{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.999731255038968,
  "eval_steps": 500,
  "global_step": 930,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010749798441279225,
      "grad_norm": 22.998251618632413,
      "learning_rate": 1.0752688172043012e-07,
      "loss": 1.3109,
      "step": 1
    },
    {
      "epoch": 0.005374899220639613,
      "grad_norm": 21.62666674963918,
      "learning_rate": 5.376344086021506e-07,
      "loss": 1.319,
      "step": 5
    },
    {
      "epoch": 0.010749798441279226,
      "grad_norm": 8.47868515286049,
      "learning_rate": 1.0752688172043011e-06,
      "loss": 1.201,
      "step": 10
    },
    {
      "epoch": 0.01612469766191884,
      "grad_norm": 7.0045452722639245,
      "learning_rate": 1.6129032258064516e-06,
      "loss": 1.0409,
      "step": 15
    },
    {
      "epoch": 0.021499596882558453,
      "grad_norm": 2.9866055931463724,
      "learning_rate": 2.1505376344086023e-06,
      "loss": 0.9079,
      "step": 20
    },
    {
      "epoch": 0.026874496103198066,
      "grad_norm": 2.466878497235169,
      "learning_rate": 2.688172043010753e-06,
      "loss": 0.8681,
      "step": 25
    },
    {
      "epoch": 0.03224939532383768,
      "grad_norm": 2.1381469942804414,
      "learning_rate": 3.225806451612903e-06,
      "loss": 0.8354,
      "step": 30
    },
    {
      "epoch": 0.03762429454447729,
      "grad_norm": 2.3822492851866888,
      "learning_rate": 3.763440860215054e-06,
      "loss": 0.8242,
      "step": 35
    },
    {
      "epoch": 0.042999193765116905,
      "grad_norm": 2.2460422471438446,
      "learning_rate": 4.3010752688172045e-06,
      "loss": 0.8055,
      "step": 40
    },
    {
      "epoch": 0.04837409298575652,
      "grad_norm": 2.278945525704108,
      "learning_rate": 4.838709677419355e-06,
      "loss": 0.7808,
      "step": 45
    },
    {
      "epoch": 0.05374899220639613,
      "grad_norm": 2.3939238018515785,
      "learning_rate": 5.376344086021506e-06,
      "loss": 0.7718,
      "step": 50
    },
    {
      "epoch": 0.059123891427035745,
      "grad_norm": 2.4529746003373805,
      "learning_rate": 5.9139784946236566e-06,
      "loss": 0.7577,
      "step": 55
    },
    {
      "epoch": 0.06449879064767536,
      "grad_norm": 2.370219230419958,
      "learning_rate": 6.451612903225806e-06,
      "loss": 0.7406,
      "step": 60
    },
    {
      "epoch": 0.06987368986831496,
      "grad_norm": 2.5160763656892278,
      "learning_rate": 6.989247311827958e-06,
      "loss": 0.7328,
      "step": 65
    },
    {
      "epoch": 0.07524858908895458,
      "grad_norm": 2.2323762062076615,
      "learning_rate": 7.526881720430108e-06,
      "loss": 0.7049,
      "step": 70
    },
    {
      "epoch": 0.08062348830959419,
      "grad_norm": 2.1957037704755717,
      "learning_rate": 8.064516129032258e-06,
      "loss": 0.7038,
      "step": 75
    },
    {
      "epoch": 0.08599838753023381,
      "grad_norm": 2.263429588353797,
      "learning_rate": 8.602150537634409e-06,
      "loss": 0.7177,
      "step": 80
    },
    {
      "epoch": 0.09137328675087342,
      "grad_norm": 2.2491062827888757,
      "learning_rate": 9.13978494623656e-06,
      "loss": 0.7082,
      "step": 85
    },
    {
      "epoch": 0.09674818597151304,
      "grad_norm": 2.3972214541569326,
      "learning_rate": 9.67741935483871e-06,
      "loss": 0.7052,
      "step": 90
    },
    {
      "epoch": 0.10212308519215264,
      "grad_norm": 2.442294992070258,
      "learning_rate": 9.999859120828162e-06,
      "loss": 0.6871,
      "step": 95
    },
    {
      "epoch": 0.10749798441279226,
      "grad_norm": 2.1737197659282677,
      "learning_rate": 9.998274321315453e-06,
      "loss": 0.691,
      "step": 100
    },
    {
      "epoch": 0.11287288363343187,
      "grad_norm": 2.086931041290223,
      "learning_rate": 9.994929183335237e-06,
      "loss": 0.6887,
      "step": 105
    },
    {
      "epoch": 0.11824778285407149,
      "grad_norm": 2.181755095183289,
      "learning_rate": 9.989824885009142e-06,
      "loss": 0.6761,
      "step": 110
    },
    {
