{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 712,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.056179775280898875,
      "grad_norm": 1.6944109649364811,
      "learning_rate": 5.555555555555557e-06,
      "loss": 1.3057,
      "step": 5
    },
    {
      "epoch": 0.11235955056179775,
      "grad_norm": 1.0734576025480336,
      "learning_rate": 1.1111111111111113e-05,
      "loss": 1.2523,
      "step": 10
    },
    {
      "epoch": 0.16853932584269662,
      "grad_norm": 0.9680158843960001,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.1501,
      "step": 15
    },
    {
      "epoch": 0.2247191011235955,
      "grad_norm": 0.8622223090177016,
      "learning_rate": 1.9998445014500246e-05,
      "loss": 1.0298,
      "step": 20
    },
    {
      "epoch": 0.2808988764044944,
      "grad_norm": 0.7503899959880194,
      "learning_rate": 1.998095759792783e-05,
      "loss": 1.0131,
      "step": 25
    },
    {
      "epoch": 0.33707865168539325,
      "grad_norm": 0.5972359554952426,
      "learning_rate": 1.9944076920747774e-05,
      "loss": 0.9485,
      "step": 30
    },
    {
      "epoch": 0.39325842696629215,
      "grad_norm": 0.5365600358115739,
      "learning_rate": 1.98878826221941e-05,
      "loss": 0.9524,
      "step": 35
    },
    {
      "epoch": 0.449438202247191,
      "grad_norm": 0.4710678595191818,
      "learning_rate": 1.9812496046863375e-05,
      "loss": 0.9382,
      "step": 40
    },
    {
      "epoch": 0.5056179775280899,
      "grad_norm": 0.44555673749010527,
      "learning_rate": 1.971807998268615e-05,
      "loss": 0.91,
      "step": 45
    },
    {
      "epoch": 0.5617977528089888,
      "grad_norm": 0.4474821677236859,
      "learning_rate": 1.9604838309406668e-05,
      "loss": 0.8664,
      "step": 50
    },
    {
      "epoch": 0.6179775280898876,
      "grad_norm": 0.46993823824767494,
      "learning_rate": 1.947301555832993e-05,
      "loss": 0.8943,
      "step": 55
    },
    {
      "epoch": 0.6741573033707865,
      "grad_norm": 0.40887692660248126,
      "learning_rate": 1.932289638428676e-05,
      "loss": 0.8427,
      "step": 60
    },
    {
      "epoch": 0.7303370786516854,
      "grad_norm": 0.4178479209299752,
      "learning_rate": 1.9154804950957096e-05,
      "loss": 0.8765,
      "step": 65
    },
    {
      "epoch": 0.7865168539325843,
      "grad_norm": 0.4358655706142232,
      "learning_rate": 1.896910423087889e-05,
      "loss": 0.8815,
      "step": 70
    },
    {
      "epoch": 0.8426966292134831,
      "grad_norm": 0.4759614764741585,
      "learning_rate": 1.8766195221654038e-05,
      "loss": 0.9062,
      "step": 75
    },
    {
      "epoch": 0.898876404494382,
      "grad_norm": 0.40941475449933445,
      "learning_rate": 1.8546516080043956e-05,
      "loss": 0.8525,
      "step": 80
    },
    {
      "epoch": 0.9550561797752809,
      "grad_norm": 0.46163800722604575,
      "learning_rate": 1.831054117582457e-05,
      "loss": 0.8511,
      "step": 85
    },
    {
      "epoch": 1.0112359550561798,
      "grad_norm": 0.4144691948002747,
      "learning_rate": 1.805878006744379e-05,
      "loss": 0.8351,
      "step": 90
    },
    {
      "epoch": 1.0674157303370786,
      "grad_norm": 0.41318674441336783,
      "learning_rate": 1.7791776401693455e-05,
      "loss": 0.7575,
      "step": 95
