 1.7174683544303797,
"grad_norm": 0.36535472205829084,
"learning_rate": 3.313616681516231e-05,
"loss": 0.3857,
"step": 424
},
{
"epoch": 1.721518987341772,
"grad_norm": 0.35878169597994186,
"learning_rate": 3.309331466630713e-05,
"loss": 0.4008,
"step": 425
},
{
"epoch": 1.7255696202531645,
"grad_norm": 0.35943020277052695,
"learning_rate": 3.305035706569563e-05,
"loss": 0.3844,
"step": 426
},
{
"epoch": 1.729620253164557,
"grad_norm": 0.35190184602273716,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.3945,
"step": 427
},
{
"epoch": 1.7336708860759493,
"grad_norm": 0.41556535225273883,
"learning_rate": 3.296412689394864e-05,
"loss": 0.3984,
"step": 428
},
{
"epoch": 1.7377215189873416,
"grad_norm": 0.3268975572268799,
"learning_rate": 3.292085501729909e-05,
"loss": 0.3916,
"step": 429
},
{
"epoch": 1.7417721518987341,
"grad_norm": 0.39265757598858037,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.4046,
"step": 430
},
{
"epoch": 1.7458227848101266,
"grad_norm": 0.34184846275148395,
"learning_rate": 3.283399942497381e-05,
"loss": 0.3878,
"step": 431
},
{
"epoch": 1.749873417721519,
"grad_norm": 0.36819343174212416,
"learning_rate": 3.279041640882108e-05,
"loss": 0.3923,
"step": 432
},
{
"epoch": 1.7539240506329112,
"grad_norm": 0.42764241318430424,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.3957,
"step": 433
},
{
"epoch": 1.7579746835443038,
"grad_norm": 0.3883489398415802,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.3801,
"step": 434
},
{
"epoch": 1.7620253164556963,
"grad_norm": 0.40700363373748605,
"learning_rate": 3.265905069502244e-05,
"loss": 0.3956,
"step": 435
},
{
"epoch": 1.7660759493670886,
"grad_norm": 0.36052848916181107,
"learning_rate": 3.261505774420016e-05,
"loss": 0.3919,
"step": 436
},
{
"epoch": 1.7701265822784809,
"grad_norm": 0.3046368921237377,
"learning_rate": 3.257096319343707e-05,
"loss": 0.388,
"step": 437
},
{
"epoch": 1.7741772151898734,
"grad_norm": 0.3573093139530244,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.3993,
"step": 438
},
{
"epoch": 1.7782278481012659,
"grad_norm": 0.35826943118111315,
"learning_rate": 3.248247071342966e-05,
"loss": 0.3933,
"step": 439
},
{
"epoch": 1.7822784810126582,
"grad_norm": 0.313985473894811,
"learning_rate": 3.243807349689161e-05,
"loss": 0.3933,
"step": 440
},
{
"epoch": 1.7863291139240505,
"grad_norm": 0.3735361401886968,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.3893,
"step": 441
},
{
"epoch": 1.790379746835443,
"grad_norm": 0.3359577154792449,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.3923,
"step": 442
},
{
"epoch": 1.7944303797468355,
"grad_norm": 0.349846236241256,
"learning_rate": 3.230428223438075e-05,
"loss": 0.3825,
"step": 443
},
{
"epoch": 1.7984810126582278,
"grad_norm": 0.3225608177528253,
"learning_rate": 3.225948647317459e-05,
"loss": 0.39,
"step": 444
},
{
"epoch": 1.80253164556962,
"grad_norm": 0.3125356773873446,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.3816,
"step": 445
},
{
"epoch": 1.8065822784810126,
"grad_norm": 0.40356625749894726,
"learning_rate": 3.216959910368034e-05,
"loss": 0.3856,
"step": 446
},
{
"epoch": 1.8106329113924051,
"grad_norm": 0.30998424683580345,
"learning_rate": 3.212450821933277e-05,
"loss": 0.4038,
"step": 447
},
{
"epoch": 1.8146835443037974,
"grad_norm": 0.40626000145964564,
"learning_rate": 3.207931968586281e-05,
"loss": 0.3928,
"step": 448
},
{
"epoch": 1.81873417721519,
"grad_norm": 0.368463030348325,
"learning_rate": 3.203403386721272e-05,
"loss": 0.3811,
"step": 449
},
{
"epoch": 1.8227848101265822,
"grad_norm": 0.34085979142522366,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.3872,
"step": 450
},
{
"epoch": 1.8268354430379747,
"grad_norm": 0.3962933325578321,
"learning_rate": 3.194317183405573e-05,
"loss": 0.4,
"step": 451
},
{
"epoch": 1.8308860759493673,
"grad_norm": 0.31795190139507207,
"learning_rate": 3.189759635133914e-05,
"loss": 0.3976,
"step": 452
},
{
"epoch": 1.8349367088607595,
"grad_norm": 0.4298290727267516,
"learning_rate": 3.185192504701718e-05,
"loss": 0.3865,
"step": 453
},
{
"epoch": 1.8389873417721518,
"grad_norm": 0.2805880966120169,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.3906,
"step": 454
},
{
"epoch": 1.8430379746835444,
"grad_norm": 0.3945657493590119,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.3979,
"step": 455
},
{
"epoch": 1.8470886075949369,
"grad_norm": 0.36980712090697276,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.4032,
"step": 456
},
{
"epoch": 1.8511392405063292,
"grad_norm": 0.32459654202965044,
"learning_rate": 3.166828898179731e-05,
"loss": 0.4005,
"step": 457
},
{
"epoch": 1.8551898734177215,
"grad_norm": 0.3306944431890373,
"learning_rate": 3.162214410223536e-05,
"loss": 0.3918,
"step": 458
},
{
"epoch": 1.859240506329114,
"grad_norm": 0.34283568417328447,
"learning_rate": 3.157590561952257e-05,
"loss": 0.399,
"step": 459
},
{
"epoch": 1.8632911392405065,
"grad_norm": 0.3170785286555656,
"learning_rate": 3.152957390605732e-05,
"loss": 0.3998,
"step": 460
},
{
"epoch": 1.8673417721518988,
"grad_norm": 0.3359187432153367,
"learning_rate": 3.148314933498886e-05,
"loss": 0.392,
"step": 461
},
{
"epoch": 1.871392405063291,
"grad_norm": 0.3829249314235056,
"learning_rate": 3.143663228021431e-05,
"loss": 0.3921,
"step": 462
},
{
"epoch": 1.8754430379746836,
"grad_norm": 0.3162565840024606,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.406,
"step": 463
},
{
"epoch": 1.879493670886076,
"grad_norm": 0.42262378283154944,
"learning_rate": 3.134332221885661e-05,
"loss": 0.3924,
"step": 464
},
{
"epoch": 1.8835443037974684,
"grad_norm": 0.363542881857946,
"learning_rate": 3.129652996377987e-05,
"loss": 0.4047,
"step": 465
},
{
"epoch": 1.8875949367088607,
"grad_norm": 0.4404289254537531,
"learning_rate": 3.12496467280038e-05,
"loss": 0.4041,
"step": 466
},
{
"epoch": 1.8916455696202532,
"grad_norm": 0.3086886134039154,
"learning_rate": 3.120267288911952e-05,
"loss": 0.4031,
"step": 467
},
{
"epoch": 1.8956962025316457,
"grad_norm": 0.37861804525907344,
"learning_rate": 3.11556088254479e-05,
"loss": 0.4065,
"step": 468
},
{
"epoch": 1.899746835443038,
"grad_norm": 0.3173690592367404,
"learning_rate": 3.11084549160364e-05,
"loss": 0.3796,
"step": 469
},
{
"epoch": 1.9037974683544303,
"grad_norm": 0.3769876579997495,
"learning_rate": 3.106121154065615e-05,
"loss": 0.3984,
"step": 470
},
{
"epoch": 1.9078481012658228,
"grad_norm": 0.38655370803340333,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.3904,
"step": 471
},
{
"epoch": 1.9118987341772153,
"grad_norm": 0.4452682935943587,
"learning_rate": 3.096645791467348e-05,
"loss": 0.3871,
"step": 472
},
{
"epoch": 1.9159493670886076,
"grad_norm": 0.4056466506353652,
"learning_rate": 3.091894842720373e-05,
"loss": 0.394,
"step": 473
},
{
"epoch": 1.92,
"grad_norm": 0.42506055543977334,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.3884,
"step": 474
},
{
"epoch": 1.9240506329113924,
"grad_norm": 0.3614559436568493,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.4076,
"step": 475
},
{
"epoch": 1.928101265822785,
"grad_norm": 0.35183188466454196,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.3889,
"step": 476
},
{
"epoch": 1.9321518987341773,
"grad_norm": 0.3381337495602103,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.397,
"step": 477
},
{
"epoch": 1.9362025316455695,
"grad_norm": 0.3200906975563331,
"learning_rate": 3.06800895716334e-05,
"loss": 0.3949,
"step": 478
},
{
"epoch": 1.940253164556962,
"grad_norm": 0.4187021166877757,
"learning_rate": 3.063205821011029e-05,
"loss": 0.3969,
"step": 479
},
{
"epoch": 1.9443037974683546,
"grad_norm": 0.37806792245771714,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.393,
"step": 480
},
{
"epoch": 1.9483544303797469,
"grad_norm": 0.3235796857554698,
"learning_rate": 3.053573898718618e-05,
"loss": 0.3912,
"step": 481
},
{
"epoch": 1.9524050632911392,
"grad_norm": 0.3836734350358943,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.3937,
"step": 482
},
{
"epoch": 1.9564556962025317,
"grad_norm": 0.35651846516736657,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.3821,
"step": 483
},
{
"epoch": 1.9605063291139242,
"grad_norm": 0.28374523123562034,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.3866,
"step": 484
},
{
"epoch": 1.9645569620253165,
"grad_norm": 0.3710903623051989,
"learning_rate": 3.034208541656255e-05,
"loss": 0.3865,
"step": 485
},
{
"epoch": 1.9686075949367088,
"grad_norm": 0.3848407085683367,
"learning_rate": 3.029346281309218e-05,
"loss": 0.3912,
"step": 486
},
{
"epoch": 1.9726582278481013,
"grad_norm": 0.28521877290201214,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.3753,
"step": 487
},
{
"epoch": 1.9767088607594938,
"grad_norm": 0.365707149287704,
"learning_rate": 3.019596929201569e-05,
"loss": 0.3795,
"step": 488
},
{
"epoch": 1.980759493670886,
"grad_norm": 0.366024169985175,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.4018,
"step": 489
},
{
"epoch": 1.9848101265822784,
"grad_norm": 0.3364598136995359,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.3801,
"step": 490
},
{
"epoch": 1.988860759493671,
"grad_norm": 0.33638418269778436,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.3986,
"step": 491
},
{
"epoch": 1.9929113924050634,
"grad_norm": 0.3472787783205286,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.3898,
"step": 492
},
{
"epoch": 1.9969620253164557,
"grad_norm": 0.27361706128985974,
"learning_rate": 2.995080534235264e-05,
"loss": 0.3834,
"step": 493
},
{
"epoch": 2.001012658227848,
"grad_norm": 0.39536853461696997,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.3901,
"step": 494
},
{
"epoch": 2.0050632911392405,
"grad_norm": 0.3693262418333024,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.3409,
"step": 495
},
{
"epoch": 2.009113924050633,
"grad_norm": 0.5025450285519282,
"learning_rate": 2.980274210295326e-05,
"loss": 0.3522,
"step": 496
},
{
"epoch": 2.013164556962025,
"grad_norm": 0.45611609034820505,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.3422,
"step": 497
},
{
"epoch": 2.0172151898734176,
"grad_norm": 0.3803167265529588,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.3404,
"step": 498
},
{
"epoch": 2.02126582278481,
"grad_norm": 0.42290169554425056,
"learning_rate": 2.965396831793362e-05,
"loss": 0.3397,
"step": 499
},
{
"epoch": 2.0253164556962027,
"grad_norm": 0.3544187289164109,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.3363,
"step": 500
},
{
"epoch": 2.0293670886075947,
"grad_norm": 0.4353883592893753,
"learning_rate": 2.955439637024526e-05,
"loss": 0.3335,
"step": 501
},
{
"epoch": 2.0334177215189873,
"grad_norm": 0.3730791389973477,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.3434,
"step": 502
},
{
"epoch": 2.0374683544303798,
"grad_norm": 0.36237503260678283,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.333,
"step": 503
},
{
"epoch": 2.0415189873417723,
"grad_norm": 0.4505439455703962,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.3488,
"step": 504
},
{
"epoch": 2.0455696202531644,
"grad_norm": 0.4140900192165133,
"learning_rate": 2.93543322968538e-05,
"loss": 0.3364,
"step": 505
},
{
"epoch": 2.049620253164557,
"grad_norm": 0.36660884055586607,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.3419,
"step": 506
},
{
"epoch": 2.0536708860759494,
"grad_norm": 0.36533239399093453,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.3444,
"step": 507
},
{
"epoch": 2.057721518987342,
"grad_norm": 0.43793409502278347,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.3246,
"step": 508
},
{
"epoch": 2.061772151898734,
"grad_norm": 0.43298555353876034,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.3319,
"step": 509
},
{
"epoch": 2.0658227848101265,
"grad_norm": 0.31817459129717185,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.3441,
"step": 510
},
{
"epoch": 2.069873417721519,
"grad_norm": 0.5376292310605345,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.3349,
"step": 511
},
{
"epoch": 2.0739240506329115,
"grad_norm": 0.40423026535288487,
"learning_rate": 2.900133526131588e-05,
"loss": 0.3312,
"step": 512
},
{
"epoch": 2.0779746835443036,
"grad_norm": 0.33036540050882,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.3363,
"step": 513
},
{
"epoch": 2.082025316455696,
"grad_norm": 0.5097161119273677,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.3396,
"step": 514
},
{
"epoch": 2.0860759493670886,
"grad_norm": 0.4541725842426109,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.3382,
"step": 515
},
{
"epoch": 2.090126582278481,
"grad_norm": 0.31157905307153666,
"learning_rate": 2.879801882631847e-05,
"loss": 0.3344,
"step": 516
},
{
"epoch": 2.094177215189873,
"grad_norm": 0.42615411407927317,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.3416,
"step": 517
},
{
"epoch": 2.0982278481012657,
"grad_norm": 0.4099976387148203,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.3296,
"step": 518
},
{
"epoch": 2.1022784810126582,
"grad_norm": 0.30936896415058823,
"learning_rate": 2.864478605989494e-05,
"loss": 0.3417,
"step": 519
},
{
"epoch": 2.1063291139240508,
"grad_norm": 0.4039119979212538,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.3455,
"step": 520
},
{
"epoch": 2.110379746835443,
"grad_norm": 0.42126850271442123,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.3409,
"step": 521
},
{
"epoch": 2.1144303797468353,
"grad_norm": 0.2864311348429795,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.3386,
"step": 522
},
{
"epoch": 2.118481012658228,
"grad_norm": 0.3794974376149531,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.3451,
"step": 523
},
{
"epoch": 2.1225316455696204,
"grad_norm": 0.37033165141801566,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.3342,
"step": 524
},
{
"epoch": 2.1265822784810124,
