{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 1480,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.033783783783783786,
      "grad_norm": 7.88718318288716,
      "learning_rate": 3.3783783783783788e-06,
      "loss": 1.4134,
      "step": 5
    },
    {
      "epoch": 0.06756756756756757,
      "grad_norm": 3.4257520984949696,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 1.3007,
      "step": 10
    },
    {
      "epoch": 0.10135135135135136,
      "grad_norm": 2.191647863093101,
      "learning_rate": 1.0135135135135136e-05,
      "loss": 1.1684,
      "step": 15
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 2.0684474427551307,
      "learning_rate": 1.3513513513513515e-05,
      "loss": 1.0654,
      "step": 20
    },
    {
      "epoch": 0.16891891891891891,
      "grad_norm": 1.9323299755465573,
      "learning_rate": 1.6891891891891892e-05,
      "loss": 1.0117,
      "step": 25
    },
    {
      "epoch": 0.20270270270270271,
      "grad_norm": 1.7391434518836024,
      "learning_rate": 2.0270270270270273e-05,
      "loss": 0.9346,
      "step": 30
    },
    {
      "epoch": 0.23648648648648649,
      "grad_norm": 1.6145547479084126,
      "learning_rate": 2.364864864864865e-05,
      "loss": 0.9014,
      "step": 35
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 1.5944226046488468,
      "learning_rate": 2.702702702702703e-05,
      "loss": 0.8749,
      "step": 40
    },
    {
      "epoch": 0.30405405405405406,
      "grad_norm": 1.5482362822087288,
      "learning_rate": 3.0405405405405407e-05,
      "loss": 0.8438,
      "step": 45
    },
    {
      "epoch": 0.33783783783783783,
      "grad_norm": 1.6675462048279928,
      "learning_rate": 3.3783783783783784e-05,
      "loss": 0.8441,
      "step": 50
    },
    {
      "epoch": 0.3716216216216216,
      "grad_norm": 1.6457300349477628,
      "learning_rate": 3.7162162162162165e-05,
      "loss": 0.8473,
      "step": 55
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 1.5882803781793335,
      "learning_rate": 4.0540540540540545e-05,
      "loss": 0.8273,
      "step": 60
    },
    {
      "epoch": 0.4391891891891892,
      "grad_norm": 1.5140694587760402,
      "learning_rate": 4.391891891891892e-05,
      "loss": 0.8282,
      "step": 65
    },
    {
      "epoch": 0.47297297297297297,
      "grad_norm": 1.4276802673328162,
      "learning_rate": 4.72972972972973e-05,
      "loss": 0.8152,
      "step": 70
    },
    {
      "epoch": 0.5067567567567568,
      "grad_norm": 1.7535055136362578,
      "learning_rate": 4.999994383297182e-05,
      "loss": 0.847,
      "step": 75
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 1.5044025339968279,
      "learning_rate": 4.9997978016429605e-05,
      "loss": 0.7935,
      "step": 80
    },
    {
      "epoch": 0.5743243243243243,
      "grad_norm": 1.4387848137632269,
      "learning_rate": 4.9993204128893056e-05,
      "loss": 0.8142,
      "step": 85
    },
    {
      "epoch": 0.6081081081081081,
      "grad_norm": 1.3927131640530965,
      "learning_rate": 4.9985622766211935e-05,
      "loss": 0.7988,
      "step": 90
    },
    {
      "epoch": 0.6418918918918919,
      "grad_norm": 1.4019741467378986,
      "learning_rate": 4.997523487464928e-05,
      "loss": 0.7906,
      "step": 95
    },
    {
      "epoch": 0.6756756756756757,
      "grad_norm": 1.428610633837164,
      "learning_rate": 4.996204175076325e-05,
      "loss": 0.7847,
      "step": 100
    },
    {
      "epoch": 0.7094594594594594,
      "grad_norm": 1.3734086318202763,
      "learning_rate": 4.99460450412453e-05,
      "loss": 0.795,
      "step": 105
    },
    {
      "epoch": 0.7432432432432432,
      "grad_norm": 1.3099744937786044,
      "learning_rate": 4.99272467427147e-05,
      "loss": 0.7854,
      "step": 110
    },
    {
      "epoch": 0.777027027027027,
      "grad_norm": 1.2244051450025255,
      "learning_rate": 4.990564920146926e-05,
      "loss": 0.7664,
      "step": 115
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 1.2043425572756492,
      "learning_rate": 4.9881255113192526e-05,
      "loss": 0.7746,
      "step": 120
    },
    {
      "epoch": 0.8445945945945946,
      "grad_norm": 1.2619234476578853,
      "learning_rate": 4.985406752261731e-05,
      "loss": 0.7921,
      "step": 125
    },
    {
      "epoch": 0.8783783783783784,
      "grad_norm": 1.3124758494345672,
      "learning_rate": 4.982408982314565e-05,
      "loss": 0.7719,
      "step": 130
    },
    {
      "epoch": 0.9121621621621622,
      "grad_norm": 1.1738332117725407,
      "learning_rate": 4.9791325756425264e-05,
      "loss": 0.7753,
      "step": 135
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 1.2217000394827666,
      "learning_rate": 4.975577941188258e-05,
      "loss": 0.7563,
      "step": 140
    },
    {
      "epoch": 0.9797297297297297,
      "grad_norm": 1.1904603733924783,
      "learning_rate": 4.971745522621225e-05,
      "loss": 0.7574,
      "step": 145
    },
    {
      "epoch": 1.0135135135135136,
      "grad_norm": 1.4110504819460845,
      "learning_rate": 4.967635798282344e-05,
      "loss": 0.6871,
      "step": 150
    },
    {
      "epoch": 1.0472972972972974,
      "grad_norm": 1.2688773977473746,
      "learning_rate": 4.963249281124278e-05,
      "loss": 0.5628,
      "step": 155
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 1.2184047471515878,
      "learning_rate": 4.958586518647407e-05,
      "loss": 0.5482,
      "step": 160
    },
    {
      "epoch": 1.114864864864865,
      "grad_norm": 1.3477020314378312,
      "learning_rate": 4.953648092831505e-05,
      "loss": 0.5398,
      "step": 165
    },
    {
      "epoch": 1.1486486486486487,
      "grad_norm": 1.2998415055888053,
      "learning_rate": 4.9484346200630855e-05,
      "loss": 0.5502,
      "step": 170
    },
    {
      "epoch": 1.1824324324324325,
      "grad_norm": 1.2026761809862723,
      "learning_rate": 4.9429467510584794e-05,
      "loss": 0.5575,
      "step": 175
    },
    {
      "epoch": 1.2162162162162162,
      "grad_norm": 1.2249292438429276,
      "learning_rate": 4.937185170782607e-05,
      "loss": 0.5455,
      "step": 180
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.1313845209864204,
      "learning_rate": 4.931150598363494e-05,
      "loss": 0.5538,
      "step": 185
    },
    {
      "epoch": 1.2837837837837838,
