CRI-R-qwen-7B-Instruct-2 / trainer_state.json
vitaleantonio's picture
Upload CRI-R-qwen-7B-Instruct-2
208820e verified
Raw
History Blame Contribute Delete
498 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 3060,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006539153179663233,
"grad_norm": 22.375,
"learning_rate": 2.6143790849673203e-08,
"loss": 1.7853,
"step": 1
},
{
"epoch": 0.0013078306359326467,
"grad_norm": 18.75,
"learning_rate": 5.228758169934641e-08,
"loss": 1.3249,
"step": 2
},
{
"epoch": 0.00196174595389897,
"grad_norm": 16.5,
"learning_rate": 7.843137254901961e-08,
"loss": 1.4347,
"step": 3
},
{
"epoch": 0.0026156612718652933,
"grad_norm": 17.125,
"learning_rate": 1.0457516339869281e-07,
"loss": 1.1277,
"step": 4
},
{
"epoch": 0.003269576589831617,
"grad_norm": 24.375,
"learning_rate": 1.3071895424836603e-07,
"loss": 1.8236,
"step": 5
},
{
"epoch": 0.00392349190779794,
"grad_norm": 28.375,
"learning_rate": 1.5686274509803921e-07,
"loss": 2.18,
"step": 6
},
{
"epoch": 0.0045774072257642635,
"grad_norm": 17.875,
"learning_rate": 1.8300653594771245e-07,
"loss": 1.7174,
"step": 7
},
{
"epoch": 0.005231322543730587,
"grad_norm": 16.25,
"learning_rate": 2.0915032679738563e-07,
"loss": 1.1492,
"step": 8
},
{
"epoch": 0.005885237861696911,
"grad_norm": 27.25,
"learning_rate": 2.3529411764705883e-07,
"loss": 1.6498,
"step": 9
},
{
"epoch": 0.006539153179663234,
"grad_norm": 14.8125,
"learning_rate": 2.6143790849673207e-07,
"loss": 1.0398,
"step": 10
},
{
"epoch": 0.007193068497629557,
"grad_norm": 16.0,
"learning_rate": 2.8758169934640525e-07,
"loss": 1.3389,
"step": 11
},
{
"epoch": 0.00784698381559588,
"grad_norm": 20.75,
"learning_rate": 3.1372549019607843e-07,
"loss": 1.9224,
"step": 12
},
{
"epoch": 0.008500899133562204,
"grad_norm": 18.0,
"learning_rate": 3.398692810457516e-07,
"loss": 1.3336,
"step": 13
},
{
"epoch": 0.009154814451528527,
"grad_norm": 20.625,
"learning_rate": 3.660130718954249e-07,
"loss": 1.6119,
"step": 14
},
{
"epoch": 0.00980872976949485,
"grad_norm": 17.125,
"learning_rate": 3.921568627450981e-07,
"loss": 1.2061,
"step": 15
},
{
"epoch": 0.010462645087461173,
"grad_norm": 21.125,
"learning_rate": 4.1830065359477126e-07,
"loss": 1.55,
"step": 16
},
{
"epoch": 0.011116560405427496,
"grad_norm": 22.125,
"learning_rate": 4.444444444444445e-07,
"loss": 1.495,
"step": 17
},
{
"epoch": 0.011770475723393821,
"grad_norm": 23.625,
"learning_rate": 4.7058823529411767e-07,
"loss": 1.5443,
"step": 18
},
{
"epoch": 0.012424391041360144,
"grad_norm": 18.375,
"learning_rate": 4.967320261437909e-07,
"loss": 1.72,
"step": 19
},
{
"epoch": 0.013078306359326467,
"grad_norm": 20.5,
"learning_rate": 5.228758169934641e-07,
"loss": 1.5461,
"step": 20
},
{
"epoch": 0.01373222167729279,
"grad_norm": 14.5,
"learning_rate": 5.490196078431373e-07,
"loss": 1.3291,
"step": 21
},
{
"epoch": 0.014386136995259114,
"grad_norm": 23.625,
"learning_rate": 5.751633986928105e-07,
"loss": 1.4587,
"step": 22
},
{
"epoch": 0.015040052313225437,
"grad_norm": 20.25,
"learning_rate": 6.013071895424837e-07,
"loss": 1.5046,
"step": 23
},
{
"epoch": 0.01569396763119176,
"grad_norm": 17.625,
"learning_rate": 6.274509803921569e-07,
"loss": 1.9609,
"step": 24
},
{
"epoch": 0.016347882949158083,
"grad_norm": 21.375,
"learning_rate": 6.535947712418302e-07,
"loss": 1.5493,
"step": 25
},
{
"epoch": 0.017001798267124408,
"grad_norm": 18.625,
"learning_rate": 6.797385620915032e-07,
"loss": 1.4214,
"step": 26
},
{
"epoch": 0.01765571358509073,
"grad_norm": 26.375,
"learning_rate": 7.058823529411766e-07,
"loss": 2.0565,
"step": 27
},
{
"epoch": 0.018309628903057054,
"grad_norm": 23.5,
"learning_rate": 7.320261437908498e-07,
"loss": 2.0252,
"step": 28
},
{
"epoch": 0.01896354422102338,
"grad_norm": 22.625,
"learning_rate": 7.581699346405229e-07,
"loss": 1.6488,
"step": 29
},
{
"epoch": 0.0196174595389897,
"grad_norm": 16.125,
"learning_rate": 7.843137254901962e-07,
"loss": 1.3428,
"step": 30
},
{
"epoch": 0.020271374856956025,
"grad_norm": 28.25,
"learning_rate": 8.104575163398694e-07,
"loss": 1.7191,
"step": 31
},
{
"epoch": 0.020925290174922347,
"grad_norm": 17.0,
"learning_rate": 8.366013071895425e-07,
"loss": 1.1449,
"step": 32
},
{
"epoch": 0.02157920549288867,
"grad_norm": 24.5,
"learning_rate": 8.627450980392157e-07,
"loss": 2.135,
"step": 33
},
{
"epoch": 0.022233120810854993,
"grad_norm": 20.375,
"learning_rate": 8.88888888888889e-07,
"loss": 1.4919,
"step": 34
},
{
"epoch": 0.022887036128821318,
"grad_norm": 29.0,
"learning_rate": 9.150326797385621e-07,
"loss": 1.8614,
"step": 35
},
{
"epoch": 0.023540951446787643,
"grad_norm": 19.125,
"learning_rate": 9.411764705882353e-07,
"loss": 1.0191,
"step": 36
},
{
"epoch": 0.024194866764753964,
"grad_norm": 15.9375,
"learning_rate": 9.673202614379086e-07,
"loss": 1.1653,
"step": 37
},
{
"epoch": 0.02484878208272029,
"grad_norm": 22.375,
"learning_rate": 9.934640522875818e-07,
"loss": 1.7648,
"step": 38
},
{
"epoch": 0.02550269740068661,
"grad_norm": 18.625,
"learning_rate": 1.019607843137255e-06,
"loss": 1.7209,
"step": 39
},
{
"epoch": 0.026156612718652935,
"grad_norm": 20.125,
"learning_rate": 1.0457516339869283e-06,
"loss": 1.5833,
"step": 40
},
{
"epoch": 0.026810528036619256,
"grad_norm": 15.0625,
"learning_rate": 1.0718954248366013e-06,
"loss": 0.7896,
"step": 41
},
{
"epoch": 0.02746444335458558,
"grad_norm": 22.0,
"learning_rate": 1.0980392156862745e-06,
"loss": 2.0211,
"step": 42
},
{
"epoch": 0.028118358672551906,
"grad_norm": 27.125,
"learning_rate": 1.1241830065359478e-06,
"loss": 2.4727,
"step": 43
},
{
"epoch": 0.028772273990518227,
"grad_norm": 19.125,
"learning_rate": 1.150326797385621e-06,
"loss": 1.6349,
"step": 44
},
{
"epoch": 0.029426189308484552,
"grad_norm": 23.125,
"learning_rate": 1.1764705882352942e-06,
"loss": 1.6653,
"step": 45
},
{
"epoch": 0.030080104626450874,
"grad_norm": 15.0,
"learning_rate": 1.2026143790849675e-06,
"loss": 1.0016,
"step": 46
},
{
"epoch": 0.0307340199444172,
"grad_norm": 20.5,
"learning_rate": 1.2287581699346405e-06,
"loss": 1.2971,
"step": 47
},
{
"epoch": 0.03138793526238352,
"grad_norm": 15.6875,
"learning_rate": 1.2549019607843137e-06,
"loss": 1.0573,
"step": 48
},
{
"epoch": 0.03204185058034984,
"grad_norm": 18.75,
"learning_rate": 1.281045751633987e-06,
"loss": 1.3878,
"step": 49
},
{
"epoch": 0.032695765898316166,
"grad_norm": 20.875,
"learning_rate": 1.3071895424836604e-06,
"loss": 1.8702,
"step": 50
},
{
"epoch": 0.03334968121628249,
"grad_norm": 17.0,
"learning_rate": 1.3333333333333334e-06,
"loss": 1.8821,
"step": 51
},
{
"epoch": 0.034003596534248816,
"grad_norm": 21.125,
"learning_rate": 1.3594771241830064e-06,
"loss": 1.751,
"step": 52
},
{
"epoch": 0.03465751185221514,
"grad_norm": 17.625,
"learning_rate": 1.3856209150326799e-06,
"loss": 1.2839,
"step": 53
},
{
"epoch": 0.03531142717018146,
"grad_norm": 20.125,
"learning_rate": 1.4117647058823531e-06,
"loss": 1.4963,
"step": 54
},
{
"epoch": 0.03596534248814778,
"grad_norm": 16.875,
"learning_rate": 1.4379084967320261e-06,
"loss": 1.3659,
"step": 55
},
{
"epoch": 0.03661925780611411,
"grad_norm": 19.0,
"learning_rate": 1.4640522875816996e-06,
"loss": 1.9621,
"step": 56
},
{
"epoch": 0.03727317312408043,
"grad_norm": 17.125,
"learning_rate": 1.4901960784313726e-06,
"loss": 1.525,
"step": 57
},
{
"epoch": 0.03792708844204676,
"grad_norm": 16.75,
"learning_rate": 1.5163398692810458e-06,
"loss": 1.207,
"step": 58
},
{
"epoch": 0.038581003760013076,
"grad_norm": 10.0625,
"learning_rate": 1.542483660130719e-06,
"loss": 0.6293,
"step": 59
},
{
"epoch": 0.0392349190779794,
"grad_norm": 15.1875,
"learning_rate": 1.5686274509803923e-06,
"loss": 1.1944,
"step": 60
},
{
"epoch": 0.039888834395945726,
"grad_norm": 17.375,
"learning_rate": 1.5947712418300653e-06,
"loss": 2.1252,
"step": 61
},
{
"epoch": 0.04054274971391205,
"grad_norm": 18.625,
"learning_rate": 1.6209150326797388e-06,
"loss": 1.2682,
"step": 62
},
{
"epoch": 0.04119666503187837,
"grad_norm": 12.75,
"learning_rate": 1.6470588235294118e-06,
"loss": 1.1163,
"step": 63
},
{
"epoch": 0.04185058034984469,
"grad_norm": 11.8125,
"learning_rate": 1.673202614379085e-06,
"loss": 0.7913,
"step": 64
},
{
"epoch": 0.04250449566781102,
"grad_norm": 16.375,
"learning_rate": 1.6993464052287585e-06,
"loss": 1.2979,
"step": 65
},
{
"epoch": 0.04315841098577734,
"grad_norm": 14.125,
"learning_rate": 1.7254901960784315e-06,
"loss": 1.2015,
"step": 66
},
{
"epoch": 0.04381232630374367,
"grad_norm": 14.6875,
"learning_rate": 1.7516339869281045e-06,
"loss": 1.0585,
"step": 67
},
{
"epoch": 0.044466241621709986,
"grad_norm": 13.25,
"learning_rate": 1.777777777777778e-06,
"loss": 1.3912,
"step": 68
},
{
"epoch": 0.04512015693967631,
"grad_norm": 12.125,
"learning_rate": 1.8039215686274512e-06,
"loss": 1.3715,
"step": 69
},
{
"epoch": 0.045774072257642635,
"grad_norm": 12.6875,
"learning_rate": 1.8300653594771242e-06,
"loss": 1.0968,
"step": 70
},
{
"epoch": 0.04642798757560896,
"grad_norm": 9.0,
"learning_rate": 1.8562091503267977e-06,
"loss": 0.8315,
"step": 71
},
{
"epoch": 0.047081902893575285,
"grad_norm": 9.6875,
"learning_rate": 1.8823529411764707e-06,
"loss": 1.0216,
"step": 72
},
{
"epoch": 0.0477358182115416,
"grad_norm": 8.9375,
"learning_rate": 1.9084967320261437e-06,
"loss": 0.963,
"step": 73
},
{
"epoch": 0.04838973352950793,
"grad_norm": 9.9375,
"learning_rate": 1.934640522875817e-06,
"loss": 1.0332,
"step": 74
},
{
"epoch": 0.04904364884747425,
"grad_norm": 10.0625,
"learning_rate": 1.96078431372549e-06,
"loss": 1.3037,
"step": 75
},
{
"epoch": 0.04969756416544058,
"grad_norm": 6.75,
"learning_rate": 1.9869281045751636e-06,
"loss": 0.6811,
"step": 76
},
{
"epoch": 0.050351479483406895,
"grad_norm": 12.625,
"learning_rate": 2.0130718954248366e-06,
"loss": 1.2454,
"step": 77
},
{
"epoch": 0.05100539480137322,
"grad_norm": 11.125,
"learning_rate": 2.03921568627451e-06,
"loss": 1.1049,
"step": 78
},
{
"epoch": 0.051659310119339545,
"grad_norm": 11.75,
"learning_rate": 2.065359477124183e-06,
"loss": 0.8606,
"step": 79
},
{
"epoch": 0.05231322543730587,
"grad_norm": 7.15625,
"learning_rate": 2.0915032679738565e-06,
"loss": 0.8439,
"step": 80
},
{
"epoch": 0.052967140755272195,
"grad_norm": 7.46875,
"learning_rate": 2.1176470588235296e-06,
"loss": 0.7439,
"step": 81
},
{
"epoch": 0.05362105607323851,
"grad_norm": 10.4375,
"learning_rate": 2.1437908496732026e-06,
"loss": 1.5164,
"step": 82
},
{
"epoch": 0.05427497139120484,
"grad_norm": 10.75,
"learning_rate": 2.169934640522876e-06,
"loss": 1.5607,
"step": 83
},
{
"epoch": 0.05492888670917116,
"grad_norm": 9.0,
"learning_rate": 2.196078431372549e-06,
"loss": 0.8527,
"step": 84
},
{
"epoch": 0.05558280202713749,
"grad_norm": 9.375,
"learning_rate": 2.222222222222222e-06,
"loss": 1.1674,
"step": 85
},
{
"epoch": 0.05623671734510381,
"grad_norm": 10.125,
"learning_rate": 2.2483660130718955e-06,
"loss": 1.4559,
"step": 86
},
{
"epoch": 0.05689063266307013,
"grad_norm": 7.34375,
"learning_rate": 2.274509803921569e-06,
"loss": 1.0805,
"step": 87
},
{
"epoch": 0.057544547981036455,
"grad_norm": 10.0,
"learning_rate": 2.300653594771242e-06,
"loss": 1.3947,
"step": 88
},
{
"epoch": 0.05819846329900278,
"grad_norm": 7.8125,
"learning_rate": 2.3267973856209154e-06,
"loss": 1.3273,
"step": 89
},
{
"epoch": 0.058852378616969105,
"grad_norm": 8.9375,
"learning_rate": 2.3529411764705885e-06,
"loss": 1.041,
"step": 90
},
{
"epoch": 0.05950629393493542,
"grad_norm": 6.15625,
"learning_rate": 2.3790849673202615e-06,
"loss": 0.8624,
"step": 91
},
{
"epoch": 0.06016020925290175,
"grad_norm": 8.125,
"learning_rate": 2.405228758169935e-06,
"loss": 0.8563,
"step": 92
},
{
"epoch": 0.06081412457086807,
"grad_norm": 5.5,
"learning_rate": 2.431372549019608e-06,
"loss": 0.8141,
"step": 93
},
{
"epoch": 0.0614680398888344,
"grad_norm": 6.25,
"learning_rate": 2.457516339869281e-06,
"loss": 0.9837,
"step": 94
},
{
"epoch": 0.06212195520680072,
"grad_norm": 7.03125,
"learning_rate": 2.4836601307189544e-06,
"loss": 0.9888,
"step": 95
},
{
"epoch": 0.06277587052476705,
"grad_norm": 6.78125,
"learning_rate": 2.5098039215686274e-06,
"loss": 1.1952,
"step": 96
},
{
"epoch": 0.06342978584273337,
"grad_norm": 5.78125,
"learning_rate": 2.535947712418301e-06,
"loss": 0.9052,
"step": 97
},
{
"epoch": 0.06408370116069968,
"grad_norm": 7.6875,
"learning_rate": 2.562091503267974e-06,
"loss": 1.0956,
"step": 98
},
{
"epoch": 0.06473761647866601,
"grad_norm": 6.375,
"learning_rate": 2.5882352941176473e-06,
"loss": 1.0037,
"step": 99
},
{
"epoch": 0.06539153179663233,
"grad_norm": 10.9375,
"learning_rate": 2.6143790849673208e-06,
"loss": 1.4849,
"step": 100
},
{
"epoch": 0.06604544711459866,
"grad_norm": 6.4375,
"learning_rate": 2.640522875816994e-06,
"loss": 0.8732,
"step": 101
},
{
"epoch": 0.06669936243256498,
"grad_norm": 6.59375,
"learning_rate": 2.666666666666667e-06,
"loss": 0.7752,
"step": 102
},
{
"epoch": 0.0673532777505313,
"grad_norm": 5.34375,
"learning_rate": 2.69281045751634e-06,
"loss": 0.8277,
"step": 103
},
{
"epoch": 0.06800719306849763,
"grad_norm": 6.78125,
"learning_rate": 2.718954248366013e-06,
"loss": 0.8004,
"step": 104
},
{
"epoch": 0.06866110838646396,
"grad_norm": 3.84375,
"learning_rate": 2.7450980392156867e-06,
"loss": 0.5019,
"step": 105
},
{
"epoch": 0.06931502370443028,
"grad_norm": 5.5625,
"learning_rate": 2.7712418300653598e-06,
"loss": 0.9185,
"step": 106
},
{
"epoch": 0.0699689390223966,
"grad_norm": 6.21875,
"learning_rate": 2.7973856209150328e-06,
"loss": 0.6808,
"step": 107
},
{
"epoch": 0.07062285434036292,
"grad_norm": 5.40625,
"learning_rate": 2.8235294117647062e-06,
"loss": 0.872,
"step": 108
},
{
"epoch": 0.07127676965832924,
"grad_norm": 6.875,
"learning_rate": 2.8496732026143792e-06,
"loss": 1.0618,
"step": 109
},
{
"epoch": 0.07193068497629557,
"grad_norm": 6.3125,
"learning_rate": 2.8758169934640523e-06,
"loss": 0.8521,
"step": 110
},
{
"epoch": 0.07258460029426189,
"grad_norm": 7.0,
"learning_rate": 2.901960784313726e-06,
"loss": 1.1416,
"step": 111
},
{
"epoch": 0.07323851561222822,
"grad_norm": 6.59375,
"learning_rate": 2.928104575163399e-06,
"loss": 0.9626,
"step": 112
},
{
"epoch": 0.07389243093019454,
"grad_norm": 5.875,
"learning_rate": 2.954248366013072e-06,
"loss": 1.0726,
"step": 113
},
{
"epoch": 0.07454634624816087,
"grad_norm": 7.09375,
"learning_rate": 2.980392156862745e-06,
"loss": 1.4055,
"step": 114
},
{
"epoch": 0.07520026156612719,
"grad_norm": 6.8125,
"learning_rate": 3.0065359477124182e-06,
"loss": 0.8797,
"step": 115
},
{
"epoch": 0.07585417688409352,
"grad_norm": 7.375,
"learning_rate": 3.0326797385620917e-06,
"loss": 1.2011,
"step": 116
},
{
"epoch": 0.07650809220205983,
"grad_norm": 6.65625,
"learning_rate": 3.058823529411765e-06,
"loss": 1.1419,
"step": 117
},
{
"epoch": 0.07716200752002615,
"grad_norm": 4.96875,
"learning_rate": 3.084967320261438e-06,
"loss": 0.9388,
"step": 118
},
{
"epoch": 0.07781592283799248,
"grad_norm": 9.6875,
"learning_rate": 3.1111111111111116e-06,
"loss": 1.0324,
"step": 119
},
{
"epoch": 0.0784698381559588,
"grad_norm": 4.21875,
"learning_rate": 3.1372549019607846e-06,
"loss": 0.5297,
"step": 120
},
{
"epoch": 0.07912375347392513,
"grad_norm": 6.125,
"learning_rate": 3.1633986928104576e-06,
"loss": 1.1836,
"step": 121
},
{
"epoch": 0.07977766879189145,
"grad_norm": 5.9375,
"learning_rate": 3.1895424836601306e-06,
"loss": 0.8431,
"step": 122
},
{
"epoch": 0.08043158410985778,
"grad_norm": 4.9375,
"learning_rate": 3.2156862745098045e-06,
"loss": 0.8145,
"step": 123
},
{
"epoch": 0.0810854994278241,
"grad_norm": 5.59375,
"learning_rate": 3.2418300653594775e-06,
"loss": 0.8455,
"step": 124
},
{
"epoch": 0.08173941474579043,
"grad_norm": 5.03125,
"learning_rate": 3.2679738562091506e-06,
"loss": 0.9593,
"step": 125
},
{
"epoch": 0.08239333006375674,
"grad_norm": 5.46875,
"learning_rate": 3.2941176470588236e-06,
"loss": 0.9752,
"step": 126
},
{
"epoch": 0.08304724538172306,
"grad_norm": 4.5,
"learning_rate": 3.320261437908497e-06,
"loss": 0.6472,
"step": 127
},
{
"epoch": 0.08370116069968939,
"grad_norm": 5.71875,
"learning_rate": 3.34640522875817e-06,
"loss": 0.7602,
"step": 128
},
{
"epoch": 0.08435507601765571,
"grad_norm": 7.03125,
"learning_rate": 3.3725490196078435e-06,
"loss": 1.1583,
"step": 129
},
{
"epoch": 0.08500899133562204,
"grad_norm": 4.40625,
"learning_rate": 3.398692810457517e-06,
"loss": 0.3314,
"step": 130
},
{
"epoch": 0.08566290665358836,
"grad_norm": 5.25,
"learning_rate": 3.42483660130719e-06,
"loss": 0.6011,
"step": 131
},
{
"epoch": 0.08631682197155469,
"grad_norm": 7.90625,
"learning_rate": 3.450980392156863e-06,
"loss": 1.1565,
"step": 132
},
{
"epoch": 0.08697073728952101,
"grad_norm": 5.96875,
"learning_rate": 3.477124183006536e-06,
"loss": 0.9662,
"step": 133
},
{
"epoch": 0.08762465260748734,
"grad_norm": 4.0625,
"learning_rate": 3.503267973856209e-06,
"loss": 0.6753,
"step": 134
},
{
"epoch": 0.08827856792545366,
"grad_norm": 4.71875,
"learning_rate": 3.529411764705883e-06,
"loss": 0.6891,
"step": 135
},
{
"epoch": 0.08893248324341997,
"grad_norm": 5.65625,
"learning_rate": 3.555555555555556e-06,
"loss": 0.9057,
"step": 136
},
{
"epoch": 0.0895863985613863,
"grad_norm": 5.34375,
"learning_rate": 3.581699346405229e-06,
"loss": 0.7746,
"step": 137
},
{
"epoch": 0.09024031387935262,
"grad_norm": 4.40625,
"learning_rate": 3.6078431372549024e-06,
"loss": 0.66,
"step": 138
},
{
"epoch": 0.09089422919731895,
"grad_norm": 3.953125,
"learning_rate": 3.6339869281045754e-06,
"loss": 0.4679,
"step": 139
},
{
"epoch": 0.09154814451528527,
"grad_norm": 4.375,
"learning_rate": 3.6601307189542484e-06,
"loss": 0.5352,
"step": 140
},
{
"epoch": 0.0922020598332516,
"grad_norm": 5.8125,
"learning_rate": 3.6862745098039223e-06,
"loss": 0.6645,
"step": 141
},
{
"epoch": 0.09285597515121792,
"grad_norm": 5.90625,
"learning_rate": 3.7124183006535953e-06,
"loss": 1.3812,
"step": 142
},
{
"epoch": 0.09350989046918425,
"grad_norm": 4.84375,
"learning_rate": 3.7385620915032683e-06,
"loss": 0.8903,
"step": 143
},
{
"epoch": 0.09416380578715057,
"grad_norm": 4.0625,
"learning_rate": 3.7647058823529414e-06,
"loss": 0.517,
"step": 144
},
{
"epoch": 0.09481772110511688,
"grad_norm": 5.96875,
"learning_rate": 3.7908496732026144e-06,
"loss": 1.3503,
"step": 145
},
{
"epoch": 0.0954716364230832,
"grad_norm": 5.5625,
"learning_rate": 3.816993464052287e-06,
"loss": 0.7654,
"step": 146
},
{
"epoch": 0.09612555174104953,
"grad_norm": 6.75,
"learning_rate": 3.843137254901962e-06,
"loss": 1.0458,
"step": 147
},
{
"epoch": 0.09677946705901586,
"grad_norm": 6.125,
"learning_rate": 3.869281045751634e-06,
"loss": 0.9453,
"step": 148
},
{
"epoch": 0.09743338237698218,
"grad_norm": 4.125,
"learning_rate": 3.895424836601308e-06,
"loss": 0.6584,
"step": 149
},
{
"epoch": 0.0980872976949485,
"grad_norm": 3.40625,
"learning_rate": 3.92156862745098e-06,
"loss": 0.5141,
"step": 150
},
{
"epoch": 0.09874121301291483,
"grad_norm": 4.46875,
"learning_rate": 3.947712418300654e-06,
"loss": 0.7394,
"step": 151
},
{
"epoch": 0.09939512833088116,
"grad_norm": 6.8125,
"learning_rate": 3.973856209150327e-06,
"loss": 1.023,
"step": 152
},
{
"epoch": 0.10004904364884748,
"grad_norm": 3.828125,
"learning_rate": 4.000000000000001e-06,
"loss": 0.7538,
"step": 153
},
{
"epoch": 0.10070295896681379,
"grad_norm": 4.28125,
"learning_rate": 4.026143790849673e-06,
"loss": 0.6604,
"step": 154
},
{
"epoch": 0.10135687428478012,
"grad_norm": 3.75,
"learning_rate": 4.052287581699347e-06,
"loss": 0.5097,
"step": 155
},
{
"epoch": 0.10201078960274644,
"grad_norm": 8.4375,
"learning_rate": 4.07843137254902e-06,
"loss": 1.247,
"step": 156
},
{
"epoch": 0.10266470492071277,
"grad_norm": 3.453125,
"learning_rate": 4.104575163398693e-06,
"loss": 0.5514,
"step": 157
},
{
"epoch": 0.10331862023867909,
"grad_norm": 7.03125,
"learning_rate": 4.130718954248366e-06,
"loss": 1.0491,
"step": 158
},
{
"epoch": 0.10397253555664542,
"grad_norm": 3.359375,
"learning_rate": 4.15686274509804e-06,
"loss": 0.4755,
"step": 159
},
{
"epoch": 0.10462645087461174,
"grad_norm": 4.03125,
"learning_rate": 4.183006535947713e-06,
"loss": 0.609,
"step": 160
},
{
"epoch": 0.10528036619257806,
"grad_norm": 5.625,
"learning_rate": 4.209150326797386e-06,
"loss": 0.951,
"step": 161
},
{
"epoch": 0.10593428151054439,
"grad_norm": 2.78125,
"learning_rate": 4.235294117647059e-06,
"loss": 0.3454,
"step": 162
},
{
"epoch": 0.10658819682851071,
"grad_norm": 4.3125,
"learning_rate": 4.2614379084967326e-06,
"loss": 0.8485,
"step": 163
},
{
"epoch": 0.10724211214647703,
"grad_norm": 4.53125,
"learning_rate": 4.287581699346405e-06,
"loss": 0.7783,
"step": 164
},
{
"epoch": 0.10789602746444335,
"grad_norm": 5.9375,
"learning_rate": 4.313725490196079e-06,
"loss": 1.0732,
"step": 165
},
{
"epoch": 0.10854994278240968,
"grad_norm": 5.65625,
"learning_rate": 4.339869281045752e-06,
"loss": 0.9304,
"step": 166
},
{
"epoch": 0.109203858100376,
"grad_norm": 4.59375,
"learning_rate": 4.3660130718954255e-06,
"loss": 0.6748,
"step": 167
},
{
"epoch": 0.10985777341834232,
"grad_norm": 3.53125,
"learning_rate": 4.392156862745098e-06,
"loss": 0.5123,
"step": 168
},
{
"epoch": 0.11051168873630865,
"grad_norm": 4.59375,
"learning_rate": 4.4183006535947716e-06,
"loss": 0.7995,
"step": 169
},
{
"epoch": 0.11116560405427497,
"grad_norm": 4.375,
"learning_rate": 4.444444444444444e-06,
"loss": 0.6551,
"step": 170
},
{
"epoch": 0.1118195193722413,
"grad_norm": 3.15625,
"learning_rate": 4.4705882352941184e-06,
"loss": 0.5553,
"step": 171
},
{
"epoch": 0.11247343469020762,
"grad_norm": 3.078125,
"learning_rate": 4.496732026143791e-06,
"loss": 0.4769,
"step": 172
},
{
"epoch": 0.11312735000817394,
"grad_norm": 5.5625,
"learning_rate": 4.5228758169934645e-06,
"loss": 1.0379,
"step": 173
},
{
"epoch": 0.11378126532614026,
"grad_norm": 5.8125,
"learning_rate": 4.549019607843138e-06,
"loss": 1.0826,
"step": 174
},
{
"epoch": 0.11443518064410658,
"grad_norm": 4.28125,
"learning_rate": 4.5751633986928105e-06,
"loss": 0.6884,
"step": 175
},
{
"epoch": 0.11508909596207291,
"grad_norm": 7.65625,
"learning_rate": 4.601307189542484e-06,
"loss": 0.8737,
"step": 176
},
{
"epoch": 0.11574301128003923,
"grad_norm": 4.5,
"learning_rate": 4.627450980392157e-06,
"loss": 0.7343,
"step": 177
},
{
"epoch": 0.11639692659800556,
"grad_norm": 5.21875,
"learning_rate": 4.653594771241831e-06,
"loss": 1.0053,
"step": 178
},
{
"epoch": 0.11705084191597188,
"grad_norm": 7.03125,
"learning_rate": 4.6797385620915035e-06,
"loss": 1.8239,
"step": 179
},
{
"epoch": 0.11770475723393821,
"grad_norm": 5.28125,
"learning_rate": 4.705882352941177e-06,
"loss": 0.8575,
"step": 180
},
{
"epoch": 0.11835867255190453,
"grad_norm": 5.46875,
"learning_rate": 4.7320261437908495e-06,
"loss": 1.0108,
"step": 181
},
{
"epoch": 0.11901258786987085,
"grad_norm": 5.125,
"learning_rate": 4.758169934640523e-06,
"loss": 0.8172,
"step": 182
},
{
"epoch": 0.11966650318783717,
"grad_norm": 4.90625,
"learning_rate": 4.784313725490196e-06,
"loss": 1.0499,
"step": 183
},
{
"epoch": 0.1203204185058035,
"grad_norm": 5.21875,
"learning_rate": 4.81045751633987e-06,
"loss": 0.9608,
"step": 184
},
{
"epoch": 0.12097433382376982,
"grad_norm": 6.4375,
"learning_rate": 4.836601307189543e-06,
"loss": 1.1778,
"step": 185
},
{
"epoch": 0.12162824914173614,
"grad_norm": 4.0625,
"learning_rate": 4.862745098039216e-06,
"loss": 0.5224,
"step": 186
},
{
"epoch": 0.12228216445970247,
"grad_norm": 3.8125,
"learning_rate": 4.888888888888889e-06,
"loss": 0.5296,
"step": 187
},
{
"epoch": 0.1229360797776688,
"grad_norm": 4.875,
"learning_rate": 4.915032679738562e-06,
"loss": 1.05,
"step": 188
},
{
"epoch": 0.12358999509563512,
"grad_norm": 3.90625,
"learning_rate": 4.941176470588236e-06,
"loss": 0.5972,
"step": 189
},
{
"epoch": 0.12424391041360144,
"grad_norm": 3.25,
"learning_rate": 4.967320261437909e-06,
"loss": 0.5035,
"step": 190
},
{
"epoch": 0.12489782573156777,
"grad_norm": 5.6875,
"learning_rate": 4.993464052287582e-06,
"loss": 1.0502,
"step": 191
},
{
"epoch": 0.1255517410495341,
"grad_norm": 6.84375,
"learning_rate": 5.019607843137255e-06,
"loss": 1.0902,
"step": 192
},
{
"epoch": 0.12620565636750042,
"grad_norm": 3.890625,
"learning_rate": 5.045751633986928e-06,
"loss": 0.6796,
"step": 193
},
{
"epoch": 0.12685957168546674,
"grad_norm": 5.59375,
"learning_rate": 5.071895424836602e-06,
"loss": 0.8478,
"step": 194
},
{
"epoch": 0.12751348700343307,
"grad_norm": 3.765625,
"learning_rate": 5.098039215686274e-06,
"loss": 0.5782,
"step": 195
},
{
"epoch": 0.12816740232139937,
"grad_norm": 5.9375,
"learning_rate": 5.124183006535948e-06,
"loss": 0.8886,
"step": 196
},
{
"epoch": 0.1288213176393657,
"grad_norm": 4.78125,
"learning_rate": 5.15032679738562e-06,
"loss": 0.9319,
"step": 197
},
{
"epoch": 0.12947523295733201,
"grad_norm": 5.25,
"learning_rate": 5.176470588235295e-06,
"loss": 0.9604,
"step": 198
},
{
"epoch": 0.13012914827529834,
"grad_norm": 4.46875,
"learning_rate": 5.202614379084968e-06,
"loss": 0.9059,
"step": 199
},
{
"epoch": 0.13078306359326466,
"grad_norm": 4.15625,
"learning_rate": 5.2287581699346416e-06,
"loss": 0.6895,
"step": 200
},
{
"epoch": 0.131436978911231,
"grad_norm": 5.5,
"learning_rate": 5.254901960784314e-06,
"loss": 0.9966,
"step": 201
},
{
"epoch": 0.13209089422919731,
"grad_norm": 6.21875,
"learning_rate": 5.281045751633988e-06,
"loss": 1.1461,
"step": 202
},
{
"epoch": 0.13274480954716364,
"grad_norm": 3.84375,
"learning_rate": 5.30718954248366e-06,
"loss": 0.4568,
"step": 203
},
{
"epoch": 0.13339872486512996,
"grad_norm": 4.4375,
"learning_rate": 5.333333333333334e-06,
"loss": 0.6391,
"step": 204
},
{
"epoch": 0.1340526401830963,
"grad_norm": 4.59375,
"learning_rate": 5.359477124183007e-06,
"loss": 0.7027,
"step": 205
},
{
"epoch": 0.1347065555010626,
"grad_norm": 3.40625,
"learning_rate": 5.38562091503268e-06,
"loss": 0.5334,
"step": 206
},
{
"epoch": 0.13536047081902894,
"grad_norm": 4.75,
"learning_rate": 5.411764705882353e-06,
"loss": 0.6524,
"step": 207
},
{
"epoch": 0.13601438613699526,
"grad_norm": 4.90625,
"learning_rate": 5.437908496732026e-06,
"loss": 0.7142,
"step": 208
},
{
"epoch": 0.1366683014549616,
"grad_norm": 7.21875,
"learning_rate": 5.464052287581699e-06,
"loss": 1.3035,
"step": 209
},
{
"epoch": 0.1373222167729279,
"grad_norm": 6.0625,
"learning_rate": 5.4901960784313735e-06,
"loss": 0.8904,
"step": 210
},
{
"epoch": 0.13797613209089424,
"grad_norm": 6.125,
"learning_rate": 5.516339869281047e-06,
"loss": 1.1201,
"step": 211
},
{
"epoch": 0.13863004740886056,
"grad_norm": 4.8125,
"learning_rate": 5.5424836601307195e-06,
"loss": 0.772,
"step": 212
},
{
"epoch": 0.1392839627268269,
"grad_norm": 4.8125,
"learning_rate": 5.568627450980393e-06,
"loss": 0.8415,
"step": 213
},
{
"epoch": 0.1399378780447932,
"grad_norm": 4.9375,
"learning_rate": 5.5947712418300656e-06,
"loss": 0.6259,
"step": 214
},
{
"epoch": 0.1405917933627595,
"grad_norm": 5.03125,
"learning_rate": 5.620915032679739e-06,
"loss": 0.7825,
"step": 215
},
{
"epoch": 0.14124570868072583,
"grad_norm": 4.15625,
"learning_rate": 5.6470588235294125e-06,
"loss": 0.9823,
"step": 216
},
{
"epoch": 0.14189962399869216,
"grad_norm": 4.15625,
"learning_rate": 5.673202614379085e-06,
"loss": 0.6209,
"step": 217
},
{
"epoch": 0.14255353931665848,
"grad_norm": 4.6875,
"learning_rate": 5.6993464052287585e-06,
"loss": 0.8821,
"step": 218
},
{
"epoch": 0.1432074546346248,
"grad_norm": 5.15625,
"learning_rate": 5.725490196078431e-06,
"loss": 0.7635,
"step": 219
},
{
"epoch": 0.14386136995259113,
"grad_norm": 6.03125,
"learning_rate": 5.7516339869281045e-06,
"loss": 1.2596,
"step": 220
},
{
"epoch": 0.14451528527055746,
"grad_norm": 5.53125,
"learning_rate": 5.777777777777778e-06,
"loss": 0.7383,
"step": 221
},
{
"epoch": 0.14516920058852378,
"grad_norm": 5.8125,
"learning_rate": 5.803921568627452e-06,
"loss": 0.7485,
"step": 222
},
{
"epoch": 0.1458231159064901,
"grad_norm": 4.75,
"learning_rate": 5.830065359477125e-06,
"loss": 0.7227,
"step": 223
},
{
"epoch": 0.14647703122445643,
"grad_norm": 5.0,
"learning_rate": 5.856209150326798e-06,
"loss": 0.7887,
"step": 224
},
{
"epoch": 0.14713094654242276,
"grad_norm": 6.53125,
"learning_rate": 5.882352941176471e-06,
"loss": 0.9524,
"step": 225
},
{
"epoch": 0.14778486186038908,
"grad_norm": 6.46875,
"learning_rate": 5.908496732026144e-06,
"loss": 1.1014,
"step": 226
},
{
"epoch": 0.1484387771783554,
"grad_norm": 5.375,
"learning_rate": 5.934640522875818e-06,
"loss": 1.1672,
"step": 227
},
{
"epoch": 0.14909269249632173,
"grad_norm": 5.25,
"learning_rate": 5.96078431372549e-06,
"loss": 0.7312,
"step": 228
},
{
"epoch": 0.14974660781428806,
"grad_norm": 5.125,
"learning_rate": 5.986928104575164e-06,
"loss": 0.707,
"step": 229
},
{
"epoch": 0.15040052313225438,
"grad_norm": 3.96875,
"learning_rate": 6.0130718954248365e-06,
"loss": 0.5949,
"step": 230
},
{
"epoch": 0.1510544384502207,
"grad_norm": 6.40625,
"learning_rate": 6.03921568627451e-06,
"loss": 0.9902,
"step": 231
},
{
"epoch": 0.15170835376818703,
"grad_norm": 4.3125,
"learning_rate": 6.065359477124183e-06,
"loss": 0.7778,
"step": 232
},
{
"epoch": 0.15236226908615336,
"grad_norm": 5.71875,
"learning_rate": 6.091503267973856e-06,
"loss": 1.1321,
"step": 233
},
{
"epoch": 0.15301618440411965,
"grad_norm": 5.03125,
"learning_rate": 6.11764705882353e-06,
"loss": 0.7795,
"step": 234
},
{
"epoch": 0.15367009972208598,
"grad_norm": 7.0,
"learning_rate": 6.143790849673204e-06,
"loss": 1.1002,
"step": 235
},
{
"epoch": 0.1543240150400523,
"grad_norm": 3.875,
"learning_rate": 6.169934640522876e-06,
"loss": 0.5591,
"step": 236
},
{
"epoch": 0.15497793035801863,
"grad_norm": 3.984375,
"learning_rate": 6.19607843137255e-06,
"loss": 0.6337,
"step": 237
},
{
"epoch": 0.15563184567598495,
"grad_norm": 3.34375,
"learning_rate": 6.222222222222223e-06,
"loss": 0.53,
"step": 238
},
{
"epoch": 0.15628576099395128,
"grad_norm": 4.375,
"learning_rate": 6.248366013071896e-06,
"loss": 0.7318,
"step": 239
},
{
"epoch": 0.1569396763119176,
"grad_norm": 4.59375,
"learning_rate": 6.274509803921569e-06,
"loss": 0.59,
"step": 240
},
{
"epoch": 0.15759359162988393,
"grad_norm": 5.28125,
"learning_rate": 6.300653594771242e-06,
"loss": 0.7881,
"step": 241
},
{
"epoch": 0.15824750694785025,
"grad_norm": 4.625,
"learning_rate": 6.326797385620915e-06,
"loss": 0.7447,
"step": 242
},
{
"epoch": 0.15890142226581658,
"grad_norm": 5.0625,
"learning_rate": 6.352941176470589e-06,
"loss": 0.7828,
"step": 243
},
{
"epoch": 0.1595553375837829,
"grad_norm": 4.40625,
"learning_rate": 6.379084967320261e-06,
"loss": 0.7839,
"step": 244
},
{
"epoch": 0.16020925290174923,
"grad_norm": 5.28125,
"learning_rate": 6.405228758169935e-06,
"loss": 0.8127,
"step": 245
},
{
"epoch": 0.16086316821971555,
"grad_norm": 4.4375,
"learning_rate": 6.431372549019609e-06,
"loss": 0.7087,
"step": 246
},
{
"epoch": 0.16151708353768188,
"grad_norm": 4.875,
"learning_rate": 6.457516339869282e-06,
"loss": 0.6723,
"step": 247
},
{
"epoch": 0.1621709988556482,
"grad_norm": 3.671875,
"learning_rate": 6.483660130718955e-06,
"loss": 0.4365,
"step": 248
},
{
"epoch": 0.16282491417361453,
"grad_norm": 3.59375,
"learning_rate": 6.5098039215686285e-06,
"loss": 0.5081,
"step": 249
},
{
"epoch": 0.16347882949158085,
"grad_norm": 3.953125,
"learning_rate": 6.535947712418301e-06,
"loss": 0.6549,
"step": 250
},
{
"epoch": 0.16413274480954718,
"grad_norm": 4.4375,
"learning_rate": 6.5620915032679746e-06,
"loss": 0.7618,
"step": 251
},
{
"epoch": 0.16478666012751347,
"grad_norm": 4.03125,
"learning_rate": 6.588235294117647e-06,
"loss": 0.6397,
"step": 252
},
{
"epoch": 0.1654405754454798,
"grad_norm": 4.8125,
"learning_rate": 6.614379084967321e-06,
"loss": 0.7834,
"step": 253
},
{
"epoch": 0.16609449076344612,
"grad_norm": 4.1875,
"learning_rate": 6.640522875816994e-06,
"loss": 0.6746,
"step": 254
},
{
"epoch": 0.16674840608141245,
"grad_norm": 4.5,
"learning_rate": 6.666666666666667e-06,
"loss": 0.7292,
"step": 255
},
{
"epoch": 0.16740232139937877,
"grad_norm": 4.65625,
"learning_rate": 6.69281045751634e-06,
"loss": 0.7284,
"step": 256
},
{
"epoch": 0.1680562367173451,
"grad_norm": 5.0625,
"learning_rate": 6.718954248366014e-06,
"loss": 0.9511,
"step": 257
},
{
"epoch": 0.16871015203531142,
"grad_norm": 4.53125,
"learning_rate": 6.745098039215687e-06,
"loss": 0.6918,
"step": 258
},
{
"epoch": 0.16936406735327775,
"grad_norm": 3.203125,
"learning_rate": 6.7712418300653604e-06,
"loss": 0.4269,
"step": 259
},
{
"epoch": 0.17001798267124407,
"grad_norm": 5.0625,
"learning_rate": 6.797385620915034e-06,
"loss": 0.8029,
"step": 260
},
{
"epoch": 0.1706718979892104,
"grad_norm": 4.59375,
"learning_rate": 6.8235294117647065e-06,
"loss": 0.694,
"step": 261
},
{
"epoch": 0.17132581330717672,
"grad_norm": 4.3125,
"learning_rate": 6.84967320261438e-06,
"loss": 0.6182,
"step": 262
},
{
"epoch": 0.17197972862514305,
"grad_norm": 5.9375,
"learning_rate": 6.8758169934640525e-06,
"loss": 1.0165,
"step": 263
},
{
"epoch": 0.17263364394310937,
"grad_norm": 4.28125,
"learning_rate": 6.901960784313726e-06,
"loss": 0.8059,
"step": 264
},
{
"epoch": 0.1732875592610757,
"grad_norm": 5.90625,
"learning_rate": 6.928104575163399e-06,
"loss": 1.0684,
"step": 265
},
{
"epoch": 0.17394147457904202,
"grad_norm": 4.625,
"learning_rate": 6.954248366013072e-06,
"loss": 0.8969,
"step": 266
},
{
"epoch": 0.17459538989700835,
"grad_norm": 5.1875,
"learning_rate": 6.9803921568627454e-06,
"loss": 0.6325,
"step": 267
},
{
"epoch": 0.17524930521497467,
"grad_norm": 3.484375,
"learning_rate": 7.006535947712418e-06,
"loss": 0.535,
"step": 268
},
{
"epoch": 0.175903220532941,
"grad_norm": 3.921875,
"learning_rate": 7.032679738562092e-06,
"loss": 0.6261,
"step": 269
},
{
"epoch": 0.17655713585090732,
"grad_norm": 3.59375,
"learning_rate": 7.058823529411766e-06,
"loss": 0.49,
"step": 270
},
{
"epoch": 0.17721105116887362,
"grad_norm": 4.65625,
"learning_rate": 7.084967320261439e-06,
"loss": 0.6293,
"step": 271
},
{
"epoch": 0.17786496648683994,
"grad_norm": 4.375,
"learning_rate": 7.111111111111112e-06,
"loss": 0.816,
"step": 272
},
{
"epoch": 0.17851888180480627,
"grad_norm": 5.625,
"learning_rate": 7.137254901960785e-06,
"loss": 1.0568,
"step": 273
},
{
"epoch": 0.1791727971227726,
"grad_norm": 6.53125,
"learning_rate": 7.163398692810458e-06,
"loss": 0.9127,
"step": 274
},
{
"epoch": 0.17982671244073892,
"grad_norm": 4.125,
"learning_rate": 7.189542483660131e-06,
"loss": 0.637,
"step": 275
},
{
"epoch": 0.18048062775870524,
"grad_norm": 5.09375,
"learning_rate": 7.215686274509805e-06,
"loss": 0.7204,
"step": 276
},
{
"epoch": 0.18113454307667157,
"grad_norm": 4.8125,
"learning_rate": 7.241830065359477e-06,
"loss": 0.847,
"step": 277
},
{
"epoch": 0.1817884583946379,
"grad_norm": 4.09375,
"learning_rate": 7.267973856209151e-06,
"loss": 0.8382,
"step": 278
},
{
"epoch": 0.18244237371260422,
"grad_norm": 4.125,
"learning_rate": 7.294117647058823e-06,
"loss": 0.5354,
"step": 279
},
{
"epoch": 0.18309628903057054,
"grad_norm": 5.28125,
"learning_rate": 7.320261437908497e-06,
"loss": 0.9527,
"step": 280
},
{
"epoch": 0.18375020434853687,
"grad_norm": 3.5,
"learning_rate": 7.346405228758171e-06,
"loss": 0.5157,
"step": 281
},
{
"epoch": 0.1844041196665032,
"grad_norm": 2.890625,
"learning_rate": 7.372549019607845e-06,
"loss": 0.3743,
"step": 282
},
{
"epoch": 0.18505803498446952,
"grad_norm": 6.03125,
"learning_rate": 7.398692810457517e-06,
"loss": 0.7489,
"step": 283
},
{
"epoch": 0.18571195030243584,
"grad_norm": 4.28125,
"learning_rate": 7.424836601307191e-06,
"loss": 0.6514,
"step": 284
},
{
"epoch": 0.18636586562040217,
"grad_norm": 4.96875,
"learning_rate": 7.450980392156863e-06,
"loss": 0.8043,
"step": 285
},
{
"epoch": 0.1870197809383685,
"grad_norm": 5.75,
"learning_rate": 7.477124183006537e-06,
"loss": 0.9505,
"step": 286
},
{
"epoch": 0.18767369625633482,
"grad_norm": 4.5625,
"learning_rate": 7.50326797385621e-06,
"loss": 0.7588,
"step": 287
},
{
"epoch": 0.18832761157430114,
"grad_norm": 3.265625,
"learning_rate": 7.529411764705883e-06,
"loss": 0.3853,
"step": 288
},
{
"epoch": 0.18898152689226747,
"grad_norm": 3.640625,
"learning_rate": 7.555555555555556e-06,
"loss": 0.5271,
"step": 289
},
{
"epoch": 0.18963544221023376,
"grad_norm": 6.5625,
"learning_rate": 7.581699346405229e-06,
"loss": 0.9335,
"step": 290
},
{
"epoch": 0.1902893575282001,
"grad_norm": 4.84375,
"learning_rate": 7.607843137254902e-06,
"loss": 0.6355,
"step": 291
},
{
"epoch": 0.1909432728461664,
"grad_norm": 5.84375,
"learning_rate": 7.633986928104575e-06,
"loss": 1.0356,
"step": 292
},
{
"epoch": 0.19159718816413274,
"grad_norm": 5.65625,
"learning_rate": 7.660130718954249e-06,
"loss": 1.033,
"step": 293
},
{
"epoch": 0.19225110348209906,
"grad_norm": 4.40625,
"learning_rate": 7.686274509803923e-06,
"loss": 0.7495,
"step": 294
},
{
"epoch": 0.1929050188000654,
"grad_norm": 4.90625,
"learning_rate": 7.712418300653596e-06,
"loss": 0.8563,
"step": 295
},
{
"epoch": 0.1935589341180317,
"grad_norm": 6.0,
"learning_rate": 7.738562091503269e-06,
"loss": 1.0563,
"step": 296
},
{
"epoch": 0.19421284943599804,
"grad_norm": 5.84375,
"learning_rate": 7.764705882352941e-06,
"loss": 1.0439,
"step": 297
},
{
"epoch": 0.19486676475396436,
"grad_norm": 6.65625,
"learning_rate": 7.790849673202615e-06,
"loss": 1.2193,
"step": 298
},
{
"epoch": 0.19552068007193069,
"grad_norm": 4.625,
"learning_rate": 7.816993464052288e-06,
"loss": 0.6662,
"step": 299
},
{
"epoch": 0.196174595389897,
"grad_norm": 4.25,
"learning_rate": 7.84313725490196e-06,
"loss": 0.6387,
"step": 300
},
{
"epoch": 0.19682851070786334,
"grad_norm": 4.71875,
"learning_rate": 7.869281045751635e-06,
"loss": 0.9447,
"step": 301
},
{
"epoch": 0.19748242602582966,
"grad_norm": 5.5,
"learning_rate": 7.895424836601308e-06,
"loss": 0.7585,
"step": 302
},
{
"epoch": 0.19813634134379599,
"grad_norm": 5.65625,
"learning_rate": 7.92156862745098e-06,
"loss": 0.8031,
"step": 303
},
{
"epoch": 0.1987902566617623,
"grad_norm": 6.15625,
"learning_rate": 7.947712418300654e-06,
"loss": 0.8829,
"step": 304
},
{
"epoch": 0.19944417197972863,
"grad_norm": 5.71875,
"learning_rate": 7.973856209150329e-06,
"loss": 0.9142,
"step": 305
},
{
"epoch": 0.20009808729769496,
"grad_norm": 7.5625,
"learning_rate": 8.000000000000001e-06,
"loss": 1.7219,
"step": 306
},
{
"epoch": 0.20075200261566128,
"grad_norm": 6.21875,
"learning_rate": 8.026143790849674e-06,
"loss": 1.0178,
"step": 307
},
{
"epoch": 0.20140591793362758,
"grad_norm": 5.34375,
"learning_rate": 8.052287581699347e-06,
"loss": 0.9982,
"step": 308
},
{
"epoch": 0.2020598332515939,
"grad_norm": 4.5625,
"learning_rate": 8.07843137254902e-06,
"loss": 0.798,
"step": 309
},
{
"epoch": 0.20271374856956023,
"grad_norm": 5.8125,
"learning_rate": 8.104575163398693e-06,
"loss": 1.3086,
"step": 310
},
{
"epoch": 0.20336766388752656,
"grad_norm": 4.3125,
"learning_rate": 8.130718954248366e-06,
"loss": 0.5503,
"step": 311
},
{
"epoch": 0.20402157920549288,
"grad_norm": 3.65625,
"learning_rate": 8.15686274509804e-06,
"loss": 0.6438,
"step": 312
},
{
"epoch": 0.2046754945234592,
"grad_norm": 3.84375,
"learning_rate": 8.183006535947713e-06,
"loss": 0.4031,
"step": 313
},
{
"epoch": 0.20532940984142553,
"grad_norm": 6.28125,
"learning_rate": 8.209150326797386e-06,
"loss": 1.0666,
"step": 314
},
{
"epoch": 0.20598332515939186,
"grad_norm": 4.875,
"learning_rate": 8.23529411764706e-06,
"loss": 0.9797,
"step": 315
},
{
"epoch": 0.20663724047735818,
"grad_norm": 4.21875,
"learning_rate": 8.261437908496732e-06,
"loss": 0.6984,
"step": 316
},
{
"epoch": 0.2072911557953245,
"grad_norm": 4.34375,
"learning_rate": 8.287581699346407e-06,
"loss": 0.8736,
"step": 317
},
{
"epoch": 0.20794507111329083,
"grad_norm": 4.0,
"learning_rate": 8.31372549019608e-06,
"loss": 0.6156,
"step": 318
},
{
"epoch": 0.20859898643125716,
"grad_norm": 5.65625,
"learning_rate": 8.339869281045752e-06,
"loss": 0.9509,
"step": 319
},
{
"epoch": 0.20925290174922348,
"grad_norm": 4.34375,
"learning_rate": 8.366013071895426e-06,
"loss": 0.8173,
"step": 320
},
{
"epoch": 0.2099068170671898,
"grad_norm": 3.875,
"learning_rate": 8.392156862745099e-06,
"loss": 0.5676,
"step": 321
},
{
"epoch": 0.21056073238515613,
"grad_norm": 5.78125,
"learning_rate": 8.418300653594771e-06,
"loss": 0.7598,
"step": 322
},
{
"epoch": 0.21121464770312245,
"grad_norm": 3.140625,
"learning_rate": 8.444444444444446e-06,
"loss": 0.4609,
"step": 323
},
{
"epoch": 0.21186856302108878,
"grad_norm": 5.34375,
"learning_rate": 8.470588235294118e-06,
"loss": 0.6789,
"step": 324
},
{
"epoch": 0.2125224783390551,
"grad_norm": 4.3125,
"learning_rate": 8.496732026143791e-06,
"loss": 0.6431,
"step": 325
},
{
"epoch": 0.21317639365702143,
"grad_norm": 3.3125,
"learning_rate": 8.522875816993465e-06,
"loss": 0.5432,
"step": 326
},
{
"epoch": 0.21383030897498773,
"grad_norm": 4.84375,
"learning_rate": 8.549019607843138e-06,
"loss": 1.008,
"step": 327
},
{
"epoch": 0.21448422429295405,
"grad_norm": 5.0,
"learning_rate": 8.57516339869281e-06,
"loss": 0.6168,
"step": 328
},
{
"epoch": 0.21513813961092038,
"grad_norm": 5.0,
"learning_rate": 8.601307189542485e-06,
"loss": 0.5513,
"step": 329
},
{
"epoch": 0.2157920549288867,
"grad_norm": 3.953125,
"learning_rate": 8.627450980392157e-06,
"loss": 0.5599,
"step": 330
},
{
"epoch": 0.21644597024685303,
"grad_norm": 7.0,
"learning_rate": 8.653594771241832e-06,
"loss": 1.3891,
"step": 331
},
{
"epoch": 0.21709988556481935,
"grad_norm": 3.34375,
"learning_rate": 8.679738562091504e-06,
"loss": 0.5349,
"step": 332
},
{
"epoch": 0.21775380088278568,
"grad_norm": 5.34375,
"learning_rate": 8.705882352941177e-06,
"loss": 0.6295,
"step": 333
},
{
"epoch": 0.218407716200752,
"grad_norm": 5.0,
"learning_rate": 8.732026143790851e-06,
"loss": 0.5849,
"step": 334
},
{
"epoch": 0.21906163151871832,
"grad_norm": 3.28125,
"learning_rate": 8.758169934640524e-06,
"loss": 0.4733,
"step": 335
},
{
"epoch": 0.21971554683668465,
"grad_norm": 6.875,
"learning_rate": 8.784313725490196e-06,
"loss": 1.281,
"step": 336
},
{
"epoch": 0.22036946215465097,
"grad_norm": 6.21875,
"learning_rate": 8.81045751633987e-06,
"loss": 0.7766,
"step": 337
},
{
"epoch": 0.2210233774726173,
"grad_norm": 3.59375,
"learning_rate": 8.836601307189543e-06,
"loss": 0.6427,
"step": 338
},
{
"epoch": 0.22167729279058362,
"grad_norm": 4.84375,
"learning_rate": 8.862745098039216e-06,
"loss": 0.6244,
"step": 339
},
{
"epoch": 0.22233120810854995,
"grad_norm": 4.5,
"learning_rate": 8.888888888888888e-06,
"loss": 0.6362,
"step": 340
},
{
"epoch": 0.22298512342651627,
"grad_norm": 6.25,
"learning_rate": 8.915032679738563e-06,
"loss": 1.2682,
"step": 341
},
{
"epoch": 0.2236390387444826,
"grad_norm": 4.15625,
"learning_rate": 8.941176470588237e-06,
"loss": 0.6027,
"step": 342
},
{
"epoch": 0.22429295406244892,
"grad_norm": 5.09375,
"learning_rate": 8.96732026143791e-06,
"loss": 0.9463,
"step": 343
},
{
"epoch": 0.22494686938041525,
"grad_norm": 4.5625,
"learning_rate": 8.993464052287582e-06,
"loss": 0.6714,
"step": 344
},
{
"epoch": 0.22560078469838157,
"grad_norm": 4.0,
"learning_rate": 9.019607843137256e-06,
"loss": 0.6505,
"step": 345
},
{
"epoch": 0.22625470001634787,
"grad_norm": 5.15625,
"learning_rate": 9.045751633986929e-06,
"loss": 1.3697,
"step": 346
},
{
"epoch": 0.2269086153343142,
"grad_norm": 5.03125,
"learning_rate": 9.071895424836602e-06,
"loss": 0.8606,
"step": 347
},
{
"epoch": 0.22756253065228052,
"grad_norm": 4.40625,
"learning_rate": 9.098039215686276e-06,
"loss": 0.8106,
"step": 348
},
{
"epoch": 0.22821644597024685,
"grad_norm": 3.640625,
"learning_rate": 9.124183006535948e-06,
"loss": 0.5782,
"step": 349
},
{
"epoch": 0.22887036128821317,
"grad_norm": 5.1875,
"learning_rate": 9.150326797385621e-06,
"loss": 0.9808,
"step": 350
},
{
"epoch": 0.2295242766061795,
"grad_norm": 6.75,
"learning_rate": 9.176470588235294e-06,
"loss": 1.1096,
"step": 351
},
{
"epoch": 0.23017819192414582,
"grad_norm": 5.1875,
"learning_rate": 9.202614379084968e-06,
"loss": 0.729,
"step": 352
},
{
"epoch": 0.23083210724211214,
"grad_norm": 3.28125,
"learning_rate": 9.228758169934642e-06,
"loss": 0.4901,
"step": 353
},
{
"epoch": 0.23148602256007847,
"grad_norm": 5.46875,
"learning_rate": 9.254901960784315e-06,
"loss": 0.8826,
"step": 354
},
{
"epoch": 0.2321399378780448,
"grad_norm": 5.0,
"learning_rate": 9.281045751633987e-06,
"loss": 0.6336,
"step": 355
},
{
"epoch": 0.23279385319601112,
"grad_norm": 5.5,
"learning_rate": 9.307189542483662e-06,
"loss": 0.9505,
"step": 356
},
{
"epoch": 0.23344776851397744,
"grad_norm": 4.1875,
"learning_rate": 9.333333333333334e-06,
"loss": 0.5205,
"step": 357
},
{
"epoch": 0.23410168383194377,
"grad_norm": 3.34375,
"learning_rate": 9.359477124183007e-06,
"loss": 0.5255,
"step": 358
},
{
"epoch": 0.2347555991499101,
"grad_norm": 5.59375,
"learning_rate": 9.385620915032681e-06,
"loss": 0.9878,
"step": 359
},
{
"epoch": 0.23540951446787642,
"grad_norm": 4.6875,
"learning_rate": 9.411764705882354e-06,
"loss": 0.7116,
"step": 360
},
{
"epoch": 0.23606342978584274,
"grad_norm": 3.8125,
"learning_rate": 9.437908496732026e-06,
"loss": 0.5263,
"step": 361
},
{
"epoch": 0.23671734510380907,
"grad_norm": 4.28125,
"learning_rate": 9.464052287581699e-06,
"loss": 0.8144,
"step": 362
},
{
"epoch": 0.2373712604217754,
"grad_norm": 3.8125,
"learning_rate": 9.490196078431373e-06,
"loss": 0.6487,
"step": 363
},
{
"epoch": 0.2380251757397417,
"grad_norm": 5.625,
"learning_rate": 9.516339869281046e-06,
"loss": 0.885,
"step": 364
},
{
"epoch": 0.23867909105770801,
"grad_norm": 3.875,
"learning_rate": 9.54248366013072e-06,
"loss": 0.5814,
"step": 365
},
{
"epoch": 0.23933300637567434,
"grad_norm": 3.890625,
"learning_rate": 9.568627450980393e-06,
"loss": 0.8265,
"step": 366
},
{
"epoch": 0.23998692169364066,
"grad_norm": 6.375,
"learning_rate": 9.594771241830067e-06,
"loss": 1.0435,
"step": 367
},
{
"epoch": 0.240640837011607,
"grad_norm": 4.25,
"learning_rate": 9.62091503267974e-06,
"loss": 0.5925,
"step": 368
},
{
"epoch": 0.24129475232957331,
"grad_norm": 5.09375,
"learning_rate": 9.647058823529412e-06,
"loss": 0.9492,
"step": 369
},
{
"epoch": 0.24194866764753964,
"grad_norm": 4.25,
"learning_rate": 9.673202614379087e-06,
"loss": 0.8301,
"step": 370
},
{
"epoch": 0.24260258296550596,
"grad_norm": 4.625,
"learning_rate": 9.69934640522876e-06,
"loss": 0.718,
"step": 371
},
{
"epoch": 0.2432564982834723,
"grad_norm": 6.28125,
"learning_rate": 9.725490196078432e-06,
"loss": 1.0038,
"step": 372
},
{
"epoch": 0.2439104136014386,
"grad_norm": 6.15625,
"learning_rate": 9.751633986928104e-06,
"loss": 0.9708,
"step": 373
},
{
"epoch": 0.24456432891940494,
"grad_norm": 6.375,
"learning_rate": 9.777777777777779e-06,
"loss": 0.997,
"step": 374
},
{
"epoch": 0.24521824423737126,
"grad_norm": 3.71875,
"learning_rate": 9.803921568627451e-06,
"loss": 0.6004,
"step": 375
},
{
"epoch": 0.2458721595553376,
"grad_norm": 4.65625,
"learning_rate": 9.830065359477124e-06,
"loss": 0.5015,
"step": 376
},
{
"epoch": 0.2465260748733039,
"grad_norm": 4.125,
"learning_rate": 9.856209150326798e-06,
"loss": 0.7668,
"step": 377
},
{
"epoch": 0.24717999019127024,
"grad_norm": 5.34375,
"learning_rate": 9.882352941176472e-06,
"loss": 0.8773,
"step": 378
},
{
"epoch": 0.24783390550923656,
"grad_norm": 3.578125,
"learning_rate": 9.908496732026145e-06,
"loss": 0.5378,
"step": 379
},
{
"epoch": 0.2484878208272029,
"grad_norm": 3.75,
"learning_rate": 9.934640522875818e-06,
"loss": 0.5267,
"step": 380
},
{
"epoch": 0.2491417361451692,
"grad_norm": 3.53125,
"learning_rate": 9.960784313725492e-06,
"loss": 0.3774,
"step": 381
},
{
"epoch": 0.24979565146313554,
"grad_norm": 5.9375,
"learning_rate": 9.986928104575165e-06,
"loss": 1.1173,
"step": 382
},
{
"epoch": 0.25044956678110186,
"grad_norm": 4.0,
"learning_rate": 1.0013071895424837e-05,
"loss": 0.4871,
