narabzad's picture
Add files using upload-large-folder tool
5072e16 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 7.0,
"eval_steps": 500,
"global_step": 875,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.008,
"grad_norm": 3.784468173980713,
"learning_rate": 0.0,
"loss": 0.8231,
"step": 1
},
{
"epoch": 0.016,
"grad_norm": 3.725552797317505,
"learning_rate": 1.7543859649122808e-07,
"loss": 1.0003,
"step": 2
},
{
"epoch": 0.024,
"grad_norm": 3.5788533687591553,
"learning_rate": 3.5087719298245616e-07,
"loss": 0.8452,
"step": 3
},
{
"epoch": 0.032,
"grad_norm": 3.4994688034057617,
"learning_rate": 5.263157894736843e-07,
"loss": 0.9015,
"step": 4
},
{
"epoch": 0.04,
"grad_norm": 3.340575695037842,
"learning_rate": 7.017543859649123e-07,
"loss": 0.91,
"step": 5
},
{
"epoch": 0.048,
"grad_norm": 3.4169387817382812,
"learning_rate": 8.771929824561404e-07,
"loss": 0.8981,
"step": 6
},
{
"epoch": 0.056,
"grad_norm": 3.199388027191162,
"learning_rate": 1.0526315789473685e-06,
"loss": 0.8082,
"step": 7
},
{
"epoch": 0.064,
"grad_norm": 3.5631041526794434,
"learning_rate": 1.2280701754385965e-06,
"loss": 1.0333,
"step": 8
},
{
"epoch": 0.072,
"grad_norm": 2.9914121627807617,
"learning_rate": 1.4035087719298246e-06,
"loss": 0.9455,
"step": 9
},
{
"epoch": 0.08,
"grad_norm": 3.418433904647827,
"learning_rate": 1.5789473684210526e-06,
"loss": 0.7966,
"step": 10
},
{
"epoch": 0.088,
"grad_norm": 2.4177350997924805,
"learning_rate": 1.7543859649122807e-06,
"loss": 0.9202,
"step": 11
},
{
"epoch": 0.096,
"grad_norm": 1.7621887922286987,
"learning_rate": 1.929824561403509e-06,
"loss": 0.8102,
"step": 12
},
{
"epoch": 0.104,
"grad_norm": 2.0125510692596436,
"learning_rate": 2.105263157894737e-06,
"loss": 0.8969,
"step": 13
},
{
"epoch": 0.112,
"grad_norm": 1.7342252731323242,
"learning_rate": 2.280701754385965e-06,
"loss": 0.8587,
"step": 14
},
{
"epoch": 0.12,
"grad_norm": 1.712204098701477,
"learning_rate": 2.456140350877193e-06,
"loss": 0.8708,
"step": 15
},
{
"epoch": 0.128,
"grad_norm": 1.412792682647705,
"learning_rate": 2.631578947368421e-06,
"loss": 0.8413,
"step": 16
},
{
"epoch": 0.136,
"grad_norm": 1.5133228302001953,
"learning_rate": 2.8070175438596493e-06,
"loss": 0.7527,
"step": 17
},
{
"epoch": 0.144,
"grad_norm": 1.6985952854156494,
"learning_rate": 2.9824561403508774e-06,
"loss": 0.9278,
"step": 18
},
{
"epoch": 0.152,
"grad_norm": 1.118333101272583,
"learning_rate": 3.157894736842105e-06,
"loss": 0.7364,
"step": 19
},
{
"epoch": 0.16,
"grad_norm": 1.1672853231430054,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.8443,
"step": 20
},
{
"epoch": 0.168,
"grad_norm": 1.0605567693710327,
"learning_rate": 3.5087719298245615e-06,
"loss": 0.714,
"step": 21
},
{
"epoch": 0.176,
"grad_norm": 1.2411671876907349,
"learning_rate": 3.6842105263157896e-06,
"loss": 0.8215,
"step": 22
},
{
"epoch": 0.184,
"grad_norm": 1.144175410270691,
"learning_rate": 3.859649122807018e-06,
"loss": 0.6955,
"step": 23
},
{
"epoch": 0.192,
"grad_norm": 1.3835996389389038,
"learning_rate": 4.035087719298246e-06,
"loss": 0.7968,
"step": 24
},
{
"epoch": 0.2,
"grad_norm": 1.1685224771499634,
"learning_rate": 4.210526315789474e-06,
"loss": 0.7425,
"step": 25
},
{
"epoch": 0.208,
"grad_norm": 1.1630430221557617,
"learning_rate": 4.385964912280702e-06,
"loss": 0.7122,
"step": 26
},
{
"epoch": 0.216,
"grad_norm": 1.054180383682251,
"learning_rate": 4.56140350877193e-06,
"loss": 0.7967,
"step": 27
},
{
"epoch": 0.224,
"grad_norm": 1.0310646295547485,
"learning_rate": 4.736842105263158e-06,
"loss": 0.6822,
"step": 28
},
{
"epoch": 0.232,
"grad_norm": 1.1047680377960205,
"learning_rate": 4.912280701754386e-06,
"loss": 0.739,
"step": 29
},
{
"epoch": 0.24,
"grad_norm": 1.712615966796875,
"learning_rate": 5.087719298245615e-06,
"loss": 0.7789,
"step": 30
},
{
"epoch": 0.248,
"grad_norm": 1.1467419862747192,
"learning_rate": 5.263157894736842e-06,
"loss": 0.7517,
"step": 31
},
{
"epoch": 0.256,
"grad_norm": 0.9748108386993408,
"learning_rate": 5.438596491228071e-06,
"loss": 0.7997,
"step": 32
},
{
"epoch": 0.264,
"grad_norm": 1.1908164024353027,
"learning_rate": 5.6140350877192985e-06,
"loss": 0.7893,
"step": 33
},
{
"epoch": 0.272,
"grad_norm": 1.160500168800354,
"learning_rate": 5.789473684210527e-06,
"loss": 0.7667,
"step": 34
},
{
"epoch": 0.28,
"grad_norm": 1.0798025131225586,
"learning_rate": 5.964912280701755e-06,
"loss": 0.6695,
"step": 35
},
{
"epoch": 0.288,
"grad_norm": 1.128517746925354,
"learning_rate": 6.140350877192983e-06,
"loss": 0.7281,
"step": 36
},
{
"epoch": 0.296,
"grad_norm": 1.1588135957717896,
"learning_rate": 6.31578947368421e-06,
"loss": 0.6856,
"step": 37
},
{
"epoch": 0.304,
"grad_norm": 0.9977500438690186,
"learning_rate": 6.491228070175439e-06,
"loss": 0.7085,
"step": 38
},
{
"epoch": 0.312,
"grad_norm": 1.0899662971496582,
"learning_rate": 6.666666666666667e-06,
"loss": 0.7503,
"step": 39
},
{
"epoch": 0.32,
"grad_norm": 1.0250076055526733,
"learning_rate": 6.842105263157896e-06,
"loss": 0.7469,
"step": 40
},
{
"epoch": 0.328,
"grad_norm": 1.0345467329025269,
"learning_rate": 7.017543859649123e-06,
"loss": 0.7034,
"step": 41
},
{
"epoch": 0.336,
"grad_norm": 0.9953707456588745,
"learning_rate": 7.192982456140352e-06,
"loss": 0.8141,
"step": 42
},
{
"epoch": 0.344,
"grad_norm": 1.167507290840149,
"learning_rate": 7.368421052631579e-06,
"loss": 0.7411,
"step": 43
},
{
"epoch": 0.352,
"grad_norm": 0.9854204058647156,
"learning_rate": 7.5438596491228074e-06,
"loss": 0.7219,
"step": 44
},
{
"epoch": 0.36,
"grad_norm": 1.1585136651992798,
"learning_rate": 7.719298245614036e-06,
"loss": 0.7666,
"step": 45
},
{
"epoch": 0.368,
"grad_norm": 0.9028245210647583,
"learning_rate": 7.894736842105265e-06,
"loss": 0.677,
"step": 46
},
{
"epoch": 0.376,
"grad_norm": 1.0796241760253906,
"learning_rate": 8.070175438596492e-06,
"loss": 0.7579,
"step": 47
},
{
"epoch": 0.384,
"grad_norm": 0.8502395153045654,
"learning_rate": 8.24561403508772e-06,
"loss": 0.6178,
"step": 48
},
{
"epoch": 0.392,
"grad_norm": 1.0221744775772095,
"learning_rate": 8.421052631578948e-06,
"loss": 0.7143,
"step": 49
},
{
"epoch": 0.4,
"grad_norm": 1.1040499210357666,
"learning_rate": 8.596491228070176e-06,
"loss": 0.739,
"step": 50
},
{
"epoch": 0.408,
"grad_norm": 1.3160624504089355,
"learning_rate": 8.771929824561405e-06,
"loss": 0.7673,
"step": 51
},
{
"epoch": 0.416,
"grad_norm": 0.9868547916412354,
"learning_rate": 8.947368421052632e-06,
"loss": 0.6158,
"step": 52
},
{
"epoch": 0.424,
"grad_norm": 0.97400963306427,
"learning_rate": 9.12280701754386e-06,
"loss": 0.7602,
"step": 53
},
{
"epoch": 0.432,
"grad_norm": 0.9297606945037842,
"learning_rate": 9.298245614035088e-06,
"loss": 0.6972,
"step": 54
},
{
"epoch": 0.44,
"grad_norm": 0.9819108247756958,
"learning_rate": 9.473684210526315e-06,
"loss": 0.6661,
"step": 55
},
{
"epoch": 0.448,
"grad_norm": 0.9263789057731628,
"learning_rate": 9.649122807017545e-06,
"loss": 0.6548,
"step": 56
},
{
"epoch": 0.456,
"grad_norm": 1.1375892162322998,
"learning_rate": 9.824561403508772e-06,
"loss": 0.7532,
"step": 57
},
{
"epoch": 0.464,
"grad_norm": 0.9523381590843201,
"learning_rate": 1e-05,
"loss": 0.7023,
"step": 58
},
{
"epoch": 0.472,
"grad_norm": 1.0329796075820923,
"learning_rate": 9.999978367986988e-06,
"loss": 0.6951,
"step": 59
},
{
"epoch": 0.48,
"grad_norm": 0.9789192080497742,
"learning_rate": 9.999913472135126e-06,
"loss": 0.7022,
"step": 60
},
{
"epoch": 0.488,
"grad_norm": 1.0567715167999268,
"learning_rate": 9.999805313005946e-06,
"loss": 0.7192,
"step": 61
},
{
"epoch": 0.496,
"grad_norm": 0.9230039715766907,
"learning_rate": 9.99965389153533e-06,
"loss": 0.6258,
"step": 62
},
{
"epoch": 0.504,
"grad_norm": 1.0457018613815308,
"learning_rate": 9.999459209033495e-06,
"loss": 0.7205,
"step": 63
},
{
"epoch": 0.512,
"grad_norm": 0.8816724419593811,
"learning_rate": 9.999221267184993e-06,
"loss": 0.6797,
"step": 64
},
{
"epoch": 0.52,
"grad_norm": 0.7928981184959412,
"learning_rate": 9.998940068048688e-06,
"loss": 0.6074,
"step": 65
},
{
"epoch": 0.528,
"grad_norm": 0.911626398563385,
"learning_rate": 9.998615614057743e-06,
"loss": 0.5725,
"step": 66
},
{
"epoch": 0.536,
"grad_norm": 1.0122913122177124,
"learning_rate": 9.998247908019594e-06,
"loss": 0.7628,
"step": 67
},
{
"epoch": 0.544,
"grad_norm": 0.9864280819892883,
"learning_rate": 9.997836953115927e-06,
"loss": 0.6184,
"step": 68
},
{
"epoch": 0.552,
"grad_norm": 1.4457077980041504,
"learning_rate": 9.997382752902658e-06,
"loss": 0.7471,
"step": 69
},
{
"epoch": 0.56,
"grad_norm": 0.9342664480209351,
"learning_rate": 9.996885311309892e-06,
"loss": 0.7304,
"step": 70
},
{
"epoch": 0.568,
"grad_norm": 1.1148232221603394,
"learning_rate": 9.996344632641895e-06,
"loss": 0.8129,
"step": 71
},
{
"epoch": 0.576,
"grad_norm": 1.0656917095184326,
"learning_rate": 9.995760721577053e-06,
"loss": 0.7355,
"step": 72
},
{
"epoch": 0.584,
"grad_norm": 1.0649811029434204,
"learning_rate": 9.995133583167833e-06,
"loss": 0.7288,
"step": 73
},
{
"epoch": 0.592,
"grad_norm": 1.1274362802505493,
"learning_rate": 9.994463222840748e-06,
"loss": 0.7243,
"step": 74
},
{
"epoch": 0.6,
"grad_norm": 0.8427581191062927,
"learning_rate": 9.993749646396286e-06,
"loss": 0.5633,
"step": 75
},
{
"epoch": 0.608,
"grad_norm": 0.98853600025177,
"learning_rate": 9.992992860008893e-06,
"loss": 0.7045,
"step": 76
},
{
"epoch": 0.616,
"grad_norm": 0.995302677154541,
"learning_rate": 9.99219287022689e-06,
"loss": 0.6268,
"step": 77
},
{
"epoch": 0.624,
"grad_norm": 0.9867312908172607,
"learning_rate": 9.991349683972435e-06,
"loss": 0.6828,
"step": 78
},
{
"epoch": 0.632,
"grad_norm": 0.9813309907913208,
"learning_rate": 9.990463308541452e-06,
"loss": 0.6601,
"step": 79
},
{
"epoch": 0.64,
"grad_norm": 0.9621861577033997,
"learning_rate": 9.989533751603578e-06,
"loss": 0.6984,
"step": 80
},
{
"epoch": 0.648,
"grad_norm": 0.988609790802002,
"learning_rate": 9.988561021202083e-06,
"loss": 0.6877,
"step": 81
},
{
"epoch": 0.656,
"grad_norm": 0.9378901720046997,
"learning_rate": 9.987545125753818e-06,
"loss": 0.6282,
"step": 82
},
{
"epoch": 0.664,
"grad_norm": 0.932298481464386,
"learning_rate": 9.986486074049131e-06,
"loss": 0.7014,
"step": 83
},
{
"epoch": 0.672,
"grad_norm": 0.9098712205886841,
"learning_rate": 9.985383875251783e-06,
"loss": 0.6461,
"step": 84
},
{
"epoch": 0.68,
"grad_norm": 0.9377619624137878,
"learning_rate": 9.98423853889889e-06,
"loss": 0.6283,
"step": 85
},
{
"epoch": 0.688,
"grad_norm": 1.2829806804656982,
"learning_rate": 9.983050074900824e-06,
"loss": 0.6713,
"step": 86
},
{
"epoch": 0.696,
"grad_norm": 0.981778085231781,
"learning_rate": 9.98181849354113e-06,
"loss": 0.643,
"step": 87
},
{
"epoch": 0.704,
"grad_norm": 1.2414823770523071,
"learning_rate": 9.980543805476447e-06,
"loss": 0.7538,
"step": 88
},
{
"epoch": 0.712,
"grad_norm": 1.1841117143630981,
"learning_rate": 9.979226021736396e-06,
"loss": 0.8119,
"step": 89
},
{
"epoch": 0.72,
"grad_norm": 1.5399696826934814,
"learning_rate": 9.977865153723508e-06,
"loss": 0.726,
"step": 90
},
{
"epoch": 0.728,
"grad_norm": 0.8497399687767029,
"learning_rate": 9.976461213213104e-06,
"loss": 0.6109,
"step": 91
},
{
"epoch": 0.736,
"grad_norm": 0.9964956045150757,
"learning_rate": 9.975014212353212e-06,
"loss": 0.6957,
"step": 92
},
{
"epoch": 0.744,
"grad_norm": 0.9907211065292358,
"learning_rate": 9.973524163664447e-06,
"loss": 0.7035,
"step": 93
},
{
"epoch": 0.752,
"grad_norm": 1.0158443450927734,
"learning_rate": 9.971991080039912e-06,
"loss": 0.764,
"step": 94
},
{
"epoch": 0.76,
"grad_norm": 1.0125267505645752,
"learning_rate": 9.970414974745077e-06,
"loss": 0.6104,
"step": 95
},
{
"epoch": 0.768,
"grad_norm": 1.0134661197662354,
"learning_rate": 9.968795861417676e-06,
"loss": 0.6558,
"step": 96
},
{
"epoch": 0.776,
"grad_norm": 1.1835541725158691,
"learning_rate": 9.967133754067581e-06,
"loss": 0.7853,
"step": 97
},
{
"epoch": 0.784,
"grad_norm": 0.9032031893730164,
"learning_rate": 9.965428667076687e-06,
"loss": 0.7235,
"step": 98
},
{
"epoch": 0.792,
"grad_norm": 1.2750600576400757,
"learning_rate": 9.963680615198774e-06,
"loss": 0.6768,
"step": 99
},
{
"epoch": 0.8,
"grad_norm": 1.0533162355422974,
"learning_rate": 9.961889613559396e-06,
"loss": 0.6036,
"step": 100
},
{
"epoch": 0.808,
"grad_norm": 0.9049885869026184,
"learning_rate": 9.960055677655743e-06,
"loss": 0.5954,
"step": 101
},
{
"epoch": 0.816,
"grad_norm": 1.2661690711975098,
"learning_rate": 9.958178823356503e-06,
"loss": 0.6608,
"step": 102
},
{
"epoch": 0.824,
"grad_norm": 0.957478940486908,
"learning_rate": 9.956259066901733e-06,
"loss": 0.6722,
"step": 103
},
{
"epoch": 0.832,
"grad_norm": 1.0017534494400024,
"learning_rate": 9.954296424902709e-06,
"loss": 0.699,
"step": 104
},
{
"epoch": 0.84,
"grad_norm": 0.9536829590797424,
"learning_rate": 9.95229091434179e-06,
"loss": 0.7978,
"step": 105
},
{
"epoch": 0.848,
"grad_norm": 0.9466567635536194,
"learning_rate": 9.950242552572272e-06,
"loss": 0.6377,
"step": 106
},
{
"epoch": 0.856,
"grad_norm": 0.8663733601570129,
"learning_rate": 9.948151357318228e-06,
"loss": 0.6943,
"step": 107
},
{
"epoch": 0.864,
"grad_norm": 0.9885082840919495,
"learning_rate": 9.946017346674362e-06,
"loss": 0.8259,
"step": 108
},
{
"epoch": 0.872,
"grad_norm": 0.9989089369773865,
"learning_rate": 9.943840539105853e-06,
"loss": 0.7219,
"step": 109
},
{
"epoch": 0.88,
"grad_norm": 0.9034218192100525,
"learning_rate": 9.941620953448195e-06,
"loss": 0.6568,
"step": 110
},
{
"epoch": 0.888,
