{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.20390824129141885,
  "eval_steps": 500,
  "global_step": 300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006796941376380628,
      "grad_norm": 9.122925758361816,
      "learning_rate": 0.0,
      "loss": 6.6013,
      "step": 1
    },
    {
      "epoch": 0.0013593882752761257,
      "grad_norm": 9.549845695495605,
      "learning_rate": 6.756756756756758e-07,
      "loss": 6.7658,
      "step": 2
    },
    {
      "epoch": 0.0027187765505522514,
      "grad_norm": 8.209335327148438,
      "learning_rate": 2.0270270270270273e-06,
      "loss": 6.5902,
      "step": 4
    },
    {
      "epoch": 0.0040781648258283775,
      "grad_norm": 6.113947868347168,
      "learning_rate": 3.3783783783783788e-06,
      "loss": 6.5818,
      "step": 6
    },
    {
      "epoch": 0.005437553101104503,
      "grad_norm": 6.703476428985596,
      "learning_rate": 4.72972972972973e-06,
      "loss": 6.5232,
      "step": 8
    },
    {
      "epoch": 0.006796941376380629,
      "grad_norm": 13.405858039855957,
      "learning_rate": 6.081081081081082e-06,
      "loss": 6.5521,
      "step": 10
    },
    {
      "epoch": 0.008156329651656755,
      "grad_norm": 11.226860046386719,
      "learning_rate": 7.432432432432433e-06,
      "loss": 6.584,
      "step": 12
    },
    {
      "epoch": 0.009515717926932881,
      "grad_norm": 7.006751537322998,
      "learning_rate": 8.783783783783785e-06,
      "loss": 6.518,
      "step": 14
    },
    {
      "epoch": 0.010875106202209005,
      "grad_norm": 6.206234455108643,
      "learning_rate": 1.0135135135135136e-05,
      "loss": 6.4523,
      "step": 16
    },
    {
      "epoch": 0.012234494477485132,
      "grad_norm": 3.75468111038208,
      "learning_rate": 1.1486486486486488e-05,
      "loss": 6.4488,
      "step": 18
    },
    {
      "epoch": 0.013593882752761258,
      "grad_norm": 3.5509755611419678,
      "learning_rate": 1.2837837837837838e-05,
      "loss": 6.3353,
      "step": 20
    },
    {
      "epoch": 0.014953271028037384,
      "grad_norm": 2.838531494140625,
      "learning_rate": 1.4189189189189189e-05,
      "loss": 6.2737,
      "step": 22
    },
    {
      "epoch": 0.01631265930331351,
      "grad_norm": 2.849353790283203,
      "learning_rate": 1.554054054054054e-05,
      "loss": 6.2386,
      "step": 24
    },
    {
      "epoch": 0.017672047578589634,
      "grad_norm": 3.192340850830078,
      "learning_rate": 1.6891891891891892e-05,
      "loss": 6.1459,
      "step": 26
    },
    {
      "epoch": 0.019031435853865762,
      "grad_norm": 3.079922914505005,
      "learning_rate": 1.8243243243243244e-05,
      "loss": 6.1783,
      "step": 28
    },
    {
      "epoch": 0.020390824129141887,
      "grad_norm": 3.689027786254883,
      "learning_rate": 1.9594594594594595e-05,
      "loss": 5.9851,
      "step": 30
    },
    {
      "epoch": 0.02175021240441801,
      "grad_norm": 2.39050555229187,
      "learning_rate": 2.0945945945945947e-05,
      "loss": 6.0281,
      "step": 32
    },
    {
      "epoch": 0.02310960067969414,
      "grad_norm": 2.3905773162841797,
      "learning_rate": 2.2297297297297298e-05,
      "loss": 6.0399,
      "step": 34
    },
    {
      "epoch": 0.024468988954970263,
      "grad_norm": 2.676403045654297,