"step": 383
},
{
"epoch": 0.2511034820990682,
"grad_norm": 3.265625,
"learning_rate": 1.003921568627451e-05,
"loss": 0.3388,
"step": 384
},
{
"epoch": 0.2517573974170345,
"grad_norm": 4.03125,
"learning_rate": 1.0065359477124184e-05,
"loss": 0.464,
"step": 385
},
{
"epoch": 0.25241131273500084,
"grad_norm": 4.71875,
"learning_rate": 1.0091503267973857e-05,
"loss": 0.7948,
"step": 386
},
{
"epoch": 0.25306522805296716,
"grad_norm": 4.0,
"learning_rate": 1.011764705882353e-05,
"loss": 0.6752,
"step": 387
},
{
"epoch": 0.2537191433709335,
"grad_norm": 5.53125,
"learning_rate": 1.0143790849673203e-05,
"loss": 0.8921,
"step": 388
},
{
"epoch": 0.2543730586888998,
"grad_norm": 3.78125,
"learning_rate": 1.0169934640522876e-05,
"loss": 0.5517,
"step": 389
},
{
"epoch": 0.25502697400686614,
"grad_norm": 4.96875,
"learning_rate": 1.0196078431372549e-05,
"loss": 0.8911,
"step": 390
},
{
"epoch": 0.25568088932483246,
"grad_norm": 5.5,
"learning_rate": 1.0222222222222223e-05,
"loss": 0.8359,
"step": 391
},
{
"epoch": 0.25633480464279873,
"grad_norm": 3.4375,
"learning_rate": 1.0248366013071896e-05,
"loss": 0.536,
"step": 392
},
{
"epoch": 0.25698871996076506,
"grad_norm": 4.78125,
"learning_rate": 1.0274509803921568e-05,
"loss": 0.8284,
"step": 393
},
{
"epoch": 0.2576426352787314,
"grad_norm": 5.5625,
"learning_rate": 1.030065359477124e-05,
"loss": 0.7924,
"step": 394
},
{
"epoch": 0.2582965505966977,
"grad_norm": 5.28125,
"learning_rate": 1.0326797385620917e-05,
"loss": 0.9475,
"step": 395
},
{
"epoch": 0.25895046591466403,
"grad_norm": 3.984375,
"learning_rate": 1.035294117647059e-05,
"loss": 0.5109,
"step": 396
},
{
"epoch": 0.25960438123263035,
"grad_norm": 4.5625,
"learning_rate": 1.0379084967320264e-05,
"loss": 0.7566,
"step": 397
},
{
"epoch": 0.2602582965505967,
"grad_norm": 4.15625,
"learning_rate": 1.0405228758169936e-05,
"loss": 0.5593,
"step": 398
},
{
"epoch": 0.260912211868563,
"grad_norm": 4.3125,
"learning_rate": 1.0431372549019609e-05,
"loss": 0.6165,
"step": 399
},
{
"epoch": 0.26156612718652933,
"grad_norm": 5.0,
"learning_rate": 1.0457516339869283e-05,
"loss": 0.8043,
"step": 400
},
{
"epoch": 0.26222004250449565,
"grad_norm": 3.953125,
"learning_rate": 1.0483660130718956e-05,
"loss": 0.5931,
"step": 401
},
{
"epoch": 0.262873957822462,
"grad_norm": 4.9375,
"learning_rate": 1.0509803921568628e-05,
"loss": 0.7563,
"step": 402
},
{
"epoch": 0.2635278731404283,
"grad_norm": 6.90625,
"learning_rate": 1.0535947712418303e-05,
"loss": 1.3466,
"step": 403
},
{
"epoch": 0.26418178845839463,
"grad_norm": 3.4375,
"learning_rate": 1.0562091503267975e-05,
"loss": 0.5283,
"step": 404
},
{
"epoch": 0.26483570377636095,
"grad_norm": 3.765625,
"learning_rate": 1.0588235294117648e-05,
"loss": 0.3773,
"step": 405
},
{
"epoch": 0.2654896190943273,
"grad_norm": 6.40625,
"learning_rate": 1.061437908496732e-05,
"loss": 1.484,
"step": 406
},
{
"epoch": 0.2661435344122936,
"grad_norm": 3.6875,
"learning_rate": 1.0640522875816995e-05,
"loss": 0.5294,
"step": 407
},
{
"epoch": 0.26679744973025993,
"grad_norm": 5.4375,
"learning_rate": 1.0666666666666667e-05,
"loss": 0.7722,
"step": 408
},
{
"epoch": 0.26745136504822625,
"grad_norm": 4.375,
"learning_rate": 1.069281045751634e-05,
"loss": 0.9535,
"step": 409
},
{
"epoch": 0.2681052803661926,
"grad_norm": 4.25,
"learning_rate": 1.0718954248366014e-05,
"loss": 0.735,
"step": 410
},
{
"epoch": 0.2687591956841589,
"grad_norm": 4.53125,
"learning_rate": 1.0745098039215687e-05,
"loss": 0.6053,
"step": 411
},
{
"epoch": 0.2694131110021252,
"grad_norm": 3.6875,
"learning_rate": 1.077124183006536e-05,
"loss": 0.6995,
"step": 412
},
{
"epoch": 0.27006702632009155,
"grad_norm": 4.3125,
"learning_rate": 1.0797385620915034e-05,
"loss": 0.6014,
"step": 413
},
{
"epoch": 0.2707209416380579,
"grad_norm": 2.9375,
"learning_rate": 1.0823529411764706e-05,
"loss": 0.4536,
"step": 414
},
{
"epoch": 0.2713748569560242,
"grad_norm": 4.59375,
"learning_rate": 1.0849673202614379e-05,
"loss": 0.6981,
"step": 415
},
{
"epoch": 0.2720287722739905,
"grad_norm": 6.46875,
"learning_rate": 1.0875816993464051e-05,
"loss": 1.3937,
"step": 416
},
{
"epoch": 0.27268268759195685,
"grad_norm": 4.15625,
"learning_rate": 1.0901960784313726e-05,
"loss": 0.7647,
"step": 417
},
{
"epoch": 0.2733366029099232,
"grad_norm": 4.59375,
"learning_rate": 1.0928104575163398e-05,
"loss": 0.704,
"step": 418
},
{
"epoch": 0.2739905182278895,
"grad_norm": 5.875,
"learning_rate": 1.0954248366013074e-05,
"loss": 0.9878,
"step": 419
},
{
"epoch": 0.2746444335458558,
"grad_norm": 4.625,
"learning_rate": 1.0980392156862747e-05,
"loss": 0.6714,
"step": 420
},
{
"epoch": 0.27529834886382215,
"grad_norm": 4.34375,
"learning_rate": 1.100653594771242e-05,
"loss": 0.9399,
"step": 421
},
{
"epoch": 0.2759522641817885,
"grad_norm": 5.53125,
"learning_rate": 1.1032679738562094e-05,
"loss": 0.8217,
"step": 422
},
{
"epoch": 0.2766061794997548,
"grad_norm": 4.53125,
"learning_rate": 1.1058823529411766e-05,
"loss": 0.6653,
"step": 423
},
{
"epoch": 0.2772600948177211,
"grad_norm": 3.296875,
"learning_rate": 1.1084967320261439e-05,
"loss": 0.4958,
"step": 424
},
{
"epoch": 0.27791401013568745,
"grad_norm": 5.34375,
"learning_rate": 1.1111111111111113e-05,
"loss": 0.9351,
"step": 425
},
{
"epoch": 0.2785679254536538,
"grad_norm": 6.8125,
"learning_rate": 1.1137254901960786e-05,
"loss": 1.5588,
"step": 426
},
{
"epoch": 0.2792218407716201,
"grad_norm": 3.296875,
"learning_rate": 1.1163398692810459e-05,
"loss": 0.4091,
"step": 427
},
{
"epoch": 0.2798757560895864,
"grad_norm": 5.34375,
"learning_rate": 1.1189542483660131e-05,
"loss": 0.5979,
"step": 428
},
{
"epoch": 0.2805296714075527,
"grad_norm": 4.15625,
"learning_rate": 1.1215686274509805e-05,
"loss": 0.6911,
"step": 429
},
{
"epoch": 0.281183586725519,
"grad_norm": 8.0625,
"learning_rate": 1.1241830065359478e-05,
"loss": 1.5556,
"step": 430
},
{
"epoch": 0.28183750204348534,
"grad_norm": 5.3125,
"learning_rate": 1.126797385620915e-05,
"loss": 0.7323,
"step": 431
},
{
"epoch": 0.28249141736145167,
"grad_norm": 3.953125,
"learning_rate": 1.1294117647058825e-05,
"loss": 0.7486,
"step": 432
},
{
"epoch": 0.283145332679418,
"grad_norm": 3.75,
"learning_rate": 1.1320261437908498e-05,
"loss": 0.4753,
"step": 433
},
{
"epoch": 0.2837992479973843,
"grad_norm": 4.125,
"learning_rate": 1.134640522875817e-05,
"loss": 0.5799,
"step": 434
},
{
"epoch": 0.28445316331535064,
"grad_norm": 3.953125,
"learning_rate": 1.1372549019607844e-05,
"loss": 0.4395,
"step": 435
},
{
"epoch": 0.28510707863331697,
"grad_norm": 5.40625,
"learning_rate": 1.1398692810457517e-05,
"loss": 0.945,
"step": 436
},
{
"epoch": 0.2857609939512833,
"grad_norm": 5.4375,
"learning_rate": 1.142483660130719e-05,
"loss": 0.9442,
"step": 437
},
{
"epoch": 0.2864149092692496,
"grad_norm": 4.25,
"learning_rate": 1.1450980392156862e-05,
"loss": 0.573,
"step": 438
},
{
"epoch": 0.28706882458721594,
"grad_norm": 5.34375,
"learning_rate": 1.1477124183006536e-05,
"loss": 0.7825,
"step": 439
},
{
"epoch": 0.28772273990518227,
"grad_norm": 5.5625,
"learning_rate": 1.1503267973856209e-05,
"loss": 1.3477,
"step": 440
},
{
"epoch": 0.2883766552231486,
"grad_norm": 4.0,
"learning_rate": 1.1529411764705882e-05,
"loss": 0.617,
"step": 441
},
{
"epoch": 0.2890305705411149,
"grad_norm": 4.40625,
"learning_rate": 1.1555555555555556e-05,
"loss": 0.7197,
"step": 442
},
{
"epoch": 0.28968448585908124,
"grad_norm": 4.125,
"learning_rate": 1.158169934640523e-05,
"loss": 0.5806,
"step": 443
},
{
"epoch": 0.29033840117704757,
"grad_norm": 4.59375,
"learning_rate": 1.1607843137254905e-05,
"loss": 0.5216,
"step": 444
},
{
"epoch": 0.2909923164950139,
"grad_norm": 6.78125,
"learning_rate": 1.1633986928104577e-05,
"loss": 1.2764,
"step": 445
},
{
"epoch": 0.2916462318129802,
"grad_norm": 3.8125,
"learning_rate": 1.166013071895425e-05,
"loss": 0.5399,
"step": 446
},
{
"epoch": 0.29230014713094654,
"grad_norm": 6.03125,
"learning_rate": 1.1686274509803922e-05,
"loss": 0.919,
"step": 447
},
{
"epoch": 0.29295406244891287,
"grad_norm": 3.1875,
"learning_rate": 1.1712418300653597e-05,
"loss": 0.4369,
"step": 448
},
{
"epoch": 0.2936079777668792,
"grad_norm": 3.09375,
"learning_rate": 1.173856209150327e-05,
"loss": 0.4569,
"step": 449
},
{
"epoch": 0.2942618930848455,
"grad_norm": 3.9375,
"learning_rate": 1.1764705882352942e-05,
"loss": 0.6275,
"step": 450
},
{
"epoch": 0.29491580840281184,
"grad_norm": 4.0625,
"learning_rate": 1.1790849673202616e-05,
"loss": 0.7875,
"step": 451
},
{
"epoch": 0.29556972372077817,
"grad_norm": 5.1875,
"learning_rate": 1.1816993464052289e-05,
"loss": 0.628,
"step": 452
},
{
"epoch": 0.2962236390387445,
"grad_norm": 4.6875,
"learning_rate": 1.1843137254901961e-05,
"loss": 0.6575,
"step": 453
},
{
"epoch": 0.2968775543567108,
"grad_norm": 4.0,
"learning_rate": 1.1869281045751636e-05,
"loss": 0.6567,
"step": 454
},
{
"epoch": 0.29753146967467714,
"grad_norm": 4.0625,
"learning_rate": 1.1895424836601308e-05,
"loss": 0.6079,
"step": 455
},
{
"epoch": 0.29818538499264347,
"grad_norm": 4.125,
"learning_rate": 1.192156862745098e-05,
"loss": 0.6131,
"step": 456
},
{
"epoch": 0.2988393003106098,
"grad_norm": 3.078125,
"learning_rate": 1.1947712418300655e-05,
"loss": 0.3501,
"step": 457
},
{
"epoch": 0.2994932156285761,
"grad_norm": 5.90625,
"learning_rate": 1.1973856209150328e-05,
"loss": 0.7122,
"step": 458
},
{
"epoch": 0.30014713094654244,
"grad_norm": 3.546875,
"learning_rate": 1.2e-05,
"loss": 0.502,
"step": 459
},
{
"epoch": 0.30080104626450876,
"grad_norm": 4.09375,
"learning_rate": 1.2026143790849673e-05,
"loss": 0.4544,
"step": 460
},
{
"epoch": 0.3014549615824751,
"grad_norm": 5.15625,
"learning_rate": 1.2052287581699347e-05,
"loss": 0.676,
"step": 461
},
{
"epoch": 0.3021088769004414,
"grad_norm": 4.1875,
"learning_rate": 1.207843137254902e-05,
"loss": 0.6687,
"step": 462
},
{
"epoch": 0.30276279221840774,
"grad_norm": 4.6875,
"learning_rate": 1.2104575163398692e-05,
"loss": 0.7542,
"step": 463
},
{
"epoch": 0.30341670753637406,
"grad_norm": 6.71875,
"learning_rate": 1.2130718954248367e-05,
"loss": 0.9416,
"step": 464
},
{
"epoch": 0.3040706228543404,
"grad_norm": 4.28125,
"learning_rate": 1.215686274509804e-05,
"loss": 0.5418,
"step": 465
},
{
"epoch": 0.3047245381723067,
"grad_norm": 3.484375,
"learning_rate": 1.2183006535947712e-05,
"loss": 0.5504,
"step": 466
},
{
"epoch": 0.305378453490273,
"grad_norm": 5.25,
"learning_rate": 1.2209150326797388e-05,
"loss": 0.9925,
"step": 467
},
{
"epoch": 0.3060323688082393,
"grad_norm": 5.375,
"learning_rate": 1.223529411764706e-05,
"loss": 0.766,
"step": 468
},
{
"epoch": 0.30668628412620563,
"grad_norm": 3.578125,
"learning_rate": 1.2261437908496733e-05,
"loss": 0.4903,
"step": 469
},
{
"epoch": 0.30734019944417196,
"grad_norm": 4.03125,
"learning_rate": 1.2287581699346407e-05,
"loss": 0.6422,
"step": 470
},
{
"epoch": 0.3079941147621383,
"grad_norm": 5.40625,
"learning_rate": 1.231372549019608e-05,
"loss": 0.9172,
"step": 471
},
{
"epoch": 0.3086480300801046,
"grad_norm": 4.40625,
"learning_rate": 1.2339869281045753e-05,
"loss": 0.5337,
"step": 472
},
{
"epoch": 0.30930194539807093,
"grad_norm": 6.28125,
"learning_rate": 1.2366013071895427e-05,
"loss": 0.6589,
"step": 473
},
{
"epoch": 0.30995586071603726,
"grad_norm": 5.6875,
"learning_rate": 1.23921568627451e-05,
"loss": 1.0247,
"step": 474
},
{
"epoch": 0.3106097760340036,
"grad_norm": 4.625,
"learning_rate": 1.2418300653594772e-05,
"loss": 0.5602,
"step": 475
},
{
"epoch": 0.3112636913519699,
"grad_norm": 6.375,
"learning_rate": 1.2444444444444446e-05,
"loss": 1.0996,
"step": 476
},
{
"epoch": 0.31191760666993623,
"grad_norm": 4.1875,
"learning_rate": 1.2470588235294119e-05,
"loss": 0.4376,
"step": 477
},
{
"epoch": 0.31257152198790256,
"grad_norm": 4.34375,
"learning_rate": 1.2496732026143792e-05,
"loss": 0.7795,
"step": 478
},
{
"epoch": 0.3132254373058689,
"grad_norm": 5.5625,
"learning_rate": 1.2522875816993466e-05,
"loss": 0.9096,
"step": 479
},
{
"epoch": 0.3138793526238352,
"grad_norm": 5.625,
"learning_rate": 1.2549019607843138e-05,
"loss": 0.9408,
"step": 480
},
{
"epoch": 0.31453326794180153,
"grad_norm": 4.9375,
"learning_rate": 1.2575163398692811e-05,
"loss": 0.7616,
"step": 481
},
{
"epoch": 0.31518718325976786,
"grad_norm": 5.34375,
"learning_rate": 1.2601307189542484e-05,
"loss": 0.733,
"step": 482
},
{
"epoch": 0.3158410985777342,
"grad_norm": 2.5625,
"learning_rate": 1.2627450980392158e-05,
"loss": 0.3602,
"step": 483
},
{
"epoch": 0.3164950138957005,
"grad_norm": 5.65625,
"learning_rate": 1.265359477124183e-05,
"loss": 0.6671,
"step": 484
},
{
"epoch": 0.31714892921366683,
"grad_norm": 4.15625,
"learning_rate": 1.2679738562091503e-05,
"loss": 0.4821,
"step": 485
},
{
"epoch": 0.31780284453163316,
"grad_norm": 4.1875,
"learning_rate": 1.2705882352941177e-05,
"loss": 0.7277,
"step": 486
},
{
"epoch": 0.3184567598495995,
"grad_norm": 5.375,
"learning_rate": 1.273202614379085e-05,
"loss": 0.7381,
"step": 487
},
{
"epoch": 0.3191106751675658,
"grad_norm": 4.53125,
"learning_rate": 1.2758169934640523e-05,
"loss": 0.7067,
"step": 488
},
{
"epoch": 0.31976459048553213,
"grad_norm": 5.75,
"learning_rate": 1.2784313725490197e-05,
"loss": 0.6068,
"step": 489
},
{
"epoch": 0.32041850580349845,
"grad_norm": 4.875,
"learning_rate": 1.281045751633987e-05,
"loss": 0.8502,
"step": 490
},
{
"epoch": 0.3210724211214648,
"grad_norm": 6.96875,
"learning_rate": 1.2836601307189544e-05,
"loss": 1.248,
"step": 491
},
{
"epoch": 0.3217263364394311,
"grad_norm": 4.125,
"learning_rate": 1.2862745098039218e-05,
"loss": 0.5831,
"step": 492
},
{
"epoch": 0.32238025175739743,
"grad_norm": 3.21875,
"learning_rate": 1.288888888888889e-05,
"loss": 0.4113,
"step": 493
},
{
"epoch": 0.32303416707536375,
"grad_norm": 5.1875,
"learning_rate": 1.2915032679738563e-05,
"loss": 0.9455,
"step": 494
},
{
"epoch": 0.3236880823933301,
"grad_norm": 5.3125,
"learning_rate": 1.2941176470588238e-05,
"loss": 0.8783,
"step": 495
},
{
"epoch": 0.3243419977112964,
"grad_norm": 4.9375,
"learning_rate": 1.296732026143791e-05,
"loss": 0.9331,
"step": 496
},
{
"epoch": 0.32499591302926273,
"grad_norm": 4.46875,
"learning_rate": 1.2993464052287583e-05,
"loss": 0.7988,
"step": 497
},
{
"epoch": 0.32564982834722905,
"grad_norm": 4.90625,
"learning_rate": 1.3019607843137257e-05,
"loss": 0.7535,
"step": 498
},
{
"epoch": 0.3263037436651954,
"grad_norm": 4.625,
"learning_rate": 1.304575163398693e-05,
"loss": 0.7557,
"step": 499
},
{
"epoch": 0.3269576589831617,
"grad_norm": 3.265625,
"learning_rate": 1.3071895424836602e-05,
"loss": 0.4799,
"step": 500
},
{
"epoch": 0.32761157430112803,
"grad_norm": 4.5625,
"learning_rate": 1.3098039215686275e-05,
"loss": 0.7014,
"step": 501
},
{
"epoch": 0.32826548961909435,
"grad_norm": 3.125,
"learning_rate": 1.3124183006535949e-05,
"loss": 0.4667,
"step": 502
},
{
"epoch": 0.3289194049370607,
"grad_norm": 4.25,
"learning_rate": 1.3150326797385622e-05,
"loss": 0.8005,
"step": 503
},
{
"epoch": 0.32957332025502695,
"grad_norm": 6.3125,
"learning_rate": 1.3176470588235294e-05,
"loss": 1.0702,
"step": 504
},
{
"epoch": 0.33022723557299327,
"grad_norm": 3.4375,
"learning_rate": 1.3202614379084969e-05,
"loss": 0.5992,
"step": 505
},
{
"epoch": 0.3308811508909596,
"grad_norm": 4.28125,
"learning_rate": 1.3228758169934641e-05,
"loss": 0.4368,
"step": 506
},
{
"epoch": 0.3315350662089259,
"grad_norm": 3.9375,
"learning_rate": 1.3254901960784314e-05,
"loss": 0.7551,
"step": 507
},
{
"epoch": 0.33218898152689225,
"grad_norm": 4.84375,
"learning_rate": 1.3281045751633988e-05,
"loss": 0.9635,
"step": 508
},
{
"epoch": 0.33284289684485857,
"grad_norm": 4.71875,
"learning_rate": 1.330718954248366e-05,
"loss": 0.8593,
"step": 509
},
{
"epoch": 0.3334968121628249,
"grad_norm": 6.21875,
"learning_rate": 1.3333333333333333e-05,
"loss": 1.2897,
"step": 510
},
{
"epoch": 0.3341507274807912,
"grad_norm": 3.875,
"learning_rate": 1.3359477124183008e-05,
"loss": 0.7316,
"step": 511
},
{
"epoch": 0.33480464279875755,
"grad_norm": 3.53125,
"learning_rate": 1.338562091503268e-05,
"loss": 0.4614,
"step": 512
},
{
"epoch": 0.33545855811672387,
"grad_norm": 8.8125,
"learning_rate": 1.3411764705882353e-05,
"loss": 1.8679,
"step": 513
},
{
"epoch": 0.3361124734346902,
"grad_norm": 4.21875,
"learning_rate": 1.3437908496732029e-05,
"loss": 0.6266,
"step": 514
},
{
"epoch": 0.3367663887526565,
"grad_norm": 5.6875,
"learning_rate": 1.3464052287581701e-05,
"loss": 0.8198,
"step": 515
},
{
"epoch": 0.33742030407062285,
"grad_norm": 4.6875,
"learning_rate": 1.3490196078431374e-05,
"loss": 0.9571,
"step": 516
},
{
"epoch": 0.33807421938858917,
"grad_norm": 4.9375,
"learning_rate": 1.3516339869281048e-05,
"loss": 0.7718,
"step": 517
},
{
"epoch": 0.3387281347065555,
"grad_norm": 3.84375,
"learning_rate": 1.3542483660130721e-05,
"loss": 0.5037,
"step": 518
},
{
"epoch": 0.3393820500245218,
"grad_norm": 3.21875,
"learning_rate": 1.3568627450980393e-05,
"loss": 0.4967,
"step": 519
},
{
"epoch": 0.34003596534248814,
"grad_norm": 5.25,
"learning_rate": 1.3594771241830068e-05,
"loss": 0.7059,
"step": 520
},
{
"epoch": 0.34068988066045447,
"grad_norm": 5.53125,
"learning_rate": 1.362091503267974e-05,
"loss": 0.7243,
"step": 521
},
{
"epoch": 0.3413437959784208,
"grad_norm": 4.46875,
"learning_rate": 1.3647058823529413e-05,
"loss": 0.7595,
"step": 522
},
{
"epoch": 0.3419977112963871,
"grad_norm": 4.4375,
"learning_rate": 1.3673202614379086e-05,
"loss": 0.7224,
"step": 523
},
{
"epoch": 0.34265162661435344,
"grad_norm": 3.609375,
"learning_rate": 1.369934640522876e-05,
"loss": 0.699,
"step": 524
},
{
"epoch": 0.34330554193231977,
"grad_norm": 4.21875,
"learning_rate": 1.3725490196078432e-05,
"loss": 0.7635,
"step": 525
},
{
"epoch": 0.3439594572502861,
"grad_norm": 4.5,
"learning_rate": 1.3751633986928105e-05,
"loss": 0.7832,
"step": 526
},
{
"epoch": 0.3446133725682524,
"grad_norm": 5.15625,
"learning_rate": 1.377777777777778e-05,
"loss": 0.8386,
"step": 527
},
{
"epoch": 0.34526728788621874,
"grad_norm": 5.34375,
"learning_rate": 1.3803921568627452e-05,
"loss": 1.1034,
"step": 528
},
{
"epoch": 0.34592120320418507,
"grad_norm": 4.8125,
"learning_rate": 1.3830065359477125e-05,
"loss": 0.6836,
"step": 529
},
{
"epoch": 0.3465751185221514,
"grad_norm": 4.8125,
"learning_rate": 1.3856209150326799e-05,
"loss": 0.7655,
"step": 530
},
{
"epoch": 0.3472290338401177,
"grad_norm": 4.5,
"learning_rate": 1.3882352941176471e-05,
"loss": 0.5553,
"step": 531
},
{
"epoch": 0.34788294915808404,
"grad_norm": 5.40625,
"learning_rate": 1.3908496732026144e-05,
"loss": 0.5882,
"step": 532
},
{
"epoch": 0.34853686447605037,
"grad_norm": 3.640625,
"learning_rate": 1.3934640522875818e-05,
"loss": 0.5727,
"step": 533
},
{
"epoch": 0.3491907797940167,
"grad_norm": 4.40625,
"learning_rate": 1.3960784313725491e-05,
"loss": 0.8736,
"step": 534
},
{
"epoch": 0.349844695111983,
"grad_norm": 5.34375,
"learning_rate": 1.3986928104575163e-05,
"loss": 0.5816,
"step": 535
},
{
"epoch": 0.35049861042994934,
"grad_norm": 4.875,
"learning_rate": 1.4013071895424836e-05,
"loss": 0.7251,
"step": 536
},
{
"epoch": 0.35115252574791567,
"grad_norm": 3.703125,
"learning_rate": 1.403921568627451e-05,
"loss": 0.62,
"step": 537
},
{
"epoch": 0.351806441065882,
"grad_norm": 4.46875,
"learning_rate": 1.4065359477124185e-05,
"loss": 0.6138,
"step": 538
},
{
"epoch": 0.3524603563838483,
"grad_norm": 3.59375,
"learning_rate": 1.4091503267973859e-05,
"loss": 0.5634,
"step": 539
},
{
"epoch": 0.35311427170181464,
"grad_norm": 4.96875,
"learning_rate": 1.4117647058823532e-05,
"loss": 0.73,
"step": 540
},
{
"epoch": 0.3537681870197809,
"grad_norm": 3.9375,
"learning_rate": 1.4143790849673204e-05,
"loss": 0.6003,
"step": 541
},
{
"epoch": 0.35442210233774724,
"grad_norm": 5.65625,
"learning_rate": 1.4169934640522878e-05,
"loss": 1.0472,
"step": 542
},
{
"epoch": 0.35507601765571356,
"grad_norm": 3.5625,
"learning_rate": 1.4196078431372551e-05,
"loss": 0.4215,
"step": 543
},
{
"epoch": 0.3557299329736799,
"grad_norm": 5.5625,
"learning_rate": 1.4222222222222224e-05,
"loss": 0.93,
"step": 544
},
{
"epoch": 0.3563838482916462,
"grad_norm": 3.625,
"learning_rate": 1.4248366013071896e-05,
"loss": 0.6136,
"step": 545
},
{
"epoch": 0.35703776360961254,
"grad_norm": 3.09375,
"learning_rate": 1.427450980392157e-05,
"loss": 0.3565,
"step": 546
},
{
"epoch": 0.35769167892757886,
"grad_norm": 4.71875,
"learning_rate": 1.4300653594771243e-05,
"loss": 0.6536,
"step": 547
},
{
"epoch": 0.3583455942455452,
"grad_norm": 4.6875,
"learning_rate": 1.4326797385620916e-05,
"loss": 0.8062,
"step": 548
},
{
"epoch": 0.3589995095635115,
"grad_norm": 3.4375,
"learning_rate": 1.435294117647059e-05,
"loss": 0.5485,
"step": 549
},
{
"epoch": 0.35965342488147783,
"grad_norm": 5.5625,
"learning_rate": 1.4379084967320263e-05,
"loss": 0.8227,
"step": 550
},
{
"epoch": 0.36030734019944416,
"grad_norm": 4.1875,
"learning_rate": 1.4405228758169935e-05,
"loss": 0.7039,
"step": 551
},
{
"epoch": 0.3609612555174105,
"grad_norm": 4.25,
"learning_rate": 1.443137254901961e-05,
"loss": 0.8103,
"step": 552
},
{
"epoch": 0.3616151708353768,
"grad_norm": 3.984375,
"learning_rate": 1.4457516339869282e-05,
"loss": 0.6487,
"step": 553
},
{
"epoch": 0.36226908615334313,
"grad_norm": 3.203125,
"learning_rate": 1.4483660130718955e-05,
"loss": 0.4815,
"step": 554
},
{
"epoch": 0.36292300147130946,
"grad_norm": 3.828125,
"learning_rate": 1.4509803921568629e-05,
"loss": 0.449,
"step": 555
},
{
"epoch": 0.3635769167892758,
"grad_norm": 4.0625,
"learning_rate": 1.4535947712418302e-05,
"loss": 0.5781,
"step": 556
},
{
"epoch": 0.3642308321072421,
"grad_norm": 4.875,
"learning_rate": 1.4562091503267974e-05,
"loss": 1.0315,
"step": 557
},
{
"epoch": 0.36488474742520843,
"grad_norm": 3.765625,
"learning_rate": 1.4588235294117647e-05,
"loss": 0.5251,
"step": 558
},
{
"epoch": 0.36553866274317476,
"grad_norm": 5.75,
"learning_rate": 1.4614379084967321e-05,
"loss": 1.0653,
"step": 559
},
{
"epoch": 0.3661925780611411,
"grad_norm": 4.75,
"learning_rate": 1.4640522875816994e-05,
"loss": 0.9446,
"step": 560
},
{
"epoch": 0.3668464933791074,
"grad_norm": 3.640625,
"learning_rate": 1.4666666666666666e-05,
"loss": 0.6336,
"step": 561
},
{
"epoch": 0.36750040869707373,
"grad_norm": 4.40625,
"learning_rate": 1.4692810457516342e-05,
"loss": 0.7986,
"step": 562
},
{
"epoch": 0.36815432401504006,
"grad_norm": 3.953125,
"learning_rate": 1.4718954248366015e-05,
"loss": 0.6941,
"step": 563
},
{
"epoch": 0.3688082393330064,
"grad_norm": 3.65625,
"learning_rate": 1.474509803921569e-05,
"loss": 0.4157,
"step": 564
},
{
"epoch": 0.3694621546509727,
"grad_norm": 3.53125,
"learning_rate": 1.4771241830065362e-05,
"loss": 0.7389,
"step": 565
},
{
"epoch": 0.37011606996893903,
"grad_norm": 4.21875,
"learning_rate": 1.4797385620915034e-05,
"loss": 0.7562,
"step": 566
},
{
"epoch": 0.37076998528690536,
"grad_norm": 3.796875,
"learning_rate": 1.4823529411764707e-05,
"loss": 0.5432,
"step": 567
},
{
"epoch": 0.3714239006048717,
"grad_norm": 4.78125,
"learning_rate": 1.4849673202614381e-05,
"loss": 0.7261,
"step": 568
},
{
"epoch": 0.372077815922838,
"grad_norm": 3.09375,
"learning_rate": 1.4875816993464054e-05,
"loss": 0.5192,
"step": 569
},
{
"epoch": 0.37273173124080433,
"grad_norm": 6.0625,
"learning_rate": 1.4901960784313726e-05,
"loss": 1.4136,
"step": 570
},
{
"epoch": 0.37338564655877066,
"grad_norm": 4.25,
"learning_rate": 1.49281045751634e-05,
"loss": 0.7003,
"step": 571
},
{
"epoch": 0.374039561876737,
"grad_norm": 5.53125,
"learning_rate": 1.4954248366013073e-05,
"loss": 0.6006,
"step": 572
},
{
"epoch": 0.3746934771947033,
"grad_norm": 3.328125,
"learning_rate": 1.4980392156862746e-05,
"loss": 0.4577,
"step": 573
},
{
"epoch": 0.37534739251266963,
"grad_norm": 6.9375,
"learning_rate": 1.500653594771242e-05,
"loss": 1.0642,
"step": 574
},
{
"epoch": 0.37600130783063596,
"grad_norm": 4.9375,
"learning_rate": 1.5032679738562093e-05,
"loss": 0.6655,
"step": 575
},
{
"epoch": 0.3766552231486023,
"grad_norm": 4.96875,
"learning_rate": 1.5058823529411765e-05,
"loss": 1.0192,
"step": 576
},
{
"epoch": 0.3773091384665686,
"grad_norm": 4.21875,
"learning_rate": 1.5084967320261438e-05,
"loss": 0.5627,
"step": 577
},
{
"epoch": 0.37796305378453493,
"grad_norm": 4.21875,
"learning_rate": 1.5111111111111112e-05,
"loss": 0.717,
"step": 578
},
{
"epoch": 0.3786169691025012,
"grad_norm": 5.4375,
"learning_rate": 1.5137254901960785e-05,
"loss": 1.2168,
"step": 579
},
{
"epoch": 0.3792708844204675,
"grad_norm": 5.03125,
"learning_rate": 1.5163398692810458e-05,
"loss": 0.7631,
"step": 580
},
{
"epoch": 0.37992479973843385,
"grad_norm": 5.03125,
"learning_rate": 1.5189542483660132e-05,
"loss": 0.9894,
"step": 581
},
{
"epoch": 0.3805787150564002,
"grad_norm": 5.09375,
"learning_rate": 1.5215686274509804e-05,
"loss": 0.8453,
"step": 582
},
{
"epoch": 0.3812326303743665,
"grad_norm": 4.6875,
"learning_rate": 1.5241830065359477e-05,
"loss": 0.6673,
"step": 583
},
{
"epoch": 0.3818865456923328,
"grad_norm": 3.765625,
"learning_rate": 1.526797385620915e-05,
"loss": 0.7228,
"step": 584
},
{
"epoch": 0.38254046101029915,
"grad_norm": 4.1875,
"learning_rate": 1.5294117647058822e-05,
"loss": 0.664,
"step": 585
},
{
"epoch": 0.3831943763282655,
"grad_norm": 3.421875,
"learning_rate": 1.5320261437908498e-05,
"loss": 0.5511,
"step": 586
},
{
"epoch": 0.3838482916462318,
"grad_norm": 3.8125,
"learning_rate": 1.534640522875817e-05,
"loss": 0.4787,
"step": 587
},
{
"epoch": 0.3845022069641981,
"grad_norm": 5.15625,
"learning_rate": 1.5372549019607847e-05,
"loss": 1.1154,
"step": 588
},
{
"epoch": 0.38515612228216445,
"grad_norm": 4.8125,
"learning_rate": 1.539869281045752e-05,
"loss": 0.7246,
"step": 589
},
{
"epoch": 0.3858100376001308,
"grad_norm": 5.25,
"learning_rate": 1.5424836601307192e-05,
"loss": 0.8271,
"step": 590
},
{
"epoch": 0.3864639529180971,
"grad_norm": 2.890625,
"learning_rate": 1.5450980392156865e-05,
"loss": 0.3758,
"step": 591
},
{
"epoch": 0.3871178682360634,
"grad_norm": 6.125,
"learning_rate": 1.5477124183006537e-05,
"loss": 0.9127,
"step": 592
},
{
"epoch": 0.38777178355402975,
"grad_norm": 4.8125,
"learning_rate": 1.550326797385621e-05,
"loss": 0.6839,
"step": 593
},
{
"epoch": 0.3884256988719961,
"grad_norm": 4.9375,
"learning_rate": 1.5529411764705882e-05,
"loss": 0.8352,
"step": 594
},
{
"epoch": 0.3890796141899624,
"grad_norm": 4.4375,
"learning_rate": 1.555555555555556e-05,
"loss": 0.7525,
"step": 595
},
{
"epoch": 0.3897335295079287,
"grad_norm": 3.796875,
"learning_rate": 1.558169934640523e-05,
"loss": 0.6019,
"step": 596
},
{
"epoch": 0.39038744482589505,
"grad_norm": 4.21875,
"learning_rate": 1.5607843137254904e-05,
"loss": 0.6474,
"step": 597
},
{
"epoch": 0.39104136014386137,
"grad_norm": 4.5,
"learning_rate": 1.5633986928104576e-05,
"loss": 0.9984,
"step": 598
},
{
"epoch": 0.3916952754618277,
"grad_norm": 4.78125,
"learning_rate": 1.566013071895425e-05,
"loss": 0.9227,
"step": 599
},
{
"epoch": 0.392349190779794,
"grad_norm": 3.359375,
"learning_rate": 1.568627450980392e-05,
"loss": 0.5438,
"step": 600
},
{
"epoch": 0.39300310609776035,
"grad_norm": 6.1875,
"learning_rate": 1.5712418300653597e-05,
"loss": 0.6924,
"step": 601
},
{
"epoch": 0.39365702141572667,
"grad_norm": 3.59375,
"learning_rate": 1.573856209150327e-05,
"loss": 0.6807,
"step": 602
},
{
"epoch": 0.394310936733693,
"grad_norm": 4.4375,
"learning_rate": 1.5764705882352943e-05,
"loss": 0.7607,
"step": 603
},
{
"epoch": 0.3949648520516593,
"grad_norm": 4.15625,
"learning_rate": 1.5790849673202615e-05,
"loss": 0.7465,
"step": 604
},
{
"epoch": 0.39561876736962565,
"grad_norm": 5.5625,
"learning_rate": 1.5816993464052288e-05,
"loss": 0.8777,
"step": 605
},
{
"epoch": 0.39627268268759197,
"grad_norm": 4.78125,
"learning_rate": 1.584313725490196e-05,
"loss": 0.7425,
"step": 606
},
{
"epoch": 0.3969265980055583,
"grad_norm": 2.59375,
"learning_rate": 1.5869281045751633e-05,
"loss": 0.3947,
"step": 607
},
{
"epoch": 0.3975805133235246,
"grad_norm": 4.96875,
"learning_rate": 1.589542483660131e-05,
"loss": 0.4741,
"step": 608
},
{
"epoch": 0.39823442864149095,
"grad_norm": 6.375,
"learning_rate": 1.592156862745098e-05,
"loss": 1.4103,
"step": 609
},
{
"epoch": 0.39888834395945727,
"grad_norm": 3.859375,
"learning_rate": 1.5947712418300657e-05,
"loss": 0.5789,
"step": 610
},
{
"epoch": 0.3995422592774236,
"grad_norm": 5.4375,
"learning_rate": 1.597385620915033e-05,
"loss": 1.2326,
"step": 611
},
{
"epoch": 0.4001961745953899,
"grad_norm": 4.65625,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.829,
"step": 612
},
{
"epoch": 0.40085008991335624,
"grad_norm": 5.71875,
"learning_rate": 1.6026143790849675e-05,
"loss": 0.6386,
"step": 613
},
{
"epoch": 0.40150400523132257,
"grad_norm": 4.625,
"learning_rate": 1.6052287581699348e-05,
"loss": 0.9493,
"step": 614
},
{
"epoch": 0.4021579205492889,
"grad_norm": 3.328125,
"learning_rate": 1.607843137254902e-05,
"loss": 0.4558,
"step": 615
},
{
"epoch": 0.40281183586725516,
"grad_norm": 4.46875,
"learning_rate": 1.6104575163398693e-05,
"loss": 0.9592,
"step": 616
},
{
"epoch": 0.4034657511852215,
"grad_norm": 5.65625,
"learning_rate": 1.613071895424837e-05,
"loss": 0.938,
"step": 617
},
{
"epoch": 0.4041196665031878,
"grad_norm": 5.71875,
"learning_rate": 1.615686274509804e-05,
"loss": 0.9973,
"step": 618
},
{
"epoch": 0.40477358182115414,
"grad_norm": 3.390625,
"learning_rate": 1.6183006535947714e-05,
"loss": 0.3333,
"step": 619
},
{
"epoch": 0.40542749713912046,
"grad_norm": 3.65625,
"learning_rate": 1.6209150326797387e-05,
"loss": 0.6231,
"step": 620
},
{
"epoch": 0.4060814124570868,
"grad_norm": 5.3125,
"learning_rate": 1.623529411764706e-05,
"loss": 0.7472,
"step": 621
},
{
"epoch": 0.4067353277750531,
"grad_norm": 4.6875,
"learning_rate": 1.6261437908496732e-05,
"loss": 0.6525,
"step": 622
},
{
"epoch": 0.40738924309301944,
"grad_norm": 4.71875,
"learning_rate": 1.6287581699346408e-05,
"loss": 0.696,
"step": 623
},
{
"epoch": 0.40804315841098576,
"grad_norm": 7.34375,
"learning_rate": 1.631372549019608e-05,
"loss": 0.945,
"step": 624
},
{
"epoch": 0.4086970737289521,
"grad_norm": 4.46875,
"learning_rate": 1.6339869281045753e-05,
"loss": 0.6496,
"step": 625
},
{
"epoch": 0.4093509890469184,
"grad_norm": 4.96875,
"learning_rate": 1.6366013071895426e-05,
"loss": 1.0453,
"step": 626
},
{
"epoch": 0.41000490436488474,
"grad_norm": 4.28125,
"learning_rate": 1.63921568627451e-05,
"loss": 0.5734,
"step": 627
},
{
"epoch": 0.41065881968285106,
"grad_norm": 4.28125,
"learning_rate": 1.641830065359477e-05,
"loss": 0.4982,
"step": 628
},
{
"epoch": 0.4113127350008174,
"grad_norm": 3.625,
"learning_rate": 1.6444444444444444e-05,
"loss": 0.5462,
"step": 629
},
{
"epoch": 0.4119666503187837,
"grad_norm": 3.671875,
"learning_rate": 1.647058823529412e-05,
"loss": 0.5412,
"step": 630
},
{
"epoch": 0.41262056563675004,
"grad_norm": 3.9375,
"learning_rate": 1.6496732026143792e-05,
"loss": 0.6907,
"step": 631
},
{
"epoch": 0.41327448095471636,
"grad_norm": 7.03125,
"learning_rate": 1.6522875816993465e-05,
"loss": 1.0787,
"step": 632
},
{
"epoch": 0.4139283962726827,
"grad_norm": 2.71875,
"learning_rate": 1.6549019607843137e-05,
"loss": 0.4138,
"step": 633
},
{
"epoch": 0.414582311590649,
"grad_norm": 6.4375,
"learning_rate": 1.6575163398692813e-05,
"loss": 0.7656,
"step": 634
},
{
"epoch": 0.41523622690861534,
"grad_norm": 3.75,
"learning_rate": 1.6601307189542486e-05,
"loss": 0.625,
"step": 635
},
{
"epoch": 0.41589014222658166,
"grad_norm": 4.46875,
"learning_rate": 1.662745098039216e-05,
"loss": 0.5313,
"step": 636
},
{
"epoch": 0.416544057544548,
"grad_norm": 4.90625,
"learning_rate": 1.665359477124183e-05,
"loss": 0.7668,
"step": 637
},
{
"epoch": 0.4171979728625143,
"grad_norm": 3.96875,
"learning_rate": 1.6679738562091504e-05,
"loss": 0.6051,
"step": 638
},
{
"epoch": 0.41785188818048064,
"grad_norm": 3.890625,
"learning_rate": 1.670588235294118e-05,
"loss": 0.5768,
"step": 639
},
{
"epoch": 0.41850580349844696,
"grad_norm": 4.0,
"learning_rate": 1.6732026143790852e-05,
"loss": 0.5606,
"step": 640
},
{
"epoch": 0.4191597188164133,
"grad_norm": 4.34375,
"learning_rate": 1.6758169934640525e-05,
"loss": 0.6251,
"step": 641
},
{
"epoch": 0.4198136341343796,
"grad_norm": 4.0625,
"learning_rate": 1.6784313725490198e-05,
"loss": 0.4984,
"step": 642
},
{
"epoch": 0.42046754945234593,
"grad_norm": 3.640625,
"learning_rate": 1.681045751633987e-05,
"loss": 0.6556,
"step": 643
},
{
"epoch": 0.42112146477031226,
"grad_norm": 4.96875,
"learning_rate": 1.6836601307189543e-05,
"loss": 0.8022,
"step": 644
},
{
"epoch": 0.4217753800882786,
"grad_norm": 5.125,
"learning_rate": 1.686274509803922e-05,
"loss": 0.7004,
"step": 645
},
{
"epoch": 0.4224292954062449,
"grad_norm": 5.34375,
"learning_rate": 1.688888888888889e-05,
"loss": 0.6712,
"step": 646
},
{
"epoch": 0.42308321072421123,
"grad_norm": 4.9375,
"learning_rate": 1.6915032679738564e-05,
"loss": 0.9576,
"step": 647
},
{
"epoch": 0.42373712604217756,
"grad_norm": 3.515625,
"learning_rate": 1.6941176470588237e-05,
"loss": 0.603,
"step": 648
},
{
"epoch": 0.4243910413601439,
"grad_norm": 5.03125,
"learning_rate": 1.696732026143791e-05,
"loss": 0.8925,
"step": 649
},
{
"epoch": 0.4250449566781102,
"grad_norm": 4.9375,
"learning_rate": 1.6993464052287582e-05,
"loss": 1.1251,
"step": 650
},
{
"epoch": 0.42569887199607653,
"grad_norm": 4.65625,
"learning_rate": 1.7019607843137254e-05,
"loss": 0.7117,
"step": 651
},
{
"epoch": 0.42635278731404286,
"grad_norm": 5.53125,
"learning_rate": 1.704575163398693e-05,
"loss": 0.8326,
"step": 652
},
{
"epoch": 0.4270067026320091,
"grad_norm": 2.453125,
"learning_rate": 1.7071895424836603e-05,
"loss": 0.3195,
"step": 653
},
{
"epoch": 0.42766061794997545,
"grad_norm": 3.453125,
"learning_rate": 1.7098039215686275e-05,
"loss": 0.4999,
"step": 654
},
{
"epoch": 0.4283145332679418,
"grad_norm": 3.359375,
"learning_rate": 1.7124183006535948e-05,
"loss": 0.4422,
"step": 655
},
{
"epoch": 0.4289684485859081,
"grad_norm": 3.921875,
"learning_rate": 1.715032679738562e-05,
"loss": 0.5043,
"step": 656
},
{
"epoch": 0.4296223639038744,
"grad_norm": 4.375,
"learning_rate": 1.7176470588235293e-05,
"loss": 0.6563,
"step": 657
},
{
"epoch": 0.43027627922184075,
"grad_norm": 4.15625,
"learning_rate": 1.720261437908497e-05,
"loss": 0.3584,
"step": 658
},
{
"epoch": 0.4309301945398071,
"grad_norm": 4.5,
"learning_rate": 1.7228758169934642e-05,
"loss": 0.7562,
"step": 659
},
{
"epoch": 0.4315841098577734,
"grad_norm": 4.84375,
"learning_rate": 1.7254901960784314e-05,
"loss": 0.5622,
"step": 660
},
{
"epoch": 0.4322380251757397,
"grad_norm": 4.90625,
"learning_rate": 1.728104575163399e-05,
"loss": 0.9408,
"step": 661
},
{
"epoch": 0.43289194049370605,
"grad_norm": 4.46875,
"learning_rate": 1.7307189542483663e-05,
"loss": 0.7223,
"step": 662
},
{
"epoch": 0.4335458558116724,
"grad_norm": 4.4375,
"learning_rate": 1.7333333333333336e-05,
"loss": 0.5794,
"step": 663
},
{
"epoch": 0.4341997711296387,
"grad_norm": 5.0625,
"learning_rate": 1.7359477124183008e-05,
"loss": 0.5965,
"step": 664
},
{
"epoch": 0.434853686447605,
"grad_norm": 5.0,
"learning_rate": 1.738562091503268e-05,
"loss": 0.8629,
"step": 665
},
{
"epoch": 0.43550760176557135,
"grad_norm": 3.328125,
"learning_rate": 1.7411764705882353e-05,
"loss": 0.3594,
"step": 666
},
{
"epoch": 0.4361615170835377,
"grad_norm": 6.46875,
"learning_rate": 1.743790849673203e-05,
"loss": 0.8079,
"step": 667
},
{
"epoch": 0.436815432401504,
"grad_norm": 8.4375,
"learning_rate": 1.7464052287581702e-05,
"loss": 1.1402,
"step": 668
},
{
"epoch": 0.4374693477194703,
"grad_norm": 4.03125,
"learning_rate": 1.7490196078431375e-05,
"loss": 0.6158,
"step": 669
},
{
"epoch": 0.43812326303743665,
"grad_norm": 5.0625,
"learning_rate": 1.7516339869281047e-05,
"loss": 0.8871,
"step": 670
},
{
"epoch": 0.438777178355403,
"grad_norm": 3.734375,
"learning_rate": 1.754248366013072e-05,
"loss": 0.6452,
"step": 671
},
{
"epoch": 0.4394310936733693,
"grad_norm": 3.703125,
"learning_rate": 1.7568627450980392e-05,
"loss": 0.5196,
"step": 672
},
{
"epoch": 0.4400850089913356,
"grad_norm": 3.921875,
"learning_rate": 1.7594771241830065e-05,
"loss": 0.4932,
"step": 673
},
{
"epoch": 0.44073892430930195,
"grad_norm": 5.6875,
"learning_rate": 1.762091503267974e-05,
"loss": 0.896,
"step": 674
},
{
"epoch": 0.4413928396272683,
"grad_norm": 3.25,
"learning_rate": 1.7647058823529414e-05,
"loss": 0.3271,
"step": 675
},
{
"epoch": 0.4420467549452346,
"grad_norm": 4.65625,
"learning_rate": 1.7673202614379086e-05,
"loss": 0.6974,
"step": 676
},
{
"epoch": 0.4427006702632009,
"grad_norm": 5.46875,
"learning_rate": 1.769934640522876e-05,
"loss": 0.7297,
"step": 677
},
{
"epoch": 0.44335458558116725,
"grad_norm": 4.59375,
"learning_rate": 1.772549019607843e-05,
"loss": 0.8317,
"step": 678
},
{
"epoch": 0.4440085008991336,
"grad_norm": 4.96875,
"learning_rate": 1.7751633986928104e-05,
"loss": 0.8311,
"step": 679
},
{
"epoch": 0.4446624162170999,
"grad_norm": 4.6875,
"learning_rate": 1.7777777777777777e-05,
"loss": 0.8124,
"step": 680
},
{
"epoch": 0.4453163315350662,
"grad_norm": 3.53125,
"learning_rate": 1.7803921568627453e-05,
"loss": 0.5505,
"step": 681
},
{
"epoch": 0.44597024685303255,
"grad_norm": 3.609375,
"learning_rate": 1.7830065359477125e-05,
"loss": 0.472,
"step": 682
},
{
"epoch": 0.4466241621709989,
"grad_norm": 5.6875,
"learning_rate": 1.78562091503268e-05,
"loss": 1.0624,
"step": 683
},
{
"epoch": 0.4472780774889652,
"grad_norm": 6.375,
"learning_rate": 1.7882352941176474e-05,
"loss": 1.0482,
"step": 684
},
{
"epoch": 0.4479319928069315,
"grad_norm": 6.53125,
"learning_rate": 1.7908496732026146e-05,
"loss": 1.0284,
"step": 685
},
{
"epoch": 0.44858590812489785,
"grad_norm": 5.5625,
"learning_rate": 1.793464052287582e-05,
"loss": 0.8013,
"step": 686
},
{
"epoch": 0.4492398234428642,
"grad_norm": 3.78125,
"learning_rate": 1.796078431372549e-05,
"loss": 0.6488,
"step": 687
},
{
"epoch": 0.4498937387608305,
"grad_norm": 5.40625,
"learning_rate": 1.7986928104575164e-05,
"loss": 0.8196,
"step": 688
},
{
"epoch": 0.4505476540787968,
"grad_norm": 3.3125,
"learning_rate": 1.801307189542484e-05,
"loss": 0.4819,
"step": 689
},
{
"epoch": 0.45120156939676315,
"grad_norm": 5.21875,
"learning_rate": 1.8039215686274513e-05,
"loss": 0.754,
"step": 690
},
{
"epoch": 0.4518554847147294,
"grad_norm": 4.375,
"learning_rate": 1.8065359477124185e-05,
"loss": 0.8142,
"step": 691
},
{
"epoch": 0.45250940003269574,
"grad_norm": 3.90625,
"learning_rate": 1.8091503267973858e-05,
"loss": 0.6999,
"step": 692
},
{
"epoch": 0.45316331535066207,
"grad_norm": 3.65625,
"learning_rate": 1.811764705882353e-05,
"loss": 0.4873,
"step": 693
},
{
"epoch": 0.4538172306686284,
"grad_norm": 4.25,
"learning_rate": 1.8143790849673203e-05,
"loss": 0.8854,
"step": 694
},
{
"epoch": 0.4544711459865947,
"grad_norm": 5.09375,
"learning_rate": 1.8169934640522876e-05,
"loss": 0.7774,
"step": 695
},
{
"epoch": 0.45512506130456104,
"grad_norm": 4.53125,
"learning_rate": 1.8196078431372552e-05,
"loss": 0.6882,
"step": 696
},
{
"epoch": 0.45577897662252737,
"grad_norm": 6.875,
"learning_rate": 1.8222222222222224e-05,
"loss": 0.9632,
"step": 697
},
{
"epoch": 0.4564328919404937,
"grad_norm": 3.90625,
"learning_rate": 1.8248366013071897e-05,
"loss": 0.5186,
"step": 698
},
{
"epoch": 0.45708680725846,
"grad_norm": 5.15625,
"learning_rate": 1.827450980392157e-05,
"loss": 0.6747,
"step": 699
},
{
"epoch": 0.45774072257642634,
"grad_norm": 4.875,
"learning_rate": 1.8300653594771242e-05,
"loss": 0.6845,
"step": 700
},
{
"epoch": 0.45839463789439266,
"grad_norm": 6.71875,
"learning_rate": 1.8326797385620915e-05,
"loss": 1.0584,
"step": 701
},
{
"epoch": 0.459048553212359,
"grad_norm": 4.1875,
"learning_rate": 1.8352941176470587e-05,
"loss": 0.6146,
"step": 702
},
{
"epoch": 0.4597024685303253,
"grad_norm": 4.1875,
"learning_rate": 1.8379084967320263e-05,
"loss": 0.5963,
"step": 703
},
{
"epoch": 0.46035638384829164,
"grad_norm": 4.78125,
"learning_rate": 1.8405228758169936e-05,
"loss": 0.5975,
"step": 704
},
{
"epoch": 0.46101029916625796,
"grad_norm": 5.9375,
"learning_rate": 1.843137254901961e-05,
"loss": 0.9296,
"step": 705
},
{
"epoch": 0.4616642144842243,
"grad_norm": 5.78125,
"learning_rate": 1.8457516339869284e-05,
"loss": 1.1027,
"step": 706
},
{
"epoch": 0.4623181298021906,
"grad_norm": 5.0,
"learning_rate": 1.8483660130718957e-05,
"loss": 0.8167,
"step": 707
},
{
"epoch": 0.46297204512015694,
"grad_norm": 5.46875,
"learning_rate": 1.850980392156863e-05,
"loss": 0.9061,
"step": 708
},
{
"epoch": 0.46362596043812326,
"grad_norm": 5.3125,
"learning_rate": 1.8535947712418302e-05,
"loss": 0.8982,
"step": 709
},
{
"epoch": 0.4642798757560896,
"grad_norm": 4.53125,
"learning_rate": 1.8562091503267975e-05,
"loss": 0.7445,
"step": 710
},
{
"epoch": 0.4649337910740559,
"grad_norm": 4.0625,
"learning_rate": 1.8588235294117647e-05,
"loss": 0.4744,
"step": 711
},
{
"epoch": 0.46558770639202224,
"grad_norm": 5.75,
"learning_rate": 1.8614379084967323e-05,
"loss": 0.8194,
"step": 712
},
{
"epoch": 0.46624162170998856,
"grad_norm": 3.4375,
"learning_rate": 1.8640522875816996e-05,
"loss": 0.4436,
"step": 713
},
{
"epoch": 0.4668955370279549,
"grad_norm": 5.3125,
"learning_rate": 1.866666666666667e-05,
"loss": 0.8432,
"step": 714
},
{
"epoch": 0.4675494523459212,
"grad_norm": 4.46875,
"learning_rate": 1.869281045751634e-05,
"loss": 0.7775,
"step": 715
},
{
"epoch": 0.46820336766388754,
"grad_norm": 4.53125,
"learning_rate": 1.8718954248366014e-05,
"loss": 0.6029,
"step": 716
},
{
"epoch": 0.46885728298185386,
"grad_norm": 4.375,
"learning_rate": 1.8745098039215686e-05,
"loss": 0.6191,
"step": 717
},
{
"epoch": 0.4695111982998202,
"grad_norm": 4.21875,
"learning_rate": 1.8771241830065362e-05,
"loss": 0.5841,
"step": 718
},
{
"epoch": 0.4701651136177865,
"grad_norm": 4.0625,
"learning_rate": 1.8797385620915035e-05,
"loss": 0.6045,
"step": 719
},
{
"epoch": 0.47081902893575284,
"grad_norm": 4.75,
"learning_rate": 1.8823529411764708e-05,
"loss": 0.975,
"step": 720
},
{
"epoch": 0.47147294425371916,
"grad_norm": 3.78125,
"learning_rate": 1.884967320261438e-05,
"loss": 0.5637,
"step": 721
},
{
"epoch": 0.4721268595716855,
"grad_norm": 4.34375,
"learning_rate": 1.8875816993464053e-05,
"loss": 0.7553,
"step": 722
},
{
"epoch": 0.4727807748896518,
"grad_norm": 3.1875,
"learning_rate": 1.8901960784313725e-05,
"loss": 0.4431,
"step": 723
},
{
"epoch": 0.47343469020761814,
"grad_norm": 3.84375,
"learning_rate": 1.8928104575163398e-05,
"loss": 0.5161,
"step": 724
},
{
"epoch": 0.47408860552558446,
"grad_norm": 5.34375,
"learning_rate": 1.8954248366013074e-05,
"loss": 0.9629,
"step": 725
},
{
"epoch": 0.4747425208435508,
"grad_norm": 4.53125,
"learning_rate": 1.8980392156862747e-05,
"loss": 0.7033,
"step": 726
},
{
"epoch": 0.4753964361615171,
"grad_norm": 4.65625,
"learning_rate": 1.900653594771242e-05,
"loss": 0.7945,
"step": 727
},
{
"epoch": 0.4760503514794834,
"grad_norm": 3.953125,
"learning_rate": 1.9032679738562092e-05,
"loss": 0.6442,
"step": 728
},
{
"epoch": 0.4767042667974497,
"grad_norm": 4.09375,
"learning_rate": 1.9058823529411764e-05,
"loss": 0.6704,
"step": 729
},
{
"epoch": 0.47735818211541603,
"grad_norm": 5.40625,
"learning_rate": 1.908496732026144e-05,
"loss": 0.8781,
"step": 730
},
{
"epoch": 0.47801209743338235,
"grad_norm": 6.03125,
"learning_rate": 1.9111111111111113e-05,
"loss": 0.9963,
"step": 731
},
{
"epoch": 0.4786660127513487,
"grad_norm": 5.125,
"learning_rate": 1.9137254901960786e-05,
"loss": 0.7436,
"step": 732
},
{
"epoch": 0.479319928069315,
"grad_norm": 5.125,
"learning_rate": 1.9163398692810458e-05,
"loss": 0.8493,
"step": 733
},
{
"epoch": 0.47997384338728133,
"grad_norm": 4.5625,
"learning_rate": 1.9189542483660134e-05,
"loss": 0.8925,
"step": 734
},
{
"epoch": 0.48062775870524765,
"grad_norm": 2.859375,
"learning_rate": 1.9215686274509807e-05,
"loss": 0.3353,
"step": 735
},
{
"epoch": 0.481281674023214,
"grad_norm": 5.8125,
"learning_rate": 1.924183006535948e-05,
"loss": 0.9909,
"step": 736
},
{
"epoch": 0.4819355893411803,
"grad_norm": 4.34375,
"learning_rate": 1.9267973856209152e-05,
"loss": 0.8178,
"step": 737
},
{
"epoch": 0.48258950465914663,
"grad_norm": 4.90625,
"learning_rate": 1.9294117647058825e-05,
"loss": 0.6734,
"step": 738
},
{
"epoch": 0.48324341997711295,
"grad_norm": 3.09375,
"learning_rate": 1.9320261437908497e-05,
"loss": 0.4815,
"step": 739
},
{
"epoch": 0.4838973352950793,
"grad_norm": 2.78125,
"learning_rate": 1.9346405228758173e-05,
"loss": 0.408,
"step": 740
},
{
"epoch": 0.4845512506130456,
"grad_norm": 5.40625,
"learning_rate": 1.9372549019607846e-05,
"loss": 0.8701,
"step": 741
},
{
"epoch": 0.48520516593101193,
"grad_norm": 3.90625,
"learning_rate": 1.939869281045752e-05,
"loss": 0.7292,
"step": 742
},
{
"epoch": 0.48585908124897825,
"grad_norm": 3.796875,
"learning_rate": 1.942483660130719e-05,
"loss": 0.5935,
"step": 743
},
{
"epoch": 0.4865129965669446,
"grad_norm": 4.90625,
"learning_rate": 1.9450980392156864e-05,
"loss": 0.7148,
"step": 744
},
{
"epoch": 0.4871669118849109,
"grad_norm": 4.09375,
"learning_rate": 1.9477124183006536e-05,
"loss": 0.4538,
"step": 745
},
{
"epoch": 0.4878208272028772,
"grad_norm": 5.09375,
"learning_rate": 1.950326797385621e-05,
"loss": 0.9908,
"step": 746
},
{
"epoch": 0.48847474252084355,
"grad_norm": 4.125,
"learning_rate": 1.9529411764705885e-05,
"loss": 0.5186,
"step": 747
},
{
"epoch": 0.4891286578388099,
"grad_norm": 4.28125,
"learning_rate": 1.9555555555555557e-05,
"loss": 0.4834,
"step": 748
},
{
"epoch": 0.4897825731567762,
"grad_norm": 3.4375,
"learning_rate": 1.958169934640523e-05,
"loss": 0.5935,
"step": 749
},
{
"epoch": 0.4904364884747425,
"grad_norm": 4.8125,
"learning_rate": 1.9607843137254903e-05,
"loss": 0.7119,
"step": 750
},
{
"epoch": 0.49109040379270885,
"grad_norm": 6.71875,
"learning_rate": 1.9633986928104575e-05,
"loss": 0.7678,
"step": 751
},
{
"epoch": 0.4917443191106752,
"grad_norm": 6.625,
"learning_rate": 1.9660130718954248e-05,
"loss": 1.0624,
"step": 752
},
{
"epoch": 0.4923982344286415,
"grad_norm": 3.953125,
"learning_rate": 1.9686274509803924e-05,
"loss": 0.4809,
"step": 753
},
{
"epoch": 0.4930521497466078,
"grad_norm": 5.875,
"learning_rate": 1.9712418300653596e-05,
"loss": 0.664,
"step": 754
},
{
"epoch": 0.49370606506457415,
"grad_norm": 5.5,
"learning_rate": 1.973856209150327e-05,
"loss": 0.721,
"step": 755
},
{
"epoch": 0.4943599803825405,
"grad_norm": 4.375,
"learning_rate": 1.9764705882352945e-05,
"loss": 0.5999,
"step": 756
},
{
"epoch": 0.4950138957005068,
"grad_norm": 3.46875,
"learning_rate": 1.9790849673202617e-05,
"loss": 0.4176,
"step": 757
},
{
"epoch": 0.4956678110184731,
"grad_norm": 5.875,
"learning_rate": 1.981699346405229e-05,
"loss": 0.8356,
"step": 758
},
{
"epoch": 0.49632172633643945,
"grad_norm": 2.875,
"learning_rate": 1.9843137254901963e-05,
"loss": 0.4127,
"step": 759
},
{
"epoch": 0.4969756416544058,
"grad_norm": 3.640625,
"learning_rate": 1.9869281045751635e-05,
"loss": 0.5169,
"step": 760
},
{
"epoch": 0.4976295569723721,
"grad_norm": 5.15625,
"learning_rate": 1.9895424836601308e-05,
"loss": 0.8059,
"step": 761
},
{
"epoch": 0.4982834722903384,
"grad_norm": 5.5,
"learning_rate": 1.9921568627450984e-05,
"loss": 0.9276,
"step": 762
},
{
"epoch": 0.49893738760830475,
"grad_norm": 9.0625,
"learning_rate": 1.9947712418300656e-05,
"loss": 1.4035,
"step": 763
},
{
"epoch": 0.4995913029262711,
"grad_norm": 3.734375,
"learning_rate": 1.997385620915033e-05,
"loss": 0.6315,
"step": 764
},
{
"epoch": 0.5002452182442374,
"grad_norm": 4.125,
"learning_rate": 2e-05,
"loss": 0.5016,
"step": 765
},
{
"epoch": 0.5008991335622037,
"grad_norm": 5.25,
"learning_rate": 1.9999998957459666e-05,
"loss": 0.4525,
"step": 766
},
{
