{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.6117247238742566,
  "eval_steps": 500,
  "global_step": 900,
  "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
    },
    {
      "epoch": 0.20526762956669498,
      "grad_norm": 1.33147394657135,
      "learning_rate": 9.926298140921221e-05,
      "loss": 3.8581,
      "step": 302
    },
    {
      "epoch": 0.2066270178419711,
      "grad_norm": 1.1469197273254395,
      "learning_rate": 9.924363605575746e-05,
      "loss": 3.8449,
      "step": 304
    },
    {
      "epoch": 0.20798640611724725,
      "grad_norm": 1.3006025552749634,
      "learning_rate": 9.922404202546691e-05,
      "loss": 3.8268,
      "step": 306
    },
    {
      "epoch": 0.20934579439252338,
      "grad_norm": 1.4287155866622925,
      "learning_rate": 9.9204199417289e-05,
      "loss": 3.7724,
      "step": 308
    },
    {
      "epoch": 0.2107051826677995,
      "grad_norm": 1.5455858707427979,
      "learning_rate": 9.918410833142748e-05,
      "loss": 3.7289,
      "step": 310
    },
    {
      "epoch": 0.21206457094307563,
      "grad_norm": 2.112565517425537,
      "learning_rate": 9.91637688693409e-05,
      "loss": 3.7394,
      "step": 312
    },
    {
      "epoch": 0.21342395921835175,
      "grad_norm": 1.6835887432098389,
      "learning_rate": 9.914318113374208e-05,
      "loss": 3.7487,
      "step": 314
    },
    {
      "epoch": 0.21478334749362787,
      "grad_norm": 1.4296996593475342,
      "learning_rate": 9.912234522859761e-05,
      "loss": 3.7731,
      "step": 316
    },
    {
      "epoch": 0.216142735768904,
      "grad_norm": 1.7142002582550049,
      "learning_rate": 9.910126125912733e-05,
      "loss": 3.8279,
      "step": 318
    },
    {
      "epoch": 0.21750212404418012,
      "grad_norm": 1.5620222091674805,
      "learning_rate": 9.907992933180376e-05,
      "loss": 3.8279,
      "step": 320
    },
    {
      "epoch": 0.21886151231945625,
      "grad_norm": 1.3424922227859497,
      "learning_rate": 9.905834955435162e-05,
      "loss": 3.7903,
      "step": 322
    },
    {
      "epoch": 0.22022090059473237,
      "grad_norm": 1.4565094709396362,
      "learning_rate": 9.903652203574722e-05,
      "loss": 3.793,
      "step": 324
    },
    {
      "epoch": 0.2215802888700085,
      "grad_norm": 1.7999119758605957,
      "learning_rate": 9.901444688621801e-05,
      "loss": 3.7354,
      "step": 326
    },
    {
      "epoch": 0.22293967714528462,
      "grad_norm": 1.4900187253952026,
      "learning_rate": 9.899212421724187e-05,
      "loss": 3.7323,
      "step": 328
    },
    {
      "epoch": 0.22429906542056074,
      "grad_norm": 1.4624853134155273,
      "learning_rate": 9.896955414154669e-05,
      "loss": 3.7207,
      "step": 330
    },
    {
      "epoch": 0.22565845369583687,
      "grad_norm": 1.7633172273635864,
      "learning_rate": 9.894673677310972e-05,
      "loss": 3.7566,
      "step": 332
    },
    {
      "epoch": 0.227017841971113,
      "grad_norm": 1.172234296798706,
      "learning_rate": 9.892367222715709e-05,
      "loss": 3.7376,
      "step": 334
    },
    {
      "epoch": 0.22837723024638912,
      "grad_norm": 1.537023901939392,
      "learning_rate": 9.890036062016306e-05,
      "loss": 3.7157,
      "step": 336
    },
    {
      "epoch": 0.22973661852166524,
      "grad_norm": 1.3012125492095947,
      "learning_rate": 9.887680206984959e-05,
      "loss": 3.6776,
      "step": 338
    },
    {
      "epoch": 0.23109600679694137,
      "grad_norm": 1.1854647397994995,
      "learning_rate": 9.885299669518569e-05,
      "loss": 3.6635,
      "step": 340
    },
    {
      "epoch": 0.2324553950722175,
      "grad_norm": 1.1112992763519287,
      "learning_rate": 9.882894461638676e-05,
      "loss": 3.7341,
      "step": 342
    },
    {
      "epoch": 0.23381478334749362,
      "grad_norm": 1.1130858659744263,
      "learning_rate": 9.88046459549141e-05,
      "loss": 3.6422,
      "step": 344
    },
    {
      "epoch": 0.23517417162276974,
      "grad_norm": 1.441116213798523,
      "learning_rate": 9.878010083347419e-05,
      "loss": 3.6886,
      "step": 346
    },
    {
      "epoch": 0.2365335598980459,
      "grad_norm": 1.3626590967178345,
      "learning_rate": 9.875530937601816e-05,
      "loss": 3.7735,
      "step": 348
    },
    {
      "epoch": 0.23789294817332202,
      "grad_norm": 1.2444162368774414,
      "learning_rate": 9.873027170774109e-05,
      "loss": 3.7312,
      "step": 350
    },
    {
      "epoch": 0.23925233644859814,
      "grad_norm": 1.3234375715255737,
      "learning_rate": 9.87049879550814e-05,
      "loss": 3.748,
      "step": 352
    },
    {
      "epoch": 0.24061172472387427,
      "grad_norm": 1.333979606628418,
      "learning_rate": 9.867945824572024e-05,
      "loss": 3.6207,
      "step": 354
    },
    {
      "epoch": 0.2419711129991504,
      "grad_norm": 1.0207340717315674,
      "learning_rate": 9.865368270858082e-05,
      "loss": 3.7018,
      "step": 356
    },
    {
      "epoch": 0.24333050127442651,
      "grad_norm": 1.098137378692627,
      "learning_rate": 9.862766147382774e-05,
      "loss": 3.6689,
      "step": 358
    },
    {
      "epoch": 0.24468988954970264,
      "grad_norm": 1.1118202209472656,
      "learning_rate": 9.860139467286638e-05,
      "loss": 3.7185,
      "step": 360
    },
    {
      "epoch": 0.24604927782497876,
      "grad_norm": 1.4026211500167847,
      "learning_rate": 9.857488243834219e-05,
      "loss": 3.6949,
      "step": 362
    },
    {
      "epoch": 0.2474086661002549,
      "grad_norm": 1.528132677078247,
      "learning_rate": 9.85481249041401e-05,
      "loss": 3.5872,
      "step": 364
    },
    {
      "epoch": 0.248768054375531,
      "grad_norm": 1.4865642786026,
      "learning_rate": 9.852112220538367e-05,
      "loss": 3.6044,
      "step": 366
    },
    {
      "epoch": 0.25012744265080716,
      "grad_norm": 1.1037031412124634,
      "learning_rate": 9.849387447843467e-05,
      "loss": 3.7614,
      "step": 368
    },
    {
      "epoch": 0.25148683092608326,
      "grad_norm": 1.0702588558197021,
      "learning_rate": 9.846638186089214e-05,
      "loss": 3.6226,
      "step": 370
    },
    {
      "epoch": 0.2528462192013594,
      "grad_norm": 0.970947802066803,
      "learning_rate": 9.843864449159182e-05,
      "loss": 3.6127,
      "step": 372
    },
    {
      "epoch": 0.2542056074766355,
      "grad_norm": 1.1656701564788818,
      "learning_rate": 9.841066251060543e-05,