      "epoch": 0.1236226820747111,
      "grad_norm": 2.1430422893188195,
      "learning_rate": 9.982963224016152e-06,
      "loss": 0.6748,
      "step": 115
    },
    {
      "epoch": 0.12899758129535072,
      "grad_norm": 2.0944055561946047,
      "learning_rate": 9.974346616959476e-06,
      "loss": 0.6782,
      "step": 120
    },
    {
      "epoch": 0.13437248051599032,
      "grad_norm": 2.7234289381522694,
      "learning_rate": 9.963978098515468e-06,
      "loss": 0.6717,
      "step": 125
    },
    {
      "epoch": 0.13974737973662993,
      "grad_norm": 2.2612331983977265,
      "learning_rate": 9.951861320364822e-06,
      "loss": 0.671,
      "step": 130
    },
    {
      "epoch": 0.14512227895726956,
      "grad_norm": 2.349078940825244,
      "learning_rate": 9.938000549906509e-06,
      "loss": 0.6628,
      "step": 135
    },
    {
      "epoch": 0.15049717817790917,
      "grad_norm": 2.213915224858813,
      "learning_rate": 9.922400668754833e-06,
      "loss": 0.6578,
      "step": 140
    },
    {
      "epoch": 0.15587207739854878,
      "grad_norm": 2.0247679479977947,
      "learning_rate": 9.905067171020185e-06,
      "loss": 0.6621,
      "step": 145
    },
    {
      "epoch": 0.16124697661918838,
      "grad_norm": 2.175459644155833,
      "learning_rate": 9.88600616137407e-06,
      "loss": 0.6539,
      "step": 150
    },
    {
      "epoch": 0.16662187583982802,
      "grad_norm": 1.9231116317571189,
      "learning_rate": 9.86522435289912e-06,
      "loss": 0.6486,
      "step": 155
    },
    {
      "epoch": 0.17199677506046762,
      "grad_norm": 1.9429182926189486,
      "learning_rate": 9.8427290647248e-06,
      "loss": 0.6446,
      "step": 160
    },
    {
      "epoch": 0.17737167428110723,
      "grad_norm": 2.029616741174372,
      "learning_rate": 9.818528219449705e-06,
      "loss": 0.6393,
      "step": 165
    },
    {
      "epoch": 0.18274657350174683,
      "grad_norm": 1.954236053756045,
      "learning_rate": 9.792630340351301e-06,
      "loss": 0.6505,
      "step": 170
    },
    {
      "epoch": 0.18812147272238647,
      "grad_norm": 2.0238728155049994,
      "learning_rate": 9.765044548384113e-06,
      "loss": 0.6345,
      "step": 175
    },
    {
      "epoch": 0.19349637194302607,
      "grad_norm": 1.9933506258733436,
      "learning_rate": 9.735780558967434e-06,
      "loss": 0.6321,
      "step": 180
    },
    {
      "epoch": 0.19887127116366568,
      "grad_norm": 2.159514610870395,
      "learning_rate": 9.70484867856365e-06,
      "loss": 0.6307,
      "step": 185
    },
    {
      "epoch": 0.2042461703843053,
      "grad_norm": 1.8959726839851836,
      "learning_rate": 9.67225980104841e-06,
      "loss": 0.6344,
      "step": 190
    },
    {
      "epoch": 0.20962106960494492,
      "grad_norm": 1.9231014581432677,
      "learning_rate": 9.638025403873939e-06,
      "loss": 0.6354,
      "step": 195
    },
    {
      "epoch": 0.21499596882558453,
      "grad_norm": 1.9470579040757032,
      "learning_rate": 9.602157544026785e-06,
      "loss": 0.6213,
      "step": 200
    },
    {
      "epoch": 0.22037086804622413,
      "grad_norm": 2.1141633711901293,
      "learning_rate": 9.564668853781483e-06,
      "loss": 0.6287,
      "step": 205
    },
    {
      "epoch": 0.22574576726686374,
      "grad_norm": 2.094733969067925,
      "learning_rate": 9.525572536251608e-06,
      "loss": 0.6184,
      "step": 210
    },
    {
      "epoch": 0.23112066648750335,
      "grad_norm": 1.8594092068711054,
      "learning_rate": 9.484882360739772e-06,
      "loss": 0.6049,
      "step": 215
    },
    {
      "epoch": 0.23649556570814298,
      "grad_norm": 1.8042100641048069,
      "learning_rate": 9.442612657888237e-06,
      "loss": 0.604,
      "step": 220
    },
    {
      "epoch": 0.24187046492878259,
      "grad_norm": 1.9491363567278153,
      "learning_rate": 9.398778314631801e-06,
      "loss": 0.5935,
      "step": 225
    },
    {
      "epoch": 0.2472453641494222,
      "grad_norm": 2.0060700754274112,
      "learning_rate": 9.353394768954791e-06,
      "loss": 0.5973,
      "step": 230
    },