      "learning_rate": 2.364864864864865e-05,
      "loss": 6.0026,
      "step": 36
    },
    {
      "epoch": 0.025828377230246388,
      "grad_norm": 2.220277786254883,
      "learning_rate": 2.5e-05,
      "loss": 5.9469,
      "step": 38
    },
    {
      "epoch": 0.027187765505522515,
      "grad_norm": 3.7453274726867676,
      "learning_rate": 2.635135135135135e-05,
      "loss": 5.804,
      "step": 40
    },
    {
      "epoch": 0.02854715378079864,
      "grad_norm": 4.522032737731934,
      "learning_rate": 2.7702702702702704e-05,
      "loss": 5.8153,
      "step": 42
    },
    {
      "epoch": 0.029906542056074768,
      "grad_norm": 3.07928204536438,
      "learning_rate": 2.9054054054054052e-05,
      "loss": 5.8357,
      "step": 44
    },
    {
      "epoch": 0.031265930331350895,
      "grad_norm": 3.2400898933410645,
      "learning_rate": 3.0405405405405407e-05,
      "loss": 5.8705,
      "step": 46
    },
    {
      "epoch": 0.03262531860662702,
      "grad_norm": 5.057046890258789,
      "learning_rate": 3.175675675675676e-05,
      "loss": 5.6707,
      "step": 48
    },
    {
      "epoch": 0.033984706881903144,
      "grad_norm": 4.462399005889893,
      "learning_rate": 3.310810810810811e-05,
      "loss": 5.7012,
      "step": 50
    },
    {
      "epoch": 0.03534409515717927,
      "grad_norm": 3.095761299133301,
      "learning_rate": 3.445945945945946e-05,
      "loss": 5.6685,
      "step": 52
    },
    {
      "epoch": 0.03670348343245539,
      "grad_norm": 3.478303909301758,
      "learning_rate": 3.581081081081081e-05,
      "loss": 5.6353,
      "step": 54
    },
    {
      "epoch": 0.038062871707731524,
      "grad_norm": 4.6464433670043945,
      "learning_rate": 3.7162162162162165e-05,
      "loss": 5.6277,
      "step": 56
    },
    {
      "epoch": 0.03942225998300765,
      "grad_norm": 4.2293572425842285,
      "learning_rate": 3.851351351351351e-05,
      "loss": 5.5346,
      "step": 58
    },
    {
      "epoch": 0.04078164825828377,
      "grad_norm": 4.188422679901123,
      "learning_rate": 3.986486486486487e-05,
      "loss": 5.5544,
      "step": 60
    },
    {
      "epoch": 0.0421410365335599,
      "grad_norm": 3.0673420429229736,
      "learning_rate": 4.1216216216216216e-05,
      "loss": 5.53,
      "step": 62
    },
    {
      "epoch": 0.04350042480883602,
      "grad_norm": 3.3032662868499756,
      "learning_rate": 4.256756756756757e-05,
      "loss": 5.5605,
      "step": 64
    },
    {
      "epoch": 0.044859813084112146,
      "grad_norm": 3.896825075149536,
      "learning_rate": 4.391891891891892e-05,
      "loss": 5.4221,
      "step": 66
    },
    {
      "epoch": 0.04621920135938828,
      "grad_norm": 4.151010990142822,
      "learning_rate": 4.5270270270270274e-05,
      "loss": 5.3967,
      "step": 68
    },
    {
      "epoch": 0.0475785896346644,
      "grad_norm": 3.938117265701294,
      "learning_rate": 4.662162162162162e-05,
      "loss": 5.4716,
      "step": 70
    },
    {
      "epoch": 0.048937977909940526,
      "grad_norm": 3.1217191219329834,
      "learning_rate": 4.797297297297298e-05,
      "loss": 5.4567,
      "step": 72
    },
    {
      "epoch": 0.05029736618521665,
      "grad_norm": 3.293020725250244,
      "learning_rate": 4.9324324324324325e-05,