"grad_norm": 0.39718204918633276,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.3366,
"step": 525
},
{
"epoch": 2.130632911392405,
"grad_norm": 0.37551701977415364,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.3348,
"step": 526
},
{
"epoch": 2.1346835443037975,
"grad_norm": 0.3536853468074591,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.3403,
"step": 527
},
{
"epoch": 2.13873417721519,
"grad_norm": 0.3750506936338626,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.3355,
"step": 528
},
{
"epoch": 2.1427848101265825,
"grad_norm": 0.34079525352843115,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.3398,
"step": 529
},
{
"epoch": 2.1468354430379746,
"grad_norm": 0.37315693496678604,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.3345,
"step": 530
},
{
"epoch": 2.150886075949367,
"grad_norm": 0.35640147220907653,
"learning_rate": 2.802570062143278e-05,
"loss": 0.3391,
"step": 531
},
{
"epoch": 2.1549367088607596,
"grad_norm": 0.326323863799756,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.336,
"step": 532
},
{
"epoch": 2.1589873417721517,
"grad_norm": 0.30495646404427096,
"learning_rate": 2.792159532078314e-05,
"loss": 0.3416,
"step": 533
},
{
"epoch": 2.163037974683544,
"grad_norm": 0.3879535536451282,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.3418,
"step": 534
},
{
"epoch": 2.1670886075949367,
"grad_norm": 0.34239962789044215,
"learning_rate": 2.781723482288857e-05,
"loss": 0.3386,
"step": 535
},
{
"epoch": 2.1711392405063292,
"grad_norm": 0.30239409960328245,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.3384,
"step": 536
},
{
"epoch": 2.1751898734177217,
"grad_norm": 0.4381851199699727,
"learning_rate": 2.771262248976272e-05,
"loss": 0.3453,
"step": 537
},
{
"epoch": 2.179240506329114,
"grad_norm": 0.27838389901018257,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.342,
"step": 538
},
{
"epoch": 2.1832911392405063,
"grad_norm": 0.3637473567194387,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.3305,
"step": 539
},
{
"epoch": 2.187341772151899,
"grad_norm": 0.3295653380953245,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.3352,
"step": 540
},
{
"epoch": 2.191392405063291,
"grad_norm": 0.33537838672836995,
"learning_rate": 2.7502655806328e-05,
"loss": 0.3448,
"step": 541
},
{
"epoch": 2.1954430379746834,
"grad_norm": 0.36696574056362224,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.3421,
"step": 542
},
{
"epoch": 2.199493670886076,
"grad_norm": 0.3027064564865914,
"learning_rate": 2.739730822017673e-05,
"loss": 0.3413,
"step": 543
},
{
"epoch": 2.2035443037974685,
"grad_norm": 0.28598565136978654,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.3478,
"step": 544
},
{
"epoch": 2.207594936708861,
"grad_norm": 0.34885555600063844,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.3466,
"step": 545
},
{
"epoch": 2.211645569620253,
"grad_norm": 0.28571950545448044,
"learning_rate": 2.723884107774236e-05,
"loss": 0.3467,
"step": 546
},
{
"epoch": 2.2156962025316456,
"grad_norm": 0.35018224331587006,
"learning_rate": 2.718590152796239e-05,
"loss": 0.3361,
"step": 547
},
{
"epoch": 2.219746835443038,
"grad_norm": 0.3337415570026271,
"learning_rate": 2.71329041039194e-05,
"loss": 0.3425,
"step": 548
},
{
"epoch": 2.22379746835443,
"grad_norm": 0.33337745691461346,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.3433,
"step": 549
},
{
"epoch": 2.2278481012658227,
"grad_norm": 0.3128929849598542,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.3413,
"step": 550
},
{
"epoch": 2.231898734177215,
"grad_norm": 0.31338631446087406,
"learning_rate": 2.697356885686189e-05,
"loss": 0.3393,
"step": 551
},
{
"epoch": 2.2359493670886077,
"grad_norm": 0.32285390496802224,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.3443,
"step": 552
},
{
"epoch": 2.24,
"grad_norm": 0.34751158690373946,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.3487,
"step": 553
},
{
"epoch": 2.2440506329113923,
"grad_norm": 0.30647082211715226,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.3395,
"step": 554
},
{
"epoch": 2.248101265822785,
"grad_norm": 0.34381757601234614,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.3438,
"step": 555
},
{
"epoch": 2.2521518987341773,
"grad_norm": 0.3021279658430541,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.3458,
"step": 556
},
{
"epoch": 2.2562025316455694,
"grad_norm": 0.3183831927115024,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.3431,
"step": 557
},
{
"epoch": 2.260253164556962,
"grad_norm": 0.3264751033030605,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.3397,
"step": 558
},
{
"epoch": 2.2643037974683544,
"grad_norm": 0.33080937639987595,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.3535,
"step": 559
},
{
"epoch": 2.268354430379747,
"grad_norm": 0.31542782368799205,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.3414,
"step": 560
},
{
"epoch": 2.2724050632911394,
"grad_norm": 0.34385995533036184,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.3394,
"step": 561
},
{
"epoch": 2.2764556962025315,
"grad_norm": 0.25694614322414283,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.3347,
"step": 562
},
{
"epoch": 2.280506329113924,
"grad_norm": 0.30199995221106685,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.3369,
"step": 563
},
{
"epoch": 2.2845569620253166,
"grad_norm": 0.287459307339179,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.346,
"step": 564
},
{
"epoch": 2.2886075949367086,
"grad_norm": 0.29637150188262606,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.3366,
"step": 565
},
{
"epoch": 2.292658227848101,
"grad_norm": 0.3586010615699321,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.3474,
"step": 566
},
{
"epoch": 2.2967088607594937,
"grad_norm": 0.2805357647910659,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.3372,
"step": 567
},
{
"epoch": 2.300759493670886,
"grad_norm": 0.3408994723568919,
"learning_rate": 2.606146264569603e-05,
"loss": 0.339,
"step": 568
},
{
"epoch": 2.3048101265822787,
"grad_norm": 0.32468774206444107,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.3539,
"step": 569
},
{
"epoch": 2.3088607594936708,
"grad_norm": 0.3273340033639118,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.3425,
"step": 570
},
{
"epoch": 2.3129113924050633,
"grad_norm": 0.31329974415518336,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.336,
"step": 571
},
{
"epoch": 2.316962025316456,
"grad_norm": 0.3368888631192898,
"learning_rate": 2.584472010606015e-05,
"loss": 0.3296,
"step": 572
},
{
"epoch": 2.321012658227848,
"grad_norm": 0.3221269141496823,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.3391,
"step": 573
},
{
"epoch": 2.3250632911392404,
"grad_norm": 0.30067588833584685,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.3337,
"step": 574
},
{
"epoch": 2.329113924050633,
"grad_norm": 0.31060210476935507,
"learning_rate": 2.568166741322048e-05,
"loss": 0.3444,
"step": 575
},
{
"epoch": 2.3331645569620254,
"grad_norm": 0.3583593733744927,
"learning_rate": 2.56272244128509e-05,
"loss": 0.3502,
"step": 576
},
{
"epoch": 2.337215189873418,
"grad_norm": 0.2946352276951408,
"learning_rate": 2.55727360915895e-05,
"loss": 0.35,
"step": 577
},
{
"epoch": 2.34126582278481,
"grad_norm": 0.34811051348157884,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.3384,
"step": 578
},
{
"epoch": 2.3453164556962025,
"grad_norm": 0.2930624889502892,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.3438,
"step": 579
},
{
"epoch": 2.349367088607595,
"grad_norm": 0.3520751738028624,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.3468,
"step": 580
},
{
"epoch": 2.353417721518987,
"grad_norm": 0.29180560077831535,
"learning_rate": 2.535433837985559e-05,
"loss": 0.3409,
"step": 581
},
{
"epoch": 2.3574683544303796,
"grad_norm": 0.29362403826230665,
"learning_rate": 2.529963004393324e-05,
"loss": 0.3447,
"step": 582
},
{
"epoch": 2.361518987341772,
"grad_norm": 0.3368002275992509,
"learning_rate": 2.524487902551908e-05,
"loss": 0.3371,
"step": 583
},
{
"epoch": 2.3655696202531646,
"grad_norm": 0.28259578403540214,
"learning_rate": 2.519008576557029e-05,
"loss": 0.3415,
"step": 584
},
{
"epoch": 2.369620253164557,
"grad_norm": 0.3512996274065245,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.334,
"step": 585
},
{
"epoch": 2.3736708860759492,
"grad_norm": 0.2789888525173277,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.3343,
"step": 586
},
{
"epoch": 2.3777215189873417,
"grad_norm": 0.2957032476152624,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.3499,
"step": 587
},
{
"epoch": 2.3817721518987343,
"grad_norm": 0.30977308142483645,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.3447,
"step": 588
},
{
"epoch": 2.3858227848101268,
"grad_norm": 0.27987378482770864,
"learning_rate": 2.491550129992345e-05,
"loss": 0.3399,
"step": 589
},
{
"epoch": 2.389873417721519,
"grad_norm": 0.3019873343435055,
"learning_rate": 2.486046386967024e-05,
"loss": 0.3409,
"step": 590
},
{
"epoch": 2.3939240506329114,
"grad_norm": 0.34063013898065797,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.3393,
"step": 591
},
{
"epoch": 2.397974683544304,
"grad_norm": 0.33196116821537336,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.3429,
"step": 592
},
{
"epoch": 2.4020253164556964,
"grad_norm": 0.36854499654567346,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.3491,
"step": 593
},
{
"epoch": 2.4060759493670885,
"grad_norm": 0.31982732533293423,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.3323,
"step": 594
},
{
"epoch": 2.410126582278481,
"grad_norm": 0.34943185725760684,
"learning_rate": 2.458469841202444e-05,
"loss": 0.3455,
"step": 595
},
{
"epoch": 2.4141772151898735,
"grad_norm": 0.3130131138906674,
"learning_rate": 2.452943276859226e-05,
"loss": 0.343,
"step": 596
},
{
"epoch": 2.418227848101266,
"grad_norm": 0.31074418197999704,
"learning_rate": 2.447413064573125e-05,
"loss": 0.3464,
"step": 597
},
{
"epoch": 2.422278481012658,
"grad_norm": 0.29107110909520245,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.3492,
"step": 598
},
{
"epoch": 2.4263291139240506,
"grad_norm": 0.26320567176713516,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.3303,
"step": 599
},
{
"epoch": 2.430379746835443,
"grad_norm": 0.3116051172579589,
"learning_rate": 2.430800985597963e-05,
"loss": 0.3491,
"step": 600
},
{
"epoch": 2.4344303797468356,
"grad_norm": 0.3043032411050649,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.3466,
"step": 601
},
{
"epoch": 2.4384810126582277,
"grad_norm": 0.3193258036134264,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.3449,
"step": 602
},
{
"epoch": 2.44253164556962,
"grad_norm": 0.31047071317021835,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.3472,
"step": 603
},
{
"epoch": 2.4465822784810127,
"grad_norm": 0.2916986848806381,
"learning_rate": 2.408603181104957e-05,
"loss": 0.3374,
"step": 604
},
{
"epoch": 2.4506329113924052,
"grad_norm": 0.326370022769883,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.358,
"step": 605
},
{
"epoch": 2.4546835443037973,
"grad_norm": 0.334405296260276,
"learning_rate": 2.397484355008521e-05,
"loss": 0.3411,
"step": 606
},
{
"epoch": 2.45873417721519,
"grad_norm": 0.32810578139313834,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.3506,
"step": 607
},
{
"epoch": 2.4627848101265823,
"grad_norm": 0.32816548924972744,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.335,
"step": 608
},
{
"epoch": 2.466835443037975,
"grad_norm": 0.36016254450137214,
"learning_rate": 2.380782218329337e-05,
"loss": 0.3479,
"step": 609
},
{
"epoch": 2.470886075949367,
"grad_norm": 0.3323069547511776,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.3414,
"step": 610
},
{
"epoch": 2.4749367088607594,
"grad_norm": 0.31369000379946194,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.3511,
"step": 611
},
{
"epoch": 2.478987341772152,
"grad_norm": 0.3026640772371867,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.3458,
"step": 612
},
{
"epoch": 2.4830379746835445,
"grad_norm": 0.3366413836834664,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.3435,
"step": 613
},
{
"epoch": 2.4870886075949366,
"grad_norm": 0.29693762088323716,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.3483,
"step": 614
},
{
"epoch": 2.491139240506329,
"grad_norm": 0.3676826215043034,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.3382,
"step": 615
},
{
"epoch": 2.4951898734177216,
"grad_norm": 0.28139837535962237,
"learning_rate": 2.341705325992901e-05,
"loss": 0.3413,
"step": 616
},
{
"epoch": 2.499240506329114,
"grad_norm": 0.3618245897963192,
"learning_rate": 2.336111544604222e-05,
"loss": 0.3306,
"step": 617
},
{
"epoch": 2.503291139240506,
"grad_norm": 0.36608562349658,
"learning_rate": 2.33051505621937e-05,
"loss": 0.3553,
"step": 618
},
{
"epoch": 2.5073417721518987,
"grad_norm": 0.35964412705076726,
"learning_rate": 2.324915905911693e-05,
"loss": 0.3586,
"step": 619
},
{
"epoch": 2.511392405063291,
"grad_norm": 0.33139045086577495,
"learning_rate": 2.319314138775977e-05,
"loss": 0.3331,
"step": 620
},
{
"epoch": 2.5154430379746833,
"grad_norm": 0.2709082028295046,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.3278,
"step": 621
},
{
"epoch": 2.519493670886076,
"grad_norm": 0.3254637650161838,
"learning_rate": 2.308102934504593e-05,
"loss": 0.3454,
"step": 622
},
{
"epoch": 2.5235443037974683,
"grad_norm": 0.2636247776369819,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.343,
"step": 623
},
{
"epoch": 2.527594936708861,
"grad_norm": 0.3148119552746377,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.3466,
"step": 624
},
{
"epoch": 2.5316455696202533,
"grad_norm": 0.28256120371933335,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.3343,
"step": 625
},
{
"epoch": 2.5356962025316454,
"grad_norm": 0.308799157119591,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.3504,