"grad_norm": 0.9170519709587097,
"learning_rate": 9.939358608907026e-06,
"loss": 0.6996,
"step": 111
},
{
"epoch": 0.896,
"grad_norm": 1.0670710802078247,
"learning_rate": 9.937053525057977e-06,
"loss": 0.6088,
"step": 112
},
{
"epoch": 0.904,
"grad_norm": 1.0475540161132812,
"learning_rate": 9.934705721846487e-06,
"loss": 0.7299,
"step": 113
},
{
"epoch": 0.912,
"grad_norm": 0.8856591582298279,
"learning_rate": 9.932315219587641e-06,
"loss": 0.6026,
"step": 114
},
{
"epoch": 0.92,
"grad_norm": 0.9878740906715393,
"learning_rate": 9.92988203896599e-06,
"loss": 0.5906,
"step": 115
},
{
"epoch": 0.928,
"grad_norm": 0.9897602796554565,
"learning_rate": 9.927406201035368e-06,
"loss": 0.6622,
"step": 116
},
{
"epoch": 0.936,
"grad_norm": 1.0537142753601074,
"learning_rate": 9.924887727218724e-06,
"loss": 0.696,
"step": 117
},
{
"epoch": 0.944,
"grad_norm": 0.9410306811332703,
"learning_rate": 9.922326639307918e-06,
"loss": 0.6053,
"step": 118
},
{
"epoch": 0.952,
"grad_norm": 0.9250759482383728,
"learning_rate": 9.919722959463545e-06,
"loss": 0.702,
"step": 119
},
{
"epoch": 0.96,
"grad_norm": 0.9259200692176819,
"learning_rate": 9.917076710214739e-06,
"loss": 0.6032,
"step": 120
},
{
"epoch": 0.968,
"grad_norm": 1.0778374671936035,
"learning_rate": 9.914387914458983e-06,
"loss": 0.6911,
"step": 121
},
{
"epoch": 0.976,
"grad_norm": 0.8513089418411255,
"learning_rate": 9.911656595461899e-06,
"loss": 0.6422,
"step": 122
},
{
"epoch": 0.984,
"grad_norm": 1.1916236877441406,
"learning_rate": 9.908882776857057e-06,
"loss": 0.7315,
"step": 123
},
{
"epoch": 0.992,
"grad_norm": 0.9031611680984497,
"learning_rate": 9.906066482645774e-06,
"loss": 0.6034,
"step": 124
},
{
"epoch": 1.0,
"grad_norm": 0.8987178802490234,
"learning_rate": 9.903207737196892e-06,
"loss": 0.7292,
"step": 125
},
{
"epoch": 1.008,
"grad_norm": 1.241353988647461,
"learning_rate": 9.900306565246579e-06,
"loss": 0.7679,
"step": 126
},
{
"epoch": 1.016,
"grad_norm": 1.0967661142349243,
"learning_rate": 9.89736299189811e-06,
"loss": 0.6499,
"step": 127
},
{
"epoch": 1.024,
"grad_norm": 0.8532967567443848,
"learning_rate": 9.894377042621654e-06,
"loss": 0.6416,
"step": 128
},
{
"epoch": 1.032,
"grad_norm": 0.951801061630249,
"learning_rate": 9.891348743254046e-06,
"loss": 0.6654,
"step": 129
},
{
"epoch": 1.04,
"grad_norm": 1.057618498802185,
"learning_rate": 9.888278119998573e-06,
"loss": 0.623,
"step": 130
},
{
"epoch": 1.048,
"grad_norm": 0.8845769166946411,
"learning_rate": 9.885165199424738e-06,
"loss": 0.5969,
"step": 131
},
{
"epoch": 1.056,
"grad_norm": 0.8258888721466064,
"learning_rate": 9.882010008468038e-06,
"loss": 0.4811,
"step": 132
},
{
"epoch": 1.064,
"grad_norm": 0.9165515899658203,
"learning_rate": 9.878812574429722e-06,
"loss": 0.6168,
"step": 133
},
{
"epoch": 1.072,
"grad_norm": 0.8288542032241821,
"learning_rate": 9.875572924976568e-06,
"loss": 0.5462,
"step": 134
},
{
"epoch": 1.08,
"grad_norm": 1.180208444595337,
"learning_rate": 9.87229108814063e-06,
"loss": 0.5903,
"step": 135
},
{
"epoch": 1.088,
"grad_norm": 0.9072532653808594,
"learning_rate": 9.868967092319003e-06,
"loss": 0.5509,
"step": 136
},
{
"epoch": 1.096,
"grad_norm": 1.2131155729293823,
"learning_rate": 9.865600966273576e-06,
"loss": 0.5167,
"step": 137
},
{
"epoch": 1.104,
"grad_norm": 1.149257779121399,
"learning_rate": 9.86219273913078e-06,
"loss": 0.5793,
"step": 138
},
{
"epoch": 1.112,
"grad_norm": 0.8809230923652649,
"learning_rate": 9.858742440381343e-06,
"loss": 0.5457,
"step": 139
},
{
"epoch": 1.12,
"grad_norm": 0.8877855539321899,
"learning_rate": 9.855250099880026e-06,
"loss": 0.5863,
"step": 140
},
{
"epoch": 1.1280000000000001,
"grad_norm": 1.0398492813110352,
"learning_rate": 9.851715747845372e-06,
"loss": 0.6525,
"step": 141
},
{
"epoch": 1.1360000000000001,
"grad_norm": 0.9238712191581726,
"learning_rate": 9.848139414859441e-06,
"loss": 0.586,
"step": 142
},
{
"epoch": 1.144,
"grad_norm": 0.9197869896888733,
"learning_rate": 9.844521131867546e-06,
"loss": 0.4541,
"step": 143
},
{
"epoch": 1.152,
"grad_norm": 0.9942979216575623,
"learning_rate": 9.840860930177984e-06,
"loss": 0.655,
"step": 144
},
{
"epoch": 1.16,
"grad_norm": 1.1098957061767578,
"learning_rate": 9.837158841461767e-06,
"loss": 0.603,
"step": 145
},
{
"epoch": 1.168,
"grad_norm": 0.9269440770149231,
"learning_rate": 9.833414897752346e-06,
"loss": 0.5672,
"step": 146
},
{
"epoch": 1.176,
"grad_norm": 1.0234994888305664,
"learning_rate": 9.829629131445342e-06,
"loss": 0.5307,
"step": 147
},
{
"epoch": 1.184,
"grad_norm": 0.9524175524711609,
"learning_rate": 9.825801575298248e-06,
"loss": 0.5078,
"step": 148
},
{
"epoch": 1.192,
"grad_norm": 0.9059958457946777,
"learning_rate": 9.821932262430164e-06,
"loss": 0.5077,
"step": 149
},
{
"epoch": 1.2,
"grad_norm": 0.9674916863441467,
"learning_rate": 9.818021226321502e-06,
"loss": 0.554,
"step": 150
},
{
"epoch": 1.208,
"grad_norm": 0.9286289215087891,
"learning_rate": 9.814068500813692e-06,
"loss": 0.5633,
"step": 151
},
{
"epoch": 1.216,
"grad_norm": 0.9656635522842407,
"learning_rate": 9.8100741201089e-06,
"loss": 0.5439,
"step": 152
},
{
"epoch": 1.224,
"grad_norm": 0.9533044099807739,
"learning_rate": 9.806038118769724e-06,
"loss": 0.5286,
"step": 153
},
{
"epoch": 1.232,
"grad_norm": 0.9019532203674316,
"learning_rate": 9.801960531718898e-06,
"loss": 0.477,
"step": 154
},
{
"epoch": 1.24,
"grad_norm": 0.9925954341888428,
"learning_rate": 9.797841394238987e-06,
"loss": 0.6048,
"step": 155
},
{
"epoch": 1.248,
"grad_norm": 1.058506965637207,
"learning_rate": 9.793680741972084e-06,
"loss": 0.5777,
"step": 156
},
{
"epoch": 1.256,
"grad_norm": 0.9570199847221375,
"learning_rate": 9.789478610919508e-06,
"loss": 0.5514,
"step": 157
},
{
"epoch": 1.264,
"grad_norm": 0.9736652970314026,
"learning_rate": 9.785235037441473e-06,
"loss": 0.5932,
"step": 158
},
{
"epoch": 1.272,
"grad_norm": 0.9769409894943237,
"learning_rate": 9.780950058256802e-06,
"loss": 0.519,
"step": 159
},
{
"epoch": 1.28,
"grad_norm": 0.854789137840271,
"learning_rate": 9.77662371044258e-06,
"loss": 0.5516,
"step": 160
},
{
"epoch": 1.288,
"grad_norm": 1.084175944328308,
"learning_rate": 9.77225603143385e-06,
"loss": 0.7489,
"step": 161
},
{
"epoch": 1.296,
"grad_norm": 0.9743218421936035,
"learning_rate": 9.767847059023292e-06,
"loss": 0.6241,
"step": 162
},
{
"epoch": 1.304,
"grad_norm": 0.7581115365028381,
"learning_rate": 9.763396831360884e-06,
"loss": 0.586,
"step": 163
},
{
"epoch": 1.312,
"grad_norm": 0.9065977931022644,
"learning_rate": 9.75890538695358e-06,
"loss": 0.592,
"step": 164
},
{
"epoch": 1.32,
"grad_norm": 0.892532467842102,
"learning_rate": 9.75437276466497e-06,
"loss": 0.6491,
"step": 165
},
{
"epoch": 1.328,
"grad_norm": 1.1423097848892212,
"learning_rate": 9.749799003714954e-06,
"loss": 0.5284,
"step": 166
},
{
"epoch": 1.336,
"grad_norm": 0.9437663555145264,
"learning_rate": 9.745184143679398e-06,
"loss": 0.5462,
"step": 167
},
{
"epoch": 1.3439999999999999,
"grad_norm": 1.9607871770858765,
"learning_rate": 9.74052822448978e-06,
"loss": 0.6264,
"step": 168
},
{
"epoch": 1.3519999999999999,
"grad_norm": 0.952648937702179,
"learning_rate": 9.735831286432869e-06,
"loss": 0.6131,
"step": 169
},
{
"epoch": 1.3599999999999999,
"grad_norm": 0.9836501479148865,
"learning_rate": 9.731093370150349e-06,
"loss": 0.5708,
"step": 170
},
{
"epoch": 1.3679999999999999,
"grad_norm": 0.9113344550132751,
"learning_rate": 9.72631451663849e-06,
"loss": 0.441,
"step": 171
},
{
"epoch": 1.376,
"grad_norm": 0.7625031471252441,
"learning_rate": 9.721494767247779e-06,
"loss": 0.5014,
"step": 172
},
{
"epoch": 1.384,
"grad_norm": 0.841316819190979,
"learning_rate": 9.71663416368257e-06,
"loss": 0.5227,
"step": 173
},
{
"epoch": 1.392,
"grad_norm": 0.9969580173492432,
"learning_rate": 9.71173274800072e-06,
"loss": 0.5641,
"step": 174
},
{
"epoch": 1.4,
"grad_norm": 0.746874988079071,
"learning_rate": 9.70679056261322e-06,
"loss": 0.5006,
"step": 175
},
{
"epoch": 1.408,
"grad_norm": 1.0710760354995728,
"learning_rate": 9.70180765028384e-06,
"loss": 0.5583,
"step": 176
},
{
"epoch": 1.416,
"grad_norm": 0.9600836038589478,
"learning_rate": 9.696784054128749e-06,
"loss": 0.6547,
"step": 177
},
{
"epoch": 1.424,
"grad_norm": 0.9146630764007568,
"learning_rate": 9.691719817616148e-06,
"loss": 0.6432,
"step": 178
},
{
"epoch": 1.432,
"grad_norm": 1.3010084629058838,
"learning_rate": 9.686614984565888e-06,
"loss": 0.6419,
"step": 179
},
{
"epoch": 1.44,
"grad_norm": 0.8703446388244629,
"learning_rate": 9.681469599149093e-06,
"loss": 0.6048,
"step": 180
},
{
"epoch": 1.448,
"grad_norm": 0.9326883554458618,
"learning_rate": 9.676283705887783e-06,
"loss": 0.5797,
"step": 181
},
{
"epoch": 1.456,
"grad_norm": 0.8628938794136047,
"learning_rate": 9.671057349654481e-06,
"loss": 0.5471,
"step": 182
},
{
"epoch": 1.464,
"grad_norm": 0.8387150168418884,
"learning_rate": 9.66579057567183e-06,
"loss": 0.4629,
"step": 183
},
{
"epoch": 1.472,
"grad_norm": 0.9188330769538879,
"learning_rate": 9.660483429512198e-06,
"loss": 0.5265,
"step": 184
},
{
"epoch": 1.48,
"grad_norm": 0.8853181600570679,
"learning_rate": 9.65513595709729e-06,
"loss": 0.5754,
"step": 185
},
{
"epoch": 1.488,
"grad_norm": 0.8418251872062683,
"learning_rate": 9.649748204697741e-06,
"loss": 0.5653,
"step": 186
},
{
"epoch": 1.496,
"grad_norm": 1.0010942220687866,
"learning_rate": 9.644320218932723e-06,
"loss": 0.6375,
"step": 187
},
{
"epoch": 1.504,
"grad_norm": 0.9441695213317871,
"learning_rate": 9.63885204676954e-06,
"loss": 0.5651,
"step": 188
},
{
"epoch": 1.512,
"grad_norm": 0.774837076663971,
"learning_rate": 9.63334373552322e-06,
"loss": 0.4978,
"step": 189
},
{
"epoch": 1.52,
"grad_norm": 1.0337467193603516,
"learning_rate": 9.627795332856107e-06,
"loss": 0.7022,
"step": 190
},
{
"epoch": 1.528,
"grad_norm": 1.0459073781967163,
"learning_rate": 9.622206886777448e-06,
"loss": 0.6576,
"step": 191
},
{
"epoch": 1.536,
"grad_norm": 0.8906083106994629,
"learning_rate": 9.616578445642982e-06,
"loss": 0.5239,
"step": 192
},
{
"epoch": 1.544,
"grad_norm": 0.7993283271789551,
"learning_rate": 9.61091005815451e-06,
"loss": 0.5833,
"step": 193
},
{
"epoch": 1.552,
"grad_norm": 0.9584801197052002,
"learning_rate": 9.605201773359485e-06,
"loss": 0.5593,
"step": 194
},
{
"epoch": 1.56,
"grad_norm": 1.0783772468566895,
"learning_rate": 9.599453640650585e-06,
"loss": 0.5807,
"step": 195
},
{
"epoch": 1.568,
"grad_norm": 0.9267758727073669,
"learning_rate": 9.59366570976528e-06,
"loss": 0.5253,
"step": 196
},
{
"epoch": 1.576,
"grad_norm": 0.879656195640564,
"learning_rate": 9.587838030785413e-06,
"loss": 0.5637,
"step": 197
},
{
"epoch": 1.584,
"grad_norm": 0.7291843891143799,
"learning_rate": 9.581970654136752e-06,
"loss": 0.4746,
"step": 198
},
{
"epoch": 1.592,
"grad_norm": 1.1042965650558472,
"learning_rate": 9.576063630588563e-06,
"loss": 0.5335,
"step": 199
},
{
"epoch": 1.6,
"grad_norm": 0.89472496509552,
"learning_rate": 9.570117011253173e-06,
"loss": 0.5244,
"step": 200
},
{
"epoch": 1.608,
"grad_norm": 0.899229109287262,
"learning_rate": 9.56413084758552e-06,
"loss": 0.5962,
"step": 201
},
{
"epoch": 1.616,
"grad_norm": 0.9164184331893921,
"learning_rate": 9.55810519138271e-06,
"loss": 0.5423,
"step": 202
},
{
"epoch": 1.624,
"grad_norm": 0.949131429195404,
"learning_rate": 9.552040094783575e-06,
"loss": 0.6698,
"step": 203
},
{
"epoch": 1.6320000000000001,
"grad_norm": 1.300114393234253,
"learning_rate": 9.545935610268213e-06,
"loss": 0.7455,
"step": 204
},
{
"epoch": 1.6400000000000001,
"grad_norm": 0.9564195871353149,
"learning_rate": 9.53979179065754e-06,
"loss": 0.6041,
"step": 205
},
{
"epoch": 1.6480000000000001,
"grad_norm": 0.8021354675292969,
"learning_rate": 9.533608689112827e-06,
"loss": 0.5842,
"step": 206
},
{
"epoch": 1.6560000000000001,
"grad_norm": 0.8193942904472351,
"learning_rate": 9.527386359135254e-06,
"loss": 0.4656,
"step": 207
},
{
"epoch": 1.6640000000000001,
"grad_norm": 0.9571919441223145,
"learning_rate": 9.521124854565425e-06,
"loss": 0.6065,
"step": 208
},
{
"epoch": 1.6720000000000002,
"grad_norm": 0.6706253886222839,
"learning_rate": 9.514824229582922e-06,
"loss": 0.5117,
"step": 209
},
{
"epoch": 1.6800000000000002,
"grad_norm": 0.896847665309906,
"learning_rate": 9.508484538705823e-06,
"loss": 0.4671,
"step": 210
},
{
"epoch": 1.688,
"grad_norm": 1.0854939222335815,
"learning_rate": 9.50210583679024e-06,
"loss": 0.5669,
"step": 211
},
{
"epoch": 1.696,
"grad_norm": 0.83414626121521,
"learning_rate": 9.495688179029838e-06,
"loss": 0.5015,
"step": 212
},
{
"epoch": 1.704,
"grad_norm": 0.8416348695755005,
"learning_rate": 9.48923162095536e-06,
"loss": 0.5467,
"step": 213
},
{
"epoch": 1.712,
"grad_norm": 0.8415360450744629,
"learning_rate": 9.482736218434144e-06,
"loss": 0.5002,
"step": 214
},
{
"epoch": 1.72,
"grad_norm": 0.8762681484222412,
"learning_rate": 9.476202027669644e-06,
"loss": 0.5915,
"step": 215
},
{
"epoch": 1.728,
"grad_norm": 0.9399668574333191,
"learning_rate": 9.469629105200937e-06,
"loss": 0.6083,
"step": 216
},
{
"epoch": 1.736,
"grad_norm": 0.8347932696342468,
"learning_rate": 9.463017507902245e-06,
"loss": 0.5588,
"step": 217
},
{
"epoch": 1.744,
"grad_norm": 0.8872931003570557,
"learning_rate": 9.45636729298243e-06,
"loss": 0.5585,
"step": 218
},
{
"epoch": 1.752,
"grad_norm": 0.9400362968444824,
"learning_rate": 9.449678517984503e-06,
"loss": 0.4949,
"step": 219
},
{
"epoch": 1.76,
"grad_norm": 0.9362907409667969,
"learning_rate": 9.442951240785135e-06,