      "loss": 5.4291,
      "step": 74
    },
    {
      "epoch": 0.051656754460492775,
      "grad_norm": 3.9366047382354736,
      "learning_rate": 5.067567567567568e-05,
      "loss": 5.378,
      "step": 76
    },
    {
      "epoch": 0.053016142735768906,
      "grad_norm": 4.825038909912109,
      "learning_rate": 5.202702702702703e-05,
      "loss": 5.3462,
      "step": 78
    },
    {
      "epoch": 0.05437553101104503,
      "grad_norm": 4.513136386871338,
      "learning_rate": 5.337837837837838e-05,
      "loss": 5.4209,
      "step": 80
    },
    {
      "epoch": 0.055734919286321155,
      "grad_norm": 4.524239540100098,
      "learning_rate": 5.472972972972973e-05,
      "loss": 5.3715,
      "step": 82
    },
    {
      "epoch": 0.05709430756159728,
      "grad_norm": 5.1905317306518555,
      "learning_rate": 5.6081081081081086e-05,
      "loss": 5.2334,
      "step": 84
    },
    {
      "epoch": 0.058453695836873404,
      "grad_norm": 4.657945156097412,
      "learning_rate": 5.7432432432432434e-05,
      "loss": 5.2899,
      "step": 86
    },
    {
      "epoch": 0.059813084112149535,
      "grad_norm": 3.7982685565948486,
      "learning_rate": 5.878378378378379e-05,
      "loss": 5.2191,
      "step": 88
    },
    {
      "epoch": 0.06117247238742566,
      "grad_norm": 3.5835001468658447,
      "learning_rate": 6.013513513513514e-05,
      "loss": 5.1858,
      "step": 90
    },
    {
      "epoch": 0.06253186066270179,
      "grad_norm": 4.594094276428223,
      "learning_rate": 6.14864864864865e-05,
      "loss": 5.2013,
      "step": 92
    },
    {
      "epoch": 0.06389124893797792,
      "grad_norm": 3.8048019409179688,
      "learning_rate": 6.283783783783784e-05,
      "loss": 5.1493,
      "step": 94
    },
    {
      "epoch": 0.06525063721325404,
      "grad_norm": 3.9920341968536377,
      "learning_rate": 6.41891891891892e-05,
      "loss": 5.0612,
      "step": 96
    },
    {
      "epoch": 0.06661002548853016,
      "grad_norm": 3.4856226444244385,
      "learning_rate": 6.554054054054054e-05,
      "loss": 5.1978,
      "step": 98
    },
    {
      "epoch": 0.06796941376380629,
      "grad_norm": 3.485684871673584,
      "learning_rate": 6.68918918918919e-05,
      "loss": 5.2438,
      "step": 100
    },
    {
      "epoch": 0.06932880203908241,
      "grad_norm": 2.92802095413208,
      "learning_rate": 6.824324324324325e-05,
      "loss": 5.0203,
      "step": 102
    },
    {
      "epoch": 0.07068819031435854,
      "grad_norm": 3.472078561782837,
      "learning_rate": 6.95945945945946e-05,
      "loss": 5.1175,
      "step": 104
    },
    {
      "epoch": 0.07204757858963466,
      "grad_norm": 3.5529918670654297,
      "learning_rate": 7.094594594594594e-05,
      "loss": 5.1952,
      "step": 106
    },
    {
      "epoch": 0.07340696686491079,
      "grad_norm": 5.627261638641357,
      "learning_rate": 7.229729729729731e-05,
      "loss": 5.0469,
      "step": 108
    },
    {
      "epoch": 0.07476635514018691,
      "grad_norm": 4.3943305015563965,
      "learning_rate": 7.364864864864865e-05,
      "loss": 5.0147,
      "step": 110
    },
    {
      "epoch": 0.07612574341546305,
      "grad_norm": 2.405991792678833,
      "learning_rate": 7.500000000000001e-05,