"step": 626
},
{
"epoch": 2.539746835443038,
"grad_norm": 0.2945744878347889,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.3318,
"step": 627
},
{
"epoch": 2.5437974683544304,
"grad_norm": 0.30243627623113545,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.3448,
"step": 628
},
{
"epoch": 2.547848101265823,
"grad_norm": 0.3617154121883429,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.3452,
"step": 629
},
{
"epoch": 2.5518987341772155,
"grad_norm": 0.29145058417975844,
"learning_rate": 2.26316247733093e-05,
"loss": 0.3381,
"step": 630
},
{
"epoch": 2.5559493670886075,
"grad_norm": 0.3078127182138568,
"learning_rate": 2.257534907110328e-05,
"loss": 0.3415,
"step": 631
},
{
"epoch": 2.56,
"grad_norm": 0.3303910800537368,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.3463,
"step": 632
},
{
"epoch": 2.5640506329113926,
"grad_norm": 0.2916829313029925,
"learning_rate": 2.246273589557294e-05,
"loss": 0.3338,
"step": 633
},
{
"epoch": 2.5681012658227846,
"grad_norm": 0.3382811693955486,
"learning_rate": 2.240639932921966e-05,
"loss": 0.3371,
"step": 634
},
{
"epoch": 2.572151898734177,
"grad_norm": 0.2810397112498718,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.3516,
"step": 635
},
{
"epoch": 2.5762025316455697,
"grad_norm": 0.30182436221998715,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.3416,
"step": 636
},
{
"epoch": 2.580253164556962,
"grad_norm": 0.26048777800748213,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.3393,
"step": 637
},
{
"epoch": 2.5843037974683547,
"grad_norm": 0.2685814331647914,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.3339,
"step": 638
},
{
"epoch": 2.5883544303797468,
"grad_norm": 0.2719642154330346,
"learning_rate": 2.212443486523819e-05,
"loss": 0.3408,
"step": 639
},
{
"epoch": 2.5924050632911393,
"grad_norm": 0.2584908291914648,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.3448,
"step": 640
},
{
"epoch": 2.596455696202532,
"grad_norm": 0.2976504420621647,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.3372,
"step": 641
},
{
"epoch": 2.600506329113924,
"grad_norm": 0.2731000483309335,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.3323,
"step": 642
},
{
"epoch": 2.6045569620253164,
"grad_norm": 0.3146564011474982,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.3479,
"step": 643
},
{
"epoch": 2.608607594936709,
"grad_norm": 0.2537166135699787,
"learning_rate": 2.184204266048803e-05,
"loss": 0.3388,
"step": 644
},
{
"epoch": 2.6126582278481014,
"grad_norm": 0.27969172456128144,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.3441,
"step": 645
},
{
"epoch": 2.616708860759494,
"grad_norm": 0.2603328265549545,
"learning_rate": 2.17289787446841e-05,
"loss": 0.3479,
"step": 646
},
{
"epoch": 2.620759493670886,
"grad_norm": 0.2954541389256923,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.3403,
"step": 647
},
{
"epoch": 2.6248101265822785,
"grad_norm": 0.27205189910496175,
"learning_rate": 2.161585912916385e-05,
"loss": 0.3306,
"step": 648
},
{
"epoch": 2.628860759493671,
"grad_norm": 0.26259835669285203,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.3483,
"step": 649
},
{
"epoch": 2.632911392405063,
"grad_norm": 0.2471728366718138,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.3252,
"step": 650
},
{
"epoch": 2.6369620253164556,
"grad_norm": 0.2780868998190715,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.3431,
"step": 651
},
{
"epoch": 2.641012658227848,
"grad_norm": 0.2472629766832129,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.3325,
"step": 652
},
{
"epoch": 2.6450632911392407,
"grad_norm": 0.2729267620959595,
"learning_rate": 2.133284032321232e-05,
"loss": 0.3417,
"step": 653
},
{
"epoch": 2.649113924050633,
"grad_norm": 0.27463654570211976,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.3402,
"step": 654
},
{
"epoch": 2.6531645569620252,
"grad_norm": 0.24632592221945174,
"learning_rate": 2.121955446741306e-05,
"loss": 0.335,
"step": 655
},
{
"epoch": 2.6572151898734178,
"grad_norm": 0.3378797084707074,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.3468,
"step": 656
},
{
"epoch": 2.6612658227848103,
"grad_norm": 0.24366198780253076,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.3429,
"step": 657
},
{
"epoch": 2.6653164556962023,
"grad_norm": 0.32970350339324145,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.3515,
"step": 658
},
{
"epoch": 2.669367088607595,
"grad_norm": 0.27756894226534584,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.3398,
"step": 659
},
{
"epoch": 2.6734177215189874,
"grad_norm": 0.2984552076919266,
"learning_rate": 2.093617592756158e-05,
"loss": 0.3575,
"step": 660
},
{
"epoch": 2.67746835443038,
"grad_norm": 0.2621693070566985,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.3347,
"step": 661
},
{
"epoch": 2.6815189873417724,
"grad_norm": 0.30876787970309266,
"learning_rate": 2.082276853708978e-05,
"loss": 0.3443,
"step": 662
},
{
"epoch": 2.6855696202531645,
"grad_norm": 0.260953843402673,
"learning_rate": 2.076605467380071e-05,
"loss": 0.3396,
"step": 663
},
{
"epoch": 2.689620253164557,
"grad_norm": 0.26410371576440533,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.3347,
"step": 664
},
{
"epoch": 2.6936708860759495,
"grad_norm": 0.26131164620905856,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.3359,
"step": 665
},
{
"epoch": 2.6977215189873416,
"grad_norm": 0.2601147681720199,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.3422,
"step": 666
},
{
"epoch": 2.701772151898734,
"grad_norm": 0.2560149571668363,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.3401,
"step": 667
},
{
"epoch": 2.7058227848101266,
"grad_norm": 0.25784584378894654,
"learning_rate": 2.048240194884496e-05,
"loss": 0.3267,
"step": 668
},
{
"epoch": 2.709873417721519,
"grad_norm": 0.2724379385077366,
"learning_rate": 2.042565792045695e-05,
"loss": 0.3438,
"step": 669
},
{
"epoch": 2.7139240506329116,
"grad_norm": 0.2664594930759751,
"learning_rate": 2.036891046387857e-05,
"loss": 0.3452,
"step": 670
},
{
"epoch": 2.7179746835443037,
"grad_norm": 0.5380533432939872,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.3408,
"step": 671
},
{
"epoch": 2.7220253164556962,
"grad_norm": 0.32849445765923274,
"learning_rate": 2.025540709431948e-05,
"loss": 0.3387,
"step": 672
},
{
"epoch": 2.7260759493670887,
"grad_norm": 0.32047350969514293,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.3402,
"step": 673
},
{
"epoch": 2.730126582278481,
"grad_norm": 0.30794233724179515,
"learning_rate": 2.014189549672245e-05,
"loss": 0.3482,
"step": 674
},
{
"epoch": 2.7341772151898733,
"grad_norm": 0.2908004320822143,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.3447,
"step": 675
},
{
"epoch": 2.738227848101266,
"grad_norm": 0.28604215152829665,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.3482,
"step": 676
},
{
"epoch": 2.7422784810126584,
"grad_norm": 0.25837141964736565,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.3386,
"step": 677
},
{
"epoch": 2.746329113924051,
"grad_norm": 0.2779938884321906,
"learning_rate": 1.991486224484123e-05,
"loss": 0.342,
"step": 678
},
{
"epoch": 2.750379746835443,
"grad_norm": 0.24798233633882708,
"learning_rate": 1.985810450327756e-05,
"loss": 0.3355,
"step": 679
},
{
"epoch": 2.7544303797468355,
"grad_norm": 0.2655416724487501,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.3384,
"step": 680
},
{
"epoch": 2.758481012658228,
"grad_norm": 0.24326905980316324,
"learning_rate": 1.974459290568053e-05,
"loss": 0.3379,
"step": 681
},
{
"epoch": 2.76253164556962,
"grad_norm": 0.2989104815770004,
"learning_rate": 1.968783996385397e-05,
"loss": 0.3446,
"step": 682
},
{
"epoch": 2.7665822784810126,
"grad_norm": 0.2697704959512499,
"learning_rate": 1.963108953612143e-05,
"loss": 0.3439,
"step": 683
},
{
"epoch": 2.770632911392405,
"grad_norm": 0.2749139623197672,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.3441,
"step": 684
},
{
"epoch": 2.7746835443037976,
"grad_norm": 0.24787502832482325,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.3456,
"step": 685
},
{
"epoch": 2.77873417721519,
"grad_norm": 0.2732959112931191,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.3407,
"step": 686
},
{
"epoch": 2.782784810126582,
"grad_norm": 0.23292254500725945,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.3422,
"step": 687
},
{
"epoch": 2.7868354430379747,
"grad_norm": 0.28849908683152503,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.3297,
"step": 688
},
{
"epoch": 2.790886075949367,
"grad_norm": 0.24619780924122936,
"learning_rate": 1.92906653591874e-05,
"loss": 0.3323,
"step": 689
},
{
"epoch": 2.7949367088607593,
"grad_norm": 0.27607351614505415,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.3394,
"step": 690
},
{
"epoch": 2.798987341772152,
"grad_norm": 0.2352319577064066,
"learning_rate": 1.917723146291022e-05,
"loss": 0.3306,
"step": 691
},
{
"epoch": 2.8030379746835443,
"grad_norm": 0.25969551024643817,
"learning_rate": 1.912052422608584e-05,
"loss": 0.338,
"step": 692
},
{
"epoch": 2.807088607594937,
"grad_norm": 0.22984375101669802,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.3383,
"step": 693
},
{
"epoch": 2.8111392405063294,
"grad_norm": 0.2649979592636697,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.3404,
"step": 694
},
{
"epoch": 2.8151898734177214,
"grad_norm": 0.2551402579817862,
"learning_rate": 1.895044684123479e-05,
"loss": 0.3377,
"step": 695
},
{
"epoch": 2.819240506329114,
"grad_norm": 0.26389068689216827,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.3515,
"step": 696
},
{
"epoch": 2.8232911392405065,
"grad_norm": 0.2795419494876548,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.3529,
"step": 697
},
{
"epoch": 2.8273417721518985,
"grad_norm": 0.22203934922969656,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.332,
"step": 698
},
{
"epoch": 2.831392405063291,
"grad_norm": 0.22797035138923968,
"learning_rate": 1.872379746550743e-05,
"loss": 0.3356,
"step": 699
},
{
"epoch": 2.8354430379746836,
"grad_norm": 0.2434073695122467,
"learning_rate": 1.866715967678769e-05,
"loss": 0.3238,
"step": 700
},
{
"epoch": 2.839493670886076,
"grad_norm": 0.2122103542408331,
"learning_rate": 1.861053262258067e-05,
"loss": 0.328,
"step": 701
},
{
"epoch": 2.8435443037974686,
"grad_norm": 0.25623731889418194,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.3511,
"step": 702
},
{
"epoch": 2.8475949367088607,
"grad_norm": 0.21411941270638776,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.3373,
"step": 703
},
{
"epoch": 2.851645569620253,
"grad_norm": 0.28643867403899403,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.3411,
"step": 704
},
{
"epoch": 2.8556962025316457,
"grad_norm": 0.27898291749262594,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.3392,
"step": 705
},
{
"epoch": 2.8597468354430378,
"grad_norm": 0.2537783201596236,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.3438,
"step": 706
},
{
"epoch": 2.8637974683544303,
"grad_norm": 0.2888655520089863,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.331,
"step": 707
},
{
"epoch": 2.867848101265823,
"grad_norm": 0.24423432369913342,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.3492,
"step": 708
},
{
"epoch": 2.8718987341772153,
"grad_norm": 0.3182662275468079,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.3336,
"step": 709
},
{
"epoch": 2.875949367088608,
"grad_norm": 0.23641241758085038,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.3378,
"step": 710
},
{
"epoch": 2.88,
"grad_norm": 0.2851612831265895,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.3345,
"step": 711
},
{
"epoch": 2.8840506329113924,
"grad_norm": 0.2672465200521031,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.342,
"step": 712
},
{
"epoch": 2.888101265822785,
"grad_norm": 0.28462332514831373,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.3331,
"step": 713
},
{
"epoch": 2.892151898734177,
"grad_norm": 0.26036948024106193,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.3354,
"step": 714
},
{
"epoch": 2.8962025316455695,
"grad_norm": 0.26450818235475704,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.3372,
"step": 715
},
{
"epoch": 2.900253164556962,
"grad_norm": 0.2418163788588456,
"learning_rate": 1.776272483901182e-05,
"loss": 0.329,
"step": 716
},
{
"epoch": 2.9043037974683545,
"grad_norm": 0.28281014364790075,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.3442,
"step": 717
},
{
"epoch": 2.908354430379747,
"grad_norm": 0.2301815577660532,
"learning_rate": 1.764995661794301e-05,
"loss": 0.3386,
"step": 718
},
{
"epoch": 2.912405063291139,
"grad_norm": 0.2426329493725031,
"learning_rate": 1.759360067078035e-05,
"loss": 0.3353,
"step": 719
},
{
"epoch": 2.9164556962025316,
"grad_norm": 0.23794688008722698,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.3536,
"step": 720
},
{
"epoch": 2.920506329113924,
"grad_norm": 0.2510285565409903,
"learning_rate": 1.748094737261012e-05,
"loss": 0.3406,
"step": 721
},
{
"epoch": 2.9245569620253162,
"grad_norm": 0.23695518391214765,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.3335,
"step": 722
},
{
"epoch": 2.9286075949367087,
"grad_norm": 0.24613375748658445,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.3421,
"step": 723
},
{
"epoch": 2.9326582278481013,
"grad_norm": 0.23689418474597299,
"learning_rate": 1.731212071922883e-05,
"loss": 0.3326,
"step": 724
},
{
"epoch": 2.9367088607594938,
"grad_norm": 0.2945393651912843,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.3368,
"step": 725
},
{
"epoch": 2.9407594936708863,
"grad_norm": 0.24133947335457262,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.3405,
"step": 726
},
{
"epoch": 2.9448101265822784,
"grad_norm": 0.23495049587443773,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.3354,