    {
      "epoch": 0.2526202633700618,
      "grad_norm": 2.323587224840248,
      "learning_rate": 9.30647800445397e-06,
      "loss": 0.602,
      "step": 235
    },
    {
      "epoch": 0.25799516259070143,
      "grad_norm": 2.2542109995103985,
      "learning_rate": 9.258044544709276e-06,
      "loss": 0.6005,
      "step": 240
    },
    {
      "epoch": 0.26337006181134104,
      "grad_norm": 2.205069286015657,
      "learning_rate": 9.208111447464407e-06,
      "loss": 0.5937,
      "step": 245
    },
    {
      "epoch": 0.26874496103198064,
      "grad_norm": 1.979112533465105,
      "learning_rate": 9.156696298619266e-06,
      "loss": 0.5905,
      "step": 250
    },
    {
      "epoch": 0.27411986025262025,
      "grad_norm": 2.171300635630568,
      "learning_rate": 9.103817206036383e-06,
      "loss": 0.5885,
      "step": 255
    },
    {
      "epoch": 0.27949475947325986,
      "grad_norm": 1.9064161905725763,
      "learning_rate": 9.049492793163539e-06,
      "loss": 0.5898,
      "step": 260
    },
    {
      "epoch": 0.28486965869389946,
      "grad_norm": 1.9617828705521112,
      "learning_rate": 8.993742192474773e-06,
      "loss": 0.5831,
      "step": 265
    },
    {
      "epoch": 0.2902445579145391,
      "grad_norm": 1.8575944912932347,
      "learning_rate": 8.936585038732143e-06,
      "loss": 0.5751,
      "step": 270
    },
    {
      "epoch": 0.29561945713517873,
      "grad_norm": 1.9607189835672174,
      "learning_rate": 8.878041462070556e-06,
      "loss": 0.5795,
      "step": 275
    },
    {
      "epoch": 0.30099435635581834,
      "grad_norm": 1.9827526904209154,
      "learning_rate": 8.818132080908178e-06,
      "loss": 0.5756,
      "step": 280
    },
    {
      "epoch": 0.30636925557645794,
      "grad_norm": 1.9289388619958865,
      "learning_rate": 8.756877994684818e-06,
      "loss": 0.5797,
      "step": 285
    },
    {
      "epoch": 0.31174415479709755,
      "grad_norm": 1.9423515794251163,
      "learning_rate": 8.694300776430958e-06,
      "loss": 0.5654,
      "step": 290
    },
    {
      "epoch": 0.31711905401773716,
      "grad_norm": 2.133199145110774,
      "learning_rate": 8.630422465169947e-06,
      "loss": 0.5586,
      "step": 295
    },
    {
      "epoch": 0.32249395323837676,
      "grad_norm": 2.2051473612841845,
      "learning_rate": 8.565265558156101e-06,
      "loss": 0.5674,
      "step": 300
    },
    {
      "epoch": 0.32786885245901637,
      "grad_norm": 1.94675668742253,
      "learning_rate": 8.498853002951414e-06,
      "loss": 0.5572,
      "step": 305
    },
    {
      "epoch": 0.33324375167965603,
      "grad_norm": 2.046407155545796,
      "learning_rate": 8.43120818934367e-06,
      "loss": 0.5535,
      "step": 310
    },
    {
      "epoch": 0.33861865090029564,
      "grad_norm": 2.0046384820796486,
      "learning_rate": 8.362354941108803e-06,
      "loss": 0.5577,
      "step": 315
    },
    {
      "epoch": 0.34399355012093524,
      "grad_norm": 2.0243084631372223,
      "learning_rate": 8.292317507620438e-06,
      "loss": 0.566,
      "step": 320
    },
    {
      "epoch": 0.34936844934157485,
      "grad_norm": 1.8780763393652544,
      "learning_rate": 8.221120555309511e-06,
      "loss": 0.5471,
      "step": 325
    },
    {
      "epoch": 0.35474334856221446,
      "grad_norm": 2.0125466037004967,
      "learning_rate": 8.148789158977012e-06,
      "loss": 0.55,
      "step": 330
    },
    {
      "epoch": 0.36011824778285406,
      "grad_norm": 2.111876068514048,
      "learning_rate": 8.075348792962924e-06,
      "loss": 0.542,
      "step": 335
    },
    {
      "epoch": 0.36549314700349367,
      "grad_norm": 1.831727662600693,
      "learning_rate": 8.000825322174424e-06,
      "loss": 0.541,
      "step": 340
    },
    {
      "epoch": 0.3708680462241333,
      "grad_norm": 1.957058515389012,
      "learning_rate": 7.925244992976538e-06,
      "loss": 0.525,
      "step": 345
    },
    {
      "epoch": 0.37624294544477294,
      "grad_norm": 2.016623611705091,
      "learning_rate": 7.848634423948468e-06,
      "loss": 0.5294,