      "loss": 5.0281,
      "step": 112
    },
    {
      "epoch": 0.07748513169073917,
      "grad_norm": 3.361250162124634,
      "learning_rate": 7.635135135135135e-05,
      "loss": 4.9389,
      "step": 114
    },
    {
      "epoch": 0.0788445199660153,
      "grad_norm": 3.5558111667633057,
      "learning_rate": 7.77027027027027e-05,
      "loss": 4.9327,
      "step": 116
    },
    {
      "epoch": 0.08020390824129142,
      "grad_norm": 3.6313676834106445,
      "learning_rate": 7.905405405405406e-05,
      "loss": 4.9751,
      "step": 118
    },
    {
      "epoch": 0.08156329651656755,
      "grad_norm": 3.766629219055176,
      "learning_rate": 8.040540540540541e-05,
      "loss": 4.9362,
      "step": 120
    },
    {
      "epoch": 0.08292268479184367,
      "grad_norm": 3.8239798545837402,
      "learning_rate": 8.175675675675675e-05,
      "loss": 4.9906,
      "step": 122
    },
    {
      "epoch": 0.0842820730671198,
      "grad_norm": 3.1650514602661133,
      "learning_rate": 8.310810810810811e-05,
      "loss": 4.8465,
      "step": 124
    },
    {
      "epoch": 0.08564146134239592,
      "grad_norm": 3.0690271854400635,
      "learning_rate": 8.445945945945946e-05,
      "loss": 4.8147,
      "step": 126
    },
    {
      "epoch": 0.08700084961767204,
      "grad_norm": 3.1290276050567627,
      "learning_rate": 8.581081081081082e-05,
      "loss": 4.8495,
      "step": 128
    },
    {
      "epoch": 0.08836023789294817,
      "grad_norm": 3.1333677768707275,
      "learning_rate": 8.716216216216216e-05,
      "loss": 4.8112,
      "step": 130
    },
    {
      "epoch": 0.08971962616822429,
      "grad_norm": 2.8959381580352783,
      "learning_rate": 8.851351351351352e-05,
      "loss": 4.7989,
      "step": 132
    },
    {
      "epoch": 0.09107901444350043,
      "grad_norm": 2.715139389038086,
      "learning_rate": 8.986486486486487e-05,
      "loss": 4.7689,
      "step": 134
    },
    {
      "epoch": 0.09243840271877656,
      "grad_norm": 2.3525729179382324,
      "learning_rate": 9.121621621621623e-05,
      "loss": 4.7503,
      "step": 136
    },
    {
      "epoch": 0.09379779099405268,
      "grad_norm": 2.5053319931030273,
      "learning_rate": 9.256756756756757e-05,
      "loss": 4.8267,
      "step": 138
    },
    {
      "epoch": 0.0951571792693288,
      "grad_norm": 3.2830920219421387,
      "learning_rate": 9.391891891891892e-05,
      "loss": 4.734,
      "step": 140
    },
    {
      "epoch": 0.09651656754460493,
      "grad_norm": 3.367637872695923,
      "learning_rate": 9.527027027027028e-05,
      "loss": 4.6487,
      "step": 142
    },
    {
      "epoch": 0.09787595581988105,
      "grad_norm": 4.157845973968506,
      "learning_rate": 9.662162162162163e-05,
      "loss": 4.7186,
      "step": 144
    },
    {
      "epoch": 0.09923534409515718,
      "grad_norm": 3.549011707305908,
      "learning_rate": 9.797297297297297e-05,
      "loss": 4.7013,
      "step": 146
    },
    {
      "epoch": 0.1005947323704333,
      "grad_norm": 2.438737392425537,
      "learning_rate": 9.932432432432433e-05,
      "loss": 4.6463,
      "step": 148
    },
    {
      "epoch": 0.10195412064570943,
      "grad_norm": 2.62125301361084,
      "learning_rate": 9.999996843793759e-05,
      "loss": 4.6266,