      "loss": 3.694,
      "step": 374
    },
    {
      "epoch": 0.25556499575191166,
      "grad_norm": 1.2996894121170044,
      "learning_rate": 9.838243605924001e-05,
      "loss": 3.6226,
      "step": 376
    },
    {
      "epoch": 0.25692438402718776,
      "grad_norm": 1.392196536064148,
      "learning_rate": 9.835396528003707e-05,
      "loss": 3.6542,
      "step": 378
    },
    {
      "epoch": 0.2582837723024639,
      "grad_norm": 1.3840879201889038,
      "learning_rate": 9.832525031677205e-05,
      "loss": 3.6416,
      "step": 380
    },
    {
      "epoch": 0.25964316057774,
      "grad_norm": 1.5829066038131714,
      "learning_rate": 9.829629131445342e-05,
      "loss": 3.6383,
      "step": 382
    },
    {
      "epoch": 0.26100254885301616,
      "grad_norm": 1.260533332824707,
      "learning_rate": 9.826708841932209e-05,
      "loss": 3.6034,
      "step": 384
    },
    {
      "epoch": 0.26236193712829226,
      "grad_norm": 1.3130146265029907,
      "learning_rate": 9.823764177885059e-05,
      "loss": 3.5935,
      "step": 386
    },
    {
      "epoch": 0.2637213254035684,
      "grad_norm": 1.4189637899398804,
      "learning_rate": 9.820795154174235e-05,
      "loss": 3.6792,
      "step": 388
    },
    {
      "epoch": 0.2650807136788445,
      "grad_norm": 1.3872414827346802,
      "learning_rate": 9.817801785793092e-05,
      "loss": 3.6149,
      "step": 390
    },
    {
      "epoch": 0.26644010195412066,
      "grad_norm": 1.487898826599121,
      "learning_rate": 9.814784087857927e-05,
      "loss": 3.6161,
      "step": 392
    },
    {
      "epoch": 0.26779949022939675,
      "grad_norm": 1.374002456665039,
      "learning_rate": 9.8117420756079e-05,
      "loss": 3.6394,
      "step": 394
    },
    {
      "epoch": 0.2691588785046729,
      "grad_norm": 0.9476630091667175,
      "learning_rate": 9.808675764404953e-05,
      "loss": 3.5447,
      "step": 396
    },
    {
      "epoch": 0.270518266779949,
      "grad_norm": 1.2160744667053223,
      "learning_rate": 9.805585169733738e-05,
      "loss": 3.552,
      "step": 398
    },
    {
      "epoch": 0.27187765505522515,
      "grad_norm": 1.2432382106781006,
      "learning_rate": 9.802470307201538e-05,
      "loss": 3.5518,
      "step": 400
    },
    {
      "epoch": 0.27323704333050125,
      "grad_norm": 1.0426836013793945,
      "learning_rate": 9.799331192538185e-05,
      "loss": 3.6109,
      "step": 402
    },
    {
      "epoch": 0.2745964316057774,
      "grad_norm": 1.1359163522720337,
      "learning_rate": 9.796167841595986e-05,
      "loss": 3.563,
      "step": 404
    },
    {
      "epoch": 0.2759558198810535,
      "grad_norm": 1.2553869485855103,
      "learning_rate": 9.792980270349633e-05,
      "loss": 3.6323,
      "step": 406
    },
    {
      "epoch": 0.27731520815632965,
      "grad_norm": 1.0180846452713013,
      "learning_rate": 9.789768494896132e-05,
      "loss": 3.4775,
      "step": 408
    },
    {
      "epoch": 0.2786745964316058,
      "grad_norm": 1.2593415975570679,
      "learning_rate": 9.786532531454722e-05,
      "loss": 3.6039,
      "step": 410
    },
    {
      "epoch": 0.2800339847068819,
      "grad_norm": 1.2103174924850464,
      "learning_rate": 9.783272396366784e-05,
      "loss": 3.6399,
      "step": 412
    },
    {
      "epoch": 0.28139337298215805,
      "grad_norm": 1.3131142854690552,
      "learning_rate": 9.77998810609577e-05,
      "loss": 3.6405,
      "step": 414
    },
    {
      "epoch": 0.28275276125743415,
      "grad_norm": 1.4719713926315308,
      "learning_rate": 9.77667967722711e-05,
      "loss": 3.6138,
      "step": 416
    },
    {
      "epoch": 0.2841121495327103,
      "grad_norm": 1.4857118129730225,
      "learning_rate": 9.773347126468128e-05,
      "loss": 3.5811,
      "step": 418
    },
    {
      "epoch": 0.2854715378079864,
      "grad_norm": 1.424742579460144,
      "learning_rate": 9.769990470647974e-05,
      "loss": 3.5766,
      "step": 420
    },
    {
      "epoch": 0.28683092608326255,
      "grad_norm": 1.399685263633728,
      "learning_rate": 9.766609726717515e-05,
      "loss": 3.5816,
      "step": 422
    },
    {
      "epoch": 0.28819031435853865,
      "grad_norm": 1.0692592859268188,
      "learning_rate": 9.763204911749267e-05,
      "loss": 3.5316,
      "step": 424
    },
    {
      "epoch": 0.2895497026338148,
      "grad_norm": 0.9437915682792664,
      "learning_rate": 9.759776042937302e-05,
      "loss": 3.5464,
      "step": 426
    },
    {
      "epoch": 0.2909090909090909,
      "grad_norm": 1.3703209161758423,
      "learning_rate": 9.756323137597159e-05,
      "loss": 3.5578,
      "step": 428
    },
    {
      "epoch": 0.29226847918436705,
      "grad_norm": 1.31071138381958,
      "learning_rate": 9.752846213165767e-05,
      "loss": 3.6392,
      "step": 430
    },
    {
      "epoch": 0.29362786745964314,
      "grad_norm": 0.9742053747177124,
      "learning_rate": 9.749345287201343e-05,
      "loss": 3.5328,
      "step": 432
    },
    {
      "epoch": 0.2949872557349193,
      "grad_norm": 0.9459298253059387,
      "learning_rate": 9.745820377383314e-05,
      "loss": 3.4811,
      "step": 434
    },
    {
      "epoch": 0.2963466440101954,
      "grad_norm": 1.5522956848144531,
      "learning_rate": 9.74227150151222e-05,
      "loss": 3.5842,
      "step": 436
    },
    {
      "epoch": 0.29770603228547154,
      "grad_norm": 1.2024612426757812,
      "learning_rate": 9.738698677509632e-05,
      "loss": 3.5591,
      "step": 438
    },
    {
      "epoch": 0.29906542056074764,
      "grad_norm": 1.2312219142913818,
      "learning_rate": 9.735101923418054e-05,
      "loss": 3.5347,
      "step": 440
    },
    {
      "epoch": 0.3004248088360238,
      "grad_norm": 1.215908169746399,
      "learning_rate": 9.731481257400838e-05,
      "loss": 3.5233,
      "step": 442
    },
    {
      "epoch": 0.3017841971112999,
      "grad_norm": 1.025688886642456,
      "learning_rate": 9.727836697742086e-05,
      "loss": 3.4792,
      "step": 444
    },
    {
      "epoch": 0.30314358538657604,
      "grad_norm": 0.8059235215187073,
      "learning_rate": 9.724168262846566e-05,
      "loss": 3.4884,
      "step": 446
    },
    {
      "epoch": 0.3045029736618522,
      "grad_norm": 1.061319351196289,
      "learning_rate": 9.720475971239609e-05,
      "loss": 3.5429,
      "step": 448
    },
    {
      "epoch": 0.3058623619371283,
      "grad_norm": 1.2992972135543823,