      "step": 350
    },
    {
      "epoch": 0.38161784466541254,
      "grad_norm": 1.9380904809426653,
      "learning_rate": 7.7710205965088e-06,
      "loss": 0.5324,
      "step": 355
    },
    {
      "epoch": 0.38699274388605215,
      "grad_norm": 2.0054637192864746,
      "learning_rate": 7.692430845412946e-06,
      "loss": 0.5276,
      "step": 360
    },
    {
      "epoch": 0.39236764310669175,
      "grad_norm": 2.201718543185445,
      "learning_rate": 7.612892849126132e-06,
      "loss": 0.5286,
      "step": 365
    },
    {
      "epoch": 0.39774254232733136,
      "grad_norm": 1.9346693875946135,
      "learning_rate": 7.532434620075349e-06,
      "loss": 0.5217,
      "step": 370
    },
    {
      "epoch": 0.40311744154797097,
      "grad_norm": 2.046737331443054,
      "learning_rate": 7.451084494783668e-06,
      "loss": 0.5196,
      "step": 375
    },
    {
      "epoch": 0.4084923407686106,
      "grad_norm": 1.9926284827905993,
      "learning_rate": 7.368871123890425e-06,
      "loss": 0.5186,
      "step": 380
    },
    {
      "epoch": 0.4138672399892502,
      "grad_norm": 2.0164701987233133,
      "learning_rate": 7.285823462060776e-06,
      "loss": 0.5155,
      "step": 385
    },
    {
      "epoch": 0.41924213920988984,
      "grad_norm": 1.821494525655949,
      "learning_rate": 7.201970757788172e-06,
      "loss": 0.5094,
      "step": 390
    },
    {
      "epoch": 0.42461703843052945,
      "grad_norm": 2.0350416485078533,
      "learning_rate": 7.117342543093358e-06,
      "loss": 0.5051,
      "step": 395
    },
    {
      "epoch": 0.42999193765116905,
      "grad_norm": 1.9177286245297331,
      "learning_rate": 7.031968623123503e-06,
      "loss": 0.5037,
      "step": 400
    },
    {
      "epoch": 0.43536683687180866,
      "grad_norm": 1.8060468183436844,
      "learning_rate": 6.945879065655164e-06,
      "loss": 0.5091,
      "step": 405
    },
    {
      "epoch": 0.44074173609244827,
      "grad_norm": 2.0324990194845425,
      "learning_rate": 6.859104190504725e-06,
      "loss": 0.5081,
      "step": 410
    },
    {
      "epoch": 0.4461166353130879,
      "grad_norm": 1.9781738403258928,
      "learning_rate": 6.771674558850088e-06,
      "loss": 0.4978,
      "step": 415
    },
    {
      "epoch": 0.4514915345337275,
      "grad_norm": 1.967394802071695,
      "learning_rate": 6.6836209624673575e-06,
      "loss": 0.4993,
      "step": 420
    },
    {
      "epoch": 0.4568664337543671,
      "grad_norm": 1.8301107915867578,
      "learning_rate": 6.5949744128863026e-06,
      "loss": 0.4951,
      "step": 425
    },
    {
      "epoch": 0.4622413329750067,
      "grad_norm": 1.8666451752143274,
      "learning_rate": 6.5057661304684314e-06,
      "loss": 0.4929,
      "step": 430
    },
    {
      "epoch": 0.46761623219564635,
      "grad_norm": 1.820566898851988,
      "learning_rate": 6.41602753341152e-06,
      "loss": 0.4925,
      "step": 435
    },
    {
      "epoch": 0.47299113141628596,
      "grad_norm": 1.972020976387126,
      "learning_rate": 6.32579022668446e-06,
      "loss": 0.4869,
      "step": 440
    },
    {
      "epoch": 0.47836603063692557,
      "grad_norm": 1.9421491239061908,
      "learning_rate": 6.235085990896317e-06,
      "loss": 0.4757,
      "step": 445
    },
    {
      "epoch": 0.48374092985756517,
      "grad_norm": 1.9181960459695213,
      "learning_rate": 6.143946771103561e-06,
      "loss": 0.4941,
      "step": 450
    },
    {
      "epoch": 0.4891158290782048,
      "grad_norm": 1.9219139761794142,
      "learning_rate": 6.052404665559342e-06,
      "loss": 0.468,
      "step": 455
    },
    {
      "epoch": 0.4944907282988444,
      "grad_norm": 1.9586485250210308,
      "learning_rate": 5.960491914408846e-06,
      "loss": 0.4704,
      "step": 460
    },
    {
      "epoch": 0.499865627519484,
      "grad_norm": 1.978155481583312,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.4891,
      "step": 465
    },
    {
      "epoch": 0.5052405267401237,
      "grad_norm": 1.9491669694735005,
      "learning_rate": 5.775684077156133e-06,
      "loss": 0.4689,