      "step": 150
    },
    {
      "epoch": 0.10331350892098555,
      "grad_norm": 2.5557775497436523,
      "learning_rate": 9.999971594167742e-05,
      "loss": 4.6659,
      "step": 152
    },
    {
      "epoch": 0.10467289719626169,
      "grad_norm": 2.435065746307373,
      "learning_rate": 9.999921095043215e-05,
      "loss": 4.6833,
      "step": 154
    },
    {
      "epoch": 0.10603228547153781,
      "grad_norm": 2.715564012527466,
      "learning_rate": 9.999845346675197e-05,
      "loss": 4.6256,
      "step": 156
    },
    {
      "epoch": 0.10739167374681394,
      "grad_norm": 2.129850149154663,
      "learning_rate": 9.999744349446207e-05,
      "loss": 4.4834,
      "step": 158
    },
    {
      "epoch": 0.10875106202209006,
      "grad_norm": 2.3702259063720703,
      "learning_rate": 9.99961810386628e-05,
      "loss": 4.5664,
      "step": 160
    },
    {
      "epoch": 0.11011045029736619,
      "grad_norm": 1.9370046854019165,
      "learning_rate": 9.999466610572944e-05,
      "loss": 4.5847,
      "step": 162
    },
    {
      "epoch": 0.11146983857264231,
      "grad_norm": 2.4077095985412598,
      "learning_rate": 9.999289870331232e-05,
      "loss": 4.6685,
      "step": 164
    },
    {
      "epoch": 0.11282922684791843,
      "grad_norm": 3.0132172107696533,
      "learning_rate": 9.999087884033666e-05,
      "loss": 4.5605,
      "step": 166
    },
    {
      "epoch": 0.11418861512319456,
      "grad_norm": 3.6081573963165283,
      "learning_rate": 9.998860652700263e-05,
      "loss": 4.4315,
      "step": 168
    },
    {
      "epoch": 0.11554800339847068,
      "grad_norm": 2.687088966369629,
      "learning_rate": 9.998608177478525e-05,
      "loss": 4.5634,
      "step": 170
    },
    {
      "epoch": 0.11690739167374681,
      "grad_norm": 2.3163015842437744,
      "learning_rate": 9.998330459643437e-05,
      "loss": 4.3725,
      "step": 172
    },
    {
      "epoch": 0.11826677994902295,
      "grad_norm": 2.576303720474243,
      "learning_rate": 9.998027500597451e-05,
      "loss": 4.4502,
      "step": 174
    },
    {
      "epoch": 0.11962616822429907,
      "grad_norm": 3.0173189640045166,
      "learning_rate": 9.997699301870488e-05,
      "loss": 4.4904,
      "step": 176
    },
    {
      "epoch": 0.1209855564995752,
      "grad_norm": 1.8845309019088745,
      "learning_rate": 9.99734586511993e-05,
      "loss": 4.4285,
      "step": 178
    },
    {
      "epoch": 0.12234494477485132,
      "grad_norm": 1.8597114086151123,
      "learning_rate": 9.996967192130606e-05,
      "loss": 4.4114,
      "step": 180
    },
    {
      "epoch": 0.12370433305012744,
      "grad_norm": 1.9403643608093262,
      "learning_rate": 9.996563284814788e-05,
      "loss": 4.3586,
      "step": 182
    },
    {
      "epoch": 0.12506372132540358,
      "grad_norm": 2.1628377437591553,
      "learning_rate": 9.99613414521218e-05,
      "loss": 4.4004,
      "step": 184
    },
    {
      "epoch": 0.1264231096006797,
      "grad_norm": 2.213683843612671,
      "learning_rate": 9.995679775489906e-05,
      "loss": 4.4017,
      "step": 186
    },
    {
      "epoch": 0.12778249787595583,
      "grad_norm": 1.9236798286437988,
      "learning_rate": 9.995200177942499e-05,
      "loss": 4.3356,
      "step": 188
    },