"epoch": 0.50155304888017,
"grad_norm": 4.0,
"learning_rate": 1.9999995829838888e-05,
"loss": 0.4469,
"step": 767
},
{
"epoch": 0.5022069641981364,
"grad_norm": 4.21875,
"learning_rate": 1.9999990617138306e-05,
"loss": 0.7654,
"step": 768
},
{
"epoch": 0.5028608795161027,
"grad_norm": 4.875,
"learning_rate": 1.999998331935902e-05,
"loss": 0.72,
"step": 769
},
{
"epoch": 0.503514794834069,
"grad_norm": 5.125,
"learning_rate": 1.9999973936502538e-05,
"loss": 0.8575,
"step": 770
},
{
"epoch": 0.5041687101520353,
"grad_norm": 5.1875,
"learning_rate": 1.999996246857083e-05,
"loss": 0.6663,
"step": 771
},
{
"epoch": 0.5048226254700017,
"grad_norm": 5.53125,
"learning_rate": 1.999994891556628e-05,
"loss": 0.8865,
"step": 772
},
{
"epoch": 0.505476540787968,
"grad_norm": 5.46875,
"learning_rate": 1.9999933277491715e-05,
"loss": 0.75,
"step": 773
},
{
"epoch": 0.5061304561059343,
"grad_norm": 3.078125,
"learning_rate": 1.9999915554350395e-05,
"loss": 0.4187,
"step": 774
},
{
"epoch": 0.5067843714239006,
"grad_norm": 3.40625,
"learning_rate": 1.999989574614602e-05,
"loss": 0.4447,
"step": 775
},
{
"epoch": 0.507438286741867,
"grad_norm": 3.953125,
"learning_rate": 1.9999873852882714e-05,
"loss": 0.6833,
"step": 776
},
{
"epoch": 0.5080922020598333,
"grad_norm": 4.15625,
"learning_rate": 1.999984987456504e-05,
"loss": 0.5476,
"step": 777
},
{
"epoch": 0.5087461173777996,
"grad_norm": 4.46875,
"learning_rate": 1.9999823811198006e-05,
"loss": 0.6096,
"step": 778
},
{
"epoch": 0.509400032695766,
"grad_norm": 3.921875,
"learning_rate": 1.9999795662787043e-05,
"loss": 0.5435,
"step": 779
},
{
"epoch": 0.5100539480137323,
"grad_norm": 3.5625,
"learning_rate": 1.9999765429338018e-05,
"loss": 0.5435,
"step": 780
},
{
"epoch": 0.5107078633316986,
"grad_norm": 4.1875,
"learning_rate": 1.9999733110857237e-05,
"loss": 0.6079,
"step": 781
},
{
"epoch": 0.5113617786496649,
"grad_norm": 4.125,
"learning_rate": 1.999969870735144e-05,
"loss": 0.5706,
"step": 782
},
{
"epoch": 0.5120156939676312,
"grad_norm": 5.09375,
"learning_rate": 1.9999662218827794e-05,
"loss": 0.9238,
"step": 783
},
{
"epoch": 0.5126696092855975,
"grad_norm": 5.9375,
"learning_rate": 1.9999623645293915e-05,
"loss": 1.1608,
"step": 784
},
{
"epoch": 0.5133235246035638,
"grad_norm": 4.75,
"learning_rate": 1.999958298675784e-05,
"loss": 0.7148,
"step": 785
},
{
"epoch": 0.5139774399215301,
"grad_norm": 5.40625,
"learning_rate": 1.9999540243228052e-05,
"loss": 0.8556,
"step": 786
},
{
"epoch": 0.5146313552394964,
"grad_norm": 4.34375,
"learning_rate": 1.999949541471346e-05,
"loss": 0.6286,
"step": 787
},
{
"epoch": 0.5152852705574628,
"grad_norm": 3.734375,
"learning_rate": 1.999944850122341e-05,
"loss": 0.4143,
"step": 788
},
{
"epoch": 0.5159391858754291,
"grad_norm": 3.984375,
"learning_rate": 1.9999399502767686e-05,
"loss": 0.4789,
"step": 789
},
{
"epoch": 0.5165931011933954,
"grad_norm": 4.21875,
"learning_rate": 1.9999348419356507e-05,
"loss": 0.6679,
"step": 790
},
{
"epoch": 0.5172470165113617,
"grad_norm": 3.96875,
"learning_rate": 1.9999295251000517e-05,
"loss": 0.5674,
"step": 791
},
{
"epoch": 0.5179009318293281,
"grad_norm": 3.03125,
"learning_rate": 1.999923999771081e-05,
"loss": 0.2701,
"step": 792
},
{
"epoch": 0.5185548471472944,
"grad_norm": 4.625,
"learning_rate": 1.9999182659498905e-05,
"loss": 0.7045,
"step": 793
},
{
"epoch": 0.5192087624652607,
"grad_norm": 4.28125,
"learning_rate": 1.999912323637675e-05,
"loss": 0.7503,
"step": 794
},
{
"epoch": 0.519862677783227,
"grad_norm": 4.375,
"learning_rate": 1.9999061728356743e-05,
"loss": 0.6953,
"step": 795
},
{
"epoch": 0.5205165931011934,
"grad_norm": 5.34375,
"learning_rate": 1.999899813545171e-05,
"loss": 0.6895,
"step": 796
},
{
"epoch": 0.5211705084191597,
"grad_norm": 4.46875,
"learning_rate": 1.9998932457674904e-05,
"loss": 0.6713,
"step": 797
},
{
"epoch": 0.521824423737126,
"grad_norm": 4.0625,
"learning_rate": 1.999886469504002e-05,
"loss": 0.5524,
"step": 798
},
{
"epoch": 0.5224783390550923,
"grad_norm": 4.4375,
"learning_rate": 1.9998794847561197e-05,
"loss": 0.6095,
"step": 799
},
{
"epoch": 0.5231322543730587,
"grad_norm": 4.15625,
"learning_rate": 1.9998722915252987e-05,
"loss": 0.5745,
"step": 800
},
{
"epoch": 0.523786169691025,
"grad_norm": 4.21875,
"learning_rate": 1.9998648898130396e-05,
"loss": 0.69,
"step": 801
},
{
"epoch": 0.5244400850089913,
"grad_norm": 4.5,
"learning_rate": 1.9998572796208848e-05,
"loss": 0.6694,
"step": 802
},
{
"epoch": 0.5250940003269576,
"grad_norm": 5.90625,
"learning_rate": 1.9998494609504223e-05,
"loss": 0.884,
"step": 803
},
{
"epoch": 0.525747915644924,
"grad_norm": 4.625,
"learning_rate": 1.9998414338032817e-05,
"loss": 0.8186,
"step": 804
},
{
"epoch": 0.5264018309628903,
"grad_norm": 4.40625,
"learning_rate": 1.999833198181137e-05,
"loss": 0.6467,
"step": 805
},
{
"epoch": 0.5270557462808566,
"grad_norm": 4.5625,
"learning_rate": 1.9998247540857045e-05,
"loss": 0.7452,
"step": 806
},
{
"epoch": 0.5277096615988229,
"grad_norm": 6.59375,
"learning_rate": 1.999816101518746e-05,
"loss": 1.2817,
"step": 807
},
{
"epoch": 0.5283635769167893,
"grad_norm": 4.125,
"learning_rate": 1.9998072404820648e-05,
"loss": 0.577,
"step": 808
},
{
"epoch": 0.5290174922347556,
"grad_norm": 3.953125,
"learning_rate": 1.9997981709775096e-05,
"loss": 0.4899,
"step": 809
},
{
"epoch": 0.5296714075527219,
"grad_norm": 5.21875,
"learning_rate": 1.99978889300697e-05,
"loss": 0.8487,
"step": 810
},
{
"epoch": 0.5303253228706882,
"grad_norm": 6.625,
"learning_rate": 1.9997794065723818e-05,
"loss": 1.3262,
"step": 811
},
{
"epoch": 0.5309792381886546,
"grad_norm": 4.75,
"learning_rate": 1.9997697116757222e-05,
"loss": 0.6515,
"step": 812
},
{
"epoch": 0.5316331535066209,
"grad_norm": 7.25,
"learning_rate": 1.999759808319013e-05,
"loss": 0.4887,
"step": 813
},
{
"epoch": 0.5322870688245872,
"grad_norm": 4.46875,
"learning_rate": 1.999749696504319e-05,
"loss": 0.8626,
"step": 814
},
{
"epoch": 0.5329409841425535,
"grad_norm": 5.75,
"learning_rate": 1.9997393762337486e-05,
"loss": 0.6561,
"step": 815
},
{
"epoch": 0.5335948994605199,
"grad_norm": 4.34375,
"learning_rate": 1.999728847509454e-05,
"loss": 0.5553,
"step": 816
},
{
"epoch": 0.5342488147784862,
"grad_norm": 3.078125,
"learning_rate": 1.9997181103336302e-05,
"loss": 0.4297,
"step": 817
},
{
"epoch": 0.5349027300964525,
"grad_norm": 6.03125,
"learning_rate": 1.9997071647085158e-05,
"loss": 1.0966,
"step": 818
},
{
"epoch": 0.5355566454144188,
"grad_norm": 4.625,
"learning_rate": 1.9996960106363932e-05,
"loss": 0.7236,
"step": 819
},
{
"epoch": 0.5362105607323852,
"grad_norm": 3.625,
"learning_rate": 1.9996846481195884e-05,
"loss": 0.5967,
"step": 820
},
{
"epoch": 0.5368644760503515,
"grad_norm": 6.40625,
"learning_rate": 1.99967307716047e-05,
"loss": 1.2281,
"step": 821
},
{
"epoch": 0.5375183913683178,
"grad_norm": 3.515625,
"learning_rate": 1.9996612977614518e-05,
"loss": 0.4814,
"step": 822
},
{
"epoch": 0.5381723066862841,
"grad_norm": 4.28125,
"learning_rate": 1.9996493099249884e-05,
"loss": 0.6375,
"step": 823
},
{
"epoch": 0.5388262220042505,
"grad_norm": 4.25,
"learning_rate": 1.9996371136535803e-05,
"loss": 0.5511,
"step": 824
},
{
"epoch": 0.5394801373222168,
"grad_norm": 3.78125,
"learning_rate": 1.9996247089497703e-05,
"loss": 0.4829,
"step": 825
},
{
"epoch": 0.5401340526401831,
"grad_norm": 4.0,
"learning_rate": 1.999612095816145e-05,
"loss": 0.5908,
"step": 826
},
{
"epoch": 0.5407879679581494,
"grad_norm": 3.03125,
"learning_rate": 1.9995992742553342e-05,
"loss": 0.466,
"step": 827
},
{
"epoch": 0.5414418832761158,
"grad_norm": 3.546875,
"learning_rate": 1.9995862442700112e-05,
"loss": 0.5842,
"step": 828
},
{
"epoch": 0.5420957985940821,
"grad_norm": 4.4375,
"learning_rate": 1.9995730058628928e-05,
"loss": 0.7486,
"step": 829
},
{
"epoch": 0.5427497139120484,
"grad_norm": 4.125,
"learning_rate": 1.9995595590367397e-05,
"loss": 0.4746,
"step": 830
},
{
"epoch": 0.5434036292300147,
"grad_norm": 4.8125,
"learning_rate": 1.9995459037943556e-05,
"loss": 0.7395,
"step": 831
},
{
"epoch": 0.544057544547981,
"grad_norm": 5.875,
"learning_rate": 1.999532040138587e-05,
"loss": 0.9984,
"step": 832
},
{
"epoch": 0.5447114598659474,
"grad_norm": 4.09375,
"learning_rate": 1.9995179680723256e-05,
"loss": 0.6687,
"step": 833
},
{
"epoch": 0.5453653751839137,
"grad_norm": 5.8125,
"learning_rate": 1.999503687598505e-05,
"loss": 1.459,
"step": 834
},
{
"epoch": 0.54601929050188,
"grad_norm": 2.9375,
"learning_rate": 1.9994891987201033e-05,
"loss": 0.3902,
"step": 835
},
{
"epoch": 0.5466732058198464,
"grad_norm": 5.09375,
"learning_rate": 1.9994745014401407e-05,
"loss": 0.7955,
"step": 836
},
{
"epoch": 0.5473271211378127,
"grad_norm": 3.546875,
"learning_rate": 1.999459595761682e-05,
"loss": 0.5094,
"step": 837
},
{
"epoch": 0.547981036455779,
"grad_norm": 3.765625,
"learning_rate": 1.9994444816878354e-05,
"loss": 0.4971,
"step": 838
},
{
"epoch": 0.5486349517737453,
"grad_norm": 4.03125,
"learning_rate": 1.9994291592217527e-05,
"loss": 0.6496,
"step": 839
},
{
"epoch": 0.5492888670917117,
"grad_norm": 7.03125,
"learning_rate": 1.999413628366628e-05,
"loss": 1.149,
"step": 840
},
{
"epoch": 0.549942782409678,
"grad_norm": 3.4375,
"learning_rate": 1.9993978891256997e-05,
"loss": 0.4271,
"step": 841
},
{
"epoch": 0.5505966977276443,
"grad_norm": 4.71875,
"learning_rate": 1.99938194150225e-05,
"loss": 0.7827,
"step": 842
},
{
"epoch": 0.5512506130456106,
"grad_norm": 4.375,
"learning_rate": 1.9993657854996033e-05,
"loss": 0.4376,
"step": 843
},
{
"epoch": 0.551904528363577,
"grad_norm": 3.765625,
"learning_rate": 1.9993494211211292e-05,
"loss": 0.51,
"step": 844
},
{
"epoch": 0.5525584436815433,
"grad_norm": 5.28125,
"learning_rate": 1.9993328483702393e-05,
"loss": 0.9663,
"step": 845
},
{
"epoch": 0.5532123589995096,
"grad_norm": 3.6875,
"learning_rate": 1.9993160672503896e-05,
"loss": 0.6493,
"step": 846
},
{
"epoch": 0.5538662743174759,
"grad_norm": 5.125,
"learning_rate": 1.9992990777650784e-05,
"loss": 0.8078,
"step": 847
},
{
"epoch": 0.5545201896354423,
"grad_norm": 5.09375,
"learning_rate": 1.9992818799178485e-05,
"loss": 0.8382,
"step": 848
},
{
"epoch": 0.5551741049534086,
"grad_norm": 4.09375,
"learning_rate": 1.9992644737122863e-05,
"loss": 0.5921,
"step": 849
},
{
"epoch": 0.5558280202713749,
"grad_norm": 4.40625,
"learning_rate": 1.9992468591520198e-05,
"loss": 0.6852,
"step": 850
},
{
"epoch": 0.5564819355893412,
"grad_norm": 4.09375,
"learning_rate": 1.9992290362407232e-05,
"loss": 0.5955,
"step": 851
},
{
"epoch": 0.5571358509073076,
"grad_norm": 5.09375,
"learning_rate": 1.9992110049821118e-05,
"loss": 0.9488,
"step": 852
},
{
"epoch": 0.5577897662252739,
"grad_norm": 5.34375,
"learning_rate": 1.9991927653799458e-05,
"loss": 0.5971,
"step": 853
},
{
"epoch": 0.5584436815432402,
"grad_norm": 4.1875,
"learning_rate": 1.999174317438028e-05,
"loss": 0.5554,
"step": 854
},
{
"epoch": 0.5590975968612065,
"grad_norm": 3.03125,
"learning_rate": 1.999155661160205e-05,
"loss": 0.4821,
"step": 855
},
{
"epoch": 0.5597515121791728,
"grad_norm": 7.125,
"learning_rate": 1.9991367965503667e-05,
"loss": 1.1827,
"step": 856
},
{
"epoch": 0.5604054274971392,
"grad_norm": 4.78125,
"learning_rate": 1.9991177236124467e-05,
"loss": 0.7167,
"step": 857
},
{
"epoch": 0.5610593428151054,
"grad_norm": 6.46875,
"learning_rate": 1.999098442350422e-05,
"loss": 1.2377,
"step": 858
},
{
"epoch": 0.5617132581330717,
"grad_norm": 6.09375,
"learning_rate": 1.9990789527683127e-05,
"loss": 0.9011,
"step": 859
},
{
"epoch": 0.562367173451038,
"grad_norm": 4.25,
"learning_rate": 1.9990592548701824e-05,
"loss": 0.684,
"step": 860
},
{
"epoch": 0.5630210887690044,
"grad_norm": 4.78125,
"learning_rate": 1.9990393486601385e-05,
"loss": 0.6646,
"step": 861
},
{
"epoch": 0.5636750040869707,
"grad_norm": 6.09375,
"learning_rate": 1.9990192341423313e-05,
"loss": 0.7725,
"step": 862
},
{
"epoch": 0.564328919404937,
"grad_norm": 3.703125,
"learning_rate": 1.998998911320955e-05,
"loss": 0.5421,
"step": 863
},
{
"epoch": 0.5649828347229033,
"grad_norm": 4.03125,
"learning_rate": 1.9989783802002477e-05,
"loss": 0.5503,
"step": 864
},
{
"epoch": 0.5656367500408697,
"grad_norm": 6.0,
"learning_rate": 1.9989576407844894e-05,
"loss": 0.9568,
"step": 865
},
{
"epoch": 0.566290665358836,
"grad_norm": 4.5,
"learning_rate": 1.9989366930780046e-05,
"loss": 0.5848,
"step": 866
},
{
"epoch": 0.5669445806768023,
"grad_norm": 4.5625,
"learning_rate": 1.9989155370851613e-05,
"loss": 0.6084,
"step": 867
},
{
"epoch": 0.5675984959947686,
"grad_norm": 3.6875,
"learning_rate": 1.9988941728103707e-05,
"loss": 0.5925,
"step": 868
},
{
"epoch": 0.568252411312735,
"grad_norm": 5.71875,
"learning_rate": 1.9988726002580874e-05,
"loss": 0.8152,
"step": 869
},
{
"epoch": 0.5689063266307013,
"grad_norm": 4.40625,
"learning_rate": 1.9988508194328093e-05,
"loss": 0.7085,
"step": 870
},
{
"epoch": 0.5695602419486676,
"grad_norm": 6.90625,
"learning_rate": 1.998828830339078e-05,
"loss": 1.0088,
"step": 871
},
{
"epoch": 0.5702141572666339,
"grad_norm": 3.09375,
"learning_rate": 1.9988066329814785e-05,
"loss": 0.3421,
"step": 872
},
{
"epoch": 0.5708680725846003,
"grad_norm": 6.21875,
"learning_rate": 1.9987842273646388e-05,
"loss": 1.2265,
"step": 873
},
{
"epoch": 0.5715219879025666,
"grad_norm": 3.0625,
"learning_rate": 1.998761613493231e-05,
"loss": 0.4234,
"step": 874
},
{
"epoch": 0.5721759032205329,
"grad_norm": 4.25,
"learning_rate": 1.9987387913719698e-05,
"loss": 0.6454,
"step": 875
},
{
"epoch": 0.5728298185384992,
"grad_norm": 5.5625,
"learning_rate": 1.9987157610056144e-05,
"loss": 0.686,
"step": 876
},
{
"epoch": 0.5734837338564656,
"grad_norm": 4.9375,
"learning_rate": 1.9986925223989665e-05,
"loss": 0.718,
"step": 877
},
{
"epoch": 0.5741376491744319,
"grad_norm": 5.59375,
"learning_rate": 1.998669075556872e-05,
"loss": 0.9246,
"step": 878
},
{
"epoch": 0.5747915644923982,
"grad_norm": 4.375,
"learning_rate": 1.998645420484219e-05,
"loss": 0.7671,
"step": 879
},
{
"epoch": 0.5754454798103645,
"grad_norm": 4.03125,
"learning_rate": 1.99862155718594e-05,
"loss": 0.3656,
"step": 880
},
{
"epoch": 0.5760993951283309,
"grad_norm": 3.40625,
"learning_rate": 1.998597485667011e-05,
"loss": 0.4916,
"step": 881
},
{
"epoch": 0.5767533104462972,
"grad_norm": 3.4375,
"learning_rate": 1.998573205932451e-05,
"loss": 0.4893,
"step": 882
},
{
"epoch": 0.5774072257642635,
"grad_norm": 4.59375,
"learning_rate": 1.998548717987322e-05,
"loss": 0.5658,
"step": 883
},
{
"epoch": 0.5780611410822298,
"grad_norm": 4.59375,
"learning_rate": 1.9985240218367305e-05,
"loss": 0.7497,
"step": 884
},
{
"epoch": 0.5787150564001962,
"grad_norm": 5.0,
"learning_rate": 1.998499117485826e-05,
"loss": 0.8593,
"step": 885
},
{
"epoch": 0.5793689717181625,
"grad_norm": 3.796875,
"learning_rate": 1.9984740049398007e-05,
"loss": 0.6118,
"step": 886
},
{
"epoch": 0.5800228870361288,
"grad_norm": 4.5625,
"learning_rate": 1.9984486842038912e-05,
"loss": 0.6922,
"step": 887
},
{
"epoch": 0.5806768023540951,
"grad_norm": 4.65625,
"learning_rate": 1.998423155283377e-05,
"loss": 0.6188,
"step": 888
},
{
"epoch": 0.5813307176720615,
"grad_norm": 3.4375,
"learning_rate": 1.9983974181835808e-05,
"loss": 0.3675,
"step": 889
},
{
"epoch": 0.5819846329900278,
"grad_norm": 4.25,
"learning_rate": 1.9983714729098695e-05,
"loss": 0.5906,
"step": 890
},
{
"epoch": 0.5826385483079941,
"grad_norm": 5.125,
"learning_rate": 1.998345319467652e-05,
"loss": 0.8265,
"step": 891
},
{
"epoch": 0.5832924636259604,
"grad_norm": 3.5,
"learning_rate": 1.9983189578623824e-05,
"loss": 0.472,
"step": 892
},
{
"epoch": 0.5839463789439268,
"grad_norm": 4.90625,
"learning_rate": 1.998292388099557e-05,
"loss": 0.7463,
"step": 893
},
{
"epoch": 0.5846002942618931,
"grad_norm": 3.5625,
"learning_rate": 1.998265610184716e-05,
"loss": 0.5639,
"step": 894
},
{
"epoch": 0.5852542095798594,
"grad_norm": 4.65625,
"learning_rate": 1.9982386241234426e-05,
"loss": 0.7212,
"step": 895
},
{
"epoch": 0.5859081248978257,
"grad_norm": 3.140625,
"learning_rate": 1.9982114299213634e-05,
"loss": 0.3914,
"step": 896
},
{
"epoch": 0.5865620402157921,
"grad_norm": 5.25,
"learning_rate": 1.9981840275841492e-05,
"loss": 1.0893,
"step": 897
},
{
"epoch": 0.5872159555337584,
"grad_norm": 5.09375,
"learning_rate": 1.998156417117513e-05,
"loss": 1.2687,
"step": 898
},
{
"epoch": 0.5878698708517247,
"grad_norm": 3.609375,
"learning_rate": 1.998128598527212e-05,
"loss": 0.4849,
"step": 899
},
{
"epoch": 0.588523786169691,
"grad_norm": 4.6875,
"learning_rate": 1.998100571819047e-05,
"loss": 0.5872,
"step": 900
},
{
"epoch": 0.5891777014876574,
"grad_norm": 5.59375,
"learning_rate": 1.998072336998861e-05,
"loss": 1.1697,
"step": 901
},
{
"epoch": 0.5898316168056237,
"grad_norm": 5.375,
"learning_rate": 1.998043894072542e-05,
"loss": 1.0084,
"step": 902
},
{
"epoch": 0.59048553212359,
"grad_norm": 5.4375,
"learning_rate": 1.99801524304602e-05,
"loss": 0.592,
"step": 903
},
{
"epoch": 0.5911394474415563,
"grad_norm": 3.453125,
"learning_rate": 1.9979863839252693e-05,
"loss": 0.6599,
"step": 904
},
{
"epoch": 0.5917933627595227,
"grad_norm": 4.1875,
"learning_rate": 1.997957316716307e-05,
"loss": 0.698,
"step": 905
},
{
"epoch": 0.592447278077489,
"grad_norm": 4.6875,
"learning_rate": 1.997928041425194e-05,
"loss": 0.5678,
"step": 906
},
{
"epoch": 0.5931011933954553,
"grad_norm": 6.09375,
"learning_rate": 1.9978985580580348e-05,
"loss": 0.9745,
"step": 907
},
{
"epoch": 0.5937551087134216,
"grad_norm": 4.8125,
"learning_rate": 1.9978688666209766e-05,
"loss": 0.8067,
"step": 908
},
{
"epoch": 0.594409024031388,
"grad_norm": 4.1875,
"learning_rate": 1.9978389671202096e-05,
"loss": 0.5082,
"step": 909
},
{
"epoch": 0.5950629393493543,
"grad_norm": 5.28125,
"learning_rate": 1.9978088595619692e-05,
"loss": 0.9207,
"step": 910
},
{
"epoch": 0.5957168546673206,
"grad_norm": 5.09375,
"learning_rate": 1.9977785439525328e-05,
"loss": 0.6936,
"step": 911
},
{
"epoch": 0.5963707699852869,
"grad_norm": 3.859375,
"learning_rate": 1.9977480202982208e-05,
"loss": 0.6724,
"step": 912
},
{
"epoch": 0.5970246853032533,
"grad_norm": 4.34375,
"learning_rate": 1.9977172886053982e-05,
"loss": 0.4092,
"step": 913
},
{
"epoch": 0.5976786006212196,
"grad_norm": 3.109375,
"learning_rate": 1.997686348880473e-05,
"loss": 0.542,
"step": 914
},
{
"epoch": 0.5983325159391859,
"grad_norm": 4.21875,
"learning_rate": 1.997655201129896e-05,
"loss": 0.6357,
"step": 915
},
{
"epoch": 0.5989864312571522,
"grad_norm": 4.8125,
"learning_rate": 1.9976238453601614e-05,
"loss": 0.5035,
"step": 916
},
{
"epoch": 0.5996403465751186,
"grad_norm": 3.65625,
"learning_rate": 1.9975922815778082e-05,
"loss": 0.6335,
"step": 917
},
{
"epoch": 0.6002942618930849,
"grad_norm": 5.34375,
"learning_rate": 1.9975605097894165e-05,
"loss": 0.9339,
"step": 918
},
{
"epoch": 0.6009481772110512,
"grad_norm": 4.0625,
"learning_rate": 1.997528530001612e-05,
"loss": 0.7369,
"step": 919
},
{
"epoch": 0.6016020925290175,
"grad_norm": 6.46875,
"learning_rate": 1.9974963422210623e-05,
"loss": 0.6741,
"step": 920
},
{
"epoch": 0.6022560078469839,
"grad_norm": 4.09375,
"learning_rate": 1.9974639464544785e-05,
"loss": 0.7056,
"step": 921
},
{
"epoch": 0.6029099231649502,
"grad_norm": 6.875,
"learning_rate": 1.9974313427086157e-05,
"loss": 0.8128,
"step": 922
},
{
"epoch": 0.6035638384829165,
"grad_norm": 4.28125,
"learning_rate": 1.9973985309902722e-05,
"loss": 0.7754,
"step": 923
},
{
"epoch": 0.6042177538008828,
"grad_norm": 3.8125,
"learning_rate": 1.9973655113062895e-05,
"loss": 0.4393,
"step": 924
},
{
"epoch": 0.6048716691188492,
"grad_norm": 3.71875,
"learning_rate": 1.9973322836635517e-05,
"loss": 0.4985,
"step": 925
},
{
"epoch": 0.6055255844368155,
"grad_norm": 3.78125,
"learning_rate": 1.997298848068988e-05,
"loss": 0.5001,
"step": 926
},
{
"epoch": 0.6061794997547818,
"grad_norm": 3.328125,
"learning_rate": 1.9972652045295698e-05,
"loss": 0.4583,
"step": 927
},
{
"epoch": 0.6068334150727481,
"grad_norm": 3.859375,
"learning_rate": 1.997231353052312e-05,
"loss": 0.5878,
"step": 928
},
{
"epoch": 0.6074873303907145,
"grad_norm": 4.25,
"learning_rate": 1.997197293644272e-05,
"loss": 0.6283,
"step": 929
},
{
"epoch": 0.6081412457086808,
"grad_norm": 6.4375,
"learning_rate": 1.997163026312553e-05,
"loss": 0.7797,
"step": 930
},
{
"epoch": 0.6087951610266471,
"grad_norm": 4.9375,
"learning_rate": 1.997128551064299e-05,
"loss": 0.8994,
"step": 931
},
{
"epoch": 0.6094490763446134,
"grad_norm": 3.03125,
"learning_rate": 1.9970938679066984e-05,
"loss": 0.3363,
"step": 932
},
{
"epoch": 0.6101029916625796,
"grad_norm": 4.84375,
"learning_rate": 1.9970589768469833e-05,
"loss": 0.8361,
"step": 933
},
{
"epoch": 0.610756906980546,
"grad_norm": 5.28125,
"learning_rate": 1.9970238778924286e-05,
"loss": 0.7096,
"step": 934
},
{
"epoch": 0.6114108222985123,
"grad_norm": 4.15625,
"learning_rate": 1.996988571050353e-05,
"loss": 0.5665,
"step": 935
},
{
"epoch": 0.6120647376164786,
"grad_norm": 5.90625,
"learning_rate": 1.9969530563281177e-05,
"loss": 0.9535,
"step": 936
},
{
"epoch": 0.6127186529344449,
"grad_norm": 7.125,
"learning_rate": 1.9969173337331283e-05,
"loss": 0.7664,
"step": 937
},
{
"epoch": 0.6133725682524113,
"grad_norm": 2.609375,
"learning_rate": 1.996881403272833e-05,
"loss": 0.3357,
"step": 938
},
{
"epoch": 0.6140264835703776,
"grad_norm": 3.828125,
"learning_rate": 1.996845264954723e-05,
"loss": 0.5738,
"step": 939
},
{
"epoch": 0.6146803988883439,
"grad_norm": 4.78125,
"learning_rate": 1.996808918786335e-05,
"loss": 0.7155,
"step": 940
},
{
"epoch": 0.6153343142063102,
"grad_norm": 3.765625,
"learning_rate": 1.9967723647752463e-05,
"loss": 0.7199,
"step": 941
},
{
"epoch": 0.6159882295242766,
"grad_norm": 3.96875,
"learning_rate": 1.9967356029290786e-05,
"loss": 0.6848,
"step": 942
},
{
"epoch": 0.6166421448422429,
"grad_norm": 5.0,
"learning_rate": 1.9966986332554973e-05,
"loss": 0.7918,
"step": 943
},
{
"epoch": 0.6172960601602092,
"grad_norm": 5.4375,
"learning_rate": 1.9966614557622112e-05,
"loss": 0.7112,
"step": 944
},
{
"epoch": 0.6179499754781755,
"grad_norm": 3.984375,
"learning_rate": 1.9966240704569722e-05,
"loss": 0.5883,
"step": 945
},
{
"epoch": 0.6186038907961419,
"grad_norm": 4.4375,
"learning_rate": 1.996586477347575e-05,
"loss": 0.6443,
"step": 946
},
{
"epoch": 0.6192578061141082,
"grad_norm": 4.40625,
"learning_rate": 1.9965486764418577e-05,
"loss": 0.6256,
"step": 947
},
{
"epoch": 0.6199117214320745,
"grad_norm": 3.390625,
"learning_rate": 1.996510667747703e-05,
"loss": 0.4967,
"step": 948
},
{
"epoch": 0.6205656367500408,
"grad_norm": 2.828125,
"learning_rate": 1.9964724512730352e-05,
"loss": 0.3946,
"step": 949
},
{
"epoch": 0.6212195520680072,
"grad_norm": 4.59375,
"learning_rate": 1.9964340270258233e-05,
"loss": 0.8035,
"step": 950
},
{
"epoch": 0.6218734673859735,
"grad_norm": 5.8125,
"learning_rate": 1.996395395014079e-05,
"loss": 0.9117,
"step": 951
},
{
"epoch": 0.6225273827039398,
"grad_norm": 3.640625,
"learning_rate": 1.9963565552458572e-05,
"loss": 0.4545,
"step": 952
},
{
"epoch": 0.6231812980219061,
"grad_norm": 3.796875,
"learning_rate": 1.996317507729256e-05,
"loss": 0.565,
"step": 953
},
{
"epoch": 0.6238352133398725,
"grad_norm": 4.3125,
"learning_rate": 1.996278252472418e-05,
"loss": 0.6327,
"step": 954
},
{
"epoch": 0.6244891286578388,
"grad_norm": 4.03125,
"learning_rate": 1.9962387894835277e-05,
"loss": 0.828,
"step": 955
},
{
"epoch": 0.6251430439758051,
"grad_norm": 4.59375,
"learning_rate": 1.996199118770813e-05,
"loss": 0.6541,
"step": 956
},
{
"epoch": 0.6257969592937714,
"grad_norm": 6.3125,
"learning_rate": 1.996159240342547e-05,
"loss": 0.7501,
"step": 957
},
{
"epoch": 0.6264508746117378,
"grad_norm": 4.0625,
"learning_rate": 1.996119154207043e-05,
"loss": 0.4816,
"step": 958
},
{
"epoch": 0.6271047899297041,
"grad_norm": 4.5,
"learning_rate": 1.9960788603726603e-05,
"loss": 0.8893,
"step": 959
},
{
"epoch": 0.6277587052476704,
"grad_norm": 3.4375,
"learning_rate": 1.9960383588478e-05,
"loss": 0.5005,
"step": 960
},
{
"epoch": 0.6284126205656367,
"grad_norm": 4.0625,
"learning_rate": 1.9959976496409073e-05,
"loss": 0.493,
"step": 961
},
{
"epoch": 0.6290665358836031,
"grad_norm": 4.90625,
"learning_rate": 1.9959567327604704e-05,
"loss": 0.6515,
"step": 962
},
{
"epoch": 0.6297204512015694,
"grad_norm": 3.125,
"learning_rate": 1.9959156082150205e-05,
"loss": 0.4173,
"step": 963
},
{
"epoch": 0.6303743665195357,
"grad_norm": 3.46875,
"learning_rate": 1.9958742760131325e-05,
"loss": 0.4823,
"step": 964
},
{
"epoch": 0.631028281837502,
"grad_norm": 3.484375,
"learning_rate": 1.9958327361634248e-05,
"loss": 0.4804,
"step": 965
},
{
"epoch": 0.6316821971554684,
"grad_norm": 3.84375,
"learning_rate": 1.9957909886745588e-05,
"loss": 0.4664,
"step": 966
},
{
"epoch": 0.6323361124734347,
"grad_norm": 4.34375,
"learning_rate": 1.9957490335552386e-05,
"loss": 0.5754,
"step": 967
},
{
"epoch": 0.632990027791401,
"grad_norm": 4.78125,
"learning_rate": 1.9957068708142127e-05,
"loss": 1.0389,
"step": 968
},
{
"epoch": 0.6336439431093673,
"grad_norm": 5.75,
"learning_rate": 1.995664500460272e-05,
"loss": 0.8707,
"step": 969
},
{
"epoch": 0.6342978584273337,
"grad_norm": 2.5,
"learning_rate": 1.9956219225022514e-05,
"loss": 0.3593,
"step": 970
},
{
"epoch": 0.6349517737453,
"grad_norm": 8.0625,
"learning_rate": 1.9955791369490288e-05,
"loss": 1.0368,
"step": 971
},
{
"epoch": 0.6356056890632663,
"grad_norm": 5.375,
"learning_rate": 1.9955361438095254e-05,
"loss": 0.8584,
"step": 972
},
{
"epoch": 0.6362596043812326,
"grad_norm": 4.84375,
"learning_rate": 1.995492943092705e-05,
"loss": 0.6599,
"step": 973
},
{
"epoch": 0.636913519699199,
"grad_norm": 5.15625,
"learning_rate": 1.995449534807576e-05,
"loss": 0.8362,
"step": 974
},
{
"epoch": 0.6375674350171653,
"grad_norm": 5.84375,
"learning_rate": 1.9954059189631883e-05,
"loss": 0.8764,
"step": 975
},
{
"epoch": 0.6382213503351316,
"grad_norm": 2.734375,
"learning_rate": 1.9953620955686376e-05,
"loss": 0.4303,
"step": 976
},
{
"epoch": 0.6388752656530979,
"grad_norm": 4.40625,
"learning_rate": 1.9953180646330604e-05,
"loss": 0.6384,
"step": 977
},
{
"epoch": 0.6395291809710643,
"grad_norm": 3.53125,
"learning_rate": 1.9952738261656382e-05,
"loss": 0.5877,
"step": 978
},
{
"epoch": 0.6401830962890306,
"grad_norm": 5.90625,
"learning_rate": 1.9952293801755946e-05,
"loss": 0.938,
"step": 979
},
{
"epoch": 0.6408370116069969,
"grad_norm": 4.21875,
"learning_rate": 1.995184726672197e-05,
"loss": 0.6501,
"step": 980
},
{
"epoch": 0.6414909269249632,
"grad_norm": 4.0,
"learning_rate": 1.995139865664756e-05,
"loss": 0.5492,
"step": 981
},
{
"epoch": 0.6421448422429296,
"grad_norm": 2.75,
"learning_rate": 1.9950947971626257e-05,
"loss": 0.3914,
"step": 982
},
{
"epoch": 0.6427987575608959,
"grad_norm": 6.0,
"learning_rate": 1.995049521175203e-05,
"loss": 0.8864,
"step": 983
},
{
"epoch": 0.6434526728788622,
"grad_norm": 5.3125,
"learning_rate": 1.995004037711929e-05,
"loss": 0.6174,
"step": 984
},
{
"epoch": 0.6441065881968285,
"grad_norm": 4.34375,
"learning_rate": 1.9949583467822863e-05,
"loss": 0.3753,
"step": 985
},
{
"epoch": 0.6447605035147949,
"grad_norm": 3.78125,
"learning_rate": 1.994912448395802e-05,
"loss": 0.4583,
"step": 986
},
{
"epoch": 0.6454144188327612,
"grad_norm": 5.4375,
"learning_rate": 1.9948663425620473e-05,
"loss": 0.6973,
"step": 987
},
{
"epoch": 0.6460683341507275,
"grad_norm": 4.375,
"learning_rate": 1.9948200292906344e-05,
"loss": 0.8518,
"step": 988
},
{
"epoch": 0.6467222494686938,
"grad_norm": 6.625,
"learning_rate": 1.994773508591221e-05,
"loss": 1.0712,
"step": 989
},
{
"epoch": 0.6473761647866602,
"grad_norm": 4.375,
"learning_rate": 1.9947267804735064e-05,
"loss": 0.5485,
"step": 990
},
{
"epoch": 0.6480300801046265,
"grad_norm": 6.5625,
"learning_rate": 1.9946798449472336e-05,
"loss": 1.2496,
"step": 991
},
{
"epoch": 0.6486839954225928,
"grad_norm": 5.03125,
"learning_rate": 1.9946327020221898e-05,
"loss": 0.601,
"step": 992
},
{
"epoch": 0.6493379107405591,
"grad_norm": 4.09375,
"learning_rate": 1.9945853517082043e-05,
"loss": 0.6031,
"step": 993
},
{
"epoch": 0.6499918260585255,
"grad_norm": 5.28125,
"learning_rate": 1.9945377940151498e-05,
"loss": 0.8796,
"step": 994
},
{
"epoch": 0.6506457413764918,
"grad_norm": 6.8125,
"learning_rate": 1.9944900289529427e-05,
"loss": 0.9228,
"step": 995
},
{
"epoch": 0.6512996566944581,
"grad_norm": 3.984375,
"learning_rate": 1.9944420565315427e-05,
"loss": 0.8225,
"step": 996
},
{
"epoch": 0.6519535720124244,
"grad_norm": 4.5,
"learning_rate": 1.9943938767609515e-05,
"loss": 0.6592,
"step": 997
},
{
"epoch": 0.6526074873303908,
"grad_norm": 5.0,
"learning_rate": 1.994345489651216e-05,
"loss": 0.904,
"step": 998
},
{
"epoch": 0.6532614026483571,
"grad_norm": 3.03125,
"learning_rate": 1.9942968952124248e-05,
"loss": 0.3448,
"step": 999
},
{
"epoch": 0.6539153179663234,
"grad_norm": 3.515625,
"learning_rate": 1.9942480934547103e-05,
"loss": 0.4805,
"step": 1000
},
{
"epoch": 0.6545692332842897,
"grad_norm": 4.8125,
"learning_rate": 1.9941990843882482e-05,
"loss": 0.685,
"step": 1001
},
{
"epoch": 0.6552231486022561,
"grad_norm": 2.953125,
"learning_rate": 1.9941498680232573e-05,
"loss": 0.4924,
"step": 1002
},
{
"epoch": 0.6558770639202224,
"grad_norm": 5.4375,
"learning_rate": 1.994100444369999e-05,
"loss": 0.9571,
"step": 1003
},
{
"epoch": 0.6565309792381887,
"grad_norm": 4.875,
"learning_rate": 1.9940508134387793e-05,
"loss": 0.6047,
"step": 1004
},
{
"epoch": 0.657184894556155,
"grad_norm": 5.0,
"learning_rate": 1.9940009752399462e-05,
"loss": 0.5699,
"step": 1005
},
{
"epoch": 0.6578388098741214,
"grad_norm": 4.9375,
"learning_rate": 1.9939509297838915e-05,
"loss": 1.1176,
"step": 1006
},
{
"epoch": 0.6584927251920877,
"grad_norm": 6.34375,
"learning_rate": 1.9939006770810502e-05,
"loss": 0.9447,
"step": 1007
},
{
"epoch": 0.6591466405100539,
"grad_norm": 4.34375,
"learning_rate": 1.9938502171419002e-05,
"loss": 0.6341,
"step": 1008
},
{
"epoch": 0.6598005558280202,
"grad_norm": 4.21875,
"learning_rate": 1.993799549976963e-05,
"loss": 0.6712,
"step": 1009
},
{
"epoch": 0.6604544711459865,
"grad_norm": 5.15625,
"learning_rate": 1.9937486755968027e-05,
"loss": 0.9834,
"step": 1010
},
{
"epoch": 0.6611083864639529,
"grad_norm": 3.296875,
"learning_rate": 1.9936975940120276e-05,
"loss": 0.387,
"step": 1011
},
{
"epoch": 0.6617623017819192,
"grad_norm": 4.3125,
"learning_rate": 1.9936463052332887e-05,
"loss": 0.5781,
"step": 1012
},
{
"epoch": 0.6624162170998855,
"grad_norm": 3.578125,
"learning_rate": 1.9935948092712792e-05,
"loss": 0.5726,
"step": 1013
},
{
"epoch": 0.6630701324178518,
"grad_norm": 4.71875,
"learning_rate": 1.993543106136737e-05,
"loss": 0.8957,
"step": 1014
},
{
"epoch": 0.6637240477358182,
"grad_norm": 3.90625,
"learning_rate": 1.993491195840443e-05,
"loss": 0.6109,
"step": 1015
},
{
"epoch": 0.6643779630537845,
"grad_norm": 5.90625,
"learning_rate": 1.9934390783932204e-05,
"loss": 0.8616,
"step": 1016
},
{
"epoch": 0.6650318783717508,
"grad_norm": 4.59375,
"learning_rate": 1.9933867538059362e-05,
"loss": 0.9335,
"step": 1017
},
{
"epoch": 0.6656857936897171,
"grad_norm": 5.15625,
"learning_rate": 1.9933342220895e-05,
"loss": 0.8178,
"step": 1018
},
{
"epoch": 0.6663397090076835,
"grad_norm": 5.65625,
"learning_rate": 1.9932814832548665e-05,
"loss": 1.248,
"step": 1019
},
{
"epoch": 0.6669936243256498,
"grad_norm": 4.90625,
"learning_rate": 1.993228537313031e-05,
"loss": 0.6167,
"step": 1020
},
{
"epoch": 0.6676475396436161,
"grad_norm": 5.0625,
"learning_rate": 1.993175384275033e-05,
"loss": 0.6346,
"step": 1021
},
{
"epoch": 0.6683014549615824,
"grad_norm": 3.890625,
"learning_rate": 1.9931220241519564e-05,
"loss": 0.6665,
"step": 1022
},
{
"epoch": 0.6689553702795488,
"grad_norm": 3.46875,
"learning_rate": 1.9930684569549265e-05,
"loss": 0.4437,
"step": 1023
},
{
"epoch": 0.6696092855975151,
"grad_norm": 4.65625,
"learning_rate": 1.9930146826951124e-05,
"loss": 0.9239,
"step": 1024
},
{
"epoch": 0.6702632009154814,
"grad_norm": 3.0625,
"learning_rate": 1.9929607013837268e-05,
"loss": 0.3079,
"step": 1025
},
{
"epoch": 0.6709171162334477,
"grad_norm": 4.1875,
"learning_rate": 1.992906513032025e-05,
"loss": 0.689,
"step": 1026
},
{
"epoch": 0.6715710315514141,
"grad_norm": 3.90625,
"learning_rate": 1.992852117651306e-05,
"loss": 0.5696,
"step": 1027
},
{
"epoch": 0.6722249468693804,
"grad_norm": 5.34375,
"learning_rate": 1.992797515252911e-05,
"loss": 0.4639,
"step": 1028
},
{
"epoch": 0.6728788621873467,
"grad_norm": 3.0625,
"learning_rate": 1.992742705848226e-05,
"loss": 0.4805,
"step": 1029
},
{
"epoch": 0.673532777505313,
"grad_norm": 5.0,
"learning_rate": 1.9926876894486787e-05,
"loss": 0.5762,
"step": 1030
},
{
"epoch": 0.6741866928232794,
"grad_norm": 3.578125,
"learning_rate": 1.9926324660657403e-05,
"loss": 0.4997,
"step": 1031
},
{
"epoch": 0.6748406081412457,
"grad_norm": 4.09375,
"learning_rate": 1.9925770357109256e-05,
"loss": 0.5333,
"step": 1032
},
{
"epoch": 0.675494523459212,
"grad_norm": 3.96875,
"learning_rate": 1.9925213983957926e-05,
"loss": 0.5319,
"step": 1033
},
{
"epoch": 0.6761484387771783,
"grad_norm": 5.34375,
"learning_rate": 1.992465554131941e-05,
"loss": 0.6658,
"step": 1034
},
{
"epoch": 0.6768023540951447,
"grad_norm": 3.703125,
"learning_rate": 1.992409502931016e-05,
"loss": 0.6798,
"step": 1035
},
{
"epoch": 0.677456269413111,
"grad_norm": 4.5,
"learning_rate": 1.992353244804704e-05,
"loss": 0.6656,
"step": 1036
},
{
"epoch": 0.6781101847310773,
"grad_norm": 3.890625,
"learning_rate": 1.9922967797647357e-05,
"loss": 0.5322,
"step": 1037
},
{
"epoch": 0.6787641000490436,
"grad_norm": 4.375,
"learning_rate": 1.992240107822884e-05,
"loss": 0.5492,
"step": 1038
},
{
"epoch": 0.67941801536701,
"grad_norm": 4.03125,
"learning_rate": 1.992183228990966e-05,
"loss": 0.5365,
"step": 1039
},
{
"epoch": 0.6800719306849763,
"grad_norm": 3.625,
"learning_rate": 1.992126143280841e-05,
"loss": 0.5934,
"step": 1040
},
{
"epoch": 0.6807258460029426,
"grad_norm": 4.75,
"learning_rate": 1.992068850704412e-05,
"loss": 0.7762,
"step": 1041
},
{
"epoch": 0.6813797613209089,
"grad_norm": 5.5,
"learning_rate": 1.992011351273625e-05,
"loss": 0.8475,
"step": 1042
},
{
"epoch": 0.6820336766388753,
"grad_norm": 4.3125,
"learning_rate": 1.9919536450004695e-05,
"loss": 0.5899,
"step": 1043
},
{
"epoch": 0.6826875919568416,
"grad_norm": 3.515625,
"learning_rate": 1.9918957318969768e-05,
"loss": 0.3123,
"step": 1044
},
{
"epoch": 0.6833415072748079,
"grad_norm": 2.765625,
"learning_rate": 1.991837611975223e-05,
"loss": 0.2633,
"step": 1045
},
{
"epoch": 0.6839954225927742,
"grad_norm": 5.96875,
"learning_rate": 1.9917792852473263e-05,
"loss": 0.7355,
"step": 1046
},
{
"epoch": 0.6846493379107406,
"grad_norm": 3.75,
"learning_rate": 1.991720751725448e-05,
"loss": 0.8369,
"step": 1047
},
{
"epoch": 0.6853032532287069,
"grad_norm": 4.3125,
"learning_rate": 1.9916620114217934e-05,
"loss": 0.8012,
"step": 1048
},
{
"epoch": 0.6859571685466732,
"grad_norm": 3.25,
"learning_rate": 1.9916030643486098e-05,
"loss": 0.4809,
"step": 1049
},
{
"epoch": 0.6866110838646395,
"grad_norm": 4.65625,
"learning_rate": 1.9915439105181887e-05,
"loss": 0.6411,
"step": 1050
},
{
"epoch": 0.6872649991826059,
"grad_norm": 7.40625,
"learning_rate": 1.9914845499428635e-05,
"loss": 0.8177,
"step": 1051
},
{
"epoch": 0.6879189145005722,
"grad_norm": 4.6875,
"learning_rate": 1.9914249826350118e-05,
"loss": 0.8298,
"step": 1052
},
{
"epoch": 0.6885728298185385,
"grad_norm": 3.46875,
"learning_rate": 1.9913652086070535e-05,
"loss": 0.4278,
"step": 1053
},
{
"epoch": 0.6892267451365048,
"grad_norm": 4.4375,
"learning_rate": 1.9913052278714527e-05,
"loss": 0.718,
"step": 1054
},
{
"epoch": 0.6898806604544712,
"grad_norm": 4.125,
"learning_rate": 1.9912450404407152e-05,
"loss": 0.6031,
"step": 1055
},
{
"epoch": 0.6905345757724375,
"grad_norm": 5.15625,
"learning_rate": 1.9911846463273907e-05,
"loss": 0.5463,
"step": 1056
},
{
"epoch": 0.6911884910904038,
"grad_norm": 3.046875,
"learning_rate": 1.9911240455440718e-05,
"loss": 0.4352,
"step": 1057
},
{
"epoch": 0.6918424064083701,
"grad_norm": 5.1875,
"learning_rate": 1.9910632381033946e-05,
"loss": 0.5608,
"step": 1058
},
{
"epoch": 0.6924963217263365,
"grad_norm": 4.125,
"learning_rate": 1.9910022240180374e-05,
"loss": 0.5663,
"step": 1059
},
{
"epoch": 0.6931502370443028,
"grad_norm": 4.96875,
"learning_rate": 1.9909410033007228e-05,
"loss": 0.8075,
"step": 1060
},
{
"epoch": 0.6938041523622691,
"grad_norm": 6.125,
"learning_rate": 1.9908795759642154e-05,
"loss": 1.3843,
"step": 1061
},
{
"epoch": 0.6944580676802354,
"grad_norm": 4.125,
"learning_rate": 1.990817942021323e-05,
"loss": 0.5968,
"step": 1062
},
{
"epoch": 0.6951119829982018,
"grad_norm": 3.90625,
"learning_rate": 1.9907561014848974e-05,
"loss": 0.4465,
"step": 1063
},
{
"epoch": 0.6957658983161681,
"grad_norm": 3.859375,
"learning_rate": 1.9906940543678326e-05,
"loss": 0.87,
"step": 1064
},
{
"epoch": 0.6964198136341344,
"grad_norm": 4.59375,
"learning_rate": 1.990631800683066e-05,
"loss": 0.9706,
"step": 1065
},
{
"epoch": 0.6970737289521007,
"grad_norm": 3.703125,
"learning_rate": 1.990569340443577e-05,
"loss": 0.5227,
"step": 1066
},
{
"epoch": 0.6977276442700671,
"grad_norm": 6.0625,
"learning_rate": 1.990506673662391e-05,
"loss": 1.177,
"step": 1067
},
{
"epoch": 0.6983815595880334,
"grad_norm": 3.890625,
"learning_rate": 1.9904438003525733e-05,
"loss": 0.5243,
"step": 1068
},
{
"epoch": 0.6990354749059997,
"grad_norm": 3.375,
"learning_rate": 1.9903807205272333e-05,
"loss": 0.446,
"step": 1069
},
{
"epoch": 0.699689390223966,
"grad_norm": 4.5,
"learning_rate": 1.9903174341995243e-05,
"loss": 0.7413,
"step": 1070
},
{
"epoch": 0.7003433055419324,
"grad_norm": 5.34375,
"learning_rate": 1.9902539413826415e-05,
"loss": 0.933,
"step": 1071
},
{
"epoch": 0.7009972208598987,
"grad_norm": 3.71875,
"learning_rate": 1.990190242089824e-05,
"loss": 0.474,
"step": 1072
},
{
"epoch": 0.701651136177865,
"grad_norm": 4.21875,
"learning_rate": 1.990126336334354e-05,
"loss": 0.5378,
"step": 1073
},
{
"epoch": 0.7023050514958313,
"grad_norm": 4.625,
"learning_rate": 1.990062224129555e-05,
"loss": 0.7578,
"step": 1074
},
{
"epoch": 0.7029589668137977,
"grad_norm": 3.421875,
"learning_rate": 1.9899979054887964e-05,
"loss": 0.4882,
"step": 1075
},
{
"epoch": 0.703612882131764,
"grad_norm": 4.15625,
"learning_rate": 1.9899333804254885e-05,
"loss": 0.7977,
"step": 1076
},
{
"epoch": 0.7042667974497303,
"grad_norm": 3.140625,
"learning_rate": 1.9898686489530852e-05,
"loss": 0.4417,
"step": 1077
},
{
"epoch": 0.7049207127676966,
"grad_norm": 4.34375,
"learning_rate": 1.9898037110850835e-05,
"loss": 0.4985,
"step": 1078
},
{
"epoch": 0.705574628085663,
"grad_norm": 4.75,
"learning_rate": 1.989738566835024e-05,
"loss": 0.6424,
"step": 1079
},
{
"epoch": 0.7062285434036293,
"grad_norm": 4.1875,
"learning_rate": 1.989673216216489e-05,
"loss": 0.5638,
"step": 1080
},
{
"epoch": 0.7068824587215956,
"grad_norm": 3.734375,
"learning_rate": 1.9896076592431054e-05,
"loss": 0.417,
"step": 1081
},
{
"epoch": 0.7075363740395618,
"grad_norm": 4.3125,
"learning_rate": 1.989541895928542e-05,
"loss": 0.7654,
"step": 1082
},
{
"epoch": 0.7081902893575281,
"grad_norm": 5.71875,
"learning_rate": 1.9894759262865114e-05,
"loss": 0.7763,
"step": 1083
},
{
"epoch": 0.7088442046754945,
"grad_norm": 4.84375,
"learning_rate": 1.989409750330768e-05,
"loss": 0.7535,
"step": 1084
},
{
"epoch": 0.7094981199934608,
"grad_norm": 6.96875,
"learning_rate": 1.9893433680751105e-05,
"loss": 0.7835,
"step": 1085
},
{
"epoch": 0.7101520353114271,
"grad_norm": 5.09375,
"learning_rate": 1.98927677953338e-05,
"loss": 0.6426,
"step": 1086
},
{
"epoch": 0.7108059506293934,
"grad_norm": 5.9375,
"learning_rate": 1.9892099847194608e-05,
"loss": 0.9646,
"step": 1087
},
{
"epoch": 0.7114598659473598,
"grad_norm": 3.90625,
"learning_rate": 1.9891429836472805e-05,
"loss": 0.4663,
"step": 1088
},
{
"epoch": 0.7121137812653261,
"grad_norm": 3.5,
"learning_rate": 1.989075776330809e-05,
"loss": 0.3986,
"step": 1089
},
{
"epoch": 0.7127676965832924,
"grad_norm": 3.71875,
"learning_rate": 1.989008362784059e-05,
"loss": 0.4106,
"step": 1090
},
{
"epoch": 0.7134216119012587,
"grad_norm": 4.5,
"learning_rate": 1.988940743021088e-05,
"loss": 0.6797,
"step": 1091
},
{
"epoch": 0.7140755272192251,
"grad_norm": 4.96875,
"learning_rate": 1.9888729170559948e-05,
"loss": 0.8449,
"step": 1092
},
{
"epoch": 0.7147294425371914,
"grad_norm": 5.46875,
"learning_rate": 1.988804884902921e-05,
"loss": 0.6937,
"step": 1093
},
{
"epoch": 0.7153833578551577,
"grad_norm": 5.59375,
"learning_rate": 1.9887366465760523e-05,
"loss": 0.9909,
"step": 1094
},
{
"epoch": 0.716037273173124,
"grad_norm": 5.46875,
"learning_rate": 1.9886682020896172e-05,
"loss": 0.8319,
"step": 1095
},
{
"epoch": 0.7166911884910904,
"grad_norm": 5.75,
"learning_rate": 1.988599551457887e-05,
"loss": 0.8728,
"step": 1096
},
{
"epoch": 0.7173451038090567,
"grad_norm": 5.09375,
"learning_rate": 1.9885306946951753e-05,
"loss": 0.5158,
"step": 1097
},
{
"epoch": 0.717999019127023,
"grad_norm": 5.15625,
"learning_rate": 1.98846163181584e-05,
"loss": 0.5721,
"step": 1098
},
{
"epoch": 0.7186529344449893,
"grad_norm": 5.46875,
"learning_rate": 1.9883923628342807e-05,
"loss": 1.0006,
"step": 1099
},
{
"epoch": 0.7193068497629557,
"grad_norm": 6.25,
"learning_rate": 1.9883228877649407e-05,
"loss": 0.7545,
"step": 1100
},
{
"epoch": 0.719960765080922,
"grad_norm": 3.90625,
"learning_rate": 1.988253206622306e-05,
"loss": 0.559,
"step": 1101
},
{
"epoch": 0.7206146803988883,
"grad_norm": 5.34375,
"learning_rate": 1.9881833194209063e-05,
"loss": 0.634,
"step": 1102
},
{
"epoch": 0.7212685957168546,
"grad_norm": 4.59375,
"learning_rate": 1.9881132261753126e-05,
"loss": 0.8305,
"step": 1103
},
{
"epoch": 0.721922511034821,
"grad_norm": 5.53125,
"learning_rate": 1.9880429269001407e-05,
"loss": 0.6211,
"step": 1104
},
{
"epoch": 0.7225764263527873,
"grad_norm": 3.328125,
"learning_rate": 1.9879724216100488e-05,
"loss": 0.3974,
"step": 1105
},
{
"epoch": 0.7232303416707536,
"grad_norm": 7.25,
"learning_rate": 1.987901710319737e-05,
"loss": 1.0916,
"step": 1106
},
{
"epoch": 0.7238842569887199,
"grad_norm": 6.625,
"learning_rate": 1.9878307930439497e-05,
"loss": 1.0786,
"step": 1107
},
{
"epoch": 0.7245381723066863,
"grad_norm": 3.859375,
"learning_rate": 1.9877596697974732e-05,
"loss": 0.4618,
"step": 1108
},
{
"epoch": 0.7251920876246526,
"grad_norm": 6.21875,
"learning_rate": 1.9876883405951378e-05,
"loss": 1.01,
"step": 1109
},
{
"epoch": 0.7258460029426189,
"grad_norm": 4.125,
"learning_rate": 1.987616805451816e-05,
"loss": 0.5793,
"step": 1110
},
{
"epoch": 0.7264999182605852,
"grad_norm": 5.21875,
"learning_rate": 1.987545064382424e-05,
"loss": 0.6738,
"step": 1111
},
{
"epoch": 0.7271538335785516,
"grad_norm": 4.125,
"learning_rate": 1.9874731174019195e-05,
"loss": 0.5042,
"step": 1112
},
{
"epoch": 0.7278077488965179,
"grad_norm": 4.71875,
"learning_rate": 1.9874009645253044e-05,
"loss": 0.7292,
"step": 1113
},
{
"epoch": 0.7284616642144842,
"grad_norm": 3.28125,
"learning_rate": 1.9873286057676232e-05,
"loss": 0.2973,
"step": 1114
},
{
"epoch": 0.7291155795324505,
"grad_norm": 3.875,
"learning_rate": 1.9872560411439634e-05,
"loss": 0.5845,
"step": 1115
},
{
"epoch": 0.7297694948504169,
"grad_norm": 4.3125,
"learning_rate": 1.9871832706694553e-05,
"loss": 0.5467,
"step": 1116
},
{
"epoch": 0.7304234101683832,
"grad_norm": 4.40625,
"learning_rate": 1.9871102943592717e-05,
"loss": 0.6854,
"step": 1117
},
{
"epoch": 0.7310773254863495,
"grad_norm": 4.5625,
"learning_rate": 1.987037112228629e-05,
"loss": 0.4087,
"step": 1118
},
{
"epoch": 0.7317312408043158,
"grad_norm": 7.3125,
"learning_rate": 1.986963724292787e-05,
"loss": 1.1244,
"step": 1119
},
{
"epoch": 0.7323851561222822,
"grad_norm": 5.1875,
"learning_rate": 1.9868901305670465e-05,
"loss": 0.7025,
"step": 1120
},
{
"epoch": 0.7330390714402485,
"grad_norm": 5.0,
"learning_rate": 1.9868163310667527e-05,
"loss": 0.637,
"step": 1121
},
{
"epoch": 0.7336929867582148,
"grad_norm": 4.5,
"learning_rate": 1.9867423258072936e-05,
"loss": 0.6376,
"step": 1122
},
{
"epoch": 0.7343469020761811,
"grad_norm": 5.5,
"learning_rate": 1.9866681148041005e-05,
"loss": 0.7703,
"step": 1123
},
{
"epoch": 0.7350008173941475,
"grad_norm": 4.5,
"learning_rate": 1.9865936980726457e-05,
"loss": 0.6379,
"step": 1124
},
{
"epoch": 0.7356547327121138,
"grad_norm": 4.21875,
"learning_rate": 1.9865190756284467e-05,
"loss": 0.6159,
"step": 1125
},
{
"epoch": 0.7363086480300801,
"grad_norm": 4.0,
"learning_rate": 1.9864442474870625e-05,
"loss": 0.4685,
"step": 1126
},
{
"epoch": 0.7369625633480464,
"grad_norm": 4.84375,
"learning_rate": 1.9863692136640952e-05,
"loss": 0.9719,
"step": 1127
},
{
"epoch": 0.7376164786660128,
"grad_norm": 6.5,
"learning_rate": 1.9862939741751904e-05,
"loss": 0.8208,
"step": 1128
},
{
"epoch": 0.7382703939839791,
"grad_norm": 4.6875,
"learning_rate": 1.986218529036036e-05,
"loss": 0.665,
"step": 1129
},
{
"epoch": 0.7389243093019454,
"grad_norm": 3.3125,
"learning_rate": 1.9861428782623626e-05,
"loss": 0.5091,
"step": 1130
},
{
"epoch": 0.7395782246199117,
"grad_norm": 4.0625,
"learning_rate": 1.986067021869944e-05,
"loss": 0.4906,
"step": 1131
},
{
"epoch": 0.7402321399378781,
"grad_norm": 3.796875,
"learning_rate": 1.9859909598745975e-05,
"loss": 0.3969,
"step": 1132
},
{
"epoch": 0.7408860552558444,
"grad_norm": 4.53125,
"learning_rate": 1.985914692292182e-05,
"loss": 0.6145,
"step": 1133
},
{
"epoch": 0.7415399705738107,
"grad_norm": 3.75,
"learning_rate": 1.9858382191386004e-05,
"loss": 0.428,
"step": 1134
},
{
"epoch": 0.742193885891777,
"grad_norm": 3.5,
"learning_rate": 1.9857615404297972e-05,
"loss": 0.385,
"step": 1135
},
{
"epoch": 0.7428478012097434,
"grad_norm": 5.78125,
"learning_rate": 1.9856846561817616e-05,
"loss": 1.0006,
"step": 1136
},
{
"epoch": 0.7435017165277097,
"grad_norm": 4.09375,
"learning_rate": 1.9856075664105238e-05,
"loss": 0.3184,
"step": 1137
},
{
"epoch": 0.744155631845676,
"grad_norm": 5.53125,
"learning_rate": 1.985530271132158e-05,
"loss": 1.005,
"step": 1138
},
{
"epoch": 0.7448095471636423,
"grad_norm": 3.78125,
"learning_rate": 1.9854527703627804e-05,
"loss": 0.3148,
"step": 1139
},
{
"epoch": 0.7454634624816087,
"grad_norm": 4.59375,
"learning_rate": 1.9853750641185512e-05,
"loss": 0.5231,
"step": 1140
},
{
"epoch": 0.746117377799575,
"grad_norm": 3.78125,
"learning_rate": 1.9852971524156722e-05,
"loss": 0.5369,
"step": 1141
},
{
"epoch": 0.7467712931175413,
"grad_norm": 4.59375,
"learning_rate": 1.985219035270389e-05,
"loss": 0.5627,
"step": 1142
},
{
"epoch": 0.7474252084355076,
"grad_norm": 5.0,
"learning_rate": 1.9851407126989897e-05,
"loss": 0.5977,
"step": 1143
},
{
"epoch": 0.748079123753474,
"grad_norm": 4.25,
"learning_rate": 1.9850621847178047e-05,
"loss": 0.5823,
"step": 1144
},
{
"epoch": 0.7487330390714403,
"grad_norm": 6.03125,
"learning_rate": 1.9849834513432084e-05,
"loss": 1.0281,
"step": 1145
},
{
"epoch": 0.7493869543894066,
"grad_norm": 3.59375,
"learning_rate": 1.984904512591617e-05,
"loss": 0.5137,
"step": 1146
},
{
"epoch": 0.7500408697073729,
"grad_norm": 3.609375,
"learning_rate": 1.9848253684794895e-05,
"loss": 0.4052,
"step": 1147
},
{
"epoch": 0.7506947850253393,
"grad_norm": 3.828125,
"learning_rate": 1.9847460190233286e-05,
"loss": 0.4234,
"step": 1148
},
{
"epoch": 0.7513487003433056,
"grad_norm": 4.09375,
"learning_rate": 1.9846664642396793e-05,
"loss": 0.5227,
"step": 1149
},
{
"epoch": 0.7520026156612719,
"grad_norm": 3.34375,
"learning_rate": 1.984586704145129e-05,
"loss": 0.3631,
"step": 1150
},
{