      "grad_norm": 1.4304332039232173,
      "learning_rate": 4.9248437870025035e-05,
      "loss": 0.5717,
      "step": 190
    },
    {
      "epoch": 1.3175675675675675,
      "grad_norm": 1.182690720837129,
      "learning_rate": 4.9182655238803365e-05,
      "loss": 0.572,
      "step": 195
    },
    {
      "epoch": 1.3513513513513513,
      "grad_norm": 1.2493523187984534,
      "learning_rate": 4.911416630058772e-05,
      "loss": 0.5695,
      "step": 200
    },
    {
      "epoch": 1.385135135135135,
      "grad_norm": 1.179425944327011,
      "learning_rate": 4.9042979603781884e-05,
      "loss": 0.5529,
      "step": 205
    },
    {
      "epoch": 1.4189189189189189,
      "grad_norm": 1.163910491590739,
      "learning_rate": 4.896910403350873e-05,
      "loss": 0.5587,
      "step": 210
    },
    {
      "epoch": 1.4527027027027026,
      "grad_norm": 1.1221216566249026,
      "learning_rate": 4.8892548810501146e-05,
      "loss": 0.5525,
      "step": 215
    },
    {
      "epoch": 1.4864864864864864,
      "grad_norm": 1.1088888471413794,
      "learning_rate": 4.88133234899512e-05,
      "loss": 0.5568,
      "step": 220
    },
    {
      "epoch": 1.5202702702702702,
      "grad_norm": 1.2086677185836778,
      "learning_rate": 4.873143796031752e-05,
      "loss": 0.5662,
      "step": 225
    },
    {
      "epoch": 1.554054054054054,
      "grad_norm": 1.0554325577468497,
      "learning_rate": 4.864690244209105e-05,
      "loss": 0.5743,
      "step": 230
    },
    {
      "epoch": 1.5878378378378377,
      "grad_norm": 1.3999106221209277,
      "learning_rate": 4.855972748651939e-05,
      "loss": 0.5592,
      "step": 235
    },
    {
      "epoch": 1.6216216216216215,
      "grad_norm": 1.2793227178602764,
      "learning_rate": 4.8469923974289874e-05,
      "loss": 0.573,
      "step": 240
    },
    {
      "epoch": 1.6554054054054053,
      "grad_norm": 1.2318194624385215,
      "learning_rate": 4.837750311417146e-05,
      "loss": 0.5725,
      "step": 245
    },
    {
      "epoch": 1.689189189189189,
      "grad_norm": 1.120766793347924,
      "learning_rate": 4.828247644161577e-05,
      "loss": 0.5499,
      "step": 250
    },
    {
      "epoch": 1.722972972972973,
      "grad_norm": 1.0623533741469207,
      "learning_rate": 4.8184855817317226e-05,
      "loss": 0.5625,
      "step": 255
    },
    {
      "epoch": 1.7567567567567568,
      "grad_norm": 1.0753940197543652,
      "learning_rate": 4.808465342573274e-05,
      "loss": 0.5564,
      "step": 260
    },
    {
      "epoch": 1.7905405405405406,
      "grad_norm": 1.0275909563953955,
      "learning_rate": 4.7981881773560886e-05,
      "loss": 0.5798,
      "step": 265
    },
    {
      "epoch": 1.8243243243243243,
      "grad_norm": 1.1046724912105061,
      "learning_rate": 4.787655368818087e-05,
      "loss": 0.5671,
      "step": 270
    },
    {
      "epoch": 1.8581081081081081,
      "grad_norm": 1.135196328730013,
      "learning_rate": 4.77686823160515e-05,
      "loss": 0.57,
      "step": 275
    },
    {
      "epoch": 1.8918918918918919,
      "grad_norm": 1.1328460560052764,
      "learning_rate": 4.765828112107034e-05,
      "loss": 0.5598,
      "step": 280
    },
    {
      "epoch": 1.9256756756756757,
      "grad_norm": 1.057018544001217,
      "learning_rate": 4.75453638828932e-05,
      "loss": 0.5652,
      "step": 285
    },
    {
      "epoch": 1.9594594594594594,
      "grad_norm": 1.0247880097301743,
      "learning_rate": 4.742994469521421e-05,
      "loss": 0.5666,
      "step": 290
    },
    {
      "epoch": 1.9932432432432432,
      "grad_norm": 1.0262429261302537,
      "learning_rate": 4.7312037964006806e-05,
      "loss": 0.5584,
      "step": 295
    },
    {
      "epoch": 2.027027027027027,
      "grad_norm": 1.0741602970862851,
      "learning_rate": 4.719165840572557e-05,
      "loss": 0.3677,
      "step": 300
    },
    {
      "epoch": 2.060810810810811,
      "grad_norm": 1.2516874229222916,
      "learning_rate": 4.7068821045469464e-05,
      "loss": 0.3279,
      "step": 305
    },
    {
      "epoch": 2.0945945945945947,
      "grad_norm": 1.0871177485021244,
      "learning_rate": 4.694354121510644e-05,
      "loss": 0.3068,
      "step": 310
    },
    {
      "epoch": 2.1283783783783785,
      "grad_norm": 1.1397824469739213,
      "learning_rate": 4.681583455135985e-05,
      "loss": 0.3019,
      "step": 315
    },
    {
      "epoch": 2.1621621621621623,
      "grad_norm": 1.1255156379678857,
      "learning_rate": 4.668571699385668e-05,
      "loss": 0.3092,
      "step": 320
    },
    {
      "epoch": 2.195945945945946,
      "grad_norm": 1.1550071128975854,
      "learning_rate": 4.655320478313817e-05,
      "loss": 0.3022,
      "step": 325
    },
    {
      "epoch": 2.22972972972973,
      "grad_norm": 1.1643046015475467,
      "learning_rate": 4.641831445863265e-05,
      "loss": 0.3087,
      "step": 330
    },
    {
      "epoch": 2.2635135135135136,
      "grad_norm": 1.1831197518010665,
      "learning_rate": 4.628106285659124e-05,
      "loss": 0.3118,
      "step": 335
    },
    {
      "epoch": 2.2972972972972974,
      "grad_norm": 1.072743749774166,
      "learning_rate": 4.614146710798645e-05,
      "loss": 0.3122,
      "step": 340
    },
    {
      "epoch": 2.331081081081081,
      "grad_norm": 1.1662799236597767,
      "learning_rate": 4.599954463637394e-05,
      "loss": 0.3133,
      "step": 345
    },
    {
      "epoch": 2.364864864864865,
      "grad_norm": 1.0873311978796223,
      "learning_rate": 4.585531315571788e-05,
      "loss": 0.3159,
      "step": 350
    },
    {
      "epoch": 2.3986486486486487,
      "grad_norm": 1.0778273660741753,
      "learning_rate": 4.570879066817991e-05,
      "loss": 0.317,
      "step": 355
    },
    {
      "epoch": 2.4324324324324325,
      "grad_norm": 1.1320054405466773,
      "learning_rate": 4.555999546187229e-05,
      "loss": 0.3246,
      "step": 360
    },
    {
      "epoch": 2.4662162162162162,
      "grad_norm": 1.083386698244907,
      "learning_rate": 4.5408946108575215e-05,
      "loss": 0.3301,
      "step": 365
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.0471457006769278,
      "learning_rate": 4.5255661461418854e-05,
      "loss": 0.3279,
      "step": 370
    },
    {
      "epoch": 2.5337837837837838,
      "grad_norm": 1.194275932040919,
      "learning_rate": 4.510016065253016e-05,