    {
      "epoch": 0.12914188615123195,
      "grad_norm": 2.8310718536376953,
      "learning_rate": 9.994695354991892e-05,
      "loss": 4.2476,
      "step": 190
    },
    {
      "epoch": 0.13050127442650808,
      "grad_norm": 2.613215446472168,
      "learning_rate": 9.994165309187406e-05,
      "loss": 4.4249,
      "step": 192
    },
    {
      "epoch": 0.1318606627017842,
      "grad_norm": 3.2933475971221924,
      "learning_rate": 9.993610043205735e-05,
      "loss": 4.359,
      "step": 194
    },
    {
      "epoch": 0.13322005097706033,
      "grad_norm": 2.660553455352783,
      "learning_rate": 9.993029559850932e-05,
      "loss": 4.3591,
      "step": 196
    },
    {
      "epoch": 0.13457943925233645,
      "grad_norm": 2.223825693130493,
      "learning_rate": 9.992423862054397e-05,
      "loss": 4.2638,
      "step": 198
    },
    {
      "epoch": 0.13593882752761258,
      "grad_norm": 1.6391338109970093,
      "learning_rate": 9.991792952874857e-05,
      "loss": 4.2506,
      "step": 200
    },
    {
      "epoch": 0.1372982158028887,
      "grad_norm": 1.568050742149353,
      "learning_rate": 9.991136835498363e-05,
      "loss": 4.1789,
      "step": 202
    },
    {
      "epoch": 0.13865760407816483,
      "grad_norm": 1.8366698026657104,
      "learning_rate": 9.990455513238257e-05,
      "loss": 4.2361,
      "step": 204
    },
    {
      "epoch": 0.14001699235344095,
      "grad_norm": 2.0478951930999756,
      "learning_rate": 9.98974898953517e-05,
      "loss": 4.2613,
      "step": 206
    },
    {
      "epoch": 0.14137638062871707,
      "grad_norm": 1.7681331634521484,
      "learning_rate": 9.989017267956994e-05,
      "loss": 4.2437,
      "step": 208
    },
    {
      "epoch": 0.1427357689039932,
      "grad_norm": 2.2257468700408936,
      "learning_rate": 9.988260352198872e-05,
      "loss": 4.1724,
      "step": 210
    },
    {
      "epoch": 0.14409515717926932,
      "grad_norm": 1.6590179204940796,
      "learning_rate": 9.987478246083175e-05,
      "loss": 4.1619,
      "step": 212
    },
    {
      "epoch": 0.14545454545454545,
      "grad_norm": 2.029710292816162,
      "learning_rate": 9.986670953559482e-05,
      "loss": 4.2611,
      "step": 214
    },
    {
      "epoch": 0.14681393372982157,
      "grad_norm": 1.7355066537857056,
      "learning_rate": 9.985838478704563e-05,
      "loss": 4.222,
      "step": 216
    },
    {
      "epoch": 0.1481733220050977,
      "grad_norm": 1.91265869140625,
      "learning_rate": 9.984980825722356e-05,
      "loss": 4.0887,
      "step": 218
    },
    {
      "epoch": 0.14953271028037382,
      "grad_norm": 2.1522412300109863,
      "learning_rate": 9.984097998943947e-05,
      "loss": 4.1331,
      "step": 220
    },
    {
      "epoch": 0.15089209855564995,
      "grad_norm": 1.7838095426559448,
      "learning_rate": 9.983190002827546e-05,
      "loss": 4.0928,
      "step": 222
    },
    {
      "epoch": 0.1522514868309261,
      "grad_norm": 1.8782153129577637,
      "learning_rate": 9.982256841958472e-05,
      "loss": 4.2071,
      "step": 224
    },
    {
      "epoch": 0.15361087510620222,
      "grad_norm": 2.179396390914917,
      "learning_rate": 9.981298521049118e-05,
      "loss": 4.0642,
      "step": 226
    },
    {