"loss": 0.4697,
"step": 220
},
{
"epoch": 1.768,
"grad_norm": 0.9293828010559082,
"learning_rate": 9.436185519594145e-06,
"loss": 0.5465,
"step": 221
},
{
"epoch": 1.776,
"grad_norm": 0.9569793939590454,
"learning_rate": 9.429381412954e-06,
"loss": 0.5107,
"step": 222
},
{
"epoch": 1.784,
"grad_norm": 0.8464429974555969,
"learning_rate": 9.422538979739307e-06,
"loss": 0.5828,
"step": 223
},
{
"epoch": 1.792,
"grad_norm": 0.9483685493469238,
"learning_rate": 9.415658279156312e-06,
"loss": 0.5515,
"step": 224
},
{
"epoch": 1.8,
"grad_norm": 0.8740932941436768,
"learning_rate": 9.408739370742372e-06,
"loss": 0.5731,
"step": 225
},
{
"epoch": 1.808,
"grad_norm": 0.8768803477287292,
"learning_rate": 9.401782314365458e-06,
"loss": 0.5504,
"step": 226
},
{
"epoch": 1.8159999999999998,
"grad_norm": 0.9519500732421875,
"learning_rate": 9.39478717022362e-06,
"loss": 0.5498,
"step": 227
},
{
"epoch": 1.8239999999999998,
"grad_norm": 1.1657662391662598,
"learning_rate": 9.387753998844482e-06,
"loss": 0.7251,
"step": 228
},
{
"epoch": 1.8319999999999999,
"grad_norm": 0.9255201816558838,
"learning_rate": 9.380682861084703e-06,
"loss": 0.5701,
"step": 229
},
{
"epoch": 1.8399999999999999,
"grad_norm": 0.8490859270095825,
"learning_rate": 9.37357381812946e-06,
"loss": 0.5303,
"step": 230
},
{
"epoch": 1.8479999999999999,
"grad_norm": 0.9158905148506165,
"learning_rate": 9.366426931491917e-06,
"loss": 0.6711,
"step": 231
},
{
"epoch": 1.8559999999999999,
"grad_norm": 0.8813873529434204,
"learning_rate": 9.359242263012693e-06,
"loss": 0.5088,
"step": 232
},
{
"epoch": 1.8639999999999999,
"grad_norm": 0.9613426327705383,
"learning_rate": 9.352019874859326e-06,
"loss": 0.5749,
"step": 233
},
{
"epoch": 1.8719999999999999,
"grad_norm": 0.7892547845840454,
"learning_rate": 9.344759829525734e-06,
"loss": 0.4288,
"step": 234
},
{
"epoch": 1.88,
"grad_norm": 0.9048332571983337,
"learning_rate": 9.33746218983167e-06,
"loss": 0.5524,
"step": 235
},
{
"epoch": 1.888,
"grad_norm": 1.0441476106643677,
"learning_rate": 9.330127018922195e-06,
"loss": 0.5697,
"step": 236
},
{
"epoch": 1.896,
"grad_norm": 0.8699061274528503,
"learning_rate": 9.32275438026711e-06,
"loss": 0.5447,
"step": 237
},
{
"epoch": 1.904,
"grad_norm": 0.8975421786308289,
"learning_rate": 9.315344337660422e-06,
"loss": 0.4788,
"step": 238
},
{
"epoch": 1.912,
"grad_norm": 0.8610514402389526,
"learning_rate": 9.307896955219787e-06,
"loss": 0.5311,
"step": 239
},
{
"epoch": 1.92,
"grad_norm": 0.9291175007820129,
"learning_rate": 9.300412297385954e-06,
"loss": 0.4566,
"step": 240
},
{
"epoch": 1.928,
"grad_norm": 0.8736976981163025,
"learning_rate": 9.29289042892221e-06,
"loss": 0.5223,
"step": 241
},
{
"epoch": 1.936,
"grad_norm": 0.9177389144897461,
"learning_rate": 9.285331414913816e-06,
"loss": 0.5391,
"step": 242
},
{
"epoch": 1.944,
"grad_norm": 0.8681254386901855,
"learning_rate": 9.277735320767449e-06,
"loss": 0.4491,
"step": 243
},
{
"epoch": 1.952,
"grad_norm": 0.9195399284362793,
"learning_rate": 9.270102212210632e-06,
"loss": 0.5366,
"step": 244
},
{
"epoch": 1.96,
"grad_norm": 1.0078879594802856,
"learning_rate": 9.262432155291167e-06,
"loss": 0.5684,
"step": 245
},
{
"epoch": 1.968,
"grad_norm": 0.9095841646194458,
"learning_rate": 9.254725216376562e-06,
"loss": 0.559,
"step": 246
},
{
"epoch": 1.976,
"grad_norm": 0.8708402514457703,
"learning_rate": 9.246981462153456e-06,
"loss": 0.5294,
"step": 247
},
{
"epoch": 1.984,
"grad_norm": 0.8832100033760071,
"learning_rate": 9.239200959627048e-06,
"loss": 0.652,
"step": 248
},
{
"epoch": 1.992,
"grad_norm": 0.8539401292800903,
"learning_rate": 9.231383776120512e-06,
"loss": 0.5354,
"step": 249
},
{
"epoch": 2.0,
"grad_norm": 0.8361132144927979,
"learning_rate": 9.223529979274411e-06,
"loss": 0.5124,
"step": 250
},
{
"epoch": 2.008,
"grad_norm": 1.3811569213867188,
"learning_rate": 9.215639637046121e-06,
"loss": 0.5122,
"step": 251
},
{
"epoch": 2.016,
"grad_norm": 1.0794579982757568,
"learning_rate": 9.207712817709237e-06,
"loss": 0.4453,
"step": 252
},
{
"epoch": 2.024,
"grad_norm": 0.8161504864692688,
"learning_rate": 9.19974958985298e-06,
"loss": 0.4507,
"step": 253
},
{
"epoch": 2.032,
"grad_norm": 0.934089183807373,
"learning_rate": 9.191750022381613e-06,
"loss": 0.3668,
"step": 254
},
{
"epoch": 2.04,
"grad_norm": 0.9390188455581665,
"learning_rate": 9.183714184513832e-06,
"loss": 0.4384,
"step": 255
},
{
"epoch": 2.048,
"grad_norm": 1.3397053480148315,
"learning_rate": 9.175642145782179e-06,
"loss": 0.4417,
"step": 256
},
{
"epoch": 2.056,
"grad_norm": 0.9079061150550842,
"learning_rate": 9.16753397603243e-06,
"loss": 0.4005,
"step": 257
},
{
"epoch": 2.064,
"grad_norm": 1.0313668251037598,
"learning_rate": 9.159389745423003e-06,
"loss": 0.4249,
"step": 258
},
{
"epoch": 2.072,
"grad_norm": 0.6768733859062195,
"learning_rate": 9.151209524424333e-06,
"loss": 0.3698,
"step": 259
},
{
"epoch": 2.08,
"grad_norm": 0.8785944581031799,
"learning_rate": 9.142993383818284e-06,
"loss": 0.35,
"step": 260
},
{
"epoch": 2.088,
"grad_norm": 1.2951953411102295,
"learning_rate": 9.134741394697517e-06,
"loss": 0.5264,
"step": 261
},
{
"epoch": 2.096,
"grad_norm": 0.9182500839233398,
"learning_rate": 9.126453628464889e-06,
"loss": 0.3614,
"step": 262
},
{
"epoch": 2.104,
"grad_norm": 1.2565897703170776,
"learning_rate": 9.118130156832823e-06,
"loss": 0.4656,
"step": 263
},
{
"epoch": 2.112,
"grad_norm": 0.857447624206543,
"learning_rate": 9.109771051822702e-06,
"loss": 0.388,
"step": 264
},
{
"epoch": 2.12,
"grad_norm": 0.8584538102149963,
"learning_rate": 9.10137638576423e-06,
"loss": 0.3833,
"step": 265
},
{
"epoch": 2.128,
"grad_norm": 1.0366853475570679,
"learning_rate": 9.09294623129482e-06,
"loss": 0.3688,
"step": 266
},
{
"epoch": 2.136,
"grad_norm": 0.8696836233139038,
"learning_rate": 9.084480661358954e-06,
"loss": 0.4356,
"step": 267
},
{
"epoch": 2.144,
"grad_norm": 0.8424058556556702,
"learning_rate": 9.07597974920756e-06,
"loss": 0.4523,
"step": 268
},
{
"epoch": 2.152,
"grad_norm": 0.8936377763748169,
"learning_rate": 9.067443568397378e-06,
"loss": 0.4195,
"step": 269
},
{
"epoch": 2.16,
"grad_norm": 0.7706946730613708,
"learning_rate": 9.058872192790314e-06,
"loss": 0.4114,
"step": 270
},
{
"epoch": 2.168,
"grad_norm": 0.8530108332633972,
"learning_rate": 9.05026569655281e-06,
"loss": 0.4485,
"step": 271
},
{
"epoch": 2.176,
"grad_norm": 0.9020074009895325,
"learning_rate": 9.041624154155208e-06,
"loss": 0.4264,
"step": 272
},
{
"epoch": 2.184,
"grad_norm": 0.6583293676376343,
"learning_rate": 9.032947640371086e-06,
"loss": 0.3326,
"step": 273
},
{
"epoch": 2.192,
"grad_norm": 0.8414630889892578,
"learning_rate": 9.02423623027663e-06,
"loss": 0.3968,
"step": 274
},
{
"epoch": 2.2,
"grad_norm": 0.9140536785125732,
"learning_rate": 9.01548999924997e-06,
"loss": 0.3917,
"step": 275
},
{
"epoch": 2.208,
"grad_norm": 0.8281865119934082,
"learning_rate": 9.006709022970547e-06,
"loss": 0.4069,
"step": 276
},
{
"epoch": 2.216,
"grad_norm": 0.7317336201667786,
"learning_rate": 8.997893377418432e-06,
"loss": 0.4185,
"step": 277
},
{
"epoch": 2.224,
"grad_norm": 1.2639456987380981,
"learning_rate": 8.98904313887369e-06,
"loss": 0.3394,
"step": 278
},
{
"epoch": 2.232,
"grad_norm": 0.8909697532653809,
"learning_rate": 8.980158383915714e-06,
"loss": 0.4178,
"step": 279
},
{
"epoch": 2.24,
"grad_norm": 0.8912489414215088,
"learning_rate": 8.971239189422555e-06,
"loss": 0.4751,
"step": 280
},
{
"epoch": 2.248,
"grad_norm": 0.8743604421615601,
"learning_rate": 8.962285632570266e-06,
"loss": 0.438,
"step": 281
},
{
"epoch": 2.2560000000000002,
"grad_norm": 0.9688166379928589,
"learning_rate": 8.953297790832231e-06,
"loss": 0.4437,
"step": 282
},
{
"epoch": 2.2640000000000002,
"grad_norm": 0.9198762774467468,
"learning_rate": 8.944275741978495e-06,
"loss": 0.4303,
"step": 283
},
{
"epoch": 2.2720000000000002,
"grad_norm": 0.8178912997245789,
"learning_rate": 8.935219564075087e-06,
"loss": 0.4714,
"step": 284
},
{
"epoch": 2.2800000000000002,
"grad_norm": 1.30088472366333,
"learning_rate": 8.92612933548335e-06,
"loss": 0.3948,
"step": 285
},
{
"epoch": 2.288,
"grad_norm": 1.0883452892303467,
"learning_rate": 8.917005134859263e-06,
"loss": 0.3697,
"step": 286
},
{
"epoch": 2.296,
"grad_norm": 1.3085649013519287,
"learning_rate": 8.907847041152757e-06,
"loss": 0.4761,
"step": 287
},
{
"epoch": 2.304,
"grad_norm": 0.8386606574058533,
"learning_rate": 8.89865513360703e-06,
"loss": 0.3663,
"step": 288
},
{
"epoch": 2.312,
"grad_norm": 0.8883110284805298,
"learning_rate": 8.889429491757872e-06,
"loss": 0.4755,
"step": 289
},
{
"epoch": 2.32,
"grad_norm": 0.8336955308914185,
"learning_rate": 8.88017019543296e-06,
"loss": 0.3742,
"step": 290
},
{
"epoch": 2.328,
"grad_norm": 0.7842932939529419,
"learning_rate": 8.870877324751186e-06,
"loss": 0.3338,
"step": 291
},
{
"epoch": 2.336,
"grad_norm": 0.8498310446739197,
"learning_rate": 8.861550960121946e-06,
"loss": 0.4814,
"step": 292
},
{
"epoch": 2.344,
"grad_norm": 0.785335123538971,
"learning_rate": 8.852191182244456e-06,
"loss": 0.3093,
"step": 293
},
{
"epoch": 2.352,
"grad_norm": 0.758841872215271,
"learning_rate": 8.842798072107055e-06,
"loss": 0.4555,
"step": 294
},
{
"epoch": 2.36,
"grad_norm": 0.8311530947685242,
"learning_rate": 8.833371710986493e-06,
"loss": 0.3642,
"step": 295
},
{
"epoch": 2.368,
"grad_norm": 0.8695397973060608,
"learning_rate": 8.823912180447237e-06,
"loss": 0.3611,
"step": 296
},
{
"epoch": 2.376,
"grad_norm": 0.9586389660835266,
"learning_rate": 8.81441956234076e-06,
"loss": 0.4375,
"step": 297
},
{
"epoch": 2.384,
"grad_norm": 1.3389581441879272,
"learning_rate": 8.804893938804839e-06,
"loss": 0.4707,
"step": 298
},
{
"epoch": 2.392,
"grad_norm": 0.8372800946235657,
"learning_rate": 8.795335392262841e-06,
"loss": 0.4676,
"step": 299
},
{
"epoch": 2.4,
"grad_norm": 0.9432985186576843,
"learning_rate": 8.785744005423003e-06,
"loss": 0.5151,
"step": 300
},
{
"epoch": 2.408,
"grad_norm": 0.7734491229057312,
"learning_rate": 8.77611986127773e-06,
"loss": 0.4515,
"step": 301
},
{
"epoch": 2.416,
"grad_norm": 0.7202035784721375,
"learning_rate": 8.766463043102864e-06,
"loss": 0.3825,
"step": 302
},
{
"epoch": 2.424,
"grad_norm": 0.9535182118415833,
"learning_rate": 8.756773634456975e-06,
"loss": 0.5219,
"step": 303
},
{
"epoch": 2.432,
"grad_norm": 0.7026752233505249,
"learning_rate": 8.747051719180626e-06,
"loss": 0.36,
"step": 304
},
{
"epoch": 2.44,
"grad_norm": 0.7914669513702393,
"learning_rate": 8.737297381395657e-06,
"loss": 0.4205,
"step": 305
},
{
"epoch": 2.448,
"grad_norm": 0.7313674688339233,
"learning_rate": 8.727510705504453e-06,
"loss": 0.4484,
"step": 306
},
{
"epoch": 2.456,
"grad_norm": 0.8241586089134216,
"learning_rate": 8.717691776189214e-06,
"loss": 0.3453,
"step": 307
},
{
"epoch": 2.464,
"grad_norm": 0.9311320185661316,
"learning_rate": 8.707840678411223e-06,
"loss": 0.3825,
"step": 308
},
{
"epoch": 2.472,
"grad_norm": 0.8270311951637268,
"learning_rate": 8.69795749741011e-06,
"loss": 0.4327,
"step": 309
},
{
"epoch": 2.48,
"grad_norm": 0.8111954927444458,
"learning_rate": 8.688042318703111e-06,
"loss": 0.3797,
"step": 310
},
{
"epoch": 2.488,
"grad_norm": 0.8930568695068359,
"learning_rate": 8.678095228084343e-06,
"loss": 0.3948,
"step": 311
},
{
"epoch": 2.496,
"grad_norm": 1.5294451713562012,
"learning_rate": 8.66811631162404e-06,
"loss": 0.4458,
"step": 312
},
{
"epoch": 2.504,
"grad_norm": 1.0527986288070679,
"learning_rate": 8.65810565566782e-06,
"loss": 0.37,
"step": 313
},
{
"epoch": 2.512,
"grad_norm": 0.79234379529953,
"learning_rate": 8.648063346835943e-06,
"loss": 0.4103,
"step": 314
},
{
"epoch": 2.52,
"grad_norm": 0.8093475103378296,
"learning_rate": 8.637989472022548e-06,
"loss": 0.3947,
"step": 315
},
{
"epoch": 2.528,
"grad_norm": 0.9096853733062744,
"learning_rate": 8.627884118394913e-06,
"loss": 0.3167,
"step": 316
},
{
"epoch": 2.536,
"grad_norm": 0.7235214114189148,
"learning_rate": 8.617747373392697e-06,
"loss": 0.405,
"step": 317
},
{
"epoch": 2.544,
"grad_norm": 0.8731074929237366,
"learning_rate": 8.607579324727175e-06,
"loss": 0.4647,
"step": 318
},
{
"epoch": 2.552,
"grad_norm": 0.8666344881057739,
"learning_rate": 8.597380060380493e-06,
"loss": 0.3891,
"step": 319
},
{
"epoch": 2.56,
"grad_norm": 1.1291053295135498,
"learning_rate": 8.5871496686049e-06,
"loss": 0.461,
"step": 320
},
{
"epoch": 2.568,
"grad_norm": 0.9848377108573914,
"learning_rate": 8.576888237921983e-06,
"loss": 0.4388,
"step": 321
},
{
"epoch": 2.576,
"grad_norm": 0.8755123019218445,
"learning_rate": 8.566595857121902e-06,
"loss": 0.4386,
"step": 322
},
{
"epoch": 2.584,
"grad_norm": 1.2679446935653687,
"learning_rate": 8.556272615262623e-06,
"loss": 0.3827,
"step": 323
},
{
"epoch": 2.592,
"grad_norm": 1.340916395187378,
"learning_rate": 8.545918601669147e-06,
"loss": 0.4543,
"step": 324
},
{
"epoch": 2.6,
"grad_norm": 0.7855986952781677,
"learning_rate": 8.535533905932739e-06,
"loss": 0.3923,
"step": 325
},
{
"epoch": 2.608,
"grad_norm": 0.829623818397522,
"learning_rate": 8.525118617910144e-06,
"loss": 0.4529,
"step": 326
},
{
"epoch": 2.616,
"grad_norm": 0.932386040687561,
"learning_rate": 8.514672827722824e-06,
"loss": 0.4716,
"step": 327
},
{
"epoch": 2.624,
"grad_norm": 0.864211916923523,
"learning_rate": 8.504196625756166e-06,
"loss": 0.3825,
"step": 328
},
{
"epoch": 2.632,
"grad_norm": 0.7904481291770935,
"learning_rate": 8.493690102658703e-06,
"loss": 0.3582,
"step": 329
},
{
"epoch": 2.64,