"step": 727
},
{
"epoch": 2.948860759493671,
"grad_norm": 0.221199466412082,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.3263,
"step": 728
},
{
"epoch": 2.9529113924050634,
"grad_norm": 0.2272206389883426,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.3388,
"step": 729
},
{
"epoch": 2.9569620253164555,
"grad_norm": 0.2539543706431719,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.3418,
"step": 730
},
{
"epoch": 2.961012658227848,
"grad_norm": 0.22905076120618073,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.3312,
"step": 731
},
{
"epoch": 2.9650632911392405,
"grad_norm": 0.24670721366404885,
"learning_rate": 1.686290200071915e-05,
"loss": 0.3374,
"step": 732
},
{
"epoch": 2.969113924050633,
"grad_norm": 0.24013307184064592,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.3441,
"step": 733
},
{
"epoch": 2.9731645569620255,
"grad_norm": 0.23075581371288023,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.3281,
"step": 734
},
{
"epoch": 2.9772151898734176,
"grad_norm": 0.21402123357452313,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.3371,
"step": 735
},
{
"epoch": 2.98126582278481,
"grad_norm": 0.24135024561271387,
"learning_rate": 1.663888455395778e-05,
"loss": 0.3356,
"step": 736
},
{
"epoch": 2.9853164556962026,
"grad_norm": 0.2148485209248323,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.3385,
"step": 737
},
{
"epoch": 2.9893670886075947,
"grad_norm": 0.24362080183989535,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.3465,
"step": 738
},
{
"epoch": 2.993417721518987,
"grad_norm": 0.2576226704329281,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.3325,
"step": 739
},
{
"epoch": 2.9974683544303797,
"grad_norm": 0.22176454826710973,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.3393,
"step": 740
},
{
"epoch": 3.0015189873417722,
"grad_norm": 0.2894070076846216,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.3229,
"step": 741
},
{
"epoch": 3.0055696202531648,
"grad_norm": 0.2862364906340537,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.2946,
"step": 742
},
{
"epoch": 3.009620253164557,
"grad_norm": 0.5908487179177049,
"learning_rate": 1.624791353909428e-05,
"loss": 0.2848,
"step": 743
},
{
"epoch": 3.0136708860759494,
"grad_norm": 0.38094896540573203,
"learning_rate": 1.619217781670663e-05,
"loss": 0.2811,
"step": 744
},
{
"epoch": 3.017721518987342,
"grad_norm": 0.4049991893152948,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.2991,
"step": 745
},
{
"epoch": 3.0217721518987344,
"grad_norm": 0.3302748543568189,
"learning_rate": 1.608079882360134e-05,
"loss": 0.2861,
"step": 746
},
{
"epoch": 3.0258227848101265,
"grad_norm": 0.37450505831655667,
"learning_rate": 1.60251564499148e-05,
"loss": 0.2843,
"step": 747
},
{
"epoch": 3.029873417721519,
"grad_norm": 0.3509320328150259,
"learning_rate": 1.596954608907176e-05,
"loss": 0.2956,
"step": 748
},
{
"epoch": 3.0339240506329115,
"grad_norm": 0.3342926281177519,
"learning_rate": 1.591396818895043e-05,
"loss": 0.2939,
"step": 749
},
{
"epoch": 3.037974683544304,
"grad_norm": 0.318132178870383,
"learning_rate": 1.585842319716759e-05,
"loss": 0.294,
"step": 750
},
{
"epoch": 3.042025316455696,
"grad_norm": 0.2728368019675022,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.2918,
"step": 751
},
{
"epoch": 3.0460759493670886,
"grad_norm": 0.3247798481233589,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.2878,
"step": 752
},
{
"epoch": 3.050126582278481,
"grad_norm": 0.29713601953949165,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.2897,
"step": 753
},
{
"epoch": 3.0541772151898736,
"grad_norm": 0.27703521716198587,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.2709,
"step": 754
},
{
"epoch": 3.0582278481012657,
"grad_norm": 0.31878484296956083,
"learning_rate": 1.558120751116291e-05,
"loss": 0.2952,
"step": 755
},
{
"epoch": 3.062278481012658,
"grad_norm": 0.2808319087880098,
"learning_rate": 1.552586935426876e-05,
"loss": 0.29,
"step": 756
},
{
"epoch": 3.0663291139240507,
"grad_norm": 0.295702665842535,
"learning_rate": 1.547056723140774e-05,
"loss": 0.2919,
"step": 757
},
{
"epoch": 3.0703797468354432,
"grad_norm": 0.2725845410267604,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.285,
"step": 758
},
{
"epoch": 3.0744303797468353,
"grad_norm": 0.2816308455284789,
"learning_rate": 1.536007286907411e-05,
"loss": 0.2954,
"step": 759
},
{
"epoch": 3.078481012658228,
"grad_norm": 0.2864125877901859,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.288,
"step": 760
},
{
"epoch": 3.0825316455696203,
"grad_norm": 0.27209585969378636,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.2969,
"step": 761
},
{
"epoch": 3.086582278481013,
"grad_norm": 0.2896333314736055,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.2838,
"step": 762
},
{
"epoch": 3.090632911392405,
"grad_norm": 0.23838588684343753,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.2873,
"step": 763
},
{
"epoch": 3.0946835443037974,
"grad_norm": 0.2746102648849077,
"learning_rate": 1.508449870007656e-05,
"loss": 0.2864,
"step": 764
},
{
"epoch": 3.09873417721519,
"grad_norm": 0.24214196331364873,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.2859,
"step": 765
},
{
"epoch": 3.1027848101265825,
"grad_norm": 0.24617456130350532,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.2768,
"step": 766
},
{
"epoch": 3.1068354430379745,
"grad_norm": 0.27383851432777145,
"learning_rate": 1.4919625713405e-05,
"loss": 0.29,
"step": 767
},
{
"epoch": 3.110886075949367,
"grad_norm": 0.24065735473570812,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.2908,
"step": 768
},
{
"epoch": 3.1149367088607596,
"grad_norm": 0.2843616226813786,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.2851,
"step": 769
},
{
"epoch": 3.118987341772152,
"grad_norm": 0.23886727506480138,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.2882,
"step": 770
},
{
"epoch": 3.123037974683544,
"grad_norm": 0.25867637177687597,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.2983,
"step": 771
},
{
"epoch": 3.1270886075949367,
"grad_norm": 0.2394489678803083,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.2794,
"step": 772
},
{
"epoch": 3.131139240506329,
"grad_norm": 0.2567621458687313,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.2863,
"step": 773
},
{
"epoch": 3.1351898734177217,
"grad_norm": 0.24823271488580953,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.2843,
"step": 774
},
{
"epoch": 3.1392405063291138,
"grad_norm": 0.24709490370663198,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.282,
"step": 775
},
{
"epoch": 3.1432911392405063,
"grad_norm": 0.230630042652721,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.2908,
"step": 776
},
{
"epoch": 3.147341772151899,
"grad_norm": 0.26031931292445587,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.2861,
"step": 777
},
{
"epoch": 3.1513924050632913,
"grad_norm": 0.23205540753837636,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.2862,
"step": 778
},
{
"epoch": 3.1554430379746834,
"grad_norm": 0.23533451414284243,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.2905,
"step": 779
},
{
"epoch": 3.159493670886076,
"grad_norm": 0.22677046027644715,
"learning_rate": 1.420958430225303e-05,
"loss": 0.2907,
"step": 780
},
{
"epoch": 3.1635443037974684,
"grad_norm": 0.24466070797109413,
"learning_rate": 1.415527989393985e-05,
"loss": 0.2947,
"step": 781
},
{
"epoch": 3.167594936708861,
"grad_norm": 0.23899456463976343,
"learning_rate": 1.410102255819891e-05,
"loss": 0.2988,
"step": 782
},
{
"epoch": 3.171645569620253,
"grad_norm": 0.24491121048512224,
"learning_rate": 1.404681273201131e-05,
"loss": 0.2845,
"step": 783
},
{
"epoch": 3.1756962025316455,
"grad_norm": 0.24961991943879333,
"learning_rate": 1.399265085197556e-05,
"loss": 0.2881,
"step": 784
},
{
"epoch": 3.179746835443038,
"grad_norm": 0.24520127371999784,
"learning_rate": 1.393853735430398e-05,
"loss": 0.2866,
"step": 785
},
{
"epoch": 3.1837974683544306,
"grad_norm": 0.21978029931459328,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.2883,
"step": 786
},
{
"epoch": 3.1878481012658226,
"grad_norm": 0.2289108799060373,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.2808,
"step": 787
},
{
"epoch": 3.191898734177215,
"grad_norm": 0.227553757579659,
"learning_rate": 1.377649151173163e-05,
"loss": 0.2836,
"step": 788
},
{
"epoch": 3.1959493670886077,
"grad_norm": 0.21194653491154936,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.2784,
"step": 789
},
{
"epoch": 3.2,
"grad_norm": 0.2438795249274544,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.2768,
"step": 790
},
{
"epoch": 3.2040506329113922,
"grad_norm": 0.21757996891187387,
"learning_rate": 1.361489677627324e-05,
"loss": 0.2849,
"step": 791
},
{
"epoch": 3.2081012658227848,
"grad_norm": 0.23247813594369454,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.2839,
"step": 792
},
{
"epoch": 3.2121518987341773,
"grad_norm": 0.23269175416419738,
"learning_rate": 1.350742335331241e-05,
"loss": 0.289,
"step": 793
},
{
"epoch": 3.21620253164557,
"grad_norm": 0.22642791788110248,
"learning_rate": 1.345376486102198e-05,
"loss": 0.2983,
"step": 794
},
{
"epoch": 3.220253164556962,
"grad_norm": 0.23932561067486613,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.2957,
"step": 795
},
{
"epoch": 3.2243037974683544,
"grad_norm": 0.24032255635096197,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.2829,
"step": 796
},
{
"epoch": 3.228354430379747,
"grad_norm": 0.24420464952783222,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.2825,
"step": 797
},
{
"epoch": 3.2324050632911394,
"grad_norm": 0.21659438205803125,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.2862,
"step": 798
},
{
"epoch": 3.2364556962025315,
"grad_norm": 0.2503255215089602,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.2945,
"step": 799
},
{
"epoch": 3.240506329113924,
"grad_norm": 0.24088260380585452,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.2939,
"step": 800
},
{
"epoch": 3.2445569620253165,
"grad_norm": 0.278661647870566,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.2907,
"step": 801
},
{
"epoch": 3.248607594936709,
"grad_norm": 0.24231510014065025,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.2853,
"step": 802
},
{
"epoch": 3.252658227848101,
"grad_norm": 0.2459669700034695,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.2976,
"step": 803
},
{
"epoch": 3.2567088607594936,
"grad_norm": 0.26615294763992425,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.2893,
"step": 804
},
{
"epoch": 3.260759493670886,
"grad_norm": 0.22550571932469646,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.2884,
"step": 805
},
{
"epoch": 3.2648101265822786,
"grad_norm": 0.21956369201395065,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.2769,
"step": 806
},
{
"epoch": 3.2688607594936707,
"grad_norm": 0.22674343642820993,
"learning_rate": 1.276115892225764e-05,
"loss": 0.2939,
"step": 807
},
{
"epoch": 3.2729113924050632,
"grad_norm": 0.20930891418204697,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.2882,
"step": 808
},
{
"epoch": 3.2769620253164558,
"grad_norm": 0.22862191685093022,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.2955,
"step": 809
},
{
"epoch": 3.2810126582278483,
"grad_norm": 0.20070396595598147,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.2841,
"step": 810
},
{
"epoch": 3.2850632911392403,
"grad_norm": 0.2228258723657111,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.29,
"step": 811
},
{
"epoch": 3.289113924050633,
"grad_norm": 0.21558004486807106,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.2969,
"step": 812
},
{
"epoch": 3.2931645569620254,
"grad_norm": 0.22728730076346948,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.2967,
"step": 813
},
{
"epoch": 3.297215189873418,
"grad_norm": 0.22564452403158208,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.282,
"step": 814
},
{
"epoch": 3.30126582278481,
"grad_norm": 0.19552523809847108,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.2907,
"step": 815
},
{
"epoch": 3.3053164556962025,
"grad_norm": 0.20773508325586099,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.2809,
"step": 816
},
{
"epoch": 3.309367088607595,
"grad_norm": 0.216687152070374,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.2844,
"step": 817
},
{
"epoch": 3.3134177215189875,
"grad_norm": 0.2104372219951085,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.2882,
"step": 818
},
{
"epoch": 3.3174683544303796,
"grad_norm": 0.20763550832735847,
"learning_rate": 1.213055323844297e-05,
"loss": 0.2746,
"step": 819
},
{
"epoch": 3.321518987341772,
"grad_norm": 0.2321951731304462,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.284,
"step": 820
},
{
"epoch": 3.3255696202531646,
"grad_norm": 0.20823629229919685,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.2925,
"step": 821
},
{
"epoch": 3.329620253164557,
"grad_norm": 0.20581147484920675,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.2988,
"step": 822
},
{
"epoch": 3.333670886075949,
"grad_norm": 0.21732412322263192,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.2839,
"step": 823
},
{
"epoch": 3.3377215189873417,
"grad_norm": 0.217497939537192,
"learning_rate": 1.187045262895488e-05,
"loss": 0.2926,
"step": 824
},
{
"epoch": 3.3417721518987342,
"grad_norm": 0.20851150365891796,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.2851,
"step": 825
},
{
"epoch": 3.3458227848101267,
"grad_norm": 0.20254906333526965,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.288,
"step": 826
},
{
"epoch": 3.349873417721519,
"grad_norm": 0.2234091860062816,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.2777,
"step": 827
},
{
"epoch": 3.3539240506329113,
"grad_norm": 0.22913473350457303,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.291,