"epoch": 0.7526565309792382,
"grad_norm": 5.25,
"learning_rate": 1.984506738756309e-05,
"loss": 1.2075,
"step": 1151
},
{
"epoch": 0.7533104462972046,
"grad_norm": 3.796875,
"learning_rate": 1.9844265680898917e-05,
"loss": 0.4715,
"step": 1152
},
{
"epoch": 0.7539643616151709,
"grad_norm": 4.28125,
"learning_rate": 1.9843461921625942e-05,
"loss": 0.5403,
"step": 1153
},
{
"epoch": 0.7546182769331372,
"grad_norm": 4.875,
"learning_rate": 1.9842656109911753e-05,
"loss": 0.7448,
"step": 1154
},
{
"epoch": 0.7552721922511035,
"grad_norm": 3.96875,
"learning_rate": 1.984184824592437e-05,
"loss": 0.6446,
"step": 1155
},
{
"epoch": 0.7559261075690699,
"grad_norm": 4.53125,
"learning_rate": 1.9841038329832237e-05,
"loss": 0.6427,
"step": 1156
},
{
"epoch": 0.7565800228870361,
"grad_norm": 6.09375,
"learning_rate": 1.9840226361804223e-05,
"loss": 1.4863,
"step": 1157
},
{
"epoch": 0.7572339382050024,
"grad_norm": 3.703125,
"learning_rate": 1.983941234200964e-05,
"loss": 0.5256,
"step": 1158
},
{
"epoch": 0.7578878535229687,
"grad_norm": 4.65625,
"learning_rate": 1.983859627061821e-05,
"loss": 0.6581,
"step": 1159
},
{
"epoch": 0.758541768840935,
"grad_norm": 4.78125,
"learning_rate": 1.9837778147800092e-05,
"loss": 0.6791,
"step": 1160
},
{
"epoch": 0.7591956841589014,
"grad_norm": 5.84375,
"learning_rate": 1.9836957973725877e-05,
"loss": 0.9089,
"step": 1161
},
{
"epoch": 0.7598495994768677,
"grad_norm": 4.0,
"learning_rate": 1.983613574856657e-05,
"loss": 0.7796,
"step": 1162
},
{
"epoch": 0.760503514794834,
"grad_norm": 3.3125,
"learning_rate": 1.9835311472493616e-05,
"loss": 0.4999,
"step": 1163
},
{
"epoch": 0.7611574301128003,
"grad_norm": 4.375,
"learning_rate": 1.9834485145678883e-05,
"loss": 0.7209,
"step": 1164
},
{
"epoch": 0.7618113454307667,
"grad_norm": 3.359375,
"learning_rate": 1.983365676829466e-05,
"loss": 0.4103,
"step": 1165
},
{
"epoch": 0.762465260748733,
"grad_norm": 3.984375,
"learning_rate": 1.9832826340513684e-05,
"loss": 0.6533,
"step": 1166
},
{
"epoch": 0.7631191760666993,
"grad_norm": 3.546875,
"learning_rate": 1.9831993862509096e-05,
"loss": 0.4123,
"step": 1167
},
{
"epoch": 0.7637730913846656,
"grad_norm": 4.125,
"learning_rate": 1.9831159334454475e-05,
"loss": 0.5468,
"step": 1168
},
{
"epoch": 0.764427006702632,
"grad_norm": 5.375,
"learning_rate": 1.9830322756523827e-05,
"loss": 0.5445,
"step": 1169
},
{
"epoch": 0.7650809220205983,
"grad_norm": 2.609375,
"learning_rate": 1.982948412889159e-05,
"loss": 0.1975,
"step": 1170
},
{
"epoch": 0.7657348373385646,
"grad_norm": 3.59375,
"learning_rate": 1.9828643451732618e-05,
"loss": 0.5713,
"step": 1171
},
{
"epoch": 0.766388752656531,
"grad_norm": 5.0625,
"learning_rate": 1.9827800725222203e-05,
"loss": 0.7995,
"step": 1172
},
{
"epoch": 0.7670426679744973,
"grad_norm": 4.40625,
"learning_rate": 1.9826955949536062e-05,
"loss": 0.6437,
"step": 1173
},
{
"epoch": 0.7676965832924636,
"grad_norm": 4.375,
"learning_rate": 1.9826109124850333e-05,
"loss": 0.6973,
"step": 1174
},
{
"epoch": 0.7683504986104299,
"grad_norm": 3.984375,
"learning_rate": 1.9825260251341586e-05,
"loss": 0.4613,
"step": 1175
},
{
"epoch": 0.7690044139283962,
"grad_norm": 3.5625,
"learning_rate": 1.9824409329186823e-05,
"loss": 0.4911,
"step": 1176
},
{
"epoch": 0.7696583292463626,
"grad_norm": 4.28125,
"learning_rate": 1.9823556358563464e-05,
"loss": 0.4457,
"step": 1177
},
{
"epoch": 0.7703122445643289,
"grad_norm": 4.40625,
"learning_rate": 1.9822701339649358e-05,
"loss": 0.4608,
"step": 1178
},
{
"epoch": 0.7709661598822952,
"grad_norm": 3.25,
"learning_rate": 1.9821844272622793e-05,
"loss": 0.3109,
"step": 1179
},
{
"epoch": 0.7716200752002615,
"grad_norm": 4.6875,
"learning_rate": 1.9820985157662462e-05,
"loss": 0.6808,
"step": 1180
},
{
"epoch": 0.7722739905182279,
"grad_norm": 3.5625,
"learning_rate": 1.9820123994947505e-05,
"loss": 0.538,
"step": 1181
},
{
"epoch": 0.7729279058361942,
"grad_norm": 3.875,
"learning_rate": 1.981926078465748e-05,
"loss": 0.588,
"step": 1182
},
{
"epoch": 0.7735818211541605,
"grad_norm": 4.46875,
"learning_rate": 1.9818395526972373e-05,
"loss": 0.581,
"step": 1183
},
{
"epoch": 0.7742357364721268,
"grad_norm": 3.71875,
"learning_rate": 1.9817528222072595e-05,
"loss": 0.5263,
"step": 1184
},
{
"epoch": 0.7748896517900932,
"grad_norm": 3.8125,
"learning_rate": 1.981665887013899e-05,
"loss": 0.3681,
"step": 1185
},
{
"epoch": 0.7755435671080595,
"grad_norm": 3.859375,
"learning_rate": 1.9815787471352824e-05,
"loss": 0.4546,
"step": 1186
},
{
"epoch": 0.7761974824260258,
"grad_norm": 4.40625,
"learning_rate": 1.981491402589579e-05,
"loss": 0.6558,
"step": 1187
},
{
"epoch": 0.7768513977439921,
"grad_norm": 5.09375,
"learning_rate": 1.9814038533950008e-05,
"loss": 0.5277,
"step": 1188
},
{
"epoch": 0.7775053130619585,
"grad_norm": 4.25,
"learning_rate": 1.9813160995698023e-05,
"loss": 0.7088,
"step": 1189
},
{
"epoch": 0.7781592283799248,
"grad_norm": 2.9375,
"learning_rate": 1.9812281411322812e-05,
"loss": 0.2773,
"step": 1190
},
{
"epoch": 0.7788131436978911,
"grad_norm": 7.375,
"learning_rate": 1.9811399781007775e-05,
"loss": 0.8488,
"step": 1191
},
{
"epoch": 0.7794670590158574,
"grad_norm": 4.46875,
"learning_rate": 1.9810516104936743e-05,
"loss": 0.5473,
"step": 1192
},
{
"epoch": 0.7801209743338238,
"grad_norm": 2.84375,
"learning_rate": 1.980963038329396e-05,
"loss": 0.2899,
"step": 1193
},
{
"epoch": 0.7807748896517901,
"grad_norm": 2.953125,
"learning_rate": 1.9808742616264114e-05,
"loss": 0.3808,
"step": 1194
},
{
"epoch": 0.7814288049697564,
"grad_norm": 4.53125,
"learning_rate": 1.9807852804032306e-05,
"loss": 0.7729,
"step": 1195
},
{
"epoch": 0.7820827202877227,
"grad_norm": 3.5625,
"learning_rate": 1.9806960946784076e-05,
"loss": 0.5398,
"step": 1196
},
{
"epoch": 0.7827366356056891,
"grad_norm": 4.59375,
"learning_rate": 1.9806067044705375e-05,
"loss": 0.8246,
"step": 1197
},
{
"epoch": 0.7833905509236554,
"grad_norm": 4.3125,
"learning_rate": 1.9805171097982595e-05,
"loss": 0.6711,
"step": 1198
},
{
"epoch": 0.7840444662416217,
"grad_norm": 3.5,
"learning_rate": 1.9804273106802546e-05,
"loss": 0.3875,
"step": 1199
},
{
"epoch": 0.784698381559588,
"grad_norm": 6.90625,
"learning_rate": 1.9803373071352467e-05,
"loss": 1.0515,
"step": 1200
},
{
"epoch": 0.7853522968775544,
"grad_norm": 5.625,
"learning_rate": 1.9802470991820023e-05,
"loss": 0.7506,
"step": 1201
},
{
"epoch": 0.7860062121955207,
"grad_norm": 4.0,
"learning_rate": 1.9801566868393303e-05,
"loss": 0.6214,
"step": 1202
},
{
"epoch": 0.786660127513487,
"grad_norm": 4.4375,
"learning_rate": 1.9800660701260826e-05,
"loss": 0.6577,
"step": 1203
},
{
"epoch": 0.7873140428314533,
"grad_norm": 4.65625,
"learning_rate": 1.9799752490611534e-05,
"loss": 0.5119,
"step": 1204
},
{
"epoch": 0.7879679581494197,
"grad_norm": 4.5,
"learning_rate": 1.9798842236634797e-05,
"loss": 0.6337,
"step": 1205
},
{
"epoch": 0.788621873467386,
"grad_norm": 2.90625,
"learning_rate": 1.979792993952041e-05,
"loss": 0.3248,
"step": 1206
},
{
"epoch": 0.7892757887853523,
"grad_norm": 4.125,
"learning_rate": 1.979701559945859e-05,
"loss": 0.4343,
"step": 1207
},
{
"epoch": 0.7899297041033186,
"grad_norm": 3.671875,
"learning_rate": 1.9796099216639995e-05,
"loss": 0.6919,
"step": 1208
},
{
"epoch": 0.790583619421285,
"grad_norm": 3.65625,
"learning_rate": 1.9795180791255687e-05,
"loss": 0.5202,
"step": 1209
},
{
"epoch": 0.7912375347392513,
"grad_norm": 4.25,
"learning_rate": 1.9794260323497175e-05,
"loss": 0.5914,
"step": 1210
},
{
"epoch": 0.7918914500572176,
"grad_norm": 7.34375,
"learning_rate": 1.9793337813556372e-05,
"loss": 0.8604,
"step": 1211
},
{
"epoch": 0.7925453653751839,
"grad_norm": 3.03125,
"learning_rate": 1.9792413261625644e-05,
"loss": 0.4264,
"step": 1212
},
{
"epoch": 0.7931992806931503,
"grad_norm": 6.65625,
"learning_rate": 1.979148666789775e-05,
"loss": 0.9233,
"step": 1213
},
{
"epoch": 0.7938531960111166,
"grad_norm": 4.25,
"learning_rate": 1.9790558032565908e-05,
"loss": 0.4091,
"step": 1214
},
{
"epoch": 0.7945071113290829,
"grad_norm": 4.0,
"learning_rate": 1.9789627355823737e-05,
"loss": 0.5195,
"step": 1215
},
{
"epoch": 0.7951610266470492,
"grad_norm": 4.71875,
"learning_rate": 1.9788694637865292e-05,
"loss": 0.6803,
"step": 1216
},
{
"epoch": 0.7958149419650156,
"grad_norm": 4.4375,
"learning_rate": 1.9787759878885054e-05,
"loss": 0.6369,
"step": 1217
},
{
"epoch": 0.7964688572829819,
"grad_norm": 5.1875,
"learning_rate": 1.9786823079077927e-05,
"loss": 0.7595,
"step": 1218
},
{
"epoch": 0.7971227726009482,
"grad_norm": 4.53125,
"learning_rate": 1.978588423863924e-05,
"loss": 0.6496,
"step": 1219
},
{
"epoch": 0.7977766879189145,
"grad_norm": 4.25,
"learning_rate": 1.9784943357764754e-05,
"loss": 0.6233,
"step": 1220
},
{
"epoch": 0.7984306032368809,
"grad_norm": 4.25,
"learning_rate": 1.978400043665064e-05,
"loss": 0.5194,
"step": 1221
},
{
"epoch": 0.7990845185548472,
"grad_norm": 5.84375,
"learning_rate": 1.9783055475493516e-05,
"loss": 0.9662,
"step": 1222
},
{
"epoch": 0.7997384338728135,
"grad_norm": 4.59375,
"learning_rate": 1.9782108474490408e-05,
"loss": 0.6733,
"step": 1223
},
{
"epoch": 0.8003923491907798,
"grad_norm": 3.234375,
"learning_rate": 1.9781159433838774e-05,
"loss": 0.3441,
"step": 1224
},
{
"epoch": 0.8010462645087462,
"grad_norm": 3.921875,
"learning_rate": 1.9780208353736493e-05,
"loss": 0.5661,
"step": 1225
},
{
"epoch": 0.8017001798267125,
"grad_norm": 4.90625,
"learning_rate": 1.9779255234381883e-05,
"loss": 0.6459,
"step": 1226
},
{
"epoch": 0.8023540951446788,
"grad_norm": 3.375,
"learning_rate": 1.9778300075973668e-05,
"loss": 0.4205,
"step": 1227
},
{
"epoch": 0.8030080104626451,
"grad_norm": 2.234375,
"learning_rate": 1.9777342878711006e-05,
"loss": 0.2515,
"step": 1228
},
{
"epoch": 0.8036619257806115,
"grad_norm": 5.09375,
"learning_rate": 1.977638364279349e-05,
"loss": 0.6933,
"step": 1229
},
{
"epoch": 0.8043158410985778,
"grad_norm": 3.671875,
"learning_rate": 1.9775422368421118e-05,
"loss": 0.2568,
"step": 1230
},
{
"epoch": 0.804969756416544,
"grad_norm": 5.875,
"learning_rate": 1.9774459055794328e-05,
"loss": 0.6884,
"step": 1231
},
{
"epoch": 0.8056236717345103,
"grad_norm": 3.828125,
"learning_rate": 1.977349370511398e-05,
"loss": 0.6073,
"step": 1232
},
{
"epoch": 0.8062775870524767,
"grad_norm": 4.5625,
"learning_rate": 1.9772526316581354e-05,
"loss": 0.7383,
"step": 1233
},
{
"epoch": 0.806931502370443,
"grad_norm": 2.46875,
"learning_rate": 1.9771556890398156e-05,
"loss": 0.3692,
"step": 1234
},
{
"epoch": 0.8075854176884093,
"grad_norm": 5.46875,
"learning_rate": 1.9770585426766528e-05,
"loss": 0.8585,
"step": 1235
},
{
"epoch": 0.8082393330063756,
"grad_norm": 4.75,
"learning_rate": 1.9769611925889016e-05,
"loss": 0.64,
"step": 1236
},
{
"epoch": 0.808893248324342,
"grad_norm": 4.84375,
"learning_rate": 1.9768636387968616e-05,
"loss": 0.6219,
"step": 1237
},
{
"epoch": 0.8095471636423083,
"grad_norm": 4.625,
"learning_rate": 1.9767658813208725e-05,
"loss": 0.8045,
"step": 1238
},
{
"epoch": 0.8102010789602746,
"grad_norm": 3.453125,
"learning_rate": 1.976667920181318e-05,
"loss": 0.4122,
"step": 1239
},
{
"epoch": 0.8108549942782409,
"grad_norm": 3.53125,
"learning_rate": 1.9765697553986237e-05,
"loss": 0.5164,
"step": 1240
},
{
"epoch": 0.8115089095962073,
"grad_norm": 4.15625,
"learning_rate": 1.976471386993258e-05,
"loss": 0.8459,
"step": 1241
},
{
"epoch": 0.8121628249141736,
"grad_norm": 3.234375,
"learning_rate": 1.976372814985731e-05,
"loss": 0.3772,
"step": 1242
},
{
"epoch": 0.8128167402321399,
"grad_norm": 3.40625,
"learning_rate": 1.9762740393965963e-05,
"loss": 0.4406,
"step": 1243
},
{
"epoch": 0.8134706555501062,
"grad_norm": 5.25,
"learning_rate": 1.976175060246449e-05,
"loss": 1.0501,
"step": 1244
},
{
"epoch": 0.8141245708680726,
"grad_norm": 3.90625,
"learning_rate": 1.9760758775559275e-05,
"loss": 0.4366,
"step": 1245
},
{
"epoch": 0.8147784861860389,
"grad_norm": 3.34375,
"learning_rate": 1.9759764913457113e-05,
"loss": 0.4725,
"step": 1246
},
{
"epoch": 0.8154324015040052,
"grad_norm": 6.0,
"learning_rate": 1.9758769016365242e-05,
"loss": 0.719,
"step": 1247
},
{
"epoch": 0.8160863168219715,
"grad_norm": 8.6875,
"learning_rate": 1.975777108449131e-05,
"loss": 1.6058,
"step": 1248
},
{
"epoch": 0.8167402321399378,
"grad_norm": 3.65625,
"learning_rate": 1.9756771118043393e-05,
"loss": 0.5823,
"step": 1249
},
{
"epoch": 0.8173941474579042,
"grad_norm": 4.59375,
"learning_rate": 1.9755769117229993e-05,
"loss": 0.7355,
"step": 1250
},
{
"epoch": 0.8180480627758705,
"grad_norm": 2.921875,
"learning_rate": 1.9754765082260035e-05,
"loss": 0.438,
"step": 1251
},
{
"epoch": 0.8187019780938368,
"grad_norm": 4.25,
"learning_rate": 1.9753759013342875e-05,
"loss": 0.6023,
"step": 1252
},
{
"epoch": 0.8193558934118031,
"grad_norm": 2.984375,
"learning_rate": 1.9752750910688278e-05,
"loss": 0.3565,
"step": 1253
},
{
"epoch": 0.8200098087297695,
"grad_norm": 4.21875,
"learning_rate": 1.975174077450644e-05,
"loss": 0.6343,
"step": 1254
},
{
"epoch": 0.8206637240477358,
"grad_norm": 4.25,
"learning_rate": 1.975072860500799e-05,
"loss": 0.8848,
"step": 1255
},
{
"epoch": 0.8213176393657021,
"grad_norm": 4.6875,
"learning_rate": 1.974971440240397e-05,
"loss": 0.8236,
"step": 1256
},
{
"epoch": 0.8219715546836684,
"grad_norm": 4.53125,
"learning_rate": 1.9748698166905847e-05,
"loss": 0.5921,
"step": 1257
},
{
"epoch": 0.8226254700016348,
"grad_norm": 3.5,
"learning_rate": 1.974767989872552e-05,
"loss": 0.6978,
"step": 1258
},
{
"epoch": 0.8232793853196011,
"grad_norm": 4.34375,
"learning_rate": 1.9746659598075303e-05,
"loss": 0.503,
"step": 1259
},
{
"epoch": 0.8239333006375674,
"grad_norm": 2.984375,
"learning_rate": 1.9745637265167937e-05,
"loss": 0.4463,
"step": 1260
},
{
"epoch": 0.8245872159555337,
"grad_norm": 3.109375,
"learning_rate": 1.9744612900216588e-05,
"loss": 0.3345,
"step": 1261
},
{
"epoch": 0.8252411312735001,
"grad_norm": 3.453125,
"learning_rate": 1.9743586503434842e-05,
"loss": 0.4771,
"step": 1262
},
{
"epoch": 0.8258950465914664,
"grad_norm": 4.96875,
"learning_rate": 1.9742558075036712e-05,
"loss": 0.6727,
"step": 1263
},
{
"epoch": 0.8265489619094327,
"grad_norm": 3.71875,
"learning_rate": 1.974152761523663e-05,
"loss": 0.5154,
"step": 1264
},
{
"epoch": 0.827202877227399,
"grad_norm": 3.765625,
"learning_rate": 1.9740495124249462e-05,
"loss": 0.4207,
"step": 1265
},
{
"epoch": 0.8278567925453654,
"grad_norm": 4.5625,
"learning_rate": 1.973946060229049e-05,
"loss": 0.6274,
"step": 1266
},
{
"epoch": 0.8285107078633317,
"grad_norm": 4.3125,
"learning_rate": 1.973842404957541e-05,
"loss": 0.4892,
"step": 1267
},
{
"epoch": 0.829164623181298,
"grad_norm": 4.40625,
"learning_rate": 1.9737385466320366e-05,
"loss": 0.4182,
"step": 1268
},
{
"epoch": 0.8298185384992643,
"grad_norm": 6.125,
"learning_rate": 1.9736344852741905e-05,
"loss": 1.1654,
"step": 1269
},
{
"epoch": 0.8304724538172307,
"grad_norm": 3.21875,
"learning_rate": 1.9735302209057e-05,
"loss": 0.352,
"step": 1270
},
{
"epoch": 0.831126369135197,
"grad_norm": 3.84375,
"learning_rate": 1.973425753548305e-05,
"loss": 0.5972,
"step": 1271
},
{
"epoch": 0.8317802844531633,
"grad_norm": 5.34375,
"learning_rate": 1.9733210832237887e-05,
"loss": 0.854,
"step": 1272
},
{
"epoch": 0.8324341997711296,
"grad_norm": 6.5,
"learning_rate": 1.9732162099539748e-05,
"loss": 0.8069,
"step": 1273
},
{
"epoch": 0.833088115089096,
"grad_norm": 6.53125,
"learning_rate": 1.9731111337607304e-05,
"loss": 0.87,
"step": 1274
},
{
"epoch": 0.8337420304070623,
"grad_norm": 4.3125,
"learning_rate": 1.9730058546659653e-05,
"loss": 0.4515,
"step": 1275
},
{
"epoch": 0.8343959457250286,
"grad_norm": 4.59375,
"learning_rate": 1.97290037269163e-05,
"loss": 0.7724,
"step": 1276
},
{
"epoch": 0.835049861042995,
"grad_norm": 6.0,
"learning_rate": 1.9727946878597193e-05,
"loss": 0.8362,
"step": 1277
},
{
"epoch": 0.8357037763609613,
"grad_norm": 3.046875,
"learning_rate": 1.9726888001922693e-05,
"loss": 0.3663,
"step": 1278
},
{
"epoch": 0.8363576916789276,
"grad_norm": 3.734375,
"learning_rate": 1.9725827097113578e-05,
"loss": 0.3654,
"step": 1279
},
{
"epoch": 0.8370116069968939,
"grad_norm": 4.625,
"learning_rate": 1.9724764164391063e-05,
"loss": 0.6039,
"step": 1280
},
{
"epoch": 0.8376655223148602,
"grad_norm": 5.78125,
"learning_rate": 1.9723699203976768e-05,
"loss": 1.0668,
"step": 1281
},
{
"epoch": 0.8383194376328266,
"grad_norm": 5.78125,
"learning_rate": 1.9722632216092752e-05,
"loss": 0.9367,
"step": 1282
},
{
"epoch": 0.8389733529507929,
"grad_norm": 2.9375,
"learning_rate": 1.9721563200961494e-05,
"loss": 0.3356,
"step": 1283
},
{
"epoch": 0.8396272682687592,
"grad_norm": 4.65625,
"learning_rate": 1.9720492158805884e-05,
"loss": 0.4986,
"step": 1284
},
{
"epoch": 0.8402811835867255,
"grad_norm": 4.875,
"learning_rate": 1.971941908984925e-05,
"loss": 0.6653,
"step": 1285
},
{
"epoch": 0.8409350989046919,
"grad_norm": 4.9375,
"learning_rate": 1.971834399431533e-05,
"loss": 0.7116,
"step": 1286
},
{
"epoch": 0.8415890142226582,
"grad_norm": 6.71875,
"learning_rate": 1.9717266872428295e-05,
"loss": 0.9576,
"step": 1287
},
{
"epoch": 0.8422429295406245,
"grad_norm": 4.40625,
"learning_rate": 1.9716187724412728e-05,
"loss": 0.567,
"step": 1288
},
{
"epoch": 0.8428968448585908,
"grad_norm": 4.625,
"learning_rate": 1.9715106550493645e-05,
"loss": 0.442,
"step": 1289
},
{
"epoch": 0.8435507601765572,
"grad_norm": 3.96875,
"learning_rate": 1.971402335089648e-05,
"loss": 0.5429,
"step": 1290
},
{
"epoch": 0.8442046754945235,
"grad_norm": 3.0625,
"learning_rate": 1.9712938125847086e-05,
"loss": 0.3932,
"step": 1291
},
{
"epoch": 0.8448585908124898,
"grad_norm": 4.46875,
"learning_rate": 1.971185087557174e-05,
"loss": 0.7773,
"step": 1292
},
{
"epoch": 0.8455125061304561,
"grad_norm": 3.875,
"learning_rate": 1.9710761600297147e-05,
"loss": 0.7628,
"step": 1293
},
{
"epoch": 0.8461664214484225,
"grad_norm": 3.71875,
"learning_rate": 1.9709670300250425e-05,
"loss": 0.5101,
"step": 1294
},
{
"epoch": 0.8468203367663888,
"grad_norm": 3.140625,
"learning_rate": 1.9708576975659123e-05,
"loss": 0.5427,
"step": 1295
},
{
"epoch": 0.8474742520843551,
"grad_norm": 4.28125,
"learning_rate": 1.9707481626751206e-05,
"loss": 0.632,
"step": 1296
},
{
"epoch": 0.8481281674023214,
"grad_norm": 4.0625,
"learning_rate": 1.9706384253755063e-05,
"loss": 0.4592,
"step": 1297
},
{
"epoch": 0.8487820827202878,
"grad_norm": 4.59375,
"learning_rate": 1.97052848568995e-05,
"loss": 0.579,
"step": 1298
},
{
"epoch": 0.8494359980382541,
"grad_norm": 5.46875,
"learning_rate": 1.9704183436413764e-05,
"loss": 1.0889,
"step": 1299
},
{
"epoch": 0.8500899133562204,
"grad_norm": 4.375,
"learning_rate": 1.9703079992527495e-05,
"loss": 0.7516,
"step": 1300
},
{
"epoch": 0.8507438286741867,
"grad_norm": 3.296875,
"learning_rate": 1.970197452547078e-05,
"loss": 0.3716,
"step": 1301
},
{
"epoch": 0.8513977439921531,
"grad_norm": 3.28125,
"learning_rate": 1.970086703547411e-05,
"loss": 0.371,
"step": 1302
},
{
"epoch": 0.8520516593101194,
"grad_norm": 4.09375,
"learning_rate": 1.9699757522768413e-05,
"loss": 0.51,
"step": 1303
},
{
"epoch": 0.8527055746280857,
"grad_norm": 4.34375,
"learning_rate": 1.969864598758503e-05,
"loss": 0.6076,
"step": 1304
},
{
"epoch": 0.853359489946052,
"grad_norm": 5.375,
"learning_rate": 1.9697532430155718e-05,
"loss": 0.934,
"step": 1305
},
{
"epoch": 0.8540134052640183,
"grad_norm": 4.78125,
"learning_rate": 1.969641685071267e-05,
"loss": 0.9857,
"step": 1306
},
{
"epoch": 0.8546673205819846,
"grad_norm": 2.578125,
"learning_rate": 1.9695299249488488e-05,
"loss": 0.2168,
"step": 1307
},
{
"epoch": 0.8553212358999509,
"grad_norm": 9.375,
"learning_rate": 1.969417962671621e-05,
"loss": 0.8594,
"step": 1308
},
{
"epoch": 0.8559751512179172,
"grad_norm": 4.96875,
"learning_rate": 1.9693057982629277e-05,
"loss": 0.7069,
"step": 1309
},
{
"epoch": 0.8566290665358836,
"grad_norm": 5.125,
"learning_rate": 1.9691934317461565e-05,
"loss": 1.0717,
"step": 1310
},
{
"epoch": 0.8572829818538499,
"grad_norm": 5.46875,
"learning_rate": 1.9690808631447368e-05,
"loss": 0.7704,
"step": 1311
},
{
"epoch": 0.8579368971718162,
"grad_norm": 7.25,
"learning_rate": 1.9689680924821395e-05,
"loss": 1.098,
"step": 1312
},
{
"epoch": 0.8585908124897825,
"grad_norm": 4.09375,
"learning_rate": 1.9688551197818788e-05,
"loss": 0.5938,
"step": 1313
},
{
"epoch": 0.8592447278077489,
"grad_norm": 4.53125,
"learning_rate": 1.9687419450675104e-05,
"loss": 0.7635,
"step": 1314
},
{
"epoch": 0.8598986431257152,
"grad_norm": 3.53125,
"learning_rate": 1.9686285683626316e-05,
"loss": 0.3596,
"step": 1315
},
{
"epoch": 0.8605525584436815,
"grad_norm": 4.40625,
"learning_rate": 1.9685149896908832e-05,
"loss": 0.6028,
"step": 1316
},
{
"epoch": 0.8612064737616478,
"grad_norm": 3.6875,
"learning_rate": 1.968401209075946e-05,
"loss": 0.496,
"step": 1317
},
{
"epoch": 0.8618603890796142,
"grad_norm": 3.28125,
"learning_rate": 1.9682872265415458e-05,
"loss": 0.4101,
"step": 1318
},
{
"epoch": 0.8625143043975805,
"grad_norm": 4.875,
"learning_rate": 1.9681730421114475e-05,
"loss": 0.481,
"step": 1319
},
{
"epoch": 0.8631682197155468,
"grad_norm": 5.59375,
"learning_rate": 1.9680586558094604e-05,
"loss": 0.5441,
"step": 1320
},
{
"epoch": 0.8638221350335131,
"grad_norm": 4.0,
"learning_rate": 1.9679440676594342e-05,
"loss": 0.4592,
"step": 1321
},
{
"epoch": 0.8644760503514795,
"grad_norm": 5.0625,
"learning_rate": 1.967829277685262e-05,
"loss": 0.6525,
"step": 1322
},
{
"epoch": 0.8651299656694458,
"grad_norm": 4.03125,
"learning_rate": 1.9677142859108784e-05,
"loss": 0.367,
"step": 1323
},
{
"epoch": 0.8657838809874121,
"grad_norm": 4.09375,
"learning_rate": 1.96759909236026e-05,
"loss": 0.5479,
"step": 1324
},
{
"epoch": 0.8664377963053784,
"grad_norm": 5.0,
"learning_rate": 1.9674836970574253e-05,
"loss": 0.7971,
"step": 1325
},
{
"epoch": 0.8670917116233448,
"grad_norm": 4.34375,
"learning_rate": 1.9673681000264358e-05,
"loss": 0.6036,
"step": 1326
},
{
"epoch": 0.8677456269413111,
"grad_norm": 5.875,
"learning_rate": 1.9672523012913937e-05,
"loss": 0.9767,
"step": 1327
},
{
"epoch": 0.8683995422592774,
"grad_norm": 4.28125,
"learning_rate": 1.9671363008764445e-05,
"loss": 0.666,
"step": 1328
},
{
"epoch": 0.8690534575772437,
"grad_norm": 5.125,
"learning_rate": 1.967020098805775e-05,
"loss": 0.8879,
"step": 1329
},
{
"epoch": 0.86970737289521,
"grad_norm": 4.40625,
"learning_rate": 1.966903695103614e-05,
"loss": 0.576,
"step": 1330
},
{
"epoch": 0.8703612882131764,
"grad_norm": 4.15625,
"learning_rate": 1.966787089794233e-05,
"loss": 0.4973,
"step": 1331
},
{
"epoch": 0.8710152035311427,
"grad_norm": 5.875,
"learning_rate": 1.9666702829019455e-05,
"loss": 1.0025,
"step": 1332
},
{
"epoch": 0.871669118849109,
"grad_norm": 3.75,
"learning_rate": 1.966553274451106e-05,
"loss": 0.4209,
"step": 1333
},
{
"epoch": 0.8723230341670754,
"grad_norm": 4.875,
"learning_rate": 1.9664360644661118e-05,
"loss": 0.7864,
"step": 1334
},
{
"epoch": 0.8729769494850417,
"grad_norm": 4.0625,
"learning_rate": 1.9663186529714022e-05,
"loss": 0.6316,
"step": 1335
},
{
"epoch": 0.873630864803008,
"grad_norm": 7.34375,
"learning_rate": 1.9662010399914588e-05,
"loss": 0.8036,
"step": 1336
},
{
"epoch": 0.8742847801209743,
"grad_norm": 5.25,
"learning_rate": 1.9660832255508045e-05,
"loss": 0.937,
"step": 1337
},
{
"epoch": 0.8749386954389407,
"grad_norm": 4.84375,
"learning_rate": 1.9659652096740042e-05,
"loss": 0.8778,
"step": 1338
},
{
"epoch": 0.875592610756907,
"grad_norm": 4.15625,
"learning_rate": 1.9658469923856664e-05,
"loss": 0.4902,
"step": 1339
},
{
"epoch": 0.8762465260748733,
"grad_norm": 4.1875,
"learning_rate": 1.965728573710439e-05,
"loss": 0.7317,
"step": 1340
},
{
"epoch": 0.8769004413928396,
"grad_norm": 3.671875,
"learning_rate": 1.965609953673014e-05,
"loss": 0.6844,
"step": 1341
},
{
"epoch": 0.877554356710806,
"grad_norm": 5.90625,
"learning_rate": 1.965491132298124e-05,
"loss": 0.5328,
"step": 1342
},
{
"epoch": 0.8782082720287723,
"grad_norm": 5.96875,
"learning_rate": 1.9653721096105448e-05,
"loss": 0.8051,
"step": 1343
},
{
"epoch": 0.8788621873467386,
"grad_norm": 3.46875,
"learning_rate": 1.965252885635094e-05,
"loss": 0.5046,
"step": 1344
},
{
"epoch": 0.8795161026647049,
"grad_norm": 5.0,
"learning_rate": 1.9651334603966298e-05,
"loss": 0.6676,
"step": 1345
},
{
"epoch": 0.8801700179826712,
"grad_norm": 5.21875,
"learning_rate": 1.9650138339200535e-05,
"loss": 0.766,
"step": 1346
},
{
"epoch": 0.8808239333006376,
"grad_norm": 4.34375,
"learning_rate": 1.9648940062303085e-05,
"loss": 0.6907,
"step": 1347
},
{
"epoch": 0.8814778486186039,
"grad_norm": 3.671875,
"learning_rate": 1.96477397735238e-05,
"loss": 0.5437,
"step": 1348
},
{
"epoch": 0.8821317639365702,
"grad_norm": 4.03125,
"learning_rate": 1.9646537473112946e-05,
"loss": 0.6664,
"step": 1349
},
{
"epoch": 0.8827856792545365,
"grad_norm": 5.0625,
"learning_rate": 1.9645333161321213e-05,
"loss": 0.8064,
"step": 1350
},
{
"epoch": 0.8834395945725029,
"grad_norm": 4.03125,
"learning_rate": 1.964412683839971e-05,
"loss": 0.5848,
"step": 1351
},
{
"epoch": 0.8840935098904692,
"grad_norm": 2.90625,
"learning_rate": 1.9642918504599967e-05,
"loss": 0.3608,
"step": 1352
},
{
"epoch": 0.8847474252084355,
"grad_norm": 3.625,
"learning_rate": 1.964170816017393e-05,
"loss": 0.8004,
"step": 1353
},
{
"epoch": 0.8854013405264018,
"grad_norm": 5.25,
"learning_rate": 1.9640495805373965e-05,
"loss": 0.6615,
"step": 1354
},
{
"epoch": 0.8860552558443682,
"grad_norm": 5.59375,
"learning_rate": 1.9639281440452858e-05,
"loss": 0.9144,
"step": 1355
},
{
"epoch": 0.8867091711623345,
"grad_norm": 5.09375,
"learning_rate": 1.9638065065663815e-05,
"loss": 0.9773,
"step": 1356
},
{
"epoch": 0.8873630864803008,
"grad_norm": 3.4375,
"learning_rate": 1.963684668126046e-05,
"loss": 0.4712,
"step": 1357
},
{
"epoch": 0.8880170017982671,
"grad_norm": 4.84375,
"learning_rate": 1.9635626287496833e-05,
"loss": 0.7769,
"step": 1358
},
{
"epoch": 0.8886709171162335,
"grad_norm": 3.828125,
"learning_rate": 1.9634403884627394e-05,
"loss": 0.6143,
"step": 1359
},
{
"epoch": 0.8893248324341998,
"grad_norm": 4.09375,
"learning_rate": 1.9633179472907032e-05,
"loss": 0.5514,
"step": 1360
},
{
"epoch": 0.8899787477521661,
"grad_norm": 5.125,
"learning_rate": 1.963195305259104e-05,
"loss": 0.5352,
"step": 1361
},
{
"epoch": 0.8906326630701324,
"grad_norm": 4.0625,
"learning_rate": 1.9630724623935146e-05,
"loss": 0.4738,
"step": 1362
},
{
"epoch": 0.8912865783880988,
"grad_norm": 5.3125,
"learning_rate": 1.9629494187195473e-05,
"loss": 0.8231,
"step": 1363
},
{
"epoch": 0.8919404937060651,
"grad_norm": 5.03125,
"learning_rate": 1.9628261742628587e-05,
"loss": 0.894,
"step": 1364
},
{
"epoch": 0.8925944090240314,
"grad_norm": 4.03125,
"learning_rate": 1.962702729049146e-05,
"loss": 0.6571,
"step": 1365
},
{
"epoch": 0.8932483243419977,
"grad_norm": 6.0,
"learning_rate": 1.9625790831041484e-05,
"loss": 0.8724,
"step": 1366
},
{
"epoch": 0.8939022396599641,
"grad_norm": 5.75,
"learning_rate": 1.9624552364536472e-05,
"loss": 0.6192,
"step": 1367
},
{
"epoch": 0.8945561549779304,
"grad_norm": 5.46875,
"learning_rate": 1.962331189123466e-05,
"loss": 0.872,
"step": 1368
},
{
"epoch": 0.8952100702958967,
"grad_norm": 4.09375,
"learning_rate": 1.9622069411394685e-05,
"loss": 0.6936,
"step": 1369
},
{
"epoch": 0.895863985613863,
"grad_norm": 4.125,
"learning_rate": 1.962082492527562e-05,
"loss": 0.5064,
"step": 1370
},
{
"epoch": 0.8965179009318294,
"grad_norm": 5.0625,
"learning_rate": 1.961957843313695e-05,
"loss": 0.7711,
"step": 1371
},
{
"epoch": 0.8971718162497957,
"grad_norm": 3.953125,
"learning_rate": 1.9618329935238582e-05,
"loss": 0.4953,
"step": 1372
},
{
"epoch": 0.897825731567762,
"grad_norm": 4.4375,
"learning_rate": 1.961707943184083e-05,
"loss": 0.541,
"step": 1373
},
{
"epoch": 0.8984796468857283,
"grad_norm": 3.75,
"learning_rate": 1.9615826923204446e-05,
"loss": 0.4206,
"step": 1374
},
{
"epoch": 0.8991335622036947,
"grad_norm": 6.65625,
"learning_rate": 1.9614572409590575e-05,
"loss": 1.1008,
"step": 1375
},
{
"epoch": 0.899787477521661,
"grad_norm": 4.21875,
"learning_rate": 1.96133158912608e-05,
"loss": 0.5554,
"step": 1376
},
{
"epoch": 0.9004413928396273,
"grad_norm": 3.6875,
"learning_rate": 1.9612057368477117e-05,
"loss": 0.4283,
"step": 1377
},
{
"epoch": 0.9010953081575936,
"grad_norm": 5.78125,
"learning_rate": 1.9610796841501935e-05,
"loss": 0.7818,
"step": 1378
},
{
"epoch": 0.90174922347556,
"grad_norm": 4.03125,
"learning_rate": 1.9609534310598083e-05,
"loss": 0.6435,
"step": 1379
},
{
"epoch": 0.9024031387935263,
"grad_norm": 4.59375,
"learning_rate": 1.9608269776028813e-05,
"loss": 0.6275,
"step": 1380
},
{
"epoch": 0.9030570541114925,
"grad_norm": 3.375,
"learning_rate": 1.960700323805779e-05,
"loss": 0.4377,
"step": 1381
},
{
"epoch": 0.9037109694294588,
"grad_norm": 5.78125,
"learning_rate": 1.960573469694909e-05,
"loss": 0.5706,
"step": 1382
},
{
"epoch": 0.9043648847474252,
"grad_norm": 3.109375,
"learning_rate": 1.960446415296722e-05,
"loss": 0.3354,
"step": 1383
},
{
"epoch": 0.9050188000653915,
"grad_norm": 4.6875,
"learning_rate": 1.9603191606377105e-05,
"loss": 0.6345,
"step": 1384
},
{
"epoch": 0.9056727153833578,
"grad_norm": 3.96875,
"learning_rate": 1.960191705744407e-05,
"loss": 0.6971,
"step": 1385
},
{
"epoch": 0.9063266307013241,
"grad_norm": 3.15625,
"learning_rate": 1.9600640506433873e-05,
"loss": 0.4258,
"step": 1386
},
{
"epoch": 0.9069805460192905,
"grad_norm": 3.40625,
"learning_rate": 1.9599361953612685e-05,
"loss": 0.4723,
"step": 1387
},
{
"epoch": 0.9076344613372568,
"grad_norm": 4.46875,
"learning_rate": 1.95980813992471e-05,
"loss": 0.7193,
"step": 1388
},
{
"epoch": 0.9082883766552231,
"grad_norm": 3.40625,
"learning_rate": 1.9596798843604114e-05,
"loss": 0.4142,
"step": 1389
},
{
"epoch": 0.9089422919731894,
"grad_norm": 4.15625,
"learning_rate": 1.9595514286951155e-05,
"loss": 0.6368,
"step": 1390
},
{
"epoch": 0.9095962072911558,
"grad_norm": 4.375,
"learning_rate": 1.9594227729556064e-05,
"loss": 0.4688,
"step": 1391
},
{
"epoch": 0.9102501226091221,
"grad_norm": 3.875,
"learning_rate": 1.95929391716871e-05,
"loss": 0.5326,
"step": 1392
},
{
"epoch": 0.9109040379270884,
"grad_norm": 4.8125,
"learning_rate": 1.9591648613612936e-05,
"loss": 0.9176,
"step": 1393
},
{
"epoch": 0.9115579532450547,
"grad_norm": 6.40625,
"learning_rate": 1.9590356055602662e-05,
"loss": 1.0696,
"step": 1394
},
{
"epoch": 0.912211868563021,
"grad_norm": 4.15625,
"learning_rate": 1.958906149792579e-05,
"loss": 0.4847,
"step": 1395
},
{
"epoch": 0.9128657838809874,
"grad_norm": 3.34375,
"learning_rate": 1.958776494085224e-05,
"loss": 0.4474,
"step": 1396
},
{
"epoch": 0.9135196991989537,
"grad_norm": 4.90625,
"learning_rate": 1.9586466384652366e-05,
"loss": 0.4446,
"step": 1397
},
{
"epoch": 0.91417361451692,
"grad_norm": 3.625,
"learning_rate": 1.9585165829596917e-05,
"loss": 0.4773,
"step": 1398
},
{
"epoch": 0.9148275298348864,
"grad_norm": 4.65625,
"learning_rate": 1.9583863275957072e-05,
"loss": 0.653,
"step": 1399
},
{
"epoch": 0.9154814451528527,
"grad_norm": 3.578125,
"learning_rate": 1.9582558724004423e-05,
"loss": 0.4315,
"step": 1400
},
{
"epoch": 0.916135360470819,
"grad_norm": 6.25,
"learning_rate": 1.958125217401098e-05,
"loss": 1.4404,
"step": 1401
},
{
"epoch": 0.9167892757887853,
"grad_norm": 6.0625,
"learning_rate": 1.9579943626249178e-05,
"loss": 0.7897,
"step": 1402
},
{
"epoch": 0.9174431911067517,
"grad_norm": 3.796875,
"learning_rate": 1.9578633080991843e-05,
"loss": 0.4445,
"step": 1403
},
{
"epoch": 0.918097106424718,
"grad_norm": 6.3125,
"learning_rate": 1.9577320538512246e-05,
"loss": 0.7226,
"step": 1404
},
{
"epoch": 0.9187510217426843,
"grad_norm": 4.15625,
"learning_rate": 1.957600599908406e-05,
"loss": 0.6411,
"step": 1405
},
{
"epoch": 0.9194049370606506,
"grad_norm": 5.21875,
"learning_rate": 1.957468946298138e-05,
"loss": 0.6734,
"step": 1406
},
{
"epoch": 0.920058852378617,
"grad_norm": 4.09375,
"learning_rate": 1.957337093047871e-05,
"loss": 0.7218,
"step": 1407
},
{
"epoch": 0.9207127676965833,
"grad_norm": 3.65625,
"learning_rate": 1.9572050401850972e-05,
"loss": 0.5851,
"step": 1408
},
{
"epoch": 0.9213666830145496,
"grad_norm": 5.71875,
"learning_rate": 1.9570727877373514e-05,
"loss": 0.7319,
"step": 1409
},
{
"epoch": 0.9220205983325159,
"grad_norm": 3.265625,
"learning_rate": 1.956940335732209e-05,
"loss": 0.4833,
"step": 1410
},
{
"epoch": 0.9226745136504823,
"grad_norm": 5.0,
"learning_rate": 1.9568076841972875e-05,
"loss": 0.8887,
"step": 1411
},
{
"epoch": 0.9233284289684486,
"grad_norm": 3.40625,
"learning_rate": 1.9566748331602453e-05,
"loss": 0.54,
"step": 1412
},
{
"epoch": 0.9239823442864149,
"grad_norm": 5.71875,
"learning_rate": 1.9565417826487835e-05,
"loss": 0.6719,
"step": 1413
},
{
"epoch": 0.9246362596043812,
"grad_norm": 4.625,
"learning_rate": 1.9564085326906434e-05,
"loss": 0.5223,
"step": 1414
},
{
"epoch": 0.9252901749223476,
"grad_norm": 3.6875,
"learning_rate": 1.9562750833136096e-05,
"loss": 0.5198,
"step": 1415
},
{
"epoch": 0.9259440902403139,
"grad_norm": 4.5,
"learning_rate": 1.9561414345455073e-05,
"loss": 0.5871,
"step": 1416
},
{
"epoch": 0.9265980055582802,
"grad_norm": 4.125,
"learning_rate": 1.9560075864142028e-05,
"loss": 0.5454,
"step": 1417
},
{
"epoch": 0.9272519208762465,
"grad_norm": 4.8125,
"learning_rate": 1.9558735389476044e-05,
"loss": 0.5979,
"step": 1418
},
{
"epoch": 0.9279058361942129,
"grad_norm": 5.09375,
"learning_rate": 1.955739292173663e-05,
"loss": 0.9165,
"step": 1419
},
{
"epoch": 0.9285597515121792,
"grad_norm": 3.484375,
"learning_rate": 1.9556048461203692e-05,
"loss": 0.4034,
"step": 1420
},
{
"epoch": 0.9292136668301455,
"grad_norm": 7.65625,
"learning_rate": 1.9554702008157567e-05,
"loss": 1.1863,
"step": 1421
},
{
"epoch": 0.9298675821481118,
"grad_norm": 3.4375,
"learning_rate": 1.9553353562878997e-05,
"loss": 0.6109,
"step": 1422
},
{
"epoch": 0.9305214974660782,
"grad_norm": 3.015625,
"learning_rate": 1.9552003125649145e-05,
"loss": 0.5343,
"step": 1423
},
{
"epoch": 0.9311754127840445,
"grad_norm": 3.625,
"learning_rate": 1.9550650696749593e-05,
"loss": 0.487,
"step": 1424
},
{
"epoch": 0.9318293281020108,
"grad_norm": 3.296875,
"learning_rate": 1.9549296276462326e-05,
"loss": 0.504,
"step": 1425
},
{
"epoch": 0.9324832434199771,
"grad_norm": 4.625,
"learning_rate": 1.9547939865069755e-05,
"loss": 0.8877,
"step": 1426
},
{
"epoch": 0.9331371587379435,
"grad_norm": 3.921875,
"learning_rate": 1.9546581462854705e-05,
"loss": 0.7203,
"step": 1427
},
{
"epoch": 0.9337910740559098,
"grad_norm": 3.71875,
"learning_rate": 1.9545221070100407e-05,
"loss": 0.5545,
"step": 1428
},
{
"epoch": 0.9344449893738761,
"grad_norm": 7.125,
"learning_rate": 1.9543858687090517e-05,
"loss": 0.824,
"step": 1429
},
{
"epoch": 0.9350989046918424,
"grad_norm": 2.6875,
"learning_rate": 1.954249431410911e-05,
"loss": 0.3832,
"step": 1430
},
{
"epoch": 0.9357528200098088,
"grad_norm": 3.78125,
"learning_rate": 1.9541127951440656e-05,
"loss": 0.7981,
"step": 1431
},
{
"epoch": 0.9364067353277751,
"grad_norm": 5.90625,
"learning_rate": 1.953975959937006e-05,
"loss": 0.6708,
"step": 1432
},
{
"epoch": 0.9370606506457414,
"grad_norm": 3.734375,
"learning_rate": 1.9538389258182638e-05,
"loss": 0.5624,
"step": 1433
},
{
"epoch": 0.9377145659637077,
"grad_norm": 5.34375,
"learning_rate": 1.953701692816411e-05,
"loss": 0.883,
"step": 1434
},
{
"epoch": 0.938368481281674,
"grad_norm": 3.640625,
"learning_rate": 1.9535642609600625e-05,
"loss": 0.5844,
"step": 1435
},
{
"epoch": 0.9390223965996404,
"grad_norm": 3.71875,
"learning_rate": 1.953426630277873e-05,
"loss": 0.5852,
"step": 1436
},
{
"epoch": 0.9396763119176067,
"grad_norm": 5.09375,
"learning_rate": 1.9532888007985408e-05,
"loss": 0.8348,
"step": 1437
},
{
"epoch": 0.940330227235573,
"grad_norm": 4.53125,
"learning_rate": 1.953150772550803e-05,
"loss": 0.4851,
"step": 1438
},
{
"epoch": 0.9409841425535393,
"grad_norm": 3.6875,
"learning_rate": 1.9530125455634407e-05,
"loss": 0.6099,
"step": 1439
},
{
"epoch": 0.9416380578715057,
"grad_norm": 3.109375,
"learning_rate": 1.952874119865275e-05,
"loss": 0.3754,
"step": 1440
},
{
"epoch": 0.942291973189472,
"grad_norm": 4.25,
"learning_rate": 1.9527354954851693e-05,
"loss": 0.4043,
"step": 1441
},
{
"epoch": 0.9429458885074383,
"grad_norm": 8.4375,
"learning_rate": 1.952596672452027e-05,
"loss": 1.1566,
"step": 1442
},
{
"epoch": 0.9435998038254046,
"grad_norm": 3.859375,
"learning_rate": 1.9524576507947942e-05,
"loss": 0.4023,
"step": 1443
},
{
"epoch": 0.944253719143371,
"grad_norm": 5.1875,
"learning_rate": 1.952318430542458e-05,
"loss": 0.6585,
"step": 1444
},
{
"epoch": 0.9449076344613373,
"grad_norm": 4.03125,
"learning_rate": 1.9521790117240472e-05,
"loss": 0.7138,
"step": 1445
},
{
"epoch": 0.9455615497793036,
"grad_norm": 4.96875,
"learning_rate": 1.9520393943686313e-05,
"loss": 0.6334,
"step": 1446
},
{
"epoch": 0.94621546509727,
"grad_norm": 5.4375,
"learning_rate": 1.951899578505322e-05,
"loss": 0.9144,
"step": 1447
},
{
"epoch": 0.9468693804152363,
"grad_norm": 2.859375,
"learning_rate": 1.9517595641632723e-05,
"loss": 0.3517,
"step": 1448
},
{
"epoch": 0.9475232957332026,
"grad_norm": 4.1875,
"learning_rate": 1.951619351371676e-05,
"loss": 0.5874,
"step": 1449
},
{
"epoch": 0.9481772110511689,
"grad_norm": 6.3125,
"learning_rate": 1.951478940159768e-05,
"loss": 0.7009,
"step": 1450
},
{
"epoch": 0.9488311263691352,
"grad_norm": 4.625,
"learning_rate": 1.951338330556826e-05,
"loss": 0.5022,
"step": 1451
},
{
"epoch": 0.9494850416871016,
"grad_norm": 3.953125,
"learning_rate": 1.951197522592168e-05,
"loss": 0.6131,
"step": 1452
},
{
"epoch": 0.9501389570050679,
"grad_norm": 4.53125,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.5976,
"step": 1453
},
{
"epoch": 0.9507928723230342,
"grad_norm": 3.640625,
"learning_rate": 1.950915311695184e-05,
"loss": 0.4705,
"step": 1454
},
{
"epoch": 0.9514467876410004,
"grad_norm": 2.640625,
"learning_rate": 1.9507739088217007e-05,
"loss": 0.3424,
"step": 1455
},
{
"epoch": 0.9521007029589668,
"grad_norm": 4.75,
"learning_rate": 1.950632307704188e-05,
"loss": 0.9069,
"step": 1456
},
{
"epoch": 0.9527546182769331,
"grad_norm": 2.828125,
"learning_rate": 1.950490508372171e-05,
"loss": 0.3473,
"step": 1457
},
{
"epoch": 0.9534085335948994,
"grad_norm": 5.8125,
"learning_rate": 1.9503485108552157e-05,
"loss": 0.8924,
"step": 1458
},
{
"epoch": 0.9540624489128657,
"grad_norm": 3.578125,
"learning_rate": 1.9502063151829297e-05,
"loss": 0.3127,
"step": 1459
},
{
"epoch": 0.9547163642308321,
"grad_norm": 4.6875,
"learning_rate": 1.950063921384962e-05,
"loss": 0.8086,
"step": 1460
},
{
"epoch": 0.9553702795487984,
"grad_norm": 4.09375,
"learning_rate": 1.9499213294910025e-05,
"loss": 0.5399,
"step": 1461
},
{
"epoch": 0.9560241948667647,
"grad_norm": 3.328125,
"learning_rate": 1.9497785395307834e-05,
"loss": 0.6244,
"step": 1462
},
{
"epoch": 0.956678110184731,
"grad_norm": 3.3125,
"learning_rate": 1.9496355515340773e-05,
"loss": 0.32,
"step": 1463
},
{
"epoch": 0.9573320255026974,
"grad_norm": 3.1875,
"learning_rate": 1.9494923655306987e-05,
"loss": 0.3978,
"step": 1464
},
{
"epoch": 0.9579859408206637,
"grad_norm": 5.90625,
"learning_rate": 1.949348981550502e-05,
"loss": 0.7608,
"step": 1465
},
{
"epoch": 0.95863985613863,
"grad_norm": 5.625,
"learning_rate": 1.949205399623385e-05,
"loss": 0.8117,
"step": 1466
},
{
"epoch": 0.9592937714565963,
"grad_norm": 4.0,
"learning_rate": 1.9490616197792852e-05,
"loss": 0.3998,
"step": 1467
},
{
"epoch": 0.9599476867745627,
"grad_norm": 3.90625,
"learning_rate": 1.948917642048182e-05,
"loss": 0.5844,
"step": 1468
},
{
"epoch": 0.960601602092529,
"grad_norm": 4.65625,
"learning_rate": 1.9487734664600956e-05,
"loss": 0.6231,
"step": 1469
},
{
"epoch": 0.9612555174104953,
"grad_norm": 3.5625,
"learning_rate": 1.9486290930450886e-05,
"loss": 0.4554,
"step": 1470
},
{
"epoch": 0.9619094327284616,
"grad_norm": 5.59375,
"learning_rate": 1.9484845218332632e-05,
"loss": 0.8052,
"step": 1471
},
{
"epoch": 0.962563348046428,
"grad_norm": 4.5,
"learning_rate": 1.9483397528547633e-05,
"loss": 0.7266,
"step": 1472
},
{
"epoch": 0.9632172633643943,
"grad_norm": 6.59375,
"learning_rate": 1.9481947861397758e-05,
"loss": 0.8459,
"step": 1473
},
{
"epoch": 0.9638711786823606,
"grad_norm": 4.1875,
"learning_rate": 1.948049621718526e-05,
"loss": 0.5965,
"step": 1474
},
{
"epoch": 0.9645250940003269,
"grad_norm": 5.25,
"learning_rate": 1.9479042596212833e-05,
"loss": 0.9731,
"step": 1475
},
{
"epoch": 0.9651790093182933,
"grad_norm": 4.5,
"learning_rate": 1.9477586998783555e-05,
"loss": 0.4942,
"step": 1476
},
{
"epoch": 0.9658329246362596,
"grad_norm": 5.25,
"learning_rate": 1.9476129425200935e-05,
"loss": 0.4065,
"step": 1477
},
{
"epoch": 0.9664868399542259,
"grad_norm": 4.15625,
"learning_rate": 1.9474669875768893e-05,
"loss": 0.6536,
"step": 1478
},
{
"epoch": 0.9671407552721922,
"grad_norm": 3.796875,
"learning_rate": 1.947320835079175e-05,
"loss": 0.517,
"step": 1479
},
{
"epoch": 0.9677946705901586,
"grad_norm": 3.53125,
"learning_rate": 1.947174485057425e-05,
"loss": 0.4812,
"step": 1480
},
{
"epoch": 0.9684485859081249,
"grad_norm": 5.40625,
"learning_rate": 1.9470279375421545e-05,
"loss": 0.7359,
"step": 1481
},
{
"epoch": 0.9691025012260912,
"grad_norm": 4.75,
"learning_rate": 1.9468811925639196e-05,
"loss": 0.6756,
"step": 1482
},
{
"epoch": 0.9697564165440575,
"grad_norm": 4.15625,
"learning_rate": 1.9467342501533176e-05,
"loss": 0.4946,
"step": 1483
},
{
"epoch": 0.9704103318620239,
"grad_norm": 3.96875,
"learning_rate": 1.9465871103409875e-05,
"loss": 0.583,
"step": 1484
},
{
"epoch": 0.9710642471799902,
"grad_norm": 3.75,
"learning_rate": 1.9464397731576093e-05,
"loss": 0.6647,
"step": 1485
},
{
"epoch": 0.9717181624979565,
"grad_norm": 4.84375,
"learning_rate": 1.9462922386339042e-05,
"loss": 0.7904,
"step": 1486
},
{
"epoch": 0.9723720778159228,
"grad_norm": 4.3125,
"learning_rate": 1.9461445068006333e-05,
"loss": 0.537,
"step": 1487
},
{
"epoch": 0.9730259931338892,
"grad_norm": 3.671875,
"learning_rate": 1.945996577688601e-05,
"loss": 0.5162,
"step": 1488
},
{
"epoch": 0.9736799084518555,
"grad_norm": 4.28125,
"learning_rate": 1.945848451328651e-05,
"loss": 0.5582,
"step": 1489
},
{
"epoch": 0.9743338237698218,
"grad_norm": 4.90625,
"learning_rate": 1.945700127751669e-05,
"loss": 0.8845,
"step": 1490
},
{
"epoch": 0.9749877390877881,
"grad_norm": 4.4375,
"learning_rate": 1.945551606988582e-05,
"loss": 0.6739,
"step": 1491
},
{
"epoch": 0.9756416544057545,
"grad_norm": 4.625,
"learning_rate": 1.9454028890703574e-05,
"loss": 0.7845,
"step": 1492
},
{
"epoch": 0.9762955697237208,
"grad_norm": 4.75,
"learning_rate": 1.945253974028004e-05,
"loss": 0.647,
"step": 1493
},
{
"epoch": 0.9769494850416871,
"grad_norm": 3.359375,
"learning_rate": 1.9451048618925723e-05,
"loss": 0.4261,
"step": 1494
},
{
"epoch": 0.9776034003596534,
"grad_norm": 3.75,
"learning_rate": 1.944955552695153e-05,
"loss": 0.4878,
"step": 1495
},
{
"epoch": 0.9782573156776198,
"grad_norm": 4.8125,
"learning_rate": 1.944806046466878e-05,
"loss": 0.6081,
"step": 1496
},
{
"epoch": 0.9789112309955861,
"grad_norm": 5.21875,
"learning_rate": 1.9446563432389215e-05,
"loss": 0.8683,
"step": 1497
},
{
"epoch": 0.9795651463135524,
"grad_norm": 3.5,
"learning_rate": 1.944506443042497e-05,
"loss": 0.5905,
"step": 1498
},
{
"epoch": 0.9802190616315187,
"grad_norm": 6.25,
"learning_rate": 1.9443563459088598e-05,
"loss": 0.6727,
"step": 1499
},
{
"epoch": 0.980872976949485,
"grad_norm": 5.46875,
"learning_rate": 1.9442060518693072e-05,
"loss": 0.5534,
"step": 1500
},
{
"epoch": 0.9815268922674514,
"grad_norm": 2.90625,
"learning_rate": 1.9440555609551758e-05,
"loss": 0.3002,
"step": 1501
},
{
"epoch": 0.9821808075854177,
"grad_norm": 3.859375,
"learning_rate": 1.9439048731978445e-05,
"loss": 0.5486,
"step": 1502
},
{
"epoch": 0.982834722903384,
"grad_norm": 4.9375,
"learning_rate": 1.9437539886287332e-05,
"loss": 0.7604,
"step": 1503
},
{
"epoch": 0.9834886382213504,
"grad_norm": 4.78125,
"learning_rate": 1.9436029072793026e-05,
"loss": 0.4906,
"step": 1504
},
{
"epoch": 0.9841425535393167,
"grad_norm": 4.8125,
"learning_rate": 1.943451629181054e-05,
"loss": 0.6075,
"step": 1505
},
{
"epoch": 0.984796468857283,
"grad_norm": 3.921875,
"learning_rate": 1.94330015436553e-05,
"loss": 0.4458,
"step": 1506
},
{
"epoch": 0.9854503841752493,
"grad_norm": 7.6875,
"learning_rate": 1.943148482864315e-05,
"loss": 1.041,
"step": 1507
},
{
"epoch": 0.9861042994932157,
"grad_norm": 2.859375,
"learning_rate": 1.9429966147090326e-05,
"loss": 0.3697,
"step": 1508
},
{
"epoch": 0.986758214811182,
"grad_norm": 3.671875,
"learning_rate": 1.9428445499313496e-05,
"loss": 0.4429,
"step": 1509
},
{
"epoch": 0.9874121301291483,
"grad_norm": 5.3125,
"learning_rate": 1.9426922885629723e-05,
"loss": 0.8543,
"step": 1510
},
{
"epoch": 0.9880660454471146,
"grad_norm": 3.265625,
"learning_rate": 1.942539830635649e-05,
"loss": 0.5574,
"step": 1511
},
{
"epoch": 0.988719960765081,
"grad_norm": 3.859375,
"learning_rate": 1.942387176181167e-05,
"loss": 0.4815,
"step": 1512
},
{
"epoch": 0.9893738760830473,
"grad_norm": 4.625,
"learning_rate": 1.9422343252313574e-05,
"loss": 0.899,
"step": 1513
},
{
"epoch": 0.9900277914010136,
"grad_norm": 4.375,
"learning_rate": 1.9420812778180903e-05,
"loss": 0.7372,
"step": 1514
},
{
"epoch": 0.9906817067189799,
"grad_norm": 4.34375,
"learning_rate": 1.941928033973277e-05,
"loss": 0.6701,
"step": 1515
},
{
"epoch": 0.9913356220369463,
"grad_norm": 4.53125,
"learning_rate": 1.9417745937288707e-05,
"loss": 0.8026,
"step": 1516
},
{
"epoch": 0.9919895373549126,
"grad_norm": 3.546875,
"learning_rate": 1.9416209571168648e-05,
"loss": 0.684,
"step": 1517
},
{
"epoch": 0.9926434526728789,
"grad_norm": 5.0,
"learning_rate": 1.9414671241692936e-05,
"loss": 0.6559,
"step": 1518
},
{
"epoch": 0.9932973679908452,
"grad_norm": 4.28125,
"learning_rate": 1.9413130949182323e-05,
"loss": 0.3972,
"step": 1519
},
{
"epoch": 0.9939512833088116,
"grad_norm": 4.3125,
"learning_rate": 1.9411588693957976e-05,
"loss": 0.4497,
"step": 1520
},
{
"epoch": 0.9946051986267779,
"grad_norm": 3.40625,
"learning_rate": 1.9410044476341462e-05,
"loss": 0.4881,
"step": 1521
},
{
"epoch": 0.9952591139447442,
"grad_norm": 4.65625,
"learning_rate": 1.940849829665477e-05,
"loss": 0.4933,
"step": 1522
},
{
"epoch": 0.9959130292627105,
"grad_norm": 4.625,
"learning_rate": 1.940695015522029e-05,
"loss": 0.5533,
"step": 1523
},
{
"epoch": 0.9965669445806769,
"grad_norm": 4.25,
"learning_rate": 1.940540005236082e-05,
"loss": 0.8025,
"step": 1524
},
{
"epoch": 0.9972208598986432,
"grad_norm": 3.78125,
"learning_rate": 1.940384798839957e-05,
"loss": 0.5927,
"step": 1525
},
{
"epoch": 0.9978747752166095,
"grad_norm": 4.4375,
"learning_rate": 1.9402293963660155e-05,
"loss": 0.7094,
"step": 1526
},
{
"epoch": 0.9985286905345758,
"grad_norm": 4.53125,
"learning_rate": 1.9400737978466602e-05,
"loss": 0.4626,
"step": 1527
},
{
"epoch": 0.9991826058525421,
"grad_norm": 5.59375,
"learning_rate": 1.9399180033143353e-05,
"loss": 0.7961,
"step": 1528
},
{
"epoch": 0.9998365211705085,
"grad_norm": 5.625,
"learning_rate": 1.9397620128015244e-05,
"loss": 0.8528,
"step": 1529
},
{
"epoch": 1.0,
"grad_norm": 2.671875,
"learning_rate": 1.939605826340753e-05,
"loss": 0.0832,
"step": 1530
},
{
"epoch": 1.0006539153179663,
"grad_norm": 1.8359375,
"learning_rate": 1.9394494439645877e-05,
"loss": 0.1649,
"step": 1531
},
{
"epoch": 1.0013078306359326,
"grad_norm": 4.9375,
"learning_rate": 1.9392928657056348e-05,
"loss": 0.5409,
"step": 1532
},