    },
    {
      "epoch": 1.1235955056179776,
      "grad_norm": 0.4140804142512328,
      "learning_rate": 1.7510106739771698e-05,
      "loss": 0.7629,
      "step": 100
    },
    {
      "epoch": 1.1797752808988764,
      "grad_norm": 0.3888169360375874,
      "learning_rate": 1.7214379312270786e-05,
      "loss": 0.7725,
      "step": 105
    },
    {
      "epoch": 1.2359550561797752,
      "grad_norm": 0.39830672831462005,
      "learning_rate": 1.690523270577883e-05,
      "loss": 0.7957,
      "step": 110
    },
    {
      "epoch": 1.2921348314606742,
      "grad_norm": 0.4025052364368211,
      "learning_rate": 1.6583334483931465e-05,
      "loss": 0.8151,
      "step": 115
    },
    {
      "epoch": 1.348314606741573,
      "grad_norm": 0.40954244286961317,
      "learning_rate": 1.6249379745891277e-05,
      "loss": 0.7625,
      "step": 120
    },
    {
      "epoch": 1.404494382022472,
      "grad_norm": 0.38363991825348154,
      "learning_rate": 1.590408962536759e-05,
      "loss": 0.768,
      "step": 125
    },
    {
      "epoch": 1.4606741573033708,
      "grad_norm": 0.4202167638100542,
      "learning_rate": 1.5548209733417974e-05,
      "loss": 0.7333,
      "step": 130
    },
    {
      "epoch": 1.5168539325842696,
      "grad_norm": 0.39549093654722417,
      "learning_rate": 1.5182508548393917e-05,
      "loss": 0.7561,
      "step": 135
    },
    {
      "epoch": 1.5730337078651684,
      "grad_norm": 0.38558554907475995,
      "learning_rate": 1.4807775756507423e-05,
      "loss": 0.7659,
      "step": 140
    },
    {
      "epoch": 1.6292134831460674,
      "grad_norm": 0.40811098318454503,
      "learning_rate": 1.4424820546601892e-05,
      "loss": 0.7592,
      "step": 145
    },
    {
      "epoch": 1.6853932584269664,
      "grad_norm": 0.4075116805060462,
      "learning_rate": 1.4034469862809474e-05,
      "loss": 0.7738,
      "step": 150
    },
    {
      "epoch": 1.7415730337078652,
      "grad_norm": 0.4178748191013212,
      "learning_rate": 1.363756661886812e-05,
      "loss": 0.7677,
      "step": 155
    },
    {
      "epoch": 1.797752808988764,
      "grad_norm": 0.41587042744707226,
      "learning_rate": 1.32349678779542e-05,
      "loss": 0.7555,
      "step": 160
    },
    {
      "epoch": 1.8539325842696628,
      "grad_norm": 0.4186538870477955,
      "learning_rate": 1.282754300196122e-05,
      "loss": 0.7731,
      "step": 165
    },
    {
      "epoch": 1.9101123595505618,
      "grad_norm": 0.40473262826860434,
      "learning_rate": 1.2416171774220976e-05,
      "loss": 0.7644,
      "step": 170
    },
    {
      "epoch": 1.9662921348314608,
      "grad_norm": 0.38178573402418625,
      "learning_rate": 1.2001742499720888e-05,
      "loss": 0.7687,
      "step": 175
    },
    {
      "epoch": 2.0224719101123596,
      "grad_norm": 0.3819894080787757,
      "learning_rate": 1.1585150086919896e-05,
      "loss": 0.7067,
      "step": 180
    },
    {
      "epoch": 2.0786516853932584,
      "grad_norm": 0.394658925096467,
      "learning_rate": 1.1167294115304992e-05,
      "loss": 0.6749,
      "step": 185
    },
    {
      "epoch": 2.134831460674157,
      "grad_norm": 0.39767908125309154,
      "learning_rate": 1.0749076892861227e-05,
      "loss": 0.6544,
      "step": 190
    },
    {
      "epoch": 2.191011235955056,