"grad_norm": 0.6928600072860718,
"learning_rate": 8.483153349341336e-06,
"loss": 0.3512,
"step": 330
},
{
"epoch": 2.648,
"grad_norm": 0.7888721227645874,
"learning_rate": 8.472586456976534e-06,
"loss": 0.4814,
"step": 331
},
{
"epoch": 2.656,
"grad_norm": 0.8113727569580078,
"learning_rate": 8.461989516997565e-06,
"loss": 0.4022,
"step": 332
},
{
"epoch": 2.664,
"grad_norm": 0.9936776757240295,
"learning_rate": 8.45136262109768e-06,
"loss": 0.6013,
"step": 333
},
{
"epoch": 2.672,
"grad_norm": 0.6121838688850403,
"learning_rate": 8.440705861229344e-06,
"loss": 0.3144,
"step": 334
},
{
"epoch": 2.68,
"grad_norm": 0.9799290299415588,
"learning_rate": 8.430019329603423e-06,
"loss": 0.4724,
"step": 335
},
{
"epoch": 2.6879999999999997,
"grad_norm": 0.8076691627502441,
"learning_rate": 8.41930311868839e-06,
"loss": 0.404,
"step": 336
},
{
"epoch": 2.6959999999999997,
"grad_norm": 0.9041216373443604,
"learning_rate": 8.408557321209534e-06,
"loss": 0.4434,
"step": 337
},
{
"epoch": 2.7039999999999997,
"grad_norm": 0.8356102108955383,
"learning_rate": 8.397782030148147e-06,
"loss": 0.4153,
"step": 338
},
{
"epoch": 2.7119999999999997,
"grad_norm": 1.0063518285751343,
"learning_rate": 8.386977338740724e-06,
"loss": 0.421,
"step": 339
},
{
"epoch": 2.7199999999999998,
"grad_norm": 0.741091787815094,
"learning_rate": 8.376143340478153e-06,
"loss": 0.3746,
"step": 340
},
{
"epoch": 2.7279999999999998,
"grad_norm": 0.992077112197876,
"learning_rate": 8.365280129104912e-06,
"loss": 0.4036,
"step": 341
},
{
"epoch": 2.7359999999999998,
"grad_norm": 0.8580658435821533,
"learning_rate": 8.354387798618254e-06,
"loss": 0.3937,
"step": 342
},
{
"epoch": 2.7439999999999998,
"grad_norm": 0.9947491884231567,
"learning_rate": 8.34346644326739e-06,
"loss": 0.3879,
"step": 343
},
{
"epoch": 2.752,
"grad_norm": 0.8819236159324646,
"learning_rate": 8.332516157552684e-06,
"loss": 0.4296,
"step": 344
},
{
"epoch": 2.76,
"grad_norm": 0.7174811363220215,
"learning_rate": 8.321537036224822e-06,
"loss": 0.3896,
"step": 345
},
{
"epoch": 2.768,
"grad_norm": 0.7424646019935608,
"learning_rate": 8.310529174284004e-06,
"loss": 0.3654,
"step": 346
},
{
"epoch": 2.776,
"grad_norm": 0.9448553323745728,
"learning_rate": 8.299492666979114e-06,
"loss": 0.3425,
"step": 347
},
{
"epoch": 2.784,
"grad_norm": 0.7875136137008667,
"learning_rate": 8.288427609806899e-06,
"loss": 0.3703,
"step": 348
},
{
"epoch": 2.792,
"grad_norm": 0.9348728060722351,
"learning_rate": 8.277334098511147e-06,
"loss": 0.4224,
"step": 349
},
{
"epoch": 2.8,
"grad_norm": 0.7805924415588379,
"learning_rate": 8.266212229081846e-06,
"loss": 0.3679,
"step": 350
},
{
"epoch": 2.808,
"grad_norm": 0.8468799591064453,
"learning_rate": 8.255062097754371e-06,
"loss": 0.4477,
"step": 351
},
{
"epoch": 2.816,
"grad_norm": 0.7618088126182556,
"learning_rate": 8.243883801008632e-06,
"loss": 0.4523,
"step": 352
},
{
"epoch": 2.824,
"grad_norm": 0.7418726086616516,
"learning_rate": 8.232677435568252e-06,
"loss": 0.4286,
"step": 353
},
{
"epoch": 2.832,
"grad_norm": 0.7962421774864197,
"learning_rate": 8.221443098399733e-06,
"loss": 0.3583,
"step": 354
},
{
"epoch": 2.84,
"grad_norm": 0.9815137982368469,
"learning_rate": 8.210180886711603e-06,
"loss": 0.4421,
"step": 355
},
{
"epoch": 2.848,
"grad_norm": 0.8846978545188904,
"learning_rate": 8.198890897953586e-06,
"loss": 0.4625,
"step": 356
},
{
"epoch": 2.856,
"grad_norm": 0.7743475437164307,
"learning_rate": 8.187573229815757e-06,
"loss": 0.3382,
"step": 357
},
{
"epoch": 2.864,
"grad_norm": 0.9756518602371216,
"learning_rate": 8.176227980227693e-06,
"loss": 0.4967,
"step": 358
},
{
"epoch": 2.872,
"grad_norm": 0.919818103313446,
"learning_rate": 8.164855247357628e-06,
"loss": 0.4667,
"step": 359
},
{
"epoch": 2.88,
"grad_norm": 0.7409064173698425,
"learning_rate": 8.153455129611605e-06,
"loss": 0.3253,
"step": 360
},
{
"epoch": 2.888,
"grad_norm": 0.8657407760620117,
"learning_rate": 8.142027725632622e-06,
"loss": 0.4967,
"step": 361
},
{
"epoch": 2.896,
"grad_norm": 0.8034350275993347,
"learning_rate": 8.130573134299782e-06,
"loss": 0.4267,
"step": 362
},
{
"epoch": 2.904,
"grad_norm": 0.8715494871139526,
"learning_rate": 8.119091454727427e-06,
"loss": 0.4406,
"step": 363
},
{
"epoch": 2.912,
"grad_norm": 0.6950966715812683,
"learning_rate": 8.107582786264299e-06,
"loss": 0.3652,
"step": 364
},
{
"epoch": 2.92,
"grad_norm": 0.8034759163856506,
"learning_rate": 8.09604722849266e-06,
"loss": 0.4174,
"step": 365
},
{
"epoch": 2.928,
"grad_norm": 0.9261465072631836,
"learning_rate": 8.084484881227449e-06,
"loss": 0.348,
"step": 366
},
{
"epoch": 2.936,
"grad_norm": 0.804601788520813,
"learning_rate": 8.072895844515398e-06,
"loss": 0.4119,
"step": 367
},
{
"epoch": 2.944,
"grad_norm": 0.9597411751747131,
"learning_rate": 8.061280218634192e-06,
"loss": 0.4271,
"step": 368
},
{
"epoch": 2.952,
"grad_norm": 0.9602273106575012,
"learning_rate": 8.049638104091575e-06,
"loss": 0.4448,
"step": 369
},
{
"epoch": 2.96,
"grad_norm": 0.8276174664497375,
"learning_rate": 8.037969601624495e-06,
"loss": 0.3886,
"step": 370
},
{
"epoch": 2.968,
"grad_norm": 0.9530600309371948,
"learning_rate": 8.026274812198235e-06,
"loss": 0.4353,
"step": 371
},
{
"epoch": 2.976,
"grad_norm": 0.8712935447692871,
"learning_rate": 8.014553837005527e-06,
"loss": 0.4336,
"step": 372
},
{
"epoch": 2.984,
"grad_norm": 0.7948423624038696,
"learning_rate": 8.002806777465685e-06,
"loss": 0.4402,
"step": 373
},
{
"epoch": 2.992,
"grad_norm": 0.8660122156143188,
"learning_rate": 7.99103373522373e-06,
"loss": 0.3639,
"step": 374
},
{
"epoch": 3.0,
"grad_norm": 0.9771859049797058,
"learning_rate": 7.9792348121495e-06,
"loss": 0.4096,
"step": 375
},
{
"epoch": 3.008,
"grad_norm": 0.9802374243736267,
"learning_rate": 7.967410110336782e-06,
"loss": 0.336,
"step": 376
},
{
"epoch": 3.016,
"grad_norm": 1.075129747390747,
"learning_rate": 7.955559732102414e-06,
"loss": 0.2861,
"step": 377
},
{
"epoch": 3.024,
"grad_norm": 0.8979294896125793,
"learning_rate": 7.943683779985412e-06,
"loss": 0.3531,
"step": 378
},
{
"epoch": 3.032,
"grad_norm": 0.9654936790466309,
"learning_rate": 7.931782356746076e-06,
"loss": 0.2724,
"step": 379
},
{
"epoch": 3.04,
"grad_norm": 0.9817936420440674,
"learning_rate": 7.919855565365102e-06,
"loss": 0.2525,
"step": 380
},
{
"epoch": 3.048,
"grad_norm": 1.1742961406707764,
"learning_rate": 7.907903509042696e-06,
"loss": 0.2952,
"step": 381
},
{
"epoch": 3.056,
"grad_norm": 0.9033184051513672,
"learning_rate": 7.895926291197667e-06,
"loss": 0.2767,
"step": 382
},
{
"epoch": 3.064,
"grad_norm": 0.7659105062484741,
"learning_rate": 7.883924015466554e-06,
"loss": 0.2231,
"step": 383
},
{
"epoch": 3.072,
"grad_norm": 0.8166160583496094,
"learning_rate": 7.871896785702707e-06,
"loss": 0.2325,
"step": 384
},
{
"epoch": 3.08,
"grad_norm": 0.7707303762435913,
"learning_rate": 7.859844705975405e-06,
"loss": 0.1768,
"step": 385
},
{
"epoch": 3.088,
"grad_norm": 1.0003777742385864,
"learning_rate": 7.847767880568944e-06,
"loss": 0.3651,
"step": 386
},
{
"epoch": 3.096,
"grad_norm": 0.845194399356842,
"learning_rate": 7.835666413981744e-06,
"loss": 0.2742,
"step": 387
},
{
"epoch": 3.104,
"grad_norm": 0.7469092607498169,
"learning_rate": 7.823540410925434e-06,
"loss": 0.2141,
"step": 388
},
{
"epoch": 3.112,
"grad_norm": 0.8582166433334351,
"learning_rate": 7.811389976323963e-06,
"loss": 0.2813,
"step": 389
},
{
"epoch": 3.12,
"grad_norm": 0.8023694753646851,
"learning_rate": 7.799215215312667e-06,
"loss": 0.2098,
"step": 390
},
{
"epoch": 3.128,
"grad_norm": 0.8036847114562988,
"learning_rate": 7.787016233237387e-06,
"loss": 0.2091,
"step": 391
},
{
"epoch": 3.136,
"grad_norm": 0.9070097804069519,
"learning_rate": 7.774793135653537e-06,
"loss": 0.2785,
"step": 392
},
{
"epoch": 3.144,
"grad_norm": 0.8162369132041931,
"learning_rate": 7.7625460283252e-06,
"loss": 0.2755,
"step": 393
},
{
"epoch": 3.152,
"grad_norm": 0.8396410942077637,
"learning_rate": 7.750275017224208e-06,
"loss": 0.3056,
"step": 394
},
{
"epoch": 3.16,
"grad_norm": 0.7635102272033691,
"learning_rate": 7.737980208529232e-06,
"loss": 0.2045,
"step": 395
},
{
"epoch": 3.168,
"grad_norm": 0.7944246530532837,
"learning_rate": 7.725661708624855e-06,
"loss": 0.2589,
"step": 396
},
{
"epoch": 3.176,
"grad_norm": 0.8792198300361633,
"learning_rate": 7.713319624100657e-06,
"loss": 0.2622,
"step": 397
},
{
"epoch": 3.184,
"grad_norm": 0.8552615642547607,
"learning_rate": 7.700954061750295e-06,
"loss": 0.2594,
"step": 398
},
{
"epoch": 3.192,
"grad_norm": 0.7373338341712952,
"learning_rate": 7.688565128570564e-06,
"loss": 0.2032,
"step": 399
},
{
"epoch": 3.2,
"grad_norm": 1.1373937129974365,
"learning_rate": 7.676152931760496e-06,
"loss": 0.3086,
"step": 400
},
{
"epoch": 3.208,
"grad_norm": 0.9050979614257812,
"learning_rate": 7.663717578720412e-06,
"loss": 0.2816,
"step": 401
},
{
"epoch": 3.216,
"grad_norm": 0.8409582376480103,
"learning_rate": 7.651259177050996e-06,
"loss": 0.2636,
"step": 402
},
{
"epoch": 3.224,
"grad_norm": 0.8253751397132874,
"learning_rate": 7.638777834552372e-06,
"loss": 0.2511,
"step": 403
},
{
"epoch": 3.232,
"grad_norm": 0.9281233549118042,
"learning_rate": 7.626273659223166e-06,
"loss": 0.2718,
"step": 404
},
{
"epoch": 3.24,
"grad_norm": 0.745572566986084,
"learning_rate": 7.61374675925957e-06,
"loss": 0.2602,
"step": 405
},
{
"epoch": 3.248,
"grad_norm": 0.7207661867141724,
"learning_rate": 7.601197243054411e-06,
"loss": 0.2428,
"step": 406
},
{
"epoch": 3.2560000000000002,
"grad_norm": 0.804275631904602,
"learning_rate": 7.588625219196208e-06,
"loss": 0.239,
"step": 407
},
{
"epoch": 3.2640000000000002,
"grad_norm": 0.730238676071167,
"learning_rate": 7.576030796468233e-06,
"loss": 0.2988,
"step": 408
},
{
"epoch": 3.2720000000000002,
"grad_norm": 0.7774324417114258,
"learning_rate": 7.563414083847573e-06,
"loss": 0.2133,
"step": 409
},
{
"epoch": 3.2800000000000002,
"grad_norm": 0.9310129880905151,
"learning_rate": 7.5507751905041885e-06,
"loss": 0.3192,
"step": 410
},
{
"epoch": 3.288,
"grad_norm": 0.9868288636207581,
"learning_rate": 7.538114225799955e-06,
"loss": 0.3086,
"step": 411
},
{
"epoch": 3.296,
"grad_norm": 0.7649028301239014,
"learning_rate": 7.525431299287737e-06,
"loss": 0.2505,
"step": 412
},
{
"epoch": 3.304,
"grad_norm": 0.7084949016571045,
"learning_rate": 7.512726520710429e-06,
"loss": 0.2534,
"step": 413
},
{
"epoch": 3.312,
"grad_norm": 0.8004085421562195,
"learning_rate": 7.500000000000001e-06,
"loss": 0.2705,
"step": 414
},
{
"epoch": 3.32,
"grad_norm": 0.8937138319015503,
"learning_rate": 7.4872518472765594e-06,
"loss": 0.1749,
"step": 415
},
{
"epoch": 3.328,
"grad_norm": 0.792624831199646,
"learning_rate": 7.474482172847391e-06,
"loss": 0.2763,
"step": 416
},
{
"epoch": 3.336,
"grad_norm": 0.8165427446365356,
"learning_rate": 7.461691087205993e-06,
"loss": 0.3184,
"step": 417
},
{
"epoch": 3.344,
"grad_norm": 0.7113141417503357,
"learning_rate": 7.4488787010311425e-06,
"loss": 0.2336,
"step": 418
},
{
"epoch": 3.352,
"grad_norm": 0.9322896599769592,
"learning_rate": 7.436045125185923e-06,
"loss": 0.3307,
"step": 419
},
{
"epoch": 3.36,
"grad_norm": 0.7401679754257202,
"learning_rate": 7.423190470716761e-06,
"loss": 0.2605,
"step": 420
},
{
"epoch": 3.368,
"grad_norm": 0.8005726337432861,
"learning_rate": 7.4103148488524824e-06,
"loss": 0.2684,
"step": 421
},
{
"epoch": 3.376,
"grad_norm": 0.7309401631355286,
"learning_rate": 7.3974183710033334e-06,
"loss": 0.2729,
"step": 422
},
{
"epoch": 3.384,
"grad_norm": 0.9116814136505127,
"learning_rate": 7.384501148760024e-06,
"loss": 0.3469,
"step": 423
},
{
"epoch": 3.392,
"grad_norm": 0.8202308416366577,
"learning_rate": 7.371563293892761e-06,
"loss": 0.3268,
"step": 424
},
{
"epoch": 3.4,
"grad_norm": 0.8762120604515076,
"learning_rate": 7.3586049183502875e-06,
"loss": 0.2884,
"step": 425
},
{
"epoch": 3.408,
"grad_norm": 0.7986353635787964,
"learning_rate": 7.345626134258897e-06,
"loss": 0.2388,
"step": 426
},
{
"epoch": 3.416,
"grad_norm": 0.880988597869873,
"learning_rate": 7.3326270539214826e-06,
"loss": 0.2725,
"step": 427
},
{
"epoch": 3.424,
"grad_norm": 0.8958916664123535,
"learning_rate": 7.319607789816555e-06,
"loss": 0.251,
"step": 428
},
{
"epoch": 3.432,
"grad_norm": 0.753807008266449,
"learning_rate": 7.306568454597269e-06,
"loss": 0.2943,
"step": 429
},
{
"epoch": 3.44,
"grad_norm": 0.8259589672088623,
"learning_rate": 7.293509161090453e-06,
"loss": 0.255,
"step": 430
},
{
"epoch": 3.448,
"grad_norm": 0.725982666015625,
"learning_rate": 7.28043002229563e-06,
"loss": 0.244,
"step": 431
},
{
"epoch": 3.456,
"grad_norm": 0.8970488905906677,
"learning_rate": 7.2673311513840395e-06,
"loss": 0.3301,
"step": 432
},
{
"epoch": 3.464,
"grad_norm": 0.7942942380905151,
"learning_rate": 7.2542126616976596e-06,
"loss": 0.2654,
"step": 433
},
{
"epoch": 3.472,
"grad_norm": 0.9445860981941223,
"learning_rate": 7.241074666748228e-06,
"loss": 0.2441,
"step": 434
},
{
"epoch": 3.48,
"grad_norm": 0.9545169472694397,
"learning_rate": 7.227917280216254e-06,
"loss": 0.2609,
"step": 435
},
{
"epoch": 3.488,
"grad_norm": 0.8117399215698242,
"learning_rate": 7.214740615950041e-06,
"loss": 0.3145,
"step": 436
},
{
"epoch": 3.496,
"grad_norm": 0.8447906970977783,
"learning_rate": 7.201544787964698e-06,
"loss": 0.2937,
"step": 437
},
{
"epoch": 3.504,
"grad_norm": 0.8552675247192383,
"learning_rate": 7.188329910441154e-06,
"loss": 0.2547,
"step": 438
},
{
"epoch": 3.512,
"grad_norm": 0.7751868367195129,