      "learning_rate": 9.716759841567025e-05,
      "loss": 3.4973,
      "step": 450
    },
    {
      "epoch": 0.30722175021240444,
      "grad_norm": 1.206768274307251,
      "learning_rate": 9.713019892595003e-05,
      "loss": 3.5063,
      "step": 452
    },
    {
      "epoch": 0.30858113848768054,
      "grad_norm": 0.9437035918235779,
      "learning_rate": 9.709256143210015e-05,
      "loss": 3.4601,
      "step": 454
    },
    {
      "epoch": 0.3099405267629567,
      "grad_norm": 1.4230656623840332,
      "learning_rate": 9.705468612418727e-05,
      "loss": 3.5248,
      "step": 456
    },
    {
      "epoch": 0.3112999150382328,
      "grad_norm": 1.3039008378982544,
      "learning_rate": 9.701657319347902e-05,
      "loss": 3.3868,
      "step": 458
    },
    {
      "epoch": 0.31265930331350894,
      "grad_norm": 1.0053726434707642,
      "learning_rate": 9.69782228324429e-05,
      "loss": 3.4689,
      "step": 460
    },
    {
      "epoch": 0.31401869158878504,
      "grad_norm": 1.0519505739212036,
      "learning_rate": 9.693963523474554e-05,
      "loss": 3.5104,
      "step": 462
    },
    {
      "epoch": 0.3153780798640612,
      "grad_norm": 1.2941850423812866,
      "learning_rate": 9.690081059525154e-05,
      "loss": 3.4889,
      "step": 464
    },
    {
      "epoch": 0.3167374681393373,
      "grad_norm": 1.2811554670333862,
      "learning_rate": 9.686174911002253e-05,
      "loss": 3.5412,
      "step": 466
    },
    {
      "epoch": 0.31809685641461344,
      "grad_norm": 1.0608913898468018,
      "learning_rate": 9.682245097631622e-05,
      "loss": 3.5063,
      "step": 468
    },
    {
      "epoch": 0.31945624468988953,
      "grad_norm": 0.7853614687919617,
      "learning_rate": 9.678291639258537e-05,
      "loss": 3.5436,
      "step": 470
    },
    {
      "epoch": 0.3208156329651657,
      "grad_norm": 1.3205409049987793,
      "learning_rate": 9.674314555847682e-05,
      "loss": 3.5236,
      "step": 472
    },
    {
      "epoch": 0.3221750212404418,
      "grad_norm": 1.1064730882644653,
      "learning_rate": 9.670313867483041e-05,
      "loss": 3.4756,
      "step": 474
    },
    {
      "epoch": 0.32353440951571794,
      "grad_norm": 0.9121582508087158,
      "learning_rate": 9.666289594367803e-05,
      "loss": 3.5151,
      "step": 476
    },
    {
      "epoch": 0.32489379779099403,
      "grad_norm": 0.9929459691047668,
      "learning_rate": 9.662241756824261e-05,
      "loss": 3.4621,
      "step": 478
    },
    {
      "epoch": 0.3262531860662702,
      "grad_norm": 1.1451283693313599,
      "learning_rate": 9.658170375293703e-05,
      "loss": 3.4464,
      "step": 480
    },
    {
      "epoch": 0.3276125743415463,
      "grad_norm": 1.161496639251709,
      "learning_rate": 9.654075470336317e-05,
      "loss": 3.4189,
      "step": 482
    },
    {
      "epoch": 0.32897196261682243,
      "grad_norm": 1.0276612043380737,
      "learning_rate": 9.649957062631078e-05,
      "loss": 3.4705,
      "step": 484
    },
    {
      "epoch": 0.33033135089209853,
      "grad_norm": 1.061078667640686,
      "learning_rate": 9.645815172975649e-05,
      "loss": 3.4573,
      "step": 486
    },
    {
      "epoch": 0.3316907391673747,
      "grad_norm": 1.14658522605896,
      "learning_rate": 9.641649822286278e-05,
      "loss": 3.4265,
      "step": 488
    },
    {
      "epoch": 0.33305012744265083,
      "grad_norm": 1.1178537607192993,
      "learning_rate": 9.637461031597686e-05,
      "loss": 3.4919,
      "step": 490
    },
    {
      "epoch": 0.33440951571792693,
      "grad_norm": 1.1290613412857056,
      "learning_rate": 9.633248822062968e-05,
      "loss": 3.5057,
      "step": 492
    },
    {
      "epoch": 0.3357689039932031,
      "grad_norm": 0.9476689696311951,
      "learning_rate": 9.629013214953478e-05,
      "loss": 3.4467,
      "step": 494
    },
    {
      "epoch": 0.3371282922684792,
      "grad_norm": 1.0658568143844604,
      "learning_rate": 9.624754231658731e-05,
      "loss": 3.4675,
      "step": 496
    },
    {
      "epoch": 0.33848768054375533,
      "grad_norm": 0.9237107634544373,
      "learning_rate": 9.620471893686287e-05,
      "loss": 3.5161,
      "step": 498
    },
    {
      "epoch": 0.33984706881903143,
      "grad_norm": 1.1166672706604004,
      "learning_rate": 9.616166222661646e-05,
      "loss": 3.4474,
      "step": 500
    },
    {
      "epoch": 0.3412064570943076,
      "grad_norm": 1.3173065185546875,
      "learning_rate": 9.611837240328138e-05,
      "loss": 3.4247,
      "step": 502
    },
    {
      "epoch": 0.3425658453695837,
      "grad_norm": 0.8932580947875977,
      "learning_rate": 9.607484968546813e-05,
      "loss": 3.425,
      "step": 504
    },
    {
      "epoch": 0.34392523364485983,
      "grad_norm": 1.3925460577011108,
      "learning_rate": 9.603109429296333e-05,
      "loss": 3.4579,
      "step": 506
    },
    {
      "epoch": 0.3452846219201359,
      "grad_norm": 1.265743374824524,
      "learning_rate": 9.598710644672859e-05,
      "loss": 3.584,
      "step": 508
    },
    {
      "epoch": 0.3466440101954121,
      "grad_norm": 1.1315666437149048,
      "learning_rate": 9.594288636889936e-05,
      "loss": 3.3599,
      "step": 510
    },
    {
      "epoch": 0.3480033984706882,
      "grad_norm": 1.1435672044754028,
      "learning_rate": 9.589843428278388e-05,
      "loss": 3.4324,
      "step": 512
    },
    {
      "epoch": 0.3493627867459643,
      "grad_norm": 1.0228195190429688,
      "learning_rate": 9.5853750412862e-05,
      "loss": 3.3526,
      "step": 514
    },
    {
      "epoch": 0.3507221750212404,
      "grad_norm": 0.9447354674339294,
      "learning_rate": 9.580883498478406e-05,
      "loss": 3.4243,
      "step": 516
    },
    {
      "epoch": 0.3520815632965166,
      "grad_norm": 0.9552397131919861,
      "learning_rate": 9.576368822536976e-05,
      "loss": 3.3066,
      "step": 518
    },
    {
      "epoch": 0.35344095157179267,
      "grad_norm": 1.1417309045791626,
      "learning_rate": 9.571831036260699e-05,
      "loss": 3.3925,
      "step": 520
    },
    {
      "epoch": 0.3548003398470688,
      "grad_norm": 1.0355608463287354,
      "learning_rate": 9.567270162565073e-05,
      "loss": 3.4224,
      "step": 522
    },
    {
      "epoch": 0.3561597281223449,
      "grad_norm": 0.9213873744010925,
      "learning_rate": 9.562686224482182e-05,
      "loss": 3.4091,
      "step": 524
    },
