{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 23.053184509277344,
      "learning_rate": 0.0,
      "loss": 3.0474,
      "step": 1
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 37.69414520263672,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 3.2925,
      "step": 2
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 5.540323734283447,
      "learning_rate": 6.666666666666667e-05,
      "loss": 3.0089,
      "step": 3
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 4.970560073852539,
      "learning_rate": 0.0001,
      "loss": 2.6052,
      "step": 4
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 3.1140971183776855,
      "learning_rate": 0.00013333333333333334,
      "loss": 2.1661,
      "step": 5
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 5.0788140296936035,
      "learning_rate": 0.00016666666666666666,
      "loss": 2.7634,
      "step": 6
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 3.610426425933838,
      "learning_rate": 0.0002,
      "loss": 2.5581,
      "step": 7
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 3.3719804286956787,
      "learning_rate": 0.00023333333333333333,
      "loss": 2.3932,
      "step": 8
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 3.6790554523468018,
      "learning_rate": 0.0002666666666666667,
      "loss": 2.7487,
      "step": 9
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 3.152188301086426,
      "learning_rate": 0.0003,
      "loss": 2.5796,
      "step": 10
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 3.6318979263305664,
      "learning_rate": 0.0003333333333333333,
      "loss": 2.6359,
      "step": 11
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 2.668596029281616,
      "learning_rate": 0.00036666666666666667,
      "loss": 2.7351,
      "step": 12
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 5.233928680419922,
      "learning_rate": 0.0004,
      "loss": 2.4318,
      "step": 13
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 3.2564899921417236,
      "learning_rate": 0.00043333333333333337,
      "loss": 2.0882,
      "step": 14
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 3.2959020137786865,
      "learning_rate": 0.00046666666666666666,
      "loss": 2.3181,
      "step": 15
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 3.4378912448883057,
      "learning_rate": 0.0005,
      "loss": 2.7332,
      "step": 16
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 3.6707003116607666,
      "learning_rate": 0.0004998292650357557,
      "loss": 2.5919,
      "step": 17
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 3.861492872238159,
      "learning_rate": 0.0004993172933464471,
      "loss": 2.9812,
      "step": 18
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 5.971181869506836,
      "learning_rate": 0.0004984647842238185,
      "loss": 2.7555,
      "step": 19
    },
    {
      "epoch": 1.0,
      "grad_norm": 4.519949436187744,
      "learning_rate": 0.0004972729020927865,
      "loss": 2.4604,
      "step": 20
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 6.372091770172119,
      "learning_rate": 0.0004957432749209755,
      "loss": 1.9803,
      "step": 21
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 3.0549910068511963,
      "learning_rate": 0.0004938779919951092,
      "loss": 1.7481,
      "step": 22
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 3.166491746902466,
      "learning_rate": 0.0004916796010672969,
      "loss": 1.7108,
      "step": 23
    },
    {
      "epoch": 1.2025316455696202,
      "grad_norm": 5.136909008026123,
      "learning_rate": 0.0004891511048751102,
      "loss": 1.9706,
      "step": 24
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 3.662278413772583,
      "learning_rate": 0.00048629595704020493,
      "loss": 1.451,
      "step": 25
    },
    {
      "epoch": 1.3037974683544304,
      "grad_norm": 3.243025302886963,
      "learning_rate": 0.00048311805735108893,
      "loss": 1.882,
      "step": 26
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 4.813160419464111,
      "learning_rate": 0.0004796217464364808,
      "loss": 1.9157,
      "step": 27
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 3.516997814178467,
      "learning_rate": 0.0004758117998365322,
      "loss": 1.8105,
      "step": 28
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 3.336134672164917,
      "learning_rate": 0.00047169342148001547,
      "loss": 1.7735,
      "step": 29
    },
    {
      "epoch": 1.5063291139240507,
      "grad_norm": 3.748394727706909,
      "learning_rate": 0.0004672722365763821,
      "loss": 1.6359,
      "step": 30
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 4.215877056121826,
      "learning_rate": 0.0004625542839324036,
      "loss": 1.557,
      "step": 31
    },
    {
      "epoch": 1.6075949367088609,
      "grad_norm": 3.181777000427246,
      "learning_rate": 0.00045754600770388763,
      "loss": 1.6716,
      "step": 32
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 5.638424873352051,
      "learning_rate": 0.0004522542485937369,
      "loss": 1.8503,
      "step": 33
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 3.092801570892334,
      "learning_rate": 0.0004466862345083708,
      "loss": 1.7874,
      "step": 34
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 3.7433102130889893,
      "learning_rate": 0.0004408495706852758,