"learning_rate": 7.175096097725169e-06,
"loss": 0.3318,
"step": 439
},
{
"epoch": 3.52,
"grad_norm": 2.058791160583496,
"learning_rate": 7.161843464326349e-06,
"loss": 0.2861,
"step": 440
},
{
"epoch": 3.528,
"grad_norm": 0.8021336197853088,
"learning_rate": 7.148572124917148e-06,
"loss": 0.3147,
"step": 441
},
{
"epoch": 3.536,
"grad_norm": 0.9582990407943726,
"learning_rate": 7.135282194331881e-06,
"loss": 0.3402,
"step": 442
},
{
"epoch": 3.544,
"grad_norm": 0.7796058058738708,
"learning_rate": 7.121973787565727e-06,
"loss": 0.2842,
"step": 443
},
{
"epoch": 3.552,
"grad_norm": 0.6856812834739685,
"learning_rate": 7.1086470197737405e-06,
"loss": 0.2571,
"step": 444
},
{
"epoch": 3.56,
"grad_norm": 0.7525624632835388,
"learning_rate": 7.095302006269842e-06,
"loss": 0.2708,
"step": 445
},
{
"epoch": 3.568,
"grad_norm": 0.6989469528198242,
"learning_rate": 7.0819388625258385e-06,
"loss": 0.2104,
"step": 446
},
{
"epoch": 3.576,
"grad_norm": 0.861801028251648,
"learning_rate": 7.06855770417041e-06,
"loss": 0.2718,
"step": 447
},
{
"epoch": 3.584,
"grad_norm": 0.719370424747467,
"learning_rate": 7.05515864698811e-06,
"loss": 0.2787,
"step": 448
},
{
"epoch": 3.592,
"grad_norm": 0.6387084126472473,
"learning_rate": 7.041741806918372e-06,
"loss": 0.2023,
"step": 449
},
{
"epoch": 3.6,
"grad_norm": 0.7619750499725342,
"learning_rate": 7.028307300054499e-06,
"loss": 0.2164,
"step": 450
},
{
"epoch": 3.608,
"grad_norm": 0.8766146302223206,
"learning_rate": 7.014855242642662e-06,
"loss": 0.3432,
"step": 451
},
{
"epoch": 3.616,
"grad_norm": 0.9562556147575378,
"learning_rate": 7.0013857510808934e-06,
"loss": 0.3581,
"step": 452
},
{
"epoch": 3.624,
"grad_norm": 0.7040983438491821,
"learning_rate": 6.987898941918082e-06,
"loss": 0.2897,
"step": 453
},
{
"epoch": 3.632,
"grad_norm": 0.8851937055587769,
"learning_rate": 6.974394931852957e-06,
"loss": 0.2532,
"step": 454
},
{
"epoch": 3.64,
"grad_norm": 0.9052807092666626,
"learning_rate": 6.960873837733089e-06,
"loss": 0.3085,
"step": 455
},
{
"epoch": 3.648,
"grad_norm": 0.8503252267837524,
"learning_rate": 6.94733577655387e-06,
"loss": 0.3077,
"step": 456
},
{
"epoch": 3.656,
"grad_norm": 0.7360372543334961,
"learning_rate": 6.933780865457508e-06,
"loss": 0.2262,
"step": 457
},
{
"epoch": 3.664,
"grad_norm": 0.813845157623291,
"learning_rate": 6.920209221732007e-06,
"loss": 0.3619,
"step": 458
},
{
"epoch": 3.672,
"grad_norm": 0.8746098875999451,
"learning_rate": 6.90662096281016e-06,
"loss": 0.3011,
"step": 459
},
{
"epoch": 3.68,
"grad_norm": 0.828517496585846,
"learning_rate": 6.893016206268518e-06,
"loss": 0.3173,
"step": 460
},
{
"epoch": 3.6879999999999997,
"grad_norm": 0.7118755578994751,
"learning_rate": 6.879395069826394e-06,
"loss": 0.3181,
"step": 461
},
{
"epoch": 3.6959999999999997,
"grad_norm": 1.315356969833374,
"learning_rate": 6.865757671344827e-06,
"loss": 0.3079,
"step": 462
},
{
"epoch": 3.7039999999999997,
"grad_norm": 0.8539767861366272,
"learning_rate": 6.85210412882557e-06,
"loss": 0.3023,
"step": 463
},
{
"epoch": 3.7119999999999997,
"grad_norm": 0.7354301810264587,
"learning_rate": 6.838434560410064e-06,
"loss": 0.2199,
"step": 464
},
{
"epoch": 3.7199999999999998,
"grad_norm": 0.8764749765396118,
"learning_rate": 6.824749084378428e-06,
"loss": 0.3109,
"step": 465
},
{
"epoch": 3.7279999999999998,
"grad_norm": 0.8957481980323792,
"learning_rate": 6.811047819148413e-06,
"loss": 0.3485,
"step": 466
},
{
"epoch": 3.7359999999999998,
"grad_norm": 0.9940058588981628,
"learning_rate": 6.7973308832744035e-06,
"loss": 0.2532,
"step": 467
},
{
"epoch": 3.7439999999999998,
"grad_norm": 0.7741367816925049,
"learning_rate": 6.783598395446371e-06,
"loss": 0.2186,
"step": 468
},
{
"epoch": 3.752,
"grad_norm": 0.7404496073722839,
"learning_rate": 6.769850474488859e-06,
"loss": 0.2587,
"step": 469
},
{
"epoch": 3.76,
"grad_norm": 0.9231681227684021,
"learning_rate": 6.756087239359948e-06,
"loss": 0.3029,
"step": 470
},
{
"epoch": 3.768,
"grad_norm": 0.7261999845504761,
"learning_rate": 6.742308809150232e-06,
"loss": 0.2329,
"step": 471
},
{
"epoch": 3.776,
"grad_norm": 0.8459004163742065,
"learning_rate": 6.728515303081782e-06,
"loss": 0.2336,
"step": 472
},
{
"epoch": 3.784,
"grad_norm": 0.9713183641433716,
"learning_rate": 6.714706840507122e-06,
"loss": 0.266,
"step": 473
},
{
"epoch": 3.792,
"grad_norm": 0.7366663217544556,
"learning_rate": 6.700883540908185e-06,
"loss": 0.3162,
"step": 474
},
{
"epoch": 3.8,
"grad_norm": 0.8059794306755066,
"learning_rate": 6.687045523895292e-06,
"loss": 0.2596,
"step": 475
},
{
"epoch": 3.808,
"grad_norm": 0.9890598654747009,
"learning_rate": 6.673192909206109e-06,
"loss": 0.3798,
"step": 476
},
{
"epoch": 3.816,
"grad_norm": 0.7894034385681152,
"learning_rate": 6.6593258167046115e-06,
"loss": 0.3121,
"step": 477
},
{
"epoch": 3.824,
"grad_norm": 0.8938712477684021,
"learning_rate": 6.64544436638005e-06,
"loss": 0.2633,
"step": 478
},
{
"epoch": 3.832,
"grad_norm": 0.738814115524292,
"learning_rate": 6.63154867834591e-06,
"loss": 0.2544,
"step": 479
},
{
"epoch": 3.84,
"grad_norm": 0.9438901543617249,
"learning_rate": 6.617638872838874e-06,
"loss": 0.2366,
"step": 480
},
{
"epoch": 3.848,
"grad_norm": 1.0699201822280884,
"learning_rate": 6.603715070217779e-06,
"loss": 0.3156,
"step": 481
},
{
"epoch": 3.856,
"grad_norm": 0.753672182559967,
"learning_rate": 6.589777390962575e-06,
"loss": 0.3096,
"step": 482
},
{
"epoch": 3.864,
"grad_norm": 1.0672876834869385,
"learning_rate": 6.5758259556732896e-06,
"loss": 0.2431,
"step": 483
},
{
"epoch": 3.872,
"grad_norm": 0.756453275680542,
"learning_rate": 6.561860885068972e-06,
"loss": 0.2151,
"step": 484
},
{
"epoch": 3.88,
"grad_norm": 0.8135131001472473,
"learning_rate": 6.547882299986658e-06,
"loss": 0.3089,
"step": 485
},
{
"epoch": 3.888,
"grad_norm": 0.7398325204849243,
"learning_rate": 6.53389032138032e-06,
"loss": 0.326,
"step": 486
},
{
"epoch": 3.896,
"grad_norm": 0.8907092809677124,
"learning_rate": 6.519885070319827e-06,
"loss": 0.2609,
"step": 487
},
{
"epoch": 3.904,
"grad_norm": 0.9507347345352173,
"learning_rate": 6.505866667989884e-06,
"loss": 0.3,
"step": 488
},
{
"epoch": 3.912,
"grad_norm": 0.7755032777786255,
"learning_rate": 6.491835235688999e-06,
"loss": 0.2554,
"step": 489
},
{
"epoch": 3.92,
"grad_norm": 0.9234123229980469,
"learning_rate": 6.477790894828422e-06,
"loss": 0.3371,
"step": 490
},
{
"epoch": 3.928,
"grad_norm": 0.811630129814148,
"learning_rate": 6.463733766931096e-06,
"loss": 0.2166,
"step": 491
},
{
"epoch": 3.936,
"grad_norm": 0.6761499047279358,
"learning_rate": 6.449663973630613e-06,
"loss": 0.2017,
"step": 492
},
{
"epoch": 3.944,
"grad_norm": 0.7352895736694336,
"learning_rate": 6.435581636670154e-06,
"loss": 0.2395,
"step": 493
},
{
"epoch": 3.952,
"grad_norm": 0.8231508731842041,
"learning_rate": 6.421486877901436e-06,
"loss": 0.2746,
"step": 494
},
{
"epoch": 3.96,
"grad_norm": 0.7367643713951111,
"learning_rate": 6.407379819283661e-06,
"loss": 0.2449,
"step": 495
},
{
"epoch": 3.968,
"grad_norm": 0.7580301761627197,
"learning_rate": 6.393260582882462e-06,
"loss": 0.2871,
"step": 496
},
{
"epoch": 3.976,
"grad_norm": 0.9079432487487793,
"learning_rate": 6.379129290868837e-06,
"loss": 0.2737,
"step": 497
},
{
"epoch": 3.984,
"grad_norm": 0.9549089074134827,
"learning_rate": 6.364986065518106e-06,
"loss": 0.4168,
"step": 498
},
{
"epoch": 3.992,
"grad_norm": 0.900466799736023,
"learning_rate": 6.350831029208844e-06,
"loss": 0.3325,
"step": 499
},
{
"epoch": 4.0,
"grad_norm": 0.8359453678131104,
"learning_rate": 6.336664304421818e-06,
"loss": 0.2557,
"step": 500
},
{
"epoch": 4.008,
"grad_norm": 1.3857921361923218,
"learning_rate": 6.322486013738942e-06,
"loss": 0.2166,
"step": 501
},
{
"epoch": 4.016,
"grad_norm": 1.0960663557052612,
"learning_rate": 6.308296279842204e-06,
"loss": 0.1362,
"step": 502
},
{
"epoch": 4.024,
"grad_norm": 1.209594488143921,
"learning_rate": 6.294095225512604e-06,
"loss": 0.1746,
"step": 503
},
{
"epoch": 4.032,
"grad_norm": 0.7824774980545044,
"learning_rate": 6.279882973629101e-06,
"loss": 0.1169,
"step": 504
},
{
"epoch": 4.04,
"grad_norm": 1.3840663433074951,
"learning_rate": 6.265659647167542e-06,
"loss": 0.2345,
"step": 505
},
{
"epoch": 4.048,
"grad_norm": 1.2887827157974243,
"learning_rate": 6.2514253691996e-06,
"loss": 0.18,
"step": 506
},
{
"epoch": 4.056,
"grad_norm": 1.167967677116394,
"learning_rate": 6.237180262891709e-06,
"loss": 0.1821,
"step": 507
},
{
"epoch": 4.064,
"grad_norm": 1.2175976037979126,
"learning_rate": 6.222924451504001e-06,
"loss": 0.1129,
"step": 508
},
{
"epoch": 4.072,
"grad_norm": 1.063623070716858,
"learning_rate": 6.208658058389232e-06,
"loss": 0.1169,
"step": 509
},
{
"epoch": 4.08,
"grad_norm": 0.8346629738807678,
"learning_rate": 6.194381206991723e-06,
"loss": 0.1423,
"step": 510
},
{
"epoch": 4.088,
"grad_norm": 0.9246945381164551,
"learning_rate": 6.180094020846291e-06,
"loss": 0.141,
"step": 511
},
{
"epoch": 4.096,
"grad_norm": 0.8086335062980652,
"learning_rate": 6.165796623577171e-06,
"loss": 0.0987,
"step": 512
},
{
"epoch": 4.104,
"grad_norm": 0.7477990984916687,
"learning_rate": 6.15148913889696e-06,
"loss": 0.0986,
"step": 513
},
{
"epoch": 4.112,
"grad_norm": 0.836835503578186,
"learning_rate": 6.1371716906055336e-06,
"loss": 0.1293,
"step": 514
},
{
"epoch": 4.12,
"grad_norm": 0.9984874725341797,
"learning_rate": 6.122844402588982e-06,
"loss": 0.1184,
"step": 515
},
{
"epoch": 4.128,
"grad_norm": 0.7542848587036133,
"learning_rate": 6.10850739881854e-06,
"loss": 0.1256,
"step": 516
},
{
"epoch": 4.136,
"grad_norm": 1.12161123752594,
"learning_rate": 6.094160803349508e-06,
"loss": 0.1174,
"step": 517
},
{
"epoch": 4.144,
"grad_norm": 0.9414368867874146,
"learning_rate": 6.079804740320181e-06,
"loss": 0.1302,
"step": 518
},
{
"epoch": 4.152,
"grad_norm": 0.9736228585243225,
"learning_rate": 6.065439333950776e-06,
"loss": 0.1686,
"step": 519
},
{
"epoch": 4.16,
"grad_norm": 0.6382215023040771,
"learning_rate": 6.051064708542357e-06,
"loss": 0.0873,
"step": 520
},
{
"epoch": 4.168,
"grad_norm": 0.7092307209968567,
"learning_rate": 6.036680988475756e-06,
"loss": 0.0862,
"step": 521
},
{
"epoch": 4.176,
"grad_norm": 0.7188888788223267,
"learning_rate": 6.022288298210502e-06,
"loss": 0.1247,
"step": 522
},
{
"epoch": 4.184,
"grad_norm": 0.7763389945030212,
"learning_rate": 6.00788676228374e-06,
"loss": 0.0686,
"step": 523
},
{
"epoch": 4.192,
"grad_norm": 0.7515356540679932,
"learning_rate": 5.993476505309154e-06,
"loss": 0.1354,
"step": 524
},
{
"epoch": 4.2,
"grad_norm": 0.9005271792411804,
"learning_rate": 5.979057651975893e-06,
"loss": 0.1756,
"step": 525
},
{
"epoch": 4.208,
"grad_norm": 0.7862653136253357,
"learning_rate": 5.964630327047485e-06,
"loss": 0.1146,
"step": 526
},
{
"epoch": 4.216,
"grad_norm": 0.8641824126243591,
"learning_rate": 5.9501946553607615e-06,
"loss": 0.1673,
"step": 527
},
{
"epoch": 4.224,
"grad_norm": 1.1484317779541016,
"learning_rate": 5.935750761824777e-06,
"loss": 0.1787,
"step": 528
},
{
"epoch": 4.232,
"grad_norm": 0.9002270102500916,
"learning_rate": 5.921298771419731e-06,
"loss": 0.1587,
"step": 529
},
{
"epoch": 4.24,
"grad_norm": 0.951327383518219,
"learning_rate": 5.906838809195879e-06,
"loss": 0.1974,
"step": 530
},
{
"epoch": 4.248,
"grad_norm": 0.7668549418449402,
"learning_rate": 5.8923710002724595e-06,
"loss": 0.1783,
"step": 531
},
{
"epoch": 4.256,
"grad_norm": 0.773178219795227,
"learning_rate": 5.877895469836604e-06,
"loss": 0.1061,
"step": 532
},
{
"epoch": 4.264,
"grad_norm": 0.7705480456352234,
"learning_rate": 5.863412343142258e-06,
"loss": 0.142,
"step": 533
},
{
"epoch": 4.272,
"grad_norm": 0.9089683294296265,
"learning_rate": 5.848921745509094e-06,
"loss": 0.1697,
"step": 534
},
{
"epoch": 4.28,
"grad_norm": 0.7634876370429993,
"learning_rate": 5.8344238023214305e-06,
"loss": 0.1261,
"step": 535
},
{
"epoch": 4.288,
"grad_norm": 0.7811418175697327,
"learning_rate": 5.819918639027149e-06,
"loss": 0.122,
"step": 536
},
{
"epoch": 4.296,
"grad_norm": 0.7748238444328308,
"learning_rate": 5.805406381136598e-06,
"loss": 0.1597,
"step": 537
},
{
"epoch": 4.304,
"grad_norm": 0.8066184520721436,
"learning_rate": 5.790887154221521e-06,
"loss": 0.1219,
"step": 538
},
{
"epoch": 4.312,
"grad_norm": 0.7770021557807922,
"learning_rate": 5.776361083913959e-06,
"loss": 0.1352,
"step": 539
},
{
"epoch": 4.32,
"grad_norm": 0.7021104097366333,
"learning_rate": 5.7618282959051685e-06,
"loss": 0.1294,
"step": 540
},
{
"epoch": 4.328,
"grad_norm": 0.7303136587142944,
"learning_rate": 5.747288915944533e-06,
"loss": 0.1032,
"step": 541
},
{
"epoch": 4.336,
"grad_norm": 0.758567750453949,
"learning_rate": 5.7327430698384775e-06,
"loss": 0.1577,
"step": 542
},
{
"epoch": 4.344,
"grad_norm": 0.8246206641197205,
"learning_rate": 5.718190883449373e-06,
"loss": 0.0904,
"step": 543
},
{
"epoch": 4.352,
"grad_norm": 0.9011938571929932,
"learning_rate": 5.703632482694453e-06,
"loss": 0.1311,
"step": 544
},
{
"epoch": 4.36,
"grad_norm": 0.6359648704528809,
"learning_rate": 5.689067993544726e-06,
"loss": 0.1199,
"step": 545
},
{
"epoch": 4.368,
"grad_norm": 0.8066957592964172,
"learning_rate": 5.674497542023875e-06,
"loss": 0.1491,
"step": 546
},
{
"epoch": 4.376,
"grad_norm": 0.7746307253837585,
"learning_rate": 5.659921254207183e-06,
"loss": 0.1438,
"step": 547
},
{
"epoch": 4.384,
"grad_norm": 0.7862228751182556,
"learning_rate": 5.645339256220427e-06,
"loss": 0.145,
"step": 548
},
{
"epoch": 4.392,