    {
      "epoch": 0.3575191163976211,
      "grad_norm": 0.97687828540802,
      "learning_rate": 9.558079245160584e-05,
      "loss": 3.3404,
      "step": 526
    },
    {
      "epoch": 0.35887850467289717,
      "grad_norm": 1.0606272220611572,
      "learning_rate": 9.553449247865199e-05,
      "loss": 3.4489,
      "step": 528
    },
    {
      "epoch": 0.3602378929481733,
      "grad_norm": 1.0698935985565186,
      "learning_rate": 9.548796255977175e-05,
      "loss": 3.4235,
      "step": 530
    },
    {
      "epoch": 0.3615972812234495,
      "grad_norm": 1.0898542404174805,
      "learning_rate": 9.544120292993795e-05,
      "loss": 3.4447,
      "step": 532
    },
    {
      "epoch": 0.36295666949872557,
      "grad_norm": 1.1248620748519897,
      "learning_rate": 9.539421382528331e-05,
      "loss": 3.4496,
      "step": 534
    },
    {
      "epoch": 0.3643160577740017,
      "grad_norm": 1.1177901029586792,
      "learning_rate": 9.534699548309948e-05,
      "loss": 3.3291,
      "step": 536
    },
    {
      "epoch": 0.3656754460492778,
      "grad_norm": 0.8826277256011963,
      "learning_rate": 9.529954814183572e-05,
      "loss": 3.4179,
      "step": 538
    },
    {
      "epoch": 0.36703483432455397,
      "grad_norm": 0.8299089074134827,
      "learning_rate": 9.525187204109767e-05,
      "loss": 3.2932,
      "step": 540
    },
    {
      "epoch": 0.36839422259983007,
      "grad_norm": 1.1234545707702637,
      "learning_rate": 9.520396742164624e-05,
      "loss": 3.3519,
      "step": 542
    },
    {
      "epoch": 0.3697536108751062,
      "grad_norm": 0.8834955096244812,
      "learning_rate": 9.515583452539633e-05,
      "loss": 3.3694,
      "step": 544
    },
    {
      "epoch": 0.3711129991503823,
      "grad_norm": 1.1295942068099976,
      "learning_rate": 9.510747359541562e-05,
      "loss": 3.4234,
      "step": 546
    },
    {
      "epoch": 0.37247238742565847,
      "grad_norm": 1.1070072650909424,
      "learning_rate": 9.505888487592333e-05,
      "loss": 3.2804,
      "step": 548
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 0.8992867469787598,
      "learning_rate": 9.501006861228903e-05,
      "loss": 3.3935,
      "step": 550
    },
    {
      "epoch": 0.3751911639762107,
      "grad_norm": 0.9362753033638,
      "learning_rate": 9.496102505103135e-05,
      "loss": 3.4367,
      "step": 552
    },
    {
      "epoch": 0.3765505522514868,
      "grad_norm": 0.997207760810852,
      "learning_rate": 9.491175443981677e-05,
      "loss": 3.3499,
      "step": 554
    },
    {
      "epoch": 0.37790994052676297,
      "grad_norm": 0.9916525483131409,
      "learning_rate": 9.486225702745833e-05,
      "loss": 3.3746,
      "step": 556
    },
    {
      "epoch": 0.37926932880203906,
      "grad_norm": 0.9965975284576416,
      "learning_rate": 9.481253306391445e-05,
      "loss": 3.2899,
      "step": 558
    },
    {
      "epoch": 0.3806287170773152,
      "grad_norm": 0.9485024213790894,
      "learning_rate": 9.476258280028753e-05,
      "loss": 3.2999,
      "step": 560
    },
    {
      "epoch": 0.3819881053525913,
      "grad_norm": 0.9940890669822693,
      "learning_rate": 9.471240648882288e-05,
      "loss": 3.2989,
      "step": 562
    },
    {
      "epoch": 0.38334749362786746,
      "grad_norm": 0.8807222843170166,
      "learning_rate": 9.466200438290724e-05,
      "loss": 3.3442,
      "step": 564
    },
    {
      "epoch": 0.38470688190314356,
      "grad_norm": 0.923343300819397,
      "learning_rate": 9.461137673706768e-05,
      "loss": 3.3407,
      "step": 566
    },
    {
      "epoch": 0.3860662701784197,
      "grad_norm": 1.0074143409729004,
      "learning_rate": 9.456052380697015e-05,
      "loss": 3.3164,
      "step": 568
    },
    {
      "epoch": 0.3874256584536958,
      "grad_norm": 0.9250668883323669,
      "learning_rate": 9.450944584941831e-05,
      "loss": 3.3366,
      "step": 570
    },
    {
      "epoch": 0.38878504672897196,
      "grad_norm": 0.9320377707481384,
      "learning_rate": 9.44581431223522e-05,
      "loss": 3.2337,
      "step": 572
    },
    {
      "epoch": 0.3901444350042481,
      "grad_norm": 0.8707028031349182,
      "learning_rate": 9.440661588484691e-05,
      "loss": 3.3565,
      "step": 574
    },
    {
      "epoch": 0.3915038232795242,
      "grad_norm": 0.9949326515197754,
      "learning_rate": 9.43548643971113e-05,
      "loss": 3.3029,
      "step": 576
    },
    {
      "epoch": 0.39286321155480036,
      "grad_norm": 0.7374237775802612,
      "learning_rate": 9.430288892048666e-05,
      "loss": 3.3016,
      "step": 578
    },
    {
      "epoch": 0.39422259983007646,
      "grad_norm": 1.2260855436325073,
      "learning_rate": 9.425068971744547e-05,
      "loss": 3.4076,
      "step": 580
    },
    {
      "epoch": 0.3955819881053526,
      "grad_norm": 1.042262315750122,
      "learning_rate": 9.419826705158994e-05,
      "loss": 3.4558,
      "step": 582
    },
    {
      "epoch": 0.3969413763806287,
      "grad_norm": 0.9673371911048889,
      "learning_rate": 9.414562118765077e-05,
      "loss": 3.3727,
      "step": 584
    },
    {
      "epoch": 0.39830076465590486,
      "grad_norm": 0.8633317947387695,
      "learning_rate": 9.40927523914858e-05,
      "loss": 3.3029,
      "step": 586
    },
    {
      "epoch": 0.39966015293118096,
      "grad_norm": 1.0391029119491577,
      "learning_rate": 9.40396609300787e-05,
      "loss": 3.3455,
      "step": 588
    },
    {
      "epoch": 0.4010195412064571,
      "grad_norm": 0.9045621752738953,
      "learning_rate": 9.398634707153752e-05,
      "loss": 3.365,
      "step": 590
    },
    {
      "epoch": 0.4023789294817332,
      "grad_norm": 1.0436588525772095,
      "learning_rate": 9.393281108509342e-05,
      "loss": 3.3427,
      "step": 592
    },
    {
      "epoch": 0.40373831775700936,
      "grad_norm": 0.9874547719955444,
      "learning_rate": 9.387905324109934e-05,
      "loss": 3.2496,
      "step": 594
    },
    {
      "epoch": 0.40509770603228545,
      "grad_norm": 0.7234767079353333,
      "learning_rate": 9.382507381102849e-05,
      "loss": 3.2806,
      "step": 596
    },
    {
      "epoch": 0.4064570943075616,
      "grad_norm": 0.7901805639266968,
      "learning_rate": 9.377087306747315e-05,
      "loss": 3.2479,
      "step": 598
    },
    {
      "epoch": 0.4078164825828377,
      "grad_norm": 0.8309746980667114,
      "learning_rate": 9.37164512841432e-05,