"step": 828
},
{
"epoch": 3.357974683544304,
"grad_norm": 0.21839741337992352,
"learning_rate": 1.161198884952377e-05,
"loss": 0.2872,
"step": 829
},
{
"epoch": 3.3620253164556964,
"grad_norm": 0.23304749591955606,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.2957,
"step": 830
},
{
"epoch": 3.3660759493670884,
"grad_norm": 0.22088813155430967,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.2828,
"step": 831
},
{
"epoch": 3.370126582278481,
"grad_norm": 0.20970946825838624,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.2734,
"step": 832
},
{
"epoch": 3.3741772151898735,
"grad_norm": 0.24941124937721768,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.2962,
"step": 833
},
{
"epoch": 3.378227848101266,
"grad_norm": 0.21108653676014,
"learning_rate": 1.135521394010506e-05,
"loss": 0.2801,
"step": 834
},
{
"epoch": 3.382278481012658,
"grad_norm": 0.21999135204239417,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.2949,
"step": 835
},
{
"epoch": 3.3863291139240506,
"grad_norm": 0.2375213445079912,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.2839,
"step": 836
},
{
"epoch": 3.390379746835443,
"grad_norm": 0.21501711882397329,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.2862,
"step": 837
},
{
"epoch": 3.3944303797468356,
"grad_norm": 0.22592813895732008,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.2806,
"step": 838
},
{
"epoch": 3.3984810126582277,
"grad_norm": 0.22415082927505794,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.2979,
"step": 839
},
{
"epoch": 3.40253164556962,
"grad_norm": 0.2030553478952714,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.2794,
"step": 840
},
{
"epoch": 3.4065822784810127,
"grad_norm": 0.20243514632679677,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.2889,
"step": 841
},
{
"epoch": 3.410632911392405,
"grad_norm": 0.21902768604549688,
"learning_rate": 1.0948015846744e-05,
"loss": 0.2865,
"step": 842
},
{
"epoch": 3.4146835443037973,
"grad_norm": 0.1911816411916076,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.29,
"step": 843
},
{
"epoch": 3.41873417721519,
"grad_norm": 0.21814632367129166,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.29,
"step": 844
},
{
"epoch": 3.4227848101265823,
"grad_norm": 0.19975677320192708,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.2809,
"step": 845
},
{
"epoch": 3.426835443037975,
"grad_norm": 0.20315271022516787,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.2842,
"step": 846
},
{
"epoch": 3.430886075949367,
"grad_norm": 0.20508833239123136,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.2806,
"step": 847
},
{
"epoch": 3.4349367088607594,
"grad_norm": 0.2138781256422824,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.2832,
"step": 848
},
{
"epoch": 3.438987341772152,
"grad_norm": 0.2234202671066357,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.2819,
"step": 849
},
{
"epoch": 3.4430379746835444,
"grad_norm": 0.20806558159247346,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.2911,
"step": 850
},
{
"epoch": 3.4470886075949365,
"grad_norm": 0.20656291632191376,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.2832,
"step": 851
},
{
"epoch": 3.451139240506329,
"grad_norm": 0.24073758049263555,
"learning_rate": 1.044560362975474e-05,
"loss": 0.2957,
"step": 852
},
{
"epoch": 3.4551898734177215,
"grad_norm": 0.21242672196521287,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.2943,
"step": 853
},
{
"epoch": 3.459240506329114,
"grad_norm": 0.21082739514774118,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.297,
"step": 854
},
{
"epoch": 3.463291139240506,
"grad_norm": 0.2164858452482678,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.2944,
"step": 855
},
{
"epoch": 3.4673417721518986,
"grad_norm": 0.21434179164018755,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.2974,
"step": 856
},
{
"epoch": 3.471392405063291,
"grad_norm": 0.20864357776596912,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.298,
"step": 857
},
{
"epoch": 3.4754430379746837,
"grad_norm": 0.20748270177666833,
"learning_rate": 1.014782400334433e-05,
"loss": 0.2858,
"step": 858
},
{
"epoch": 3.479493670886076,
"grad_norm": 0.2140559779975167,
"learning_rate": 1.009846945771296e-05,
"loss": 0.2887,
"step": 859
},
{
"epoch": 3.4835443037974683,
"grad_norm": 0.20130698224131413,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.2911,
"step": 860
},
{
"epoch": 3.487594936708861,
"grad_norm": 0.19878933290083703,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.2818,
"step": 861
},
{
"epoch": 3.4916455696202533,
"grad_norm": 0.21195836653616051,
"learning_rate": 9.950885880977891e-06,
"loss": 0.278,
"step": 862
},
{
"epoch": 3.4956962025316454,
"grad_norm": 0.2052761129913914,
"learning_rate": 9.901852696139382e-06,
"loss": 0.2897,
"step": 863
},
{
"epoch": 3.499746835443038,
"grad_norm": 0.4473095473586543,
"learning_rate": 9.852900840391027e-06,
"loss": 0.2879,
"step": 864
},
{
"epoch": 3.5037974683544304,
"grad_norm": 0.22269614814293334,
"learning_rate": 9.804030707984313e-06,
"loss": 0.287,
"step": 865
},
{
"epoch": 3.507848101265823,
"grad_norm": 0.19634628482547706,
"learning_rate": 9.755242692512599e-06,
"loss": 0.276,
"step": 866
},
{
"epoch": 3.5118987341772154,
"grad_norm": 0.23394481883954685,
"learning_rate": 9.70653718690782e-06,
"loss": 0.2853,
"step": 867
},
{
"epoch": 3.5159493670886075,
"grad_norm": 0.2471347710963604,
"learning_rate": 9.657914583437454e-06,
"loss": 0.2912,
"step": 868
},
{
"epoch": 3.52,
"grad_norm": 0.21215229797931662,
"learning_rate": 9.609375273701246e-06,
"loss": 0.294,
"step": 869
},
{
"epoch": 3.5240506329113925,
"grad_norm": 0.2584054600474304,
"learning_rate": 9.560919648628133e-06,
"loss": 0.2914,
"step": 870
},
{
"epoch": 3.5281012658227846,
"grad_norm": 0.20040468629930014,
"learning_rate": 9.512548098473047e-06,
"loss": 0.2834,
"step": 871
},
{
"epoch": 3.532151898734177,
"grad_norm": 0.21599135685082027,
"learning_rate": 9.464261012813825e-06,
"loss": 0.2821,
"step": 872
},
{
"epoch": 3.5362025316455696,
"grad_norm": 0.235216264217567,
"learning_rate": 9.416058780547987e-06,
"loss": 0.2934,
"step": 873
},
{
"epoch": 3.540253164556962,
"grad_norm": 0.21685272677768852,
"learning_rate": 9.367941789889714e-06,
"loss": 0.281,
"step": 874
},
{
"epoch": 3.5443037974683547,
"grad_norm": 0.2187074702086221,
"learning_rate": 9.319910428366607e-06,
"loss": 0.291,
"step": 875
},
{
"epoch": 3.5483544303797467,
"grad_norm": 0.20353491110914884,
"learning_rate": 9.271965082816667e-06,
"loss": 0.2996,
"step": 876
},
{
"epoch": 3.5524050632911393,
"grad_norm": 0.21385613819368743,
"learning_rate": 9.224106139385111e-06,
"loss": 0.2901,
"step": 877
},
{
"epoch": 3.5564556962025318,
"grad_norm": 0.22096161618852916,
"learning_rate": 9.176333983521291e-06,
"loss": 0.2942,
"step": 878
},
{
"epoch": 3.560506329113924,
"grad_norm": 0.22892682741384687,
"learning_rate": 9.12864899997558e-06,
"loss": 0.2809,
"step": 879
},
{
"epoch": 3.5645569620253164,
"grad_norm": 0.21431812608567072,
"learning_rate": 9.08105157279628e-06,
"loss": 0.2893,
"step": 880
},
{
"epoch": 3.568607594936709,
"grad_norm": 0.21722551589999328,
"learning_rate": 9.03354208532653e-06,
"loss": 0.2836,
"step": 881
},
{
"epoch": 3.5726582278481014,
"grad_norm": 0.21489925634107865,
"learning_rate": 8.986120920201205e-06,
"loss": 0.2991,
"step": 882
},
{
"epoch": 3.576708860759494,
"grad_norm": 0.2045218248004575,
"learning_rate": 8.938788459343852e-06,
"loss": 0.2924,
"step": 883
},
{
"epoch": 3.580759493670886,
"grad_norm": 0.2053035611150445,
"learning_rate": 8.8915450839636e-06,
"loss": 0.2903,
"step": 884
},
{
"epoch": 3.5848101265822785,
"grad_norm": 0.21160109226457108,
"learning_rate": 8.844391174552116e-06,
"loss": 0.2908,
"step": 885
},
{
"epoch": 3.588860759493671,
"grad_norm": 0.2131220760556764,
"learning_rate": 8.797327110880479e-06,
"loss": 0.2828,
"step": 886
},
{
"epoch": 3.592911392405063,
"grad_norm": 0.20701924963522986,
"learning_rate": 8.750353271996206e-06,
"loss": 0.2913,
"step": 887
},
{
"epoch": 3.5969620253164556,
"grad_norm": 0.2097738383963327,
"learning_rate": 8.703470036220132e-06,
"loss": 0.2854,
"step": 888
},
{
"epoch": 3.601012658227848,
"grad_norm": 0.21319762634396824,
"learning_rate": 8.656677781143394e-06,
"loss": 0.2911,
"step": 889
},
{
"epoch": 3.6050632911392406,
"grad_norm": 0.21537131273150592,
"learning_rate": 8.609976883624377e-06,
"loss": 0.2871,
"step": 890
},
{
"epoch": 3.609113924050633,
"grad_norm": 0.22051967127016378,
"learning_rate": 8.563367719785698e-06,
"loss": 0.2938,
"step": 891
},
{
"epoch": 3.613164556962025,
"grad_norm": 0.20776925271372784,
"learning_rate": 8.516850665011138e-06,
"loss": 0.2848,
"step": 892
},
{
"epoch": 3.6172151898734177,
"grad_norm": 0.2115070756303823,
"learning_rate": 8.47042609394269e-06,
"loss": 0.2834,
"step": 893
},
{
"epoch": 3.6212658227848102,
"grad_norm": 0.22102928224126875,
"learning_rate": 8.424094380477432e-06,
"loss": 0.289,
"step": 894
},
{
"epoch": 3.6253164556962023,
"grad_norm": 0.1999359264954731,
"learning_rate": 8.37785589776465e-06,
"loss": 0.292,
"step": 895
},
{
"epoch": 3.629367088607595,
"grad_norm": 0.21240096025740957,
"learning_rate": 8.331711018202694e-06,
"loss": 0.2973,
"step": 896
},
{
"epoch": 3.6334177215189873,
"grad_norm": 0.21936645163103557,
"learning_rate": 8.285660113436104e-06,
"loss": 0.2845,
"step": 897
},
{
"epoch": 3.63746835443038,
"grad_norm": 0.2143159818866527,
"learning_rate": 8.239703554352527e-06,
"loss": 0.2889,
"step": 898
},
{
"epoch": 3.6415189873417724,
"grad_norm": 0.19514107900177027,
"learning_rate": 8.193841711079775e-06,
"loss": 0.2897,
"step": 899
},
{
"epoch": 3.6455696202531644,
"grad_norm": 0.22578495831311013,
"learning_rate": 8.148074952982828e-06,
"loss": 0.2857,
"step": 900
},
{
"epoch": 3.649620253164557,
"grad_norm": 0.2254046476885201,
"learning_rate": 8.102403648660859e-06,
"loss": 0.2879,
"step": 901
},
{
"epoch": 3.6536708860759495,
"grad_norm": 0.22151029165567068,
"learning_rate": 8.056828165944282e-06,
"loss": 0.2975,
"step": 902
},
{
"epoch": 3.6577215189873415,
"grad_norm": 0.19707956115533481,
"learning_rate": 8.011348871891762e-06,
"loss": 0.2828,
"step": 903
},
{
"epoch": 3.661772151898734,
"grad_norm": 0.18924761227672054,
"learning_rate": 7.965966132787287e-06,
"loss": 0.287,
"step": 904
},
{
"epoch": 3.6658227848101266,
"grad_norm": 0.22127393527601227,
"learning_rate": 7.920680314137189e-06,
"loss": 0.2846,
"step": 905
},
{
"epoch": 3.669873417721519,
"grad_norm": 0.21373718052739432,
"learning_rate": 7.875491780667246e-06,
"loss": 0.2856,
"step": 906
},
{
"epoch": 3.6739240506329116,
"grad_norm": 0.21451406493840353,
"learning_rate": 7.830400896319667e-06,
"loss": 0.2854,
"step": 907
},
{
"epoch": 3.6779746835443037,
"grad_norm": 0.22681682768253209,
"learning_rate": 7.785408024250259e-06,
"loss": 0.2894,
"step": 908
},
{
"epoch": 3.682025316455696,
"grad_norm": 0.22825171144478593,
"learning_rate": 7.74051352682542e-06,
"loss": 0.2857,
"step": 909
},
{
"epoch": 3.6860759493670887,
"grad_norm": 0.2182325681207031,
"learning_rate": 7.695717765619257e-06,
"loss": 0.2973,
"step": 910
},
{
"epoch": 3.690126582278481,
"grad_norm": 0.23295632634523386,
"learning_rate": 7.651021101410673e-06,
"loss": 0.2955,
"step": 911
},
{
"epoch": 3.6941772151898733,
"grad_norm": 0.18685974683126488,
"learning_rate": 7.606423894180464e-06,
"loss": 0.2931,
"step": 912
},
{
"epoch": 3.698227848101266,
"grad_norm": 0.2101637763992384,
"learning_rate": 7.56192650310839e-06,
"loss": 0.2836,
"step": 913
},
{
"epoch": 3.7022784810126583,
"grad_norm": 0.20920107638225885,
"learning_rate": 7.517529286570349e-06,
"loss": 0.286,
"step": 914
},
{
"epoch": 3.706329113924051,
"grad_norm": 0.19583553311256843,
"learning_rate": 7.473232602135387e-06,
"loss": 0.2853,
"step": 915
},
{
"epoch": 3.710379746835443,
"grad_norm": 0.21275160153356698,
"learning_rate": 7.429036806562935e-06,
"loss": 0.2834,
"step": 916
},
{
"epoch": 3.7144303797468354,
"grad_norm": 0.20885232373695578,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.2874,
"step": 917
},
{
"epoch": 3.718481012658228,
"grad_norm": 0.20426934025179935,
"learning_rate": 7.340949304977567e-06,
"loss": 0.2865,
"step": 918
},
{
"epoch": 3.72253164556962,
"grad_norm": 0.18606850846295084,
"learning_rate": 7.297058308409282e-06,
"loss": 0.2967,
"step": 919
},
{
"epoch": 3.7265822784810125,
"grad_norm": 0.21428185628867788,
"learning_rate": 7.25326961958704e-06,
"loss": 0.297,
"step": 920
},
{
"epoch": 3.730632911392405,
"grad_norm": 0.22676197511468604,
"learning_rate": 7.209583591178921e-06,
"loss": 0.2923,
"step": 921
},
{
"epoch": 3.7346835443037976,
"grad_norm": 0.19692566395103808,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.3006,
"step": 922
},
{
"epoch": 3.73873417721519,
"grad_norm": 0.21837106715598953,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.2904,
"step": 923
},
{
"epoch": 3.742784810126582,
"grad_norm": 0.21616497895304287,
"learning_rate": 7.079144982700909e-06,
"loss": 0.2888,
"step": 924
},
{
"epoch": 3.7468354430379747,
"grad_norm": 0.18828645091963458,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.28,
"step": 925
},
{
"epoch": 3.750886075949367,
"grad_norm": 0.2028762538829359,
"learning_rate": 6.99270564069757e-06,
"loss": 0.292,
"step": 926
},
{
"epoch": 3.7549367088607593,
"grad_norm": 0.21302071652686078,
"learning_rate": 6.949642934304375e-06,
"loss": 0.2842,
"step": 927
},
{
"epoch": 3.7589873417721518,
"grad_norm": 0.1953944161379615,
"learning_rate": 6.906685333692871e-06,
"loss": 0.2842,
"step": 928
},
{
"epoch": 3.7630379746835443,
"grad_norm": 0.19505946689433437,
"learning_rate": 6.86383318483769e-06,
"loss": 0.2908,
"step": 929
},
{
"epoch": 3.767088607594937,
"grad_norm": 0.20072262776351332,
"learning_rate": 6.821086832864139e-06,