{
"epoch": 1.001961745953899,
"grad_norm": 3.3125,
"learning_rate": 1.9391360915965426e-05,
"loss": 0.4503,
"step": 1533
},
{
"epoch": 1.0026156612718653,
"grad_norm": 3.265625,
"learning_rate": 1.9389791216699997e-05,
"loss": 0.4672,
"step": 1534
},
{
"epoch": 1.0032695765898316,
"grad_norm": 3.109375,
"learning_rate": 1.9388219559587353e-05,
"loss": 0.3329,
"step": 1535
},
{
"epoch": 1.003923491907798,
"grad_norm": 4.5625,
"learning_rate": 1.9386645944955202e-05,
"loss": 0.481,
"step": 1536
},
{
"epoch": 1.0045774072257643,
"grad_norm": 5.78125,
"learning_rate": 1.938507037313165e-05,
"loss": 0.6512,
"step": 1537
},
{
"epoch": 1.0052313225437306,
"grad_norm": 3.71875,
"learning_rate": 1.9383492844445224e-05,
"loss": 0.4758,
"step": 1538
},
{
"epoch": 1.005885237861697,
"grad_norm": 3.890625,
"learning_rate": 1.9381913359224844e-05,
"loss": 0.3795,
"step": 1539
},
{
"epoch": 1.0065391531796632,
"grad_norm": 3.71875,
"learning_rate": 1.9380331917799846e-05,
"loss": 0.3594,
"step": 1540
},
{
"epoch": 1.0071930684976296,
"grad_norm": 2.9375,
"learning_rate": 1.9378748520499977e-05,
"loss": 0.2911,
"step": 1541
},
{
"epoch": 1.007846983815596,
"grad_norm": 3.765625,
"learning_rate": 1.937716316765539e-05,
"loss": 0.3409,
"step": 1542
},
{
"epoch": 1.0085008991335622,
"grad_norm": 3.953125,
"learning_rate": 1.9375575859596634e-05,
"loss": 0.4217,
"step": 1543
},
{
"epoch": 1.0091548144515285,
"grad_norm": 3.734375,
"learning_rate": 1.9373986596654687e-05,
"loss": 0.3226,
"step": 1544
},
{
"epoch": 1.0098087297694949,
"grad_norm": 3.375,
"learning_rate": 1.9372395379160913e-05,
"loss": 0.4153,
"step": 1545
},
{
"epoch": 1.0104626450874612,
"grad_norm": 4.03125,
"learning_rate": 1.93708022074471e-05,
"loss": 0.2788,
"step": 1546
},
{
"epoch": 1.0111165604054275,
"grad_norm": 3.0625,
"learning_rate": 1.936920708184544e-05,
"loss": 0.2541,
"step": 1547
},
{
"epoch": 1.0117704757233938,
"grad_norm": 5.90625,
"learning_rate": 1.936761000268852e-05,
"loss": 0.8165,
"step": 1548
},
{
"epoch": 1.0124243910413602,
"grad_norm": 6.78125,
"learning_rate": 1.9366010970309355e-05,
"loss": 0.6086,
"step": 1549
},
{
"epoch": 1.0130783063593265,
"grad_norm": 4.65625,
"learning_rate": 1.936440998504135e-05,
"loss": 0.2726,
"step": 1550
},
{
"epoch": 1.0137322216772928,
"grad_norm": 6.09375,
"learning_rate": 1.936280704721832e-05,
"loss": 0.5245,
"step": 1551
},
{
"epoch": 1.0143861369952591,
"grad_norm": 4.875,
"learning_rate": 1.936120215717449e-05,
"loss": 0.7389,
"step": 1552
},
{
"epoch": 1.0150400523132255,
"grad_norm": 6.0625,
"learning_rate": 1.9359595315244507e-05,
"loss": 0.3956,
"step": 1553
},
{
"epoch": 1.0156939676311918,
"grad_norm": 7.8125,
"learning_rate": 1.9357986521763393e-05,
"loss": 0.7851,
"step": 1554
},
{
"epoch": 1.0163478829491581,
"grad_norm": 3.953125,
"learning_rate": 1.9356375777066604e-05,
"loss": 0.3608,
"step": 1555
},
{
"epoch": 1.0170017982671244,
"grad_norm": 5.09375,
"learning_rate": 1.9354763081489992e-05,
"loss": 0.5686,
"step": 1556
},
{
"epoch": 1.0176557135850908,
"grad_norm": 5.59375,
"learning_rate": 1.9353148435369816e-05,
"loss": 0.6191,
"step": 1557
},
{
"epoch": 1.018309628903057,
"grad_norm": 5.28125,
"learning_rate": 1.9351531839042744e-05,
"loss": 0.8533,
"step": 1558
},
{
"epoch": 1.0189635442210234,
"grad_norm": 4.03125,
"learning_rate": 1.9349913292845844e-05,
"loss": 0.4576,
"step": 1559
},
{
"epoch": 1.0196174595389897,
"grad_norm": 2.703125,
"learning_rate": 1.9348292797116605e-05,
"loss": 0.2327,
"step": 1560
},
{
"epoch": 1.020271374856956,
"grad_norm": 4.40625,
"learning_rate": 1.93466703521929e-05,
"loss": 0.4636,
"step": 1561
},
{
"epoch": 1.0209252901749224,
"grad_norm": 2.234375,
"learning_rate": 1.9345045958413038e-05,
"loss": 0.1231,
"step": 1562
},
{
"epoch": 1.0215792054928887,
"grad_norm": 3.953125,
"learning_rate": 1.934341961611571e-05,
"loss": 0.5567,
"step": 1563
},
{
"epoch": 1.022233120810855,
"grad_norm": 3.3125,
"learning_rate": 1.934179132564002e-05,
"loss": 0.2968,
"step": 1564
},
{
"epoch": 1.0228870361288214,
"grad_norm": 3.015625,
"learning_rate": 1.9340161087325483e-05,
"loss": 0.2862,
"step": 1565
},
{
"epoch": 1.0235409514467877,
"grad_norm": 3.828125,
"learning_rate": 1.9338528901512014e-05,
"loss": 0.2953,
"step": 1566
},
{
"epoch": 1.024194866764754,
"grad_norm": 3.375,
"learning_rate": 1.933689476853994e-05,
"loss": 0.4599,
"step": 1567
},
{
"epoch": 1.0248487820827203,
"grad_norm": 5.4375,
"learning_rate": 1.9335258688749987e-05,
"loss": 0.4225,
"step": 1568
},
{
"epoch": 1.0255026974006867,
"grad_norm": 3.375,
"learning_rate": 1.9333620662483295e-05,
"loss": 0.35,
"step": 1569
},
{
"epoch": 1.026156612718653,
"grad_norm": 4.75,
"learning_rate": 1.9331980690081404e-05,
"loss": 0.603,
"step": 1570
},
{
"epoch": 1.0268105280366193,
"grad_norm": 4.0625,
"learning_rate": 1.9330338771886262e-05,
"loss": 0.4836,
"step": 1571
},
{
"epoch": 1.0274644433545856,
"grad_norm": 4.65625,
"learning_rate": 1.932869490824022e-05,
"loss": 0.5231,
"step": 1572
},
{
"epoch": 1.028118358672552,
"grad_norm": 3.890625,
"learning_rate": 1.932704909948604e-05,
"loss": 0.4428,
"step": 1573
},
{
"epoch": 1.0287722739905183,
"grad_norm": 3.390625,
"learning_rate": 1.9325401345966883e-05,
"loss": 0.4017,
"step": 1574
},
{
"epoch": 1.0294261893084846,
"grad_norm": 5.65625,
"learning_rate": 1.932375164802632e-05,
"loss": 0.5376,
"step": 1575
},
{
"epoch": 1.030080104626451,
"grad_norm": 5.0,
"learning_rate": 1.932210000600833e-05,
"loss": 0.5641,
"step": 1576
},
{
"epoch": 1.0307340199444173,
"grad_norm": 5.5,
"learning_rate": 1.9320446420257292e-05,
"loss": 0.6356,
"step": 1577
},
{
"epoch": 1.0313879352623836,
"grad_norm": 3.734375,
"learning_rate": 1.9318790891117985e-05,
"loss": 0.3179,
"step": 1578
},
{
"epoch": 1.03204185058035,
"grad_norm": 4.0,
"learning_rate": 1.931713341893561e-05,
"loss": 0.2323,
"step": 1579
},
{
"epoch": 1.0326957658983162,
"grad_norm": 3.859375,
"learning_rate": 1.9315474004055757e-05,
"loss": 0.383,
"step": 1580
},
{
"epoch": 1.0333496812162826,
"grad_norm": 4.84375,
"learning_rate": 1.9313812646824432e-05,
"loss": 0.3024,
"step": 1581
},
{
"epoch": 1.034003596534249,
"grad_norm": 3.421875,
"learning_rate": 1.9312149347588035e-05,
"loss": 0.376,
"step": 1582
},
{
"epoch": 1.0346575118522152,
"grad_norm": 3.296875,
"learning_rate": 1.9310484106693385e-05,
"loss": 0.2676,
"step": 1583
},
{
"epoch": 1.0353114271701815,
"grad_norm": 4.0,
"learning_rate": 1.9308816924487694e-05,
"loss": 0.453,
"step": 1584
},
{
"epoch": 1.0359653424881479,
"grad_norm": 5.625,
"learning_rate": 1.9307147801318585e-05,
"loss": 0.5575,
"step": 1585
},
{
"epoch": 1.0366192578061142,
"grad_norm": 4.5625,
"learning_rate": 1.9305476737534075e-05,
"loss": 0.4973,
"step": 1586
},
{
"epoch": 1.0372731731240805,
"grad_norm": 3.421875,
"learning_rate": 1.930380373348261e-05,
"loss": 0.2506,
"step": 1587
},
{
"epoch": 1.0379270884420468,
"grad_norm": 5.625,
"learning_rate": 1.9302128789513013e-05,
"loss": 0.5172,
"step": 1588
},
{
"epoch": 1.0385810037600132,
"grad_norm": 5.125,
"learning_rate": 1.9300451905974523e-05,
"loss": 0.636,
"step": 1589
},
{
"epoch": 1.0392349190779795,
"grad_norm": 4.09375,
"learning_rate": 1.929877308321679e-05,
"loss": 0.6321,
"step": 1590
},
{
"epoch": 1.0398888343959458,
"grad_norm": 4.0625,
"learning_rate": 1.929709232158986e-05,
"loss": 0.3245,
"step": 1591
},
{
"epoch": 1.0405427497139121,
"grad_norm": 3.59375,
"learning_rate": 1.9295409621444183e-05,
"loss": 0.2675,
"step": 1592
},
{
"epoch": 1.0411966650318785,
"grad_norm": 5.59375,
"learning_rate": 1.929372498313062e-05,
"loss": 0.6093,
"step": 1593
},
{
"epoch": 1.0418505803498448,
"grad_norm": 5.40625,
"learning_rate": 1.9292038407000424e-05,
"loss": 0.374,
"step": 1594
},
{
"epoch": 1.0425044956678111,
"grad_norm": 3.1875,
"learning_rate": 1.9290349893405268e-05,
"loss": 0.3346,
"step": 1595
},
{
"epoch": 1.0431584109857774,
"grad_norm": 5.8125,
"learning_rate": 1.9288659442697218e-05,
"loss": 0.5781,
"step": 1596
},
{
"epoch": 1.0438123263037438,
"grad_norm": 3.84375,
"learning_rate": 1.9286967055228744e-05,
"loss": 0.341,
"step": 1597
},
{
"epoch": 1.04446624162171,
"grad_norm": 5.875,
"learning_rate": 1.9285272731352726e-05,
"loss": 0.4546,
"step": 1598
},
{
"epoch": 1.0451201569396764,
"grad_norm": 4.21875,
"learning_rate": 1.9283576471422442e-05,
"loss": 0.3662,
"step": 1599
},
{
"epoch": 1.0457740722576427,
"grad_norm": 3.8125,
"learning_rate": 1.9281878275791575e-05,
"loss": 0.2589,
"step": 1600
},
{
"epoch": 1.046427987575609,
"grad_norm": 3.796875,
"learning_rate": 1.9280178144814213e-05,
"loss": 0.4266,
"step": 1601
},
{
"epoch": 1.0470819028935754,
"grad_norm": 4.40625,
"learning_rate": 1.927847607884485e-05,
"loss": 0.3815,
"step": 1602
},
{
"epoch": 1.0477358182115415,
"grad_norm": 4.125,
"learning_rate": 1.927677207823838e-05,
"loss": 0.3813,
"step": 1603
},
{
"epoch": 1.048389733529508,
"grad_norm": 4.75,
"learning_rate": 1.9275066143350096e-05,
"loss": 0.441,
"step": 1604
},
{
"epoch": 1.0490436488474741,
"grad_norm": 5.5625,
"learning_rate": 1.9273358274535703e-05,
"loss": 0.4194,
"step": 1605
},
{
"epoch": 1.0496975641654407,
"grad_norm": 3.71875,
"learning_rate": 1.9271648472151308e-05,
"loss": 0.519,
"step": 1606
},
{
"epoch": 1.0503514794834068,
"grad_norm": 3.453125,
"learning_rate": 1.9269936736553415e-05,
"loss": 0.4147,
"step": 1607
},
{
"epoch": 1.0510053948013731,
"grad_norm": 3.21875,
"learning_rate": 1.926822306809893e-05,
"loss": 0.2866,
"step": 1608
},
{
"epoch": 1.0516593101193394,
"grad_norm": 3.515625,
"learning_rate": 1.9266507467145177e-05,
"loss": 0.3438,
"step": 1609
},
{
"epoch": 1.0523132254373058,
"grad_norm": 4.8125,
"learning_rate": 1.9264789934049867e-05,
"loss": 0.4355,
"step": 1610
},
{
"epoch": 1.052967140755272,
"grad_norm": 3.390625,
"learning_rate": 1.9263070469171116e-05,
"loss": 0.3278,
"step": 1611
},
{
"epoch": 1.0536210560732384,
"grad_norm": 3.71875,
"learning_rate": 1.9261349072867458e-05,
"loss": 0.2832,
"step": 1612
},
{
"epoch": 1.0542749713912047,
"grad_norm": 5.0625,
"learning_rate": 1.9259625745497803e-05,
"loss": 0.4408,
"step": 1613
},
{
"epoch": 1.054928886709171,
"grad_norm": 4.78125,
"learning_rate": 1.925790048742149e-05,
"loss": 0.5269,
"step": 1614
},
{
"epoch": 1.0555828020271374,
"grad_norm": 4.3125,
"learning_rate": 1.9256173298998243e-05,
"loss": 0.5133,
"step": 1615
},
{
"epoch": 1.0562367173451037,
"grad_norm": 4.0,
"learning_rate": 1.92544441805882e-05,
"loss": 0.379,
"step": 1616
},
{
"epoch": 1.05689063266307,
"grad_norm": 3.953125,
"learning_rate": 1.925271313255189e-05,
"loss": 0.2888,
"step": 1617
},
{
"epoch": 1.0575445479810364,
"grad_norm": 4.5625,
"learning_rate": 1.925098015525025e-05,
"loss": 0.4633,
"step": 1618
},
{
"epoch": 1.0581984632990027,
"grad_norm": 4.21875,
"learning_rate": 1.924924524904463e-05,
"loss": 0.4209,
"step": 1619
},
{
"epoch": 1.058852378616969,
"grad_norm": 4.3125,
"learning_rate": 1.9247508414296764e-05,
"loss": 0.4135,
"step": 1620
},
{
"epoch": 1.0595062939349353,
"grad_norm": 4.5,
"learning_rate": 1.9245769651368794e-05,
"loss": 0.5609,
"step": 1621
},
{
"epoch": 1.0601602092529017,
"grad_norm": 3.796875,
"learning_rate": 1.9244028960623275e-05,
"loss": 0.4483,
"step": 1622
},
{
"epoch": 1.060814124570868,
"grad_norm": 3.0625,
"learning_rate": 1.9242286342423147e-05,
"loss": 0.3167,
"step": 1623
},
{
"epoch": 1.0614680398888343,
"grad_norm": 4.8125,
"learning_rate": 1.9240541797131758e-05,
"loss": 0.4909,
"step": 1624
},
{
"epoch": 1.0621219552068006,
"grad_norm": 4.34375,
"learning_rate": 1.9238795325112867e-05,
"loss": 0.5859,
"step": 1625
},
{
"epoch": 1.062775870524767,
"grad_norm": 2.671875,
"learning_rate": 1.9237046926730628e-05,
"loss": 0.311,
"step": 1626
},
{
"epoch": 1.0634297858427333,
"grad_norm": 4.71875,
"learning_rate": 1.9235296602349592e-05,
"loss": 0.4103,
"step": 1627
},
{
"epoch": 1.0640837011606996,
"grad_norm": 3.140625,
"learning_rate": 1.9233544352334713e-05,
"loss": 0.2443,
"step": 1628
},
{
"epoch": 1.064737616478666,
"grad_norm": 3.75,
"learning_rate": 1.9231790177051354e-05,
"loss": 0.3833,
"step": 1629
},
{
"epoch": 1.0653915317966323,
"grad_norm": 4.34375,
"learning_rate": 1.9230034076865272e-05,
"loss": 0.3941,
"step": 1630
},
{
"epoch": 1.0660454471145986,
"grad_norm": 4.0,
"learning_rate": 1.9228276052142635e-05,
"loss": 0.5496,
"step": 1631
},
{
"epoch": 1.066699362432565,
"grad_norm": 3.734375,
"learning_rate": 1.9226516103249996e-05,
"loss": 0.4533,
"step": 1632
},
{
"epoch": 1.0673532777505312,
"grad_norm": 3.296875,
"learning_rate": 1.9224754230554325e-05,
"loss": 0.3128,
"step": 1633
},
{
"epoch": 1.0680071930684976,
"grad_norm": 3.9375,
"learning_rate": 1.922299043442298e-05,
"loss": 0.1853,
"step": 1634
},
{
"epoch": 1.0686611083864639,
"grad_norm": 4.25,
"learning_rate": 1.9221224715223735e-05,
"loss": 0.3258,
"step": 1635
},
{
"epoch": 1.0693150237044302,
"grad_norm": 4.125,
"learning_rate": 1.921945707332475e-05,
"loss": 0.3446,
"step": 1636
},
{
"epoch": 1.0699689390223965,
"grad_norm": 4.59375,
"learning_rate": 1.9217687509094596e-05,
"loss": 0.4845,
"step": 1637
},
{
"epoch": 1.0706228543403629,
"grad_norm": 3.8125,
"learning_rate": 1.921591602290224e-05,
"loss": 0.4086,
"step": 1638
},
{
"epoch": 1.0712767696583292,
"grad_norm": 3.859375,
"learning_rate": 1.9214142615117052e-05,
"loss": 0.2485,
"step": 1639
},
{
"epoch": 1.0719306849762955,
"grad_norm": 4.4375,
"learning_rate": 1.9212367286108802e-05,
"loss": 0.2616,
"step": 1640
},
{
"epoch": 1.0725846002942618,
"grad_norm": 4.15625,
"learning_rate": 1.921059003624766e-05,
"loss": 0.2743,
"step": 1641
},
{
"epoch": 1.0732385156122282,
"grad_norm": 4.875,
"learning_rate": 1.9208810865904196e-05,
"loss": 0.4426,
"step": 1642
},
{
"epoch": 1.0738924309301945,
"grad_norm": 5.21875,
"learning_rate": 1.9207029775449383e-05,
"loss": 0.6329,
"step": 1643
},
{
"epoch": 1.0745463462481608,
"grad_norm": 4.15625,
"learning_rate": 1.9205246765254592e-05,
"loss": 0.3414,
"step": 1644
},
{
"epoch": 1.0752002615661271,
"grad_norm": 3.90625,
"learning_rate": 1.9203461835691596e-05,
"loss": 0.3469,
"step": 1645
},
{
"epoch": 1.0758541768840935,
"grad_norm": 4.03125,
"learning_rate": 1.9201674987132564e-05,
"loss": 0.3712,
"step": 1646
},
{
"epoch": 1.0765080922020598,
"grad_norm": 5.15625,
"learning_rate": 1.9199886219950073e-05,
"loss": 0.5181,
"step": 1647
},
{
"epoch": 1.077162007520026,
"grad_norm": 3.25,
"learning_rate": 1.9198095534517093e-05,
"loss": 0.4119,
"step": 1648
},
{
"epoch": 1.0778159228379924,
"grad_norm": 5.0625,
"learning_rate": 1.9196302931206995e-05,
"loss": 0.6093,
"step": 1649
},
{
"epoch": 1.0784698381559588,
"grad_norm": 4.125,
"learning_rate": 1.919450841039355e-05,
"loss": 0.3899,
"step": 1650
},
{
"epoch": 1.079123753473925,
"grad_norm": 5.25,
"learning_rate": 1.919271197245094e-05,
"loss": 0.8004,
"step": 1651
},
{
"epoch": 1.0797776687918914,
"grad_norm": 3.5625,
"learning_rate": 1.9190913617753724e-05,
"loss": 0.4489,
"step": 1652
},
{
"epoch": 1.0804315841098577,
"grad_norm": 6.09375,
"learning_rate": 1.9189113346676878e-05,
"loss": 0.9255,
"step": 1653
},
{
"epoch": 1.081085499427824,
"grad_norm": 3.09375,
"learning_rate": 1.9187311159595775e-05,
"loss": 0.2869,
"step": 1654
},
{
"epoch": 1.0817394147457904,
"grad_norm": 4.9375,
"learning_rate": 1.9185507056886184e-05,
"loss": 0.3752,
"step": 1655
},
{
"epoch": 1.0823933300637567,
"grad_norm": 4.5625,
"learning_rate": 1.918370103892428e-05,
"loss": 0.4958,
"step": 1656
},
{
"epoch": 1.083047245381723,
"grad_norm": 3.578125,
"learning_rate": 1.9181893106086626e-05,
"loss": 0.3923,
"step": 1657
},
{
"epoch": 1.0837011606996894,
"grad_norm": 4.65625,
"learning_rate": 1.9180083258750188e-05,
"loss": 0.6374,
"step": 1658
},
{
"epoch": 1.0843550760176557,
"grad_norm": 4.34375,
"learning_rate": 1.917827149729234e-05,
"loss": 0.3447,
"step": 1659
},
{
"epoch": 1.085008991335622,
"grad_norm": 4.78125,
"learning_rate": 1.9176457822090847e-05,
"loss": 0.379,
"step": 1660
},
{
"epoch": 1.0856629066535883,
"grad_norm": 7.6875,
"learning_rate": 1.9174642233523876e-05,
"loss": 0.8173,
"step": 1661
},
{
"epoch": 1.0863168219715547,
"grad_norm": 3.296875,
"learning_rate": 1.9172824731969992e-05,
"loss": 0.3723,
"step": 1662
},
{
"epoch": 1.086970737289521,
"grad_norm": 4.21875,
"learning_rate": 1.9171005317808155e-05,
"loss": 0.4662,
"step": 1663
},
{
"epoch": 1.0876246526074873,
"grad_norm": 5.6875,
"learning_rate": 1.9169183991417732e-05,
"loss": 0.4691,
"step": 1664
},
{
"epoch": 1.0882785679254536,
"grad_norm": 3.96875,
"learning_rate": 1.916736075317848e-05,
"loss": 0.4088,
"step": 1665
},
{
"epoch": 1.08893248324342,
"grad_norm": 5.21875,
"learning_rate": 1.9165535603470565e-05,
"loss": 0.4255,
"step": 1666
},
{
"epoch": 1.0895863985613863,
"grad_norm": 3.640625,
"learning_rate": 1.916370854267454e-05,
"loss": 0.3258,
"step": 1667
},
{
"epoch": 1.0902403138793526,
"grad_norm": 5.21875,
"learning_rate": 1.916187957117136e-05,
"loss": 0.639,
"step": 1668
},
{
"epoch": 1.090894229197319,
"grad_norm": 4.59375,
"learning_rate": 1.9160048689342386e-05,
"loss": 0.6161,
"step": 1669
},
{
"epoch": 1.0915481445152853,
"grad_norm": 9.5,
"learning_rate": 1.915821589756937e-05,
"loss": 1.3438,
"step": 1670
},
{
"epoch": 1.0922020598332516,
"grad_norm": 4.53125,
"learning_rate": 1.9156381196234464e-05,
"loss": 0.4627,
"step": 1671
},
{
"epoch": 1.092855975151218,
"grad_norm": 5.03125,
"learning_rate": 1.9154544585720218e-05,
"loss": 0.5367,
"step": 1672
},
{
"epoch": 1.0935098904691842,
"grad_norm": 4.375,
"learning_rate": 1.9152706066409575e-05,
"loss": 0.55,
"step": 1673
},
{
"epoch": 1.0941638057871506,
"grad_norm": 4.15625,
"learning_rate": 1.915086563868589e-05,
"loss": 0.4368,
"step": 1674
},
{
"epoch": 1.0948177211051169,
"grad_norm": 5.90625,
"learning_rate": 1.91490233029329e-05,
"loss": 0.6335,
"step": 1675
},
{
"epoch": 1.0954716364230832,
"grad_norm": 4.46875,
"learning_rate": 1.9147179059534752e-05,
"loss": 0.431,
"step": 1676
},
{
"epoch": 1.0961255517410495,
"grad_norm": 4.21875,
"learning_rate": 1.9145332908875984e-05,
"loss": 0.3023,
"step": 1677
},
{
"epoch": 1.0967794670590159,
"grad_norm": 5.75,
"learning_rate": 1.914348485134153e-05,
"loss": 0.4379,
"step": 1678
},
{
"epoch": 1.0974333823769822,
"grad_norm": 3.125,
"learning_rate": 1.9141634887316726e-05,
"loss": 0.3477,
"step": 1679
},
{
"epoch": 1.0980872976949485,
"grad_norm": 4.25,
"learning_rate": 1.9139783017187308e-05,
"loss": 0.5346,
"step": 1680
},
{
"epoch": 1.0987412130129148,
"grad_norm": 3.890625,
"learning_rate": 1.91379292413394e-05,
"loss": 0.4935,
"step": 1681
},
{
"epoch": 1.0993951283308812,
"grad_norm": 5.40625,
"learning_rate": 1.9136073560159535e-05,
"loss": 0.3976,
"step": 1682
},
{
"epoch": 1.1000490436488475,
"grad_norm": 5.1875,
"learning_rate": 1.9134215974034636e-05,
"loss": 0.7272,
"step": 1683
},
{
"epoch": 1.1007029589668138,
"grad_norm": 5.5625,
"learning_rate": 1.9132356483352022e-05,
"loss": 0.863,
"step": 1684
},
{
"epoch": 1.1013568742847801,
"grad_norm": 3.953125,
"learning_rate": 1.9130495088499417e-05,
"loss": 0.2939,
"step": 1685
},
{
"epoch": 1.1020107896027465,
"grad_norm": 4.53125,
"learning_rate": 1.9128631789864932e-05,
"loss": 0.3124,
"step": 1686
},
{
"epoch": 1.1026647049207128,
"grad_norm": 3.390625,
"learning_rate": 1.9126766587837086e-05,
"loss": 0.1571,
"step": 1687
},
{
"epoch": 1.103318620238679,
"grad_norm": 5.84375,
"learning_rate": 1.9124899482804776e-05,
"loss": 0.5439,
"step": 1688
},
{
"epoch": 1.1039725355566454,
"grad_norm": 5.0625,
"learning_rate": 1.912303047515732e-05,
"loss": 0.4246,
"step": 1689
},
{
"epoch": 1.1046264508746118,
"grad_norm": 4.375,
"learning_rate": 1.912115956528442e-05,
"loss": 0.3892,
"step": 1690
},
{
"epoch": 1.105280366192578,
"grad_norm": 3.875,
"learning_rate": 1.911928675357617e-05,
"loss": 0.4029,
"step": 1691
},
{
"epoch": 1.1059342815105444,
"grad_norm": 6.15625,
"learning_rate": 1.9117412040423072e-05,
"loss": 0.5176,
"step": 1692
},
{
"epoch": 1.1065881968285107,
"grad_norm": 3.34375,
"learning_rate": 1.9115535426216018e-05,
"loss": 0.2995,
"step": 1693
},
{
"epoch": 1.107242112146477,
"grad_norm": 4.90625,
"learning_rate": 1.911365691134629e-05,
"loss": 0.4247,
"step": 1694
},
{
"epoch": 1.1078960274644434,
"grad_norm": 4.34375,
"learning_rate": 1.911177649620558e-05,
"loss": 0.4234,
"step": 1695
},
{
"epoch": 1.1085499427824097,
"grad_norm": 3.640625,
"learning_rate": 1.9109894181185973e-05,
"loss": 0.3691,
"step": 1696
},
{
"epoch": 1.109203858100376,
"grad_norm": 4.21875,
"learning_rate": 1.9108009966679944e-05,
"loss": 0.3509,
"step": 1697
},
{
"epoch": 1.1098577734183424,
"grad_norm": 6.15625,
"learning_rate": 1.9106123853080366e-05,
"loss": 0.5001,
"step": 1698
},
{
"epoch": 1.1105116887363087,
"grad_norm": 6.53125,
"learning_rate": 1.9104235840780507e-05,
"loss": 0.6305,
"step": 1699
},
{
"epoch": 1.111165604054275,
"grad_norm": 5.125,
"learning_rate": 1.9102345930174034e-05,
"loss": 0.4128,
"step": 1700
},
{
"epoch": 1.1118195193722413,
"grad_norm": 4.59375,
"learning_rate": 1.910045412165501e-05,
"loss": 0.4622,
"step": 1701
},
{
"epoch": 1.1124734346902077,
"grad_norm": 2.34375,
"learning_rate": 1.9098560415617892e-05,
"loss": 0.1521,
"step": 1702
},
{
"epoch": 1.113127350008174,
"grad_norm": 6.125,
"learning_rate": 1.909666481245753e-05,
"loss": 0.4579,
"step": 1703
},
{
"epoch": 1.1137812653261403,
"grad_norm": 4.90625,
"learning_rate": 1.909476731256918e-05,
"loss": 0.4009,
"step": 1704
},
{
"epoch": 1.1144351806441066,
"grad_norm": 5.0625,
"learning_rate": 1.909286791634848e-05,
"loss": 0.7575,
"step": 1705
},
{
"epoch": 1.115089095962073,
"grad_norm": 4.8125,
"learning_rate": 1.9090966624191468e-05,
"loss": 0.6515,
"step": 1706
},
{
"epoch": 1.1157430112800393,
"grad_norm": 5.71875,
"learning_rate": 1.908906343649458e-05,
"loss": 0.4333,
"step": 1707
},
{
"epoch": 1.1163969265980056,
"grad_norm": 3.46875,
"learning_rate": 1.908715835365465e-05,
"loss": 0.2782,
"step": 1708
},
{
"epoch": 1.117050841915972,
"grad_norm": 3.328125,
"learning_rate": 1.9085251376068898e-05,
"loss": 0.317,
"step": 1709
},
{
"epoch": 1.1177047572339383,
"grad_norm": 4.6875,
"learning_rate": 1.908334250413495e-05,
"loss": 0.4316,
"step": 1710
},
{
"epoch": 1.1183586725519046,
"grad_norm": 4.59375,
"learning_rate": 1.9081431738250815e-05,
"loss": 0.429,
"step": 1711
},
{
"epoch": 1.119012587869871,
"grad_norm": 5.75,
"learning_rate": 1.907951907881491e-05,
"loss": 0.4571,
"step": 1712
},
{
"epoch": 1.1196665031878372,
"grad_norm": 4.0,
"learning_rate": 1.907760452622603e-05,
"loss": 0.502,
"step": 1713
},
{
"epoch": 1.1203204185058036,
"grad_norm": 4.34375,
"learning_rate": 1.907568808088338e-05,
"loss": 0.3742,
"step": 1714
},
{
"epoch": 1.1209743338237699,
"grad_norm": 4.8125,
"learning_rate": 1.907376974318656e-05,
"loss": 0.3999,
"step": 1715
},
{
"epoch": 1.1216282491417362,
"grad_norm": 5.5,
"learning_rate": 1.907184951353555e-05,
"loss": 0.7605,
"step": 1716
},
{
"epoch": 1.1222821644597025,
"grad_norm": 5.84375,
"learning_rate": 1.9069927392330742e-05,
"loss": 0.6978,
"step": 1717
},
{
"epoch": 1.1229360797776688,
"grad_norm": 5.8125,
"learning_rate": 1.9068003379972904e-05,
"loss": 0.6585,
"step": 1718
},
{
"epoch": 1.1235899950956352,
"grad_norm": 6.375,
"learning_rate": 1.9066077476863216e-05,
"loss": 0.6394,
"step": 1719
},
{
"epoch": 1.1242439104136015,
"grad_norm": 4.8125,
"learning_rate": 1.9064149683403238e-05,
"loss": 0.4759,
"step": 1720
},
{
"epoch": 1.1248978257315678,
"grad_norm": 2.53125,
"learning_rate": 1.906221999999494e-05,
"loss": 0.2171,
"step": 1721
},
{
"epoch": 1.1255517410495341,
"grad_norm": 4.1875,
"learning_rate": 1.9060288427040664e-05,
"loss": 0.4865,
"step": 1722
},
{
"epoch": 1.1262056563675005,
"grad_norm": 2.90625,
"learning_rate": 1.9058354964943167e-05,
"loss": 0.3124,
"step": 1723
},
{
"epoch": 1.1268595716854668,
"grad_norm": 4.03125,
"learning_rate": 1.905641961410559e-05,
"loss": 0.4099,
"step": 1724
},
{
"epoch": 1.1275134870034331,
"grad_norm": 6.84375,
"learning_rate": 1.905448237493147e-05,
"loss": 0.7502,
"step": 1725
},
{
"epoch": 1.1281674023213994,
"grad_norm": 4.875,
"learning_rate": 1.9052543247824732e-05,
"loss": 0.4905,
"step": 1726
},
{
"epoch": 1.1288213176393658,
"grad_norm": 3.46875,
"learning_rate": 1.9050602233189704e-05,
"loss": 0.3497,
"step": 1727
},
{
"epoch": 1.129475232957332,
"grad_norm": 4.0625,
"learning_rate": 1.9048659331431104e-05,
"loss": 0.2488,
"step": 1728
},
{
"epoch": 1.1301291482752984,
"grad_norm": 5.96875,
"learning_rate": 1.904671454295404e-05,
"loss": 0.6958,
"step": 1729
},
{
"epoch": 1.1307830635932647,
"grad_norm": 4.25,
"learning_rate": 1.9044767868164018e-05,
"loss": 0.2931,
"step": 1730
},
{
"epoch": 1.131436978911231,
"grad_norm": 6.53125,
"learning_rate": 1.9042819307466934e-05,
"loss": 0.8022,
"step": 1731
},
{
"epoch": 1.1320908942291974,
"grad_norm": 3.953125,
"learning_rate": 1.904086886126908e-05,
"loss": 0.3343,
"step": 1732
},
{
"epoch": 1.1327448095471637,
"grad_norm": 4.6875,
"learning_rate": 1.9038916529977136e-05,
"loss": 0.4148,
"step": 1733
},
{
"epoch": 1.13339872486513,
"grad_norm": 4.375,
"learning_rate": 1.9036962313998183e-05,
"loss": 0.3113,
"step": 1734
},
{
"epoch": 1.1340526401830964,
"grad_norm": 4.3125,
"learning_rate": 1.903500621373969e-05,
"loss": 0.4413,
"step": 1735
},
{
"epoch": 1.1347065555010627,
"grad_norm": 4.5625,
"learning_rate": 1.9033048229609524e-05,
"loss": 0.3487,
"step": 1736
},
{
"epoch": 1.135360470819029,
"grad_norm": 3.765625,
"learning_rate": 1.903108836201593e-05,
"loss": 0.2774,
"step": 1737
},
{
"epoch": 1.1360143861369953,
"grad_norm": 3.671875,
"learning_rate": 1.9029126611367563e-05,
"loss": 0.3175,
"step": 1738
},
{
"epoch": 1.1366683014549617,
"grad_norm": 4.03125,
"learning_rate": 1.9027162978073463e-05,
"loss": 0.4118,
"step": 1739
},
{
"epoch": 1.137322216772928,
"grad_norm": 3.875,
"learning_rate": 1.9025197462543063e-05,
"loss": 0.2846,
"step": 1740
},
{
"epoch": 1.1379761320908943,
"grad_norm": 3.703125,
"learning_rate": 1.9023230065186192e-05,
"loss": 0.3581,
"step": 1741
},
{
"epoch": 1.1386300474088606,
"grad_norm": 3.28125,
"learning_rate": 1.902126078641306e-05,
"loss": 0.2813,
"step": 1742
},
{
"epoch": 1.139283962726827,
"grad_norm": 4.03125,
"learning_rate": 1.9019289626634286e-05,
"loss": 0.3153,
"step": 1743
},
{
"epoch": 1.1399378780447933,
"grad_norm": 4.9375,
"learning_rate": 1.901731658626087e-05,
"loss": 0.4182,
"step": 1744
},
{
"epoch": 1.1405917933627596,
"grad_norm": 3.046875,
"learning_rate": 1.9015341665704206e-05,
"loss": 0.2581,
"step": 1745
},
{
"epoch": 1.1412457086807257,
"grad_norm": 4.6875,
"learning_rate": 1.9013364865376084e-05,
"loss": 0.4289,
"step": 1746
},
{
"epoch": 1.1418996239986923,
"grad_norm": 5.0,
"learning_rate": 1.9011386185688674e-05,
"loss": 0.4027,
"step": 1747
},
{
"epoch": 1.1425535393166584,
"grad_norm": 5.0625,
"learning_rate": 1.9009405627054556e-05,
"loss": 0.5724,
"step": 1748
},
{
"epoch": 1.143207454634625,
"grad_norm": 4.90625,
"learning_rate": 1.9007423189886687e-05,
"loss": 0.3581,
"step": 1749
},
{
"epoch": 1.143861369952591,
"grad_norm": 3.703125,
"learning_rate": 1.9005438874598425e-05,
"loss": 0.2439,
"step": 1750
},
{
"epoch": 1.1445152852705576,
"grad_norm": 5.6875,
"learning_rate": 1.9003452681603517e-05,
"loss": 0.661,
"step": 1751
},
{
"epoch": 1.1451692005885237,
"grad_norm": 4.1875,
"learning_rate": 1.900146461131609e-05,
"loss": 0.3718,
"step": 1752
},
{
"epoch": 1.1458231159064902,
"grad_norm": 3.78125,
"learning_rate": 1.8999474664150686e-05,
"loss": 0.2928,
"step": 1753
},
{
"epoch": 1.1464770312244563,
"grad_norm": 4.15625,
"learning_rate": 1.8997482840522218e-05,
"loss": 0.2709,
"step": 1754
},
{
"epoch": 1.1471309465424229,
"grad_norm": 4.40625,
"learning_rate": 1.8995489140845995e-05,
"loss": 0.4094,
"step": 1755
},
{
"epoch": 1.147784861860389,
"grad_norm": 4.46875,
"learning_rate": 1.8993493565537727e-05,
"loss": 0.2629,
"step": 1756
},
{
"epoch": 1.1484387771783555,
"grad_norm": 3.703125,
"learning_rate": 1.89914961150135e-05,
"loss": 0.4529,
"step": 1757
},
{
"epoch": 1.1490926924963216,
"grad_norm": 3.015625,
"learning_rate": 1.8989496789689803e-05,
"loss": 0.3615,
"step": 1758
},
{
"epoch": 1.1497466078142882,
"grad_norm": 3.8125,
"learning_rate": 1.898749558998351e-05,
"loss": 0.3419,
"step": 1759
},
{
"epoch": 1.1504005231322543,
"grad_norm": 6.125,
"learning_rate": 1.898549251631189e-05,
"loss": 0.7083,
"step": 1760
},
{
"epoch": 1.1510544384502208,
"grad_norm": 3.9375,
"learning_rate": 1.8983487569092594e-05,
"loss": 0.3401,
"step": 1761
},
{
"epoch": 1.151708353768187,
"grad_norm": 4.21875,
"learning_rate": 1.8981480748743674e-05,
"loss": 0.4549,
"step": 1762
},
{
"epoch": 1.1523622690861535,
"grad_norm": 6.625,
"learning_rate": 1.897947205568357e-05,
"loss": 0.8424,
"step": 1763
},
{
"epoch": 1.1530161844041196,
"grad_norm": 2.546875,
"learning_rate": 1.8977461490331106e-05,
"loss": 0.1166,
"step": 1764
},
{
"epoch": 1.153670099722086,
"grad_norm": 6.15625,
"learning_rate": 1.8975449053105505e-05,
"loss": 0.7031,
"step": 1765
},
{
"epoch": 1.1543240150400522,
"grad_norm": 5.875,
"learning_rate": 1.8973434744426375e-05,
"loss": 0.5132,
"step": 1766
},
{
"epoch": 1.1549779303580185,
"grad_norm": 4.8125,
"learning_rate": 1.8971418564713714e-05,
"loss": 0.4821,
"step": 1767
},
{
"epoch": 1.1556318456759849,
"grad_norm": 3.34375,
"learning_rate": 1.896940051438791e-05,
"loss": 0.2415,
"step": 1768
},
{
"epoch": 1.1562857609939512,
"grad_norm": 4.875,
"learning_rate": 1.896738059386975e-05,
"loss": 0.5278,
"step": 1769
},
{
"epoch": 1.1569396763119175,
"grad_norm": 5.5625,
"learning_rate": 1.89653588035804e-05,
"loss": 0.5758,
"step": 1770
},
{
"epoch": 1.1575935916298838,
"grad_norm": 4.9375,
"learning_rate": 1.8963335143941416e-05,
"loss": 0.5217,
"step": 1771
},
{
"epoch": 1.1582475069478502,
"grad_norm": 8.3125,
"learning_rate": 1.896130961537475e-05,
"loss": 0.6752,
"step": 1772
},
{
"epoch": 1.1589014222658165,
"grad_norm": 4.75,
"learning_rate": 1.8959282218302746e-05,
"loss": 0.3104,
"step": 1773
},
{
"epoch": 1.1595553375837828,
"grad_norm": 3.953125,
"learning_rate": 1.8957252953148125e-05,
"loss": 0.5391,
"step": 1774
},
{
"epoch": 1.1602092529017491,
"grad_norm": 5.71875,
"learning_rate": 1.895522182033401e-05,
"loss": 0.6122,
"step": 1775
},
{
"epoch": 1.1608631682197155,
"grad_norm": 4.65625,
"learning_rate": 1.8953188820283905e-05,
"loss": 0.5364,
"step": 1776
},
{
"epoch": 1.1615170835376818,
"grad_norm": 4.65625,
"learning_rate": 1.895115395342171e-05,
"loss": 0.3107,
"step": 1777
},
{
"epoch": 1.1621709988556481,
"grad_norm": 4.09375,
"learning_rate": 1.894911722017171e-05,
"loss": 0.3487,
"step": 1778
},
{
"epoch": 1.1628249141736144,
"grad_norm": 8.125,
"learning_rate": 1.894707862095858e-05,
"loss": 1.0426,
"step": 1779
},
{
"epoch": 1.1634788294915808,
"grad_norm": 5.21875,
"learning_rate": 1.8945038156207386e-05,
"loss": 0.5502,
"step": 1780
},
{
"epoch": 1.164132744809547,
"grad_norm": 4.34375,
"learning_rate": 1.894299582634358e-05,
"loss": 0.4497,
"step": 1781
},
{
"epoch": 1.1647866601275134,
"grad_norm": 5.625,
"learning_rate": 1.8940951631793e-05,
"loss": 0.4388,
"step": 1782
},
{
"epoch": 1.1654405754454797,
"grad_norm": 5.46875,
"learning_rate": 1.8938905572981884e-05,
"loss": 0.3637,
"step": 1783
},
{
"epoch": 1.166094490763446,
"grad_norm": 5.6875,
"learning_rate": 1.893685765033685e-05,
"loss": 0.5709,
"step": 1784
},
{
"epoch": 1.1667484060814124,
"grad_norm": 3.859375,
"learning_rate": 1.8934807864284904e-05,
"loss": 0.2923,
"step": 1785
},
{
"epoch": 1.1674023213993787,
"grad_norm": 6.15625,
"learning_rate": 1.8932756215253446e-05,
"loss": 0.5836,
"step": 1786
},
{
"epoch": 1.168056236717345,
"grad_norm": 5.375,
"learning_rate": 1.893070270367026e-05,
"loss": 0.5417,
"step": 1787
},
{
"epoch": 1.1687101520353114,
"grad_norm": 3.203125,
"learning_rate": 1.8928647329963516e-05,
"loss": 0.3354,
"step": 1788
},
{
"epoch": 1.1693640673532777,
"grad_norm": 5.15625,
"learning_rate": 1.8926590094561784e-05,
"loss": 0.4202,
"step": 1789
},
{
"epoch": 1.170017982671244,
"grad_norm": 4.28125,
"learning_rate": 1.8924530997894005e-05,
"loss": 0.3976,
"step": 1790
},
{
"epoch": 1.1706718979892103,
"grad_norm": 6.1875,
"learning_rate": 1.8922470040389528e-05,
"loss": 0.5129,
"step": 1791
},
{
"epoch": 1.1713258133071767,
"grad_norm": 5.15625,
"learning_rate": 1.892040722247807e-05,
"loss": 0.4063,
"step": 1792
},
{
"epoch": 1.171979728625143,
"grad_norm": 5.0,
"learning_rate": 1.8918342544589747e-05,
"loss": 0.551,
"step": 1793
},
{
"epoch": 1.1726336439431093,
"grad_norm": 5.875,
"learning_rate": 1.8916276007155065e-05,
"loss": 0.5138,
"step": 1794
},
{
"epoch": 1.1732875592610756,
"grad_norm": 4.78125,
"learning_rate": 1.8914207610604908e-05,
"loss": 0.364,
"step": 1795
},
{
"epoch": 1.173941474579042,
"grad_norm": 5.21875,
"learning_rate": 1.8912137355370558e-05,
"loss": 0.3154,
"step": 1796
},
{
"epoch": 1.1745953898970083,
"grad_norm": 4.8125,
"learning_rate": 1.891006524188368e-05,
"loss": 0.6029,
"step": 1797
},
{
"epoch": 1.1752493052149746,
"grad_norm": 5.8125,
"learning_rate": 1.8907991270576325e-05,
"loss": 0.7767,
"step": 1798
},
{
"epoch": 1.175903220532941,
"grad_norm": 4.9375,
"learning_rate": 1.890591544188093e-05,
"loss": 0.5119,
"step": 1799
},
{
"epoch": 1.1765571358509073,
"grad_norm": 5.65625,
"learning_rate": 1.8903837756230326e-05,
"loss": 0.4669,
"step": 1800
},
{
"epoch": 1.1772110511688736,
"grad_norm": 3.203125,
"learning_rate": 1.8901758214057726e-05,
"loss": 0.2536,
"step": 1801
},
{
"epoch": 1.17786496648684,
"grad_norm": 4.53125,
"learning_rate": 1.889967681579673e-05,
"loss": 0.3883,
"step": 1802
},
{
"epoch": 1.1785188818048062,
"grad_norm": 3.078125,
"learning_rate": 1.8897593561881326e-05,
"loss": 0.2112,
"step": 1803
},
{
"epoch": 1.1791727971227726,
"grad_norm": 5.0,
"learning_rate": 1.8895508452745894e-05,
"loss": 0.5707,
"step": 1804
},
{
"epoch": 1.179826712440739,
"grad_norm": 3.859375,
"learning_rate": 1.889342148882519e-05,
"loss": 0.3054,
"step": 1805
},
{
"epoch": 1.1804806277587052,
"grad_norm": 5.28125,
"learning_rate": 1.889133267055437e-05,
"loss": 0.5753,
"step": 1806
},
{
"epoch": 1.1811345430766715,
"grad_norm": 5.59375,
"learning_rate": 1.888924199836896e-05,
"loss": 0.7539,
"step": 1807
},
{
"epoch": 1.1817884583946379,
"grad_norm": 4.78125,
"learning_rate": 1.8887149472704888e-05,
"loss": 0.627,
"step": 1808
},
{
"epoch": 1.1824423737126042,
"grad_norm": 4.6875,
"learning_rate": 1.8885055093998465e-05,
"loss": 0.4261,
"step": 1809
},
{
"epoch": 1.1830962890305705,
"grad_norm": 3.96875,
"learning_rate": 1.8882958862686377e-05,
"loss": 0.3042,
"step": 1810
},
{
"epoch": 1.1837502043485368,
"grad_norm": 3.796875,
"learning_rate": 1.8880860779205715e-05,
"loss": 0.2819,
"step": 1811
},
{
"epoch": 1.1844041196665032,
"grad_norm": 4.4375,
"learning_rate": 1.887876084399394e-05,
"loss": 0.3086,
"step": 1812
},
{
"epoch": 1.1850580349844695,
"grad_norm": 4.625,
"learning_rate": 1.8876659057488905e-05,
"loss": 0.4658,
"step": 1813
},
{
"epoch": 1.1857119503024358,
"grad_norm": 4.3125,
"learning_rate": 1.8874555420128854e-05,
"loss": 0.3848,
"step": 1814
},
{
"epoch": 1.1863658656204021,
"grad_norm": 2.765625,
"learning_rate": 1.887244993235241e-05,
"loss": 0.2182,
"step": 1815
},
{
"epoch": 1.1870197809383685,
"grad_norm": 4.6875,
"learning_rate": 1.8870342594598585e-05,
"loss": 0.4541,
"step": 1816
},
{
"epoch": 1.1876736962563348,
"grad_norm": 4.09375,
"learning_rate": 1.8868233407306774e-05,
"loss": 0.3568,
"step": 1817
},
{
"epoch": 1.1883276115743011,
"grad_norm": 4.8125,
"learning_rate": 1.8866122370916763e-05,
"loss": 0.5352,
"step": 1818
},
{
"epoch": 1.1889815268922674,
"grad_norm": 7.1875,
"learning_rate": 1.8864009485868716e-05,
"loss": 0.5747,
"step": 1819
},
{
"epoch": 1.1896354422102338,
"grad_norm": 7.5,
"learning_rate": 1.8861894752603187e-05,
"loss": 0.7813,
"step": 1820
},
{
"epoch": 1.1902893575282,
"grad_norm": 2.875,
"learning_rate": 1.8859778171561118e-05,
"loss": 0.2441,
"step": 1821
},
{
"epoch": 1.1909432728461664,
"grad_norm": 2.859375,
"learning_rate": 1.885765974318383e-05,
"loss": 0.1836,
"step": 1822
},
{
"epoch": 1.1915971881641327,
"grad_norm": 5.0625,
"learning_rate": 1.8855539467913034e-05,
"loss": 0.5176,
"step": 1823
},
{
"epoch": 1.192251103482099,
"grad_norm": 5.0625,
"learning_rate": 1.8853417346190827e-05,
"loss": 0.6172,
"step": 1824
},
{
"epoch": 1.1929050188000654,
"grad_norm": 6.0,
"learning_rate": 1.8851293378459685e-05,
"loss": 0.6559,
"step": 1825
},
{
"epoch": 1.1935589341180317,
"grad_norm": 3.453125,
"learning_rate": 1.8849167565162472e-05,
"loss": 0.3288,
"step": 1826
},
{
"epoch": 1.194212849435998,
"grad_norm": 5.375,
"learning_rate": 1.8847039906742442e-05,
"loss": 0.519,
"step": 1827
},
{
"epoch": 1.1948667647539644,
"grad_norm": 7.125,
"learning_rate": 1.884491040364322e-05,
"loss": 0.6269,
"step": 1828
},
{
"epoch": 1.1955206800719307,
"grad_norm": 6.28125,
"learning_rate": 1.8842779056308834e-05,
"loss": 0.9508,
"step": 1829
},
{
"epoch": 1.196174595389897,
"grad_norm": 2.859375,
"learning_rate": 1.8840645865183685e-05,
"loss": 0.1906,
"step": 1830
},
{
"epoch": 1.1968285107078633,
"grad_norm": 4.0625,
"learning_rate": 1.8838510830712554e-05,
"loss": 0.4802,
"step": 1831
},
{
"epoch": 1.1974824260258297,
"grad_norm": 3.25,
"learning_rate": 1.8836373953340625e-05,
"loss": 0.3923,
"step": 1832
},
{
"epoch": 1.198136341343796,
"grad_norm": 5.34375,
"learning_rate": 1.8834235233513437e-05,
"loss": 0.7116,
"step": 1833
},
{
"epoch": 1.1987902566617623,
"grad_norm": 2.59375,
"learning_rate": 1.8832094671676947e-05,
"loss": 0.2161,
"step": 1834
},
{
"epoch": 1.1994441719797286,
"grad_norm": 4.28125,
"learning_rate": 1.8829952268277472e-05,
"loss": 0.5296,
"step": 1835
},
{
"epoch": 1.200098087297695,
"grad_norm": 4.8125,
"learning_rate": 1.882780802376172e-05,
"loss": 0.4934,
"step": 1836
},
{
"epoch": 1.2007520026156613,
"grad_norm": 6.78125,
"learning_rate": 1.8825661938576784e-05,
"loss": 0.8144,
"step": 1837
},
{
"epoch": 1.2014059179336276,
"grad_norm": 4.875,
"learning_rate": 1.8823514013170138e-05,
"loss": 0.7268,
"step": 1838
},
{
"epoch": 1.202059833251594,
"grad_norm": 6.46875,
"learning_rate": 1.8821364247989645e-05,
"loss": 0.8158,
"step": 1839
},
{
"epoch": 1.2027137485695603,
"grad_norm": 7.15625,
"learning_rate": 1.881921264348355e-05,
"loss": 0.7695,
"step": 1840
},
{
"epoch": 1.2033676638875266,
"grad_norm": 4.875,
"learning_rate": 1.881705920010048e-05,
"loss": 0.4648,
"step": 1841
},
{
"epoch": 1.204021579205493,
"grad_norm": 3.328125,
"learning_rate": 1.8814903918289438e-05,
"loss": 0.4097,
"step": 1842
},
{
"epoch": 1.2046754945234592,
"grad_norm": 5.34375,
"learning_rate": 1.8812746798499822e-05,
"loss": 0.4559,
"step": 1843
},
{
"epoch": 1.2053294098414256,
"grad_norm": 3.75,
"learning_rate": 1.881058784118141e-05,
"loss": 0.3152,
"step": 1844
},
{
"epoch": 1.2059833251593919,
"grad_norm": 3.140625,
"learning_rate": 1.8808427046784365e-05,
"loss": 0.2467,
"step": 1845
},
{
"epoch": 1.2066372404773582,
"grad_norm": 4.0625,
"learning_rate": 1.8806264415759222e-05,
"loss": 0.2896,
"step": 1846
},
{
"epoch": 1.2072911557953245,
"grad_norm": 3.921875,
"learning_rate": 1.8804099948556916e-05,
"loss": 0.3453,
"step": 1847
},
{
"epoch": 1.2079450711132909,
"grad_norm": 4.71875,
"learning_rate": 1.8801933645628744e-05,
"loss": 0.5751,
"step": 1848
},
{
"epoch": 1.2085989864312572,
"grad_norm": 4.5625,
"learning_rate": 1.879976550742641e-05,
"loss": 0.5926,
"step": 1849
},
{
"epoch": 1.2092529017492235,
"grad_norm": 5.03125,
"learning_rate": 1.879759553440198e-05,
"loss": 0.6948,
"step": 1850
},
{
"epoch": 1.2099068170671898,
"grad_norm": 4.6875,
"learning_rate": 1.879542372700792e-05,
"loss": 0.4818,
"step": 1851
},
{
"epoch": 1.2105607323851562,
"grad_norm": 4.9375,
"learning_rate": 1.879325008569706e-05,
"loss": 0.6441,
"step": 1852
},
{
"epoch": 1.2112146477031225,
"grad_norm": 6.3125,
"learning_rate": 1.8791074610922624e-05,
"loss": 0.626,
"step": 1853
},
{
"epoch": 1.2118685630210888,
"grad_norm": 3.90625,
"learning_rate": 1.8788897303138215e-05,
"loss": 0.3648,
"step": 1854
},
{
"epoch": 1.2125224783390551,
"grad_norm": 5.5625,
"learning_rate": 1.878671816279783e-05,
"loss": 0.5313,
"step": 1855
},
{
"epoch": 1.2131763936570215,
"grad_norm": 5.34375,
"learning_rate": 1.878453719035582e-05,
"loss": 0.5782,
"step": 1856
},
{
"epoch": 1.2138303089749878,
"grad_norm": 5.34375,
"learning_rate": 1.878235438626695e-05,
"loss": 0.7719,
"step": 1857
},
{
"epoch": 1.214484224292954,
"grad_norm": 3.65625,
"learning_rate": 1.8780169750986342e-05,
"loss": 0.1953,
"step": 1858
},
{
"epoch": 1.2151381396109204,
"grad_norm": 6.03125,
"learning_rate": 1.8777983284969513e-05,
"loss": 0.6721,
"step": 1859
},
{
"epoch": 1.2157920549288868,
"grad_norm": 3.296875,
"learning_rate": 1.8775794988672364e-05,
"loss": 0.2816,
"step": 1860
},
{
"epoch": 1.216445970246853,
"grad_norm": 4.25,
"learning_rate": 1.877360486255117e-05,
"loss": 0.492,
"step": 1861
},
{
"epoch": 1.2170998855648194,
"grad_norm": 5.28125,
"learning_rate": 1.8771412907062586e-05,
"loss": 0.5339,
"step": 1862
},
{
"epoch": 1.2177538008827857,
"grad_norm": 4.09375,
"learning_rate": 1.8769219122663654e-05,
"loss": 0.4807,
"step": 1863
},
{
"epoch": 1.218407716200752,
"grad_norm": 3.296875,
"learning_rate": 1.87670235098118e-05,
"loss": 0.2866,
"step": 1864
},
{
"epoch": 1.2190616315187184,
"grad_norm": 3.828125,
"learning_rate": 1.876482606896482e-05,
"loss": 0.3427,
"step": 1865
},
{
"epoch": 1.2197155468366847,
"grad_norm": 7.125,
"learning_rate": 1.8762626800580905e-05,
"loss": 0.971,
"step": 1866
},
{
"epoch": 1.220369462154651,
"grad_norm": 4.3125,
"learning_rate": 1.8760425705118614e-05,
"loss": 0.3703,
"step": 1867
},
{
"epoch": 1.2210233774726174,
"grad_norm": 2.65625,
"learning_rate": 1.87582227830369e-05,
"loss": 0.1433,
"step": 1868
},
{
"epoch": 1.2216772927905837,
"grad_norm": 4.125,
"learning_rate": 1.8756018034795084e-05,
"loss": 0.2775,
"step": 1869
},
{
"epoch": 1.22233120810855,
"grad_norm": 3.390625,
"learning_rate": 1.8753811460852876e-05,
"loss": 0.2552,
"step": 1870
},
{
"epoch": 1.2229851234265163,
"grad_norm": 3.234375,
"learning_rate": 1.8751603061670368e-05,
"loss": 0.2434,
"step": 1871
},
{
"epoch": 1.2236390387444827,
"grad_norm": 3.625,
"learning_rate": 1.874939283770802e-05,
"loss": 0.3092,
"step": 1872
},
{
"epoch": 1.224292954062449,
"grad_norm": 5.03125,
"learning_rate": 1.874718078942669e-05,
"loss": 0.54,
"step": 1873
},
{
"epoch": 1.2249468693804153,
"grad_norm": 4.34375,
"learning_rate": 1.874496691728761e-05,
"loss": 0.4759,
"step": 1874
},
{
"epoch": 1.2256007846983816,
"grad_norm": 5.75,
"learning_rate": 1.8742751221752377e-05,
"loss": 0.497,
"step": 1875
},
{
"epoch": 1.226254700016348,
"grad_norm": 6.5625,
"learning_rate": 1.8740533703282994e-05,
"loss": 0.6583,
"step": 1876
},
{
"epoch": 1.2269086153343143,
"grad_norm": 4.53125,
"learning_rate": 1.8738314362341828e-05,
"loss": 0.3153,
"step": 1877
},
{
"epoch": 1.2275625306522806,
"grad_norm": 4.3125,
"learning_rate": 1.8736093199391626e-05,
"loss": 0.4116,
"step": 1878
},
{
"epoch": 1.228216445970247,
"grad_norm": 5.9375,
"learning_rate": 1.8733870214895523e-05,
"loss": 0.4747,
"step": 1879
},
{
"epoch": 1.2288703612882133,
"grad_norm": 3.34375,
"learning_rate": 1.8731645409317022e-05,
"loss": 0.1829,
"step": 1880
},
{
"epoch": 1.2295242766061796,
"grad_norm": 3.375,
"learning_rate": 1.8729418783120025e-05,
"loss": 0.288,
"step": 1881
},
{
"epoch": 1.230178191924146,
"grad_norm": 5.0,
"learning_rate": 1.872719033676879e-05,
"loss": 0.4688,
"step": 1882
},
{
"epoch": 1.2308321072421122,
"grad_norm": 5.5,
"learning_rate": 1.8724960070727974e-05,
"loss": 0.3791,
"step": 1883
},
{
"epoch": 1.2314860225600786,
"grad_norm": 5.21875,
"learning_rate": 1.87227279854626e-05,
"loss": 0.6097,
"step": 1884
},
{
"epoch": 1.2321399378780449,
"grad_norm": 5.1875,
"learning_rate": 1.872049408143808e-05,
"loss": 0.4637,
"step": 1885
},
{
"epoch": 1.2327938531960112,
"grad_norm": 3.515625,
"learning_rate": 1.8718258359120196e-05,
"loss": 0.2699,
"step": 1886
},
{
"epoch": 1.2334477685139775,
"grad_norm": 5.375,
"learning_rate": 1.871602081897512e-05,
"loss": 0.6195,
"step": 1887
},
{
"epoch": 1.2341016838319439,
"grad_norm": 4.53125,
"learning_rate": 1.8713781461469392e-05,
"loss": 0.5253,
"step": 1888
},
{
"epoch": 1.2347555991499102,
"grad_norm": 5.15625,
"learning_rate": 1.8711540287069937e-05,
"loss": 0.5069,
"step": 1889
},
{
"epoch": 1.2354095144678765,
"grad_norm": 6.90625,
"learning_rate": 1.8709297296244063e-05,
"loss": 0.6096,
"step": 1890
},
{
"epoch": 1.2360634297858428,
"grad_norm": 5.84375,
"learning_rate": 1.8707052489459447e-05,
"loss": 0.4621,
"step": 1891
},
{
"epoch": 1.2367173451038092,
"grad_norm": 5.75,
"learning_rate": 1.8704805867184147e-05,
"loss": 0.5406,
"step": 1892
},
{
"epoch": 1.2373712604217755,
"grad_norm": 3.4375,
"learning_rate": 1.8702557429886607e-05,
"loss": 0.275,
"step": 1893
},
{
"epoch": 1.2380251757397418,
"grad_norm": 4.4375,
"learning_rate": 1.8700307178035643e-05,
"loss": 0.2549,
"step": 1894
},
{
"epoch": 1.238679091057708,
"grad_norm": 7.09375,
"learning_rate": 1.869805511210045e-05,
"loss": 0.5705,
"step": 1895
},
{
"epoch": 1.2393330063756745,
"grad_norm": 4.21875,
"learning_rate": 1.8695801232550603e-05,
"loss": 0.255,
"step": 1896
},
{
"epoch": 1.2399869216936406,
"grad_norm": 5.1875,
"learning_rate": 1.8693545539856052e-05,
"loss": 0.2976,
"step": 1897
},
{
"epoch": 1.240640837011607,
"grad_norm": 3.734375,
"learning_rate": 1.869128803448713e-05,
"loss": 0.4428,
"step": 1898
},
{
"epoch": 1.2412947523295732,
"grad_norm": 4.15625,
"learning_rate": 1.868902871691454e-05,
"loss": 0.3686,
"step": 1899
},
{
"epoch": 1.2419486676475398,
"grad_norm": 4.25,
"learning_rate": 1.8686767587609376e-05,
"loss": 0.3735,
"step": 1900
},
{
"epoch": 1.2426025829655059,
"grad_norm": 3.96875,
"learning_rate": 1.8684504647043093e-05,
"loss": 0.4075,
"step": 1901
},
{
"epoch": 1.2432564982834724,
"grad_norm": 5.0625,
"learning_rate": 1.868223989568754e-05,
"loss": 0.5168,
"step": 1902
},
{
"epoch": 1.2439104136014385,
"grad_norm": 5.09375,
"learning_rate": 1.8679973334014927e-05,
"loss": 0.764,
"step": 1903
},
{
"epoch": 1.244564328919405,
"grad_norm": 5.78125,
"learning_rate": 1.8677704962497856e-05,
"loss": 0.5123,
"step": 1904
},
{
"epoch": 1.2452182442373712,
"grad_norm": 3.953125,
"learning_rate": 1.8675434781609305e-05,
"loss": 0.3374,
"step": 1905
},
{
"epoch": 1.2458721595553377,
"grad_norm": 4.15625,
"learning_rate": 1.8673162791822617e-05,
"loss": 0.2769,
"step": 1906
},
{
"epoch": 1.2465260748733038,
"grad_norm": 5.9375,
"learning_rate": 1.8670888993611526e-05,
"loss": 0.4565,
"step": 1907
},
{
"epoch": 1.2471799901912703,
"grad_norm": 6.40625,
"learning_rate": 1.866861338745013e-05,
"loss": 0.4852,
"step": 1908
},
{
"epoch": 1.2478339055092365,
"grad_norm": 5.34375,
"learning_rate": 1.8666335973812922e-05,
"loss": 0.5433,
"step": 1909
},
{
"epoch": 1.248487820827203,
"grad_norm": 2.640625,
"learning_rate": 1.8664056753174754e-05,
"loss": 0.1548,
"step": 1910
},
{
"epoch": 1.249141736145169,
"grad_norm": 5.03125,
"learning_rate": 1.866177572601086e-05,
"loss": 0.5246,
"step": 1911
},
{
"epoch": 1.2497956514631356,
"grad_norm": 6.84375,
"learning_rate": 1.865949289279686e-05,
"loss": 0.5556,
"step": 1912
},
{
"epoch": 1.2504495667811018,
"grad_norm": 5.125,
"learning_rate": 1.8657208254008735e-05,
"loss": 0.4132,
"step": 1913
},
{
"epoch": 1.2511034820990683,
"grad_norm": 3.984375,
"learning_rate": 1.865492181012286e-05,
"loss": 0.2818,
"step": 1914