      "grad_norm": 0.42673180505937613,
      "learning_rate": 1.0331401507649868e-05,
      "loss": 0.6896,
      "step": 195
    },
    {
      "epoch": 2.247191011235955,
      "grad_norm": 0.408239645067392,
      "learning_rate": 9.915169877702096e-06,
      "loss": 0.726,
      "step": 200
    },
    {
      "epoch": 2.303370786516854,
      "grad_norm": 0.39464002130518294,
      "learning_rate": 9.501280803439204e-06,
      "loss": 0.6655,
      "step": 205
    },
    {
      "epoch": 2.359550561797753,
      "grad_norm": 0.40978323940865435,
      "learning_rate": 9.090628026824941e-06,
      "loss": 0.6764,
      "step": 210
    },
    {
      "epoch": 2.4157303370786516,
      "grad_norm": 0.38292295583739444,
      "learning_rate": 8.684098301440903e-06,
      "loss": 0.6719,
      "step": 215
    },
    {
      "epoch": 2.4719101123595504,
      "grad_norm": 0.38693445253625824,
      "learning_rate": 8.2825694776525e-06,
      "loss": 0.6584,
      "step": 220
    },
    {
      "epoch": 2.5280898876404496,
      "grad_norm": 0.3994341136157858,
      "learning_rate": 7.886908607000321e-06,
      "loss": 0.6971,
      "step": 225
    },
    {
      "epoch": 2.5842696629213484,
      "grad_norm": 0.40129558631958756,
      "learning_rate": 7.497970069910192e-06,
      "loss": 0.6747,
      "step": 230
    },
    {
      "epoch": 2.640449438202247,
      "grad_norm": 0.39014270480667235,
      "learning_rate": 7.116593730764929e-06,
      "loss": 0.6796,
      "step": 235
    },
    {
      "epoch": 2.696629213483146,
      "grad_norm": 0.36998791500930744,
      "learning_rate": 6.743603124321712e-06,
      "loss": 0.6902,
      "step": 240
    },
    {
      "epoch": 2.752808988764045,
      "grad_norm": 0.3970482683621501,
      "learning_rate": 6.379803677391223e-06,
      "loss": 0.6934,
      "step": 245
    },
    {
      "epoch": 2.808988764044944,
      "grad_norm": 0.3861644580574831,
      "learning_rate": 6.02598096961865e-06,
      "loss": 0.6614,
      "step": 250
    },
    {
      "epoch": 2.865168539325843,
      "grad_norm": 0.3941319315860406,
      "learning_rate": 5.682899037122178e-06,
      "loss": 0.6839,
      "step": 255
    },
    {
      "epoch": 2.9213483146067416,
      "grad_norm": 0.3898881332551341,
      "learning_rate": 5.351298722652064e-06,
      "loss": 0.6667,
      "step": 260
    },
    {
      "epoch": 2.9775280898876404,
      "grad_norm": 0.40076378703515086,
      "learning_rate": 5.031896075832846e-06,
      "loss": 0.6708,
      "step": 265
    },
    {
      "epoch": 3.033707865168539,
      "grad_norm": 0.41791394545484606,
      "learning_rate": 4.725380806943299e-06,
      "loss": 0.6575,
      "step": 270
    },
    {
      "epoch": 3.0898876404494384,
      "grad_norm": 0.41375267265638194,
      "learning_rate": 4.432414797572894e-06,
      "loss": 0.6318,
      "step": 275
    },
    {
      "epoch": 3.146067415730337,
      "grad_norm": 0.4218475144549854,
      "learning_rate": 4.153630671370821e-06,
      "loss": 0.6342,
      "step": 280
    },
    {
      "epoch": 3.202247191011236,
      "grad_norm": 0.4164065437650473,
      "learning_rate": 3.889630427973951e-06,
      "loss": 0.6274,
      "step": 285
    },
    {
      "epoch": 3.258426966292135,
      "grad_norm": 0.41742766148238786,
      "learning_rate": 3.6409841430635166e-06,