      "loss": 0.3352,
      "step": 375
    },
    {
      "epoch": 2.5675675675675675,
      "grad_norm": 1.1631565157473671,
      "learning_rate": 4.4942463090644896e-05,
      "loss": 0.3344,
      "step": 380
    },
    {
      "epoch": 2.6013513513513513,
      "grad_norm": 1.081711556346097,
      "learning_rate": 4.478258845868522e-05,
      "loss": 0.3377,
      "step": 385
    },
    {
      "epoch": 2.635135135135135,
      "grad_norm": 1.0805097188404547,
      "learning_rate": 4.462055671130289e-05,
      "loss": 0.33,
      "step": 390
    },
    {
      "epoch": 2.668918918918919,
      "grad_norm": 1.145565895850181,
      "learning_rate": 4.445638807238865e-05,
      "loss": 0.3366,
      "step": 395
    },
    {
      "epoch": 2.7027027027027026,
      "grad_norm": 1.171407891549564,
      "learning_rate": 4.4290103032548094e-05,
      "loss": 0.339,
      "step": 400
    },
    {
      "epoch": 2.7364864864864864,
      "grad_norm": 1.1785865313557242,
      "learning_rate": 4.412172234654399e-05,
      "loss": 0.3304,
      "step": 405
    },
    {
      "epoch": 2.77027027027027,
      "grad_norm": 1.1284728994654754,
      "learning_rate": 4.395126703070589e-05,
      "loss": 0.3408,
      "step": 410
    },
    {
      "epoch": 2.804054054054054,
      "grad_norm": 1.323962121907871,
      "learning_rate": 4.3778758360306986e-05,
      "loss": 0.3308,
      "step": 415
    },
    {
      "epoch": 2.8378378378378377,
      "grad_norm": 1.0966747305504794,
      "learning_rate": 4.360421786690862e-05,
      "loss": 0.3342,
      "step": 420
    },
    {
      "epoch": 2.8716216216216215,
      "grad_norm": 1.0939892147638888,
      "learning_rate": 4.3427667335672815e-05,
      "loss": 0.3362,
      "step": 425
    },
    {
      "epoch": 2.9054054054054053,
      "grad_norm": 1.167765121050444,
      "learning_rate": 4.324912880264326e-05,
      "loss": 0.3434,
      "step": 430
    },
    {
      "epoch": 2.939189189189189,
      "grad_norm": 1.0893669336722744,
      "learning_rate": 4.30686245519948e-05,
      "loss": 0.3359,
      "step": 435
    },
    {
      "epoch": 2.972972972972973,
      "grad_norm": 1.0889900953503875,
      "learning_rate": 4.288617711325207e-05,
      "loss": 0.3371,
      "step": 440
    },
    {
      "epoch": 3.0067567567567566,
      "grad_norm": 0.9033482081348249,
      "learning_rate": 4.2701809258477575e-05,
      "loss": 0.3102,
      "step": 445
    },
    {
      "epoch": 3.0405405405405403,
      "grad_norm": 1.1429246602555858,
      "learning_rate": 4.251554399942928e-05,
      "loss": 0.1664,
      "step": 450
    },
    {
      "epoch": 3.074324324324324,
      "grad_norm": 1.184595266249209,
      "learning_rate": 4.2327404584688495e-05,
      "loss": 0.1576,
      "step": 455
    },
    {
      "epoch": 3.108108108108108,
      "grad_norm": 1.0975552788282228,
      "learning_rate": 4.21374144967581e-05,
      "loss": 0.1592,
      "step": 460
    },
    {
      "epoch": 3.141891891891892,
      "grad_norm": 1.013488196717002,
      "learning_rate": 4.194559744913157e-05,
      "loss": 0.1676,
      "step": 465
    },
    {
      "epoch": 3.175675675675676,
      "grad_norm": 1.0306883925555368,
      "learning_rate": 4.1751977383333224e-05,
      "loss": 0.1632,
      "step": 470
    },
    {
      "epoch": 3.2094594594594597,
      "grad_norm": 1.0008903278259533,
      "learning_rate": 4.1556578465929955e-05,
      "loss": 0.1635,
      "step": 475
    },
    {
      "epoch": 3.2432432432432434,
      "grad_norm": 1.078775253383146,
      "learning_rate": 4.1359425085514906e-05,
      "loss": 0.1666,
      "step": 480
    },
    {
      "epoch": 3.277027027027027,
      "grad_norm": 0.9648675976996084,
      "learning_rate": 4.1160541849663404e-05,
      "loss": 0.1679,
      "step": 485
    },
    {
      "epoch": 3.310810810810811,
      "grad_norm": 1.0220021288871572,
      "learning_rate": 4.095995358186162e-05,
      "loss": 0.1647,
      "step": 490
    },
    {
      "epoch": 3.3445945945945947,
      "grad_norm": 1.0044230404884802,
      "learning_rate": 4.0757685318408186e-05,
      "loss": 0.1692,
      "step": 495
    },
    {
      "epoch": 3.3783783783783785,
      "grad_norm": 1.0332007031632586,
      "learning_rate": 4.055376230528936e-05,
      "loss": 0.1708,
      "step": 500
    },
    {
      "epoch": 3.4121621621621623,
      "grad_norm": 1.0652288806309902,
      "learning_rate": 4.034820999502794e-05,
      "loss": 0.1761,
      "step": 505
    },
    {
      "epoch": 3.445945945945946,
      "grad_norm": 0.9831050553466202,
      "learning_rate": 4.0141054043506406e-05,
      "loss": 0.1776,
      "step": 510
    },
    {
      "epoch": 3.47972972972973,
      "grad_norm": 1.0164952889432723,
      "learning_rate": 3.993232030676473e-05,
      "loss": 0.1698,
      "step": 515
    },
    {
      "epoch": 3.5135135135135136,
      "grad_norm": 1.0260487525381927,
      "learning_rate": 3.972203483777315e-05,
      "loss": 0.1765,
      "step": 520
    },
    {
      "epoch": 3.5472972972972974,
      "grad_norm": 1.0117156887089442,
      "learning_rate": 3.95102238831804e-05,
      "loss": 0.1772,
      "step": 525
    },
    {
      "epoch": 3.581081081081081,
      "grad_norm": 1.0260708971757766,
      "learning_rate": 3.929691388003772e-05,
      "loss": 0.1779,
      "step": 530
    },
    {
      "epoch": 3.614864864864865,
      "grad_norm": 1.015688481709973,
      "learning_rate": 3.908213145249916e-05,
      "loss": 0.178,
      "step": 535
    },
    {
      "epoch": 3.6486486486486487,
      "grad_norm": 0.9753179528236033,
      "learning_rate": 3.886590340849852e-05,
      "loss": 0.1815,
      "step": 540
    },
    {
      "epoch": 3.6824324324324325,
      "grad_norm": 0.9844555678681653,
      "learning_rate": 3.864825673640326e-05,
      "loss": 0.1815,
      "step": 545
    },
    {
      "epoch": 3.7162162162162162,
      "grad_norm": 1.048277705372721,
      "learning_rate": 3.842921860164607e-05,
      "loss": 0.1806,
      "step": 550
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.944833986198349,
      "learning_rate": 3.8208816343334156e-05,
      "loss": 0.1779,
      "step": 555
    },
    {
      "epoch": 3.7837837837837838,
      "grad_norm": 1.040289500445584,
      "learning_rate": 3.798707747083694e-05,
      "loss": 0.1834,
      "step": 560
    },
    {