      "step": 470
    },
    {
      "epoch": 0.5106154259607633,
      "grad_norm": 1.9616828525043124,
      "learning_rate": 5.682854078386882e-06,
      "loss": 0.4722,
      "step": 475
    },
    {
      "epoch": 0.5159903251814029,
      "grad_norm": 1.9497396606270447,
      "learning_rate": 5.5897835857542315e-06,
      "loss": 0.4624,
      "step": 480
    },
    {
      "epoch": 0.5213652244020425,
      "grad_norm": 1.895936510041223,
      "learning_rate": 5.496505377684858e-06,
      "loss": 0.4519,
      "step": 485
    },
    {
      "epoch": 0.5267401236226821,
      "grad_norm": 1.97427958718019,
      "learning_rate": 5.4030523057605865e-06,
      "loss": 0.4671,
      "step": 490
    },
    {
      "epoch": 0.5321150228433217,
      "grad_norm": 1.99020514417267,
      "learning_rate": 5.30945728314841e-06,
      "loss": 0.4555,
      "step": 495
    },
    {
      "epoch": 0.5374899220639613,
      "grad_norm": 1.834260157432595,
      "learning_rate": 5.215753273008828e-06,
      "loss": 0.4592,
      "step": 500
    },
    {
      "epoch": 0.5428648212846009,
      "grad_norm": 2.07183045007662,
      "learning_rate": 5.1219732768865744e-06,
      "loss": 0.4544,
      "step": 505
    },
    {
      "epoch": 0.5482397205052405,
      "grad_norm": 1.8925742071728788,
      "learning_rate": 5.0281503230878304e-06,
      "loss": 0.4503,
      "step": 510
    },
    {
      "epoch": 0.5536146197258801,
      "grad_norm": 2.120435800008408,
      "learning_rate": 4.934317455048005e-06,
      "loss": 0.4432,
      "step": 515
    },
    {
      "epoch": 0.5589895189465197,
      "grad_norm": 2.078137912583167,
      "learning_rate": 4.840507719694202e-06,
      "loss": 0.4454,
      "step": 520
    },
    {
      "epoch": 0.5643644181671593,
      "grad_norm": 1.8218423489133921,
      "learning_rate": 4.746754155806437e-06,
      "loss": 0.4353,
      "step": 525
    },
    {
      "epoch": 0.5697393173877989,
      "grad_norm": 1.886033793318668,
      "learning_rate": 4.6530897823817425e-06,
      "loss": 0.4518,
      "step": 530
    },
    {
      "epoch": 0.5751142166084386,
      "grad_norm": 1.837947070178225,
      "learning_rate": 4.559547587005227e-06,
      "loss": 0.4491,
      "step": 535
    },
    {
      "epoch": 0.5804891158290783,
      "grad_norm": 1.8242868271131116,
      "learning_rate": 4.466160514232206e-06,
      "loss": 0.4405,
      "step": 540
    },
    {
      "epoch": 0.5858640150497179,
      "grad_norm": 1.9410093391418528,
      "learning_rate": 4.3729614539854815e-06,
      "loss": 0.4323,
      "step": 545
    },
    {
      "epoch": 0.5912389142703575,
      "grad_norm": 2.079990387721551,
      "learning_rate": 4.279983229971863e-06,
      "loss": 0.4347,
      "step": 550
    },
    {
      "epoch": 0.5966138134909971,
      "grad_norm": 1.8661086060382346,
      "learning_rate": 4.187258588122019e-06,
      "loss": 0.4309,
      "step": 555
    },
    {
      "epoch": 0.6019887127116367,
      "grad_norm": 1.9549179181995389,
      "learning_rate": 4.094820185057701e-06,
      "loss": 0.434,
      "step": 560
    },
    {
      "epoch": 0.6073636119322763,
      "grad_norm": 1.9438750319747926,
      "learning_rate": 4.002700576590441e-06,
      "loss": 0.417,
      "step": 565
    },
    {
      "epoch": 0.6127385111529159,
      "grad_norm": 1.86309458236793,
      "learning_rate": 3.910932206255742e-06,
      "loss": 0.4196,
      "step": 570
    },
    {
      "epoch": 0.6181134103735555,
      "grad_norm": 1.931200769910855,
      "learning_rate": 3.819547393886816e-06,
      "loss": 0.4089,
      "step": 575
    },
    {
      "epoch": 0.6234883095941951,
      "grad_norm": 1.9236481963943084,
      "learning_rate": 3.7285783242318773e-06,
      "loss": 0.4176,
      "step": 580
    },
    {
      "epoch": 0.6288632088148347,
      "grad_norm": 1.987655423768294,
      "learning_rate": 3.6380570356190346e-06,
      "loss": 0.4111,
      "step": 585
    },
    {
      "epoch": 0.6342381080354743,
      "grad_norm": 1.9140625569934329,
      "learning_rate": 3.548015408672723e-06,
      "loss": 0.4179,