      "loss": 0.6078,
      "step": 290
    },
    {
      "epoch": 3.3146067415730336,
      "grad_norm": 0.4472934678743176,
      "learning_rate": 3.408228737357575e-06,
      "loss": 0.6137,
      "step": 295
    },
    {
      "epoch": 3.370786516853933,
      "grad_norm": 0.3983175026213215,
      "learning_rate": 3.191866817197539e-06,
      "loss": 0.6268,
      "step": 300
    },
    {
      "epoch": 3.4269662921348316,
      "grad_norm": 0.41447625127416166,
      "learning_rate": 2.9923655892323144e-06,
      "loss": 0.635,
      "step": 305
    },
    {
      "epoch": 3.4831460674157304,
      "grad_norm": 0.3967828745027063,
      "learning_rate": 2.8101558515436506e-06,
      "loss": 0.6181,
      "step": 310
    },
    {
      "epoch": 3.539325842696629,
      "grad_norm": 0.39144720476319333,
      "learning_rate": 2.645631063391285e-06,
      "loss": 0.6002,
      "step": 315
    },
    {
      "epoch": 3.595505617977528,
      "grad_norm": 0.40314596605224257,
      "learning_rate": 2.4991464955866314e-06,
      "loss": 0.6046,
      "step": 320
    },
    {
      "epoch": 3.6516853932584272,
      "grad_norm": 0.3979575371097757,
      "learning_rate": 2.371018463329651e-06,
      "loss": 0.6091,
      "step": 325
    },
    {
      "epoch": 3.7078651685393256,
      "grad_norm": 0.4108961213792489,
      "learning_rate": 2.261523643165532e-06,
      "loss": 0.6445,
      "step": 330
    },
    {
      "epoch": 3.764044943820225,
      "grad_norm": 0.40810633042253025,
      "learning_rate": 2.1708984755361205e-06,
      "loss": 0.6311,
      "step": 335
    },
    {
      "epoch": 3.8202247191011236,
      "grad_norm": 0.39820144430518867,
      "learning_rate": 2.0993386542161944e-06,
      "loss": 0.6154,
      "step": 340
    },
    {
      "epoch": 3.8764044943820224,
      "grad_norm": 0.4070293846963812,
      "learning_rate": 2.0469987037371005e-06,
      "loss": 0.6193,
      "step": 345
    },
    {
      "epoch": 3.932584269662921,
      "grad_norm": 0.40297827384808366,
      "learning_rate": 2.013991645710262e-06,
      "loss": 0.6471,
      "step": 350
    },
    {
      "epoch": 3.98876404494382,
      "grad_norm": 0.4066161050267969,
      "learning_rate": 2.0003887547710647e-06,
      "loss": 0.6469,
      "step": 355
    },
    {
      "epoch": 4.044943820224719,
      "grad_norm": 0.4074217412543327,
      "learning_rate": 1.1585150086919896e-05,
      "loss": 0.6009,
      "step": 360
    },
    {
      "epoch": 4.101123595505618,
      "grad_norm": 0.4271665264127198,
      "learning_rate": 1.137632369284973e-05,
      "loss": 0.6207,
      "step": 365
    },
    {
      "epoch": 4.157303370786517,
      "grad_norm": 0.43109911858630096,
      "learning_rate": 1.1167294115304992e-05,
      "loss": 0.5911,
      "step": 370
    },
    {
      "epoch": 4.213483146067416,
      "grad_norm": 0.43839912783355806,
      "learning_rate": 1.0958174212861062e-05,
      "loss": 0.6011,
      "step": 375
    },
    {
      "epoch": 4.269662921348314,
      "grad_norm": 0.4190687745455898,
      "learning_rate": 1.0749076892861227e-05,
      "loss": 0.5983,
      "step": 380
    },
    {
      "epoch": 4.325842696629214,
      "grad_norm": 0.4272292754884261,
      "learning_rate": 1.0540115050456152e-05,
      "loss": 0.6298,
      "step": 385
    },
    {