      "epoch": 3.8175675675675675,
      "grad_norm": 1.0148235826363465,
      "learning_rate": 3.776402966035251e-05,
      "loss": 0.1813,
      "step": 565
    },
    {
      "epoch": 3.8513513513513513,
      "grad_norm": 1.0354290160169566,
      "learning_rate": 3.753970075145322e-05,
      "loss": 0.1873,
      "step": 570
    },
    {
      "epoch": 3.885135135135135,
      "grad_norm": 0.9811743323945232,
      "learning_rate": 3.731411874361094e-05,
      "loss": 0.1816,
      "step": 575
    },
    {
      "epoch": 3.918918918918919,
      "grad_norm": 1.0592857150968156,
      "learning_rate": 3.7087311792702265e-05,
      "loss": 0.1845,
      "step": 580
    },
    {
      "epoch": 3.9527027027027026,
      "grad_norm": 1.0208471032409192,
      "learning_rate": 3.685930820749433e-05,
      "loss": 0.1845,
      "step": 585
    },
    {
      "epoch": 3.9864864864864864,
      "grad_norm": 0.9595882251320016,
      "learning_rate": 3.663013644611139e-05,
      "loss": 0.1838,
      "step": 590
    },
    {
      "epoch": 4.02027027027027,
      "grad_norm": 0.7755330080020648,
      "learning_rate": 3.639982511248289e-05,
      "loss": 0.1227,
      "step": 595
    },
    {
      "epoch": 4.054054054054054,
      "grad_norm": 1.071486984932284,
      "learning_rate": 3.616840295277328e-05,
      "loss": 0.0858,
      "step": 600
    },
    {
      "epoch": 4.087837837837838,
      "grad_norm": 0.9888149620846987,
      "learning_rate": 3.593589885179405e-05,
      "loss": 0.1026,
      "step": 605
    },
    {
      "epoch": 4.121621621621622,
      "grad_norm": 0.8798397434906335,
      "learning_rate": 3.5702341829398525e-05,
      "loss": 0.0893,
      "step": 610
    },
    {
      "epoch": 4.155405405405405,
      "grad_norm": 0.8340558360563065,
      "learning_rate": 3.5467761036859736e-05,
      "loss": 0.088,
      "step": 615
    },
    {
      "epoch": 4.1891891891891895,
      "grad_norm": 0.8542849914021129,
      "learning_rate": 3.523218575323198e-05,
      "loss": 0.0889,
      "step": 620
    },
    {
      "epoch": 4.222972972972973,
      "grad_norm": 0.7824925632671172,
      "learning_rate": 3.499564538169629e-05,
      "loss": 0.0878,
      "step": 625
    },
    {
      "epoch": 4.256756756756757,
      "grad_norm": 0.8320323585811136,
      "learning_rate": 3.475816944589058e-05,
      "loss": 0.085,
      "step": 630
    },
    {
      "epoch": 4.29054054054054,
      "grad_norm": 0.8910368815999778,
      "learning_rate": 3.451978758622458e-05,
      "loss": 0.0867,
      "step": 635
    },
    {
      "epoch": 4.324324324324325,
      "grad_norm": 0.7680094672237836,
      "learning_rate": 3.4280529556180404e-05,
      "loss": 0.0843,
      "step": 640
    },
    {
      "epoch": 4.358108108108108,
      "grad_norm": 0.7876976983343839,
      "learning_rate": 3.4040425218598755e-05,
      "loss": 0.082,
      "step": 645
    },
    {
      "epoch": 4.391891891891892,
      "grad_norm": 0.7806203325513844,
      "learning_rate": 3.379950454195172e-05,
      "loss": 0.0858,
      "step": 650
    },
    {
      "epoch": 4.425675675675675,
      "grad_norm": 0.8497300133325525,
      "learning_rate": 3.355779759660223e-05,
      "loss": 0.0865,
      "step": 655
    },
    {
      "epoch": 4.45945945945946,
      "grad_norm": 0.8045892464140774,
      "learning_rate": 3.331533455105084e-05,
      "loss": 0.0884,
      "step": 660
    },
    {
      "epoch": 4.493243243243243,
      "grad_norm": 0.8043812626087955,
      "learning_rate": 3.307214566817027e-05,
      "loss": 0.0873,
      "step": 665
    },
    {
      "epoch": 4.527027027027027,
      "grad_norm": 0.818456912899916,
      "learning_rate": 3.2828261301428206e-05,
      "loss": 0.0826,
      "step": 670
    },
    {
      "epoch": 4.5608108108108105,
      "grad_norm": 0.8709358249179351,
      "learning_rate": 3.2583711891098665e-05,
      "loss": 0.0853,
      "step": 675
    },
    {
      "epoch": 4.594594594594595,
      "grad_norm": 1.0291242064597803,
      "learning_rate": 3.23385279604627e-05,
      "loss": 0.0884,
      "step": 680
    },
    {
      "epoch": 4.628378378378378,
      "grad_norm": 0.8178096901426156,
      "learning_rate": 3.209274011199861e-05,
      "loss": 0.0885,
      "step": 685
    },
    {
      "epoch": 4.662162162162162,
      "grad_norm": 0.820260682103539,
      "learning_rate": 3.18463790235623e-05,
      "loss": 0.0857,
      "step": 690
    },
    {
      "epoch": 4.695945945945946,
      "grad_norm": 0.8436781864137983,
      "learning_rate": 3.159947544455828e-05,
      "loss": 0.0888,
      "step": 695
    },
    {
      "epoch": 4.72972972972973,
      "grad_norm": 0.812296374311864,
      "learning_rate": 3.135206019210167e-05,
      "loss": 0.09,
      "step": 700
    },
    {
      "epoch": 4.763513513513513,
      "grad_norm": 0.8198887539752774,
      "learning_rate": 3.110416414717181e-05,
      "loss": 0.0838,
      "step": 705
    },
    {
      "epoch": 4.797297297297297,
      "grad_norm": 0.8397920306499081,
      "learning_rate": 3.085581825075782e-05,
      "loss": 0.0872,
      "step": 710
    },
    {
      "epoch": 4.831081081081081,
      "grad_norm": 0.8535599719695989,
      "learning_rate": 3.060705349999677e-05,
      "loss": 0.0863,
      "step": 715
    },
    {
      "epoch": 4.864864864864865,
      "grad_norm": 0.7806619662076235,
      "learning_rate": 3.0357900944304774e-05,
      "loss": 0.0848,
      "step": 720
    },
    {
      "epoch": 4.898648648648649,
      "grad_norm": 0.8367795703413348,
      "learning_rate": 3.0108391681501564e-05,
      "loss": 0.0864,
      "step": 725
    },
    {
      "epoch": 4.9324324324324325,
      "grad_norm": 0.8489755117918106,
      "learning_rate": 2.9858556853929048e-05,
      "loss": 0.0873,
      "step": 730
    },
    {
      "epoch": 4.966216216216216,
      "grad_norm": 0.8697546963962745,
      "learning_rate": 2.96084276445643e-05,
      "loss": 0.0882,
      "step": 735
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.9273587991097837,
      "learning_rate": 2.9358035273127483e-05,
      "loss": 0.0841,
      "step": 740
    },
    {
      "epoch": 5.033783783783784,
      "grad_norm": 0.5921650943795248,
      "learning_rate": 2.910741099218514e-05,
      "loss": 0.0413,
      "step": 745
    },
    {
      "epoch": 5.0675675675675675,
      "grad_norm": 0.5907114773407223,