      "step": 590
    },
    {
      "epoch": 0.6396130072561139,
      "grad_norm": 1.914275720117943,
      "learning_rate": 3.4584851550857007e-06,
      "loss": 0.4101,
      "step": 595
    },
    {
      "epoch": 0.6449879064767535,
      "grad_norm": 1.9920083731014693,
      "learning_rate": 3.3694978064505258e-06,
      "loss": 0.4096,
      "step": 600
    },
    {
      "epoch": 0.6503628056973931,
      "grad_norm": 1.8555446650582728,
      "learning_rate": 3.2810847031544703e-06,
      "loss": 0.4134,
      "step": 605
    },
    {
      "epoch": 0.6557377049180327,
      "grad_norm": 1.847095213057005,
      "learning_rate": 3.193276983341773e-06,
      "loss": 0.4044,
      "step": 610
    },
    {
      "epoch": 0.6611126041386725,
      "grad_norm": 1.8348823558923015,
      "learning_rate": 3.10610557194712e-06,
      "loss": 0.4081,
      "step": 615
    },
    {
      "epoch": 0.6664875033593121,
      "grad_norm": 1.7814726492285784,
      "learning_rate": 3.019601169804216e-06,
      "loss": 0.4054,
      "step": 620
    },
    {
      "epoch": 0.6718624025799517,
      "grad_norm": 1.9239354085532425,
      "learning_rate": 2.9337942428332787e-06,
      "loss": 0.4072,
      "step": 625
    },
    {
      "epoch": 0.6772373018005913,
      "grad_norm": 1.8174216015392752,
      "learning_rate": 2.848715011311271e-06,
      "loss": 0.3882,
      "step": 630
    },
    {
      "epoch": 0.6826122010212309,
      "grad_norm": 1.8355921561483057,
      "learning_rate": 2.764393439228643e-06,
      "loss": 0.4032,
      "step": 635
    },
    {
      "epoch": 0.6879871002418705,
      "grad_norm": 1.8428986241754186,
      "learning_rate": 2.6808592237363364e-06,
      "loss": 0.3938,
      "step": 640
    },
    {
      "epoch": 0.6933619994625101,
      "grad_norm": 1.930494807022806,
      "learning_rate": 2.5981417846867753e-06,
      "loss": 0.3827,
      "step": 645
    },
    {
      "epoch": 0.6987368986831497,
      "grad_norm": 1.9965194348229882,
      "learning_rate": 2.5162702542724924e-06,
      "loss": 0.4025,
      "step": 650
    },
    {
      "epoch": 0.7041117979037893,
      "grad_norm": 1.9343155927695501,
      "learning_rate": 2.4352734667661073e-06,
      "loss": 0.3931,
      "step": 655
    },
    {
      "epoch": 0.7094866971244289,
      "grad_norm": 1.8913804765293543,
      "learning_rate": 2.3551799483651894e-06,
      "loss": 0.4008,
      "step": 660
    },
    {
      "epoch": 0.7148615963450685,
      "grad_norm": 1.877731033002426,
      "learning_rate": 2.2760179071456356e-06,
      "loss": 0.3836,
      "step": 665
    },
    {
      "epoch": 0.7202364955657081,
      "grad_norm": 2.271932216399522,
      "learning_rate": 2.1978152231271077e-06,
      "loss": 0.3899,
      "step": 670
    },
    {
      "epoch": 0.7256113947863477,
      "grad_norm": 1.8362601272256027,
      "learning_rate": 2.120599438453968e-06,
      "loss": 0.3935,
      "step": 675
    },
    {
      "epoch": 0.7309862940069873,
      "grad_norm": 1.8391327367572157,
      "learning_rate": 2.044397747695247e-06,
      "loss": 0.39,
      "step": 680
    },
    {
      "epoch": 0.7363611932276269,
      "grad_norm": 1.8672651878485813,
      "learning_rate": 1.969236988267005e-06,
      "loss": 0.3829,
      "step": 685
    },
    {
      "epoch": 0.7417360924482665,
      "grad_norm": 1.8664855719301263,
      "learning_rate": 1.8951436309804766e-06,
      "loss": 0.3826,
      "step": 690
    },
    {
      "epoch": 0.7471109916689062,
      "grad_norm": 1.8648482987681003,
      "learning_rate": 1.8221437707193424e-06,
      "loss": 0.3796,
      "step": 695
    },
    {
      "epoch": 0.7524858908895459,
      "grad_norm": 1.8823945570583507,
      "learning_rate": 1.7502631172493878e-06,
      "loss": 0.3784,
      "step": 700
    },
    {
      "epoch": 0.7578607901101855,
      "grad_norm": 1.8429277019145163,
      "learning_rate": 1.6795269861638041e-06,
      "loss": 0.3808,
      "step": 705
    },
    {
      "epoch": 0.7632356893308251,
      "grad_norm": 1.8821109309531021,
      "learning_rate": 1.6099602899673083e-06,