      "loss": 3.2532,
      "step": 600
    },
    {
      "epoch": 0.40917587085811385,
      "grad_norm": 0.7645075917243958,
      "learning_rate": 9.366180873586475e-05,
      "loss": 3.3121,
      "step": 602
    },
    {
      "epoch": 0.41053525913338995,
      "grad_norm": 0.8399624824523926,
      "learning_rate": 9.360694569857873e-05,
      "loss": 3.2899,
      "step": 604
    },
    {
      "epoch": 0.4118946474086661,
      "grad_norm": 0.7422559857368469,
      "learning_rate": 9.355186244933959e-05,
      "loss": 3.3512,
      "step": 606
    },
    {
      "epoch": 0.4132540356839422,
      "grad_norm": 0.8472399115562439,
      "learning_rate": 9.349655926631375e-05,
      "loss": 3.3467,
      "step": 608
    },
    {
      "epoch": 0.41461342395921835,
      "grad_norm": 0.745278000831604,
      "learning_rate": 9.344103642877837e-05,
      "loss": 3.2806,
      "step": 610
    },
    {
      "epoch": 0.4159728122344945,
      "grad_norm": 0.9684063196182251,
      "learning_rate": 9.338529421711977e-05,
      "loss": 3.2831,
      "step": 612
    },
    {
      "epoch": 0.4173322005097706,
      "grad_norm": 0.8448832631111145,
      "learning_rate": 9.332933291283215e-05,
      "loss": 3.2381,
      "step": 614
    },
    {
      "epoch": 0.41869158878504675,
      "grad_norm": 0.9104022979736328,
      "learning_rate": 9.327315279851605e-05,
      "loss": 3.3765,
      "step": 616
    },
    {
      "epoch": 0.42005097706032285,
      "grad_norm": 0.9372128248214722,
      "learning_rate": 9.321675415787707e-05,
      "loss": 3.3079,
      "step": 618
    },
    {
      "epoch": 0.421410365335599,
      "grad_norm": 0.8389849066734314,
      "learning_rate": 9.316013727572429e-05,
      "loss": 3.3161,
      "step": 620
    },
    {
      "epoch": 0.4227697536108751,
      "grad_norm": 0.982002854347229,
      "learning_rate": 9.31033024379689e-05,
      "loss": 3.2549,
      "step": 622
    },
    {
      "epoch": 0.42412914188615125,
      "grad_norm": 0.9176488518714905,
      "learning_rate": 9.304624993162276e-05,
      "loss": 3.2436,
      "step": 624
    },
    {
      "epoch": 0.42548853016142735,
      "grad_norm": 0.8668674826622009,
      "learning_rate": 9.298898004479697e-05,
      "loss": 3.2616,
      "step": 626
    },
    {
      "epoch": 0.4268479184367035,
      "grad_norm": 0.9916755557060242,
      "learning_rate": 9.293149306670032e-05,
      "loss": 3.2799,
      "step": 628
    },
    {
      "epoch": 0.4282073067119796,
      "grad_norm": 0.843906819820404,
      "learning_rate": 9.287378928763798e-05,
      "loss": 3.2975,
      "step": 630
    },
    {
      "epoch": 0.42956669498725575,
      "grad_norm": 0.9572488069534302,
      "learning_rate": 9.281586899900985e-05,
      "loss": 3.2525,
      "step": 632
    },
    {
      "epoch": 0.43092608326253184,
      "grad_norm": 0.9162222743034363,
      "learning_rate": 9.275773249330927e-05,
      "loss": 3.3031,
      "step": 634
    },
    {
      "epoch": 0.432285471537808,
      "grad_norm": 1.137613296508789,
      "learning_rate": 9.269938006412142e-05,
      "loss": 3.2803,
      "step": 636
    },
    {
      "epoch": 0.4336448598130841,
      "grad_norm": 0.9166907072067261,
      "learning_rate": 9.26408120061219e-05,
      "loss": 3.3167,
      "step": 638
    },
    {
      "epoch": 0.43500424808836025,
      "grad_norm": 0.8309205174446106,
      "learning_rate": 9.258202861507518e-05,
      "loss": 3.2414,
      "step": 640
    },
    {
      "epoch": 0.43636363636363634,
      "grad_norm": 0.9707709550857544,
      "learning_rate": 9.252303018783324e-05,
      "loss": 3.3043,
      "step": 642
    },
    {
      "epoch": 0.4377230246389125,
      "grad_norm": 0.7566142678260803,
      "learning_rate": 9.246381702233385e-05,
      "loss": 3.2545,
      "step": 644
    },
    {
      "epoch": 0.4390824129141886,
      "grad_norm": 1.1095722913742065,
      "learning_rate": 9.240438941759926e-05,
      "loss": 3.3787,
      "step": 646
    },
    {
      "epoch": 0.44044180118946474,
      "grad_norm": 0.8704729080200195,
      "learning_rate": 9.234474767373465e-05,
      "loss": 3.2712,
      "step": 648
    },
    {
      "epoch": 0.44180118946474084,
      "grad_norm": 0.7083353400230408,
      "learning_rate": 9.228489209192652e-05,
      "loss": 3.3055,
      "step": 650
    },
    {
      "epoch": 0.443160577740017,
      "grad_norm": 0.8583936095237732,
      "learning_rate": 9.222482297444131e-05,
      "loss": 3.2406,
      "step": 652
    },
    {
      "epoch": 0.44451996601529314,
      "grad_norm": 0.9450501203536987,
      "learning_rate": 9.216454062462374e-05,
      "loss": 3.2526,
      "step": 654
    },
    {
      "epoch": 0.44587935429056924,
      "grad_norm": 0.7373863458633423,
      "learning_rate": 9.210404534689536e-05,
      "loss": 3.2554,
      "step": 656
    },
    {
      "epoch": 0.4472387425658454,
      "grad_norm": 1.0712296962738037,
      "learning_rate": 9.2043337446753e-05,
      "loss": 3.3109,
      "step": 658
    },
    {
      "epoch": 0.4485981308411215,
      "grad_norm": 0.8983728885650635,
      "learning_rate": 9.198241723076719e-05,
      "loss": 3.2369,
      "step": 660
    },
    {
      "epoch": 0.44995751911639764,
      "grad_norm": 0.8862040042877197,
      "learning_rate": 9.192128500658068e-05,
      "loss": 3.2532,
      "step": 662
    },
    {
      "epoch": 0.45131690739167374,
      "grad_norm": 0.8523765206336975,
      "learning_rate": 9.185994108290682e-05,
      "loss": 3.2559,
      "step": 664
    },
    {
      "epoch": 0.4526762956669499,
      "grad_norm": 0.9425597190856934,
      "learning_rate": 9.179838576952802e-05,
      "loss": 3.2814,
      "step": 666
    },
    {
      "epoch": 0.454035683942226,
      "grad_norm": 0.7966018915176392,
      "learning_rate": 9.173661937729421e-05,
      "loss": 3.2376,
      "step": 668
    },
    {
      "epoch": 0.45539507221750214,
      "grad_norm": 0.7609128952026367,
      "learning_rate": 9.167464221812126e-05,
      "loss": 3.1981,
      "step": 670
    },
    {
      "epoch": 0.45675446049277824,
      "grad_norm": 0.7252809405326843,
      "learning_rate": 9.161245460498936e-05,
      "loss": 3.3001,
      "step": 672
    },
    {
      "epoch": 0.4581138487680544,
      "grad_norm": 0.5679906606674194,
      "learning_rate": 9.155005685194152e-05,
      "loss": 3.2215,
      "step": 674
    },
    {
      "epoch": 0.4594732370433305,
      "grad_norm": 0.8006565570831299,