      "learning_rate": 2.8856586083249487e-05,
      "loss": 0.0403,
      "step": 750
    },
    {
      "epoch": 5.101351351351352,
      "grad_norm": 0.6392965483185691,
      "learning_rate": 2.860559185287397e-05,
      "loss": 0.0428,
      "step": 755
    },
    {
      "epoch": 5.135135135135135,
      "grad_norm": 0.615093395874597,
      "learning_rate": 2.83544596287458e-05,
      "loss": 0.0407,
      "step": 760
    },
    {
      "epoch": 5.168918918918919,
      "grad_norm": 0.5517225991486159,
      "learning_rate": 2.8103220755775776e-05,
      "loss": 0.0407,
      "step": 765
    },
    {
      "epoch": 5.202702702702703,
      "grad_norm": 0.6527163748976097,
      "learning_rate": 2.785190659218604e-05,
      "loss": 0.0412,
      "step": 770
    },
    {
      "epoch": 5.236486486486487,
      "grad_norm": 0.7026500987111242,
      "learning_rate": 2.760054850559603e-05,
      "loss": 0.0395,
      "step": 775
    },
    {
      "epoch": 5.27027027027027,
      "grad_norm": 0.6630081178017392,
      "learning_rate": 2.7349177869107462e-05,
      "loss": 0.0414,
      "step": 780
    },
    {
      "epoch": 5.304054054054054,
      "grad_norm": 0.6042837378418635,
      "learning_rate": 2.709782605738842e-05,
      "loss": 0.0376,
      "step": 785
    },
    {
      "epoch": 5.337837837837838,
      "grad_norm": 0.5784226647104688,
      "learning_rate": 2.684652444275741e-05,
      "loss": 0.0375,
      "step": 790
    },
    {
      "epoch": 5.371621621621622,
      "grad_norm": 0.5793156044017487,
      "learning_rate": 2.6595304391267605e-05,
      "loss": 0.0399,
      "step": 795
    },
    {
      "epoch": 5.405405405405405,
      "grad_norm": 0.5633550607829896,
      "learning_rate": 2.634419725879193e-05,
      "loss": 0.0369,
      "step": 800
    },
    {
      "epoch": 5.4391891891891895,
      "grad_norm": 0.5743268844062026,
      "learning_rate": 2.60932343871094e-05,
      "loss": 0.0383,
      "step": 805
    },
    {
      "epoch": 5.472972972972973,
      "grad_norm": 0.5540015085887268,
      "learning_rate": 2.58424470999932e-05,
      "loss": 0.038,
      "step": 810
    },
    {
      "epoch": 5.506756756756757,
      "grad_norm": 0.6197385576025258,
      "learning_rate": 2.5591866699301055e-05,
      "loss": 0.0382,
      "step": 815
    },
    {
      "epoch": 5.54054054054054,
      "grad_norm": 0.6051779134752197,
      "learning_rate": 2.534152446106825e-05,
      "loss": 0.0406,
      "step": 820
    },
    {
      "epoch": 5.574324324324325,
      "grad_norm": 0.5933052010470478,
      "learning_rate": 2.5091451631604033e-05,
      "loss": 0.0369,
      "step": 825
    },
    {
      "epoch": 5.608108108108108,
      "grad_norm": 0.5910320454195225,
      "learning_rate": 2.4841679423591523e-05,
      "loss": 0.0375,
      "step": 830
    },
    {
      "epoch": 5.641891891891892,
      "grad_norm": 0.6105134316423255,
      "learning_rate": 2.4592239012191977e-05,
      "loss": 0.0388,
      "step": 835
    },
    {
      "epoch": 5.675675675675675,
      "grad_norm": 0.5811817695566549,
      "learning_rate": 2.4343161531153647e-05,
      "loss": 0.0373,
      "step": 840
    },
    {
      "epoch": 5.70945945945946,
      "grad_norm": 0.5984767792779337,
      "learning_rate": 2.4094478068925884e-05,
      "loss": 0.038,
      "step": 845
    },
    {
      "epoch": 5.743243243243243,
      "grad_norm": 0.611272629887631,
      "learning_rate": 2.3846219664778824e-05,
      "loss": 0.0383,
      "step": 850
    },
    {
      "epoch": 5.777027027027027,
      "grad_norm": 0.577343977284889,
      "learning_rate": 2.3598417304929226e-05,
      "loss": 0.0393,
      "step": 855
    },
    {
      "epoch": 5.8108108108108105,
      "grad_norm": 0.5654212108410479,
      "learning_rate": 2.3351101918672985e-05,
      "loss": 0.0374,
      "step": 860
    },
    {
      "epoch": 5.844594594594595,
      "grad_norm": 0.5998738235875581,
      "learning_rate": 2.3104304374524704e-05,
      "loss": 0.0386,
      "step": 865
    },
    {
      "epoch": 5.878378378378378,
      "grad_norm": 0.5685581527847907,
      "learning_rate": 2.2858055476364822e-05,
      "loss": 0.0385,
      "step": 870
    },
    {
      "epoch": 5.912162162162162,
      "grad_norm": 0.5823002111910494,
      "learning_rate": 2.2612385959594877e-05,
      "loss": 0.0381,
      "step": 875
    },
    {
      "epoch": 5.945945945945946,
      "grad_norm": 0.5839824378413123,
      "learning_rate": 2.2367326487301317e-05,
      "loss": 0.0413,
      "step": 880
    },
    {
      "epoch": 5.97972972972973,
      "grad_norm": 0.588588124064064,
      "learning_rate": 2.2122907646428214e-05,
      "loss": 0.0379,
      "step": 885
    },
    {
      "epoch": 6.013513513513513,
      "grad_norm": 0.35563132656032165,
      "learning_rate": 2.1879159943959686e-05,
      "loss": 0.0291,
      "step": 890
    },
    {
      "epoch": 6.047297297297297,
      "grad_norm": 0.3877538214743096,
      "learning_rate": 2.1636113803112097e-05,
      "loss": 0.0183,
      "step": 895
    },
    {
      "epoch": 6.081081081081081,
      "grad_norm": 0.3726851485403488,
      "learning_rate": 2.139379955953686e-05,
      "loss": 0.0173,
      "step": 900
    },
    {
      "epoch": 6.114864864864865,
      "grad_norm": 0.4065314568151127,
      "learning_rate": 2.1152247457534065e-05,
      "loss": 0.017,
      "step": 905
    },
    {
      "epoch": 6.148648648648648,
      "grad_norm": 0.4169040951424247,
      "learning_rate": 2.0911487646277623e-05,
      "loss": 0.016,
      "step": 910
    },
    {
      "epoch": 6.1824324324324325,
      "grad_norm": 0.39785314778930814,
      "learning_rate": 2.067155017605212e-05,
      "loss": 0.0157,
      "step": 915
    },
    {
      "epoch": 6.216216216216216,
      "grad_norm": 0.3835573573672649,
      "learning_rate": 2.0432464994502203e-05,
      "loss": 0.0157,
      "step": 920
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.42586691255540654,
      "learning_rate": 2.0194261942894628e-05,
      "loss": 0.0157,
      "step": 925
    },
    {
      "epoch": 6.283783783783784,
      "grad_norm": 0.43136052279972165,
      "learning_rate": 1.995697075239365e-05,
      "loss": 0.0157,
      "step": 930
    },
    {
      "epoch": 6.3175675675675675,
      "grad_norm": 0.4409662721946437,
      "learning_rate": 1.972062104035017e-05,
      "loss": 0.0177,
      "step": 935