      "epoch": 4.382022471910112,
      "grad_norm": 0.4188166195248199,
      "learning_rate": 1.0331401507649868e-05,
      "loss": 0.614,
      "step": 390
    },
    {
      "epoch": 4.438202247191011,
      "grad_norm": 0.46031386712411543,
      "learning_rate": 1.012304895238529e-05,
      "loss": 0.5934,
      "step": 395
    },
    {
      "epoch": 4.49438202247191,
      "grad_norm": 0.4198160623739865,
      "learning_rate": 9.915169877702096e-06,
      "loss": 0.5816,
      "step": 400
    },
    {
      "epoch": 4.550561797752809,
      "grad_norm": 0.41877479862746064,
      "learning_rate": 9.707876520999864e-06,
      "loss": 0.6071,
      "step": 405
    },
    {
      "epoch": 4.606741573033708,
      "grad_norm": 0.445785010864897,
      "learning_rate": 9.501280803439204e-06,
      "loss": 0.5816,
      "step": 410
    },
    {
      "epoch": 4.662921348314606,
      "grad_norm": 0.4481588101026953,
      "learning_rate": 9.295494269513659e-06,
      "loss": 0.6006,
      "step": 415
    },
    {
      "epoch": 4.719101123595506,
      "grad_norm": 0.4260960071459342,
      "learning_rate": 9.090628026824941e-06,
      "loss": 0.6357,
      "step": 420
    },
    {
      "epoch": 4.775280898876405,
      "grad_norm": 0.4192037207998735,
      "learning_rate": 8.886792686094096e-06,
      "loss": 0.6006,
      "step": 425
    },
    {
      "epoch": 4.831460674157303,
      "grad_norm": 0.43566168980415704,
      "learning_rate": 8.684098301440903e-06,
      "loss": 0.6267,
      "step": 430
    },
    {
      "epoch": 4.887640449438202,
      "grad_norm": 0.42117325698617175,
      "learning_rate": 8.482654310963817e-06,
      "loss": 0.595,
      "step": 435
    },
    {
      "epoch": 4.943820224719101,
      "grad_norm": 0.4414097385180421,
      "learning_rate": 8.2825694776525e-06,
      "loss": 0.5871,
      "step": 440
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.41448225752045526,
      "learning_rate": 8.083951830664867e-06,
      "loss": 0.5838,
      "step": 445
    },
    {
      "epoch": 5.056179775280899,
      "grad_norm": 0.44467080989005614,
      "learning_rate": 7.886908607000321e-06,
      "loss": 0.5612,
      "step": 450
    },
    {
      "epoch": 5.112359550561798,
      "grad_norm": 0.48267692647228083,
      "learning_rate": 7.691546193600702e-06,
      "loss": 0.5438,
      "step": 455
    },
    {
      "epoch": 5.168539325842697,
      "grad_norm": 0.46882770217949304,
      "learning_rate": 7.497970069910192e-06,
      "loss": 0.5145,
      "step": 460
    },
    {
      "epoch": 5.224719101123595,
      "grad_norm": 0.45016082689033277,
      "learning_rate": 7.306284750925192e-06,
      "loss": 0.5392,
      "step": 465
    },
    {
      "epoch": 5.280898876404494,
      "grad_norm": 0.4904333208280953,
      "learning_rate": 7.116593730764929e-06,
      "loss": 0.5172,
      "step": 470
    },
    {
      "epoch": 5.337078651685394,
      "grad_norm": 0.5047681974857045,
      "learning_rate": 6.928999426793234e-06,
      "loss": 0.5408,
      "step": 475
    },
    {
      "epoch": 5.393258426966292,
      "grad_norm": 0.46939981998911995,
      "learning_rate": 6.743603124321712e-06,
      "loss": 0.5408,
      "step": 480
    },
    {
      "epoch": 5.449438202247191,
      "grad_norm": 0.48583542280728176,
      "learning_rate": 6.5605049219240635e-06,