      "learning_rate": 9.148744927408193e-05,
      "loss": 3.2696,
      "step": 676
    },
    {
      "epoch": 0.46083262531860664,
      "grad_norm": 0.9909971952438354,
      "learning_rate": 9.142463218757437e-05,
      "loss": 3.3098,
      "step": 678
    },
    {
      "epoch": 0.46219201359388273,
      "grad_norm": 0.729555606842041,
      "learning_rate": 9.136160590964063e-05,
      "loss": 3.188,
      "step": 680
    },
    {
      "epoch": 0.4635514018691589,
      "grad_norm": 0.9540812969207764,
      "learning_rate": 9.129837075855887e-05,
      "loss": 3.2464,
      "step": 682
    },
    {
      "epoch": 0.464910790144435,
      "grad_norm": 0.9037480354309082,
      "learning_rate": 9.123492705366212e-05,
      "loss": 3.3134,
      "step": 684
    },
    {
      "epoch": 0.46627017841971113,
      "grad_norm": 0.7540014386177063,
      "learning_rate": 9.117127511533654e-05,
      "loss": 3.2704,
      "step": 686
    },
    {
      "epoch": 0.46762956669498723,
      "grad_norm": 0.7283887267112732,
      "learning_rate": 9.110741526501982e-05,
      "loss": 3.147,
      "step": 688
    },
    {
      "epoch": 0.4689889549702634,
      "grad_norm": 0.7288716435432434,
      "learning_rate": 9.104334782519969e-05,
      "loss": 3.2257,
      "step": 690
    },
    {
      "epoch": 0.4703483432455395,
      "grad_norm": 0.8213950991630554,
      "learning_rate": 9.097907311941208e-05,
      "loss": 3.1973,
      "step": 692
    },
    {
      "epoch": 0.47170773152081563,
      "grad_norm": 0.773909866809845,
      "learning_rate": 9.091459147223968e-05,
      "loss": 3.2165,
      "step": 694
    },
    {
      "epoch": 0.4730671197960918,
      "grad_norm": 0.9411752820014954,
      "learning_rate": 9.08499032093102e-05,
      "loss": 3.2191,
      "step": 696
    },
    {
      "epoch": 0.4744265080713679,
      "grad_norm": 0.8191618919372559,
      "learning_rate": 9.078500865729471e-05,
      "loss": 3.1823,
      "step": 698
    },
    {
      "epoch": 0.47578589634664403,
      "grad_norm": 0.7489372491836548,
      "learning_rate": 9.071990814390606e-05,
      "loss": 3.1833,
      "step": 700
    },
    {
      "epoch": 0.47714528462192013,
      "grad_norm": 0.7682393193244934,
      "learning_rate": 9.065460199789719e-05,
      "loss": 3.2761,
      "step": 702
    },
    {
      "epoch": 0.4785046728971963,
      "grad_norm": 0.9797720313072205,
      "learning_rate": 9.058909054905946e-05,
      "loss": 3.2743,
      "step": 704
    },
    {
      "epoch": 0.4798640611724724,
      "grad_norm": 0.9919414520263672,
      "learning_rate": 9.052337412822096e-05,
      "loss": 3.2964,
      "step": 706
    },
    {
      "epoch": 0.48122344944774853,
      "grad_norm": 0.8961633443832397,
      "learning_rate": 9.045745306724495e-05,
      "loss": 3.2386,
      "step": 708
    },
    {
      "epoch": 0.4825828377230246,
      "grad_norm": 0.78994220495224,
      "learning_rate": 9.039132769902808e-05,
      "loss": 3.3232,
      "step": 710
    },
    {
      "epoch": 0.4839422259983008,
      "grad_norm": 0.7165075540542603,
      "learning_rate": 9.032499835749868e-05,
      "loss": 3.1898,
      "step": 712
    },
    {
      "epoch": 0.4853016142735769,
      "grad_norm": 0.6712597608566284,
      "learning_rate": 9.02584653776152e-05,
      "loss": 3.2442,
      "step": 714
    },
    {
      "epoch": 0.48666100254885303,
      "grad_norm": 0.7340356707572937,
      "learning_rate": 9.019172909536441e-05,
      "loss": 3.2368,
      "step": 716
    },
    {
      "epoch": 0.4880203908241291,
      "grad_norm": 0.9872474074363708,
      "learning_rate": 9.012478984775977e-05,
      "loss": 3.2627,
      "step": 718
    },
    {
      "epoch": 0.4893797790994053,
      "grad_norm": 0.8943246006965637,
      "learning_rate": 9.005764797283969e-05,
      "loss": 3.1771,
      "step": 720
    },
    {
      "epoch": 0.4907391673746814,
      "grad_norm": 0.8461526036262512,
      "learning_rate": 8.99903038096658e-05,
      "loss": 3.2448,
      "step": 722
    },
    {
      "epoch": 0.4920985556499575,
      "grad_norm": 0.9530143737792969,
      "learning_rate": 8.992275769832135e-05,
      "loss": 3.1837,
      "step": 724
    },
    {
      "epoch": 0.4934579439252336,
      "grad_norm": 0.8631094098091125,
      "learning_rate": 8.985500997990933e-05,
      "loss": 3.2388,
      "step": 726
    },
    {
      "epoch": 0.4948173322005098,
      "grad_norm": 0.7376362681388855,
      "learning_rate": 8.978706099655088e-05,
      "loss": 3.2585,
      "step": 728
    },
    {
      "epoch": 0.49617672047578587,
      "grad_norm": 0.7359983325004578,
      "learning_rate": 8.97189110913835e-05,
      "loss": 3.2278,
      "step": 730
    },
    {
      "epoch": 0.497536108751062,
      "grad_norm": 0.8055249452590942,
      "learning_rate": 8.965056060855931e-05,
      "loss": 3.2128,
      "step": 732
    },
    {
      "epoch": 0.4988954970263382,
      "grad_norm": 0.8565118908882141,
      "learning_rate": 8.95820098932434e-05,
      "loss": 3.2337,
      "step": 734
    },
    {
      "epoch": 0.5002548853016143,
      "grad_norm": 0.7358447313308716,
      "learning_rate": 8.951325929161191e-05,
      "loss": 3.2258,
      "step": 736
    },
    {
      "epoch": 0.5016142735768904,
      "grad_norm": 0.8792275786399841,
      "learning_rate": 8.944430915085051e-05,
      "loss": 3.2875,
      "step": 738
    },
    {
      "epoch": 0.5029736618521665,
      "grad_norm": 0.9467582106590271,
      "learning_rate": 8.937515981915245e-05,
      "loss": 3.2911,
      "step": 740
    },
    {
      "epoch": 0.5043330501274427,
      "grad_norm": 0.7510040998458862,
      "learning_rate": 8.930581164571692e-05,
      "loss": 3.2342,
      "step": 742
    },
    {
      "epoch": 0.5056924384027188,
      "grad_norm": 0.9511517882347107,
      "learning_rate": 8.92362649807472e-05,
      "loss": 3.2036,
      "step": 744
    },
    {
      "epoch": 0.5070518266779949,
      "grad_norm": 0.7870407104492188,
      "learning_rate": 8.916652017544899e-05,
      "loss": 3.2493,
      "step": 746
    },
    {
      "epoch": 0.508411214953271,
      "grad_norm": 0.8817398548126221,
      "learning_rate": 8.909657758202857e-05,
      "loss": 3.1132,
      "step": 748
    },
    {
      "epoch": 0.5097706032285472,
      "grad_norm": 0.8300465941429138,
      "learning_rate": 8.9026437553691e-05,
      "loss": 3.2223,
      "step": 750
    },
    {
      "epoch": 0.5111299915038233,