"grad_norm": 0.7591133117675781,
"learning_rate": 5.630751674238796e-06,
"loss": 0.1081,
"step": 549
},
{
"epoch": 4.4,
"grad_norm": 1.0325382947921753,
"learning_rate": 5.616158634485793e-06,
"loss": 0.2272,
"step": 550
},
{
"epoch": 4.408,
"grad_norm": 0.6715402007102966,
"learning_rate": 5.601560263232153e-06,
"loss": 0.0886,
"step": 551
},
{
"epoch": 4.416,
"grad_norm": 0.7422818541526794,
"learning_rate": 5.5869566867947344e-06,
"loss": 0.1198,
"step": 552
},
{
"epoch": 4.424,
"grad_norm": 0.9160724878311157,
"learning_rate": 5.572348031535442e-06,
"loss": 0.1184,
"step": 553
},
{
"epoch": 4.432,
"grad_norm": 0.7711536288261414,
"learning_rate": 5.557734423860122e-06,
"loss": 0.1254,
"step": 554
},
{
"epoch": 4.44,
"grad_norm": 0.9296648502349854,
"learning_rate": 5.543115990217478e-06,
"loss": 0.2222,
"step": 555
},
{
"epoch": 4.448,
"grad_norm": 0.920160710811615,
"learning_rate": 5.528492857097966e-06,
"loss": 0.1424,
"step": 556
},
{
"epoch": 4.456,
"grad_norm": 0.788571298122406,
"learning_rate": 5.513865151032709e-06,
"loss": 0.1811,
"step": 557
},
{
"epoch": 4.464,
"grad_norm": 0.9120768904685974,
"learning_rate": 5.499232998592399e-06,
"loss": 0.116,
"step": 558
},
{
"epoch": 4.4719999999999995,
"grad_norm": 0.7307674884796143,
"learning_rate": 5.484596526386198e-06,
"loss": 0.1415,
"step": 559
},
{
"epoch": 4.48,
"grad_norm": 0.9699017405509949,
"learning_rate": 5.469955861060653e-06,
"loss": 0.2032,
"step": 560
},
{
"epoch": 4.4879999999999995,
"grad_norm": 0.8792822957038879,
"learning_rate": 5.455311129298586e-06,
"loss": 0.1764,
"step": 561
},
{
"epoch": 4.496,
"grad_norm": 0.8328776359558105,
"learning_rate": 5.44066245781801e-06,
"loss": 0.1446,
"step": 562
},
{
"epoch": 4.504,
"grad_norm": 0.7818514704704285,
"learning_rate": 5.426009973371026e-06,
"loss": 0.14,
"step": 563
},
{
"epoch": 4.5120000000000005,
"grad_norm": 0.8472930788993835,
"learning_rate": 5.4113538027427245e-06,
"loss": 0.1308,
"step": 564
},
{
"epoch": 4.52,
"grad_norm": 2.571410894393921,
"learning_rate": 5.396694072750099e-06,
"loss": 0.1595,
"step": 565
},
{
"epoch": 4.5280000000000005,
"grad_norm": 0.7498548030853271,
"learning_rate": 5.382030910240936e-06,
"loss": 0.1357,
"step": 566
},
{
"epoch": 4.536,
"grad_norm": 0.9282652735710144,
"learning_rate": 5.367364442092724e-06,
"loss": 0.1469,
"step": 567
},
{
"epoch": 4.5440000000000005,
"grad_norm": 0.8957253694534302,
"learning_rate": 5.352694795211555e-06,
"loss": 0.1435,
"step": 568
},
{
"epoch": 4.552,
"grad_norm": 0.7313421964645386,
"learning_rate": 5.338022096531028e-06,
"loss": 0.1084,
"step": 569
},
{
"epoch": 4.5600000000000005,
"grad_norm": 0.8108712434768677,
"learning_rate": 5.3233464730111426e-06,
"loss": 0.119,
"step": 570
},
{
"epoch": 4.568,
"grad_norm": 0.7092931270599365,
"learning_rate": 5.308668051637213e-06,
"loss": 0.1398,
"step": 571
},
{
"epoch": 4.576,
"grad_norm": 0.7072190046310425,
"learning_rate": 5.29398695941876e-06,
"loss": 0.1216,
"step": 572
},
{
"epoch": 4.584,
"grad_norm": 1.016250729560852,
"learning_rate": 5.279303323388413e-06,
"loss": 0.2615,
"step": 573
},
{
"epoch": 4.592,
"grad_norm": 0.6790833473205566,
"learning_rate": 5.2646172706008154e-06,
"loss": 0.0958,
"step": 574
},
{
"epoch": 4.6,
"grad_norm": 0.5775589942932129,
"learning_rate": 5.249928928131523e-06,
"loss": 0.097,
"step": 575
},
{
"epoch": 4.608,
"grad_norm": 0.7333958745002747,
"learning_rate": 5.235238423075899e-06,
"loss": 0.1343,
"step": 576
},
{
"epoch": 4.616,
"grad_norm": 0.8037298917770386,
"learning_rate": 5.220545882548024e-06,
"loss": 0.116,
"step": 577
},
{
"epoch": 4.624,
"grad_norm": 0.8875942230224609,
"learning_rate": 5.20585143367959e-06,
"loss": 0.1063,
"step": 578
},
{
"epoch": 4.632,
"grad_norm": 0.8265597820281982,
"learning_rate": 5.191155203618796e-06,
"loss": 0.1544,
"step": 579
},
{
"epoch": 4.64,
"grad_norm": 0.9249614477157593,
"learning_rate": 5.176457319529264e-06,
"loss": 0.1271,
"step": 580
},
{
"epoch": 4.648,
"grad_norm": 0.914107084274292,
"learning_rate": 5.161757908588917e-06,
"loss": 0.116,
"step": 581
},
{
"epoch": 4.656,
"grad_norm": 0.7554004192352295,
"learning_rate": 5.147057097988898e-06,
"loss": 0.1269,
"step": 582
},
{
"epoch": 4.664,
"grad_norm": 0.9406664371490479,
"learning_rate": 5.132355014932455e-06,
"loss": 0.1091,
"step": 583
},
{
"epoch": 4.672,
"grad_norm": 0.7658833265304565,
"learning_rate": 5.1176517866338495e-06,
"loss": 0.1301,
"step": 584
},
{
"epoch": 4.68,
"grad_norm": 0.6699630618095398,
"learning_rate": 5.102947540317254e-06,
"loss": 0.0935,
"step": 585
},
{
"epoch": 4.688,
"grad_norm": 0.8095746636390686,
"learning_rate": 5.088242403215644e-06,
"loss": 0.1779,
"step": 586
},
{
"epoch": 4.696,
"grad_norm": 0.7161609530448914,
"learning_rate": 5.073536502569708e-06,
"loss": 0.1411,
"step": 587
},
{
"epoch": 4.704,
"grad_norm": 0.7674558162689209,
"learning_rate": 5.058829965626742e-06,
"loss": 0.1494,
"step": 588
},
{
"epoch": 4.712,
"grad_norm": 0.8620008826255798,
"learning_rate": 5.0441229196395416e-06,
"loss": 0.1443,
"step": 589
},
{
"epoch": 4.72,
"grad_norm": 0.7442160248756409,
"learning_rate": 5.029415491865311e-06,
"loss": 0.1328,
"step": 590
},
{
"epoch": 4.728,
"grad_norm": 0.7636533975601196,
"learning_rate": 5.014707809564562e-06,
"loss": 0.1444,
"step": 591
},
{
"epoch": 4.736,
"grad_norm": 0.8142823576927185,
"learning_rate": 5e-06,
"loss": 0.1867,
"step": 592
},
{
"epoch": 4.744,
"grad_norm": 0.9690331816673279,
"learning_rate": 4.98529219043544e-06,
"loss": 0.1497,
"step": 593
},
{
"epoch": 4.752,
"grad_norm": 1.0349537134170532,
"learning_rate": 4.97058450813469e-06,
"loss": 0.1369,
"step": 594
},
{
"epoch": 4.76,
"grad_norm": 0.9176954627037048,
"learning_rate": 4.955877080360462e-06,
"loss": 0.1553,
"step": 595
},
{
"epoch": 4.768,
"grad_norm": 0.7095694541931152,
"learning_rate": 4.94117003437326e-06,
"loss": 0.1247,
"step": 596
},
{
"epoch": 4.776,
"grad_norm": 0.8611987829208374,
"learning_rate": 4.926463497430293e-06,
"loss": 0.1724,
"step": 597
},
{
"epoch": 4.784,
"grad_norm": 0.7742799520492554,
"learning_rate": 4.911757596784358e-06,
"loss": 0.123,
"step": 598
},
{
"epoch": 4.792,
"grad_norm": 0.580224335193634,
"learning_rate": 4.897052459682749e-06,
"loss": 0.1059,
"step": 599
},
{
"epoch": 4.8,
"grad_norm": 0.7499905228614807,
"learning_rate": 4.882348213366152e-06,
"loss": 0.1635,
"step": 600
},
{
"epoch": 4.808,
"grad_norm": 0.8561443090438843,
"learning_rate": 4.867644985067548e-06,
"loss": 0.0817,
"step": 601
},
{
"epoch": 4.816,
"grad_norm": 0.9102675318717957,
"learning_rate": 4.8529429020111035e-06,
"loss": 0.2103,
"step": 602
},
{
"epoch": 4.824,
"grad_norm": 0.7099835872650146,
"learning_rate": 4.838242091411085e-06,
"loss": 0.1099,
"step": 603
},
{
"epoch": 4.832,
"grad_norm": 0.897228479385376,
"learning_rate": 4.823542680470738e-06,
"loss": 0.1302,
"step": 604
},
{
"epoch": 4.84,
"grad_norm": 0.6506624817848206,
"learning_rate": 4.808844796381205e-06,
"loss": 0.12,
"step": 605
},
{
"epoch": 4.848,
"grad_norm": 0.8063788414001465,
"learning_rate": 4.794148566320412e-06,
"loss": 0.1609,
"step": 606
},
{
"epoch": 4.856,
"grad_norm": 0.7618057727813721,
"learning_rate": 4.779454117451978e-06,
"loss": 0.1449,
"step": 607
},
{
"epoch": 4.864,
"grad_norm": 0.8923940658569336,
"learning_rate": 4.7647615769241e-06,
"loss": 0.1854,
"step": 608
},
{
"epoch": 4.872,
"grad_norm": 0.7307261824607849,
"learning_rate": 4.750071071868478e-06,
"loss": 0.1354,
"step": 609
},
{
"epoch": 4.88,
"grad_norm": 0.7396050691604614,
"learning_rate": 4.7353827293991845e-06,
"loss": 0.097,
"step": 610
},
{
"epoch": 4.888,
"grad_norm": 0.7688437700271606,
"learning_rate": 4.720696676611589e-06,
"loss": 0.1447,
"step": 611
},
{
"epoch": 4.896,
"grad_norm": 0.9390592575073242,
"learning_rate": 4.706013040581242e-06,
"loss": 0.1153,
"step": 612
},
{
"epoch": 4.904,
"grad_norm": 0.7500796914100647,
"learning_rate": 4.691331948362789e-06,
"loss": 0.1404,
"step": 613
},
{
"epoch": 4.912,
"grad_norm": 0.913267195224762,
"learning_rate": 4.676653526988858e-06,
"loss": 0.1348,
"step": 614
},
{
"epoch": 4.92,
"grad_norm": 0.8776205778121948,
"learning_rate": 4.661977903468974e-06,
"loss": 0.1285,
"step": 615
},
{
"epoch": 4.928,
"grad_norm": 1.010945200920105,
"learning_rate": 4.647305204788445e-06,
"loss": 0.1646,
"step": 616
},
{
"epoch": 4.936,
"grad_norm": 0.7763349413871765,
"learning_rate": 4.632635557907277e-06,
"loss": 0.1549,
"step": 617
},
{
"epoch": 4.944,
"grad_norm": 0.7637826800346375,
"learning_rate": 4.617969089759066e-06,
"loss": 0.1077,
"step": 618
},
{
"epoch": 4.952,
"grad_norm": 0.9262024760246277,
"learning_rate": 4.603305927249902e-06,
"loss": 0.185,
"step": 619
},
{
"epoch": 4.96,
"grad_norm": 0.821549117565155,
"learning_rate": 4.588646197257278e-06,
"loss": 0.1219,
"step": 620
},
{
"epoch": 4.968,
"grad_norm": 0.7055884003639221,
"learning_rate": 4.573990026628976e-06,
"loss": 0.118,
"step": 621
},
{
"epoch": 4.976,
"grad_norm": 0.7518096566200256,
"learning_rate": 4.559337542181993e-06,
"loss": 0.168,
"step": 622
},
{
"epoch": 4.984,
"grad_norm": 0.7408725023269653,
"learning_rate": 4.544688870701416e-06,
"loss": 0.1333,
"step": 623
},
{
"epoch": 4.992,
"grad_norm": 0.6920803189277649,
"learning_rate": 4.53004413893935e-06,
"loss": 0.1177,
"step": 624
},
{
"epoch": 5.0,
"grad_norm": 0.8042945861816406,
"learning_rate": 4.5154034736138035e-06,
"loss": 0.1387,
"step": 625
},
{
"epoch": 5.008,
"grad_norm": 0.8056797981262207,
"learning_rate": 4.500767001407604e-06,
"loss": 0.0511,
"step": 626
},
{
"epoch": 5.016,
"grad_norm": 0.8114173412322998,
"learning_rate": 4.486134848967292e-06,
"loss": 0.0592,
"step": 627
},
{
"epoch": 5.024,
"grad_norm": 0.6595902442932129,
"learning_rate": 4.471507142902036e-06,
"loss": 0.0703,
"step": 628
},
{
"epoch": 5.032,
"grad_norm": 0.7525659799575806,
"learning_rate": 4.4568840097825225e-06,
"loss": 0.053,
"step": 629
},
{
"epoch": 5.04,
"grad_norm": 0.7577898502349854,
"learning_rate": 4.4422655761398785e-06,
"loss": 0.0562,
"step": 630
},
{
"epoch": 5.048,
"grad_norm": 0.66656893491745,
"learning_rate": 4.427651968464559e-06,
"loss": 0.031,
"step": 631
},
{
"epoch": 5.056,
"grad_norm": 1.139257788658142,
"learning_rate": 4.413043313205266e-06,
"loss": 0.0474,
"step": 632
},
{
"epoch": 5.064,
"grad_norm": 1.2135770320892334,
"learning_rate": 4.3984397367678475e-06,
"loss": 0.0669,
"step": 633
},
{
"epoch": 5.072,
"grad_norm": 4.926527976989746,
"learning_rate": 4.383841365514208e-06,
"loss": 0.0528,
"step": 634
},
{
"epoch": 5.08,
"grad_norm": 0.8709739446640015,
"learning_rate": 4.369248325761205e-06,
"loss": 0.0477,
"step": 635
},
{
"epoch": 5.088,
"grad_norm": 0.8318364024162292,
"learning_rate": 4.354660743779575e-06,
"loss": 0.0323,
"step": 636
},
{
"epoch": 5.096,
"grad_norm": 0.6887449026107788,
"learning_rate": 4.340078745792818e-06,
"loss": 0.0566,
"step": 637
},
{
"epoch": 5.104,
"grad_norm": 0.7675641775131226,
"learning_rate": 4.325502457976126e-06,
"loss": 0.0501,
"step": 638
},
{
"epoch": 5.112,
"grad_norm": 0.5305010080337524,
"learning_rate": 4.310932006455276e-06,
"loss": 0.04,
"step": 639
},
{
"epoch": 5.12,
"grad_norm": 0.6297786831855774,
"learning_rate": 4.296367517305548e-06,
"loss": 0.0405,
"step": 640
},
{
"epoch": 5.128,
"grad_norm": 0.7022202014923096,
"learning_rate": 4.281809116550629e-06,
"loss": 0.0622,
"step": 641
},
{
"epoch": 5.136,
"grad_norm": 0.5648162961006165,
"learning_rate": 4.267256930161523e-06,
"loss": 0.0484,
"step": 642
},
{
"epoch": 5.144,
"grad_norm": 0.6618186235427856,
"learning_rate": 4.252711084055468e-06,
"loss": 0.0682,
"step": 643
},
{
"epoch": 5.152,
"grad_norm": 0.6455008387565613,
"learning_rate": 4.238171704094833e-06,
"loss": 0.0664,
"step": 644
},
{
"epoch": 5.16,
"grad_norm": 0.6835476160049438,
"learning_rate": 4.223638916086044e-06,
"loss": 0.049,
"step": 645
},
{
"epoch": 5.168,
"grad_norm": 0.7162488102912903,
"learning_rate": 4.209112845778481e-06,
"loss": 0.0539,
"step": 646
},
{
"epoch": 5.176,
"grad_norm": 0.7715399265289307,
"learning_rate": 4.194593618863404e-06,
"loss": 0.0636,
"step": 647
},
{
"epoch": 5.184,
"grad_norm": 0.6754415035247803,
"learning_rate": 4.180081360972852e-06,
"loss": 0.0603,
"step": 648
},
{
"epoch": 5.192,
"grad_norm": 0.8835145831108093,
"learning_rate": 4.165576197678571e-06,
"loss": 0.0217,
"step": 649
},
{
"epoch": 5.2,
"grad_norm": 0.7698906660079956,
"learning_rate": 4.151078254490908e-06,
"loss": 0.0442,
"step": 650
},
{
"epoch": 5.208,
"grad_norm": 0.8498547077178955,
"learning_rate": 4.136587656857744e-06,
"loss": 0.0706,
"step": 651
},
{
"epoch": 5.216,
"grad_norm": 0.5944578051567078,
"learning_rate": 4.122104530163397e-06,
"loss": 0.0354,
"step": 652
},
{
"epoch": 5.224,
"grad_norm": 0.7195687890052795,
"learning_rate": 4.107628999727542e-06,
"loss": 0.0572,
"step": 653
},
{
"epoch": 5.232,
"grad_norm": 0.8516260981559753,
"learning_rate": 4.09316119080412e-06,
"loss": 0.068,
"step": 654
},
{
"epoch": 5.24,
"grad_norm": 0.7528324127197266,
"learning_rate": 4.0787012285802695e-06,
"loss": 0.0809,
"step": 655
},
{
"epoch": 5.248,
"grad_norm": 1.21733820438385,
"learning_rate": 4.064249238175223e-06,
"loss": 0.0333,
"step": 656
},
{
"epoch": 5.256,
"grad_norm": 0.6287680864334106,
"learning_rate": 4.04980534463924e-06,
"loss": 0.0385,
"step": 657
},
{
"epoch": 5.264,
"grad_norm": 0.7100169658660889,
"learning_rate": 4.035369672952516e-06,
"loss": 0.0469,
"step": 658
},
{
"epoch": 5.272,
"grad_norm": 0.8435373306274414,