      "loss": 1.8509,
      "step": 35
    },
    {
      "epoch": 1.810126582278481,
      "grad_norm": 5.199512004852295,
      "learning_rate": 0.00043475222930516476,
      "loss": 1.9676,
      "step": 36
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 2.9524500370025635,
      "learning_rate": 0.0004284025386029381,
      "loss": 1.6064,
      "step": 37
    },
    {
      "epoch": 1.9113924050632911,
      "grad_norm": 3.5136239528656006,
      "learning_rate": 0.00042180917149231567,
      "loss": 1.7605,
      "step": 38
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 4.008464813232422,
      "learning_rate": 0.0004149811337196807,
      "loss": 1.6991,
      "step": 39
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.3913097381591797,
      "learning_rate": 0.00040792775156331276,
      "loss": 1.9957,
      "step": 40
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 3.1906797885894775,
      "learning_rate": 0.0004006586590948141,
      "loss": 1.1241,
      "step": 41
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 72.32931518554688,
      "learning_rate": 0.0003931837850201263,
      "loss": 1.0448,
      "step": 42
    },
    {
      "epoch": 2.151898734177215,
      "grad_norm": 8.60446548461914,
      "learning_rate": 0.00038551333911811237,
      "loss": 1.2111,
      "step": 43
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 111.643798828125,
      "learning_rate": 0.00037765779829522674,
      "loss": 1.2316,
      "step": 44
    },
    {
      "epoch": 2.2531645569620253,
      "grad_norm": 69.01380157470703,
      "learning_rate": 0.00036962789227532164,
      "loss": 1.1394,
      "step": 45
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 49.87290573120117,
      "learning_rate": 0.0003614345889441346,
      "loss": 1.5392,
      "step": 46
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 6.338498115539551,
      "learning_rate": 0.0003530890793684759,
      "loss": 1.3349,
      "step": 47
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 3.9022185802459717,
      "learning_rate": 0.0003446027625105776,
      "loss": 1.4879,
      "step": 48
    },
    {
      "epoch": 2.4556962025316453,
      "grad_norm": 3.2077016830444336,
      "learning_rate": 0.00033598722965848206,
      "loss": 1.3891,
      "step": 49
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 3.5117242336273193,
      "learning_rate": 0.00032725424859373687,
      "loss": 1.0252,
      "step": 50
    },
    {
      "epoch": 2.5569620253164556,
      "grad_norm": 2.9290771484375,
      "learning_rate": 0.0003184157475180208,
      "loss": 1.1663,
      "step": 51
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 3.6337857246398926,
      "learning_rate": 0.00030948379876065467,
      "loss": 0.8949,
      "step": 52
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 3.0574092864990234,
      "learning_rate": 0.00030047060228925254,
      "loss": 1.0849,
      "step": 53
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 4.709362983703613,
      "learning_rate": 0.0002913884690460325,
      "loss": 1.5214,
      "step": 54
    },
    {
      "epoch": 2.759493670886076,
      "grad_norm": 2.733558416366577,
      "learning_rate": 0.00028224980413255084,
      "loss": 0.9746,
      "step": 55
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 2.5984580516815186,
      "learning_rate": 0.0002730670898658255,
      "loss": 0.9218,
      "step": 56
    },
    {
      "epoch": 2.8607594936708862,
      "grad_norm": 3.3489749431610107,
      "learning_rate": 0.0002638528687289925,
      "loss": 0.9848,
      "step": 57
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 3.0495173931121826,
      "learning_rate": 0.0002546197262397825,
      "loss": 1.0627,
      "step": 58
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 3.800489664077759,
      "learning_rate": 0.0002453802737602176,
      "loss": 1.193,
      "step": 59
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.3720569610595703,
      "learning_rate": 0.00023614713127100752,
      "loss": 0.5813,
      "step": 60
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 2.65234112739563,
      "learning_rate": 0.00022693291013417452,
      "loss": 0.7218,
      "step": 61
    },
    {
      "epoch": 3.1012658227848102,
      "grad_norm": 1.9882073402404785,
      "learning_rate": 0.00021775019586744925,
      "loss": 0.5474,
      "step": 62
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 2.5399203300476074,
      "learning_rate": 0.0002086115309539675,
      "loss": 0.7906,
      "step": 63
    },
    {
      "epoch": 3.2025316455696204,
      "grad_norm": 1.6655718088150024,
      "learning_rate": 0.0001995293977107475,
      "loss": 0.3726,
      "step": 64
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 2.330378770828247,
      "learning_rate": 0.00019051620123934537,
      "loss": 0.6493,
      "step": 65
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 2.360882043838501,
      "learning_rate": 0.0001815842524819793,
      "loss": 0.6553,
      "step": 66
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 2.726463556289673,
      "learning_rate": 0.00017274575140626317,
      "loss": 0.7101,
      "step": 67
    },
    {
      "epoch": 3.4050632911392404,
      "grad_norm": 2.313028573989868,
      "learning_rate": 0.00016401277034151795,
      "loss": 0.773,
      "step": 68