      "grad_norm": 0.7648670077323914,
      "learning_rate": 8.89561004446384e-05,
      "loss": 3.2196,
      "step": 752
    },
    {
      "epoch": 0.5124893797790994,
      "grad_norm": 0.6600602865219116,
      "learning_rate": 8.888556661006818e-05,
      "loss": 3.1552,
      "step": 754
    },
    {
      "epoch": 0.5138487680543755,
      "grad_norm": 0.7653377652168274,
      "learning_rate": 8.881483640617112e-05,
      "loss": 3.1453,
      "step": 756
    },
    {
      "epoch": 0.5152081563296517,
      "grad_norm": 0.793511688709259,
      "learning_rate": 8.874391019012967e-05,
      "loss": 3.1663,
      "step": 758
    },
    {
      "epoch": 0.5165675446049278,
      "grad_norm": 0.6547530889511108,
      "learning_rate": 8.867278832011617e-05,
      "loss": 3.212,
      "step": 760
    },
    {
      "epoch": 0.5179269328802039,
      "grad_norm": 0.7647628784179688,
      "learning_rate": 8.860147115529096e-05,
      "loss": 3.1953,
      "step": 762
    },
    {
      "epoch": 0.51928632115548,
      "grad_norm": 0.6164153218269348,
      "learning_rate": 8.852995905580063e-05,
      "loss": 3.1791,
      "step": 764
    },
    {
      "epoch": 0.5206457094307562,
      "grad_norm": 0.7801313996315002,
      "learning_rate": 8.845825238277614e-05,
      "loss": 3.1899,
      "step": 766
    },
    {
      "epoch": 0.5220050977060323,
      "grad_norm": 0.8477320075035095,
      "learning_rate": 8.838635149833106e-05,
      "loss": 3.2452,
      "step": 768
    },
    {
      "epoch": 0.5233644859813084,
      "grad_norm": 0.69857257604599,
      "learning_rate": 8.831425676555971e-05,
      "loss": 3.218,
      "step": 770
    },
    {
      "epoch": 0.5247238742565845,
      "grad_norm": 0.8784200549125671,
      "learning_rate": 8.824196854853533e-05,
      "loss": 3.1743,
      "step": 772
    },
    {
      "epoch": 0.5260832625318607,
      "grad_norm": 0.7521477937698364,
      "learning_rate": 8.816948721230822e-05,
      "loss": 3.204,
      "step": 774
    },
    {
      "epoch": 0.5274426508071368,
      "grad_norm": 0.7114652395248413,
      "learning_rate": 8.809681312290397e-05,
      "loss": 3.1783,
      "step": 776
    },
    {
      "epoch": 0.528802039082413,
      "grad_norm": 0.7126603126525879,
      "learning_rate": 8.802394664732152e-05,
      "loss": 3.1708,
      "step": 778
    },
    {
      "epoch": 0.530161427357689,
      "grad_norm": 0.7315563559532166,
      "learning_rate": 8.795088815353135e-05,
      "loss": 3.1273,
      "step": 780
    },
    {
      "epoch": 0.5315208156329652,
      "grad_norm": 0.7203145623207092,
      "learning_rate": 8.78776380104736e-05,
      "loss": 3.1287,
      "step": 782
    },
    {
      "epoch": 0.5328802039082413,
      "grad_norm": 0.7698382139205933,
      "learning_rate": 8.780419658805625e-05,
      "loss": 3.1671,
      "step": 784
    },
    {
      "epoch": 0.5342395921835175,
      "grad_norm": 0.7925199866294861,
      "learning_rate": 8.773056425715326e-05,
      "loss": 3.089,
      "step": 786
    },
    {
      "epoch": 0.5355989804587935,
      "grad_norm": 0.6636077761650085,
      "learning_rate": 8.765674138960261e-05,
      "loss": 3.2514,
      "step": 788
    },
    {
      "epoch": 0.5369583687340697,
      "grad_norm": 0.785798966884613,
      "learning_rate": 8.758272835820451e-05,
      "loss": 3.2389,
      "step": 790
    },
    {
      "epoch": 0.5383177570093458,
      "grad_norm": 0.7094962000846863,
      "learning_rate": 8.750852553671942e-05,
      "loss": 3.2141,
      "step": 792
    },
    {
      "epoch": 0.539677145284622,
      "grad_norm": 0.7799323797225952,
      "learning_rate": 8.743413329986632e-05,
      "loss": 3.1314,
      "step": 794
    },
    {
      "epoch": 0.541036533559898,
      "grad_norm": 0.6692888736724854,
      "learning_rate": 8.735955202332065e-05,
      "loss": 3.1683,
      "step": 796
    },
    {
      "epoch": 0.5423959218351742,
      "grad_norm": 0.7615490555763245,
      "learning_rate": 8.728478208371256e-05,
      "loss": 3.2109,
      "step": 798
    },
    {
      "epoch": 0.5437553101104503,
      "grad_norm": 0.7399694919586182,
      "learning_rate": 8.720982385862483e-05,
      "loss": 3.1026,
      "step": 800
    },
    {
      "epoch": 0.5451146983857265,
      "grad_norm": 0.6580244898796082,
      "learning_rate": 8.713467772659118e-05,
      "loss": 3.1839,
      "step": 802
    },
    {
      "epoch": 0.5464740866610025,
      "grad_norm": 0.7819783091545105,
      "learning_rate": 8.705934406709417e-05,
      "loss": 3.132,
      "step": 804
    },
    {
      "epoch": 0.5478334749362787,
      "grad_norm": 0.7655234336853027,
      "learning_rate": 8.69838232605634e-05,
      "loss": 3.1175,
      "step": 806
    },
    {
      "epoch": 0.5491928632115548,
      "grad_norm": 0.7018226385116577,
      "learning_rate": 8.690811568837352e-05,
      "loss": 3.0879,
      "step": 808
    },
    {
      "epoch": 0.550552251486831,
      "grad_norm": 0.5926535129547119,
      "learning_rate": 8.683222173284236e-05,
      "loss": 3.1408,
      "step": 810
    },
    {
      "epoch": 0.551911639762107,
      "grad_norm": 0.7037495970726013,
      "learning_rate": 8.675614177722895e-05,
      "loss": 3.1749,
      "step": 812
    },
    {
      "epoch": 0.5532710280373832,
      "grad_norm": 0.7136229872703552,
      "learning_rate": 8.667987620573163e-05,
      "loss": 3.1631,
      "step": 814
    },
    {
      "epoch": 0.5546304163126593,
      "grad_norm": 0.7994447946548462,
      "learning_rate": 8.660342540348606e-05,
      "loss": 3.149,
      "step": 816
    },
    {
      "epoch": 0.5559898045879355,
      "grad_norm": 0.5606616735458374,
      "learning_rate": 8.652678975656335e-05,
      "loss": 3.1668,
      "step": 818
    },
    {
      "epoch": 0.5573491928632116,
      "grad_norm": 0.8005346059799194,
      "learning_rate": 8.644996965196797e-05,
      "loss": 3.1792,
      "step": 820
    },
    {
      "epoch": 0.5587085811384876,
      "grad_norm": 0.6956243515014648,
      "learning_rate": 8.637296547763601e-05,
      "loss": 3.1572,
      "step": 822
    },
    {
      "epoch": 0.5600679694137638,
      "grad_norm": 0.7457271218299866,
      "learning_rate": 8.6295777622433e-05,
      "loss": 3.2719,
      "step": 824
    },
    {
      "epoch": 0.56142735768904,
      "grad_norm": 0.7460895776748657,
      "learning_rate": 8.621840647615207e-05,
      "loss": 3.1218,
      "step": 826