    },
    {
      "epoch": 6.351351351351352,
      "grad_norm": 0.3736257853696285,
      "learning_rate": 1.9485242306605028e-05,
      "loss": 0.0158,
      "step": 940
    },
    {
      "epoch": 6.385135135135135,
      "grad_norm": 0.4439209322219746,
      "learning_rate": 1.9250863929807027e-05,
      "loss": 0.0163,
      "step": 945
    },
    {
      "epoch": 6.418918918918919,
      "grad_norm": 0.44710625472954163,
      "learning_rate": 1.9017515163746058e-05,
      "loss": 0.0172,
      "step": 950
    },
    {
      "epoch": 6.452702702702703,
      "grad_norm": 0.38718415223297625,
      "learning_rate": 1.878522513370177e-05,
      "loss": 0.0154,
      "step": 955
    },
    {
      "epoch": 6.486486486486487,
      "grad_norm": 0.3939284523479193,
      "learning_rate": 1.855402283280836e-05,
      "loss": 0.0153,
      "step": 960
    },
    {
      "epoch": 6.52027027027027,
      "grad_norm": 0.3772763156064826,
      "learning_rate": 1.8323937118435786e-05,
      "loss": 0.0153,
      "step": 965
    },
    {
      "epoch": 6.554054054054054,
      "grad_norm": 0.4118743670831433,
      "learning_rate": 1.8094996708587958e-05,
      "loss": 0.0171,
      "step": 970
    },
    {
      "epoch": 6.587837837837838,
      "grad_norm": 0.3502157244987212,
      "learning_rate": 1.7867230178318334e-05,
      "loss": 0.0157,
      "step": 975
    },
    {
      "epoch": 6.621621621621622,
      "grad_norm": 0.372515539586524,
      "learning_rate": 1.7640665956163306e-05,
      "loss": 0.0162,
      "step": 980
    },
    {
      "epoch": 6.655405405405405,
      "grad_norm": 0.38425153455559596,
      "learning_rate": 1.7415332320593964e-05,
      "loss": 0.0163,
      "step": 985
    },
    {
      "epoch": 6.6891891891891895,
      "grad_norm": 0.3606383402298934,
      "learning_rate": 1.719125739648648e-05,
      "loss": 0.0161,
      "step": 990
    },
    {
      "epoch": 6.722972972972973,
      "grad_norm": 0.4056115111626112,
      "learning_rate": 1.6968469151611766e-05,
      "loss": 0.0156,
      "step": 995
    },
    {
      "epoch": 6.756756756756757,
      "grad_norm": 0.40498458349476124,
      "learning_rate": 1.6746995393144668e-05,
      "loss": 0.0151,
      "step": 1000
    },
    {
      "epoch": 6.79054054054054,
      "grad_norm": 0.42231813172022004,
      "learning_rate": 1.6526863764193228e-05,
      "loss": 0.0146,
      "step": 1005
    },
    {
      "epoch": 6.824324324324325,
      "grad_norm": 0.41971268572130555,
      "learning_rate": 1.6308101740348433e-05,
      "loss": 0.0156,
      "step": 1010
    },
    {
      "epoch": 6.858108108108108,
      "grad_norm": 0.36796570592199035,
      "learning_rate": 1.6090736626254894e-05,
      "loss": 0.0144,
      "step": 1015
    },
    {
      "epoch": 6.891891891891892,
      "grad_norm": 0.3479812131635607,
      "learning_rate": 1.5874795552202773e-05,
      "loss": 0.0142,
      "step": 1020
    },
    {
      "epoch": 6.925675675675675,
      "grad_norm": 0.3741144322926155,
      "learning_rate": 1.5660305470741603e-05,
      "loss": 0.0151,
      "step": 1025
    },
    {
      "epoch": 6.95945945945946,
      "grad_norm": 0.3846553925589328,
      "learning_rate": 1.5447293153316163e-05,
      "loss": 0.015,
      "step": 1030
    },
    {
      "epoch": 6.993243243243243,
      "grad_norm": 0.4021326062149451,
      "learning_rate": 1.523578518692505e-05,
      "loss": 0.016,
      "step": 1035
    },
    {
      "epoch": 7.027027027027027,
      "grad_norm": 0.2156653204415922,
      "learning_rate": 1.5025807970802252e-05,
      "loss": 0.008,
      "step": 1040
    },
    {
      "epoch": 7.0608108108108105,
      "grad_norm": 0.30216138321229236,
      "learning_rate": 1.481738771312209e-05,
      "loss": 0.007,
      "step": 1045
    },
    {
      "epoch": 7.094594594594595,
      "grad_norm": 0.22885301230691482,
      "learning_rate": 1.4610550427728103e-05,
      "loss": 0.0066,
      "step": 1050
    },
    {
      "epoch": 7.128378378378378,
      "grad_norm": 0.4439888190637835,
      "learning_rate": 1.4405321930886161e-05,
      "loss": 0.0069,
      "step": 1055
    },
    {
      "epoch": 7.162162162162162,
      "grad_norm": 0.26648662157316894,
      "learning_rate": 1.4201727838062181e-05,
      "loss": 0.0064,
      "step": 1060
    },
    {
      "epoch": 7.195945945945946,
      "grad_norm": 0.27177618135164877,
      "learning_rate": 1.3999793560724966e-05,
      "loss": 0.0069,
      "step": 1065
    },
    {
      "epoch": 7.22972972972973,
      "grad_norm": 0.25702708811132263,
      "learning_rate": 1.3799544303174514e-05,
      "loss": 0.0065,
      "step": 1070
    },
    {
      "epoch": 7.263513513513513,
      "grad_norm": 0.3064358578966215,
      "learning_rate": 1.3601005059396104e-05,
      "loss": 0.0068,
      "step": 1075
    },
    {
      "epoch": 7.297297297297297,
      "grad_norm": 0.27331042929230237,
      "learning_rate": 1.3404200609940754e-05,
      "loss": 0.0065,
      "step": 1080
    },
    {
      "epoch": 7.331081081081081,
      "grad_norm": 0.33619919125627085,
      "learning_rate": 1.3209155518832203e-05,
      "loss": 0.0062,
      "step": 1085
    },
    {
      "epoch": 7.364864864864865,
      "grad_norm": 0.25439965468725245,
      "learning_rate": 1.3015894130500977e-05,
      "loss": 0.006,
      "step": 1090
    },
    {
      "epoch": 7.398648648648648,
      "grad_norm": 0.2233169372767858,
      "learning_rate": 1.2824440566745865e-05,
      "loss": 0.0065,
      "step": 1095
    },
    {
      "epoch": 7.4324324324324325,
      "grad_norm": 0.3136174702135409,
      "learning_rate": 1.2634818723723174e-05,
      "loss": 0.0061,
      "step": 1100
    },
    {
      "epoch": 7.466216216216216,
      "grad_norm": 0.2029774471743054,
      "learning_rate": 1.2447052268964122e-05,
      "loss": 0.0055,
      "step": 1105
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.24374238272148147,
      "learning_rate": 1.2261164638420832e-05,
      "loss": 0.0065,
      "step": 1110
    },
    {
      "epoch": 7.533783783783784,
      "grad_norm": 0.24135246189785595,
      "learning_rate": 1.2077179033541139e-05,
      "loss": 0.0059,
      "step": 1115
    },
    {
      "epoch": 7.5675675675675675,
      "grad_norm": 0.21810076859660635,
      "learning_rate": 1.1895118418372734e-05,
      "loss": 0.0061,
      "step": 1120
    },