      "loss": 0.3752,
      "step": 710
    },
    {
      "epoch": 0.7686105885514647,
      "grad_norm": 1.8044029771514785,
      "learning_rate": 1.5415875293022181e-06,
      "loss": 0.3631,
      "step": 715
    },
    {
      "epoch": 0.7739854877721043,
      "grad_norm": 1.8714319214476052,
      "learning_rate": 1.4744327843196043e-06,
      "loss": 0.3722,
      "step": 720
    },
    {
      "epoch": 0.7793603869927439,
      "grad_norm": 1.8005842812695019,
      "learning_rate": 1.4085197061985022e-06,
      "loss": 0.3764,
      "step": 725
    },
    {
      "epoch": 0.7847352862133835,
      "grad_norm": 1.9498199256183426,
      "learning_rate": 1.3438715088162403e-06,
      "loss": 0.3717,
      "step": 730
    },
    {
      "epoch": 0.7901101854340231,
      "grad_norm": 1.7622641111168549,
      "learning_rate": 1.280510960572745e-06,
      "loss": 0.3726,
      "step": 735
    },
    {
      "epoch": 0.7954850846546627,
      "grad_norm": 1.859176007788758,
      "learning_rate": 1.2184603763717684e-06,
      "loss": 0.3581,
      "step": 740
    },
    {
      "epoch": 0.8008599838753023,
      "grad_norm": 1.8187680407734916,
      "learning_rate": 1.1577416097618138e-06,
      "loss": 0.361,
      "step": 745
    },
    {
      "epoch": 0.8062348830959419,
      "grad_norm": 1.8234490317185335,
      "learning_rate": 1.0983760452395415e-06,
      "loss": 0.3629,
      "step": 750
    },
    {
      "epoch": 0.8116097823165815,
      "grad_norm": 1.7846174067763332,
      "learning_rate": 1.040384590718399e-06,
      "loss": 0.3603,
      "step": 755
    },
    {
      "epoch": 0.8169846815372211,
      "grad_norm": 1.9539954235141312,
      "learning_rate": 9.837876701650606e-07,
      "loss": 0.3604,
      "step": 760
    },
    {
      "epoch": 0.8223595807578608,
      "grad_norm": 1.829814154109719,
      "learning_rate": 9.286052164063369e-07,
      "loss": 0.361,
      "step": 765
    },
    {
      "epoch": 0.8277344799785004,
      "grad_norm": 1.7589374129255484,
      "learning_rate": 8.748566641090433e-07,
      "loss": 0.3598,
      "step": 770
    },
    {
      "epoch": 0.83310937919914,
      "grad_norm": 1.8087859749367046,
      "learning_rate": 8.225609429353187e-07,
      "loss": 0.3575,
      "step": 775
    },
    {
      "epoch": 0.8384842784197797,
      "grad_norm": 1.659093947160247,
      "learning_rate": 7.717364708758024e-07,
      "loss": 0.357,
      "step": 780
    },
    {
      "epoch": 0.8438591776404193,
      "grad_norm": 1.7512154153875217,
      "learning_rate": 7.224011477630166e-07,
      "loss": 0.3556,
      "step": 785
    },
    {
      "epoch": 0.8492340768610589,
      "grad_norm": 1.721279609127408,
      "learning_rate": 6.745723489672412e-07,
      "loss": 0.3597,
      "step": 790
    },
    {
      "epoch": 0.8546089760816985,
      "grad_norm": 1.933232116242709,
      "learning_rate": 6.282669192770896e-07,
      "loss": 0.3583,
      "step": 795
    },
    {
      "epoch": 0.8599838753023381,
      "grad_norm": 1.807730857907691,
      "learning_rate": 5.83501166966956e-07,
      "loss": 0.3524,
      "step": 800
    },
    {
      "epoch": 0.8653587745229777,
      "grad_norm": 1.7853591098408064,
      "learning_rate": 5.402908580534233e-07,
      "loss": 0.3623,
      "step": 805
    },
    {
      "epoch": 0.8707336737436173,
      "grad_norm": 1.706116308384166,
      "learning_rate": 4.986512107426283e-07,
      "loss": 0.3485,
      "step": 810
    },
    {
      "epoch": 0.8761085729642569,
      "grad_norm": 1.8110038058348839,
      "learning_rate": 4.5859689007058896e-07,
      "loss": 0.3517,
      "step": 815
    },
    {
      "epoch": 0.8814834721848965,
      "grad_norm": 1.74966735612597,
      "learning_rate": 4.2014200273832406e-07,
      "loss": 0.3461,
      "step": 820
    },
    {
      "epoch": 0.8868583714055361,
      "grad_norm": 1.887649690807963,
      "learning_rate": 3.8330009214363197e-07,
      "loss": 0.3509,
      "step": 825
    },
    {
      "epoch": 0.8922332706261757,
      "grad_norm": 1.7866587101491602,
      "learning_rate": 3.4808413361125004e-07,