    },
    {
      "epoch": 0.5627867459643161,
      "grad_norm": 0.7265388369560242,
      "learning_rate": 8.614085242951201e-05,
      "loss": 3.1595,
      "step": 828
    },
    {
      "epoch": 0.5641461342395921,
      "grad_norm": 0.7856431007385254,
      "learning_rate": 8.606311587415518e-05,
      "loss": 3.16,
      "step": 830
    },
    {
      "epoch": 0.5655055225148683,
      "grad_norm": 0.7534942030906677,
      "learning_rate": 8.598519720264562e-05,
      "loss": 3.1308,
      "step": 832
    },
    {
      "epoch": 0.5668649107901444,
      "grad_norm": 0.6927157640457153,
      "learning_rate": 8.590709680846702e-05,
      "loss": 3.1319,
      "step": 834
    },
    {
      "epoch": 0.5682242990654206,
      "grad_norm": 0.7308480739593506,
      "learning_rate": 8.58288150860208e-05,
      "loss": 3.2368,
      "step": 836
    },
    {
      "epoch": 0.5695836873406966,
      "grad_norm": 0.691619336605072,
      "learning_rate": 8.575035243062407e-05,
      "loss": 3.0438,
      "step": 838
    },
    {
      "epoch": 0.5709430756159728,
      "grad_norm": 0.6600204110145569,
      "learning_rate": 8.567170923850759e-05,
      "loss": 3.1309,
      "step": 840
    },
    {
      "epoch": 0.572302463891249,
      "grad_norm": 0.6632899641990662,
      "learning_rate": 8.559288590681387e-05,
      "loss": 3.0793,
      "step": 842
    },
    {
      "epoch": 0.5736618521665251,
      "grad_norm": 0.7121036052703857,
      "learning_rate": 8.551388283359507e-05,
      "loss": 3.1492,
      "step": 844
    },
    {
      "epoch": 0.5750212404418011,
      "grad_norm": 0.9197178483009338,
      "learning_rate": 8.543470041781107e-05,
      "loss": 3.3029,
      "step": 846
    },
    {
      "epoch": 0.5763806287170773,
      "grad_norm": 0.6388615369796753,
      "learning_rate": 8.535533905932738e-05,
      "loss": 3.169,
      "step": 848
    },
    {
      "epoch": 0.5777400169923534,
      "grad_norm": 0.687680184841156,
      "learning_rate": 8.527579915891321e-05,
      "loss": 3.074,
      "step": 850
    },
    {
      "epoch": 0.5790994052676296,
      "grad_norm": 0.6250393390655518,
      "learning_rate": 8.519608111823931e-05,
      "loss": 3.1493,
      "step": 852
    },
    {
      "epoch": 0.5804587935429056,
      "grad_norm": 0.6668142676353455,
      "learning_rate": 8.511618533987613e-05,
      "loss": 3.236,
      "step": 854
    },
    {
      "epoch": 0.5818181818181818,
      "grad_norm": 0.7983154058456421,
      "learning_rate": 8.50361122272916e-05,
      "loss": 3.1506,
      "step": 856
    },
    {
      "epoch": 0.5831775700934579,
      "grad_norm": 0.8851556181907654,
      "learning_rate": 8.495586218484923e-05,
      "loss": 3.1296,
      "step": 858
    },
    {
      "epoch": 0.5845369583687341,
      "grad_norm": 0.8044039607048035,
      "learning_rate": 8.487543561780595e-05,
      "loss": 3.1169,
      "step": 860
    },
    {
      "epoch": 0.5858963466440102,
      "grad_norm": 0.8117210865020752,
      "learning_rate": 8.479483293231019e-05,
      "loss": 3.1419,
      "step": 862
    },
    {
      "epoch": 0.5872557349192863,
      "grad_norm": 0.7853211164474487,
      "learning_rate": 8.471405453539974e-05,
      "loss": 3.1049,
      "step": 864
    },
    {
      "epoch": 0.5886151231945624,
      "grad_norm": 0.7936228513717651,
      "learning_rate": 8.463310083499971e-05,
      "loss": 3.2174,
      "step": 866
    },
    {
      "epoch": 0.5899745114698386,
      "grad_norm": 0.69145268201828,
      "learning_rate": 8.45519722399205e-05,
      "loss": 3.1685,
      "step": 868
    },
    {
      "epoch": 0.5913338997451147,
      "grad_norm": 0.8211259841918945,
      "learning_rate": 8.447066915985568e-05,
      "loss": 3.1633,
      "step": 870
    },
    {
      "epoch": 0.5926932880203908,
      "grad_norm": 0.7517729997634888,
      "learning_rate": 8.438919200538003e-05,
      "loss": 3.1548,
      "step": 872
    },
    {
      "epoch": 0.5940526762956669,
      "grad_norm": 0.7419262528419495,
      "learning_rate": 8.43075411879473e-05,
      "loss": 3.1094,
      "step": 874
    },
    {
      "epoch": 0.5954120645709431,
      "grad_norm": 0.7104995846748352,
      "learning_rate": 8.42257171198883e-05,
      "loss": 3.1426,
      "step": 876
    },
    {
      "epoch": 0.5967714528462192,
      "grad_norm": 0.6559504270553589,
      "learning_rate": 8.414372021440868e-05,
      "loss": 3.1492,
      "step": 878
    },
    {
      "epoch": 0.5981308411214953,
      "grad_norm": 0.734095573425293,
      "learning_rate": 8.406155088558698e-05,
      "loss": 3.1309,
      "step": 880
    },
    {
      "epoch": 0.5994902293967714,
      "grad_norm": 0.5765071511268616,
      "learning_rate": 8.397920954837242e-05,
      "loss": 3.1223,
      "step": 882
    },
    {
      "epoch": 0.6008496176720476,
      "grad_norm": 0.73897385597229,
      "learning_rate": 8.389669661858284e-05,
      "loss": 3.1529,
      "step": 884
    },
    {
      "epoch": 0.6022090059473237,
      "grad_norm": 0.6814929842948914,
      "learning_rate": 8.381401251290264e-05,
      "loss": 3.0561,
      "step": 886
    },
    {
      "epoch": 0.6035683942225998,
      "grad_norm": 0.730022132396698,
      "learning_rate": 8.373115764888062e-05,
      "loss": 3.0702,
      "step": 888
    },
    {
      "epoch": 0.6049277824978759,
      "grad_norm": 0.6274394392967224,
      "learning_rate": 8.36481324449279e-05,
      "loss": 3.1253,
      "step": 890
    },
    {
      "epoch": 0.6062871707731521,
      "grad_norm": 0.7467242479324341,
      "learning_rate": 8.356493732031586e-05,
      "loss": 3.0619,
      "step": 892
    },
    {
      "epoch": 0.6076465590484282,
      "grad_norm": 0.6906052231788635,
      "learning_rate": 8.34815726951739e-05,
      "loss": 3.2207,
      "step": 894
    },
    {
      "epoch": 0.6090059473237044,
      "grad_norm": 0.7023007273674011,
      "learning_rate": 8.339803899048737e-05,
      "loss": 3.1186,
      "step": 896
    },
    {
      "epoch": 0.6103653355989804,
      "grad_norm": 0.644953727722168,
      "learning_rate": 8.331433662809555e-05,
      "loss": 3.1045,
      "step": 898
    },
    {
      "epoch": 0.6117247238742566,
      "grad_norm": 0.6254658102989197,
      "learning_rate": 8.323046603068934e-05,
      "loss": 3.0299,
      "step": 900
    }
  ],
  "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": 1.0132651008589824e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}