"loss": 0.2912,
"step": 930
},
{
"epoch": 3.7711392405063293,
"grad_norm": 0.1976104022592647,
"learning_rate": 6.77844662204546e-06,
"loss": 0.2995,
"step": 931
},
{
"epoch": 3.7751898734177214,
"grad_norm": 0.18991923896876084,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.2827,
"step": 932
},
{
"epoch": 3.779240506329114,
"grad_norm": 0.19347249871691966,
"learning_rate": 6.693485996688695e-06,
"loss": 0.2808,
"step": 933
},
{
"epoch": 3.7832911392405064,
"grad_norm": 0.1887767616233638,
"learning_rate": 6.651166266411801e-06,
"loss": 0.291,
"step": 934
},
{
"epoch": 3.7873417721518985,
"grad_norm": 0.18473433801120895,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.2867,
"step": 935
},
{
"epoch": 3.791392405063291,
"grad_norm": 0.20948619091248502,
"learning_rate": 6.566849674844711e-06,
"loss": 0.2738,
"step": 936
},
{
"epoch": 3.7954430379746835,
"grad_norm": 0.2014852460921327,
"learning_rate": 6.524853492628747e-06,
"loss": 0.285,
"step": 937
},
{
"epoch": 3.799493670886076,
"grad_norm": 0.1865419471334429,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.2873,
"step": 938
},
{
"epoch": 3.8035443037974686,
"grad_norm": 0.18861003047549588,
"learning_rate": 6.441187046486648e-06,
"loss": 0.2881,
"step": 939
},
{
"epoch": 3.8075949367088606,
"grad_norm": 0.20611365158553607,
"learning_rate": 6.399517456398567e-06,
"loss": 0.2834,
"step": 940
},
{
"epoch": 3.811645569620253,
"grad_norm": 0.2102795583984975,
"learning_rate": 6.357957402727164e-06,
"loss": 0.292,
"step": 941
},
{
"epoch": 3.8156962025316457,
"grad_norm": 0.18116769973646324,
"learning_rate": 6.316507220191395e-06,
"loss": 0.2927,
"step": 942
},
{
"epoch": 3.8197468354430377,
"grad_norm": 0.2171506109013425,
"learning_rate": 6.275167242625331e-06,
"loss": 0.2976,
"step": 943
},
{
"epoch": 3.8237974683544302,
"grad_norm": 0.21126315376887142,
"learning_rate": 6.233937802975471e-06,
"loss": 0.2832,
"step": 944
},
{
"epoch": 3.8278481012658228,
"grad_norm": 0.1836219074917592,
"learning_rate": 6.192819233298046e-06,
"loss": 0.2826,
"step": 945
},
{
"epoch": 3.8318987341772153,
"grad_norm": 0.18932330888799873,
"learning_rate": 6.151811864756383e-06,
"loss": 0.2805,
"step": 946
},
{
"epoch": 3.835949367088608,
"grad_norm": 0.21704333609784457,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.2742,
"step": 947
},
{
"epoch": 3.84,
"grad_norm": 0.20678076084214145,
"learning_rate": 6.070132051252868e-06,
"loss": 0.2909,
"step": 948
},
{
"epoch": 3.8440506329113924,
"grad_norm": 0.18038825956757615,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.2898,
"step": 949
},
{
"epoch": 3.848101265822785,
"grad_norm": 0.19780590291422845,
"learning_rate": 5.988900993811575e-06,
"loss": 0.2837,
"step": 950
},
{
"epoch": 3.852151898734177,
"grad_norm": 0.20044743070985444,
"learning_rate": 5.948454566959363e-06,
"loss": 0.2929,
"step": 951
},
{
"epoch": 3.8562025316455695,
"grad_norm": 0.22055073946049614,
"learning_rate": 5.908121309322328e-06,
"loss": 0.2942,
"step": 952
},
{
"epoch": 3.860253164556962,
"grad_norm": 0.20287403256216807,
"learning_rate": 5.867901545738976e-06,
"loss": 0.2905,
"step": 953
},
{
"epoch": 3.8643037974683545,
"grad_norm": 0.21541315684095588,
"learning_rate": 5.827795600133774e-06,
"loss": 0.2904,
"step": 954
},
{
"epoch": 3.868354430379747,
"grad_norm": 0.20534501043442335,
"learning_rate": 5.787803795514466e-06,
"loss": 0.281,
"step": 955
},
{
"epoch": 3.872405063291139,
"grad_norm": 0.19081492784836288,
"learning_rate": 5.747926453969576e-06,
"loss": 0.2917,
"step": 956
},
{
"epoch": 3.8764556962025316,
"grad_norm": 0.21272295892621443,
"learning_rate": 5.708163896665708e-06,
"loss": 0.2845,
"step": 957
},
{
"epoch": 3.880506329113924,
"grad_norm": 0.19656173327829787,
"learning_rate": 5.668516443845047e-06,
"loss": 0.2774,
"step": 958
},
{
"epoch": 3.884556962025316,
"grad_norm": 0.19262729283186422,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.2809,
"step": 959
},
{
"epoch": 3.8886075949367087,
"grad_norm": 0.20335761342778766,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.2915,
"step": 960
},
{
"epoch": 3.8926582278481012,
"grad_norm": 0.19925282944916692,
"learning_rate": 5.550267900783019e-06,
"loss": 0.2934,
"step": 961
},
{
"epoch": 3.8967088607594937,
"grad_norm": 0.19232523652943442,
"learning_rate": 5.511084049737623e-06,
"loss": 0.2845,
"step": 962
},
{
"epoch": 3.9007594936708863,
"grad_norm": 0.19058228579182784,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.2839,
"step": 963
},
{
"epoch": 3.9048101265822783,
"grad_norm": 0.19166528093580462,
"learning_rate": 5.433066737499948e-06,
"loss": 0.291,
"step": 964
},
{
"epoch": 3.908860759493671,
"grad_norm": 0.20911645315411603,
"learning_rate": 5.394233904648376e-06,
"loss": 0.2912,
"step": 965
},
{
"epoch": 3.9129113924050634,
"grad_norm": 0.19552599247791805,
"learning_rate": 5.355518704628997e-06,
"loss": 0.2881,
"step": 966
},
{
"epoch": 3.9169620253164554,
"grad_norm": 0.17272531189233795,
"learning_rate": 5.316921449248731e-06,
"loss": 0.2886,
"step": 967
},
{
"epoch": 3.921012658227848,
"grad_norm": 0.20603702043838312,
"learning_rate": 5.278442449364538e-06,
"loss": 0.2941,
"step": 968
},
{
"epoch": 3.9250632911392405,
"grad_norm": 0.1795870317291915,
"learning_rate": 5.240082014881016e-06,
"loss": 0.2754,
"step": 969
},
{
"epoch": 3.929113924050633,
"grad_norm": 0.23022799178989034,
"learning_rate": 5.201840454747822e-06,
"loss": 0.3052,
"step": 970
},
{
"epoch": 3.9331645569620255,
"grad_norm": 0.19219637433968242,
"learning_rate": 5.163718076957223e-06,
"loss": 0.2763,
"step": 971
},
{
"epoch": 3.9372151898734176,
"grad_norm": 0.18746577457297972,
"learning_rate": 5.125715188541609e-06,
"loss": 0.2761,
"step": 972
},
{
"epoch": 3.94126582278481,
"grad_norm": 0.19417478572602392,
"learning_rate": 5.087832095571021e-06,
"loss": 0.284,
"step": 973
},
{
"epoch": 3.9453164556962026,
"grad_norm": 0.21425877448611488,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.2879,
"step": 974
},
{
"epoch": 3.9493670886075947,
"grad_norm": 0.19436764761399006,
"learning_rate": 5.01242651541854e-06,
"loss": 0.2923,
"step": 975
},
{
"epoch": 3.953417721518987,
"grad_norm": 0.1857756218202181,
"learning_rate": 4.974904635542815e-06,
"loss": 0.2836,
"step": 976
},
{
"epoch": 3.9574683544303797,
"grad_norm": 0.19089454744165268,
"learning_rate": 4.937503765719582e-06,
"loss": 0.2925,
"step": 977
},
{
"epoch": 3.961518987341772,
"grad_norm": 0.2019041215889873,
"learning_rate": 4.900224207170299e-06,
"loss": 0.2915,
"step": 978
},
{
"epoch": 3.9655696202531647,
"grad_norm": 0.20466355634050235,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.2893,
"step": 979
},
{
"epoch": 3.969620253164557,
"grad_norm": 0.1904852341563188,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.2876,
"step": 980
},
{
"epoch": 3.9736708860759493,
"grad_norm": 0.1890040739732858,
"learning_rate": 4.789116396710924e-06,
"loss": 0.2912,
"step": 981
},
{
"epoch": 3.977721518987342,
"grad_norm": 0.19286242339651455,
"learning_rate": 4.752325075895368e-06,
"loss": 0.2844,
"step": 982
},
{
"epoch": 3.981772151898734,
"grad_norm": 0.19926520777099832,
"learning_rate": 4.715656557757473e-06,
"loss": 0.2934,
"step": 983
},
{
"epoch": 3.9858227848101264,
"grad_norm": 0.20085466538061034,
"learning_rate": 4.679111137620442e-06,
"loss": 0.2938,
"step": 984
},
{
"epoch": 3.989873417721519,
"grad_norm": 0.19010018407306975,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.2928,
"step": 985
},
{
"epoch": 3.9939240506329114,
"grad_norm": 0.19308717360630923,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.2773,
"step": 986
},
{
"epoch": 3.997974683544304,
"grad_norm": 0.188778993819026,
"learning_rate": 4.570216403561141e-06,
"loss": 0.2976,
"step": 987
},
{
"epoch": 4.002025316455696,
"grad_norm": 0.24545009805528187,
"learning_rate": 4.534166308795815e-06,
"loss": 0.2553,
"step": 988
},
{
"epoch": 4.006075949367089,
"grad_norm": 0.3165060610490153,
"learning_rate": 4.498240773728859e-06,
"loss": 0.2535,
"step": 989
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.22857941902949006,
"learning_rate": 4.462440087699609e-06,
"loss": 0.2423,
"step": 990
},
{
"epoch": 4.014177215189873,
"grad_norm": 0.39693182724890114,
"learning_rate": 4.426764539041861e-06,
"loss": 0.2483,
"step": 991
},
{
"epoch": 4.018227848101266,
"grad_norm": 0.30442915012846167,
"learning_rate": 4.391214415081582e-06,
"loss": 0.258,
"step": 992
},
{
"epoch": 4.022278481012658,
"grad_norm": 0.20105626712739832,
"learning_rate": 4.355790002134579e-06,
"loss": 0.2533,
"step": 993
},
{
"epoch": 4.02632911392405,
"grad_norm": 0.29767025509726414,
"learning_rate": 4.320491585504207e-06,
"loss": 0.2577,
"step": 994
},
{
"epoch": 4.030379746835443,
"grad_norm": 0.31467561977224606,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.2511,
"step": 995
},
{
"epoch": 4.034430379746835,
"grad_norm": 0.22989294383323022,
"learning_rate": 4.250273877330691e-06,
"loss": 0.2501,
"step": 996
},
{
"epoch": 4.038481012658228,
"grad_norm": 0.23223588764615827,
"learning_rate": 4.215355151311313e-06,
"loss": 0.2535,
"step": 997
},
{
"epoch": 4.04253164556962,
"grad_norm": 0.2836816358851813,
"learning_rate": 4.180563552651542e-06,
"loss": 0.2521,
"step": 998
},
{
"epoch": 4.046582278481012,
"grad_norm": 0.2245525612527193,
"learning_rate": 4.145899361558147e-06,
"loss": 0.2442,
"step": 999
},
{
"epoch": 4.050632911392405,
"grad_norm": 0.19866122664277233,
"learning_rate": 4.111362857211738e-06,
"loss": 0.2352,
"step": 1000
},
{
"epoch": 4.054683544303797,
"grad_norm": 0.22581884917246128,
"learning_rate": 4.076954317764592e-06,
"loss": 0.2472,
"step": 1001
},
{
"epoch": 4.0587341772151895,
"grad_norm": 0.20799588408604877,
"learning_rate": 4.042674020338335e-06,
"loss": 0.2521,
"step": 1002
},
{
"epoch": 4.062784810126582,
"grad_norm": 0.2144202904929127,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.2552,
"step": 1003
},
{
"epoch": 4.0668354430379745,
"grad_norm": 0.20901533971679725,
"learning_rate": 3.974499254868674e-06,
"loss": 0.2537,
"step": 1004
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.1890977240599995,
"learning_rate": 3.940605335895451e-06,
"loss": 0.2488,
"step": 1005
},
{
"epoch": 4.0749367088607595,
"grad_norm": 0.18668671039565018,
"learning_rate": 3.90684075707908e-06,
"loss": 0.2497,
"step": 1006
},
{
"epoch": 4.078987341772152,
"grad_norm": 0.21618079347355929,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.2591,
"step": 1007
},
{
"epoch": 4.083037974683545,
"grad_norm": 0.2058788288296464,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.2538,
"step": 1008
},
{
"epoch": 4.087088607594937,
"grad_norm": 0.18407367802359523,
"learning_rate": 3.806325775702304e-06,
"loss": 0.2522,
"step": 1009
},
{
"epoch": 4.091139240506329,
"grad_norm": 0.19755532351474556,
"learning_rate": 3.773081266416434e-06,
"loss": 0.2577,
"step": 1010
},
{
"epoch": 4.095189873417722,
"grad_norm": 0.19414419002446254,
"learning_rate": 3.739967446503245e-06,
"loss": 0.2484,
"step": 1011
},
{
"epoch": 4.099240506329114,
"grad_norm": 0.19318845847850008,
"learning_rate": 3.706984582656894e-06,
"loss": 0.2493,
"step": 1012
},
{
"epoch": 4.103291139240507,
"grad_norm": 0.197556349960637,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.2442,
"step": 1013
},
{
"epoch": 4.107341772151899,
"grad_norm": 0.19912470245066274,
"learning_rate": 3.641412784665648e-06,
"loss": 0.2578,
"step": 1014
},
{
"epoch": 4.111392405063291,
"grad_norm": 0.18614500819313465,
"learning_rate": 3.608824378627005e-06,
"loss": 0.2363,
"step": 1015
},
{
"epoch": 4.115443037974684,
"grad_norm": 0.19318526138420197,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.2474,
"step": 1016
},
{
"epoch": 4.119493670886076,
"grad_norm": 0.18292413104303967,
"learning_rate": 3.544043864774269e-06,
"loss": 0.2454,
"step": 1017
},
{
"epoch": 4.123544303797468,
"grad_norm": 0.18067452258324448,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.2603,
"step": 1018
},
{
"epoch": 4.127594936708861,
"grad_norm": 0.1797280719928127,
"learning_rate": 3.479793485887819e-06,
"loss": 0.2464,
"step": 1019
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.18417239336262564,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.254,
"step": 1020
},
{
"epoch": 4.135696202531646,
"grad_norm": 0.17970329926170256,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.248,
"step": 1021
},
{
"epoch": 4.139746835443038,
"grad_norm": 0.1961868197350999,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.2546,
"step": 1022
},
{
"epoch": 4.14379746835443,
"grad_norm": 0.1967443655605711,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.2539,
"step": 1023
},
{
"epoch": 4.147848101265823,
"grad_norm": 0.20443743015492924,
"learning_rate": 3.321500420335e-06,
"loss": 0.2464,
"step": 1024
},
{
"epoch": 4.151898734177215,
"grad_norm": 0.17506618523442538,
"learning_rate": 3.290243771741275e-06,
"loss": 0.2607,
"step": 1025
},
{
"epoch": 4.155949367088607,
"grad_norm": 0.18229228813073756,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.2585,
"step": 1026
},
{
"epoch": 4.16,
"grad_norm": 0.18453278720010155,
"learning_rate": 3.228134459445149e-06,
"loss": 0.2394,
"step": 1027
},
{
"epoch": 4.164050632911392,
"grad_norm": 0.18070678446725857,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.2539,
"step": 1028
},
{
"epoch": 4.168101265822785,
"grad_norm": 0.18134472342314822,
"learning_rate": 3.166565459258513e-06,
"loss": 0.2473,
"step": 1029
},
{
"epoch": 4.172151898734177,
"grad_norm": 0.18231652438356913,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.2473,
"step": 1030
},
{
"epoch": 4.176202531645569,
"grad_norm": 0.18604719101124864,
"learning_rate": 3.105538754650419e-06,
"loss": 0.2538,
"step": 1031
},
{
"epoch": 4.180253164556962,