    {
      "epoch": 7.601351351351351,
      "grad_norm": 0.24012263402490466,
      "learning_rate": 1.171500551669697e-05,
      "loss": 0.0057,
      "step": 1125
    },
    {
      "epoch": 7.635135135135135,
      "grad_norm": 0.27467072051926933,
      "learning_rate": 1.1536862809192518e-05,
      "loss": 0.006,
      "step": 1130
    },
    {
      "epoch": 7.668918918918919,
      "grad_norm": 0.23999817155533962,
      "learning_rate": 1.1360712530629513e-05,
      "loss": 0.0065,
      "step": 1135
    },
    {
      "epoch": 7.702702702702703,
      "grad_norm": 0.24863445314463017,
      "learning_rate": 1.1186576667094342e-05,
      "loss": 0.0057,
      "step": 1140
    },
    {
      "epoch": 7.736486486486487,
      "grad_norm": 0.22176049564002084,
      "learning_rate": 1.1014476953245423e-05,
      "loss": 0.0059,
      "step": 1145
    },
    {
      "epoch": 7.77027027027027,
      "grad_norm": 0.23211824524756258,
      "learning_rate": 1.0844434869600428e-05,
      "loss": 0.006,
      "step": 1150
    },
    {
      "epoch": 7.804054054054054,
      "grad_norm": 0.30542373552496616,
      "learning_rate": 1.0676471639855235e-05,
      "loss": 0.0059,
      "step": 1155
    },
    {
      "epoch": 7.837837837837838,
      "grad_norm": 0.23147089199284612,
      "learning_rate": 1.0510608228234848e-05,
      "loss": 0.006,
      "step": 1160
    },
    {
      "epoch": 7.871621621621622,
      "grad_norm": 0.24381478057877082,
      "learning_rate": 1.0346865336876809e-05,
      "loss": 0.0062,
      "step": 1165
    },
    {
      "epoch": 7.905405405405405,
      "grad_norm": 0.2432182162915693,
      "learning_rate": 1.0185263403247256e-05,
      "loss": 0.0056,
      "step": 1170
    },
    {
      "epoch": 7.9391891891891895,
      "grad_norm": 0.3613033258536312,
      "learning_rate": 1.002582259759002e-05,
      "loss": 0.0059,
      "step": 1175
    },
    {
      "epoch": 7.972972972972973,
      "grad_norm": 0.25305037196358393,
      "learning_rate": 9.868562820409103e-06,
      "loss": 0.0051,
      "step": 1180
    },
    {
      "epoch": 8.006756756756756,
      "grad_norm": 0.14099172446623198,
      "learning_rate": 9.713503699984825e-06,
      "loss": 0.0043,
      "step": 1185
    },
    {
      "epoch": 8.04054054054054,
      "grad_norm": 0.10053497170083484,
      "learning_rate": 9.560664589923895e-06,
      "loss": 0.0026,
      "step": 1190
    },
    {
      "epoch": 8.074324324324325,
      "grad_norm": 0.11988320628918524,
      "learning_rate": 9.410064566743841e-06,
      "loss": 0.0024,
      "step": 1195
    },
    {
      "epoch": 8.108108108108109,
      "grad_norm": 0.14397215841006877,
      "learning_rate": 9.261722427491953e-06,
      "loss": 0.0026,
      "step": 1200
    },
    {
      "epoch": 8.141891891891891,
      "grad_norm": 0.1386895769973641,
      "learning_rate": 9.115656687399149e-06,
      "loss": 0.0022,
      "step": 1205
    },
    {
      "epoch": 8.175675675675675,
      "grad_norm": 0.21147271188150515,
      "learning_rate": 8.971885577569058e-06,
      "loss": 0.0022,
      "step": 1210
    },
    {
      "epoch": 8.20945945945946,
      "grad_norm": 0.2758105565736383,
      "learning_rate": 8.83042704270243e-06,
      "loss": 0.0021,
      "step": 1215
    },
    {
      "epoch": 8.243243243243244,
      "grad_norm": 0.1389996580202653,
      "learning_rate": 8.691298738857432e-06,
      "loss": 0.0022,
      "step": 1220
    },
    {
      "epoch": 8.277027027027026,
      "grad_norm": 0.1580398044623882,
      "learning_rate": 8.554518031245934e-06,
      "loss": 0.0022,
      "step": 1225
    },
    {
      "epoch": 8.31081081081081,
      "grad_norm": 0.12407656196067171,
      "learning_rate": 8.420101992066028e-06,
      "loss": 0.002,
      "step": 1230
    },
    {
      "epoch": 8.344594594594595,
      "grad_norm": 0.10679978351606988,
      "learning_rate": 8.288067398371214e-06,
      "loss": 0.0022,
      "step": 1235
    },
    {
      "epoch": 8.378378378378379,
      "grad_norm": 0.16999184531962738,
      "learning_rate": 8.158430729976372e-06,
      "loss": 0.002,
      "step": 1240
    },
    {
      "epoch": 8.412162162162161,
      "grad_norm": 0.13537649324008497,
      "learning_rate": 8.031208167400833e-06,
      "loss": 0.0018,
      "step": 1245
    },
    {
      "epoch": 8.445945945945946,
      "grad_norm": 0.14335948733872103,
      "learning_rate": 7.906415589848834e-06,
      "loss": 0.0018,
      "step": 1250
    },
    {
      "epoch": 8.47972972972973,
      "grad_norm": 0.18441379623144924,
      "learning_rate": 7.78406857322756e-06,
      "loss": 0.002,
      "step": 1255
    },
    {
      "epoch": 8.513513513513514,
      "grad_norm": 0.26138037503231737,
      "learning_rate": 7.664182388203037e-06,
      "loss": 0.002,
      "step": 1260
    },
    {
      "epoch": 8.547297297297296,
      "grad_norm": 0.17556583348056748,
      "learning_rate": 7.54677199829414e-06,
      "loss": 0.0021,
      "step": 1265
    },
    {
      "epoch": 8.58108108108108,
      "grad_norm": 0.13200606559197747,
      "learning_rate": 7.4318520580049444e-06,
      "loss": 0.0022,
      "step": 1270
    },
    {
      "epoch": 8.614864864864865,
      "grad_norm": 0.09284216184237915,
      "learning_rate": 7.3194369109956e-06,
      "loss": 0.002,
      "step": 1275
    },
    {
      "epoch": 8.64864864864865,
      "grad_norm": 0.08554728261946265,
      "learning_rate": 7.209540588292083e-06,
      "loss": 0.0019,
      "step": 1280
    },
    {
      "epoch": 8.682432432432432,
      "grad_norm": 0.11659268339408016,
      "learning_rate": 7.102176806534873e-06,
      "loss": 0.0021,
      "step": 1285
    },
    {
      "epoch": 8.716216216216216,
      "grad_norm": 0.16213377803103676,
      "learning_rate": 6.9973589662669455e-06,
      "loss": 0.0022,
      "step": 1290
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.18750096193431748,
      "learning_rate": 6.8951001502612065e-06,
      "loss": 0.002,
      "step": 1295
    },
    {
      "epoch": 8.783783783783784,
      "grad_norm": 0.10613519780111069,
      "learning_rate": 6.7954131218875404e-06,
      "loss": 0.0023,
      "step": 1300
    },
    {
      "epoch": 8.817567567567568,
      "grad_norm": 0.25385399671959136,
      "learning_rate": 6.69831032351977e-06,
      "loss": 0.0023,
      "step": 1305
    },
    {
      "epoch": 8.85135135135135,