},
{
"epoch": 1.2517573974170344,
"grad_norm": 6.34375,
"learning_rate": 1.8652633561615965e-05,
"loss": 0.7699,
"step": 1915
},
{
"epoch": 1.252411312735001,
"grad_norm": 4.25,
"learning_rate": 1.865034350896518e-05,
"loss": 0.3978,
"step": 1916
},
{
"epoch": 1.253065228052967,
"grad_norm": 4.59375,
"learning_rate": 1.864805165264799e-05,
"loss": 0.2887,
"step": 1917
},
{
"epoch": 1.2537191433709336,
"grad_norm": 4.0625,
"learning_rate": 1.8645757993142274e-05,
"loss": 0.2203,
"step": 1918
},
{
"epoch": 1.2543730586888997,
"grad_norm": 3.4375,
"learning_rate": 1.864346253092627e-05,
"loss": 0.3944,
"step": 1919
},
{
"epoch": 1.2550269740068662,
"grad_norm": 3.5,
"learning_rate": 1.8641165266478606e-05,
"loss": 0.1458,
"step": 1920
},
{
"epoch": 1.2556808893248323,
"grad_norm": 4.59375,
"learning_rate": 1.8638866200278276e-05,
"loss": 0.3679,
"step": 1921
},
{
"epoch": 1.2563348046427987,
"grad_norm": 4.75,
"learning_rate": 1.863656533280466e-05,
"loss": 0.3842,
"step": 1922
},
{
"epoch": 1.256988719960765,
"grad_norm": 5.46875,
"learning_rate": 1.8634262664537505e-05,
"loss": 0.449,
"step": 1923
},
{
"epoch": 1.2576426352787313,
"grad_norm": 3.28125,
"learning_rate": 1.863195819595693e-05,
"loss": 0.2344,
"step": 1924
},
{
"epoch": 1.2582965505966976,
"grad_norm": 6.5625,
"learning_rate": 1.8629651927543447e-05,
"loss": 0.8233,
"step": 1925
},
{
"epoch": 1.258950465914664,
"grad_norm": 3.046875,
"learning_rate": 1.862734385977792e-05,
"loss": 0.2332,
"step": 1926
},
{
"epoch": 1.2596043812326303,
"grad_norm": 5.9375,
"learning_rate": 1.8625033993141606e-05,
"loss": 0.4593,
"step": 1927
},
{
"epoch": 1.2602582965505966,
"grad_norm": 4.4375,
"learning_rate": 1.8622722328116126e-05,
"loss": 0.3226,
"step": 1928
},
{
"epoch": 1.260912211868563,
"grad_norm": 4.03125,
"learning_rate": 1.8620408865183492e-05,
"loss": 0.402,
"step": 1929
},
{
"epoch": 1.2615661271865293,
"grad_norm": 3.671875,
"learning_rate": 1.8618093604826066e-05,
"loss": 0.3888,
"step": 1930
},
{
"epoch": 1.2622200425044956,
"grad_norm": 4.53125,
"learning_rate": 1.8615776547526607e-05,
"loss": 0.2518,
"step": 1931
},
{
"epoch": 1.262873957822462,
"grad_norm": 2.890625,
"learning_rate": 1.8613457693768234e-05,
"loss": 0.2422,
"step": 1932
},
{
"epoch": 1.2635278731404282,
"grad_norm": 3.71875,
"learning_rate": 1.8611137044034454e-05,
"loss": 0.2975,
"step": 1933
},
{
"epoch": 1.2641817884583946,
"grad_norm": 4.5625,
"learning_rate": 1.8608814598809134e-05,
"loss": 0.2293,
"step": 1934
},
{
"epoch": 1.264835703776361,
"grad_norm": 5.09375,
"learning_rate": 1.860649035857653e-05,
"loss": 0.5415,
"step": 1935
},
{
"epoch": 1.2654896190943272,
"grad_norm": 3.765625,
"learning_rate": 1.8604164323821257e-05,
"loss": 0.3322,
"step": 1936
},
{
"epoch": 1.2661435344122935,
"grad_norm": 5.1875,
"learning_rate": 1.8601836495028318e-05,
"loss": 0.3812,
"step": 1937
},
{
"epoch": 1.2667974497302599,
"grad_norm": 4.375,
"learning_rate": 1.859950687268308e-05,
"loss": 0.3573,
"step": 1938
},
{
"epoch": 1.2674513650482262,
"grad_norm": 3.609375,
"learning_rate": 1.8597175457271293e-05,
"loss": 0.3072,
"step": 1939
},
{
"epoch": 1.2681052803661925,
"grad_norm": 2.859375,
"learning_rate": 1.8594842249279067e-05,
"loss": 0.153,
"step": 1940
},
{
"epoch": 1.2687591956841588,
"grad_norm": 3.765625,
"learning_rate": 1.8592507249192906e-05,
"loss": 0.2034,
"step": 1941
},
{
"epoch": 1.2694131110021252,
"grad_norm": 6.1875,
"learning_rate": 1.8590170457499667e-05,
"loss": 0.6676,
"step": 1942
},
{
"epoch": 1.2700670263200915,
"grad_norm": 4.53125,
"learning_rate": 1.8587831874686598e-05,
"loss": 0.354,
"step": 1943
},
{
"epoch": 1.2707209416380578,
"grad_norm": 3.34375,
"learning_rate": 1.8585491501241307e-05,
"loss": 0.1714,
"step": 1944
},
{
"epoch": 1.2713748569560241,
"grad_norm": 5.03125,
"learning_rate": 1.858314933765178e-05,
"loss": 0.2854,
"step": 1945
},
{
"epoch": 1.2720287722739905,
"grad_norm": 5.15625,
"learning_rate": 1.858080538440638e-05,
"loss": 0.4345,
"step": 1946
},
{
"epoch": 1.2726826875919568,
"grad_norm": 3.703125,
"learning_rate": 1.857845964199384e-05,
"loss": 0.2506,
"step": 1947
},
{
"epoch": 1.2733366029099231,
"grad_norm": 5.28125,
"learning_rate": 1.8576112110903263e-05,
"loss": 0.3076,
"step": 1948
},
{
"epoch": 1.2739905182278894,
"grad_norm": 3.484375,
"learning_rate": 1.8573762791624132e-05,
"loss": 0.3637,
"step": 1949
},
{
"epoch": 1.2746444335458558,
"grad_norm": 3.96875,
"learning_rate": 1.85714116846463e-05,
"loss": 0.4045,
"step": 1950
},
{
"epoch": 1.275298348863822,
"grad_norm": 4.875,
"learning_rate": 1.8569058790459986e-05,
"loss": 0.3463,
"step": 1951
},
{
"epoch": 1.2759522641817884,
"grad_norm": 4.53125,
"learning_rate": 1.8566704109555787e-05,
"loss": 0.3541,
"step": 1952
},
{
"epoch": 1.2766061794997547,
"grad_norm": 7.3125,
"learning_rate": 1.856434764242468e-05,
"loss": 0.7119,
"step": 1953
},
{
"epoch": 1.277260094817721,
"grad_norm": 5.90625,
"learning_rate": 1.8561989389558007e-05,
"loss": 0.7424,
"step": 1954
},
{
"epoch": 1.2779140101356874,
"grad_norm": 4.53125,
"learning_rate": 1.855962935144748e-05,
"loss": 0.3664,
"step": 1955
},
{
"epoch": 1.2785679254536537,
"grad_norm": 2.625,
"learning_rate": 1.855726752858518e-05,
"loss": 0.1689,
"step": 1956
},
{
"epoch": 1.27922184077162,
"grad_norm": 5.28125,
"learning_rate": 1.8554903921463578e-05,
"loss": 0.3954,
"step": 1957
},
{
"epoch": 1.2798757560895864,
"grad_norm": 3.765625,
"learning_rate": 1.8552538530575495e-05,
"loss": 0.3754,
"step": 1958
},
{
"epoch": 1.2805296714075527,
"grad_norm": 3.109375,
"learning_rate": 1.8550171356414142e-05,
"loss": 0.2335,
"step": 1959
},
{
"epoch": 1.281183586725519,
"grad_norm": 9.1875,
"learning_rate": 1.854780239947309e-05,
"loss": 0.84,
"step": 1960
},
{
"epoch": 1.2818375020434853,
"grad_norm": 4.5625,
"learning_rate": 1.854543166024628e-05,
"loss": 0.4249,
"step": 1961
},
{
"epoch": 1.2824914173614517,
"grad_norm": 4.34375,
"learning_rate": 1.854305913922804e-05,
"loss": 0.3897,
"step": 1962
},
{
"epoch": 1.283145332679418,
"grad_norm": 3.71875,
"learning_rate": 1.854068483691305e-05,
"loss": 0.2408,
"step": 1963
},
{
"epoch": 1.2837992479973843,
"grad_norm": 5.0,
"learning_rate": 1.8538308753796385e-05,
"loss": 0.41,
"step": 1964
},
{
"epoch": 1.2844531633153506,
"grad_norm": 4.65625,
"learning_rate": 1.8535930890373467e-05,
"loss": 0.4126,
"step": 1965
},
{
"epoch": 1.285107078633317,
"grad_norm": 4.46875,
"learning_rate": 1.85335512471401e-05,
"loss": 0.6946,
"step": 1966
},
{
"epoch": 1.2857609939512833,
"grad_norm": 4.6875,
"learning_rate": 1.853116982459246e-05,
"loss": 0.3819,
"step": 1967
},
{
"epoch": 1.2864149092692496,
"grad_norm": 8.25,
"learning_rate": 1.8528786623227096e-05,
"loss": 0.9109,
"step": 1968
},
{
"epoch": 1.287068824587216,
"grad_norm": 3.578125,
"learning_rate": 1.8526401643540924e-05,
"loss": 0.3197,
"step": 1969
},
{
"epoch": 1.2877227399051823,
"grad_norm": 3.5,
"learning_rate": 1.8524014886031228e-05,
"loss": 0.3296,
"step": 1970
},
{
"epoch": 1.2883766552231486,
"grad_norm": 3.515625,
"learning_rate": 1.8521626351195666e-05,
"loss": 0.3846,
"step": 1971
},
{
"epoch": 1.289030570541115,
"grad_norm": 7.75,
"learning_rate": 1.8519236039532273e-05,
"loss": 0.8143,
"step": 1972
},
{
"epoch": 1.2896844858590812,
"grad_norm": 3.8125,
"learning_rate": 1.851684395153944e-05,
"loss": 0.3787,
"step": 1973
},
{
"epoch": 1.2903384011770476,
"grad_norm": 3.5625,
"learning_rate": 1.8514450087715944e-05,
"loss": 0.3277,
"step": 1974
},
{
"epoch": 1.290992316495014,
"grad_norm": 4.6875,
"learning_rate": 1.851205444856092e-05,
"loss": 0.5747,
"step": 1975
},
{
"epoch": 1.2916462318129802,
"grad_norm": 4.90625,
"learning_rate": 1.850965703457388e-05,
"loss": 0.3631,
"step": 1976
},
{
"epoch": 1.2923001471309465,
"grad_norm": 3.78125,
"learning_rate": 1.8507257846254707e-05,
"loss": 0.2915,
"step": 1977
},
{
"epoch": 1.2929540624489129,
"grad_norm": 4.78125,
"learning_rate": 1.8504856884103645e-05,
"loss": 0.3131,
"step": 1978
},
{
"epoch": 1.2936079777668792,
"grad_norm": 4.59375,
"learning_rate": 1.8502454148621318e-05,
"loss": 0.2863,
"step": 1979
},
{
"epoch": 1.2942618930848455,
"grad_norm": 7.25,
"learning_rate": 1.8500049640308712e-05,
"loss": 0.8575,
"step": 1980
},
{
"epoch": 1.2949158084028118,
"grad_norm": 5.1875,
"learning_rate": 1.8497643359667193e-05,
"loss": 0.3204,
"step": 1981
},
{
"epoch": 1.2955697237207782,
"grad_norm": 3.78125,
"learning_rate": 1.849523530719848e-05,
"loss": 0.3316,
"step": 1982
},
{
"epoch": 1.2962236390387445,
"grad_norm": 4.5625,
"learning_rate": 1.8492825483404683e-05,
"loss": 0.5437,
"step": 1983
},
{
"epoch": 1.2968775543567108,
"grad_norm": 4.78125,
"learning_rate": 1.8490413888788267e-05,
"loss": 0.4979,
"step": 1984
},
{
"epoch": 1.2975314696746771,
"grad_norm": 3.796875,
"learning_rate": 1.848800052385206e-05,
"loss": 0.4553,
"step": 1985
},
{
"epoch": 1.2981853849926435,
"grad_norm": 7.6875,
"learning_rate": 1.8485585389099276e-05,
"loss": 1.0042,
"step": 1986
},
{
"epoch": 1.2988393003106098,
"grad_norm": 3.515625,
"learning_rate": 1.848316848503349e-05,
"loss": 0.3589,
"step": 1987
},
{
"epoch": 1.2994932156285761,
"grad_norm": 3.6875,
"learning_rate": 1.8480749812158644e-05,
"loss": 0.2516,
"step": 1988
},
{
"epoch": 1.3001471309465424,
"grad_norm": 5.0625,
"learning_rate": 1.8478329370979047e-05,
"loss": 0.2446,
"step": 1989
},
{
"epoch": 1.3008010462645088,
"grad_norm": 4.28125,
"learning_rate": 1.8475907161999388e-05,
"loss": 0.289,
"step": 1990
},
{
"epoch": 1.301454961582475,
"grad_norm": 4.1875,
"learning_rate": 1.8473483185724712e-05,
"loss": 0.4933,
"step": 1991
},
{
"epoch": 1.3021088769004414,
"grad_norm": 6.4375,
"learning_rate": 1.847105744266044e-05,
"loss": 0.4275,
"step": 1992
},
{
"epoch": 1.3027627922184077,
"grad_norm": 6.3125,
"learning_rate": 1.8468629933312358e-05,
"loss": 0.7569,
"step": 1993
},
{
"epoch": 1.303416707536374,
"grad_norm": 5.1875,
"learning_rate": 1.8466200658186622e-05,
"loss": 0.6828,
"step": 1994
},
{
"epoch": 1.3040706228543404,
"grad_norm": 5.28125,
"learning_rate": 1.8463769617789754e-05,
"loss": 0.2777,
"step": 1995
},
{
"epoch": 1.3047245381723067,
"grad_norm": 3.703125,
"learning_rate": 1.8461336812628648e-05,
"loss": 0.458,
"step": 1996
},
{
"epoch": 1.305378453490273,
"grad_norm": 4.59375,
"learning_rate": 1.8458902243210558e-05,
"loss": 0.3809,
"step": 1997
},
{
"epoch": 1.3060323688082394,
"grad_norm": 6.46875,
"learning_rate": 1.845646591004312e-05,
"loss": 0.6737,
"step": 1998
},
{
"epoch": 1.3066862841262057,
"grad_norm": 4.6875,
"learning_rate": 1.8454027813634323e-05,
"loss": 0.4581,
"step": 1999
},
{
"epoch": 1.307340199444172,
"grad_norm": 7.4375,
"learning_rate": 1.8451587954492533e-05,
"loss": 0.4362,
"step": 2000
},
{
"epoch": 1.3079941147621383,
"grad_norm": 4.875,
"learning_rate": 1.8449146333126476e-05,
"loss": 0.5957,
"step": 2001
},
{
"epoch": 1.3086480300801047,
"grad_norm": 3.09375,
"learning_rate": 1.8446702950045252e-05,
"loss": 0.2362,
"step": 2002
},
{
"epoch": 1.309301945398071,
"grad_norm": 2.328125,
"learning_rate": 1.8444257805758327e-05,
"loss": 0.1312,
"step": 2003
},
{
"epoch": 1.3099558607160373,
"grad_norm": 3.046875,
"learning_rate": 1.8441810900775535e-05,
"loss": 0.3069,
"step": 2004
},
{
"epoch": 1.3106097760340036,
"grad_norm": 3.671875,
"learning_rate": 1.8439362235607074e-05,
"loss": 0.3313,
"step": 2005
},
{
"epoch": 1.31126369135197,
"grad_norm": 4.8125,
"learning_rate": 1.84369118107635e-05,
"loss": 0.5656,
"step": 2006
},
{
"epoch": 1.3119176066699363,
"grad_norm": 4.8125,
"learning_rate": 1.8434459626755767e-05,
"loss": 0.3542,
"step": 2007
},
{
"epoch": 1.3125715219879026,
"grad_norm": 4.1875,
"learning_rate": 1.843200568409516e-05,
"loss": 0.4698,
"step": 2008
},
{
"epoch": 1.313225437305869,
"grad_norm": 5.0625,
"learning_rate": 1.842954998329335e-05,
"loss": 0.3848,
"step": 2009
},
{
"epoch": 1.3138793526238353,
"grad_norm": 4.8125,
"learning_rate": 1.842709252486237e-05,
"loss": 0.5121,
"step": 2010
},
{
"epoch": 1.3145332679418016,
"grad_norm": 4.0625,
"learning_rate": 1.842463330931462e-05,
"loss": 0.335,
"step": 2011
},
{
"epoch": 1.315187183259768,
"grad_norm": 4.96875,
"learning_rate": 1.8422172337162865e-05,
"loss": 0.5883,
"step": 2012
},
{
"epoch": 1.3158410985777342,
"grad_norm": 4.8125,
"learning_rate": 1.8419709608920243e-05,
"loss": 0.3473,
"step": 2013
},
{
"epoch": 1.3164950138957006,
"grad_norm": 2.578125,
"learning_rate": 1.8417245125100248e-05,
"loss": 0.2446,
"step": 2014
},
{
"epoch": 1.3171489292136669,
"grad_norm": 2.828125,
"learning_rate": 1.8414778886216744e-05,
"loss": 0.195,
"step": 2015
},
{
"epoch": 1.3178028445316332,
"grad_norm": 3.65625,
"learning_rate": 1.8412310892783962e-05,
"loss": 0.3534,
"step": 2016
},
{
"epoch": 1.3184567598495995,
"grad_norm": 4.40625,
"learning_rate": 1.84098411453165e-05,
"loss": 0.3403,
"step": 2017
},
{
"epoch": 1.3191106751675659,
"grad_norm": 5.28125,
"learning_rate": 1.840736964432932e-05,
"loss": 0.4891,
"step": 2018
},
{
"epoch": 1.3197645904855322,
"grad_norm": 3.1875,
"learning_rate": 1.8404896390337756e-05,
"loss": 0.2731,
"step": 2019
},
{
"epoch": 1.3204185058034985,
"grad_norm": 3.875,
"learning_rate": 1.840242138385749e-05,
"loss": 0.4217,
"step": 2020
},
{
"epoch": 1.3210724211214648,
"grad_norm": 5.6875,
"learning_rate": 1.8399944625404585e-05,
"loss": 0.7977,
"step": 2021
},
{
"epoch": 1.3217263364394312,
"grad_norm": 4.4375,
"learning_rate": 1.839746611549547e-05,
"loss": 0.5505,
"step": 2022
},
{
"epoch": 1.3223802517573975,
"grad_norm": 4.3125,
"learning_rate": 1.8394985854646928e-05,
"loss": 0.3465,
"step": 2023
},
{
"epoch": 1.3230341670753638,
"grad_norm": 6.3125,
"learning_rate": 1.8392503843376115e-05,
"loss": 0.3455,
"step": 2024
},
{
"epoch": 1.3236880823933301,
"grad_norm": 4.84375,
"learning_rate": 1.8390020082200553e-05,
"loss": 0.2768,
"step": 2025
},
{
"epoch": 1.3243419977112965,
"grad_norm": 6.5,
"learning_rate": 1.8387534571638124e-05,
"loss": 0.5339,
"step": 2026
},
{
"epoch": 1.3249959130292628,
"grad_norm": 4.90625,
"learning_rate": 1.838504731220708e-05,
"loss": 0.4386,
"step": 2027
},
{
"epoch": 1.325649828347229,
"grad_norm": 7.15625,
"learning_rate": 1.838255830442603e-05,
"loss": 0.57,
"step": 2028
},
{
"epoch": 1.3263037436651954,
"grad_norm": 3.59375,
"learning_rate": 1.8380067548813955e-05,
"loss": 0.2516,
"step": 2029
},
{
"epoch": 1.3269576589831618,
"grad_norm": 4.5,
"learning_rate": 1.8377575045890195e-05,
"loss": 0.4564,
"step": 2030
},
{
"epoch": 1.327611574301128,
"grad_norm": 5.1875,
"learning_rate": 1.837508079617446e-05,
"loss": 0.4688,
"step": 2031
},
{
"epoch": 1.3282654896190944,
"grad_norm": 5.1875,
"learning_rate": 1.837258480018682e-05,
"loss": 0.5091,
"step": 2032
},
{
"epoch": 1.3289194049370607,
"grad_norm": 4.65625,
"learning_rate": 1.8370087058447713e-05,
"loss": 0.3689,
"step": 2033
},
{
"epoch": 1.3295733202550268,
"grad_norm": 5.09375,
"learning_rate": 1.8367587571477935e-05,
"loss": 0.4417,
"step": 2034
},
{
"epoch": 1.3302272355729934,
"grad_norm": 6.15625,
"learning_rate": 1.836508633979865e-05,
"loss": 0.6824,
"step": 2035
},
{
"epoch": 1.3308811508909595,
"grad_norm": 5.3125,
"learning_rate": 1.8362583363931385e-05,
"loss": 0.5429,
"step": 2036
},
{
"epoch": 1.331535066208926,
"grad_norm": 5.53125,
"learning_rate": 1.8360078644398028e-05,
"loss": 0.7523,
"step": 2037
},
{
"epoch": 1.3321889815268921,
"grad_norm": 6.4375,
"learning_rate": 1.835757218172084e-05,
"loss": 0.804,
"step": 2038
},
{
"epoch": 1.3328428968448587,
"grad_norm": 3.8125,
"learning_rate": 1.8355063976422433e-05,
"loss": 0.3872,
"step": 2039
},
{
"epoch": 1.3334968121628248,
"grad_norm": 6.0625,
"learning_rate": 1.8352554029025788e-05,
"loss": 0.6263,
"step": 2040
},
{
"epoch": 1.3341507274807913,
"grad_norm": 6.15625,
"learning_rate": 1.8350042340054252e-05,
"loss": 0.7653,
"step": 2041
},
{
"epoch": 1.3348046427987574,
"grad_norm": 5.78125,
"learning_rate": 1.834752891003153e-05,
"loss": 0.4468,
"step": 2042
},
{
"epoch": 1.335458558116724,
"grad_norm": 9.375,
"learning_rate": 1.8345013739481698e-05,
"loss": 0.7043,
"step": 2043
},
{
"epoch": 1.33611247343469,
"grad_norm": 4.125,
"learning_rate": 1.834249682892918e-05,
"loss": 0.3357,
"step": 2044
},
{
"epoch": 1.3367663887526566,
"grad_norm": 4.46875,
"learning_rate": 1.833997817889878e-05,
"loss": 0.2804,
"step": 2045
},
{
"epoch": 1.3374203040706227,
"grad_norm": 5.1875,
"learning_rate": 1.8337457789915656e-05,
"loss": 0.4567,
"step": 2046
},
{
"epoch": 1.3380742193885893,
"grad_norm": 4.78125,
"learning_rate": 1.833493566250532e-05,
"loss": 0.4737,
"step": 2047
},
{
"epoch": 1.3387281347065554,
"grad_norm": 5.375,
"learning_rate": 1.8332411797193672e-05,
"loss": 0.4202,
"step": 2048
},
{
"epoch": 1.339382050024522,
"grad_norm": 7.40625,
"learning_rate": 1.8329886194506948e-05,
"loss": 0.7849,
"step": 2049
},
{
"epoch": 1.340035965342488,
"grad_norm": 6.25,
"learning_rate": 1.8327358854971755e-05,
"loss": 0.7006,
"step": 2050
},
{
"epoch": 1.3406898806604546,
"grad_norm": 5.3125,
"learning_rate": 1.8324829779115068e-05,
"loss": 0.6859,
"step": 2051
},
{
"epoch": 1.3413437959784207,
"grad_norm": 6.25,
"learning_rate": 1.832229896746422e-05,
"loss": 0.8938,
"step": 2052
},
{
"epoch": 1.3419977112963872,
"grad_norm": 6.0,
"learning_rate": 1.8319766420546902e-05,
"loss": 0.51,
"step": 2053
},
{
"epoch": 1.3426516266143533,
"grad_norm": 4.5625,
"learning_rate": 1.8317232138891176e-05,
"loss": 0.3573,
"step": 2054
},
{
"epoch": 1.3433055419323199,
"grad_norm": 3.59375,
"learning_rate": 1.8314696123025456e-05,
"loss": 0.2659,
"step": 2055
},
{
"epoch": 1.343959457250286,
"grad_norm": 6.15625,
"learning_rate": 1.831215837347852e-05,
"loss": 0.7739,
"step": 2056
},
{
"epoch": 1.3446133725682525,
"grad_norm": 5.0625,
"learning_rate": 1.8309618890779513e-05,
"loss": 0.4306,
"step": 2057
},
{
"epoch": 1.3452672878862186,
"grad_norm": 3.484375,
"learning_rate": 1.8307077675457937e-05,
"loss": 0.2061,
"step": 2058
},
{
"epoch": 1.3459212032041852,
"grad_norm": 4.21875,
"learning_rate": 1.8304534728043653e-05,
"loss": 0.2685,
"step": 2059
},
{
"epoch": 1.3465751185221513,
"grad_norm": 3.8125,
"learning_rate": 1.8301990049066892e-05,
"loss": 0.5432,
"step": 2060
},
{
"epoch": 1.3472290338401178,
"grad_norm": 4.03125,
"learning_rate": 1.8299443639058238e-05,
"loss": 0.3311,
"step": 2061
},
{
"epoch": 1.347882949158084,
"grad_norm": 3.53125,
"learning_rate": 1.8296895498548635e-05,
"loss": 0.2636,
"step": 2062
},
{
"epoch": 1.3485368644760505,
"grad_norm": 5.03125,
"learning_rate": 1.8294345628069393e-05,
"loss": 0.6396,
"step": 2063
},
{
"epoch": 1.3491907797940166,
"grad_norm": 5.25,
"learning_rate": 1.8291794028152177e-05,
"loss": 0.4801,
"step": 2064
},
{
"epoch": 1.3498446951119831,
"grad_norm": 3.484375,
"learning_rate": 1.828924069932902e-05,
"loss": 0.3978,
"step": 2065
},
{
"epoch": 1.3504986104299492,
"grad_norm": 6.1875,
"learning_rate": 1.8286685642132314e-05,
"loss": 0.6958,
"step": 2066
},
{
"epoch": 1.3511525257479158,
"grad_norm": 3.125,
"learning_rate": 1.8284128857094804e-05,
"loss": 0.2136,
"step": 2067
},
{
"epoch": 1.3518064410658819,
"grad_norm": 4.875,
"learning_rate": 1.8281570344749602e-05,
"loss": 0.3536,
"step": 2068
},
{
"epoch": 1.3524603563838484,
"grad_norm": 2.796875,
"learning_rate": 1.827901010563018e-05,
"loss": 0.2265,
"step": 2069
},
{
"epoch": 1.3531142717018145,
"grad_norm": 4.3125,
"learning_rate": 1.8276448140270365e-05,
"loss": 0.3259,
"step": 2070
},
{
"epoch": 1.3537681870197809,
"grad_norm": 4.3125,
"learning_rate": 1.8273884449204348e-05,
"loss": 0.3414,
"step": 2071
},
{
"epoch": 1.3544221023377472,
"grad_norm": 4.9375,
"learning_rate": 1.8271319032966683e-05,
"loss": 0.4265,
"step": 2072
},
{
"epoch": 1.3550760176557135,
"grad_norm": 4.96875,
"learning_rate": 1.826875189209228e-05,
"loss": 0.5139,
"step": 2073
},
{
"epoch": 1.3557299329736798,
"grad_norm": 4.59375,
"learning_rate": 1.8266183027116406e-05,
"loss": 0.4905,
"step": 2074
},
{
"epoch": 1.3563838482916462,
"grad_norm": 6.3125,
"learning_rate": 1.826361243857469e-05,
"loss": 0.5678,
"step": 2075
},
{
"epoch": 1.3570377636096125,
"grad_norm": 5.90625,
"learning_rate": 1.826104012700312e-05,
"loss": 0.7462,
"step": 2076
},
{
"epoch": 1.3576916789275788,
"grad_norm": 4.6875,
"learning_rate": 1.8258466092938042e-05,
"loss": 0.4041,
"step": 2077
},
{
"epoch": 1.3583455942455451,
"grad_norm": 3.625,
"learning_rate": 1.8255890336916173e-05,
"loss": 0.3115,
"step": 2078
},
{
"epoch": 1.3589995095635115,
"grad_norm": 5.59375,
"learning_rate": 1.8253312859474563e-05,
"loss": 0.4375,
"step": 2079
},
{
"epoch": 1.3596534248814778,
"grad_norm": 5.34375,
"learning_rate": 1.825073366115065e-05,
"loss": 0.5481,
"step": 2080
},
{
"epoch": 1.360307340199444,
"grad_norm": 6.15625,
"learning_rate": 1.824815274248221e-05,
"loss": 0.6787,
"step": 2081
},
{
"epoch": 1.3609612555174104,
"grad_norm": 6.5,
"learning_rate": 1.8245570104007394e-05,
"loss": 0.4402,
"step": 2082
},
{
"epoch": 1.3616151708353768,
"grad_norm": 4.40625,
"learning_rate": 1.824298574626469e-05,
"loss": 0.4034,
"step": 2083
},
{
"epoch": 1.362269086153343,
"grad_norm": 3.84375,
"learning_rate": 1.8240399669792973e-05,
"loss": 0.3436,
"step": 2084
},
{
"epoch": 1.3629230014713094,
"grad_norm": 7.1875,
"learning_rate": 1.8237811875131446e-05,
"loss": 0.5403,
"step": 2085
},
{
"epoch": 1.3635769167892757,
"grad_norm": 5.125,
"learning_rate": 1.8235222362819693e-05,
"loss": 0.3696,
"step": 2086
},
{
"epoch": 1.364230832107242,
"grad_norm": 5.0625,
"learning_rate": 1.8232631133397647e-05,
"loss": 0.3783,
"step": 2087
},
{
"epoch": 1.3648847474252084,
"grad_norm": 4.65625,
"learning_rate": 1.82300381874056e-05,
"loss": 0.2909,
"step": 2088
},
{
"epoch": 1.3655386627431747,
"grad_norm": 4.65625,
"learning_rate": 1.8227443525384202e-05,
"loss": 0.4083,
"step": 2089
},
{
"epoch": 1.366192578061141,
"grad_norm": 4.71875,
"learning_rate": 1.8224847147874464e-05,
"loss": 0.4122,
"step": 2090
},
{
"epoch": 1.3668464933791074,
"grad_norm": 6.46875,
"learning_rate": 1.8222249055417744e-05,
"loss": 0.6137,
"step": 2091
},
{
"epoch": 1.3675004086970737,
"grad_norm": 4.28125,
"learning_rate": 1.8219649248555776e-05,
"loss": 0.4246,
"step": 2092
},
{
"epoch": 1.36815432401504,
"grad_norm": 3.609375,
"learning_rate": 1.821704772783063e-05,
"loss": 0.2983,
"step": 2093
},
{
"epoch": 1.3688082393330063,
"grad_norm": 4.40625,
"learning_rate": 1.8214444493784748e-05,
"loss": 0.4511,
"step": 2094
},
{
"epoch": 1.3694621546509727,
"grad_norm": 5.625,
"learning_rate": 1.821183954696093e-05,
"loss": 0.4283,
"step": 2095
},
{
"epoch": 1.370116069968939,
"grad_norm": 3.921875,
"learning_rate": 1.820923288790232e-05,
"loss": 0.2868,
"step": 2096
},
{
"epoch": 1.3707699852869053,
"grad_norm": 4.4375,
"learning_rate": 1.8206624517152437e-05,
"loss": 0.3739,
"step": 2097
},
{
"epoch": 1.3714239006048716,
"grad_norm": 4.125,
"learning_rate": 1.8204014435255136e-05,
"loss": 0.4141,
"step": 2098
},
{
"epoch": 1.372077815922838,
"grad_norm": 4.28125,
"learning_rate": 1.8201402642754652e-05,
"loss": 0.2279,
"step": 2099
},
{
"epoch": 1.3727317312408043,
"grad_norm": 6.28125,
"learning_rate": 1.819878914019556e-05,
"loss": 0.4616,
"step": 2100
},
{
"epoch": 1.3733856465587706,
"grad_norm": 4.84375,
"learning_rate": 1.819617392812279e-05,
"loss": 0.3735,
"step": 2101
},
{
"epoch": 1.374039561876737,
"grad_norm": 5.6875,
"learning_rate": 1.8193557007081643e-05,
"loss": 0.7108,
"step": 2102
},
{
"epoch": 1.3746934771947033,
"grad_norm": 5.8125,
"learning_rate": 1.8190938377617767e-05,
"loss": 0.2632,
"step": 2103
},
{
"epoch": 1.3753473925126696,
"grad_norm": 4.21875,
"learning_rate": 1.8188318040277165e-05,
"loss": 0.182,
"step": 2104
},
{
"epoch": 1.376001307830636,
"grad_norm": 6.09375,
"learning_rate": 1.8185695995606196e-05,
"loss": 0.7313,
"step": 2105
},
{
"epoch": 1.3766552231486022,
"grad_norm": 5.78125,
"learning_rate": 1.8183072244151584e-05,
"loss": 0.5339,
"step": 2106
},
{
"epoch": 1.3773091384665685,
"grad_norm": 5.3125,
"learning_rate": 1.81804467864604e-05,
"loss": 0.5137,
"step": 2107
},
{
"epoch": 1.3779630537845349,
"grad_norm": 4.0,
"learning_rate": 1.8177819623080075e-05,
"loss": 0.3554,
"step": 2108
},
{
"epoch": 1.3786169691025012,
"grad_norm": 6.71875,
"learning_rate": 1.8175190754558384e-05,
"loss": 0.7832,
"step": 2109
},
{
"epoch": 1.3792708844204675,
"grad_norm": 3.875,
"learning_rate": 1.817256018144348e-05,
"loss": 0.2194,
"step": 2110
},
{
"epoch": 1.3799247997384338,
"grad_norm": 3.765625,
"learning_rate": 1.816992790428385e-05,
"loss": 0.2622,
"step": 2111
},
{
"epoch": 1.3805787150564002,
"grad_norm": 3.9375,
"learning_rate": 1.816729392362835e-05,
"loss": 0.4399,
"step": 2112
},
{
"epoch": 1.3812326303743665,
"grad_norm": 3.984375,
"learning_rate": 1.8164658240026184e-05,
"loss": 0.3676,
"step": 2113
},
{
"epoch": 1.3818865456923328,
"grad_norm": 6.15625,
"learning_rate": 1.816202085402691e-05,
"loss": 0.4939,
"step": 2114
},
{
"epoch": 1.3825404610102991,
"grad_norm": 5.125,
"learning_rate": 1.8159381766180453e-05,
"loss": 0.5103,
"step": 2115
},
{
"epoch": 1.3831943763282655,
"grad_norm": 4.46875,
"learning_rate": 1.8156740977037076e-05,
"loss": 0.3793,
"step": 2116
},
{
"epoch": 1.3838482916462318,
"grad_norm": 4.21875,
"learning_rate": 1.8154098487147405e-05,
"loss": 0.4362,
"step": 2117
},
{
"epoch": 1.3845022069641981,
"grad_norm": 4.5,
"learning_rate": 1.815145429706243e-05,
"loss": 0.2675,
"step": 2118
},
{
"epoch": 1.3851561222821644,
"grad_norm": 4.15625,
"learning_rate": 1.8148808407333475e-05,
"loss": 0.4791,
"step": 2119
},
{
"epoch": 1.3858100376001308,
"grad_norm": 4.96875,
"learning_rate": 1.8146160818512236e-05,
"loss": 0.5082,
"step": 2120
},
{
"epoch": 1.386463952918097,
"grad_norm": 7.125,
"learning_rate": 1.814351153115075e-05,
"loss": 0.6934,
"step": 2121
},
{
"epoch": 1.3871178682360634,
"grad_norm": 3.625,
"learning_rate": 1.8140860545801423e-05,
"loss": 0.2469,
"step": 2122
},
{
"epoch": 1.3877717835540297,
"grad_norm": 4.6875,
"learning_rate": 1.8138207863017002e-05,
"loss": 0.3497,
"step": 2123
},
{
"epoch": 1.388425698871996,
"grad_norm": 8.5,
"learning_rate": 1.813555348335059e-05,
"loss": 1.0528,
"step": 2124
},
{
"epoch": 1.3890796141899624,
"grad_norm": 6.15625,
"learning_rate": 1.8132897407355657e-05,
"loss": 0.6329,
"step": 2125
},
{
"epoch": 1.3897335295079287,
"grad_norm": 4.59375,
"learning_rate": 1.8130239635586003e-05,
"loss": 0.4791,
"step": 2126
},
{
"epoch": 1.390387444825895,
"grad_norm": 4.0625,
"learning_rate": 1.8127580168595804e-05,
"loss": 0.4171,
"step": 2127
},
{
"epoch": 1.3910413601438614,
"grad_norm": 6.75,
"learning_rate": 1.812491900693958e-05,
"loss": 0.6384,
"step": 2128
},
{
"epoch": 1.3916952754618277,
"grad_norm": 4.875,
"learning_rate": 1.8122256151172197e-05,
"loss": 0.523,
"step": 2129
},
{
"epoch": 1.392349190779794,
"grad_norm": 5.40625,
"learning_rate": 1.811959160184889e-05,
"loss": 0.599,
"step": 2130
},
{
"epoch": 1.3930031060977603,
"grad_norm": 4.4375,
"learning_rate": 1.8116925359525236e-05,
"loss": 0.3092,
"step": 2131
},
{
"epoch": 1.3936570214157267,
"grad_norm": 3.890625,
"learning_rate": 1.811425742475717e-05,
"loss": 0.2427,
"step": 2132
},
{
"epoch": 1.394310936733693,
"grad_norm": 5.8125,
"learning_rate": 1.8111587798100974e-05,
"loss": 0.4975,
"step": 2133
},
{
"epoch": 1.3949648520516593,
"grad_norm": 2.234375,
"learning_rate": 1.8108916480113286e-05,
"loss": 0.1756,
"step": 2134
},
{
"epoch": 1.3956187673696256,
"grad_norm": 6.875,
"learning_rate": 1.8106243471351105e-05,
"loss": 0.7843,
"step": 2135
},
{
"epoch": 1.396272682687592,
"grad_norm": 4.3125,
"learning_rate": 1.810356877237177e-05,
"loss": 0.3534,
"step": 2136
},
{
"epoch": 1.3969265980055583,
"grad_norm": 4.65625,
"learning_rate": 1.8100892383732972e-05,
"loss": 0.3893,
"step": 2137
},
{
"epoch": 1.3975805133235246,
"grad_norm": 3.75,
"learning_rate": 1.809821430599277e-05,
"loss": 0.3045,
"step": 2138
},
{
"epoch": 1.398234428641491,
"grad_norm": 4.8125,
"learning_rate": 1.809553453970956e-05,
"loss": 0.4097,
"step": 2139
},
{
"epoch": 1.3988883439594573,
"grad_norm": 5.1875,
"learning_rate": 1.8092853085442093e-05,
"loss": 0.7124,
"step": 2140
},
{
"epoch": 1.3995422592774236,
"grad_norm": 5.53125,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.5841,
"step": 2141
},
{
"epoch": 1.40019617459539,
"grad_norm": 7.3125,
"learning_rate": 1.8087485115191166e-05,
"loss": 0.6909,
"step": 2142
},
{
"epoch": 1.4008500899133562,
"grad_norm": 4.25,
"learning_rate": 1.808479860032697e-05,
"loss": 0.4228,
"step": 2143
},
{
"epoch": 1.4015040052313226,
"grad_norm": 4.3125,
"learning_rate": 1.808211039971705e-05,
"loss": 0.4213,
"step": 2144
},
{
"epoch": 1.402157920549289,
"grad_norm": 4.3125,
"learning_rate": 1.8079420513921913e-05,
"loss": 0.5366,
"step": 2145
},
{
"epoch": 1.4028118358672552,
"grad_norm": 5.65625,
"learning_rate": 1.8076728943502427e-05,
"loss": 0.3706,
"step": 2146
},
{
"epoch": 1.4034657511852215,
"grad_norm": 4.53125,
"learning_rate": 1.80740356890198e-05,
"loss": 0.5166,
"step": 2147
},
{
"epoch": 1.4041196665031879,
"grad_norm": 6.78125,
"learning_rate": 1.807134075103561e-05,
"loss": 0.5706,
"step": 2148
},
{
"epoch": 1.4047735818211542,
"grad_norm": 4.4375,
"learning_rate": 1.8068644130111757e-05,
"loss": 0.465,
"step": 2149
},
{
"epoch": 1.4054274971391205,
"grad_norm": 5.0625,
"learning_rate": 1.806594582681052e-05,
"loss": 0.5925,
"step": 2150
},
{
"epoch": 1.4060814124570868,
"grad_norm": 5.0625,
"learning_rate": 1.8063245841694512e-05,
"loss": 0.4328,
"step": 2151
},
{
"epoch": 1.4067353277750532,
"grad_norm": 6.21875,
"learning_rate": 1.8060544175326704e-05,
"loss": 0.9462,
"step": 2152
},
{
"epoch": 1.4073892430930195,
"grad_norm": 4.25,
"learning_rate": 1.8057840828270413e-05,
"loss": 0.364,
"step": 2153
},
{
"epoch": 1.4080431584109858,
"grad_norm": 2.90625,
"learning_rate": 1.8055135801089306e-05,
"loss": 0.227,
"step": 2154
},
{
"epoch": 1.4086970737289521,
"grad_norm": 3.78125,
"learning_rate": 1.805242909434741e-05,
"loss": 0.2943,
"step": 2155
},
{
"epoch": 1.4093509890469185,
"grad_norm": 6.78125,
"learning_rate": 1.8049720708609094e-05,
"loss": 0.7156,
"step": 2156
},
{
"epoch": 1.4100049043648848,
"grad_norm": 6.34375,
"learning_rate": 1.8047010644439074e-05,
"loss": 0.3797,
"step": 2157
},
{
"epoch": 1.4106588196828511,
"grad_norm": 4.34375,
"learning_rate": 1.804429890240242e-05,
"loss": 0.375,
"step": 2158
},
{
"epoch": 1.4113127350008174,
"grad_norm": 7.53125,
"learning_rate": 1.804158548306456e-05,
"loss": 0.8027,
"step": 2159
},
{
"epoch": 1.4119666503187838,
"grad_norm": 4.78125,
"learning_rate": 1.8038870386991253e-05,
"loss": 0.4264,
"step": 2160
},
{
"epoch": 1.41262056563675,
"grad_norm": 3.09375,
"learning_rate": 1.8036153614748625e-05,
"loss": 0.26,
"step": 2161
},
{
"epoch": 1.4132744809547164,
"grad_norm": 5.09375,
"learning_rate": 1.8033435166903145e-05,
"loss": 0.541,
"step": 2162
},
{
"epoch": 1.4139283962726827,
"grad_norm": 5.0,
"learning_rate": 1.8030715044021628e-05,
"loss": 0.3851,
"step": 2163
},
{
"epoch": 1.414582311590649,
"grad_norm": 4.84375,
"learning_rate": 1.8027993246671243e-05,
"loss": 0.2924,
"step": 2164
},
{
"epoch": 1.4152362269086154,
"grad_norm": 5.46875,
"learning_rate": 1.802526977541951e-05,
"loss": 0.5166,
"step": 2165
},
{
"epoch": 1.4158901422265817,
"grad_norm": 4.0625,
"learning_rate": 1.8022544630834292e-05,
"loss": 0.4899,
"step": 2166
},
{
"epoch": 1.416544057544548,
"grad_norm": 5.375,
"learning_rate": 1.80198178134838e-05,
"loss": 0.4054,
"step": 2167
},
{
"epoch": 1.4171979728625144,
"grad_norm": 5.0,
"learning_rate": 1.8017089323936604e-05,
"loss": 0.2593,
"step": 2168
},
{
"epoch": 1.4178518881804807,
"grad_norm": 5.125,
"learning_rate": 1.801435916276161e-05,
"loss": 0.4517,
"step": 2169
},
{
"epoch": 1.418505803498447,
"grad_norm": 3.53125,
"learning_rate": 1.8011627330528085e-05,
"loss": 0.4364,
"step": 2170
},
{
"epoch": 1.4191597188164133,
"grad_norm": 5.5,
"learning_rate": 1.8008893827805632e-05,
"loss": 0.5262,
"step": 2171
},
{
"epoch": 1.4198136341343797,
"grad_norm": 6.96875,
"learning_rate": 1.800615865516421e-05,
"loss": 0.5427,
"step": 2172
},
{
"epoch": 1.420467549452346,
"grad_norm": 5.0625,
"learning_rate": 1.800342181317413e-05,
"loss": 0.4951,
"step": 2173
},
{
"epoch": 1.4211214647703123,
"grad_norm": 6.25,
"learning_rate": 1.8000683302406038e-05,
"loss": 0.7732,
"step": 2174
},
{
"epoch": 1.4217753800882786,
"grad_norm": 3.609375,
"learning_rate": 1.7997943123430936e-05,
"loss": 0.304,
"step": 2175
},
{
"epoch": 1.422429295406245,
"grad_norm": 5.4375,
"learning_rate": 1.7995201276820182e-05,
"loss": 0.3855,
"step": 2176
},
{
"epoch": 1.4230832107242113,
"grad_norm": 7.21875,
"learning_rate": 1.7992457763145467e-05,
"loss": 0.7412,
"step": 2177
},
{
"epoch": 1.4237371260421776,
"grad_norm": 5.375,
"learning_rate": 1.798971258297883e-05,
"loss": 0.4878,
"step": 2178
},
{
"epoch": 1.424391041360144,
"grad_norm": 3.625,
"learning_rate": 1.7986965736892674e-05,
"loss": 0.344,
"step": 2179
},
{
"epoch": 1.4250449566781103,
"grad_norm": 3.65625,
"learning_rate": 1.7984217225459736e-05,
"loss": 0.1872,
"step": 2180
},
{
"epoch": 1.4256988719960766,
"grad_norm": 4.65625,
"learning_rate": 1.7981467049253096e-05,
"loss": 0.4264,
"step": 2181
},
{
"epoch": 1.426352787314043,
"grad_norm": 4.90625,
"learning_rate": 1.7978715208846197e-05,
"loss": 0.4524,
"step": 2182
},
{
"epoch": 1.427006702632009,
"grad_norm": 5.125,
"learning_rate": 1.797596170481281e-05,
"loss": 0.366,
"step": 2183
},
{
"epoch": 1.4276606179499756,
"grad_norm": 3.953125,
"learning_rate": 1.797320653772707e-05,
"loss": 0.4595,
"step": 2184
},
{
"epoch": 1.4283145332679417,
"grad_norm": 4.96875,
"learning_rate": 1.7970449708163455e-05,
"loss": 0.4422,
"step": 2185
},
{
"epoch": 1.4289684485859082,
"grad_norm": 4.625,
"learning_rate": 1.7967691216696777e-05,
"loss": 0.5651,
"step": 2186
},
{
"epoch": 1.4296223639038743,
"grad_norm": 7.15625,
"learning_rate": 1.7964931063902208e-05,
"loss": 0.6972,
"step": 2187
},
{
"epoch": 1.4302762792218409,
"grad_norm": 6.25,
"learning_rate": 1.7962169250355262e-05,
"loss": 0.5443,
"step": 2188
},
{
"epoch": 1.430930194539807,
"grad_norm": 5.6875,
"learning_rate": 1.79594057766318e-05,
"loss": 0.8491,
"step": 2189
},
{
"epoch": 1.4315841098577735,
"grad_norm": 4.25,
"learning_rate": 1.7956640643308028e-05,
"loss": 0.2673,
"step": 2190
},
{
"epoch": 1.4322380251757396,
"grad_norm": 3.5,
"learning_rate": 1.7953873850960497e-05,
"loss": 0.484,
"step": 2191
},
{
"epoch": 1.4328919404937062,
"grad_norm": 6.21875,
"learning_rate": 1.795110540016611e-05,
"loss": 0.6586,
"step": 2192
},
{
"epoch": 1.4335458558116723,
"grad_norm": 4.09375,
"learning_rate": 1.7948335291502104e-05,
"loss": 0.6402,
"step": 2193
},
{
"epoch": 1.4341997711296388,
"grad_norm": 4.25,
"learning_rate": 1.7945563525546077e-05,
"loss": 0.4193,
"step": 2194
},
{
"epoch": 1.434853686447605,
"grad_norm": 5.125,
"learning_rate": 1.794279010287596e-05,
"loss": 0.4139,
"step": 2195
},
{
"epoch": 1.4355076017655715,
"grad_norm": 5.5,
"learning_rate": 1.7940015024070036e-05,
"loss": 0.4432,
"step": 2196
},
{
"epoch": 1.4361615170835376,
"grad_norm": 4.125,
"learning_rate": 1.7937238289706927e-05,
"loss": 0.4024,
"step": 2197
},
{
"epoch": 1.4368154324015041,
"grad_norm": 5.53125,
"learning_rate": 1.793445990036561e-05,
"loss": 0.314,
"step": 2198
},
{
"epoch": 1.4374693477194702,
"grad_norm": 5.75,
"learning_rate": 1.7931679856625398e-05,
"loss": 0.5456,
"step": 2199
},
{
"epoch": 1.4381232630374368,
"grad_norm": 7.03125,
"learning_rate": 1.7928898159065955e-05,
"loss": 0.5321,
"step": 2200
},
{
"epoch": 1.4387771783554029,
"grad_norm": 7.0625,
"learning_rate": 1.7926114808267284e-05,
"loss": 0.3437,
"step": 2201
},
{
"epoch": 1.4394310936733694,
"grad_norm": 4.34375,
"learning_rate": 1.7923329804809745e-05,
"loss": 0.4818,
"step": 2202
},
{
"epoch": 1.4400850089913355,
"grad_norm": 5.5,
"learning_rate": 1.792054314927402e-05,
"loss": 0.4846,
"step": 2203
},
{
"epoch": 1.440738924309302,
"grad_norm": 3.25,
"learning_rate": 1.791775484224116e-05,
"loss": 0.3393,
"step": 2204
},
{
"epoch": 1.4413928396272682,
"grad_norm": 3.5,
"learning_rate": 1.7914964884292543e-05,
"loss": 0.2651,
"step": 2205
},
{
"epoch": 1.4420467549452347,
"grad_norm": 4.75,
"learning_rate": 1.79121732760099e-05,
"loss": 0.4744,
"step": 2206
},
{
"epoch": 1.4427006702632008,
"grad_norm": 5.625,
"learning_rate": 1.790938001797531e-05,
"loss": 0.5476,
"step": 2207
},
{
"epoch": 1.4433545855811674,
"grad_norm": 4.125,
"learning_rate": 1.790658511077118e-05,
"loss": 0.4721,
"step": 2208
},
{
"epoch": 1.4440085008991335,
"grad_norm": 6.25,
"learning_rate": 1.7903788554980273e-05,
"loss": 0.3127,
"step": 2209
},
{
"epoch": 1.4446624162171,
"grad_norm": 3.9375,
"learning_rate": 1.7900990351185697e-05,
"loss": 0.2795,
"step": 2210
},
{
"epoch": 1.4453163315350661,
"grad_norm": 3.859375,
"learning_rate": 1.7898190499970897e-05,
"loss": 0.3123,
"step": 2211
},
{
"epoch": 1.4459702468530327,
"grad_norm": 5.46875,
"learning_rate": 1.789538900191967e-05,
"loss": 0.5746,
"step": 2212
},
{
"epoch": 1.4466241621709988,
"grad_norm": 4.09375,
"learning_rate": 1.7892585857616144e-05,
"loss": 0.3456,
"step": 2213
},
{
"epoch": 1.4472780774889653,
"grad_norm": 4.21875,
"learning_rate": 1.78897810676448e-05,
"loss": 0.4941,
"step": 2214
},
{
"epoch": 1.4479319928069314,
"grad_norm": 4.75,
"learning_rate": 1.788697463259046e-05,
"loss": 0.4685,
"step": 2215
},
{
"epoch": 1.448585908124898,
"grad_norm": 2.765625,
"learning_rate": 1.788416655303829e-05,
"loss": 0.2928,
"step": 2216
},
{
"epoch": 1.449239823442864,
"grad_norm": 5.625,
"learning_rate": 1.788135682957379e-05,
"loss": 0.3706,
"step": 2217
},
{
"epoch": 1.4498937387608306,
"grad_norm": 5.96875,
"learning_rate": 1.787854546278282e-05,
"loss": 0.3379,
"step": 2218
},
{
"epoch": 1.4505476540787967,
"grad_norm": 5.28125,
"learning_rate": 1.787573245325156e-05,
"loss": 0.6744,
"step": 2219
},
{
"epoch": 1.4512015693967633,
"grad_norm": 9.6875,
"learning_rate": 1.787291780156656e-05,
"loss": 0.7083,
"step": 2220
},
{
"epoch": 1.4518554847147294,
"grad_norm": 3.15625,
"learning_rate": 1.7870101508314686e-05,
"loss": 0.2939,
"step": 2221
},
{
"epoch": 1.4525094000326957,
"grad_norm": 4.15625,
"learning_rate": 1.7867283574083164e-05,
"loss": 0.3724,
"step": 2222
},
{
"epoch": 1.453163315350662,
"grad_norm": 3.953125,
"learning_rate": 1.7864463999459555e-05,
"loss": 0.2478,
"step": 2223
},
{
"epoch": 1.4538172306686283,
"grad_norm": 5.4375,
"learning_rate": 1.7861642785031765e-05,
"loss": 0.4748,
"step": 2224
},
{
"epoch": 1.4544711459865947,
"grad_norm": 4.28125,
"learning_rate": 1.785881993138803e-05,
"loss": 0.2419,
"step": 2225
},
{
"epoch": 1.455125061304561,
"grad_norm": 4.9375,
"learning_rate": 1.7855995439116953e-05,
"loss": 0.4652,
"step": 2226
},
{
"epoch": 1.4557789766225273,
"grad_norm": 5.125,
"learning_rate": 1.785316930880745e-05,
"loss": 0.438,
"step": 2227
},
{
"epoch": 1.4564328919404936,
"grad_norm": 3.828125,
"learning_rate": 1.78503415410488e-05,
"loss": 0.307,
"step": 2228
},
{
"epoch": 1.45708680725846,
"grad_norm": 5.5625,
"learning_rate": 1.7847512136430612e-05,
"loss": 0.4072,
"step": 2229
},
{
"epoch": 1.4577407225764263,
"grad_norm": 5.65625,
"learning_rate": 1.784468109554284e-05,
"loss": 0.3049,
"step": 2230
},
{
"epoch": 1.4583946378943926,
"grad_norm": 3.828125,
"learning_rate": 1.784184841897578e-05,
"loss": 0.391,
"step": 2231
},
{
"epoch": 1.459048553212359,
"grad_norm": 4.8125,
"learning_rate": 1.783901410732007e-05,
"loss": 0.3814,
"step": 2232
},
{
"epoch": 1.4597024685303253,
"grad_norm": 4.5,
"learning_rate": 1.7836178161166677e-05,
"loss": 0.4544,
"step": 2233
},
{
"epoch": 1.4603563838482916,
"grad_norm": 4.71875,
"learning_rate": 1.783334058110693e-05,
"loss": 0.3616,
"step": 2234
},
{
"epoch": 1.461010299166258,
"grad_norm": 4.5,
"learning_rate": 1.7830501367732484e-05,
"loss": 0.3221,
"step": 2235
},
{
"epoch": 1.4616642144842242,
"grad_norm": 4.6875,
"learning_rate": 1.7827660521635336e-05,
"loss": 0.3404,
"step": 2236
},
{
"epoch": 1.4623181298021906,
"grad_norm": 5.8125,
"learning_rate": 1.7824818043407828e-05,
"loss": 0.433,
"step": 2237
},
{
"epoch": 1.4629720451201569,
"grad_norm": 3.84375,
"learning_rate": 1.7821973933642635e-05,
"loss": 0.1666,
"step": 2238
},
{
"epoch": 1.4636259604381232,
"grad_norm": 5.59375,
"learning_rate": 1.781912819293278e-05,
"loss": 0.4518,
"step": 2239
},
{
"epoch": 1.4642798757560895,
"grad_norm": 5.21875,
"learning_rate": 1.7816280821871626e-05,
"loss": 0.4846,
"step": 2240
},
{
"epoch": 1.4649337910740559,
"grad_norm": 5.0625,
"learning_rate": 1.7813431821052867e-05,
"loss": 0.5108,
"step": 2241
},
{
"epoch": 1.4655877063920222,
"grad_norm": 4.71875,
"learning_rate": 1.7810581191070545e-05,
"loss": 0.4071,
"step": 2242
},
{
"epoch": 1.4662416217099885,
"grad_norm": 4.1875,
"learning_rate": 1.7807728932519038e-05,
"loss": 0.2578,
"step": 2243
},
{
"epoch": 1.4668955370279548,
"grad_norm": 6.75,
"learning_rate": 1.7804875045993067e-05,
"loss": 0.6314,
"step": 2244
},
{
"epoch": 1.4675494523459212,
"grad_norm": 4.5,
"learning_rate": 1.7802019532087692e-05,
"loss": 0.3506,
"step": 2245
},
{
"epoch": 1.4682033676638875,
"grad_norm": 5.0625,
"learning_rate": 1.779916239139831e-05,
"loss": 0.4146,
"step": 2246
},
{
"epoch": 1.4688572829818538,
"grad_norm": 7.84375,
"learning_rate": 1.7796303624520654e-05,
"loss": 0.754,
"step": 2247
},
{
"epoch": 1.4695111982998201,
"grad_norm": 4.5,
"learning_rate": 1.77934432320508e-05,
"loss": 0.3048,
"step": 2248
},
{
"epoch": 1.4701651136177865,
"grad_norm": 4.25,
"learning_rate": 1.779058121458517e-05,
"loss": 0.3095,
"step": 2249
},
{
"epoch": 1.4708190289357528,
"grad_norm": 5.0,
"learning_rate": 1.778771757272051e-05,
"loss": 0.5161,
"step": 2250
},
{
"epoch": 1.471472944253719,
"grad_norm": 5.65625,
"learning_rate": 1.7784852307053914e-05,
"loss": 0.4457,
"step": 2251
},
{
"epoch": 1.4721268595716854,
"grad_norm": 4.4375,
"learning_rate": 1.7781985418182818e-05,
"loss": 0.6011,
"step": 2252
},
{
"epoch": 1.4727807748896518,
"grad_norm": 6.5625,
"learning_rate": 1.7779116906704986e-05,
"loss": 0.5325,
"step": 2253
},
{
"epoch": 1.473434690207618,
"grad_norm": 4.53125,
"learning_rate": 1.777624677321853e-05,
"loss": 0.4167,
"step": 2254
},
{
"epoch": 1.4740886055255844,
"grad_norm": 4.0625,
"learning_rate": 1.777337501832189e-05,
"loss": 0.3552,
"step": 2255
},
{
"epoch": 1.4747425208435507,
"grad_norm": 4.71875,
"learning_rate": 1.7770501642613853e-05,
"loss": 0.4675,
"step": 2256
},
{
"epoch": 1.475396436161517,
"grad_norm": 3.640625,
"learning_rate": 1.7767626646693545e-05,
"loss": 0.5097,
"step": 2257
},
{
"epoch": 1.4760503514794834,
"grad_norm": 3.5625,
"learning_rate": 1.7764750031160417e-05,
"loss": 0.3028,
"step": 2258
},
{
"epoch": 1.4767042667974497,
"grad_norm": 3.625,
"learning_rate": 1.7761871796614278e-05,
"loss": 0.3012,
"step": 2259
},
{
"epoch": 1.477358182115416,
"grad_norm": 5.28125,
"learning_rate": 1.7758991943655255e-05,
"loss": 0.3402,
"step": 2260
},
{
"epoch": 1.4780120974333824,
"grad_norm": 7.15625,
"learning_rate": 1.7756110472883822e-05,
"loss": 0.6943,
"step": 2261
},
{
"epoch": 1.4786660127513487,
"grad_norm": 2.125,
"learning_rate": 1.7753227384900792e-05,
"loss": 0.1434,
"step": 2262
},
{
"epoch": 1.479319928069315,
"grad_norm": 3.390625,
"learning_rate": 1.775034268030731e-05,
"loss": 0.3775,
"step": 2263
},
{
"epoch": 1.4799738433872813,
"grad_norm": 4.53125,
"learning_rate": 1.7747456359704856e-05,
"loss": 0.3696,
"step": 2264
},
{
"epoch": 1.4806277587052477,
"grad_norm": 5.0625,
"learning_rate": 1.7744568423695257e-05,
"loss": 0.5307,
"step": 2265
},
{
"epoch": 1.481281674023214,
"grad_norm": 3.34375,
"learning_rate": 1.774167887288067e-05,
"loss": 0.2492,
"step": 2266
},
{
"epoch": 1.4819355893411803,
"grad_norm": 3.234375,
"learning_rate": 1.773878770786359e-05,
"loss": 0.2144,
"step": 2267
},
{
"epoch": 1.4825895046591466,
"grad_norm": 4.0,
"learning_rate": 1.7735894929246844e-05,
"loss": 0.2424,
"step": 2268
},
{
"epoch": 1.483243419977113,
"grad_norm": 2.671875,
"learning_rate": 1.7733000537633605e-05,
"loss": 0.1398,
"step": 2269
},
{
"epoch": 1.4838973352950793,
"grad_norm": 5.21875,
"learning_rate": 1.773010453362737e-05,
"loss": 0.4368,
"step": 2270
},
{
"epoch": 1.4845512506130456,
"grad_norm": 5.0625,
"learning_rate": 1.7727206917831988e-05,
"loss": 0.3235,
"step": 2271
},
{
"epoch": 1.485205165931012,
"grad_norm": 4.0625,
"learning_rate": 1.772430769085163e-05,
"loss": 0.3026,
"step": 2272
},
{
"epoch": 1.4858590812489783,
"grad_norm": 4.0,
"learning_rate": 1.772140685329081e-05,
"loss": 0.2587,
"step": 2273
},
{
"epoch": 1.4865129965669446,
"grad_norm": 5.875,
"learning_rate": 1.7718504405754373e-05,
"loss": 0.5082,
"step": 2274
},
{
"epoch": 1.487166911884911,
"grad_norm": 5.375,
"learning_rate": 1.7715600348847508e-05,
"loss": 0.5812,
"step": 2275
},
{
"epoch": 1.4878208272028772,
"grad_norm": 4.6875,
"learning_rate": 1.771269468317573e-05,
"loss": 0.3621,
"step": 2276
},
{
"epoch": 1.4884747425208436,
"grad_norm": 3.53125,
"learning_rate": 1.7709787409344893e-05,
"loss": 0.2482,
"step": 2277
},
{
"epoch": 1.4891286578388099,
"grad_norm": 3.828125,
"learning_rate": 1.770687852796119e-05,
"loss": 0.2793,
"step": 2278
},
{
"epoch": 1.4897825731567762,
"grad_norm": 3.484375,
"learning_rate": 1.7703968039631148e-05,
"loss": 0.2115,
"step": 2279
},
{
"epoch": 1.4904364884747425,
"grad_norm": 5.0625,
"learning_rate": 1.7701055944961624e-05,
"loss": 0.4724,
"step": 2280
},
{
"epoch": 1.4910904037927089,
"grad_norm": 6.0625,
"learning_rate": 1.7698142244559812e-05,
"loss": 0.3141,
"step": 2281
},
{
"epoch": 1.4917443191106752,
"grad_norm": 4.21875,
"learning_rate": 1.7695226939033242e-05,
"loss": 0.3516,
"step": 2282
},
{
"epoch": 1.4923982344286415,
"grad_norm": 6.5,
"learning_rate": 1.7692310028989785e-05,
"loss": 0.6596,
"step": 2283
},
{
"epoch": 1.4930521497466078,
"grad_norm": 3.359375,
"learning_rate": 1.768939151503763e-05,
"loss": 0.1663,
"step": 2284
},
{
"epoch": 1.4937060650645742,
"grad_norm": 3.875,
"learning_rate": 1.7686471397785322e-05,
"loss": 0.2344,
"step": 2285
},
{
"epoch": 1.4943599803825405,
"grad_norm": 6.59375,
"learning_rate": 1.7683549677841723e-05,
"loss": 0.5738,
"step": 2286
},
{
"epoch": 1.4950138957005068,
"grad_norm": 3.359375,
"learning_rate": 1.7680626355816036e-05,
"loss": 0.2684,
"step": 2287
},
{
"epoch": 1.4956678110184731,
"grad_norm": 4.34375,
"learning_rate": 1.7677701432317794e-05,
"loss": 0.3068,
"step": 2288
},
{
"epoch": 1.4963217263364395,
"grad_norm": 6.84375,
"learning_rate": 1.7674774907956874e-05,
"loss": 0.6352,
"step": 2289
},
{
"epoch": 1.4969756416544058,
"grad_norm": 5.40625,
"learning_rate": 1.767184678334347e-05,
"loss": 0.3149,
"step": 2290
},
{
"epoch": 1.497629556972372,
"grad_norm": 2.578125,
"learning_rate": 1.766891705908813e-05,
"loss": 0.2009,
"step": 2291
},
{
"epoch": 1.4982834722903384,
"grad_norm": 3.53125,
"learning_rate": 1.766598573580172e-05,
"loss": 0.2665,
"step": 2292
},
{
"epoch": 1.4989373876083047,
"grad_norm": 4.375,
"learning_rate": 1.7663052814095447e-05,
"loss": 0.4633,
"step": 2293
},
{
"epoch": 1.499591302926271,
"grad_norm": 5.65625,
"learning_rate": 1.7660118294580844e-05,
"loss": 0.4406,
"step": 2294
},
{
"epoch": 1.5002452182442374,
"grad_norm": 3.90625,
"learning_rate": 1.765718217786979e-05,
"loss": 0.2999,
"step": 2295
},
{
"epoch": 1.5008991335622037,
"grad_norm": 4.96875,
"learning_rate": 1.765424446457448e-05,
"loss": 0.2885,
"step": 2296
},
{
"epoch": 1.50155304888017,
"grad_norm": 5.78125,
"learning_rate": 1.7651305155307454e-05,