      "epoch": 0.15497026338147835,
      "grad_norm": 2.1441640853881836,
      "learning_rate": 9.980315044938939e-05,
      "loss": 4.0892,
      "step": 228
    },
    {
      "epoch": 0.15632965165675447,
      "grad_norm": 2.6898701190948486,
      "learning_rate": 9.979306418594417e-05,
      "loss": 4.1155,
      "step": 230
    },
    {
      "epoch": 0.1576890399320306,
      "grad_norm": 2.3028266429901123,
      "learning_rate": 9.97827264710904e-05,
      "loss": 4.1381,
      "step": 232
    },
    {
      "epoch": 0.15904842820730672,
      "grad_norm": 1.8704326152801514,
      "learning_rate": 9.977213735703283e-05,
      "loss": 4.1299,
      "step": 234
    },
    {
      "epoch": 0.16040781648258284,
      "grad_norm": 1.5334903001785278,
      "learning_rate": 9.976129689724574e-05,
      "loss": 4.1585,
      "step": 236
    },
    {
      "epoch": 0.16176720475785897,
      "grad_norm": 1.5391136407852173,
      "learning_rate": 9.975020514647267e-05,
      "loss": 4.0774,
      "step": 238
    },
    {
      "epoch": 0.1631265930331351,
      "grad_norm": 1.731969952583313,
      "learning_rate": 9.973886216072614e-05,
      "loss": 4.1801,
      "step": 240
    },
    {
      "epoch": 0.16448598130841122,
      "grad_norm": 1.355950117111206,
      "learning_rate": 9.972726799728744e-05,
      "loss": 4.1208,
      "step": 242
    },
    {
      "epoch": 0.16584536958368734,
      "grad_norm": 1.6355708837509155,
      "learning_rate": 9.971542271470625e-05,
      "loss": 4.0135,
      "step": 244
    },
    {
      "epoch": 0.16720475785896347,
      "grad_norm": 1.612067461013794,
      "learning_rate": 9.970332637280041e-05,
      "loss": 4.008,
      "step": 246
    },
    {
      "epoch": 0.1685641461342396,
      "grad_norm": 1.5609122514724731,
      "learning_rate": 9.969097903265558e-05,
      "loss": 3.9615,
      "step": 248
    },
    {
      "epoch": 0.16992353440951571,
      "grad_norm": 2.1877589225769043,
      "learning_rate": 9.967838075662495e-05,
      "loss": 4.0187,
      "step": 250
    },
    {
      "epoch": 0.17128292268479184,
      "grad_norm": 2.0836243629455566,
      "learning_rate": 9.966553160832889e-05,
      "loss": 4.0108,
      "step": 252
    },
    {
      "epoch": 0.17264231096006796,
      "grad_norm": 1.8262373208999634,
      "learning_rate": 9.96524316526547e-05,
      "loss": 3.9729,
      "step": 254
    },
    {
      "epoch": 0.1740016992353441,
      "grad_norm": 1.4357279539108276,
      "learning_rate": 9.96390809557562e-05,
      "loss": 3.9418,
      "step": 256
    },
    {
      "epoch": 0.1753610875106202,
      "grad_norm": 1.4747521877288818,
      "learning_rate": 9.962547958505346e-05,
      "loss": 4.0073,
      "step": 258
    },
    {
      "epoch": 0.17672047578589634,
      "grad_norm": 1.5109456777572632,
      "learning_rate": 9.961162760923244e-05,
      "loss": 4.0114,
      "step": 260
    },
    {
      "epoch": 0.17807986406117246,
      "grad_norm": 1.6962803602218628,
      "learning_rate": 9.959752509824462e-05,
      "loss": 3.8997,
      "step": 262
    },
    {
      "epoch": 0.17943925233644858,
      "grad_norm": 1.2874037027359009,
      "learning_rate": 9.958317212330665e-05,
      "loss": 3.9746,
      "step": 264
    },
    {
      "epoch": 0.18079864061172474,