"grad_norm": 0.19619643004531814,
"learning_rate": 3.07522937824712e-06,
"loss": 0.2631,
"step": 1032
},
{
"epoch": 4.184303797468354,
"grad_norm": 0.1793323703347004,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.2632,
"step": 1033
},
{
"epoch": 4.188354430379746,
"grad_norm": 0.18252395535391353,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.256,
"step": 1034
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.17866056572105105,
"learning_rate": 2.985120078631465e-06,
"loss": 0.2425,
"step": 1035
},
{
"epoch": 4.1964556962025314,
"grad_norm": 0.17823383828532774,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.2602,
"step": 1036
},
{
"epoch": 4.200506329113924,
"grad_norm": 0.18736541352444971,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.253,
"step": 1037
},
{
"epoch": 4.2045569620253165,
"grad_norm": 0.17784323589383358,
"learning_rate": 2.896244091927678e-06,
"loss": 0.2553,
"step": 1038
},
{
"epoch": 4.2086075949367086,
"grad_norm": 0.18650991539696488,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.2492,
"step": 1039
},
{
"epoch": 4.2126582278481015,
"grad_norm": 0.17938647387611165,
"learning_rate": 2.837681792951994e-06,
"loss": 0.2441,
"step": 1040
},
{
"epoch": 4.216708860759494,
"grad_norm": 0.17941763268567806,
"learning_rate": 2.808607860255981e-06,
"loss": 0.261,
"step": 1041
},
{
"epoch": 4.220759493670886,
"grad_norm": 0.17779645228415358,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.2519,
"step": 1042
},
{
"epoch": 4.224810126582279,
"grad_norm": 0.18444603441621515,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.2634,
"step": 1043
},
{
"epoch": 4.228860759493671,
"grad_norm": 0.17737799822616865,
"learning_rate": 2.722217735873718e-06,
"loss": 0.2508,
"step": 1044
},
{
"epoch": 4.232911392405064,
"grad_norm": 0.18028496878957073,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.2518,
"step": 1045
},
{
"epoch": 4.236962025316456,
"grad_norm": 0.18581781585946497,
"learning_rate": 2.665319697490205e-06,
"loss": 0.245,
"step": 1046
},
{
"epoch": 4.241012658227848,
"grad_norm": 0.18432483630194976,
"learning_rate": 2.637079980714945e-06,
"loss": 0.2534,
"step": 1047
},
{
"epoch": 4.245063291139241,
"grad_norm": 0.1855885581535757,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.2558,
"step": 1048
},
{
"epoch": 4.249113924050633,
"grad_norm": 0.18088691114285352,
"learning_rate": 2.581020289189571e-06,
"loss": 0.2482,
"step": 1049
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.17363655476825215,
"learning_rate": 2.553200765936501e-06,
"loss": 0.2642,
"step": 1050
},
{
"epoch": 4.257215189873418,
"grad_norm": 0.18267319008008556,
"learning_rate": 2.525521756805962e-06,
"loss": 0.2426,
"step": 1051
},
{
"epoch": 4.26126582278481,
"grad_norm": 0.18972122018079832,
"learning_rate": 2.497983484720885e-06,
"loss": 0.2564,
"step": 1052
},
{
"epoch": 4.265316455696203,
"grad_norm": 0.18388455126551653,
"learning_rate": 2.470586171470728e-06,
"loss": 0.245,
"step": 1053
},
{
"epoch": 4.269367088607595,
"grad_norm": 0.18440804468263017,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.2521,
"step": 1054
},
{
"epoch": 4.273417721518987,
"grad_norm": 0.18371031880095268,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.2606,
"step": 1055
},
{
"epoch": 4.27746835443038,
"grad_norm": 0.17927782989704294,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.2591,
"step": 1056
},
{
"epoch": 4.281518987341772,
"grad_norm": 0.17625060271820275,
"learning_rate": 2.362410902836978e-06,
"loss": 0.2514,
"step": 1057
},
{
"epoch": 4.285569620253165,
"grad_norm": 0.17671850457728433,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.2503,
"step": 1058
},
{
"epoch": 4.289620253164557,
"grad_norm": 0.18362973453087136,
"learning_rate": 2.309174704446411e-06,
"loss": 0.2541,
"step": 1059
},
{
"epoch": 4.293670886075949,
"grad_norm": 0.17862082293120035,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.2589,
"step": 1060
},
{
"epoch": 4.297721518987342,
"grad_norm": 0.17164648452413328,
"learning_rate": 2.256508422807855e-06,
"loss": 0.2409,
"step": 1061
},
{
"epoch": 4.301772151898734,
"grad_norm": 0.1762253544015869,
"learning_rate": 2.230389531696946e-06,
"loss": 0.2588,
"step": 1062
},
{
"epoch": 4.305822784810126,
"grad_norm": 0.18136285814904604,
"learning_rate": 2.204413754585857e-06,
"loss": 0.2458,
"step": 1063
},
{
"epoch": 4.309873417721519,
"grad_norm": 0.20124922797208614,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.2384,
"step": 1064
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.1751954482239485,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.2497,
"step": 1065
},
{
"epoch": 4.317974683544303,
"grad_norm": 0.17626285473120928,
"learning_rate": 2.127347193531757e-06,
"loss": 0.2469,
"step": 1066
},
{
"epoch": 4.322025316455696,
"grad_norm": 0.1802544081744511,
"learning_rate": 2.101945952921942e-06,
"loss": 0.249,
"step": 1067
},
{
"epoch": 4.326075949367088,
"grad_norm": 0.18063710832661364,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.244,
"step": 1068
},
{
"epoch": 4.330126582278481,
"grad_norm": 0.17429159208196626,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.2401,
"step": 1069
},
{
"epoch": 4.334177215189873,
"grad_norm": 0.1817926143834933,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.246,
"step": 1070
},
{
"epoch": 4.3382278481012655,
"grad_norm": 0.1781170365540815,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.2523,
"step": 1071
},
{
"epoch": 4.3422784810126585,
"grad_norm": 0.2314682253059228,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.2421,
"step": 1072
},
{
"epoch": 4.3463291139240505,
"grad_norm": 0.1931290185088278,
"learning_rate": 1.95257270594031e-06,
"loss": 0.243,
"step": 1073
},
{
"epoch": 4.3503797468354435,
"grad_norm": 0.17742487350558037,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.2466,
"step": 1074
},
{
"epoch": 4.3544303797468356,
"grad_norm": 0.18385737013544726,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.2473,
"step": 1075
},
{
"epoch": 4.358481012658228,
"grad_norm": 0.18558769525868646,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.2607,
"step": 1076
},
{
"epoch": 4.362531645569621,
"grad_norm": 0.17439546896744954,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.2601,
"step": 1077
},
{
"epoch": 4.366582278481013,
"grad_norm": 0.18834597751641827,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.2488,
"step": 1078
},
{
"epoch": 4.370632911392405,
"grad_norm": 0.1759138394736621,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.25,
"step": 1079
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.17187866977332603,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.2553,
"step": 1080
},
{
"epoch": 4.37873417721519,
"grad_norm": 0.16932399305033755,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.2613,
"step": 1081
},
{
"epoch": 4.382784810126582,
"grad_norm": 0.17984863571123402,
"learning_rate": 1.738336950434061e-06,
"loss": 0.2466,
"step": 1082
},
{
"epoch": 4.386835443037975,
"grad_norm": 0.17897411815518438,
"learning_rate": 1.715265841375957e-06,
"loss": 0.2562,
"step": 1083
},
{
"epoch": 4.390886075949367,
"grad_norm": 0.17280829754941845,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.2579,
"step": 1084
},
{
"epoch": 4.39493670886076,
"grad_norm": 0.1697684625463095,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.2465,
"step": 1085
},
{
"epoch": 4.398987341772152,
"grad_norm": 0.17014081877659284,
"learning_rate": 1.646936827927441e-06,
"loss": 0.2492,
"step": 1086
},
{
"epoch": 4.403037974683544,
"grad_norm": 0.17086230022134766,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.2524,
"step": 1087
},
{
"epoch": 4.407088607594937,
"grad_norm": 0.17483537783411784,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.2516,
"step": 1088
},
{
"epoch": 4.411139240506329,
"grad_norm": 0.1705305802910721,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.2479,
"step": 1089
},
{
"epoch": 4.415189873417722,
"grad_norm": 0.174612297216696,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.2523,
"step": 1090
},
{
"epoch": 4.419240506329114,
"grad_norm": 0.17501313825194115,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.2467,
"step": 1091
},
{
"epoch": 4.423291139240506,
"grad_norm": 0.17342775964308102,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.2456,
"step": 1092
},
{
"epoch": 4.427341772151899,
"grad_norm": 0.17017445867661404,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.2399,
"step": 1093
},
{
"epoch": 4.431392405063291,
"grad_norm": 0.1860435293945978,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.2522,
"step": 1094
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.18240720927614815,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.2631,
"step": 1095
},
{
"epoch": 4.439493670886076,
"grad_norm": 0.18682269560724557,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.249,
"step": 1096
},
{
"epoch": 4.443544303797468,
"grad_norm": 0.16849003653601258,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.2594,
"step": 1097
},
{
"epoch": 4.44759493670886,
"grad_norm": 0.1891256090236829,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.2621,
"step": 1098
},
{
"epoch": 4.451645569620253,
"grad_norm": 0.1772773298143795,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.2601,
"step": 1099
},
{
"epoch": 4.455696202531645,
"grad_norm": 0.18504348191701986,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.2502,
"step": 1100
},
{
"epoch": 4.459746835443038,
"grad_norm": 0.1762880078769516,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.2447,
"step": 1101
},
{
"epoch": 4.46379746835443,
"grad_norm": 0.17307385243133017,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.2429,
"step": 1102
},
{
"epoch": 4.467848101265822,
"grad_norm": 0.17866011581774716,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.2546,
"step": 1103
},
{
"epoch": 4.471898734177215,
"grad_norm": 0.1806184090696422,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.2615,
"step": 1104
},
{
"epoch": 4.4759493670886075,
"grad_norm": 0.179151683701198,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.242,
"step": 1105
},
{
"epoch": 4.48,
"grad_norm": 0.17585780696068282,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.2485,
"step": 1106
},
{
"epoch": 4.4840506329113925,
"grad_norm": 0.17852140667372612,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.253,
"step": 1107
},
{
"epoch": 4.488101265822785,
"grad_norm": 0.17736339835878362,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.2451,
"step": 1108
},
{
"epoch": 4.4921518987341775,
"grad_norm": 0.1807877462088156,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.2524,
"step": 1109
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.1696104931445503,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.2446,
"step": 1110
},
{
"epoch": 4.500253164556962,
"grad_norm": 0.17215290219558255,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.2498,
"step": 1111
},
{
"epoch": 4.504303797468355,
"grad_norm": 0.16710908329532845,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.2509,
"step": 1112
},
{
"epoch": 4.508354430379747,
"grad_norm": 0.2445145276193321,
"learning_rate": 1.092402023876933e-06,
"loss": 0.2498,
"step": 1113
},
{
"epoch": 4.512405063291139,
"grad_norm": 0.1691358778190376,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.2424,
"step": 1114
},
{
"epoch": 4.516455696202532,
"grad_norm": 0.1715419180322718,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.2641,
"step": 1115
},
{
"epoch": 4.520506329113924,
"grad_norm": 0.17304172151246866,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.2472,
"step": 1116
},
{
"epoch": 4.524556962025317,
"grad_norm": 0.16492876482192656,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.2569,
"step": 1117
},
{
"epoch": 4.528607594936709,
"grad_norm": 0.1766102988896944,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.245,
"step": 1118
},
{
"epoch": 4.532658227848101,
"grad_norm": 0.1668532092440531,
"learning_rate": 9.841383857091947e-07,
"loss": 0.253,
"step": 1119
},
{
"epoch": 4.536708860759494,
"grad_norm": 0.1660760768058434,
"learning_rate": 9.666296380321616e-07,
"loss": 0.2523,
"step": 1120
},
{
"epoch": 4.540759493670886,
"grad_norm": 0.1764490115928226,
"learning_rate": 9.492741825034124e-07,
"loss": 0.2529,
"step": 1121
},
{
"epoch": 4.544810126582279,
"grad_norm": 0.17038344914474965,
"learning_rate": 9.320721589013892e-07,
"loss": 0.2592,
"step": 1122
},
{
"epoch": 4.548860759493671,
"grad_norm": 0.1780447392199246,
"learning_rate": 9.150237057688339e-07,
"loss": 0.2466,
"step": 1123
},
{
"epoch": 4.552911392405063,
"grad_norm": 0.17922602114092923,
"learning_rate": 8.981289604116328e-07,
"loss": 0.2585,
"step": 1124
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.1711248232736432,
"learning_rate": 8.813880588977542e-07,
"loss": 0.2548,
"step": 1125
},
{
"epoch": 4.561012658227848,
"grad_norm": 0.18203979991369218,
"learning_rate": 8.648011360561126e-07,
"loss": 0.2624,
"step": 1126
},
{
"epoch": 4.56506329113924,
"grad_norm": 0.16886485008454272,
"learning_rate": 8.483683254755037e-07,
"loss": 0.251,
"step": 1127
},
{
"epoch": 4.569113924050633,
"grad_norm": 0.17191359979404644,
"learning_rate": 8.320897595035227e-07,
"loss": 0.2619,
"step": 1128
},
{
"epoch": 4.573164556962025,
"grad_norm": 0.17221893535237404,
"learning_rate": 8.159655692455093e-07,
"loss": 0.2595,
"step": 1129
},
{
"epoch": 4.577215189873417,
"grad_norm": 0.17169273068305263,
"learning_rate": 7.999958845634648e-07,
"loss": 0.2456,
"step": 1130
},
{
"epoch": 4.58126582278481,
"grad_norm": 0.17803500474530545,
"learning_rate": 7.841808340750478e-07,
"loss": 0.2558,
"step": 1131
},
{
"epoch": 4.585316455696202,
"grad_norm": 0.17566335495231616,
"learning_rate": 7.685205451524869e-07,