"loss": 0.5739,
"step": 2297
},
{
"epoch": 1.5022069641981364,
"grad_norm": 4.59375,
"learning_rate": 1.7648364250681583e-05,
"loss": 0.2914,
"step": 2298
},
{
"epoch": 1.5028608795161027,
"grad_norm": 6.28125,
"learning_rate": 1.7645421751310072e-05,
"loss": 0.6361,
"step": 2299
},
{
"epoch": 1.503514794834069,
"grad_norm": 6.71875,
"learning_rate": 1.7642477657806454e-05,
"loss": 0.5598,
"step": 2300
},
{
"epoch": 1.5041687101520353,
"grad_norm": 8.1875,
"learning_rate": 1.7639531970784594e-05,
"loss": 0.7778,
"step": 2301
},
{
"epoch": 1.5048226254700017,
"grad_norm": 4.28125,
"learning_rate": 1.763658469085869e-05,
"loss": 0.3894,
"step": 2302
},
{
"epoch": 1.505476540787968,
"grad_norm": 5.1875,
"learning_rate": 1.763363581864328e-05,
"loss": 0.4134,
"step": 2303
},
{
"epoch": 1.5061304561059343,
"grad_norm": 5.25,
"learning_rate": 1.763068535475322e-05,
"loss": 0.3368,
"step": 2304
},
{
"epoch": 1.5067843714239006,
"grad_norm": 4.53125,
"learning_rate": 1.7627733299803714e-05,
"loss": 0.5108,
"step": 2305
},
{
"epoch": 1.507438286741867,
"grad_norm": 3.484375,
"learning_rate": 1.7624779654410283e-05,
"loss": 0.223,
"step": 2306
},
{
"epoch": 1.5080922020598333,
"grad_norm": 4.15625,
"learning_rate": 1.7621824419188786e-05,
"loss": 0.3185,
"step": 2307
},
{
"epoch": 1.5087461173777996,
"grad_norm": 3.71875,
"learning_rate": 1.7618867594755417e-05,
"loss": 0.2575,
"step": 2308
},
{
"epoch": 1.509400032695766,
"grad_norm": 5.0,
"learning_rate": 1.7615909181726697e-05,
"loss": 0.6067,
"step": 2309
},
{
"epoch": 1.5100539480137323,
"grad_norm": 5.3125,
"learning_rate": 1.7612949180719473e-05,
"loss": 0.6687,
"step": 2310
},
{
"epoch": 1.5107078633316986,
"grad_norm": 4.125,
"learning_rate": 1.760998759235094e-05,
"loss": 0.2738,
"step": 2311
},
{
"epoch": 1.511361778649665,
"grad_norm": 3.484375,
"learning_rate": 1.76070244172386e-05,
"loss": 0.2532,
"step": 2312
},
{
"epoch": 1.5120156939676312,
"grad_norm": 5.09375,
"learning_rate": 1.7604059656000313e-05,
"loss": 0.5349,
"step": 2313
},
{
"epoch": 1.5126696092855973,
"grad_norm": 4.1875,
"learning_rate": 1.7601093309254244e-05,
"loss": 0.3029,
"step": 2314
},
{
"epoch": 1.513323524603564,
"grad_norm": 5.65625,
"learning_rate": 1.7598125377618904e-05,
"loss": 0.5348,
"step": 2315
},
{
"epoch": 1.51397743992153,
"grad_norm": 4.1875,
"learning_rate": 1.7595155861713134e-05,
"loss": 0.2674,
"step": 2316
},
{
"epoch": 1.5146313552394965,
"grad_norm": 4.46875,
"learning_rate": 1.75921847621561e-05,
"loss": 0.5274,
"step": 2317
},
{
"epoch": 1.5152852705574626,
"grad_norm": 3.203125,
"learning_rate": 1.7589212079567297e-05,
"loss": 0.2741,
"step": 2318
},
{
"epoch": 1.5159391858754292,
"grad_norm": 3.015625,
"learning_rate": 1.758623781456656e-05,
"loss": 0.2318,
"step": 2319
},
{
"epoch": 1.5165931011933953,
"grad_norm": 3.90625,
"learning_rate": 1.758326196777404e-05,
"loss": 0.3329,
"step": 2320
},
{
"epoch": 1.5172470165113618,
"grad_norm": 4.03125,
"learning_rate": 1.758028453981023e-05,
"loss": 0.2747,
"step": 2321
},
{
"epoch": 1.517900931829328,
"grad_norm": 3.5625,
"learning_rate": 1.7577305531295943e-05,
"loss": 0.373,
"step": 2322
},
{
"epoch": 1.5185548471472945,
"grad_norm": 4.46875,
"learning_rate": 1.757432494285233e-05,
"loss": 0.4361,
"step": 2323
},
{
"epoch": 1.5192087624652606,
"grad_norm": 4.90625,
"learning_rate": 1.757134277510087e-05,
"loss": 0.3804,
"step": 2324
},
{
"epoch": 1.5198626777832271,
"grad_norm": 4.34375,
"learning_rate": 1.7568359028663365e-05,
"loss": 0.3011,
"step": 2325
},
{
"epoch": 1.5205165931011932,
"grad_norm": 4.375,
"learning_rate": 1.756537370416195e-05,
"loss": 0.3906,
"step": 2326
},
{
"epoch": 1.5211705084191598,
"grad_norm": 6.96875,
"learning_rate": 1.7562386802219092e-05,
"loss": 0.8485,
"step": 2327
},
{
"epoch": 1.521824423737126,
"grad_norm": 5.0,
"learning_rate": 1.755939832345758e-05,
"loss": 0.5227,
"step": 2328
},
{
"epoch": 1.5224783390550924,
"grad_norm": 5.0625,
"learning_rate": 1.7556408268500542e-05,
"loss": 0.351,
"step": 2329
},
{
"epoch": 1.5231322543730585,
"grad_norm": 9.4375,
"learning_rate": 1.7553416637971424e-05,
"loss": 0.9022,
"step": 2330
},
{
"epoch": 1.523786169691025,
"grad_norm": 4.71875,
"learning_rate": 1.7550423432494007e-05,
"loss": 0.3311,
"step": 2331
},
{
"epoch": 1.5244400850089912,
"grad_norm": 5.8125,
"learning_rate": 1.75474286526924e-05,
"loss": 0.6049,
"step": 2332
},
{
"epoch": 1.5250940003269577,
"grad_norm": 4.53125,
"learning_rate": 1.754443229919103e-05,
"loss": 0.5469,
"step": 2333
},
{
"epoch": 1.5257479156449238,
"grad_norm": 3.78125,
"learning_rate": 1.754143437261467e-05,
"loss": 0.2707,
"step": 2334
},
{
"epoch": 1.5264018309628904,
"grad_norm": 2.90625,
"learning_rate": 1.753843487358841e-05,
"loss": 0.2334,
"step": 2335
},
{
"epoch": 1.5270557462808565,
"grad_norm": 3.046875,
"learning_rate": 1.753543380273767e-05,
"loss": 0.1857,
"step": 2336
},
{
"epoch": 1.527709661598823,
"grad_norm": 6.25,
"learning_rate": 1.7532431160688194e-05,
"loss": 0.8458,
"step": 2337
},
{
"epoch": 1.5283635769167891,
"grad_norm": 6.65625,
"learning_rate": 1.752942694806606e-05,
"loss": 0.6013,
"step": 2338
},
{
"epoch": 1.5290174922347557,
"grad_norm": 4.6875,
"learning_rate": 1.7526421165497674e-05,
"loss": 0.3637,
"step": 2339
},
{
"epoch": 1.5296714075527218,
"grad_norm": 3.90625,
"learning_rate": 1.7523413813609757e-05,
"loss": 0.2658,
"step": 2340
},
{
"epoch": 1.5303253228706883,
"grad_norm": 5.75,
"learning_rate": 1.7520404893029374e-05,
"loss": 0.5556,
"step": 2341
},
{
"epoch": 1.5309792381886544,
"grad_norm": 4.71875,
"learning_rate": 1.7517394404383906e-05,
"loss": 0.4663,
"step": 2342
},
{
"epoch": 1.531633153506621,
"grad_norm": 5.96875,
"learning_rate": 1.7514382348301063e-05,
"loss": 0.4059,
"step": 2343
},
{
"epoch": 1.532287068824587,
"grad_norm": 7.0,
"learning_rate": 1.7511368725408888e-05,
"loss": 0.4723,
"step": 2344
},
{
"epoch": 1.5329409841425536,
"grad_norm": 5.28125,
"learning_rate": 1.750835353633574e-05,
"loss": 0.5439,
"step": 2345
},
{
"epoch": 1.5335948994605197,
"grad_norm": 4.28125,
"learning_rate": 1.7505336781710314e-05,
"loss": 0.3011,
"step": 2346
},
{
"epoch": 1.5342488147784863,
"grad_norm": 4.375,
"learning_rate": 1.7502318462161626e-05,
"loss": 0.4468,
"step": 2347
},
{
"epoch": 1.5349027300964524,
"grad_norm": 2.953125,
"learning_rate": 1.7499298578319017e-05,
"loss": 0.1775,
"step": 2348
},
{
"epoch": 1.535556645414419,
"grad_norm": 3.25,
"learning_rate": 1.749627713081216e-05,
"loss": 0.3724,
"step": 2349
},
{
"epoch": 1.536210560732385,
"grad_norm": 4.125,
"learning_rate": 1.7493254120271057e-05,
"loss": 0.2553,
"step": 2350
},
{
"epoch": 1.5368644760503516,
"grad_norm": 3.59375,
"learning_rate": 1.7490229547326023e-05,
"loss": 0.309,
"step": 2351
},
{
"epoch": 1.5375183913683177,
"grad_norm": 5.09375,
"learning_rate": 1.74872034126077e-05,
"loss": 0.5343,
"step": 2352
},
{
"epoch": 1.5381723066862842,
"grad_norm": 4.84375,
"learning_rate": 1.7484175716747075e-05,
"loss": 0.9494,
"step": 2353
},
{
"epoch": 1.5388262220042503,
"grad_norm": 5.40625,
"learning_rate": 1.748114646037544e-05,
"loss": 0.3811,
"step": 2354
},
{
"epoch": 1.539480137322217,
"grad_norm": 6.8125,
"learning_rate": 1.747811564412442e-05,
"loss": 0.4373,
"step": 2355
},
{
"epoch": 1.540134052640183,
"grad_norm": 5.78125,
"learning_rate": 1.747508326862597e-05,
"loss": 0.3899,
"step": 2356
},
{
"epoch": 1.5407879679581495,
"grad_norm": 3.015625,
"learning_rate": 1.7472049334512354e-05,
"loss": 0.2166,
"step": 2357
},
{
"epoch": 1.5414418832761156,
"grad_norm": 3.96875,
"learning_rate": 1.746901384241618e-05,
"loss": 0.3114,
"step": 2358
},
{
"epoch": 1.5420957985940822,
"grad_norm": 4.34375,
"learning_rate": 1.746597679297037e-05,
"loss": 0.3474,
"step": 2359
},
{
"epoch": 1.5427497139120483,
"grad_norm": 2.9375,
"learning_rate": 1.7462938186808172e-05,
"loss": 0.2202,
"step": 2360
},
{
"epoch": 1.5434036292300148,
"grad_norm": 4.25,
"learning_rate": 1.745989802456316e-05,
"loss": 0.3439,
"step": 2361
},
{
"epoch": 1.544057544547981,
"grad_norm": 6.625,
"learning_rate": 1.7456856306869234e-05,
"loss": 0.5385,
"step": 2362
},
{
"epoch": 1.5447114598659475,
"grad_norm": 6.71875,
"learning_rate": 1.7453813034360618e-05,
"loss": 0.8461,
"step": 2363
},
{
"epoch": 1.5453653751839136,
"grad_norm": 4.6875,
"learning_rate": 1.7450768207671857e-05,
"loss": 0.4076,
"step": 2364
},
{
"epoch": 1.5460192905018801,
"grad_norm": 3.46875,
"learning_rate": 1.744772182743782e-05,
"loss": 0.2539,
"step": 2365
},
{
"epoch": 1.5466732058198462,
"grad_norm": 4.28125,
"learning_rate": 1.7444673894293705e-05,
"loss": 0.3514,
"step": 2366
},
{
"epoch": 1.5473271211378128,
"grad_norm": 4.6875,
"learning_rate": 1.744162440887503e-05,
"loss": 0.2855,
"step": 2367
},
{
"epoch": 1.547981036455779,
"grad_norm": 4.625,
"learning_rate": 1.7438573371817637e-05,
"loss": 0.3815,
"step": 2368
},
{
"epoch": 1.5486349517737454,
"grad_norm": 4.46875,
"learning_rate": 1.7435520783757693e-05,
"loss": 0.3623,
"step": 2369
},
{
"epoch": 1.5492888670917115,
"grad_norm": 4.875,
"learning_rate": 1.743246664533168e-05,
"loss": 0.4816,
"step": 2370
},
{
"epoch": 1.549942782409678,
"grad_norm": 4.125,
"learning_rate": 1.7429410957176423e-05,
"loss": 0.3022,
"step": 2371
},
{
"epoch": 1.5505966977276442,
"grad_norm": 4.875,
"learning_rate": 1.7426353719929048e-05,
"loss": 0.372,
"step": 2372
},
{
"epoch": 1.5512506130456107,
"grad_norm": 6.65625,
"learning_rate": 1.7423294934227017e-05,
"loss": 0.7791,
"step": 2373
},
{
"epoch": 1.5519045283635768,
"grad_norm": 6.21875,
"learning_rate": 1.7420234600708112e-05,
"loss": 0.4332,
"step": 2374
},
{
"epoch": 1.5525584436815434,
"grad_norm": 5.53125,
"learning_rate": 1.7417172720010436e-05,
"loss": 0.4654,
"step": 2375
},
{
"epoch": 1.5532123589995095,
"grad_norm": 5.375,
"learning_rate": 1.7414109292772416e-05,
"loss": 0.4067,
"step": 2376
},
{
"epoch": 1.553866274317476,
"grad_norm": 4.375,
"learning_rate": 1.74110443196328e-05,
"loss": 0.2961,
"step": 2377
},
{
"epoch": 1.5545201896354421,
"grad_norm": 6.0625,
"learning_rate": 1.7407977801230664e-05,
"loss": 0.6084,
"step": 2378
},
{
"epoch": 1.5551741049534087,
"grad_norm": 4.0625,
"learning_rate": 1.7404909738205397e-05,
"loss": 0.3694,
"step": 2379
},
{
"epoch": 1.5558280202713748,
"grad_norm": 6.5,
"learning_rate": 1.7401840131196716e-05,
"loss": 0.6468,
"step": 2380
},
{
"epoch": 1.5564819355893413,
"grad_norm": 5.46875,
"learning_rate": 1.7398768980844664e-05,
"loss": 0.4835,
"step": 2381
},
{
"epoch": 1.5571358509073074,
"grad_norm": 4.53125,
"learning_rate": 1.739569628778959e-05,
"loss": 0.3849,
"step": 2382
},
{
"epoch": 1.557789766225274,
"grad_norm": 5.375,
"learning_rate": 1.7392622052672185e-05,
"loss": 0.6057,
"step": 2383
},
{
"epoch": 1.55844368154324,
"grad_norm": 4.40625,
"learning_rate": 1.738954627613345e-05,
"loss": 0.4948,
"step": 2384
},
{
"epoch": 1.5590975968612066,
"grad_norm": 4.8125,
"learning_rate": 1.7386468958814706e-05,
"loss": 0.4064,
"step": 2385
},
{
"epoch": 1.5597515121791727,
"grad_norm": 3.828125,
"learning_rate": 1.73833901013576e-05,
"loss": 0.3421,
"step": 2386
},
{
"epoch": 1.5604054274971393,
"grad_norm": 4.21875,
"learning_rate": 1.7380309704404096e-05,
"loss": 0.5244,
"step": 2387
},
{
"epoch": 1.5610593428151054,
"grad_norm": 7.9375,
"learning_rate": 1.737722776859649e-05,
"loss": 0.579,
"step": 2388
},
{
"epoch": 1.5617132581330717,
"grad_norm": 4.53125,
"learning_rate": 1.7374144294577382e-05,
"loss": 0.415,
"step": 2389
},
{
"epoch": 1.562367173451038,
"grad_norm": 5.53125,
"learning_rate": 1.7371059282989705e-05,
"loss": 0.6204,
"step": 2390
},
{
"epoch": 1.5630210887690044,
"grad_norm": 3.890625,
"learning_rate": 1.7367972734476706e-05,
"loss": 0.2134,
"step": 2391
},
{
"epoch": 1.5636750040869707,
"grad_norm": 3.796875,
"learning_rate": 1.7364884649681958e-05,
"loss": 0.2046,
"step": 2392
},
{
"epoch": 1.564328919404937,
"grad_norm": 6.65625,
"learning_rate": 1.736179502924935e-05,
"loss": 0.6791,
"step": 2393
},
{
"epoch": 1.5649828347229033,
"grad_norm": 3.65625,
"learning_rate": 1.7358703873823094e-05,
"loss": 0.3279,
"step": 2394
},
{
"epoch": 1.5656367500408697,
"grad_norm": 4.96875,
"learning_rate": 1.7355611184047717e-05,
"loss": 0.3292,
"step": 2395
},
{
"epoch": 1.566290665358836,
"grad_norm": 4.71875,
"learning_rate": 1.735251696056808e-05,
"loss": 0.6034,
"step": 2396
},
{
"epoch": 1.5669445806768023,
"grad_norm": 4.125,
"learning_rate": 1.7349421204029343e-05,
"loss": 0.4317,
"step": 2397
},
{
"epoch": 1.5675984959947686,
"grad_norm": 7.4375,
"learning_rate": 1.7346323915076998e-05,
"loss": 1.0119,
"step": 2398
},
{
"epoch": 1.568252411312735,
"grad_norm": 3.875,
"learning_rate": 1.7343225094356857e-05,
"loss": 0.3956,
"step": 2399
},
{
"epoch": 1.5689063266307013,
"grad_norm": 5.625,
"learning_rate": 1.734012474251505e-05,
"loss": 0.6728,
"step": 2400
},
{
"epoch": 1.5695602419486676,
"grad_norm": 5.0625,
"learning_rate": 1.733702286019802e-05,
"loss": 0.5831,
"step": 2401
},
{
"epoch": 1.570214157266634,
"grad_norm": 4.15625,
"learning_rate": 1.7333919448052547e-05,
"loss": 0.3053,
"step": 2402
},
{
"epoch": 1.5708680725846003,
"grad_norm": 4.75,
"learning_rate": 1.7330814506725702e-05,
"loss": 0.377,
"step": 2403
},
{
"epoch": 1.5715219879025666,
"grad_norm": 3.609375,
"learning_rate": 1.7327708036864902e-05,
"loss": 0.3415,
"step": 2404
},
{
"epoch": 1.572175903220533,
"grad_norm": 5.25,
"learning_rate": 1.7324600039117862e-05,
"loss": 0.3418,
"step": 2405
},
{
"epoch": 1.5728298185384992,
"grad_norm": 5.21875,
"learning_rate": 1.7321490514132634e-05,
"loss": 0.4174,
"step": 2406
},
{
"epoch": 1.5734837338564656,
"grad_norm": 3.6875,
"learning_rate": 1.731837946255757e-05,
"loss": 0.2071,
"step": 2407
},
{
"epoch": 1.5741376491744319,
"grad_norm": 2.484375,
"learning_rate": 1.7315266885041352e-05,
"loss": 0.2235,
"step": 2408
},
{
"epoch": 1.5747915644923982,
"grad_norm": 5.65625,
"learning_rate": 1.7312152782232984e-05,
"loss": 0.5885,
"step": 2409
},
{
"epoch": 1.5754454798103645,
"grad_norm": 5.53125,
"learning_rate": 1.730903715478177e-05,
"loss": 0.5683,
"step": 2410
},
{
"epoch": 1.5760993951283309,
"grad_norm": 3.484375,
"learning_rate": 1.7305920003337355e-05,
"loss": 0.283,
"step": 2411
},
{
"epoch": 1.5767533104462972,
"grad_norm": 5.125,
"learning_rate": 1.7302801328549683e-05,
"loss": 0.331,
"step": 2412
},
{
"epoch": 1.5774072257642635,
"grad_norm": 4.46875,
"learning_rate": 1.7299681131069026e-05,
"loss": 0.4577,
"step": 2413
},
{
"epoch": 1.5780611410822298,
"grad_norm": 6.625,
"learning_rate": 1.7296559411545968e-05,
"loss": 0.4886,
"step": 2414
},
{
"epoch": 1.5787150564001962,
"grad_norm": 3.953125,
"learning_rate": 1.7293436170631414e-05,
"loss": 0.6015,
"step": 2415
},
{
"epoch": 1.5793689717181625,
"grad_norm": 5.46875,
"learning_rate": 1.7290311408976587e-05,
"loss": 0.6845,
"step": 2416
},
{
"epoch": 1.5800228870361288,
"grad_norm": 4.875,
"learning_rate": 1.728718512723302e-05,
"loss": 0.3757,
"step": 2417
},
{
"epoch": 1.5806768023540951,
"grad_norm": 4.03125,
"learning_rate": 1.7284057326052572e-05,
"loss": 0.2758,
"step": 2418
},
{
"epoch": 1.5813307176720615,
"grad_norm": 5.9375,
"learning_rate": 1.7280928006087415e-05,
"loss": 0.3769,
"step": 2419
},
{
"epoch": 1.5819846329900278,
"grad_norm": 4.84375,
"learning_rate": 1.7277797167990033e-05,
"loss": 0.3178,
"step": 2420
},
{
"epoch": 1.582638548307994,
"grad_norm": 4.46875,
"learning_rate": 1.7274664812413238e-05,
"loss": 0.269,
"step": 2421
},
{
"epoch": 1.5832924636259604,
"grad_norm": 3.796875,
"learning_rate": 1.7271530940010145e-05,
"loss": 0.419,
"step": 2422
},
{
"epoch": 1.5839463789439268,
"grad_norm": 3.859375,
"learning_rate": 1.7268395551434195e-05,
"loss": 0.3487,
"step": 2423
},
{
"epoch": 1.584600294261893,
"grad_norm": 6.21875,
"learning_rate": 1.726525864733914e-05,
"loss": 0.3626,
"step": 2424
},
{
"epoch": 1.5852542095798594,
"grad_norm": 4.75,
"learning_rate": 1.7262120228379053e-05,
"loss": 0.2808,
"step": 2425
},
{
"epoch": 1.5859081248978257,
"grad_norm": 5.34375,
"learning_rate": 1.7258980295208313e-05,
"loss": 0.475,
"step": 2426
},
{
"epoch": 1.586562040215792,
"grad_norm": 5.25,
"learning_rate": 1.7255838848481633e-05,
"loss": 0.3498,
"step": 2427
},
{
"epoch": 1.5872159555337584,
"grad_norm": 4.5625,
"learning_rate": 1.7252695888854018e-05,
"loss": 0.3264,
"step": 2428
},
{
"epoch": 1.5878698708517247,
"grad_norm": 5.5,
"learning_rate": 1.7249551416980806e-05,
"loss": 0.6514,
"step": 2429
},
{
"epoch": 1.588523786169691,
"grad_norm": 4.21875,
"learning_rate": 1.724640543351764e-05,
"loss": 0.4232,
"step": 2430
},
{
"epoch": 1.5891777014876574,
"grad_norm": 3.375,
"learning_rate": 1.7243257939120493e-05,
"loss": 0.251,
"step": 2431
},
{
"epoch": 1.5898316168056237,
"grad_norm": 7.21875,
"learning_rate": 1.7240108934445636e-05,
"loss": 0.6617,
"step": 2432
},
{
"epoch": 1.59048553212359,
"grad_norm": 5.78125,
"learning_rate": 1.723695842014966e-05,
"loss": 0.5887,
"step": 2433
},
{
"epoch": 1.5911394474415563,
"grad_norm": 4.5,
"learning_rate": 1.723380639688948e-05,
"loss": 0.3874,
"step": 2434
},
{
"epoch": 1.5917933627595227,
"grad_norm": 3.109375,
"learning_rate": 1.723065286532231e-05,
"loss": 0.2615,
"step": 2435
},
{
"epoch": 1.592447278077489,
"grad_norm": 8.1875,
"learning_rate": 1.722749782610569e-05,
"loss": 0.978,
"step": 2436
},
{
"epoch": 1.5931011933954553,
"grad_norm": 7.28125,
"learning_rate": 1.7224341279897473e-05,
"loss": 0.7855,
"step": 2437
},
{
"epoch": 1.5937551087134216,
"grad_norm": 3.515625,
"learning_rate": 1.722118322735582e-05,
"loss": 0.2365,
"step": 2438
},
{
"epoch": 1.594409024031388,
"grad_norm": 4.5625,
"learning_rate": 1.721802366913922e-05,
"loss": 0.2626,
"step": 2439
},
{
"epoch": 1.5950629393493543,
"grad_norm": 3.71875,
"learning_rate": 1.7214862605906458e-05,
"loss": 0.2982,
"step": 2440
},
{
"epoch": 1.5957168546673206,
"grad_norm": 5.78125,
"learning_rate": 1.721170003831664e-05,
"loss": 0.4901,
"step": 2441
},
{
"epoch": 1.596370769985287,
"grad_norm": 3.6875,
"learning_rate": 1.720853596702919e-05,
"loss": 0.3754,
"step": 2442
},
{
"epoch": 1.5970246853032533,
"grad_norm": 4.0625,
"learning_rate": 1.7205370392703842e-05,
"loss": 0.2826,
"step": 2443
},
{
"epoch": 1.5976786006212196,
"grad_norm": 4.65625,
"learning_rate": 1.7202203316000648e-05,
"loss": 0.3349,
"step": 2444
},
{
"epoch": 1.598332515939186,
"grad_norm": 3.421875,
"learning_rate": 1.7199034737579962e-05,
"loss": 0.2709,
"step": 2445
},
{
"epoch": 1.5989864312571522,
"grad_norm": 4.15625,
"learning_rate": 1.719586465810246e-05,
"loss": 0.4441,
"step": 2446
},
{
"epoch": 1.5996403465751186,
"grad_norm": 3.90625,
"learning_rate": 1.7192693078229134e-05,
"loss": 0.2848,
"step": 2447
},
{
"epoch": 1.6002942618930849,
"grad_norm": 3.53125,
"learning_rate": 1.7189519998621274e-05,
"loss": 0.2701,
"step": 2448
},
{
"epoch": 1.6009481772110512,
"grad_norm": 3.25,
"learning_rate": 1.7186345419940505e-05,
"loss": 0.2838,
"step": 2449
},
{
"epoch": 1.6016020925290175,
"grad_norm": 4.0,
"learning_rate": 1.7183169342848742e-05,
"loss": 0.3376,
"step": 2450
},
{
"epoch": 1.6022560078469839,
"grad_norm": 5.0625,
"learning_rate": 1.717999176800823e-05,
"loss": 0.4781,
"step": 2451
},
{
"epoch": 1.6029099231649502,
"grad_norm": 3.6875,
"learning_rate": 1.7176812696081516e-05,
"loss": 0.313,
"step": 2452
},
{
"epoch": 1.6035638384829165,
"grad_norm": 5.90625,
"learning_rate": 1.7173632127731462e-05,
"loss": 0.5668,
"step": 2453
},
{
"epoch": 1.6042177538008828,
"grad_norm": 4.4375,
"learning_rate": 1.717045006362124e-05,
"loss": 0.4045,
"step": 2454
},
{
"epoch": 1.6048716691188492,
"grad_norm": 4.96875,
"learning_rate": 1.7167266504414342e-05,
"loss": 0.539,
"step": 2455
},
{
"epoch": 1.6055255844368155,
"grad_norm": 4.53125,
"learning_rate": 1.7164081450774558e-05,
"loss": 0.3935,
"step": 2456
},
{
"epoch": 1.6061794997547818,
"grad_norm": 4.9375,
"learning_rate": 1.7160894903366006e-05,
"loss": 0.5052,
"step": 2457
},
{
"epoch": 1.6068334150727481,
"grad_norm": 4.96875,
"learning_rate": 1.7157706862853097e-05,
"loss": 0.534,
"step": 2458
},
{
"epoch": 1.6074873303907145,
"grad_norm": 3.609375,
"learning_rate": 1.715451732990057e-05,
"loss": 0.3001,
"step": 2459
},
{
"epoch": 1.6081412457086808,
"grad_norm": 6.53125,
"learning_rate": 1.715132630517347e-05,
"loss": 0.7137,
"step": 2460
},
{
"epoch": 1.608795161026647,
"grad_norm": 5.5625,
"learning_rate": 1.7148133789337145e-05,
"loss": 0.6898,
"step": 2461
},
{
"epoch": 1.6094490763446134,
"grad_norm": 4.09375,
"learning_rate": 1.7144939783057263e-05,
"loss": 0.3301,
"step": 2462
},
{
"epoch": 1.6101029916625795,
"grad_norm": 3.890625,
"learning_rate": 1.7141744286999803e-05,
"loss": 0.1765,
"step": 2463
},
{
"epoch": 1.610756906980546,
"grad_norm": 6.1875,
"learning_rate": 1.713854730183105e-05,
"loss": 0.6463,
"step": 2464
},
{
"epoch": 1.6114108222985122,
"grad_norm": 5.96875,
"learning_rate": 1.71353488282176e-05,
"loss": 0.4241,
"step": 2465
},
{
"epoch": 1.6120647376164787,
"grad_norm": 4.625,
"learning_rate": 1.713214886682636e-05,
"loss": 0.2543,
"step": 2466
},
{
"epoch": 1.6127186529344448,
"grad_norm": 3.671875,
"learning_rate": 1.7128947418324548e-05,
"loss": 0.4509,
"step": 2467
},
{
"epoch": 1.6133725682524114,
"grad_norm": 6.75,
"learning_rate": 1.7125744483379692e-05,
"loss": 0.6877,
"step": 2468
},
{
"epoch": 1.6140264835703775,
"grad_norm": 4.09375,
"learning_rate": 1.7122540062659632e-05,
"loss": 0.2936,
"step": 2469
},
{
"epoch": 1.614680398888344,
"grad_norm": 5.5625,
"learning_rate": 1.7119334156832514e-05,
"loss": 0.5913,
"step": 2470
},
{
"epoch": 1.6153343142063101,
"grad_norm": 6.96875,
"learning_rate": 1.7116126766566794e-05,
"loss": 0.7757,
"step": 2471
},
{
"epoch": 1.6159882295242767,
"grad_norm": 6.0,
"learning_rate": 1.711291789253124e-05,
"loss": 0.5675,
"step": 2472
},
{
"epoch": 1.6166421448422428,
"grad_norm": 4.53125,
"learning_rate": 1.710970753539493e-05,
"loss": 0.5969,
"step": 2473
},
{
"epoch": 1.6172960601602093,
"grad_norm": 4.46875,
"learning_rate": 1.710649569582725e-05,
"loss": 0.3588,
"step": 2474
},
{
"epoch": 1.6179499754781754,
"grad_norm": 8.6875,
"learning_rate": 1.7103282374497884e-05,
"loss": 0.6367,
"step": 2475
},
{
"epoch": 1.618603890796142,
"grad_norm": 3.984375,
"learning_rate": 1.710006757207685e-05,
"loss": 0.2543,
"step": 2476
},
{
"epoch": 1.619257806114108,
"grad_norm": 4.21875,
"learning_rate": 1.7096851289234448e-05,
"loss": 0.2611,
"step": 2477
},
{
"epoch": 1.6199117214320746,
"grad_norm": 4.90625,
"learning_rate": 1.709363352664131e-05,
"loss": 0.4685,
"step": 2478
},
{
"epoch": 1.6205656367500407,
"grad_norm": 5.96875,
"learning_rate": 1.7090414284968355e-05,
"loss": 0.414,
"step": 2479
},
{
"epoch": 1.6212195520680073,
"grad_norm": 4.03125,
"learning_rate": 1.708719356488683e-05,
"loss": 0.2501,
"step": 2480
},
{
"epoch": 1.6218734673859734,
"grad_norm": 5.09375,
"learning_rate": 1.7083971367068275e-05,
"loss": 0.437,
"step": 2481
},
{
"epoch": 1.62252738270394,
"grad_norm": 6.25,
"learning_rate": 1.7080747692184546e-05,
"loss": 0.7212,
"step": 2482
},
{
"epoch": 1.623181298021906,
"grad_norm": 4.40625,
"learning_rate": 1.7077522540907804e-05,
"loss": 0.2993,
"step": 2483
},
{
"epoch": 1.6238352133398726,
"grad_norm": 3.390625,
"learning_rate": 1.707429591391052e-05,
"loss": 0.193,
"step": 2484
},
{
"epoch": 1.6244891286578387,
"grad_norm": 4.09375,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.2797,
"step": 2485
},
{
"epoch": 1.6251430439758052,
"grad_norm": 5.03125,
"learning_rate": 1.706783823544575e-05,
"loss": 0.5681,
"step": 2486
},
{
"epoch": 1.6257969592937713,
"grad_norm": 5.71875,
"learning_rate": 1.7064607185324742e-05,
"loss": 0.3986,
"step": 2487
},
{
"epoch": 1.6264508746117379,
"grad_norm": 4.03125,
"learning_rate": 1.7061374662176146e-05,
"loss": 0.4483,
"step": 2488
},
{
"epoch": 1.627104789929704,
"grad_norm": 4.5625,
"learning_rate": 1.705814066667397e-05,
"loss": 0.5583,
"step": 2489
},
{
"epoch": 1.6277587052476705,
"grad_norm": 4.40625,
"learning_rate": 1.705490519949253e-05,
"loss": 0.3335,
"step": 2490
},
{
"epoch": 1.6284126205656366,
"grad_norm": 8.5625,
"learning_rate": 1.705166826130645e-05,
"loss": 1.0713,
"step": 2491
},
{
"epoch": 1.6290665358836032,
"grad_norm": 5.25,
"learning_rate": 1.704842985279065e-05,
"loss": 0.4527,
"step": 2492
},
{
"epoch": 1.6297204512015693,
"grad_norm": 3.75,
"learning_rate": 1.704518997462037e-05,
"loss": 0.4729,
"step": 2493
},
{
"epoch": 1.6303743665195358,
"grad_norm": 5.8125,
"learning_rate": 1.704194862747115e-05,
"loss": 0.4029,
"step": 2494
},
{
"epoch": 1.631028281837502,
"grad_norm": 7.15625,
"learning_rate": 1.7038705812018835e-05,
"loss": 0.5933,
"step": 2495
},
{
"epoch": 1.6316821971554685,
"grad_norm": 7.46875,
"learning_rate": 1.703546152893958e-05,
"loss": 0.6682,
"step": 2496
},
{
"epoch": 1.6323361124734346,
"grad_norm": 6.125,
"learning_rate": 1.7032215778909845e-05,
"loss": 0.4157,
"step": 2497
},
{
"epoch": 1.6329900277914011,
"grad_norm": 2.90625,
"learning_rate": 1.702896856260639e-05,
"loss": 0.1964,
"step": 2498
},
{
"epoch": 1.6336439431093672,
"grad_norm": 6.65625,
"learning_rate": 1.7025719880706292e-05,
"loss": 0.4789,
"step": 2499
},
{
"epoch": 1.6342978584273338,
"grad_norm": 5.375,
"learning_rate": 1.7022469733886927e-05,
"loss": 0.4753,
"step": 2500
},
{
"epoch": 1.6349517737452999,
"grad_norm": 4.9375,
"learning_rate": 1.7019218122825975e-05,
"loss": 0.5549,
"step": 2501
},
{
"epoch": 1.6356056890632664,
"grad_norm": 5.1875,
"learning_rate": 1.7015965048201418e-05,
"loss": 0.3263,
"step": 2502
},
{
"epoch": 1.6362596043812325,
"grad_norm": 5.09375,
"learning_rate": 1.7012710510691554e-05,
"loss": 0.3525,
"step": 2503
},
{
"epoch": 1.636913519699199,
"grad_norm": 6.15625,
"learning_rate": 1.700945451097498e-05,
"loss": 0.5383,
"step": 2504
},
{
"epoch": 1.6375674350171652,
"grad_norm": 4.15625,
"learning_rate": 1.70061970497306e-05,
"loss": 0.3375,
"step": 2505
},
{
"epoch": 1.6382213503351317,
"grad_norm": 3.875,
"learning_rate": 1.700293812763762e-05,
"loss": 0.3783,
"step": 2506
},
{
"epoch": 1.6388752656530978,
"grad_norm": 4.65625,
"learning_rate": 1.6999677745375546e-05,
"loss": 0.4601,
"step": 2507
},
{
"epoch": 1.6395291809710644,
"grad_norm": 4.5625,
"learning_rate": 1.69964159036242e-05,
"loss": 0.4648,
"step": 2508
},
{
"epoch": 1.6401830962890305,
"grad_norm": 3.515625,
"learning_rate": 1.69931526030637e-05,
"loss": 0.2559,
"step": 2509
},
{
"epoch": 1.640837011606997,
"grad_norm": 3.6875,
"learning_rate": 1.6989887844374473e-05,
"loss": 0.327,
"step": 2510
},
{
"epoch": 1.6414909269249631,
"grad_norm": 3.921875,
"learning_rate": 1.6986621628237244e-05,
"loss": 0.2847,
"step": 2511
},
{
"epoch": 1.6421448422429297,
"grad_norm": 2.71875,
"learning_rate": 1.6983353955333047e-05,
"loss": 0.1704,
"step": 2512
},
{
"epoch": 1.6427987575608958,
"grad_norm": 4.15625,
"learning_rate": 1.698008482634322e-05,
"loss": 0.4071,
"step": 2513
},
{
"epoch": 1.6434526728788623,
"grad_norm": 7.03125,
"learning_rate": 1.6976814241949396e-05,
"loss": 0.5005,
"step": 2514
},
{
"epoch": 1.6441065881968284,
"grad_norm": 5.0625,
"learning_rate": 1.697354220283353e-05,
"loss": 0.4081,
"step": 2515
},
{
"epoch": 1.644760503514795,
"grad_norm": 3.59375,
"learning_rate": 1.6970268709677856e-05,
"loss": 0.2658,
"step": 2516
},
{
"epoch": 1.645414418832761,
"grad_norm": 4.0,
"learning_rate": 1.6966993763164928e-05,
"loss": 0.3736,
"step": 2517
},
{
"epoch": 1.6460683341507276,
"grad_norm": 3.390625,
"learning_rate": 1.6963717363977604e-05,
"loss": 0.25,
"step": 2518
},
{
"epoch": 1.6467222494686937,
"grad_norm": 3.5625,
"learning_rate": 1.6960439512799036e-05,
"loss": 0.1866,
"step": 2519
},
{
"epoch": 1.6473761647866603,
"grad_norm": 4.34375,
"learning_rate": 1.695716021031268e-05,
"loss": 0.2979,
"step": 2520
},
{
"epoch": 1.6480300801046264,
"grad_norm": 6.1875,
"learning_rate": 1.6953879457202303e-05,
"loss": 0.6568,
"step": 2521
},
{
"epoch": 1.648683995422593,
"grad_norm": 3.59375,
"learning_rate": 1.695059725415196e-05,
"loss": 0.2959,
"step": 2522
},
{
"epoch": 1.649337910740559,
"grad_norm": 3.171875,
"learning_rate": 1.6947313601846025e-05,
"loss": 0.1853,
"step": 2523
},
{
"epoch": 1.6499918260585256,
"grad_norm": 12.625,
"learning_rate": 1.6944028500969162e-05,
"loss": 0.433,
"step": 2524
},
{
"epoch": 1.6506457413764917,
"grad_norm": 5.125,
"learning_rate": 1.6940741952206342e-05,
"loss": 0.3733,
"step": 2525
},
{
"epoch": 1.6512996566944582,
"grad_norm": 5.3125,
"learning_rate": 1.6937453956242836e-05,
"loss": 0.4937,
"step": 2526
},
{
"epoch": 1.6519535720124243,
"grad_norm": 4.03125,
"learning_rate": 1.6934164513764215e-05,
"loss": 0.2739,
"step": 2527
},
{
"epoch": 1.6526074873303909,
"grad_norm": 3.453125,
"learning_rate": 1.6930873625456362e-05,
"loss": 0.2389,
"step": 2528
},
{
"epoch": 1.653261402648357,
"grad_norm": 4.09375,
"learning_rate": 1.6927581292005445e-05,
"loss": 0.1928,
"step": 2529
},
{
"epoch": 1.6539153179663235,
"grad_norm": 4.6875,
"learning_rate": 1.692428751409795e-05,
"loss": 0.585,
"step": 2530
},
{
"epoch": 1.6545692332842896,
"grad_norm": 5.0,
"learning_rate": 1.692099229242065e-05,
"loss": 0.5313,
"step": 2531
},
{
"epoch": 1.6552231486022562,
"grad_norm": 4.84375,
"learning_rate": 1.691769562766063e-05,
"loss": 0.3923,
"step": 2532
},
{
"epoch": 1.6558770639202223,
"grad_norm": 5.28125,
"learning_rate": 1.6914397520505267e-05,
"loss": 0.3824,
"step": 2533
},
{
"epoch": 1.6565309792381888,
"grad_norm": 5.0,
"learning_rate": 1.6911097971642244e-05,
"loss": 0.4812,
"step": 2534
},
{
"epoch": 1.657184894556155,
"grad_norm": 5.53125,
"learning_rate": 1.690779698175955e-05,
"loss": 0.4113,
"step": 2535
},
{
"epoch": 1.6578388098741215,
"grad_norm": 4.125,
"learning_rate": 1.6904494551545457e-05,
"loss": 0.3789,
"step": 2536
},
{
"epoch": 1.6584927251920876,
"grad_norm": 4.46875,
"learning_rate": 1.6901190681688558e-05,
"loss": 0.4669,
"step": 2537
},
{
"epoch": 1.659146640510054,
"grad_norm": 5.53125,
"learning_rate": 1.6897885372877732e-05,
"loss": 0.2812,
"step": 2538
},
{
"epoch": 1.6598005558280202,
"grad_norm": 5.28125,
"learning_rate": 1.6894578625802157e-05,
"loss": 0.4834,
"step": 2539
},
{
"epoch": 1.6604544711459865,
"grad_norm": 6.0,
"learning_rate": 1.6891270441151327e-05,
"loss": 0.4614,
"step": 2540
},
{
"epoch": 1.6611083864639529,
"grad_norm": 4.53125,
"learning_rate": 1.6887960819615025e-05,
"loss": 0.2194,
"step": 2541
},
{
"epoch": 1.6617623017819192,
"grad_norm": 4.34375,
"learning_rate": 1.6884649761883322e-05,
"loss": 0.2039,
"step": 2542
},
{
"epoch": 1.6624162170998855,
"grad_norm": 4.84375,
"learning_rate": 1.688133726864661e-05,
"loss": 0.2917,
"step": 2543
},
{
"epoch": 1.6630701324178518,
"grad_norm": 6.96875,
"learning_rate": 1.687802334059557e-05,
"loss": 0.7402,
"step": 2544
},
{
"epoch": 1.6637240477358182,
"grad_norm": 4.75,
"learning_rate": 1.687470797842118e-05,
"loss": 0.5023,
"step": 2545
},
{
"epoch": 1.6643779630537845,
"grad_norm": 3.359375,
"learning_rate": 1.687139118281472e-05,
"loss": 0.2519,
"step": 2546
},
{
"epoch": 1.6650318783717508,
"grad_norm": 6.59375,
"learning_rate": 1.6868072954467773e-05,
"loss": 0.4992,
"step": 2547
},
{
"epoch": 1.6656857936897171,
"grad_norm": 4.4375,
"learning_rate": 1.686475329407221e-05,
"loss": 0.4386,
"step": 2548
},
{
"epoch": 1.6663397090076835,
"grad_norm": 5.6875,
"learning_rate": 1.6861432202320208e-05,
"loss": 0.3676,
"step": 2549
},
{
"epoch": 1.6669936243256498,
"grad_norm": 6.5,
"learning_rate": 1.6858109679904245e-05,
"loss": 1.1778,
"step": 2550
},
{
"epoch": 1.6676475396436161,
"grad_norm": 3.921875,
"learning_rate": 1.685478572751709e-05,
"loss": 0.339,
"step": 2551
},
{
"epoch": 1.6683014549615824,
"grad_norm": 5.84375,
"learning_rate": 1.6851460345851816e-05,
"loss": 0.3984,
"step": 2552
},
{
"epoch": 1.6689553702795488,
"grad_norm": 5.90625,
"learning_rate": 1.6848133535601795e-05,
"loss": 0.4298,
"step": 2553
},
{
"epoch": 1.669609285597515,
"grad_norm": 3.8125,
"learning_rate": 1.6844805297460687e-05,
"loss": 0.2798,
"step": 2554
},
{
"epoch": 1.6702632009154814,
"grad_norm": 5.5625,
"learning_rate": 1.684147563212246e-05,
"loss": 0.3826,
"step": 2555
},
{
"epoch": 1.6709171162334477,
"grad_norm": 5.5625,
"learning_rate": 1.683814454028138e-05,
"loss": 0.3089,
"step": 2556
},
{
"epoch": 1.671571031551414,
"grad_norm": 6.59375,
"learning_rate": 1.6834812022632e-05,
"loss": 0.7408,
"step": 2557
},
{
"epoch": 1.6722249468693804,
"grad_norm": 7.46875,
"learning_rate": 1.6831478079869176e-05,
"loss": 0.9188,
"step": 2558
},
{
"epoch": 1.6728788621873467,
"grad_norm": 3.875,
"learning_rate": 1.6828142712688068e-05,
"loss": 0.4193,
"step": 2559
},
{
"epoch": 1.673532777505313,
"grad_norm": 6.46875,
"learning_rate": 1.6824805921784122e-05,
"loss": 0.5198,
"step": 2560
},
{
"epoch": 1.6741866928232794,
"grad_norm": 4.5625,
"learning_rate": 1.682146770785309e-05,
"loss": 0.5141,
"step": 2561
},
{
"epoch": 1.6748406081412457,
"grad_norm": 3.890625,
"learning_rate": 1.6818128071591015e-05,
"loss": 0.4048,
"step": 2562
},
{
"epoch": 1.675494523459212,
"grad_norm": 5.71875,
"learning_rate": 1.6814787013694237e-05,
"loss": 0.4639,
"step": 2563
},
{
"epoch": 1.6761484387771783,
"grad_norm": 4.9375,
"learning_rate": 1.68114445348594e-05,
"loss": 0.4583,
"step": 2564
},
{
"epoch": 1.6768023540951447,
"grad_norm": 3.84375,
"learning_rate": 1.680810063578342e-05,
"loss": 0.4048,
"step": 2565
},
{
"epoch": 1.677456269413111,
"grad_norm": 4.40625,
"learning_rate": 1.6804755317163548e-05,
"loss": 0.3001,
"step": 2566
},
{
"epoch": 1.6781101847310773,
"grad_norm": 5.34375,
"learning_rate": 1.6801408579697298e-05,
"loss": 0.4298,
"step": 2567
},
{
"epoch": 1.6787641000490436,
"grad_norm": 4.625,
"learning_rate": 1.6798060424082497e-05,
"loss": 0.3892,
"step": 2568
},
{
"epoch": 1.67941801536701,
"grad_norm": 3.28125,
"learning_rate": 1.6794710851017256e-05,
"loss": 0.2836,
"step": 2569
},
{
"epoch": 1.6800719306849763,
"grad_norm": 6.15625,
"learning_rate": 1.6791359861199994e-05,
"loss": 0.8757,
"step": 2570
},
{
"epoch": 1.6807258460029426,
"grad_norm": 7.15625,
"learning_rate": 1.678800745532942e-05,
"loss": 0.7364,
"step": 2571
},
{
"epoch": 1.681379761320909,
"grad_norm": 5.34375,
"learning_rate": 1.678465363410453e-05,
"loss": 0.4924,
"step": 2572
},
{
"epoch": 1.6820336766388753,
"grad_norm": 4.9375,
"learning_rate": 1.678129839822463e-05,
"loss": 0.4368,
"step": 2573
},
{
"epoch": 1.6826875919568416,
"grad_norm": 5.1875,
"learning_rate": 1.6777941748389316e-05,
"loss": 0.2711,
"step": 2574
},
{
"epoch": 1.683341507274808,
"grad_norm": 6.9375,
"learning_rate": 1.6774583685298468e-05,
"loss": 0.5899,
"step": 2575
},
{
"epoch": 1.6839954225927742,
"grad_norm": 4.78125,
"learning_rate": 1.6771224209652274e-05,
"loss": 0.4804,
"step": 2576
},
{
"epoch": 1.6846493379107406,
"grad_norm": 4.25,
"learning_rate": 1.6767863322151212e-05,
"loss": 0.341,
"step": 2577
},
{
"epoch": 1.6853032532287069,
"grad_norm": 7.3125,
"learning_rate": 1.6764501023496055e-05,
"loss": 0.8289,
"step": 2578
},
{
"epoch": 1.6859571685466732,
"grad_norm": 5.75,
"learning_rate": 1.676113731438786e-05,
"loss": 0.4238,
"step": 2579
},
{
"epoch": 1.6866110838646395,
"grad_norm": 5.28125,
"learning_rate": 1.6757772195528002e-05,
"loss": 0.5529,
"step": 2580
},
{
"epoch": 1.6872649991826059,
"grad_norm": 5.5625,
"learning_rate": 1.675440566761813e-05,
"loss": 0.5225,
"step": 2581
},
{
"epoch": 1.6879189145005722,
"grad_norm": 7.03125,
"learning_rate": 1.675103773136019e-05,
"loss": 0.6415,
"step": 2582
},
{
"epoch": 1.6885728298185385,
"grad_norm": 4.15625,
"learning_rate": 1.6747668387456422e-05,
"loss": 0.3411,
"step": 2583
},
{
"epoch": 1.6892267451365048,
"grad_norm": 4.5625,
"learning_rate": 1.6744297636609363e-05,
"loss": 0.4101,
"step": 2584
},
{
"epoch": 1.6898806604544712,
"grad_norm": 5.375,
"learning_rate": 1.6740925479521844e-05,
"loss": 0.5726,
"step": 2585
},
{
"epoch": 1.6905345757724375,
"grad_norm": 5.25,
"learning_rate": 1.6737551916896994e-05,
"loss": 0.3703,
"step": 2586
},
{
"epoch": 1.6911884910904038,
"grad_norm": 5.59375,
"learning_rate": 1.673417694943821e-05,
"loss": 0.4041,
"step": 2587
},
{
"epoch": 1.6918424064083701,
"grad_norm": 4.46875,
"learning_rate": 1.6730800577849213e-05,
"loss": 0.3314,
"step": 2588
},
{
"epoch": 1.6924963217263365,
"grad_norm": 4.34375,
"learning_rate": 1.6727422802834e-05,
"loss": 0.5823,
"step": 2589
},
{
"epoch": 1.6931502370443028,
"grad_norm": 8.75,
"learning_rate": 1.6724043625096866e-05,
"loss": 0.8256,
"step": 2590
},
{
"epoch": 1.693804152362269,
"grad_norm": 7.3125,
"learning_rate": 1.6720663045342395e-05,
"loss": 0.6609,
"step": 2591
},
{
"epoch": 1.6944580676802354,
"grad_norm": 7.28125,
"learning_rate": 1.6717281064275467e-05,
"loss": 0.9614,
"step": 2592
},
{
"epoch": 1.6951119829982018,
"grad_norm": 4.65625,
"learning_rate": 1.671389768260125e-05,
"loss": 0.3563,
"step": 2593
},
{
"epoch": 1.695765898316168,
"grad_norm": 4.65625,
"learning_rate": 1.671051290102521e-05,
"loss": 0.3481,
"step": 2594
},
{
"epoch": 1.6964198136341344,
"grad_norm": 3.8125,
"learning_rate": 1.6707126720253097e-05,
"loss": 0.3258,
"step": 2595
},
{
"epoch": 1.6970737289521007,
"grad_norm": 4.75,
"learning_rate": 1.670373914099096e-05,
"loss": 0.4066,
"step": 2596
},
{
"epoch": 1.697727644270067,
"grad_norm": 5.3125,
"learning_rate": 1.670035016394513e-05,
"loss": 0.4834,
"step": 2597
},
{
"epoch": 1.6983815595880334,
"grad_norm": 3.3125,
"learning_rate": 1.6696959789822245e-05,
"loss": 0.1601,
"step": 2598
},
{
"epoch": 1.6990354749059997,
"grad_norm": 4.375,
"learning_rate": 1.6693568019329225e-05,
"loss": 0.3392,
"step": 2599
},
{
"epoch": 1.699689390223966,
"grad_norm": 2.921875,
"learning_rate": 1.6690174853173276e-05,
"loss": 0.2845,
"step": 2600
},
{
"epoch": 1.7003433055419324,
"grad_norm": 5.71875,
"learning_rate": 1.6686780292061903e-05,
"loss": 0.481,
"step": 2601
},
{
"epoch": 1.7009972208598987,
"grad_norm": 4.5625,
"learning_rate": 1.6683384336702898e-05,
"loss": 0.2703,
"step": 2602
},
{
"epoch": 1.701651136177865,
"grad_norm": 4.90625,
"learning_rate": 1.667998698780435e-05,
"loss": 0.3571,
"step": 2603
},
{
"epoch": 1.7023050514958313,
"grad_norm": 4.65625,
"learning_rate": 1.6676588246074627e-05,
"loss": 0.5073,
"step": 2604
},
{
"epoch": 1.7029589668137977,
"grad_norm": 4.5,
"learning_rate": 1.6673188112222394e-05,
"loss": 0.5489,
"step": 2605
},
{
"epoch": 1.703612882131764,
"grad_norm": 5.71875,
"learning_rate": 1.6669786586956616e-05,
"loss": 0.5399,
"step": 2606
},
{
"epoch": 1.7042667974497303,
"grad_norm": 4.03125,
"learning_rate": 1.6666383670986528e-05,
"loss": 0.3031,
"step": 2607
},
{
"epoch": 1.7049207127676966,
"grad_norm": 5.1875,
"learning_rate": 1.666297936502167e-05,
"loss": 0.5535,
"step": 2608
},
{
"epoch": 1.705574628085663,
"grad_norm": 3.84375,
"learning_rate": 1.6659573669771867e-05,
"loss": 0.1699,
"step": 2609
},
{
"epoch": 1.7062285434036293,
"grad_norm": 5.0625,
"learning_rate": 1.6656166585947234e-05,
"loss": 0.3948,
"step": 2610
},
{
"epoch": 1.7068824587215956,
"grad_norm": 4.59375,
"learning_rate": 1.665275811425817e-05,
"loss": 0.4746,
"step": 2611
},
{
"epoch": 1.7075363740395617,
"grad_norm": 4.1875,
"learning_rate": 1.664934825541538e-05,
"loss": 0.4626,
"step": 2612
},
{
"epoch": 1.7081902893575283,
"grad_norm": 6.8125,
"learning_rate": 1.6645937010129837e-05,
"loss": 0.9037,
"step": 2613
},
{
"epoch": 1.7088442046754944,
"grad_norm": 3.390625,
"learning_rate": 1.664252437911282e-05,
"loss": 0.215,
"step": 2614
},
{
"epoch": 1.709498119993461,
"grad_norm": 4.125,
"learning_rate": 1.6639110363075884e-05,
"loss": 0.3194,
"step": 2615
},
{
"epoch": 1.710152035311427,
"grad_norm": 3.203125,
"learning_rate": 1.6635694962730883e-05,
"loss": 0.2268,
"step": 2616
},
{
"epoch": 1.7108059506293936,
"grad_norm": 4.53125,
"learning_rate": 1.6632278178789956e-05,
"loss": 0.3103,
"step": 2617
},
{
"epoch": 1.7114598659473597,
"grad_norm": 3.453125,
"learning_rate": 1.6628860011965526e-05,
"loss": 0.2356,
"step": 2618
},
{
"epoch": 1.7121137812653262,
"grad_norm": 2.546875,
"learning_rate": 1.6625440462970314e-05,
"loss": 0.1658,
"step": 2619
},
{
"epoch": 1.7127676965832923,
"grad_norm": 5.9375,
"learning_rate": 1.6622019532517318e-05,
"loss": 0.4097,
"step": 2620
},
{
"epoch": 1.7134216119012589,
"grad_norm": 3.75,
"learning_rate": 1.6618597221319835e-05,
"loss": 0.5039,
"step": 2621
},
{
"epoch": 1.714075527219225,
"grad_norm": 5.5,
"learning_rate": 1.661517353009144e-05,
"loss": 0.4158,
"step": 2622
},
{
"epoch": 1.7147294425371915,
"grad_norm": 4.9375,
"learning_rate": 1.6611748459546e-05,
"loss": 0.7647,
"step": 2623
},
{
"epoch": 1.7153833578551576,
"grad_norm": 3.90625,
"learning_rate": 1.6608322010397673e-05,
"loss": 0.4834,
"step": 2624
},
{
"epoch": 1.7160372731731242,
"grad_norm": 7.1875,
"learning_rate": 1.66048941833609e-05,
"loss": 1.1695,
"step": 2625
},
{
"epoch": 1.7166911884910903,
"grad_norm": 5.9375,
"learning_rate": 1.6601464979150407e-05,
"loss": 0.559,
"step": 2626
},
{
"epoch": 1.7173451038090568,
"grad_norm": 7.53125,
"learning_rate": 1.6598034398481218e-05,
"loss": 1.1863,
"step": 2627
},
{
"epoch": 1.717999019127023,
"grad_norm": 4.59375,
"learning_rate": 1.6594602442068634e-05,
"loss": 0.4833,
"step": 2628
},
{
"epoch": 1.7186529344449895,
"grad_norm": 4.96875,
"learning_rate": 1.659116911062824e-05,
"loss": 0.5856,
"step": 2629
},
{
"epoch": 1.7193068497629556,
"grad_norm": 4.0,
"learning_rate": 1.6587734404875925e-05,
"loss": 0.2741,
"step": 2630
},
{
"epoch": 1.719960765080922,
"grad_norm": 9.3125,
"learning_rate": 1.6584298325527837e-05,
"loss": 0.5273,
"step": 2631
},
{
"epoch": 1.7206146803988882,
"grad_norm": 4.59375,
"learning_rate": 1.658086087330044e-05,
"loss": 0.4054,
"step": 2632
},
{
"epoch": 1.7212685957168548,
"grad_norm": 5.96875,
"learning_rate": 1.6577422048910466e-05,
"loss": 0.4402,
"step": 2633
},
{
"epoch": 1.7219225110348209,
"grad_norm": 4.9375,
"learning_rate": 1.6573981853074937e-05,
"loss": 0.4559,
"step": 2634
},
{
"epoch": 1.7225764263527874,
"grad_norm": 6.125,
"learning_rate": 1.6570540286511157e-05,
"loss": 1.0348,
"step": 2635
},
{
"epoch": 1.7232303416707535,
"grad_norm": 6.5,
"learning_rate": 1.656709734993673e-05,
"loss": 0.7541,
"step": 2636
},
{
"epoch": 1.72388425698872,
"grad_norm": 3.8125,
"learning_rate": 1.656365304406953e-05,
"loss": 0.5219,
"step": 2637
},
{
"epoch": 1.7245381723066862,
"grad_norm": 5.375,
"learning_rate": 1.656020736962772e-05,
"loss": 0.4333,
"step": 2638
},
{
"epoch": 1.7251920876246527,
"grad_norm": 5.78125,
"learning_rate": 1.655676032732976e-05,
"loss": 0.5399,
"step": 2639
},
{
"epoch": 1.7258460029426188,
"grad_norm": 5.78125,
"learning_rate": 1.655331191789438e-05,
"loss": 0.647,
"step": 2640
},
{
"epoch": 1.7264999182605854,
"grad_norm": 3.59375,
"learning_rate": 1.65498621420406e-05,
"loss": 0.2718,
"step": 2641
},
{
"epoch": 1.7271538335785515,
"grad_norm": 3.65625,
"learning_rate": 1.6546411000487726e-05,
"loss": 0.325,
"step": 2642
},
{
"epoch": 1.727807748896518,
"grad_norm": 5.40625,
"learning_rate": 1.6542958493955357e-05,
"loss": 0.5002,
"step": 2643
},
{
"epoch": 1.728461664214484,
"grad_norm": 8.625,
"learning_rate": 1.653950462316336e-05,
"loss": 1.0366,
"step": 2644
},
{
"epoch": 1.7291155795324507,
"grad_norm": 4.40625,
"learning_rate": 1.6536049388831897e-05,
"loss": 0.277,
"step": 2645
},
{
"epoch": 1.7297694948504168,
"grad_norm": 4.4375,
"learning_rate": 1.653259279168141e-05,
"loss": 0.3062,
"step": 2646
},
{
"epoch": 1.7304234101683833,
"grad_norm": 4.8125,
"learning_rate": 1.652913483243263e-05,
"loss": 0.3348,
"step": 2647
},
{
"epoch": 1.7310773254863494,
"grad_norm": 6.15625,
"learning_rate": 1.652567551180657e-05,
"loss": 0.6655,
"step": 2648
},
{
"epoch": 1.731731240804316,
"grad_norm": 4.53125,
"learning_rate": 1.6522214830524527e-05,
"loss": 0.2683,
"step": 2649
},
{
"epoch": 1.732385156122282,
"grad_norm": 5.0,
"learning_rate": 1.651875278930808e-05,
"loss": 0.5148,
"step": 2650
},
{
"epoch": 1.7330390714402486,
"grad_norm": 5.5,
"learning_rate": 1.6515289388879088e-05,
"loss": 0.4984,
"step": 2651
},
{
"epoch": 1.7336929867582147,
"grad_norm": 3.21875,
"learning_rate": 1.6511824629959706e-05,
"loss": 0.2093,
"step": 2652
},
{
"epoch": 1.7343469020761813,
"grad_norm": 4.875,
"learning_rate": 1.650835851327236e-05,
"loss": 0.3937,
"step": 2653
},
{
"epoch": 1.7350008173941474,
"grad_norm": 5.46875,
"learning_rate": 1.650489103953976e-05,
"loss": 0.6472,
"step": 2654
},
{
"epoch": 1.735654732712114,
"grad_norm": 5.09375,
"learning_rate": 1.6501422209484907e-05,
"loss": 0.3782,
"step": 2655
},
{
"epoch": 1.73630864803008,
"grad_norm": 4.34375,
"learning_rate": 1.649795202383108e-05,
"loss": 0.3921,
"step": 2656
},
{
"epoch": 1.7369625633480466,
"grad_norm": 4.25,
"learning_rate": 1.6494480483301836e-05,
"loss": 0.2881,
"step": 2657
},
{
"epoch": 1.7376164786660127,
"grad_norm": 5.0,
"learning_rate": 1.6491007588621028e-05,
"loss": 0.3763,
"step": 2658
},
{
"epoch": 1.7382703939839792,
"grad_norm": 6.0625,
"learning_rate": 1.6487533340512772e-05,
"loss": 0.6318,
"step": 2659
},
{
"epoch": 1.7389243093019453,
"grad_norm": 4.3125,
"learning_rate": 1.6484057739701485e-05,
"loss": 0.223,
"step": 2660
},
{
"epoch": 1.7395782246199119,
"grad_norm": 3.359375,
"learning_rate": 1.6480580786911854e-05,
"loss": 0.2742,
"step": 2661
},
{
"epoch": 1.740232139937878,
"grad_norm": 5.5625,
"learning_rate": 1.6477102482868856e-05,
"loss": 0.362,
"step": 2662
},
{
"epoch": 1.7408860552558445,
"grad_norm": 4.8125,
"learning_rate": 1.6473622828297735e-05,
"loss": 0.6997,
"step": 2663
},
{
"epoch": 1.7415399705738106,
"grad_norm": 6.4375,
"learning_rate": 1.6470141823924037e-05,
"loss": 0.7509,
"step": 2664
},
{
"epoch": 1.7421938858917771,
"grad_norm": 3.515625,
"learning_rate": 1.646665947047358e-05,
"loss": 0.2931,
"step": 2665
},
{
"epoch": 1.7428478012097433,
"grad_norm": 6.1875,
"learning_rate": 1.6463175768672456e-05,
"loss": 0.6228,
"step": 2666
},
{
"epoch": 1.7435017165277098,
"grad_norm": 6.84375,
"learning_rate": 1.6459690719247047e-05,
"loss": 0.7565,
"step": 2667
},
{
"epoch": 1.744155631845676,
"grad_norm": 6.125,
"learning_rate": 1.645620432292401e-05,
"loss": 0.866,
"step": 2668
},
{
"epoch": 1.7448095471636424,
"grad_norm": 3.109375,
"learning_rate": 1.6452716580430303e-05,
"loss": 0.1698,
"step": 2669
},
{
"epoch": 1.7454634624816086,
"grad_norm": 6.1875,
"learning_rate": 1.644922749249313e-05,
"loss": 1.1899,
"step": 2670
},
{
"epoch": 1.746117377799575,
"grad_norm": 5.0,
"learning_rate": 1.6445737059840007e-05,
"loss": 0.5061,
"step": 2671
},
{
"epoch": 1.7467712931175412,
"grad_norm": 3.90625,
"learning_rate": 1.6442245283198706e-05,
"loss": 0.3751,
"step": 2672
},
{
"epoch": 1.7474252084355077,
"grad_norm": 3.1875,
"learning_rate": 1.6438752163297298e-05,
"loss": 0.3441,
"step": 2673
},
{
"epoch": 1.7480791237534739,
"grad_norm": 4.28125,
"learning_rate": 1.6435257700864125e-05,
"loss": 0.3938,
"step": 2674
},
{
"epoch": 1.7487330390714404,
"grad_norm": 4.25,
"learning_rate": 1.6431761896627808e-05,
"loss": 0.4167,
"step": 2675
},
{
"epoch": 1.7493869543894065,
"grad_norm": 3.859375,
"learning_rate": 1.6428264751317257e-05,
"loss": 0.2614,
"step": 2676
},
{
"epoch": 1.750040869707373,
"grad_norm": 6.25,
"learning_rate": 1.6424766265661648e-05,
"loss": 0.8656,
"step": 2677
},
{
"epoch": 1.7506947850253392,
"grad_norm": 5.625,
"learning_rate": 1.6421266440390446e-05,
"loss": 0.4357,
"step": 2678
},
{
"epoch": 1.7513487003433057,
"grad_norm": 3.5625,
"learning_rate": 1.6417765276233394e-05,
"loss": 0.3546,
"step": 2679
},
{
"epoch": 1.7520026156612718,