"learning_rate": 4.020942348024108e-06,
"loss": 0.0836,
"step": 659
},
{
"epoch": 5.28,
"grad_norm": 0.7471747994422913,
"learning_rate": 4.0065234946908456e-06,
"loss": 0.057,
"step": 660
},
{
"epoch": 5.288,
"grad_norm": 0.6374616026878357,
"learning_rate": 3.992113237716261e-06,
"loss": 0.0501,
"step": 661
},
{
"epoch": 5.296,
"grad_norm": 1.1847115755081177,
"learning_rate": 3.977711701789499e-06,
"loss": 0.0746,
"step": 662
},
{
"epoch": 5.304,
"grad_norm": 0.5774708986282349,
"learning_rate": 3.963319011524246e-06,
"loss": 0.0347,
"step": 663
},
{
"epoch": 5.312,
"grad_norm": 0.6785467863082886,
"learning_rate": 3.948935291457645e-06,
"loss": 0.0482,
"step": 664
},
{
"epoch": 5.32,
"grad_norm": 0.55971360206604,
"learning_rate": 3.934560666049226e-06,
"loss": 0.0369,
"step": 665
},
{
"epoch": 5.328,
"grad_norm": 0.5369053483009338,
"learning_rate": 3.920195259679822e-06,
"loss": 0.0352,
"step": 666
},
{
"epoch": 5.336,
"grad_norm": 0.48250868916511536,
"learning_rate": 3.905839196650494e-06,
"loss": 0.0462,
"step": 667
},
{
"epoch": 5.344,
"grad_norm": 0.5979182720184326,
"learning_rate": 3.891492601181462e-06,
"loss": 0.0447,
"step": 668
},
{
"epoch": 5.352,
"grad_norm": 0.7114503979682922,
"learning_rate": 3.877155597411019e-06,
"loss": 0.0316,
"step": 669
},
{
"epoch": 5.36,
"grad_norm": 0.6035363078117371,
"learning_rate": 3.862828309394469e-06,
"loss": 0.0582,
"step": 670
},
{
"epoch": 5.368,
"grad_norm": 0.8595693111419678,
"learning_rate": 3.8485108611030415e-06,
"loss": 0.0796,
"step": 671
},
{
"epoch": 5.376,
"grad_norm": 0.6660138964653015,
"learning_rate": 3.834203376422831e-06,
"loss": 0.058,
"step": 672
},
{
"epoch": 5.384,
"grad_norm": 0.6106982231140137,
"learning_rate": 3.8199059791537105e-06,
"loss": 0.0403,
"step": 673
},
{
"epoch": 5.392,
"grad_norm": 0.6765463352203369,
"learning_rate": 3.805618793008279e-06,
"loss": 0.0393,
"step": 674
},
{
"epoch": 5.4,
"grad_norm": 0.7175659537315369,
"learning_rate": 3.7913419416107692e-06,
"loss": 0.0592,
"step": 675
},
{
"epoch": 5.408,
"grad_norm": 0.7095440030097961,
"learning_rate": 3.777075548496001e-06,
"loss": 0.0448,
"step": 676
},
{
"epoch": 5.416,
"grad_norm": 0.858165979385376,
"learning_rate": 3.7628197371082916e-06,
"loss": 0.0502,
"step": 677
},
{
"epoch": 5.424,
"grad_norm": 0.6832898855209351,
"learning_rate": 3.7485746308004013e-06,
"loss": 0.0405,
"step": 678
},
{
"epoch": 5.432,
"grad_norm": 0.6566227078437805,
"learning_rate": 3.7343403528324574e-06,
"loss": 0.0572,
"step": 679
},
{
"epoch": 5.44,
"grad_norm": 0.9831671118736267,
"learning_rate": 3.7201170263709004e-06,
"loss": 0.0922,
"step": 680
},
{
"epoch": 5.448,
"grad_norm": 0.6447628140449524,
"learning_rate": 3.705904774487396e-06,
"loss": 0.0433,
"step": 681
},
{
"epoch": 5.456,
"grad_norm": 0.7340384721755981,
"learning_rate": 3.6917037201577977e-06,
"loss": 0.0593,
"step": 682
},
{
"epoch": 5.464,
"grad_norm": 0.6065418124198914,
"learning_rate": 3.6775139862610577e-06,
"loss": 0.0731,
"step": 683
},
{
"epoch": 5.4719999999999995,
"grad_norm": 1.329861044883728,
"learning_rate": 3.6633356955781827e-06,
"loss": 0.0302,
"step": 684
},
{
"epoch": 5.48,
"grad_norm": 0.5943688154220581,
"learning_rate": 3.649168970791157e-06,
"loss": 0.0446,
"step": 685
},
{
"epoch": 5.4879999999999995,
"grad_norm": 0.6232944130897522,
"learning_rate": 3.635013934481895e-06,
"loss": 0.052,
"step": 686
},
{
"epoch": 5.496,
"grad_norm": 1.0451732873916626,
"learning_rate": 3.620870709131163e-06,
"loss": 0.0594,
"step": 687
},
{
"epoch": 5.504,
"grad_norm": 0.628891110420227,
"learning_rate": 3.6067394171175397e-06,
"loss": 0.0583,
"step": 688
},
{
"epoch": 5.5120000000000005,
"grad_norm": 0.7899133563041687,
"learning_rate": 3.5926201807163384e-06,
"loss": 0.0299,
"step": 689
},
{
"epoch": 5.52,
"grad_norm": 0.7041831612586975,
"learning_rate": 3.578513122098566e-06,
"loss": 0.0344,
"step": 690
},
{
"epoch": 5.5280000000000005,
"grad_norm": 0.6907998919487,
"learning_rate": 3.564418363329848e-06,
"loss": 0.0616,
"step": 691
},
{
"epoch": 5.536,
"grad_norm": 0.6608344912528992,
"learning_rate": 3.5503360263693887e-06,
"loss": 0.0471,
"step": 692
},
{
"epoch": 5.5440000000000005,
"grad_norm": 0.7428660988807678,
"learning_rate": 3.5362662330689067e-06,
"loss": 0.0541,
"step": 693
},
{
"epoch": 5.552,
"grad_norm": 0.9560613036155701,
"learning_rate": 3.5222091051715803e-06,
"loss": 0.0642,
"step": 694
},
{
"epoch": 5.5600000000000005,
"grad_norm": 0.7475034594535828,
"learning_rate": 3.5081647643110028e-06,
"loss": 0.0732,
"step": 695
},
{
"epoch": 5.568,
"grad_norm": 0.6867515444755554,
"learning_rate": 3.4941333320101173e-06,
"loss": 0.0565,
"step": 696
},
{
"epoch": 5.576,
"grad_norm": 0.5686975717544556,
"learning_rate": 3.480114929680176e-06,
"loss": 0.0384,
"step": 697
},
{
"epoch": 5.584,
"grad_norm": 0.6070277690887451,
"learning_rate": 3.466109678619681e-06,
"loss": 0.0484,
"step": 698
},
{
"epoch": 5.592,
"grad_norm": 0.6168974041938782,
"learning_rate": 3.4521177000133456e-06,
"loss": 0.0508,
"step": 699
},
{
"epoch": 5.6,
"grad_norm": 0.6029306650161743,
"learning_rate": 3.4381391149310294e-06,
"loss": 0.0502,
"step": 700
},
{
"epoch": 5.608,
"grad_norm": 0.8586588501930237,
"learning_rate": 3.4241740443267112e-06,
"loss": 0.0787,
"step": 701
},
{
"epoch": 5.616,
"grad_norm": 0.5512130856513977,
"learning_rate": 3.4102226090374246e-06,
"loss": 0.0357,
"step": 702
},
{
"epoch": 5.624,
"grad_norm": 0.7166364789009094,
"learning_rate": 3.3962849297822225e-06,
"loss": 0.0565,
"step": 703
},
{
"epoch": 5.632,
"grad_norm": 0.7719413042068481,
"learning_rate": 3.3823611271611266e-06,
"loss": 0.0626,
"step": 704
},
{
"epoch": 5.64,
"grad_norm": 1.0928401947021484,
"learning_rate": 3.368451321654091e-06,
"loss": 0.0686,
"step": 705
},
{
"epoch": 5.648,
"grad_norm": 0.7056580781936646,
"learning_rate": 3.35455563361995e-06,
"loss": 0.0646,
"step": 706
},
{
"epoch": 5.656,
"grad_norm": 0.5838544368743896,
"learning_rate": 3.3406741832953893e-06,
"loss": 0.0367,
"step": 707
},
{
"epoch": 5.664,
"grad_norm": 1.1275421380996704,
"learning_rate": 3.3268070907938915e-06,
"loss": 0.0488,
"step": 708
},
{
"epoch": 5.672,
"grad_norm": 0.8408868312835693,
"learning_rate": 3.3129544761047093e-06,
"loss": 0.0597,
"step": 709
},
{
"epoch": 5.68,
"grad_norm": 0.571863055229187,
"learning_rate": 3.2991164590918162e-06,
"loss": 0.0518,
"step": 710
},
{
"epoch": 5.688,
"grad_norm": 0.6888249516487122,
"learning_rate": 3.2852931594928804e-06,
"loss": 0.0757,
"step": 711
},
{
"epoch": 5.696,
"grad_norm": 0.5743219256401062,
"learning_rate": 3.271484696918218e-06,
"loss": 0.0268,
"step": 712
},
{
"epoch": 5.704,
"grad_norm": 0.7105786800384521,
"learning_rate": 3.2576911908497695e-06,
"loss": 0.0561,
"step": 713
},
{
"epoch": 5.712,
"grad_norm": 0.5829000473022461,
"learning_rate": 3.2439127606400546e-06,
"loss": 0.0453,
"step": 714
},
{
"epoch": 5.72,
"grad_norm": 0.6392514705657959,
"learning_rate": 3.2301495255111426e-06,
"loss": 0.0542,
"step": 715
},
{
"epoch": 5.728,
"grad_norm": 0.8063925504684448,
"learning_rate": 3.2164016045536306e-06,
"loss": 0.0693,
"step": 716
},
{
"epoch": 5.736,
"grad_norm": 0.6815555691719055,
"learning_rate": 3.202669116725598e-06,
"loss": 0.0289,
"step": 717
},
{
"epoch": 5.744,
"grad_norm": 0.6942662000656128,
"learning_rate": 3.1889521808515888e-06,
"loss": 0.0458,
"step": 718
},
{
"epoch": 5.752,
"grad_norm": 0.8222205638885498,
"learning_rate": 3.1752509156215738e-06,
"loss": 0.0606,
"step": 719
},
{
"epoch": 5.76,
"grad_norm": 0.597169816493988,
"learning_rate": 3.1615654395899377e-06,
"loss": 0.0595,
"step": 720
},
{
"epoch": 5.768,
"grad_norm": 0.542059600353241,
"learning_rate": 3.1478958711744324e-06,
"loss": 0.0302,
"step": 721
},
{
"epoch": 5.776,
"grad_norm": 0.6608166694641113,
"learning_rate": 3.1342423286551756e-06,
"loss": 0.0638,
"step": 722
},
{
"epoch": 5.784,
"grad_norm": 0.6509668231010437,
"learning_rate": 3.120604930173608e-06,
"loss": 0.0356,
"step": 723
},
{
"epoch": 5.792,
"grad_norm": 0.6087043285369873,
"learning_rate": 3.1069837937314846e-06,
"loss": 0.0617,
"step": 724
},
{
"epoch": 5.8,
"grad_norm": 0.6988217830657959,
"learning_rate": 3.093379037189842e-06,
"loss": 0.0546,
"step": 725
},
{
"epoch": 5.808,
"grad_norm": 0.5948909521102905,
"learning_rate": 3.0797907782679944e-06,
"loss": 0.0678,
"step": 726
},
{
"epoch": 5.816,
"grad_norm": 0.6464372873306274,
"learning_rate": 3.0662191345424925e-06,
"loss": 0.0562,
"step": 727
},
{
"epoch": 5.824,
"grad_norm": 1.200032353401184,
"learning_rate": 3.0526642234461313e-06,
"loss": 0.1153,
"step": 728
},
{
"epoch": 5.832,
"grad_norm": 0.7848719954490662,
"learning_rate": 3.039126162266912e-06,
"loss": 0.0774,
"step": 729
},
{
"epoch": 5.84,
"grad_norm": 0.7536558508872986,
"learning_rate": 3.0256050681470446e-06,
"loss": 0.0878,
"step": 730
},
{
"epoch": 5.848,
"grad_norm": 0.7371671795845032,
"learning_rate": 3.012101058081919e-06,
"loss": 0.0515,
"step": 731
},
{
"epoch": 5.856,
"grad_norm": 0.6955927014350891,
"learning_rate": 2.9986142489191074e-06,
"loss": 0.0448,
"step": 732
},
{
"epoch": 5.864,
"grad_norm": 0.6752610206604004,
"learning_rate": 2.9851447573573383e-06,
"loss": 0.0673,
"step": 733
},
{
"epoch": 5.872,
"grad_norm": 0.672664225101471,
"learning_rate": 2.971692699945502e-06,
"loss": 0.051,
"step": 734
},
{
"epoch": 5.88,
"grad_norm": 0.570433497428894,
"learning_rate": 2.958258193081629e-06,
"loss": 0.0477,
"step": 735
},
{
"epoch": 5.888,
"grad_norm": 0.6220299601554871,
"learning_rate": 2.9448413530118912e-06,
"loss": 0.0465,
"step": 736
},
{
"epoch": 5.896,
"grad_norm": 0.7739304900169373,
"learning_rate": 2.9314422958295906e-06,
"loss": 0.0554,
"step": 737
},
{
"epoch": 5.904,
"grad_norm": 0.5828734040260315,
"learning_rate": 2.9180611374741623e-06,
"loss": 0.0534,
"step": 738
},
{
"epoch": 5.912,
"grad_norm": 0.47254109382629395,
"learning_rate": 2.904697993730159e-06,
"loss": 0.0344,
"step": 739
},
{
"epoch": 5.92,
"grad_norm": 0.6161917448043823,
"learning_rate": 2.891352980226262e-06,
"loss": 0.0551,
"step": 740
},
{
"epoch": 5.928,
"grad_norm": 0.6161420345306396,
"learning_rate": 2.8780262124342755e-06,
"loss": 0.0627,
"step": 741
},
{
"epoch": 5.936,
"grad_norm": 0.7250885963439941,
"learning_rate": 2.8647178056681197e-06,
"loss": 0.0488,
"step": 742
},
{
"epoch": 5.944,
"grad_norm": 0.6412259340286255,
"learning_rate": 2.8514278750828537e-06,
"loss": 0.0541,
"step": 743
},
{
"epoch": 5.952,
"grad_norm": 0.5730807781219482,
"learning_rate": 2.838156535673652e-06,
"loss": 0.0416,
"step": 744
},
{
"epoch": 5.96,
"grad_norm": 0.714428722858429,
"learning_rate": 2.8249039022748315e-06,
"loss": 0.0723,
"step": 745
},
{
"epoch": 5.968,
"grad_norm": 0.625837504863739,
"learning_rate": 2.8116700895588473e-06,
"loss": 0.0464,
"step": 746
},
{
"epoch": 5.976,
"grad_norm": 0.6484919786453247,
"learning_rate": 2.798455212035305e-06,
"loss": 0.0542,
"step": 747
},
{
"epoch": 5.984,
"grad_norm": 0.930560290813446,
"learning_rate": 2.785259384049959e-06,
"loss": 0.0837,
"step": 748
},
{
"epoch": 5.992,
"grad_norm": 0.539526641368866,
"learning_rate": 2.7720827197837475e-06,
"loss": 0.0355,
"step": 749
},
{
"epoch": 6.0,
"grad_norm": 0.653667151927948,
"learning_rate": 2.7589253332517736e-06,
"loss": 0.0448,
"step": 750
},
{
"epoch": 6.008,
"grad_norm": 0.37607839703559875,
"learning_rate": 2.745787338302341e-06,
"loss": 0.0128,
"step": 751
},
{
"epoch": 6.016,
"grad_norm": 0.5014586448669434,
"learning_rate": 2.7326688486159613e-06,
"loss": 0.0292,
"step": 752
},
{
"epoch": 6.024,
"grad_norm": 0.544358491897583,
"learning_rate": 2.7195699777043723e-06,
"loss": 0.0326,
"step": 753
},
{
"epoch": 6.032,
"grad_norm": 0.2762000262737274,
"learning_rate": 2.706490838909547e-06,
"loss": 0.0077,
"step": 754
},
{
"epoch": 6.04,
"grad_norm": 0.47887730598449707,
"learning_rate": 2.6934315454027323e-06,
"loss": 0.0224,
"step": 755
},
{
"epoch": 6.048,
"grad_norm": 0.45757901668548584,
"learning_rate": 2.680392210183446e-06,
"loss": 0.0205,
"step": 756
},
{
"epoch": 6.056,
"grad_norm": 0.37720590829849243,
"learning_rate": 2.6673729460785174e-06,
"loss": 0.0232,
"step": 757
},
{
"epoch": 6.064,
"grad_norm": 0.48954445123672485,
"learning_rate": 2.6543738657411033e-06,
"loss": 0.0196,
"step": 758
},
{
"epoch": 6.072,
"grad_norm": 0.4824325740337372,
"learning_rate": 2.6413950816497146e-06,
"loss": 0.0121,
"step": 759
},
{
"epoch": 6.08,
"grad_norm": 0.4821329712867737,
"learning_rate": 2.628436706107238e-06,
"loss": 0.0182,
"step": 760
},
{
"epoch": 6.088,
"grad_norm": 0.42847394943237305,
"learning_rate": 2.6154988512399784e-06,
"loss": 0.0147,
"step": 761
},
{
"epoch": 6.096,
"grad_norm": 0.6512643098831177,
"learning_rate": 2.6025816289966703e-06,
"loss": 0.0169,
"step": 762
},
{
"epoch": 6.104,
"grad_norm": 0.5844804644584656,
"learning_rate": 2.5896851511475184e-06,
"loss": 0.0202,
"step": 763
},
{
"epoch": 6.112,
"grad_norm": 0.44382765889167786,
"learning_rate": 2.5768095292832412e-06,
"loss": 0.0214,
"step": 764
},
{
"epoch": 6.12,
"grad_norm": 0.5254351496696472,
"learning_rate": 2.5639548748140803e-06,
"loss": 0.0287,
"step": 765
},
{
"epoch": 6.128,
"grad_norm": 0.41182348132133484,
"learning_rate": 2.5511212989688587e-06,
"loss": 0.0228,
"step": 766
},
{
"epoch": 6.136,
"grad_norm": 0.411668598651886,
"learning_rate": 2.5383089127940087e-06,
"loss": 0.0133,
"step": 767
},
{
"epoch": 6.144,
"grad_norm": 0.44052034616470337,
"learning_rate": 2.525517827152614e-06,
