narabzad's picture
Add files using upload-large-folder tool
9fd509c verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 375,
"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
}
],
"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": 3.019638269101998e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}