"grad_norm": 4.25,
"learning_rate": 1.641426277392051e-05,
"loss": 0.427,
"step": 2680
},
{
"epoch": 1.7526565309792383,
"grad_norm": 5.96875,
"learning_rate": 1.6410758934182104e-05,
"loss": 0.4731,
"step": 2681
},
{
"epoch": 1.7533104462972045,
"grad_norm": 4.96875,
"learning_rate": 1.6407253757748742e-05,
"loss": 0.5371,
"step": 2682
},
{
"epoch": 1.753964361615171,
"grad_norm": 3.046875,
"learning_rate": 1.6403747245351284e-05,
"loss": 0.1959,
"step": 2683
},
{
"epoch": 1.754618276933137,
"grad_norm": 3.78125,
"learning_rate": 1.640023939772087e-05,
"loss": 0.1982,
"step": 2684
},
{
"epoch": 1.7552721922511036,
"grad_norm": 3.078125,
"learning_rate": 1.6396730215588913e-05,
"loss": 0.2847,
"step": 2685
},
{
"epoch": 1.7559261075690697,
"grad_norm": 5.75,
"learning_rate": 1.6393219699687107e-05,
"loss": 0.6827,
"step": 2686
},
{
"epoch": 1.756580022887036,
"grad_norm": 5.0625,
"learning_rate": 1.6389707850747423e-05,
"loss": 0.5482,
"step": 2687
},
{
"epoch": 1.7572339382050024,
"grad_norm": 6.5625,
"learning_rate": 1.638619466950211e-05,
"loss": 0.3707,
"step": 2688
},
{
"epoch": 1.7578878535229687,
"grad_norm": 4.8125,
"learning_rate": 1.6382680156683687e-05,
"loss": 0.3112,
"step": 2689
},
{
"epoch": 1.758541768840935,
"grad_norm": 5.28125,
"learning_rate": 1.6379164313024968e-05,
"loss": 0.382,
"step": 2690
},
{
"epoch": 1.7591956841589014,
"grad_norm": 3.6875,
"learning_rate": 1.637564713925903e-05,
"loss": 0.2672,
"step": 2691
},
{
"epoch": 1.7598495994768677,
"grad_norm": 5.78125,
"learning_rate": 1.6372128636119238e-05,
"loss": 0.6438,
"step": 2692
},
{
"epoch": 1.760503514794834,
"grad_norm": 4.28125,
"learning_rate": 1.636860880433922e-05,
"loss": 0.2837,
"step": 2693
},
{
"epoch": 1.7611574301128003,
"grad_norm": 2.828125,
"learning_rate": 1.636508764465289e-05,
"loss": 0.1825,
"step": 2694
},
{
"epoch": 1.7618113454307667,
"grad_norm": 5.9375,
"learning_rate": 1.636156515779445e-05,
"loss": 0.52,
"step": 2695
},
{
"epoch": 1.762465260748733,
"grad_norm": 4.40625,
"learning_rate": 1.635804134449835e-05,
"loss": 0.2246,
"step": 2696
},
{
"epoch": 1.7631191760666993,
"grad_norm": 6.5,
"learning_rate": 1.6354516205499345e-05,
"loss": 0.6254,
"step": 2697
},
{
"epoch": 1.7637730913846656,
"grad_norm": 3.359375,
"learning_rate": 1.635098974153245e-05,
"loss": 0.2724,
"step": 2698
},
{
"epoch": 1.764427006702632,
"grad_norm": 4.4375,
"learning_rate": 1.6347461953332962e-05,
"loss": 0.3015,
"step": 2699
},
{
"epoch": 1.7650809220205983,
"grad_norm": 6.09375,
"learning_rate": 1.6343932841636455e-05,
"loss": 0.4624,
"step": 2700
},
{
"epoch": 1.7657348373385646,
"grad_norm": 6.5,
"learning_rate": 1.634040240717878e-05,
"loss": 0.425,
"step": 2701
},
{
"epoch": 1.766388752656531,
"grad_norm": 3.625,
"learning_rate": 1.6336870650696054e-05,
"loss": 0.3165,
"step": 2702
},
{
"epoch": 1.7670426679744973,
"grad_norm": 3.359375,
"learning_rate": 1.633333757292468e-05,
"loss": 0.3692,
"step": 2703
},
{
"epoch": 1.7676965832924636,
"grad_norm": 4.46875,
"learning_rate": 1.6329803174601333e-05,
"loss": 0.272,
"step": 2704
},
{
"epoch": 1.76835049861043,
"grad_norm": 6.28125,
"learning_rate": 1.6326267456462965e-05,
"loss": 0.4734,
"step": 2705
},
{
"epoch": 1.7690044139283962,
"grad_norm": 7.84375,
"learning_rate": 1.6322730419246798e-05,
"loss": 0.6407,
"step": 2706
},
{
"epoch": 1.7696583292463626,
"grad_norm": 3.40625,
"learning_rate": 1.6319192063690338e-05,
"loss": 0.1941,
"step": 2707
},
{
"epoch": 1.770312244564329,
"grad_norm": 4.96875,
"learning_rate": 1.631565239053136e-05,
"loss": 0.3372,
"step": 2708
},
{
"epoch": 1.7709661598822952,
"grad_norm": 3.53125,
"learning_rate": 1.6312111400507907e-05,
"loss": 0.3269,
"step": 2709
},
{
"epoch": 1.7716200752002615,
"grad_norm": 6.0625,
"learning_rate": 1.6308569094358315e-05,
"loss": 0.563,
"step": 2710
},
{
"epoch": 1.7722739905182279,
"grad_norm": 4.5625,
"learning_rate": 1.6305025472821173e-05,
"loss": 0.5144,
"step": 2711
},
{
"epoch": 1.7729279058361942,
"grad_norm": 3.625,
"learning_rate": 1.6301480536635364e-05,
"loss": 0.2172,
"step": 2712
},
{
"epoch": 1.7735818211541605,
"grad_norm": 7.15625,
"learning_rate": 1.629793428654003e-05,
"loss": 0.7476,
"step": 2713
},
{
"epoch": 1.7742357364721268,
"grad_norm": 6.15625,
"learning_rate": 1.6294386723274592e-05,
"loss": 0.3746,
"step": 2714
},
{
"epoch": 1.7748896517900932,
"grad_norm": 4.25,
"learning_rate": 1.629083784757875e-05,
"loss": 0.3922,
"step": 2715
},
{
"epoch": 1.7755435671080595,
"grad_norm": 4.65625,
"learning_rate": 1.6287287660192472e-05,
"loss": 0.4536,
"step": 2716
},
{
"epoch": 1.7761974824260258,
"grad_norm": 4.09375,
"learning_rate": 1.6283736161855995e-05,
"loss": 0.287,
"step": 2717
},
{
"epoch": 1.7768513977439921,
"grad_norm": 4.78125,
"learning_rate": 1.628018335330984e-05,
"loss": 0.3152,
"step": 2718
},
{
"epoch": 1.7775053130619585,
"grad_norm": 6.6875,
"learning_rate": 1.62766292352948e-05,
"loss": 0.5527,
"step": 2719
},
{
"epoch": 1.7781592283799248,
"grad_norm": 5.0625,
"learning_rate": 1.627307380855193e-05,
"loss": 0.2801,
"step": 2720
},
{
"epoch": 1.7788131436978911,
"grad_norm": 6.25,
"learning_rate": 1.626951707382257e-05,
"loss": 0.5465,
"step": 2721
},
{
"epoch": 1.7794670590158574,
"grad_norm": 2.640625,
"learning_rate": 1.6265959031848325e-05,
"loss": 0.2187,
"step": 2722
},
{
"epoch": 1.7801209743338238,
"grad_norm": 3.5,
"learning_rate": 1.6262399683371077e-05,
"loss": 0.2206,
"step": 2723
},
{
"epoch": 1.78077488965179,
"grad_norm": 2.9375,
"learning_rate": 1.625883902913298e-05,
"loss": 0.2162,
"step": 2724
},
{
"epoch": 1.7814288049697564,
"grad_norm": 5.15625,
"learning_rate": 1.6255277069876454e-05,
"loss": 0.3755,
"step": 2725
},
{
"epoch": 1.7820827202877227,
"grad_norm": 3.953125,
"learning_rate": 1.6251713806344206e-05,
"loss": 0.3174,
"step": 2726
},
{
"epoch": 1.782736635605689,
"grad_norm": 5.0,
"learning_rate": 1.6248149239279193e-05,
"loss": 0.3674,
"step": 2727
},
{
"epoch": 1.7833905509236554,
"grad_norm": 6.3125,
"learning_rate": 1.6244583369424665e-05,
"loss": 0.5546,
"step": 2728
},
{
"epoch": 1.7840444662416217,
"grad_norm": 5.625,
"learning_rate": 1.624101619752413e-05,
"loss": 0.3282,
"step": 2729
},
{
"epoch": 1.784698381559588,
"grad_norm": 3.5625,
"learning_rate": 1.6237447724321373e-05,
"loss": 0.4294,
"step": 2730
},
{
"epoch": 1.7853522968775544,
"grad_norm": 4.6875,
"learning_rate": 1.6233877950560452e-05,
"loss": 0.3538,
"step": 2731
},
{
"epoch": 1.7860062121955207,
"grad_norm": 6.1875,
"learning_rate": 1.623030687698569e-05,
"loss": 0.6234,
"step": 2732
},
{
"epoch": 1.786660127513487,
"grad_norm": 4.03125,
"learning_rate": 1.622673450434169e-05,
"loss": 0.3857,
"step": 2733
},
{
"epoch": 1.7873140428314533,
"grad_norm": 4.4375,
"learning_rate": 1.6223160833373316e-05,
"loss": 0.4176,
"step": 2734
},
{
"epoch": 1.7879679581494197,
"grad_norm": 5.8125,
"learning_rate": 1.6219585864825706e-05,
"loss": 0.7024,
"step": 2735
},
{
"epoch": 1.788621873467386,
"grad_norm": 5.59375,
"learning_rate": 1.6216009599444275e-05,
"loss": 0.6372,
"step": 2736
},
{
"epoch": 1.7892757887853523,
"grad_norm": 4.5625,
"learning_rate": 1.6212432037974698e-05,
"loss": 0.4491,
"step": 2737
},
{
"epoch": 1.7899297041033186,
"grad_norm": 4.90625,
"learning_rate": 1.620885318116293e-05,
"loss": 0.4761,
"step": 2738
},
{
"epoch": 1.790583619421285,
"grad_norm": 5.21875,
"learning_rate": 1.6205273029755185e-05,
"loss": 0.5584,
"step": 2739
},
{
"epoch": 1.7912375347392513,
"grad_norm": 3.671875,
"learning_rate": 1.620169158449796e-05,
"loss": 0.2234,
"step": 2740
},
{
"epoch": 1.7918914500572176,
"grad_norm": 5.46875,
"learning_rate": 1.619810884613801e-05,
"loss": 0.6493,
"step": 2741
},
{
"epoch": 1.792545365375184,
"grad_norm": 3.75,
"learning_rate": 1.6194524815422373e-05,
"loss": 0.3813,
"step": 2742
},
{
"epoch": 1.7931992806931503,
"grad_norm": 4.03125,
"learning_rate": 1.6190939493098344e-05,
"loss": 0.3636,
"step": 2743
},
{
"epoch": 1.7938531960111166,
"grad_norm": 5.1875,
"learning_rate": 1.618735287991348e-05,
"loss": 0.3633,
"step": 2744
},
{
"epoch": 1.794507111329083,
"grad_norm": 3.71875,
"learning_rate": 1.618376497661564e-05,
"loss": 0.3149,
"step": 2745
},
{
"epoch": 1.7951610266470492,
"grad_norm": 4.09375,
"learning_rate": 1.618017578395292e-05,
"loss": 0.3693,
"step": 2746
},
{
"epoch": 1.7958149419650156,
"grad_norm": 4.96875,
"learning_rate": 1.617658530267369e-05,
"loss": 0.5932,
"step": 2747
},
{
"epoch": 1.796468857282982,
"grad_norm": 7.21875,
"learning_rate": 1.6172993533526602e-05,
"loss": 0.5266,
"step": 2748
},
{
"epoch": 1.7971227726009482,
"grad_norm": 5.15625,
"learning_rate": 1.6169400477260566e-05,
"loss": 0.62,
"step": 2749
},
{
"epoch": 1.7977766879189145,
"grad_norm": 4.65625,
"learning_rate": 1.6165806134624767e-05,
"loss": 0.4048,
"step": 2750
},
{
"epoch": 1.7984306032368809,
"grad_norm": 4.15625,
"learning_rate": 1.616221050636865e-05,
"loss": 0.3308,
"step": 2751
},
{
"epoch": 1.7990845185548472,
"grad_norm": 3.671875,
"learning_rate": 1.6158613593241934e-05,
"loss": 0.2648,
"step": 2752
},
{
"epoch": 1.7997384338728135,
"grad_norm": 4.875,
"learning_rate": 1.6155015395994606e-05,
"loss": 0.4015,
"step": 2753
},
{
"epoch": 1.8003923491907798,
"grad_norm": 5.6875,
"learning_rate": 1.6151415915376916e-05,
"loss": 0.5552,
"step": 2754
},
{
"epoch": 1.8010462645087462,
"grad_norm": 6.53125,
"learning_rate": 1.6147815152139385e-05,
"loss": 0.4251,
"step": 2755
},
{
"epoch": 1.8017001798267125,
"grad_norm": 4.71875,
"learning_rate": 1.6144213107032807e-05,
"loss": 0.4171,
"step": 2756
},
{
"epoch": 1.8023540951446788,
"grad_norm": 4.4375,
"learning_rate": 1.614060978080823e-05,
"loss": 0.4513,
"step": 2757
},
{
"epoch": 1.8030080104626451,
"grad_norm": 4.8125,
"learning_rate": 1.613700517421698e-05,
"loss": 0.2819,
"step": 2758
},
{
"epoch": 1.8036619257806115,
"grad_norm": 4.9375,
"learning_rate": 1.6133399288010644e-05,
"loss": 0.336,
"step": 2759
},
{
"epoch": 1.8043158410985778,
"grad_norm": 3.796875,
"learning_rate": 1.6129792122941083e-05,
"loss": 0.4022,
"step": 2760
},
{
"epoch": 1.804969756416544,
"grad_norm": 3.9375,
"learning_rate": 1.6126183679760415e-05,
"loss": 0.2594,
"step": 2761
},
{
"epoch": 1.8056236717345104,
"grad_norm": 4.40625,
"learning_rate": 1.6122573959221038e-05,
"loss": 0.5774,
"step": 2762
},
{
"epoch": 1.8062775870524765,
"grad_norm": 5.125,
"learning_rate": 1.61189629620756e-05,
"loss": 0.5567,
"step": 2763
},
{
"epoch": 1.806931502370443,
"grad_norm": 5.1875,
"learning_rate": 1.611535068907702e-05,
"loss": 0.5008,
"step": 2764
},
{
"epoch": 1.8075854176884092,
"grad_norm": 4.1875,
"learning_rate": 1.6111737140978495e-05,
"loss": 0.2958,
"step": 2765
},
{
"epoch": 1.8082393330063757,
"grad_norm": 4.15625,
"learning_rate": 1.6108122318533475e-05,
"loss": 0.2816,
"step": 2766
},
{
"epoch": 1.8088932483243418,
"grad_norm": 5.5625,
"learning_rate": 1.6104506222495678e-05,
"loss": 0.6602,
"step": 2767
},
{
"epoch": 1.8095471636423084,
"grad_norm": 3.734375,
"learning_rate": 1.610088885361909e-05,
"loss": 0.2354,
"step": 2768
},
{
"epoch": 1.8102010789602745,
"grad_norm": 5.15625,
"learning_rate": 1.6097270212657966e-05,
"loss": 0.3917,
"step": 2769
},
{
"epoch": 1.810854994278241,
"grad_norm": 4.3125,
"learning_rate": 1.6093650300366812e-05,
"loss": 0.39,
"step": 2770
},
{
"epoch": 1.8115089095962071,
"grad_norm": 4.96875,
"learning_rate": 1.609002911750042e-05,
"loss": 0.4224,
"step": 2771
},
{
"epoch": 1.8121628249141737,
"grad_norm": 4.8125,
"learning_rate": 1.6086406664813828e-05,
"loss": 0.373,
"step": 2772
},
{
"epoch": 1.8128167402321398,
"grad_norm": 4.90625,
"learning_rate": 1.6082782943062355e-05,
"loss": 0.6323,
"step": 2773
},
{
"epoch": 1.8134706555501063,
"grad_norm": 3.859375,
"learning_rate": 1.6079157953001566e-05,
"loss": 0.3049,
"step": 2774
},
{
"epoch": 1.8141245708680724,
"grad_norm": 5.21875,
"learning_rate": 1.6075531695387304e-05,
"loss": 0.325,
"step": 2775
},
{
"epoch": 1.814778486186039,
"grad_norm": 8.6875,
"learning_rate": 1.607190417097568e-05,
"loss": 0.5613,
"step": 2776
},
{
"epoch": 1.815432401504005,
"grad_norm": 3.921875,
"learning_rate": 1.606827538052305e-05,
"loss": 0.2865,
"step": 2777
},
{
"epoch": 1.8160863168219716,
"grad_norm": 3.34375,
"learning_rate": 1.6064645324786058e-05,
"loss": 0.1545,
"step": 2778
},
{
"epoch": 1.8167402321399377,
"grad_norm": 4.125,
"learning_rate": 1.6061014004521593e-05,
"loss": 0.3133,
"step": 2779
},
{
"epoch": 1.8173941474579043,
"grad_norm": 7.375,
"learning_rate": 1.6057381420486814e-05,
"loss": 0.5486,
"step": 2780
},
{
"epoch": 1.8180480627758704,
"grad_norm": 4.03125,
"learning_rate": 1.6053747573439147e-05,
"loss": 0.4003,
"step": 2781
},
{
"epoch": 1.818701978093837,
"grad_norm": 3.953125,
"learning_rate": 1.605011246413628e-05,
"loss": 0.3754,
"step": 2782
},
{
"epoch": 1.819355893411803,
"grad_norm": 2.75,
"learning_rate": 1.6046476093336157e-05,
"loss": 0.2352,
"step": 2783
},
{
"epoch": 1.8200098087297696,
"grad_norm": 4.46875,
"learning_rate": 1.6042838461797e-05,
"loss": 0.4076,
"step": 2784
},
{
"epoch": 1.8206637240477357,
"grad_norm": 4.78125,
"learning_rate": 1.603919957027727e-05,
"loss": 0.3573,
"step": 2785
},
{
"epoch": 1.8213176393657022,
"grad_norm": 4.03125,
"learning_rate": 1.6035559419535714e-05,
"loss": 0.2615,
"step": 2786
},
{
"epoch": 1.8219715546836683,
"grad_norm": 5.625,
"learning_rate": 1.6031918010331336e-05,
"loss": 0.4608,
"step": 2787
},
{
"epoch": 1.8226254700016349,
"grad_norm": 4.875,
"learning_rate": 1.6028275343423395e-05,
"loss": 0.4546,
"step": 2788
},
{
"epoch": 1.823279385319601,
"grad_norm": 6.5,
"learning_rate": 1.6024631419571418e-05,
"loss": 0.5806,
"step": 2789
},
{
"epoch": 1.8239333006375675,
"grad_norm": 6.75,
"learning_rate": 1.602098623953519e-05,
"loss": 0.5868,
"step": 2790
},
{
"epoch": 1.8245872159555336,
"grad_norm": 5.1875,
"learning_rate": 1.601733980407476e-05,
"loss": 0.3714,
"step": 2791
},
{
"epoch": 1.8252411312735002,
"grad_norm": 9.375,
"learning_rate": 1.601369211395044e-05,
"loss": 0.9932,
"step": 2792
},
{
"epoch": 1.8258950465914663,
"grad_norm": 4.40625,
"learning_rate": 1.601004316992281e-05,
"loss": 0.2976,
"step": 2793
},
{
"epoch": 1.8265489619094328,
"grad_norm": 3.625,
"learning_rate": 1.60063929727527e-05,
"loss": 0.1853,
"step": 2794
},
{
"epoch": 1.827202877227399,
"grad_norm": 4.46875,
"learning_rate": 1.6002741523201197e-05,
"loss": 0.2947,
"step": 2795
},
{
"epoch": 1.8278567925453655,
"grad_norm": 3.796875,
"learning_rate": 1.5999088822029664e-05,
"loss": 0.2546,
"step": 2796
},
{
"epoch": 1.8285107078633316,
"grad_norm": 5.21875,
"learning_rate": 1.5995434869999723e-05,
"loss": 0.442,
"step": 2797
},
{
"epoch": 1.8291646231812981,
"grad_norm": 5.15625,
"learning_rate": 1.599177966787325e-05,
"loss": 0.515,
"step": 2798
},
{
"epoch": 1.8298185384992642,
"grad_norm": 4.375,
"learning_rate": 1.5988123216412378e-05,
"loss": 0.3317,
"step": 2799
},
{
"epoch": 1.8304724538172308,
"grad_norm": 4.75,
"learning_rate": 1.598446551637951e-05,
"loss": 0.413,
"step": 2800
},
{
"epoch": 1.8311263691351969,
"grad_norm": 5.25,
"learning_rate": 1.5980806568537314e-05,
"loss": 0.5384,
"step": 2801
},
{
"epoch": 1.8317802844531634,
"grad_norm": 5.09375,
"learning_rate": 1.59771463736487e-05,
"loss": 0.5367,
"step": 2802
},
{
"epoch": 1.8324341997711295,
"grad_norm": 5.375,
"learning_rate": 1.5973484932476852e-05,
"loss": 0.6226,
"step": 2803
},
{
"epoch": 1.833088115089096,
"grad_norm": 3.578125,
"learning_rate": 1.5969822245785208e-05,
"loss": 0.1735,
"step": 2804
},
{
"epoch": 1.8337420304070622,
"grad_norm": 3.703125,
"learning_rate": 1.5966158314337472e-05,
"loss": 0.4265,
"step": 2805
},
{
"epoch": 1.8343959457250287,
"grad_norm": 4.3125,
"learning_rate": 1.59624931388976e-05,
"loss": 0.3091,
"step": 2806
},
{
"epoch": 1.8350498610429948,
"grad_norm": 3.40625,
"learning_rate": 1.595882672022981e-05,
"loss": 0.1915,
"step": 2807
},
{
"epoch": 1.8357037763609614,
"grad_norm": 4.3125,
"learning_rate": 1.595515905909858e-05,
"loss": 0.3446,
"step": 2808
},
{
"epoch": 1.8363576916789275,
"grad_norm": 5.15625,
"learning_rate": 1.5951490156268648e-05,
"loss": 0.3179,
"step": 2809
},
{
"epoch": 1.837011606996894,
"grad_norm": 4.46875,
"learning_rate": 1.5947820012505016e-05,
"loss": 0.2468,
"step": 2810
},
{
"epoch": 1.8376655223148601,
"grad_norm": 3.84375,
"learning_rate": 1.5944148628572925e-05,
"loss": 0.3148,
"step": 2811
},
{
"epoch": 1.8383194376328267,
"grad_norm": 3.78125,
"learning_rate": 1.59404760052379e-05,
"loss": 0.1768,
"step": 2812
},
{
"epoch": 1.8389733529507928,
"grad_norm": 4.15625,
"learning_rate": 1.5936802143265708e-05,
"loss": 0.2221,
"step": 2813
},
{
"epoch": 1.8396272682687593,
"grad_norm": 4.375,
"learning_rate": 1.5933127043422377e-05,
"loss": 0.3589,
"step": 2814
},
{
"epoch": 1.8402811835867254,
"grad_norm": 6.71875,
"learning_rate": 1.59294507064742e-05,
"loss": 0.4085,
"step": 2815
},
{
"epoch": 1.840935098904692,
"grad_norm": 3.59375,
"learning_rate": 1.5925773133187713e-05,
"loss": 0.2665,
"step": 2816
},
{
"epoch": 1.841589014222658,
"grad_norm": 5.46875,
"learning_rate": 1.5922094324329737e-05,
"loss": 0.6065,
"step": 2817
},
{
"epoch": 1.8422429295406246,
"grad_norm": 4.4375,
"learning_rate": 1.5918414280667316e-05,
"loss": 0.3613,
"step": 2818
},
{
"epoch": 1.8428968448585907,
"grad_norm": 3.375,
"learning_rate": 1.591473300296778e-05,
"loss": 0.1902,
"step": 2819
},
{
"epoch": 1.8435507601765573,
"grad_norm": 4.84375,
"learning_rate": 1.5911050491998695e-05,
"loss": 0.2659,
"step": 2820
},
{
"epoch": 1.8442046754945234,
"grad_norm": 5.78125,
"learning_rate": 1.5907366748527904e-05,
"loss": 0.7058,
"step": 2821
},
{
"epoch": 1.84485859081249,
"grad_norm": 3.90625,
"learning_rate": 1.5903681773323493e-05,
"loss": 0.2153,
"step": 2822
},
{
"epoch": 1.845512506130456,
"grad_norm": 4.0,
"learning_rate": 1.5899995567153807e-05,
"loss": 0.2578,
"step": 2823
},
{
"epoch": 1.8461664214484226,
"grad_norm": 4.5,
"learning_rate": 1.5896308130787455e-05,
"loss": 0.4039,
"step": 2824
},
{
"epoch": 1.8468203367663887,
"grad_norm": 5.5,
"learning_rate": 1.589261946499329e-05,
"loss": 0.2625,
"step": 2825
},
{
"epoch": 1.8474742520843552,
"grad_norm": 6.40625,
"learning_rate": 1.5888929570540437e-05,
"loss": 0.5799,
"step": 2826
},
{
"epoch": 1.8481281674023213,
"grad_norm": 4.8125,
"learning_rate": 1.5885238448198267e-05,
"loss": 0.2307,
"step": 2827
},
{
"epoch": 1.8487820827202879,
"grad_norm": 4.65625,
"learning_rate": 1.58815460987364e-05,
"loss": 0.4324,
"step": 2828
},
{
"epoch": 1.849435998038254,
"grad_norm": 4.71875,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.3467,
"step": 2829
},
{
"epoch": 1.8500899133562205,
"grad_norm": 4.09375,
"learning_rate": 1.58741577215334e-05,
"loss": 0.2631,
"step": 2830
},
{
"epoch": 1.8507438286741866,
"grad_norm": 5.96875,
"learning_rate": 1.5870461695332792e-05,
"loss": 0.625,
"step": 2831
},
{
"epoch": 1.8513977439921532,
"grad_norm": 4.59375,
"learning_rate": 1.586676444509357e-05,
"loss": 0.2877,
"step": 2832
},
{
"epoch": 1.8520516593101193,
"grad_norm": 4.34375,
"learning_rate": 1.5863065971586637e-05,
"loss": 0.3224,
"step": 2833
},
{
"epoch": 1.8527055746280858,
"grad_norm": 3.890625,
"learning_rate": 1.585936627558315e-05,
"loss": 0.3383,
"step": 2834
},
{
"epoch": 1.853359489946052,
"grad_norm": 4.09375,
"learning_rate": 1.585566535785453e-05,
"loss": 0.4232,
"step": 2835
},
{
"epoch": 1.8540134052640183,
"grad_norm": 5.8125,
"learning_rate": 1.5851963219172447e-05,
"loss": 0.4894,
"step": 2836
},
{
"epoch": 1.8546673205819846,
"grad_norm": 8.3125,
"learning_rate": 1.584825986030883e-05,
"loss": 0.6866,
"step": 2837
},
{
"epoch": 1.855321235899951,
"grad_norm": 3.8125,
"learning_rate": 1.5844555282035857e-05,
"loss": 0.2407,
"step": 2838
},
{
"epoch": 1.8559751512179172,
"grad_norm": 3.71875,
"learning_rate": 1.5840849485125957e-05,
"loss": 0.3219,
"step": 2839
},
{
"epoch": 1.8566290665358836,
"grad_norm": 5.96875,
"learning_rate": 1.583714247035183e-05,
"loss": 0.4384,
"step": 2840
},
{
"epoch": 1.8572829818538499,
"grad_norm": 3.9375,
"learning_rate": 1.58334342384864e-05,
"loss": 0.3114,
"step": 2841
},
{
"epoch": 1.8579368971718162,
"grad_norm": 4.4375,
"learning_rate": 1.5829724790302886e-05,
"loss": 0.3496,
"step": 2842
},
{
"epoch": 1.8585908124897825,
"grad_norm": 5.59375,
"learning_rate": 1.5826014126574724e-05,
"loss": 0.632,
"step": 2843
},
{
"epoch": 1.8592447278077489,
"grad_norm": 6.3125,
"learning_rate": 1.582230224807562e-05,
"loss": 0.5539,
"step": 2844
},
{
"epoch": 1.8598986431257152,
"grad_norm": 4.625,
"learning_rate": 1.581858915557953e-05,
"loss": 0.4557,
"step": 2845
},
{
"epoch": 1.8605525584436815,
"grad_norm": 6.90625,
"learning_rate": 1.5814874849860664e-05,
"loss": 0.6591,
"step": 2846
},
{
"epoch": 1.8612064737616478,
"grad_norm": 4.03125,
"learning_rate": 1.5811159331693487e-05,
"loss": 0.1641,
"step": 2847
},
{
"epoch": 1.8618603890796142,
"grad_norm": 6.0625,
"learning_rate": 1.5807442601852713e-05,
"loss": 0.5674,
"step": 2848
},
{
"epoch": 1.8625143043975805,
"grad_norm": 2.796875,
"learning_rate": 1.580372466111331e-05,
"loss": 0.2026,
"step": 2849
},
{
"epoch": 1.8631682197155468,
"grad_norm": 6.0625,
"learning_rate": 1.5800005510250498e-05,
"loss": 0.5304,
"step": 2850
},
{
"epoch": 1.8638221350335131,
"grad_norm": 4.8125,
"learning_rate": 1.579628515003975e-05,
"loss": 0.3665,
"step": 2851
},
{
"epoch": 1.8644760503514795,
"grad_norm": 3.96875,
"learning_rate": 1.579256358125679e-05,
"loss": 0.4628,
"step": 2852
},
{
"epoch": 1.8651299656694458,
"grad_norm": 4.96875,
"learning_rate": 1.57888408046776e-05,
"loss": 0.3461,
"step": 2853
},
{
"epoch": 1.865783880987412,
"grad_norm": 6.5,
"learning_rate": 1.5785116821078407e-05,
"loss": 0.751,
"step": 2854
},
{
"epoch": 1.8664377963053784,
"grad_norm": 5.75,
"learning_rate": 1.5781391631235688e-05,
"loss": 0.3803,
"step": 2855
},
{
"epoch": 1.8670917116233448,
"grad_norm": 3.5,
"learning_rate": 1.5777665235926175e-05,
"loss": 0.2405,
"step": 2856
},
{
"epoch": 1.867745626941311,
"grad_norm": 4.75,
"learning_rate": 1.5773937635926854e-05,
"loss": 0.256,
"step": 2857
},
{
"epoch": 1.8683995422592774,
"grad_norm": 3.875,
"learning_rate": 1.5770208832014967e-05,
"loss": 0.2175,
"step": 2858
},
{
"epoch": 1.8690534575772437,
"grad_norm": 3.84375,
"learning_rate": 1.5766478824967985e-05,
"loss": 0.3517,
"step": 2859
},
{
"epoch": 1.86970737289521,
"grad_norm": 5.375,
"learning_rate": 1.5762747615563653e-05,
"loss": 0.4211,
"step": 2860
},
{
"epoch": 1.8703612882131764,
"grad_norm": 6.25,
"learning_rate": 1.5759015204579958e-05,
"loss": 0.5771,
"step": 2861
},
{
"epoch": 1.8710152035311427,
"grad_norm": 4.28125,
"learning_rate": 1.575528159279514e-05,
"loss": 0.3135,
"step": 2862
},
{
"epoch": 1.871669118849109,
"grad_norm": 3.9375,
"learning_rate": 1.575154678098768e-05,
"loss": 0.2801,
"step": 2863
},
{
"epoch": 1.8723230341670754,
"grad_norm": 6.1875,
"learning_rate": 1.5747810769936323e-05,
"loss": 0.5211,
"step": 2864
},
{
"epoch": 1.8729769494850417,
"grad_norm": 8.5,
"learning_rate": 1.574407356042005e-05,
"loss": 0.8193,
"step": 2865
},
{
"epoch": 1.873630864803008,
"grad_norm": 2.859375,
"learning_rate": 1.574033515321811e-05,
"loss": 0.2843,
"step": 2866
},
{
"epoch": 1.8742847801209743,
"grad_norm": 3.890625,
"learning_rate": 1.5736595549109983e-05,
"loss": 0.2588,
"step": 2867
},
{
"epoch": 1.8749386954389407,
"grad_norm": 5.15625,
"learning_rate": 1.573285474887541e-05,
"loss": 0.5259,
"step": 2868
},
{
"epoch": 1.875592610756907,
"grad_norm": 5.3125,
"learning_rate": 1.5729112753294377e-05,
"loss": 0.5223,
"step": 2869
},
{
"epoch": 1.8762465260748733,
"grad_norm": 4.0625,
"learning_rate": 1.572536956314712e-05,
"loss": 0.5173,
"step": 2870
},
{
"epoch": 1.8769004413928396,
"grad_norm": 5.78125,
"learning_rate": 1.5721625179214127e-05,
"loss": 0.6197,
"step": 2871
},
{
"epoch": 1.877554356710806,
"grad_norm": 5.46875,
"learning_rate": 1.5717879602276123e-05,
"loss": 0.3279,
"step": 2872
},
{
"epoch": 1.8782082720287723,
"grad_norm": 6.1875,
"learning_rate": 1.5714132833114102e-05,
"loss": 0.5165,
"step": 2873
},
{
"epoch": 1.8788621873467386,
"grad_norm": 5.53125,
"learning_rate": 1.5710384872509294e-05,
"loss": 0.5288,
"step": 2874
},
{
"epoch": 1.879516102664705,
"grad_norm": 6.65625,
"learning_rate": 1.5706635721243173e-05,
"loss": 0.4987,
"step": 2875
},
{
"epoch": 1.8801700179826712,
"grad_norm": 3.34375,
"learning_rate": 1.5702885380097474e-05,
"loss": 0.2589,
"step": 2876
},
{
"epoch": 1.8808239333006376,
"grad_norm": 4.125,
"learning_rate": 1.5699133849854164e-05,
"loss": 0.2522,
"step": 2877
},
{
"epoch": 1.881477848618604,
"grad_norm": 4.59375,
"learning_rate": 1.5695381131295477e-05,
"loss": 0.5025,
"step": 2878
},
{
"epoch": 1.8821317639365702,
"grad_norm": 5.125,
"learning_rate": 1.569162722520388e-05,
"loss": 0.4331,
"step": 2879
},
{
"epoch": 1.8827856792545365,
"grad_norm": 5.59375,
"learning_rate": 1.5687872132362098e-05,
"loss": 0.4599,
"step": 2880
},
{
"epoch": 1.8834395945725029,
"grad_norm": 3.78125,
"learning_rate": 1.5684115853553094e-05,
"loss": 0.3276,
"step": 2881
},
{
"epoch": 1.8840935098904692,
"grad_norm": 7.71875,
"learning_rate": 1.568035838956008e-05,
"loss": 0.4045,
"step": 2882
},
{
"epoch": 1.8847474252084355,
"grad_norm": 4.84375,
"learning_rate": 1.567659974116652e-05,
"loss": 0.2125,
"step": 2883
},
{
"epoch": 1.8854013405264018,
"grad_norm": 3.375,
"learning_rate": 1.5672839909156126e-05,
"loss": 0.2935,
"step": 2884
},
{
"epoch": 1.8860552558443682,
"grad_norm": 3.015625,
"learning_rate": 1.5669078894312848e-05,
"loss": 0.2583,
"step": 2885
},
{
"epoch": 1.8867091711623345,
"grad_norm": 4.5,
"learning_rate": 1.5665316697420892e-05,
"loss": 0.2579,
"step": 2886
},
{
"epoch": 1.8873630864803008,
"grad_norm": 4.03125,
"learning_rate": 1.5661553319264703e-05,
"loss": 0.294,
"step": 2887
},
{
"epoch": 1.8880170017982671,
"grad_norm": 7.78125,
"learning_rate": 1.5657788760628975e-05,
"loss": 0.6656,
"step": 2888
},
{
"epoch": 1.8886709171162335,
"grad_norm": 5.71875,
"learning_rate": 1.5654023022298653e-05,
"loss": 0.493,
"step": 2889
},
{
"epoch": 1.8893248324341998,
"grad_norm": 6.09375,
"learning_rate": 1.5650256105058924e-05,
"loss": 0.4917,
"step": 2890
},
{
"epoch": 1.8899787477521661,
"grad_norm": 7.375,
"learning_rate": 1.5646488009695214e-05,
"loss": 0.5341,
"step": 2891
},
{
"epoch": 1.8906326630701324,
"grad_norm": 5.46875,
"learning_rate": 1.564271873699321e-05,
"loss": 0.3911,
"step": 2892
},
{
"epoch": 1.8912865783880988,
"grad_norm": 5.78125,
"learning_rate": 1.563894828773883e-05,
"loss": 0.6198,
"step": 2893
},
{
"epoch": 1.891940493706065,
"grad_norm": 4.65625,
"learning_rate": 1.5635176662718246e-05,
"loss": 0.2467,
"step": 2894
},
{
"epoch": 1.8925944090240314,
"grad_norm": 5.375,
"learning_rate": 1.563140386271787e-05,
"loss": 0.3692,
"step": 2895
},
{
"epoch": 1.8932483243419977,
"grad_norm": 4.03125,
"learning_rate": 1.5627629888524363e-05,
"loss": 0.4595,
"step": 2896
},
{
"epoch": 1.893902239659964,
"grad_norm": 6.28125,
"learning_rate": 1.5623854740924627e-05,
"loss": 0.7549,
"step": 2897
},
{
"epoch": 1.8945561549779304,
"grad_norm": 5.78125,
"learning_rate": 1.5620078420705808e-05,
"loss": 0.5165,
"step": 2898
},
{
"epoch": 1.8952100702958967,
"grad_norm": 2.6875,
"learning_rate": 1.561630092865531e-05,
"loss": 0.2557,
"step": 2899
},
{
"epoch": 1.895863985613863,
"grad_norm": 4.46875,
"learning_rate": 1.561252226556076e-05,
"loss": 0.2571,
"step": 2900
},
{
"epoch": 1.8965179009318294,
"grad_norm": 2.78125,
"learning_rate": 1.5608742432210044e-05,
"loss": 0.1659,
"step": 2901
},
{
"epoch": 1.8971718162497957,
"grad_norm": 3.546875,
"learning_rate": 1.5604961429391288e-05,
"loss": 0.1788,
"step": 2902
},
{
"epoch": 1.897825731567762,
"grad_norm": 4.21875,
"learning_rate": 1.5601179257892857e-05,
"loss": 0.3353,
"step": 2903
},
{
"epoch": 1.8984796468857283,
"grad_norm": 6.96875,
"learning_rate": 1.5597395918503372e-05,
"loss": 0.2772,
"step": 2904
},
{
"epoch": 1.8991335622036947,
"grad_norm": 3.921875,
"learning_rate": 1.5593611412011685e-05,
"loss": 0.2724,
"step": 2905
},
{
"epoch": 1.899787477521661,
"grad_norm": 2.828125,
"learning_rate": 1.5589825739206896e-05,
"loss": 0.2077,
"step": 2906
},
{
"epoch": 1.9004413928396273,
"grad_norm": 7.03125,
"learning_rate": 1.5586038900878354e-05,
"loss": 0.5498,
"step": 2907
},
{
"epoch": 1.9010953081575936,
"grad_norm": 4.4375,
"learning_rate": 1.5582250897815633e-05,
"loss": 0.3436,
"step": 2908
},
{
"epoch": 1.90174922347556,
"grad_norm": 3.671875,
"learning_rate": 1.5578461730808575e-05,
"loss": 0.3306,
"step": 2909
},
{
"epoch": 1.9024031387935263,
"grad_norm": 3.9375,
"learning_rate": 1.5574671400647244e-05,
"loss": 0.3749,
"step": 2910
},
{
"epoch": 1.9030570541114926,
"grad_norm": 4.875,
"learning_rate": 1.557087990812196e-05,
"loss": 0.3918,
"step": 2911
},
{
"epoch": 1.9037109694294587,
"grad_norm": 2.4375,
"learning_rate": 1.5567087254023273e-05,
"loss": 0.1341,
"step": 2912
},
{
"epoch": 1.9043648847474253,
"grad_norm": 2.53125,
"learning_rate": 1.556329343914199e-05,
"loss": 0.1517,
"step": 2913
},
{
"epoch": 1.9050188000653914,
"grad_norm": 5.53125,
"learning_rate": 1.5559498464269143e-05,
"loss": 0.6812,
"step": 2914
},
{
"epoch": 1.905672715383358,
"grad_norm": 2.453125,
"learning_rate": 1.5555702330196024e-05,
"loss": 0.135,
"step": 2915
},
{
"epoch": 1.906326630701324,
"grad_norm": 4.65625,
"learning_rate": 1.555190503771415e-05,
"loss": 0.3905,
"step": 2916
},
{
"epoch": 1.9069805460192906,
"grad_norm": 4.4375,
"learning_rate": 1.5548106587615295e-05,
"loss": 0.3245,
"step": 2917
},
{
"epoch": 1.9076344613372567,
"grad_norm": 4.34375,
"learning_rate": 1.5544306980691456e-05,
"loss": 0.246,
"step": 2918
},
{
"epoch": 1.9082883766552232,
"grad_norm": 2.9375,
"learning_rate": 1.5540506217734892e-05,
"loss": 0.1999,
"step": 2919
},
{
"epoch": 1.9089422919731893,
"grad_norm": 4.625,
"learning_rate": 1.5536704299538088e-05,
"loss": 0.4023,
"step": 2920
},
{
"epoch": 1.9095962072911559,
"grad_norm": 4.75,
"learning_rate": 1.5532901226893775e-05,
"loss": 0.3479,
"step": 2921
},
{
"epoch": 1.910250122609122,
"grad_norm": 3.34375,
"learning_rate": 1.5529097000594924e-05,
"loss": 0.3292,
"step": 2922
},
{
"epoch": 1.9109040379270885,
"grad_norm": 4.09375,
"learning_rate": 1.5525291621434745e-05,
"loss": 0.4026,
"step": 2923
},
{
"epoch": 1.9115579532450546,
"grad_norm": 6.78125,
"learning_rate": 1.552148509020669e-05,
"loss": 0.6458,
"step": 2924
},
{
"epoch": 1.9122118685630212,
"grad_norm": 4.03125,
"learning_rate": 1.551767740770446e-05,
"loss": 0.335,
"step": 2925
},
{
"epoch": 1.9128657838809873,
"grad_norm": 4.0625,
"learning_rate": 1.5513868574721978e-05,
"loss": 0.3587,
"step": 2926
},
{
"epoch": 1.9135196991989538,
"grad_norm": 7.0,
"learning_rate": 1.5510058592053422e-05,
"loss": 0.6726,
"step": 2927
},
{
"epoch": 1.91417361451692,
"grad_norm": 5.1875,
"learning_rate": 1.55062474604932e-05,
"loss": 0.565,
"step": 2928
},
{
"epoch": 1.9148275298348865,
"grad_norm": 3.765625,
"learning_rate": 1.5502435180835966e-05,
"loss": 0.258,
"step": 2929
},
{
"epoch": 1.9154814451528526,
"grad_norm": 5.78125,
"learning_rate": 1.549862175387661e-05,
"loss": 0.5377,
"step": 2930
},
{
"epoch": 1.9161353604708191,
"grad_norm": 4.90625,
"learning_rate": 1.5494807180410268e-05,
"loss": 0.5392,
"step": 2931
},
{
"epoch": 1.9167892757887852,
"grad_norm": 7.71875,
"learning_rate": 1.54909914612323e-05,
"loss": 0.9758,
"step": 2932
},
{
"epoch": 1.9174431911067518,
"grad_norm": 5.53125,
"learning_rate": 1.5487174597138314e-05,
"loss": 0.2172,
"step": 2933
},
{
"epoch": 1.9180971064247179,
"grad_norm": 4.28125,
"learning_rate": 1.5483356588924168e-05,
"loss": 0.3737,
"step": 2934
},
{
"epoch": 1.9187510217426844,
"grad_norm": 3.953125,
"learning_rate": 1.5479537437385938e-05,
"loss": 0.3086,
"step": 2935
},
{
"epoch": 1.9194049370606505,
"grad_norm": 4.40625,
"learning_rate": 1.5475717143319955e-05,
"loss": 0.4248,
"step": 2936
},
{
"epoch": 1.920058852378617,
"grad_norm": 4.84375,
"learning_rate": 1.547189570752277e-05,
"loss": 0.4027,
"step": 2937
},
{
"epoch": 1.9207127676965832,
"grad_norm": 5.125,
"learning_rate": 1.5468073130791196e-05,
"loss": 0.4885,
"step": 2938
},
{
"epoch": 1.9213666830145497,
"grad_norm": 3.84375,
"learning_rate": 1.546424941392226e-05,
"loss": 0.1807,
"step": 2939
},
{
"epoch": 1.9220205983325158,
"grad_norm": 4.40625,
"learning_rate": 1.546042455771325e-05,
"loss": 0.2951,
"step": 2940
},
{
"epoch": 1.9226745136504824,
"grad_norm": 6.03125,
"learning_rate": 1.5456598562961666e-05,
"loss": 0.6989,
"step": 2941
},
{
"epoch": 1.9233284289684485,
"grad_norm": 4.3125,
"learning_rate": 1.5452771430465267e-05,
"loss": 0.2466,
"step": 2942
},
{
"epoch": 1.923982344286415,
"grad_norm": 4.40625,
"learning_rate": 1.5448943161022042e-05,
"loss": 0.2303,
"step": 2943
},
{
"epoch": 1.9246362596043811,
"grad_norm": 6.5625,
"learning_rate": 1.544511375543021e-05,
"loss": 0.7212,
"step": 2944
},
{
"epoch": 1.9252901749223477,
"grad_norm": 4.3125,
"learning_rate": 1.544128321448824e-05,
"loss": 0.3084,
"step": 2945
},
{
"epoch": 1.9259440902403138,
"grad_norm": 5.0625,
"learning_rate": 1.5437451538994822e-05,
"loss": 0.3196,
"step": 2946
},
{
"epoch": 1.9265980055582803,
"grad_norm": 5.21875,
"learning_rate": 1.54336187297489e-05,
"loss": 0.6352,
"step": 2947
},
{
"epoch": 1.9272519208762464,
"grad_norm": 3.6875,
"learning_rate": 1.542978478754964e-05,
"loss": 0.3767,
"step": 2948
},
{
"epoch": 1.927905836194213,
"grad_norm": 4.09375,
"learning_rate": 1.542594971319645e-05,
"loss": 0.1975,
"step": 2949
},
{
"epoch": 1.928559751512179,
"grad_norm": 8.1875,
"learning_rate": 1.542211350748898e-05,
"loss": 0.763,
"step": 2950
},
{
"epoch": 1.9292136668301456,
"grad_norm": 4.96875,
"learning_rate": 1.5418276171227106e-05,
"loss": 0.5267,
"step": 2951
},
{
"epoch": 1.9298675821481117,
"grad_norm": 4.1875,
"learning_rate": 1.541443770521094e-05,
"loss": 0.3237,
"step": 2952
},
{
"epoch": 1.9305214974660783,
"grad_norm": 4.5,
"learning_rate": 1.5410598110240832e-05,
"loss": 0.5167,
"step": 2953
},
{
"epoch": 1.9311754127840444,
"grad_norm": 5.59375,
"learning_rate": 1.5406757387117376e-05,
"loss": 0.4539,
"step": 2954
},
{
"epoch": 1.931829328102011,
"grad_norm": 3.875,
"learning_rate": 1.5402915536641392e-05,
"loss": 0.2537,
"step": 2955
},
{
"epoch": 1.932483243419977,
"grad_norm": 3.625,
"learning_rate": 1.5399072559613936e-05,
"loss": 0.2882,
"step": 2956
},
{
"epoch": 1.9331371587379436,
"grad_norm": 6.84375,
"learning_rate": 1.5395228456836298e-05,
"loss": 0.6184,
"step": 2957
},
{
"epoch": 1.9337910740559097,
"grad_norm": 5.90625,
"learning_rate": 1.5391383229110005e-05,
"loss": 0.7296,
"step": 2958
},
{
"epoch": 1.9344449893738762,
"grad_norm": 4.78125,
"learning_rate": 1.5387536877236814e-05,
"loss": 0.4951,
"step": 2959
},
{
"epoch": 1.9350989046918423,
"grad_norm": 6.875,
"learning_rate": 1.538368940201873e-05,
"loss": 0.5741,
"step": 2960
},
{
"epoch": 1.9357528200098089,
"grad_norm": 4.15625,
"learning_rate": 1.537984080425798e-05,
"loss": 0.2551,
"step": 2961
},
{
"epoch": 1.936406735327775,
"grad_norm": 5.0,
"learning_rate": 1.537599108475702e-05,
"loss": 0.2708,
"step": 2962
},
{
"epoch": 1.9370606506457415,
"grad_norm": 6.375,
"learning_rate": 1.5372140244318553e-05,
"loss": 0.6566,
"step": 2963
},
{
"epoch": 1.9377145659637076,
"grad_norm": 6.0625,
"learning_rate": 1.536828828374551e-05,
"loss": 0.5313,
"step": 2964
},
{
"epoch": 1.9383684812816742,
"grad_norm": 4.09375,
"learning_rate": 1.5364435203841058e-05,
"loss": 0.4291,
"step": 2965
},
{
"epoch": 1.9390223965996403,
"grad_norm": 4.1875,
"learning_rate": 1.5360581005408595e-05,
"loss": 0.3475,
"step": 2966
},
{
"epoch": 1.9396763119176068,
"grad_norm": 5.8125,
"learning_rate": 1.5356725689251746e-05,
"loss": 0.6597,
"step": 2967
},
{
"epoch": 1.940330227235573,
"grad_norm": 4.5,
"learning_rate": 1.535286925617438e-05,
"loss": 0.2965,
"step": 2968
},
{
"epoch": 1.9409841425535395,
"grad_norm": 4.0,
"learning_rate": 1.5349011706980598e-05,
"loss": 0.3043,
"step": 2969
},
{
"epoch": 1.9416380578715056,
"grad_norm": 5.53125,
"learning_rate": 1.5345153042474723e-05,
"loss": 0.3187,
"step": 2970
},
{
"epoch": 1.942291973189472,
"grad_norm": 5.0,
"learning_rate": 1.534129326346132e-05,
"loss": 0.4415,
"step": 2971
},
{
"epoch": 1.9429458885074382,
"grad_norm": 4.6875,
"learning_rate": 1.533743237074519e-05,
"loss": 0.5067,
"step": 2972
},
{
"epoch": 1.9435998038254048,
"grad_norm": 4.6875,
"learning_rate": 1.5333570365131353e-05,
"loss": 0.3121,
"step": 2973
},
{
"epoch": 1.9442537191433709,
"grad_norm": 6.53125,
"learning_rate": 1.5329707247425072e-05,
"loss": 0.5213,
"step": 2974
},
{
"epoch": 1.9449076344613374,
"grad_norm": 5.09375,
"learning_rate": 1.5325843018431832e-05,
"loss": 0.4849,
"step": 2975
},
{
"epoch": 1.9455615497793035,
"grad_norm": 5.09375,
"learning_rate": 1.532197767895737e-05,
"loss": 0.4572,
"step": 2976
},
{
"epoch": 1.94621546509727,
"grad_norm": 4.5625,
"learning_rate": 1.5318111229807623e-05,
"loss": 0.3768,
"step": 2977
},
{
"epoch": 1.9468693804152362,
"grad_norm": 6.9375,
"learning_rate": 1.531424367178879e-05,
"loss": 0.6258,
"step": 2978
},
{
"epoch": 1.9475232957332027,
"grad_norm": 4.53125,
"learning_rate": 1.5310375005707284e-05,
"loss": 0.3077,
"step": 2979
},
{
"epoch": 1.9481772110511688,
"grad_norm": 5.59375,
"learning_rate": 1.530650523236975e-05,
"loss": 0.4776,
"step": 2980
},
{
"epoch": 1.9488311263691354,
"grad_norm": 4.1875,
"learning_rate": 1.530263435258307e-05,
"loss": 0.3434,
"step": 2981
},
{
"epoch": 1.9494850416871015,
"grad_norm": 5.65625,
"learning_rate": 1.5298762367154355e-05,
"loss": 0.4426,
"step": 2982
},
{
"epoch": 1.950138957005068,
"grad_norm": 7.0,
"learning_rate": 1.5294889276890942e-05,
"loss": 0.7048,
"step": 2983
},
{
"epoch": 1.950792872323034,
"grad_norm": 6.90625,
"learning_rate": 1.52910150826004e-05,
"loss": 0.5269,
"step": 2984
},
{
"epoch": 1.9514467876410004,
"grad_norm": 3.75,
"learning_rate": 1.5287139785090534e-05,
"loss": 0.3045,
"step": 2985
},
{
"epoch": 1.9521007029589668,
"grad_norm": 7.21875,
"learning_rate": 1.5283263385169376e-05,
"loss": 0.6063,
"step": 2986
},
{
"epoch": 1.952754618276933,
"grad_norm": 3.5625,
"learning_rate": 1.5279385883645184e-05,
"loss": 0.2593,
"step": 2987
},
{
"epoch": 1.9534085335948994,
"grad_norm": 4.5,
"learning_rate": 1.5275507281326444e-05,
"loss": 0.3916,
"step": 2988
},
{
"epoch": 1.9540624489128657,
"grad_norm": 4.84375,
"learning_rate": 1.5271627579021885e-05,
"loss": 0.4911,
"step": 2989
},
{
"epoch": 1.954716364230832,
"grad_norm": 4.96875,
"learning_rate": 1.5267746777540445e-05,
"loss": 0.306,
"step": 2990
},
{
"epoch": 1.9553702795487984,
"grad_norm": 3.5625,
"learning_rate": 1.5263864877691314e-05,
"loss": 0.3657,
"step": 2991
},
{
"epoch": 1.9560241948667647,
"grad_norm": 6.59375,
"learning_rate": 1.5259981880283894e-05,
"loss": 0.5406,
"step": 2992
},
{
"epoch": 1.956678110184731,
"grad_norm": 4.4375,
"learning_rate": 1.5256097786127822e-05,
"loss": 0.3487,
"step": 2993
},
{
"epoch": 1.9573320255026974,
"grad_norm": 4.09375,
"learning_rate": 1.525221259603296e-05,
"loss": 0.4059,
"step": 2994
},
{
"epoch": 1.9579859408206637,
"grad_norm": 5.4375,
"learning_rate": 1.5248326310809404e-05,
"loss": 0.6772,
"step": 2995
},
{
"epoch": 1.95863985613863,
"grad_norm": 2.65625,
"learning_rate": 1.524443893126748e-05,
"loss": 0.1832,
"step": 2996
},
{
"epoch": 1.9592937714565963,
"grad_norm": 3.765625,
"learning_rate": 1.524055045821773e-05,
"loss": 0.3591,
"step": 2997
},
{
"epoch": 1.9599476867745627,
"grad_norm": 5.0625,
"learning_rate": 1.5236660892470939e-05,
"loss": 0.2853,
"step": 2998
},
{
"epoch": 1.960601602092529,
"grad_norm": 3.625,
"learning_rate": 1.5232770234838108e-05,
"loss": 0.3359,
"step": 2999
},
{
"epoch": 1.9612555174104953,
"grad_norm": 4.9375,
"learning_rate": 1.522887848613047e-05,
"loss": 0.295,
"step": 3000
},
{
"epoch": 1.9619094327284616,
"grad_norm": 4.59375,
"learning_rate": 1.5224985647159489e-05,
"loss": 0.406,
"step": 3001
},
{
"epoch": 1.962563348046428,
"grad_norm": 5.71875,
"learning_rate": 1.5221091718736856e-05,
"loss": 0.611,
"step": 3002
},
{
"epoch": 1.9632172633643943,
"grad_norm": 6.59375,
"learning_rate": 1.521719670167448e-05,
"loss": 0.8252,
"step": 3003
},
{
"epoch": 1.9638711786823606,
"grad_norm": 5.15625,
"learning_rate": 1.5213300596784506e-05,
"loss": 0.3607,
"step": 3004
},
{
"epoch": 1.964525094000327,
"grad_norm": 4.21875,
"learning_rate": 1.5209403404879305e-05,
"loss": 0.2729,
"step": 3005
},
{
"epoch": 1.9651790093182933,
"grad_norm": 5.75,
"learning_rate": 1.5205505126771471e-05,
"loss": 0.4727,
"step": 3006
},
{
"epoch": 1.9658329246362596,
"grad_norm": 4.15625,
"learning_rate": 1.5201605763273826e-05,
"loss": 0.2781,
"step": 3007
},
{
"epoch": 1.966486839954226,
"grad_norm": 7.375,
"learning_rate": 1.5197705315199422e-05,
"loss": 0.6233,
"step": 3008
},
{
"epoch": 1.9671407552721922,
"grad_norm": 5.46875,
"learning_rate": 1.5193803783361531e-05,
"loss": 0.6368,
"step": 3009
},
{
"epoch": 1.9677946705901586,
"grad_norm": 5.40625,
"learning_rate": 1.5189901168573652e-05,
"loss": 0.5631,
"step": 3010
},
{
"epoch": 1.9684485859081249,
"grad_norm": 5.03125,
"learning_rate": 1.5185997471649519e-05,
"loss": 0.5366,
"step": 3011
},
{
"epoch": 1.9691025012260912,
"grad_norm": 6.09375,
"learning_rate": 1.5182092693403076e-05,
"loss": 0.6875,
"step": 3012
},
{
"epoch": 1.9697564165440575,
"grad_norm": 6.21875,
"learning_rate": 1.5178186834648509e-05,
"loss": 0.6573,
"step": 3013
},
{
"epoch": 1.9704103318620239,
"grad_norm": 3.828125,
"learning_rate": 1.5174279896200212e-05,
"loss": 0.1843,
"step": 3014
},
{
"epoch": 1.9710642471799902,
"grad_norm": 4.3125,
"learning_rate": 1.5170371878872818e-05,
"loss": 0.2257,
"step": 3015
},
{
"epoch": 1.9717181624979565,
"grad_norm": 4.75,
"learning_rate": 1.5166462783481182e-05,
"loss": 0.4537,
"step": 3016
},
{
"epoch": 1.9723720778159228,
"grad_norm": 5.4375,
"learning_rate": 1.5162552610840378e-05,
"loss": 0.4182,
"step": 3017
},
{
"epoch": 1.9730259931338892,
"grad_norm": 5.0625,
"learning_rate": 1.5158641361765711e-05,
"loss": 0.3654,
"step": 3018
},
{
"epoch": 1.9736799084518555,
"grad_norm": 5.75,
"learning_rate": 1.5154729037072708e-05,
"loss": 0.3916,
"step": 3019
},
{
"epoch": 1.9743338237698218,
"grad_norm": 3.796875,
"learning_rate": 1.5150815637577117e-05,
"loss": 0.3284,
"step": 3020
},
{
"epoch": 1.9749877390877881,
"grad_norm": 8.5625,
"learning_rate": 1.5146901164094914e-05,
"loss": 1.0165,
"step": 3021
},
{
"epoch": 1.9756416544057545,
"grad_norm": 4.09375,
"learning_rate": 1.5142985617442305e-05,
"loss": 0.2998,
"step": 3022
},
{
"epoch": 1.9762955697237208,
"grad_norm": 5.46875,
"learning_rate": 1.5139068998435707e-05,
"loss": 0.4552,
"step": 3023
},
{
"epoch": 1.976949485041687,
"grad_norm": 7.6875,
"learning_rate": 1.5135151307891766e-05,
"loss": 0.687,
"step": 3024
},
{
"epoch": 1.9776034003596534,
"grad_norm": 7.875,
"learning_rate": 1.5131232546627355e-05,
"loss": 0.4285,
"step": 3025
},
{
"epoch": 1.9782573156776198,
"grad_norm": 4.1875,
"learning_rate": 1.5127312715459565e-05,
"loss": 0.3683,
"step": 3026
},
{
"epoch": 1.978911230995586,
"grad_norm": 4.5625,
"learning_rate": 1.5123391815205715e-05,
"loss": 0.3114,
"step": 3027
},
{
"epoch": 1.9795651463135524,
"grad_norm": 5.4375,
"learning_rate": 1.5119469846683341e-05,
"loss": 0.4644,
"step": 3028
},
{
"epoch": 1.9802190616315187,
"grad_norm": 4.625,
"learning_rate": 1.5115546810710208e-05,
"loss": 0.4053,
"step": 3029
},
{
"epoch": 1.980872976949485,
"grad_norm": 4.0,
"learning_rate": 1.5111622708104304e-05,
"loss": 0.2684,
"step": 3030
},
{
"epoch": 1.9815268922674514,
"grad_norm": 4.625,
"learning_rate": 1.5107697539683828e-05,
"loss": 0.5073,
"step": 3031
},
{
"epoch": 1.9821808075854177,
"grad_norm": 6.34375,
"learning_rate": 1.5103771306267212e-05,
"loss": 0.5601,
"step": 3032
},
{
"epoch": 1.982834722903384,
"grad_norm": 2.828125,
"learning_rate": 1.5099844008673109e-05,
"loss": 0.1555,
"step": 3033
},
{
"epoch": 1.9834886382213504,
"grad_norm": 5.0,
"learning_rate": 1.509591564772039e-05,
"loss": 0.4751,
"step": 3034
},
{
"epoch": 1.9841425535393167,
"grad_norm": 4.5,
"learning_rate": 1.5091986224228156e-05,
"loss": 0.2711,
"step": 3035
},
{
"epoch": 1.984796468857283,
"grad_norm": 3.8125,
"learning_rate": 1.5088055739015715e-05,
"loss": 0.3211,
"step": 3036
},
{
"epoch": 1.9854503841752493,
"grad_norm": 4.0625,
"learning_rate": 1.5084124192902612e-05,
"loss": 0.2014,
"step": 3037
},
{
"epoch": 1.9861042994932157,
"grad_norm": 6.125,
"learning_rate": 1.5080191586708601e-05,
"loss": 0.6582,
"step": 3038
},
{
"epoch": 1.986758214811182,
"grad_norm": 6.03125,
"learning_rate": 1.5076257921253664e-05,
"loss": 0.472,
"step": 3039
},
{
"epoch": 1.9874121301291483,
"grad_norm": 7.5,
"learning_rate": 1.5072323197358003e-05,
"loss": 0.6755,
"step": 3040
},
{
"epoch": 1.9880660454471146,
"grad_norm": 4.78125,
"learning_rate": 1.5068387415842036e-05,
"loss": 0.4495,
"step": 3041
},
{
"epoch": 1.988719960765081,
"grad_norm": 4.625,
"learning_rate": 1.5064450577526412e-05,
"loss": 0.3785,
"step": 3042
},
{
"epoch": 1.9893738760830473,
"grad_norm": 3.296875,
"learning_rate": 1.5060512683231987e-05,
"loss": 0.2914,
"step": 3043
},
{
"epoch": 1.9900277914010136,
"grad_norm": 6.03125,
"learning_rate": 1.5056573733779848e-05,
"loss": 0.3983,
"step": 3044
},
{
"epoch": 1.99068170671898,
"grad_norm": 3.953125,
"learning_rate": 1.5052633729991296e-05,
"loss": 0.4056,
"step": 3045
},
{
"epoch": 1.9913356220369463,
"grad_norm": 3.765625,
"learning_rate": 1.5048692672687851e-05,
"loss": 0.6382,
"step": 3046
},
{
"epoch": 1.9919895373549126,
"grad_norm": 4.25,
"learning_rate": 1.5044750562691256e-05,
"loss": 0.2847,
"step": 3047
},
{
"epoch": 1.992643452672879,
"grad_norm": 4.90625,
"learning_rate": 1.5040807400823481e-05,
"loss": 0.4152,
"step": 3048
},
{
"epoch": 1.9932973679908452,
"grad_norm": 5.21875,
"learning_rate": 1.5036863187906696e-05,
"loss": 0.3964,
"step": 3049
},
{
"epoch": 1.9939512833088116,
"grad_norm": 4.84375,
"learning_rate": 1.5032917924763309e-05,
"loss": 0.289,
"step": 3050
},
{
"epoch": 1.9946051986267779,
"grad_norm": 6.03125,
"learning_rate": 1.502897161221593e-05,
"loss": 0.515,
"step": 3051
},
{
"epoch": 1.9952591139447442,
"grad_norm": 3.515625,
"learning_rate": 1.502502425108741e-05,
"loss": 0.2944,
"step": 3052
},
{
"epoch": 1.9959130292627105,
"grad_norm": 4.53125,
"learning_rate": 1.5021075842200796e-05,
"loss": 0.4392,
"step": 3053
},
{
"epoch": 1.9965669445806769,
"grad_norm": 5.625,
"learning_rate": 1.5017126386379364e-05,
"loss": 0.4099,
"step": 3054
},
{
"epoch": 1.9972208598986432,
"grad_norm": 2.8125,
"learning_rate": 1.501317588444661e-05,
"loss": 0.1868,
"step": 3055
},
{
"epoch": 1.9978747752166095,
"grad_norm": 4.34375,
"learning_rate": 1.5009224337226245e-05,
"loss": 0.3489,
"step": 3056
},
{
"epoch": 1.9985286905345758,
"grad_norm": 4.78125,
"learning_rate": 1.5005271745542196e-05,
"loss": 0.4906,
"step": 3057
},
{
"epoch": 1.9991826058525421,
"grad_norm": 4.3125,
"learning_rate": 1.5001318110218615e-05,
"loss": 0.3178,
"step": 3058
},
{
"epoch": 1.9998365211705085,
"grad_norm": 3.765625,
"learning_rate": 1.4997363432079863e-05,
"loss": 0.2755,
"step": 3059
},
{
"epoch": 2.0,
"grad_norm": 4.125,
"learning_rate": 1.4993407711950525e-05,
"loss": 0.0722,
"step": 3060
}
],
"logging_steps": 1.0,
"max_steps": 7645,
"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": false
},
"attributes": {}
}
},
"total_flos": 1.1313906262778112e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}