      "loss": 0.3479,
      "step": 830
    },
    {
      "epoch": 0.8976081698468154,
      "grad_norm": 1.851631043751454,
      "learning_rate": 3.1450652982307815e-07,
      "loss": 0.3502,
      "step": 835
    },
    {
      "epoch": 0.902983069067455,
      "grad_norm": 1.8437354756226652,
      "learning_rate": 2.8257910645009935e-07,
      "loss": 0.3475,
      "step": 840
    },
    {
      "epoch": 0.9083579682880946,
      "grad_norm": 1.690251790769186,
      "learning_rate": 2.523131079874963e-07,
      "loss": 0.3581,
      "step": 845
    },
    {
      "epoch": 0.9137328675087342,
      "grad_norm": 1.8644738435275017,
      "learning_rate": 2.2371919379446495e-07,
      "loss": 0.356,
      "step": 850
    },
    {
      "epoch": 0.9191077667293738,
      "grad_norm": 1.703645616104732,
      "learning_rate": 1.9680743434010385e-07,
      "loss": 0.3431,
      "step": 855
    },
    {
      "epoch": 0.9244826659500134,
      "grad_norm": 1.7632832319721445,
      "learning_rate": 1.7158730765669817e-07,
      "loss": 0.3527,
      "step": 860
    },
    {
      "epoch": 0.9298575651706531,
      "grad_norm": 1.778411516430386,
      "learning_rate": 1.480676960016636e-07,
      "loss": 0.3527,
      "step": 865
    },
    {
      "epoch": 0.9352324643912927,
      "grad_norm": 1.7461766662441864,
      "learning_rate": 1.2625688272930925e-07,
      "loss": 0.3459,
      "step": 870
    },
    {
      "epoch": 0.9406073636119323,
      "grad_norm": 1.7323066734927444,
      "learning_rate": 1.0616254937352966e-07,
      "loss": 0.3484,
      "step": 875
    },
    {
      "epoch": 0.9459822628325719,
      "grad_norm": 1.7917085698996813,
      "learning_rate": 8.779177294245044e-08,
      "loss": 0.3545,
      "step": 880
    },
    {
      "epoch": 0.9513571620532115,
      "grad_norm": 1.8063043505678373,
      "learning_rate": 7.115102342598101e-08,
      "loss": 0.3463,
      "step": 885
    },
    {
      "epoch": 0.9567320612738511,
      "grad_norm": 1.7402924935517343,
      "learning_rate": 5.6246161517158336e-08,
      "loss": 0.3443,
      "step": 890
    },
    {
      "epoch": 0.9621069604944907,
      "grad_norm": 1.7772873395565845,
      "learning_rate": 4.308243654806643e-08,
      "loss": 0.3421,
      "step": 895
    },
    {
      "epoch": 0.9674818597151303,
      "grad_norm": 1.7288228417548714,
      "learning_rate": 3.166448464108629e-08,
      "loss": 0.3493,
      "step": 900
    },
    {
      "epoch": 0.97285675893577,
      "grad_norm": 1.7547077783156886,
      "learning_rate": 2.1996327076096446e-08,
      "loss": 0.3504,
      "step": 905
    },
    {
      "epoch": 0.9782316581564096,
      "grad_norm": 1.8190476476140809,
      "learning_rate": 1.4081368874226398e-08,
      "loss": 0.3463,
      "step": 910
    },
    {
      "epoch": 0.9836065573770492,
      "grad_norm": 1.66337852034001,
      "learning_rate": 7.922397598642551e-09,
      "loss": 0.3459,
      "step": 915
    },
    {
      "epoch": 0.9889814565976888,
      "grad_norm": 1.7288826649004094,
      "learning_rate": 3.5215823727974274e-09,
      "loss": 0.3501,
      "step": 920
    },
    {
      "epoch": 0.9943563558183284,
      "grad_norm": 1.7197053184297024,
      "learning_rate": 8.804731164901991e-10,
      "loss": 0.3516,
      "step": 925
    },
    {
      "epoch": 0.999731255038968,
      "grad_norm": 1.7184366358056455,
      "learning_rate": 0.0,
      "loss": 0.3536,
      "step": 930
    },
    {
      "epoch": 0.999731255038968,
      "eval_loss": 0.27914658188819885,
      "eval_runtime": 1.1884,
      "eval_samples_per_second": 1.683,
      "eval_steps_per_second": 0.841,
      "step": 930
    },
    {
      "epoch": 0.999731255038968,
      "step": 930,
      "total_flos": 194670734868480.0,
      "train_loss": 0.5080837889384198,
      "train_runtime": 21164.3008,
      "train_samples_per_second": 1.406,
      "train_steps_per_second": 0.044
    }
  ],
  "logging_steps": 5,
  "max_steps": 930,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 194670734868480.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}