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 2.5557079315185547,
      "learning_rate": 0.00015539723748942243,
      "loss": 0.7098,
      "step": 69
    },
    {
      "epoch": 3.5063291139240507,
      "grad_norm": 2.11527681350708,
      "learning_rate": 0.00014691092063152418,
      "loss": 0.5224,
      "step": 70
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 2.06489634513855,
      "learning_rate": 0.00013856541105586545,
      "loss": 0.565,
      "step": 71
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 2.3769450187683105,
      "learning_rate": 0.0001303721077246784,
      "loss": 0.5167,
      "step": 72
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 3.1446056365966797,
      "learning_rate": 0.0001223422017047733,
      "loss": 0.738,
      "step": 73
    },
    {
      "epoch": 3.708860759493671,
      "grad_norm": 2.9032044410705566,
      "learning_rate": 0.00011448666088188764,
      "loss": 0.84,
      "step": 74
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 2.6407244205474854,
      "learning_rate": 0.00010681621497987371,
      "loss": 0.55,
      "step": 75
    },
    {
      "epoch": 3.810126582278481,
      "grad_norm": 2.3889331817626953,
      "learning_rate": 9.934134090518593e-05,
      "loss": 0.6036,
      "step": 76
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 2.591595411300659,
      "learning_rate": 9.207224843668733e-05,
      "loss": 0.4689,
      "step": 77
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 2.4760868549346924,
      "learning_rate": 8.50188662803194e-05,
      "loss": 0.4915,
      "step": 78
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 2.930954694747925,
      "learning_rate": 7.819082850768433e-05,
      "loss": 0.8307,
      "step": 79
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.1440188884735107,
      "learning_rate": 7.159746139706194e-05,
      "loss": 0.4358,
      "step": 80
    },
    {
      "epoch": 4.050632911392405,
      "grad_norm": 1.4254531860351562,
      "learning_rate": 6.524777069483526e-05,
      "loss": 0.2303,
      "step": 81
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 2.0634138584136963,
      "learning_rate": 5.9150429314724254e-05,
      "loss": 0.5029,
      "step": 82
    },
    {
      "epoch": 4.151898734177215,
      "grad_norm": 1.7933154106140137,
      "learning_rate": 5.3313765491629194e-05,
      "loss": 0.3326,
      "step": 83
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 1.1411280632019043,
      "learning_rate": 4.7745751406263163e-05,
      "loss": 0.1651,
      "step": 84
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 1.6106775999069214,
      "learning_rate": 4.245399229611238e-05,
      "loss": 0.3276,
      "step": 85
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 1.6101868152618408,
      "learning_rate": 3.7445716067596506e-05,
      "loss": 0.3073,
      "step": 86
    },
    {
      "epoch": 4.3544303797468356,
      "grad_norm": 1.1931700706481934,
      "learning_rate": 3.2727763423617915e-05,
      "loss": 0.1378,
      "step": 87
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 1.7510262727737427,
      "learning_rate": 2.8306578519984528e-05,
      "loss": 0.2117,
      "step": 88
    },
    {
      "epoch": 4.455696202531645,
      "grad_norm": 1.7942403554916382,
      "learning_rate": 2.4188200163467787e-05,
      "loss": 0.1961,
      "step": 89
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 1.7418352365493774,
      "learning_rate": 2.0378253563519245e-05,
      "loss": 0.2915,
      "step": 90
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 1.7192072868347168,
      "learning_rate": 1.6881942648911074e-05,
      "loss": 0.1873,
      "step": 91
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 1.5197887420654297,
      "learning_rate": 1.3704042959795133e-05,
      "loss": 0.3229,
      "step": 92
    },
    {
      "epoch": 4.658227848101266,
      "grad_norm": 1.1745399236679077,
      "learning_rate": 1.0848895124889818e-05,
      "loss": 0.1945,
      "step": 93
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 1.7969136238098145,
      "learning_rate": 8.320398932703144e-06,
      "loss": 0.2555,
      "step": 94
    },
    {
      "epoch": 4.759493670886076,
      "grad_norm": 1.9828176498413086,
      "learning_rate": 6.12200800489085e-06,
      "loss": 0.2931,
      "step": 95
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 1.6430472135543823,
      "learning_rate": 4.256725079024554e-06,
      "loss": 0.1705,
      "step": 96
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 1.5796780586242676,
      "learning_rate": 2.7270979072135106e-06,
      "loss": 0.2818,
      "step": 97
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 1.7523950338363647,
      "learning_rate": 1.5352157761815977e-06,
      "loss": 0.2395,
      "step": 98
    },
    {
      "epoch": 4.962025316455696,
      "grad_norm": 1.9358466863632202,
      "learning_rate": 6.827066535529947e-07,
      "loss": 0.3112,
      "step": 99
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.1313531398773193,
      "learning_rate": 1.7073496424427348e-07,
      "loss": 0.102,
      "step": 100
    }
  ],
  "logging_steps": 1,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 20,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.9265650012307456e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}