      "grad_norm": 1.4089356660842896,
      "learning_rate": 9.956856875690006e-05,
      "loss": 3.8799,
      "step": 266
    },
    {
      "epoch": 0.18215802888700086,
      "grad_norm": 1.4761899709701538,
      "learning_rate": 9.95537150727708e-05,
      "loss": 3.9084,
      "step": 268
    },
    {
      "epoch": 0.18351741716227699,
      "grad_norm": 1.2963216304779053,
      "learning_rate": 9.953861114592889e-05,
      "loss": 3.884,
      "step": 270
    },
    {
      "epoch": 0.1848768054375531,
      "grad_norm": 1.2376818656921387,
      "learning_rate": 9.952325705264806e-05,
      "loss": 3.9434,
      "step": 272
    },
    {
      "epoch": 0.18623619371282923,
      "grad_norm": 1.6393024921417236,
      "learning_rate": 9.950765287046543e-05,
      "loss": 3.9175,
      "step": 274
    },
    {
      "epoch": 0.18759558198810536,
      "grad_norm": 1.2873233556747437,
      "learning_rate": 9.949179867818099e-05,
      "loss": 3.9513,
      "step": 276
    },
    {
      "epoch": 0.18895497026338148,
      "grad_norm": 1.3314156532287598,
      "learning_rate": 9.947569455585726e-05,
      "loss": 3.9345,
      "step": 278
    },
    {
      "epoch": 0.1903143585386576,
      "grad_norm": 1.392342448234558,
      "learning_rate": 9.945934058481892e-05,
      "loss": 3.8092,
      "step": 280
    },
    {
      "epoch": 0.19167374681393373,
      "grad_norm": 1.4349101781845093,
      "learning_rate": 9.944273684765235e-05,
      "loss": 3.8548,
      "step": 282
    },
    {
      "epoch": 0.19303313508920986,
      "grad_norm": 1.2190157175064087,
      "learning_rate": 9.942588342820521e-05,
      "loss": 3.9121,
      "step": 284
    },
    {
      "epoch": 0.19439252336448598,
      "grad_norm": 1.4537711143493652,
      "learning_rate": 9.94087804115861e-05,
      "loss": 3.8502,
      "step": 286
    },
    {
      "epoch": 0.1957519116397621,
      "grad_norm": 1.6733758449554443,
      "learning_rate": 9.939142788416398e-05,
      "loss": 3.8743,
      "step": 288
    },
    {
      "epoch": 0.19711129991503823,
      "grad_norm": 1.4261025190353394,
      "learning_rate": 9.937382593356793e-05,
      "loss": 3.8947,
      "step": 290
    },
    {
      "epoch": 0.19847068819031435,
      "grad_norm": 1.6536645889282227,
      "learning_rate": 9.93559746486865e-05,
      "loss": 3.9158,
      "step": 292
    },
    {
      "epoch": 0.19983007646559048,
      "grad_norm": 1.71151864528656,
      "learning_rate": 9.933787411966742e-05,
      "loss": 3.8466,
      "step": 294
    },
    {
      "epoch": 0.2011894647408666,
      "grad_norm": 1.8195589780807495,
      "learning_rate": 9.931952443791703e-05,
      "loss": 3.8113,
      "step": 296
    },
    {
      "epoch": 0.20254885301614273,
      "grad_norm": 1.5555843114852905,
      "learning_rate": 9.930092569609996e-05,
      "loss": 3.8505,
      "step": 298
    },
    {
      "epoch": 0.20390824129141885,
      "grad_norm": 1.402797818183899,
      "learning_rate": 9.928207798813849e-05,
      "loss": 3.8856,
      "step": 300
    }
  ],
  "logging_steps": 2,
  "max_steps": 2944,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 300,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.377550336196608e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}