"loss": 0.015,
"step": 768
},
{
"epoch": 6.152,
"grad_norm": 0.4560883939266205,
"learning_rate": 2.5127481527234397e-06,
"loss": 0.0193,
"step": 769
},
{
"epoch": 6.16,
"grad_norm": 0.5043846964836121,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.024,
"step": 770
},
{
"epoch": 6.168,
"grad_norm": 0.3570071756839752,
"learning_rate": 2.487273479289574e-06,
"loss": 0.01,
"step": 771
},
{
"epoch": 6.176,
"grad_norm": 0.34656181931495667,
"learning_rate": 2.4745687007122636e-06,
"loss": 0.0118,
"step": 772
},
{
"epoch": 6.184,
"grad_norm": 0.44094350934028625,
"learning_rate": 2.4618857742000463e-06,
"loss": 0.024,
"step": 773
},
{
"epoch": 6.192,
"grad_norm": 0.4078379273414612,
"learning_rate": 2.449224809495815e-06,
"loss": 0.0267,
"step": 774
},
{
"epoch": 6.2,
"grad_norm": 0.45042550563812256,
"learning_rate": 2.436585916152426e-06,
"loss": 0.0248,
"step": 775
},
{
"epoch": 6.208,
"grad_norm": 0.5565236210823059,
"learning_rate": 2.423969203531768e-06,
"loss": 0.022,
"step": 776
},
{
"epoch": 6.216,
"grad_norm": 0.33686888217926025,
"learning_rate": 2.411374780803793e-06,
"loss": 0.0074,
"step": 777
},
{
"epoch": 6.224,
"grad_norm": 0.4109979569911957,
"learning_rate": 2.3988027569455895e-06,
"loss": 0.0114,
"step": 778
},
{
"epoch": 6.232,
"grad_norm": 0.5003446340560913,
"learning_rate": 2.3862532407404306e-06,
"loss": 0.0297,
"step": 779
},
{
"epoch": 6.24,
"grad_norm": 0.30241599678993225,
"learning_rate": 2.373726340776837e-06,
"loss": 0.0081,
"step": 780
},
{
"epoch": 6.248,
"grad_norm": 0.5940093398094177,
"learning_rate": 2.361222165447628e-06,
"loss": 0.0258,
"step": 781
},
{
"epoch": 6.256,
"grad_norm": 0.37417200207710266,
"learning_rate": 2.348740822949006e-06,
"loss": 0.011,
"step": 782
},
{
"epoch": 6.264,
"grad_norm": 0.36139291524887085,
"learning_rate": 2.33628242127959e-06,
"loss": 0.0195,
"step": 783
},
{
"epoch": 6.272,
"grad_norm": 0.4464220702648163,
"learning_rate": 2.323847068239504e-06,
"loss": 0.0182,
"step": 784
},
{
"epoch": 6.28,
"grad_norm": 0.3683570623397827,
"learning_rate": 2.3114348714294355e-06,
"loss": 0.0148,
"step": 785
},
{
"epoch": 6.288,
"grad_norm": 0.5403426885604858,
"learning_rate": 2.2990459382497086e-06,
"loss": 0.0123,
"step": 786
},
{
"epoch": 6.296,
"grad_norm": 0.47578224539756775,
"learning_rate": 2.2866803758993446e-06,
"loss": 0.0207,
"step": 787
},
{
"epoch": 6.304,
"grad_norm": 0.4909017086029053,
"learning_rate": 2.274338291375147e-06,
"loss": 0.0128,
"step": 788
},
{
"epoch": 6.312,
"grad_norm": 0.40206748247146606,
"learning_rate": 2.262019791470772e-06,
"loss": 0.0104,
"step": 789
},
{
"epoch": 6.32,
"grad_norm": 0.6169834733009338,
"learning_rate": 2.2497249827757933e-06,
"loss": 0.0316,
"step": 790
},
{
"epoch": 6.328,
"grad_norm": 0.36974218487739563,
"learning_rate": 2.2374539716748034e-06,
"loss": 0.008,
"step": 791
},
{
"epoch": 6.336,
"grad_norm": 0.455277681350708,
"learning_rate": 2.225206864346465e-06,
"loss": 0.0236,
"step": 792
},
{
"epoch": 6.344,
"grad_norm": 0.45521119236946106,
"learning_rate": 2.2129837667626147e-06,
"loss": 0.0184,
"step": 793
},
{
"epoch": 6.352,
"grad_norm": 0.542340099811554,
"learning_rate": 2.2007847846873342e-06,
"loss": 0.0257,
"step": 794
},
{
"epoch": 6.36,
"grad_norm": 0.4681219458580017,
"learning_rate": 2.188610023676041e-06,
"loss": 0.0293,
"step": 795
},
{
"epoch": 6.368,
"grad_norm": 0.4694705307483673,
"learning_rate": 2.176459589074566e-06,
"loss": 0.0147,
"step": 796
},
{
"epoch": 6.376,
"grad_norm": 0.4986151456832886,
"learning_rate": 2.164333586018259e-06,
"loss": 0.0242,
"step": 797
},
{
"epoch": 6.384,
"grad_norm": 0.3367967903614044,
"learning_rate": 2.1522321194310577e-06,
"loss": 0.0098,
"step": 798
},
{
"epoch": 6.392,
"grad_norm": 0.430538535118103,
"learning_rate": 2.1401552940245962e-06,
"loss": 0.0192,
"step": 799
},
{
"epoch": 6.4,
"grad_norm": 0.6271944046020508,
"learning_rate": 2.1281032142972933e-06,
"loss": 0.0229,
"step": 800
},
{
"epoch": 6.408,
"grad_norm": 0.33021512627601624,
"learning_rate": 2.1160759845334483e-06,
"loss": 0.0168,
"step": 801
},
{
"epoch": 6.416,
"grad_norm": 0.5376062989234924,
"learning_rate": 2.1040737088023323e-06,
"loss": 0.0154,
"step": 802
},
{
"epoch": 6.424,
"grad_norm": 0.5122645497322083,
"learning_rate": 2.0920964909573065e-06,
"loss": 0.0268,
"step": 803
},
{
"epoch": 6.432,
"grad_norm": 0.37193113565444946,
"learning_rate": 2.080144434634898e-06,
"loss": 0.0133,
"step": 804
},
{
"epoch": 6.44,
"grad_norm": 0.32150623202323914,
"learning_rate": 2.068217643253925e-06,
"loss": 0.0122,
"step": 805
},
{
"epoch": 6.448,
"grad_norm": 0.4551410675048828,
"learning_rate": 2.056316220014588e-06,
"loss": 0.0253,
"step": 806
},
{
"epoch": 6.456,
"grad_norm": 0.37979602813720703,
"learning_rate": 2.0444402678975876e-06,
"loss": 0.0175,
"step": 807
},
{
"epoch": 6.464,
"grad_norm": 0.38569003343582153,
"learning_rate": 2.0325898896632178e-06,
"loss": 0.0099,
"step": 808
},
{
"epoch": 6.4719999999999995,
"grad_norm": 0.4271693527698517,
"learning_rate": 2.0207651878505e-06,
"loss": 0.0129,
"step": 809
},
{
"epoch": 6.48,
"grad_norm": 0.3800446689128876,
"learning_rate": 2.0089662647762716e-06,
"loss": 0.0128,
"step": 810
},
{
"epoch": 6.4879999999999995,
"grad_norm": 0.4142042398452759,
"learning_rate": 1.997193222534316e-06,
"loss": 0.0158,
"step": 811
},
{
"epoch": 6.496,
"grad_norm": 0.46278542280197144,
"learning_rate": 1.9854461629944764e-06,
"loss": 0.0166,
"step": 812
},
{
"epoch": 6.504,
"grad_norm": 0.40608370304107666,
"learning_rate": 1.9737251878017678e-06,
"loss": 0.0205,
"step": 813
},
{
"epoch": 6.5120000000000005,
"grad_norm": 0.5644993185997009,
"learning_rate": 1.962030398375506e-06,
"loss": 0.0194,
"step": 814
},
{
"epoch": 6.52,
"grad_norm": 0.4298612177371979,
"learning_rate": 1.950361895908427e-06,
"loss": 0.0157,
"step": 815
},
{
"epoch": 6.5280000000000005,
"grad_norm": 0.8214244246482849,
"learning_rate": 1.9387197813658092e-06,
"loss": 0.0338,
"step": 816
},
{
"epoch": 6.536,
"grad_norm": 0.4823768436908722,
"learning_rate": 1.927104155484602e-06,
"loss": 0.0161,
"step": 817
},
{
"epoch": 6.5440000000000005,
"grad_norm": 0.45921590924263,
"learning_rate": 1.915515118772555e-06,
"loss": 0.0148,
"step": 818
},
{
"epoch": 6.552,
"grad_norm": 0.5129720568656921,
"learning_rate": 1.9039527715073424e-06,
"loss": 0.0163,
"step": 819
},
{
"epoch": 6.5600000000000005,
"grad_norm": 0.38792961835861206,
"learning_rate": 1.8924172137357038e-06,
"loss": 0.0155,
"step": 820
},
{
"epoch": 6.568,
"grad_norm": 0.5734500885009766,
"learning_rate": 1.8809085452725744e-06,
"loss": 0.0195,
"step": 821
},
{
"epoch": 6.576,
"grad_norm": 0.42012128233909607,
"learning_rate": 1.8694268657002197e-06,
"loss": 0.0179,
"step": 822
},
{
"epoch": 6.584,
"grad_norm": 0.6506406664848328,
"learning_rate": 1.8579722743673773e-06,
"loss": 0.0227,
"step": 823
},
{
"epoch": 6.592,
"grad_norm": 0.4907991886138916,
"learning_rate": 1.8465448703883959e-06,
"loss": 0.011,
"step": 824
},
{
"epoch": 6.6,
"grad_norm": 0.4725576639175415,
"learning_rate": 1.8351447526423728e-06,
"loss": 0.0149,
"step": 825
},
{
"epoch": 6.608,
"grad_norm": 0.4036758840084076,
"learning_rate": 1.8237720197723075e-06,
"loss": 0.0184,
"step": 826
},
{
"epoch": 6.616,
"grad_norm": 0.3650066554546356,
"learning_rate": 1.812426770184243e-06,
"loss": 0.015,
"step": 827
},
{
"epoch": 6.624,
"grad_norm": 0.38544753193855286,
"learning_rate": 1.8011091020464138e-06,
"loss": 0.0155,
"step": 828
},
{
"epoch": 6.632,
"grad_norm": 0.444048672914505,
"learning_rate": 1.789819113288397e-06,
"loss": 0.0177,
"step": 829
},
{
"epoch": 6.64,
"grad_norm": 0.5285355448722839,
"learning_rate": 1.7785569016002686e-06,
"loss": 0.0277,
"step": 830
},
{
"epoch": 6.648,
"grad_norm": 0.5375143885612488,
"learning_rate": 1.7673225644317487e-06,
"loss": 0.0203,
"step": 831
},
{
"epoch": 6.656,
"grad_norm": 0.32890117168426514,
"learning_rate": 1.75611619899137e-06,
"loss": 0.0144,
"step": 832
},
{
"epoch": 6.664,
"grad_norm": 0.3891383409500122,
"learning_rate": 1.7449379022456297e-06,
"loss": 0.013,
"step": 833
},
{
"epoch": 6.672,
"grad_norm": 0.4745618402957916,
"learning_rate": 1.7337877709181527e-06,
"loss": 0.0185,
"step": 834
},
{
"epoch": 6.68,
"grad_norm": 0.38062867522239685,
"learning_rate": 1.7226659014888548e-06,
"loss": 0.0102,
"step": 835
},
{
"epoch": 6.688,
"grad_norm": 0.787895917892456,
"learning_rate": 1.711572390193102e-06,
"loss": 0.0302,
"step": 836
},
{
"epoch": 6.696,
"grad_norm": 0.5411037802696228,
"learning_rate": 1.7005073330208881e-06,
"loss": 0.0249,
"step": 837
},
{
"epoch": 6.704,
"grad_norm": 0.4542792737483978,
"learning_rate": 1.689470825715998e-06,
"loss": 0.0112,
"step": 838
},
{
"epoch": 6.712,
"grad_norm": 0.5368099808692932,
"learning_rate": 1.6784629637751814e-06,
"loss": 0.0197,
"step": 839
},
{
"epoch": 6.72,
"grad_norm": 0.3907606899738312,
"learning_rate": 1.6674838424473172e-06,
"loss": 0.0131,
"step": 840
},
{
"epoch": 6.728,
"grad_norm": 0.49019739031791687,
"learning_rate": 1.6565335567326112e-06,
"loss": 0.0216,
"step": 841
},
{
"epoch": 6.736,
"grad_norm": 0.2934483587741852,
"learning_rate": 1.6456122013817477e-06,
"loss": 0.005,
"step": 842
},
{
"epoch": 6.744,
"grad_norm": 0.8866119384765625,
"learning_rate": 1.6347198708950884e-06,
"loss": 0.0194,
"step": 843
},
{
"epoch": 6.752,
"grad_norm": 0.7154141664505005,
"learning_rate": 1.6238566595218475e-06,
"loss": 0.0268,
"step": 844
},
{
"epoch": 6.76,
"grad_norm": 0.3840312361717224,
"learning_rate": 1.6130226612592787e-06,
"loss": 0.0081,
"step": 845
},
{
"epoch": 6.768,
"grad_norm": 0.513756275177002,
"learning_rate": 1.6022179698518525e-06,
"loss": 0.0203,
"step": 846
},
{
"epoch": 6.776,
"grad_norm": 0.7530612349510193,
"learning_rate": 1.591442678790467e-06,
"loss": 0.0494,
"step": 847
},
{
"epoch": 6.784,
"grad_norm": 0.29639124870300293,
"learning_rate": 1.580696881311611e-06,
"loss": 0.0076,
"step": 848
},
{
"epoch": 6.792,
"grad_norm": 0.812698483467102,
"learning_rate": 1.5699806703965787e-06,
"loss": 0.0288,
"step": 849
},
{
"epoch": 6.8,
"grad_norm": 0.47915413975715637,
"learning_rate": 1.5592941387706562e-06,
"loss": 0.0143,
"step": 850
},
{
"epoch": 6.808,
"grad_norm": 0.4331102669239044,
"learning_rate": 1.5486373789023206e-06,
"loss": 0.0149,
"step": 851
},
{
"epoch": 6.816,
"grad_norm": 0.44031015038490295,
"learning_rate": 1.538010483002435e-06,
"loss": 0.0183,
"step": 852
},
{
"epoch": 6.824,
"grad_norm": 0.4399348795413971,
"learning_rate": 1.5274135430234654e-06,
"loss": 0.0181,
"step": 853
},
{
"epoch": 6.832,
"grad_norm": 0.4304785430431366,
"learning_rate": 1.5168466506586654e-06,
"loss": 0.0173,
"step": 854
},
{
"epoch": 6.84,
"grad_norm": 0.5548763871192932,
"learning_rate": 1.506309897341297e-06,
"loss": 0.0295,
"step": 855
},
{
"epoch": 6.848,
"grad_norm": 0.4924432933330536,
"learning_rate": 1.4958033742438348e-06,
"loss": 0.0096,
"step": 856
},
{
"epoch": 6.856,
"grad_norm": 0.4524632394313812,
"learning_rate": 1.4853271722771772e-06,
"loss": 0.0211,
"step": 857
},
{
"epoch": 6.864,
"grad_norm": 0.4187138080596924,
"learning_rate": 1.4748813820898554e-06,
"loss": 0.0179,
"step": 858
},
{
"epoch": 6.872,
"grad_norm": 0.7639076113700867,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.0138,
"step": 859
},
{
"epoch": 6.88,
"grad_norm": 0.7690819501876831,
"learning_rate": 1.454081398330855e-06,
"loss": 0.0202,
"step": 860
},
{
"epoch": 6.888,
"grad_norm": 0.7614538073539734,
"learning_rate": 1.4437273847373778e-06,
"loss": 0.0132,
"step": 861
},
{
"epoch": 6.896,
"grad_norm": 0.49523505568504333,
"learning_rate": 1.4334041428781003e-06,
"loss": 0.0196,
"step": 862
},
{
"epoch": 6.904,
"grad_norm": 0.9732420444488525,
"learning_rate": 1.4231117620780188e-06,
"loss": 0.019,
"step": 863
},
{
"epoch": 6.912,
"grad_norm": 0.42634353041648865,
"learning_rate": 1.4128503313951008e-06,
"loss": 0.0231,
"step": 864
},
{
"epoch": 6.92,
"grad_norm": 0.5485000610351562,
"learning_rate": 1.4026199396195078e-06,
"loss": 0.0207,
"step": 865
},
{
"epoch": 6.928,
"grad_norm": 0.3883606791496277,
"learning_rate": 1.3924206752728282e-06,
"loss": 0.0157,
"step": 866
},
{
"epoch": 6.936,
"grad_norm": 2.5304391384124756,
"learning_rate": 1.3822526266073044e-06,
"loss": 0.0119,
"step": 867
},
{
"epoch": 6.944,
"grad_norm": 0.4002678692340851,
"learning_rate": 1.3721158816050872e-06,
"loss": 0.0158,
"step": 868
},
{
"epoch": 6.952,
"grad_norm": 0.45000985264778137,
"learning_rate": 1.3620105279774532e-06,
"loss": 0.0135,
"step": 869
},
{
"epoch": 6.96,
"grad_norm": 2.6394693851470947,
"learning_rate": 1.3519366531640589e-06,
"loss": 0.0196,
"step": 870
},
{
"epoch": 6.968,
"grad_norm": 0.5044559240341187,
"learning_rate": 1.3418943443321807e-06,
"loss": 0.0145,
"step": 871
},
{
"epoch": 6.976,
"grad_norm": 0.39515820145606995,
"learning_rate": 1.3318836883759634e-06,
"loss": 0.0137,
"step": 872
},
{
"epoch": 6.984,
"grad_norm": 0.3949025273323059,
"learning_rate": 1.3219047719156575e-06,
"loss": 0.0164,
"step": 873
},
{
"epoch": 6.992,
"grad_norm": 0.4184512495994568,
"learning_rate": 1.3119576812968893e-06,
"loss": 0.0185,
"step": 874
},
{
"epoch": 7.0,
"grad_norm": 0.36798545718193054,
"learning_rate": 1.3020425025898926e-06,
"loss": 0.0108,
"step": 875
}
],
"logging_steps": 1.0,
"max_steps": 1125,
"num_input_tokens_seen": 0,
"num_train_epochs": 9,
"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": 7.045822625779548e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}