"loss": 0.2514,
"step": 1132
},
{
"epoch": 4.589367088607595,
"grad_norm": 0.16404062650933537,
"learning_rate": 7.530151439216027e-07,
"loss": 0.2513,
"step": 1133
},
{
"epoch": 4.593417721518987,
"grad_norm": 0.17254674849564575,
"learning_rate": 7.376647552607675e-07,
"loss": 0.2522,
"step": 1134
},
{
"epoch": 4.597468354430379,
"grad_norm": 0.16401031145886935,
"learning_rate": 7.224695027998963e-07,
"loss": 0.2598,
"step": 1135
},
{
"epoch": 4.601518987341772,
"grad_norm": 0.1652330865950714,
"learning_rate": 7.07429508919466e-07,
"loss": 0.2508,
"step": 1136
},
{
"epoch": 4.605569620253164,
"grad_norm": 0.17028372518653218,
"learning_rate": 6.925448947495206e-07,
"loss": 0.2536,
"step": 1137
},
{
"epoch": 4.609620253164557,
"grad_norm": 0.1693601955051676,
"learning_rate": 6.778157801686936e-07,
"loss": 0.2629,
"step": 1138
},
{
"epoch": 4.613670886075949,
"grad_norm": 0.1659456687863225,
"learning_rate": 6.632422838032515e-07,
"loss": 0.2557,
"step": 1139
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.17580700165160165,
"learning_rate": 6.488245230261281e-07,
"loss": 0.2491,
"step": 1140
},
{
"epoch": 4.6217721518987345,
"grad_norm": 0.16593036074969028,
"learning_rate": 6.345626139559868e-07,
"loss": 0.2633,
"step": 1141
},
{
"epoch": 4.6258227848101265,
"grad_norm": 0.16839679682194816,
"learning_rate": 6.204566714562866e-07,
"loss": 0.2426,
"step": 1142
},
{
"epoch": 4.629873417721519,
"grad_norm": 0.17110454177232412,
"learning_rate": 6.06506809134344e-07,
"loss": 0.2525,
"step": 1143
},
{
"epoch": 4.633924050632912,
"grad_norm": 0.17107026180222024,
"learning_rate": 5.927131393404373e-07,
"loss": 0.2465,
"step": 1144
},
{
"epoch": 4.637974683544304,
"grad_norm": 0.16863282007582941,
"learning_rate": 5.790757731668817e-07,
"loss": 0.2405,
"step": 1145
},
{
"epoch": 4.642025316455696,
"grad_norm": 0.16915336809894208,
"learning_rate": 5.655948204471507e-07,
"loss": 0.2647,
"step": 1146
},
{
"epoch": 4.646075949367089,
"grad_norm": 0.16384061298193003,
"learning_rate": 5.522703897549875e-07,
"loss": 0.2506,
"step": 1147
},
{
"epoch": 4.650126582278481,
"grad_norm": 0.16493713942208463,
"learning_rate": 5.391025884035239e-07,
"loss": 0.2532,
"step": 1148
},
{
"epoch": 4.654177215189874,
"grad_norm": 0.1685898255941286,
"learning_rate": 5.260915224444207e-07,
"loss": 0.2478,
"step": 1149
},
{
"epoch": 4.658227848101266,
"grad_norm": 0.16446263618417067,
"learning_rate": 5.132372966670129e-07,
"loss": 0.2438,
"step": 1150
},
{
"epoch": 4.662278481012658,
"grad_norm": 0.16553426885153225,
"learning_rate": 5.005400145974704e-07,
"loss": 0.2466,
"step": 1151
},
{
"epoch": 4.666329113924051,
"grad_norm": 0.16365220715390152,
"learning_rate": 4.879997784979562e-07,
"loss": 0.2279,
"step": 1152
},
{
"epoch": 4.670379746835443,
"grad_norm": 0.17032790577399604,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.251,
"step": 1153
},
{
"epoch": 4.674430379746836,
"grad_norm": 0.17446919504437555,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.2448,
"step": 1154
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.1632495521727125,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.255,
"step": 1155
},
{
"epoch": 4.68253164556962,
"grad_norm": 0.16888728563477945,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.2542,
"step": 1156
},
{
"epoch": 4.686582278481013,
"grad_norm": 0.16861452879226235,
"learning_rate": 4.27657778143431e-07,
"loss": 0.2535,
"step": 1157
},
{
"epoch": 4.690632911392405,
"grad_norm": 0.16990402146611547,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.2543,
"step": 1158
},
{
"epoch": 4.694683544303797,
"grad_norm": 0.17451152982004325,
"learning_rate": 4.046237371675177e-07,
"loss": 0.2501,
"step": 1159
},
{
"epoch": 4.69873417721519,
"grad_norm": 0.17715901756134075,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.2528,
"step": 1160
},
{
"epoch": 4.702784810126582,
"grad_norm": 0.17524787550119852,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.2525,
"step": 1161
},
{
"epoch": 4.706835443037974,
"grad_norm": 0.17121188231643855,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.2471,
"step": 1162
},
{
"epoch": 4.710886075949367,
"grad_norm": 0.17309007358494535,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.2441,
"step": 1163
},
{
"epoch": 4.714936708860759,
"grad_norm": 0.17354018386784079,
"learning_rate": 3.498020255137813e-07,
"loss": 0.2445,
"step": 1164
},
{
"epoch": 4.718987341772152,
"grad_norm": 0.16844165906748987,
"learning_rate": 3.393121163057811e-07,
"loss": 0.247,
"step": 1165
},
{
"epoch": 4.723037974683544,
"grad_norm": 0.16944310961817555,
"learning_rate": 3.289805515751399e-07,
"loss": 0.2532,
"step": 1166
},
{
"epoch": 4.727088607594936,
"grad_norm": 0.16297575196199987,
"learning_rate": 3.188074145308573e-07,
"loss": 0.2518,
"step": 1167
},
{
"epoch": 4.731139240506329,
"grad_norm": 0.171408918119532,
"learning_rate": 3.087927871059804e-07,
"loss": 0.2394,
"step": 1168
},
{
"epoch": 4.735189873417721,
"grad_norm": 0.16910558471689793,
"learning_rate": 2.989367499569418e-07,
"loss": 0.241,
"step": 1169
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.17012798020031014,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.2609,
"step": 1170
},
{
"epoch": 4.743291139240506,
"grad_norm": 0.16791362106639846,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.2491,
"step": 1171
},
{
"epoch": 4.7473417721518985,
"grad_norm": 0.1634955187759983,
"learning_rate": 2.703209675663887e-07,
"loss": 0.2678,
"step": 1172
},
{
"epoch": 4.751392405063291,
"grad_norm": 0.16650338389829408,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.2504,
"step": 1173
},
{
"epoch": 4.7554430379746835,
"grad_norm": 0.1686444403219449,
"learning_rate": 2.520381518804471e-07,
"loss": 0.2559,
"step": 1174
},
{
"epoch": 4.759493670886076,
"grad_norm": 0.16524734614570669,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.244,
"step": 1175
},
{
"epoch": 4.7635443037974685,
"grad_norm": 0.1668187218531842,
"learning_rate": 2.343915243930317e-07,
"loss": 0.2503,
"step": 1176
},
{
"epoch": 4.767594936708861,
"grad_norm": 0.17065625639454465,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.2577,
"step": 1177
},
{
"epoch": 4.7716455696202535,
"grad_norm": 0.16900478200317276,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.2518,
"step": 1178
},
{
"epoch": 4.775696202531646,
"grad_norm": 0.16629510218141713,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.2503,
"step": 1179
},
{
"epoch": 4.779746835443038,
"grad_norm": 0.16808384162457238,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.2483,
"step": 1180
},
{
"epoch": 4.783797468354431,
"grad_norm": 0.16436629424241725,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.2514,
"step": 1181
},
{
"epoch": 4.787848101265823,
"grad_norm": 0.16644858178226654,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.2527,
"step": 1182
},
{
"epoch": 4.791898734177215,
"grad_norm": 0.1673217331036557,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.2605,
"step": 1183
},
{
"epoch": 4.795949367088608,
"grad_norm": 0.1616613962035212,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.2485,
"step": 1184
},
{
"epoch": 4.8,
"grad_norm": 0.16505208610667396,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.2505,
"step": 1185
},
{
"epoch": 4.804050632911393,
"grad_norm": 0.16893941699508638,
"learning_rate": 1.557203888516745e-07,
"loss": 0.2606,
"step": 1186
},
{
"epoch": 4.808101265822785,
"grad_norm": 0.16774750927685184,
"learning_rate": 1.487313098987131e-07,
"loss": 0.2546,
"step": 1187
},
{
"epoch": 4.812151898734177,
"grad_norm": 0.16689977741555168,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.2527,
"step": 1188
},
{
"epoch": 4.81620253164557,
"grad_norm": 0.16531237194265486,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.2522,
"step": 1189
},
{
"epoch": 4.820253164556962,
"grad_norm": 0.1699113500357116,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.2475,
"step": 1190
},
{
"epoch": 4.824303797468355,
"grad_norm": 0.1706049816439401,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.2393,
"step": 1191
},
{
"epoch": 4.828354430379747,
"grad_norm": 0.16882017887005782,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.2404,
"step": 1192
},
{
"epoch": 4.832405063291139,
"grad_norm": 0.16865368575868378,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.2462,
"step": 1193
},
{
"epoch": 4.836455696202532,
"grad_norm": 0.1621379109301859,
"learning_rate": 1.042873700093061e-07,
"loss": 0.2553,
"step": 1194
},
{
"epoch": 4.840506329113924,
"grad_norm": 0.1634530042787261,
"learning_rate": 9.857879404556291e-08,
"loss": 0.2405,
"step": 1195
},
{
"epoch": 4.844556962025316,
"grad_norm": 0.1658457563756738,
"learning_rate": 9.303050139225722e-08,
"loss": 0.2535,
"step": 1196
},
{
"epoch": 4.848607594936709,
"grad_norm": 0.1648512600602524,
"learning_rate": 8.76425367345779e-08,
"loss": 0.2609,
"step": 1197
},
{
"epoch": 4.852658227848101,
"grad_norm": 0.16811327119640435,
"learning_rate": 8.241494346644897e-08,
"loss": 0.251,
"step": 1198
},
{
"epoch": 4.856708860759493,
"grad_norm": 0.16679723069515223,
"learning_rate": 7.734776369019204e-08,
"loss": 0.2558,
"step": 1199
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.16880974036082474,
"learning_rate": 7.244103821617332e-08,
"loss": 0.2627,
"step": 1200
},
{
"epoch": 4.864810126582278,
"grad_norm": 0.17498167998977046,
"learning_rate": 6.769480656248606e-08,
"loss": 0.2501,
"step": 1201
},
{
"epoch": 4.868860759493671,
"grad_norm": 0.1766332925717399,
"learning_rate": 6.310910695462635e-08,
"loss": 0.2359,
"step": 1202
},
{
"epoch": 4.872911392405063,
"grad_norm": 0.16448633530295548,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.2598,
"step": 1203
},
{
"epoch": 4.876962025316455,
"grad_norm": 0.16582023106219768,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.2523,
"step": 1204
},
{
"epoch": 4.881012658227848,
"grad_norm": 0.16346796754173007,
"learning_rate": 5.031556326567488e-08,
"loss": 0.2663,
"step": 1205
},
{
"epoch": 4.88506329113924,
"grad_norm": 0.17003180182368718,
"learning_rate": 4.637234823364312e-08,
"loss": 0.2394,
"step": 1206
},
{
"epoch": 4.889113924050633,
"grad_norm": 0.17516477892731455,
"learning_rate": 4.258983697558838e-08,
"loss": 0.2476,
"step": 1207
},
{
"epoch": 4.8931645569620255,
"grad_norm": 0.16308304752831335,
"learning_rate": 3.896805995533548e-08,
"loss": 0.2527,
"step": 1208
},
{
"epoch": 4.8972151898734175,
"grad_norm": 0.162415928430663,
"learning_rate": 3.550704634218028e-08,
"loss": 0.2431,
"step": 1209
},
{
"epoch": 4.9012658227848105,
"grad_norm": 0.17216956053880225,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.2441,
"step": 1210
},
{
"epoch": 4.905316455696203,
"grad_norm": 0.16345515883942588,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.2448,
"step": 1211
},
{
"epoch": 4.909367088607595,
"grad_norm": 0.1667694689660845,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.2489,
"step": 1212
},
{
"epoch": 4.913417721518988,
"grad_norm": 0.165142203802428,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.2499,
"step": 1213
},
{
"epoch": 4.91746835443038,
"grad_norm": 0.16785413285542156,
"learning_rate": 2.061435966221881e-08,
"loss": 0.2531,
"step": 1214
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.1667845883779661,
"learning_rate": 1.811846652477245e-08,
"loss": 0.2541,
"step": 1215
},
{
"epoch": 4.925569620253165,
"grad_norm": 0.16794868338406627,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.2459,
"step": 1216
},
{
"epoch": 4.929620253164557,
"grad_norm": 0.16129896228310983,
"learning_rate": 1.360949303591097e-08,
"loss": 0.2498,
"step": 1217
},
{
"epoch": 4.93367088607595,
"grad_norm": 0.1696224277837463,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.2412,
"step": 1218
},
{
"epoch": 4.937721518987342,
"grad_norm": 0.166315544883146,
"learning_rate": 9.744388816668172e-09,
"loss": 0.2555,
"step": 1219
},
{
"epoch": 4.941772151898734,
"grad_norm": 0.16502721723749897,
"learning_rate": 8.05332740472009e-09,
"loss": 0.2482,
"step": 1220
},
{
"epoch": 4.945822784810127,
"grad_norm": 0.17265847315462504,
"learning_rate": 6.523278382872811e-09,
"loss": 0.2544,
"step": 1221
},
{
"epoch": 4.949873417721519,
"grad_norm": 0.17023432706011934,
"learning_rate": 5.15425407393133e-09,
"loss": 0.2699,
"step": 1222
},
{
"epoch": 4.953924050632912,
"grad_norm": 0.16619843561862016,
"learning_rate": 3.94626550383137e-09,
"loss": 0.2415,
"step": 1223
},
{
"epoch": 4.957974683544304,
"grad_norm": 0.16505800502361528,
"learning_rate": 2.899322401546112e-09,
"loss": 0.259,
"step": 1224
},
{
"epoch": 4.962025316455696,
"grad_norm": 0.16389117641936543,
"learning_rate": 2.013433199010706e-09,
"loss": 0.2579,
"step": 1225
},
{
"epoch": 4.966075949367089,
"grad_norm": 0.1663736634685331,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.2582,
"step": 1226
},
{
"epoch": 4.970126582278481,
"grad_norm": 0.16883295169024848,
"learning_rate": 7.248437353468695e-10,
"loss": 0.2462,
"step": 1227
},
{
"epoch": 4.974177215189873,
"grad_norm": 0.1691581775256991,
"learning_rate": 3.221538523412449e-10,
"loss": 0.2475,
"step": 1228
},
{
"epoch": 4.978227848101266,
"grad_norm": 0.16910044180407807,
"learning_rate": 8.053862524670663e-11,
"loss": 0.2462,
"step": 1229
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.16899449225261057,
"learning_rate": 0.0,
"loss": 0.2471,
"step": 1230
},
{
"epoch": 4.982278481012658,
"step": 1230,
"total_flos": 5.910237881712411e+18,
"train_loss": 0.35406713835834486,
"train_runtime": 54669.9787,
"train_samples_per_second": 2.89,
"train_steps_per_second": 0.022
}
],
"logging_steps": 1,
"max_steps": 1230,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.910237881712411e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}