      "grad_norm": 0.1600461171686155,
      "learning_rate": 6.603803874982687e-06,
      "loss": 0.0021,
      "step": 1310
    },
    {
      "epoch": 8.885135135135135,
      "grad_norm": 0.15964134426974025,
      "learning_rate": 6.511905572039298e-06,
      "loss": 0.0024,
      "step": 1315
    },
    {
      "epoch": 8.91891891891892,
      "grad_norm": 0.1566454771356178,
      "learning_rate": 6.422626884918559e-06,
      "loss": 0.0022,
      "step": 1320
    },
    {
      "epoch": 8.952702702702704,
      "grad_norm": 0.09927382907399851,
      "learning_rate": 6.33597895688373e-06,
      "loss": 0.0029,
      "step": 1325
    },
    {
      "epoch": 8.986486486486486,
      "grad_norm": 0.16921829031430702,
      "learning_rate": 6.2519726028415145e-06,
      "loss": 0.0022,
      "step": 1330
    },
    {
      "epoch": 9.02027027027027,
      "grad_norm": 0.07456967679569992,
      "learning_rate": 6.170618307992231e-06,
      "loss": 0.0014,
      "step": 1335
    },
    {
      "epoch": 9.054054054054054,
      "grad_norm": 0.07677701378366658,
      "learning_rate": 6.091926226521089e-06,
      "loss": 0.001,
      "step": 1340
    },
    {
      "epoch": 9.087837837837839,
      "grad_norm": 0.05783212365647212,
      "learning_rate": 6.015906180330808e-06,
      "loss": 0.001,
      "step": 1345
    },
    {
      "epoch": 9.121621621621621,
      "grad_norm": 0.06109685685531086,
      "learning_rate": 5.942567657815696e-06,
      "loss": 0.0009,
      "step": 1350
    },
    {
      "epoch": 9.155405405405405,
      "grad_norm": 0.05632608931733378,
      "learning_rate": 5.871919812677383e-06,
      "loss": 0.0009,
      "step": 1355
    },
    {
      "epoch": 9.18918918918919,
      "grad_norm": 0.08393178027149023,
      "learning_rate": 5.8039714627822754e-06,
      "loss": 0.0009,
      "step": 1360
    },
    {
      "epoch": 9.222972972972974,
      "grad_norm": 0.08764801864993857,
      "learning_rate": 5.738731089060995e-06,
      "loss": 0.0011,
      "step": 1365
    },
    {
      "epoch": 9.256756756756756,
      "grad_norm": 0.07408376282386418,
      "learning_rate": 5.676206834449797e-06,
      "loss": 0.0009,
      "step": 1370
    },
    {
      "epoch": 9.29054054054054,
      "grad_norm": 0.1377155575448204,
      "learning_rate": 5.616406502874251e-06,
      "loss": 0.001,
      "step": 1375
    },
    {
      "epoch": 9.324324324324325,
      "grad_norm": 0.04591228090384543,
      "learning_rate": 5.55933755827518e-06,
      "loss": 0.0008,
      "step": 1380
    },
    {
      "epoch": 9.358108108108109,
      "grad_norm": 0.12072204952935278,
      "learning_rate": 5.505007123677063e-06,
      "loss": 0.001,
      "step": 1385
    },
    {
      "epoch": 9.391891891891891,
      "grad_norm": 0.0723700685429284,
      "learning_rate": 5.453421980298957e-06,
      "loss": 0.0012,
      "step": 1390
    },
    {
      "epoch": 9.425675675675675,
      "grad_norm": 0.056722410222165544,
      "learning_rate": 5.4045885667081375e-06,
      "loss": 0.0008,
      "step": 1395
    },
    {
      "epoch": 9.45945945945946,
      "grad_norm": 0.11342540472708795,
      "learning_rate": 5.358512978016445e-06,
      "loss": 0.0009,
      "step": 1400
    },
    {
      "epoch": 9.493243243243244,
      "grad_norm": 0.04838998780652165,
      "learning_rate": 5.315200965119541e-06,
      "loss": 0.0008,
      "step": 1405
    },
    {
      "epoch": 9.527027027027026,
      "grad_norm": 0.14216495240988,
      "learning_rate": 5.27465793397911e-06,
      "loss": 0.001,
      "step": 1410
    },
    {
      "epoch": 9.56081081081081,
      "grad_norm": 0.053216703580214755,
      "learning_rate": 5.236888944948117e-06,
      "loss": 0.0008,
      "step": 1415
    },
    {
      "epoch": 9.594594594594595,
      "grad_norm": 0.10139816433348116,
      "learning_rate": 5.201898712139201e-06,
      "loss": 0.001,
      "step": 1420
    },
    {
      "epoch": 9.628378378378379,
      "grad_norm": 0.06402496940461881,
      "learning_rate": 5.1696916028362964e-06,
      "loss": 0.0008,
      "step": 1425
    },
    {
      "epoch": 9.662162162162161,
      "grad_norm": 0.07094878188815971,
      "learning_rate": 5.1402716369495194e-06,
      "loss": 0.0009,
      "step": 1430
    },
    {
      "epoch": 9.695945945945946,
      "grad_norm": 0.10186073339095397,
      "learning_rate": 5.113642486513428e-06,
      "loss": 0.0012,
      "step": 1435
    },
    {
      "epoch": 9.72972972972973,
      "grad_norm": 0.07465500402483634,
      "learning_rate": 5.089807475228711e-06,
      "loss": 0.0009,
      "step": 1440
    },
    {
      "epoch": 9.763513513513514,
      "grad_norm": 0.10939450461018281,
      "learning_rate": 5.06876957804733e-06,
      "loss": 0.0009,
      "step": 1445
    },
    {
      "epoch": 9.797297297297296,
      "grad_norm": 0.067342602521065,
      "learning_rate": 5.050531420801205e-06,
      "loss": 0.0009,
      "step": 1450
    },
    {
      "epoch": 9.83108108108108,
      "grad_norm": 0.049654861207641725,
      "learning_rate": 5.03509527987448e-06,
      "loss": 0.0008,
      "step": 1455
    },
    {
      "epoch": 9.864864864864865,
      "grad_norm": 0.07521612378431902,
      "learning_rate": 5.022463081919386e-06,
      "loss": 0.0009,
      "step": 1460
    },
    {
      "epoch": 9.89864864864865,
      "grad_norm": 0.10954313412188729,
      "learning_rate": 5.012636403615775e-06,
      "loss": 0.001,
      "step": 1465
    },
    {
      "epoch": 9.932432432432432,
      "grad_norm": 0.07101568129957397,
      "learning_rate": 5.005616471474332e-06,
      "loss": 0.0009,
      "step": 1470
    },
    {
      "epoch": 9.966216216216216,
      "grad_norm": 0.038606444834761555,
      "learning_rate": 5.001404161683473e-06,
      "loss": 0.0009,
      "step": 1475
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.07528867672379237,
      "learning_rate": 5e-06,
      "loss": 0.0009,
      "step": 1480
    },
    {
      "epoch": 10.0,
      "step": 1480,
      "total_flos": 1614610083348480.0,
      "train_loss": 0.20890064053127316,
      "train_runtime": 23696.909,
      "train_samples_per_second": 3.992,
      "train_steps_per_second": 0.062
    }
  ],
  "logging_steps": 5,
  "max_steps": 1480,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1614610083348480.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}