      "loss": 0.5564,
      "step": 485
    },
    {
      "epoch": 5.50561797752809,
      "grad_norm": 0.445507769617856,
      "learning_rate": 6.379803677391223e-06,
      "loss": 0.5272,
      "step": 490
    },
    {
      "epoch": 5.561797752808989,
      "grad_norm": 0.48127792996455504,
      "learning_rate": 6.201596954356362e-06,
      "loss": 0.5221,
      "step": 495
    },
    {
      "epoch": 5.617977528089888,
      "grad_norm": 0.48763800709480937,
      "learning_rate": 6.02598096961865e-06,
      "loss": 0.5017,
      "step": 500
    },
    {
      "epoch": 5.674157303370786,
      "grad_norm": 0.45141538643328166,
      "learning_rate": 5.853050541194187e-06,
      "loss": 0.5301,
      "step": 505
    },
    {
      "epoch": 5.730337078651686,
      "grad_norm": 0.46888316703181054,
      "learning_rate": 5.682899037122178e-06,
      "loss": 0.5088,
      "step": 510
    },
    {
      "epoch": 5.786516853932584,
      "grad_norm": 0.49578196304683,
      "learning_rate": 5.515618325053952e-06,
      "loss": 0.5051,
      "step": 515
    },
    {
      "epoch": 5.842696629213483,
      "grad_norm": 0.4766026956952897,
      "learning_rate": 5.351298722652064e-06,
      "loss": 0.509,
      "step": 520
    },
    {
      "epoch": 5.898876404494382,
      "grad_norm": 0.5255442360578876,
      "learning_rate": 5.190028948826304e-06,
      "loss": 0.5465,
      "step": 525
    },
    {
      "epoch": 5.955056179775281,
      "grad_norm": 0.5102538127102871,
      "learning_rate": 5.031896075832846e-06,
      "loss": 0.5186,
      "step": 530
    },
    {
      "epoch": 6.01123595505618,
      "grad_norm": 0.5022519456849229,
      "learning_rate": 4.876985482262482e-06,
      "loss": 0.5204,
      "step": 535
    },
    {
      "epoch": 6.067415730337078,
      "grad_norm": 0.46834498920984025,
      "learning_rate": 4.725380806943299e-06,
      "loss": 0.4707,
      "step": 540
    },
    {
      "epoch": 6.123595505617978,
      "grad_norm": 0.5403111606747254,
      "learning_rate": 4.577163903782655e-06,
      "loss": 0.4491,
      "step": 545
    },
    {
      "epoch": 6.179775280898877,
      "grad_norm": 0.5964108863972761,
      "learning_rate": 4.432414797572894e-06,
      "loss": 0.4709,
      "step": 550
    },
    {
      "epoch": 6.235955056179775,
      "grad_norm": 0.4994302712693112,
      "learning_rate": 4.291211640784608e-06,
      "loss": 0.4708,
      "step": 555
    },
    {
      "epoch": 6.292134831460674,
      "grad_norm": 0.49243733113864285,
      "learning_rate": 4.153630671370821e-06,
      "loss": 0.4926,
      "step": 560
    },
    {
      "epoch": 6.348314606741573,
      "grad_norm": 0.49658882787661995,
      "learning_rate": 4.019746171604824e-06,
      "loss": 0.502,
      "step": 565
    },
    {
      "epoch": 6.404494382022472,
      "grad_norm": 0.5418840142061598,
      "learning_rate": 3.889630427973951e-06,
      "loss": 0.4602,
      "step": 570
    },
    {
      "epoch": 6.460674157303371,
      "grad_norm": 0.5008443330249364,
      "learning_rate": 3.763353692150864e-06,
      "loss": 0.4665,
      "step": 575
    },
    {
      "epoch": 6.51685393258427,
      "grad_norm": 0.48589030681678613,
      "learning_rate": 3.6409841430635166e-06,
      "loss": 0.4497,
      "step": 580
    },
    {
      "epoch": 6.573033707865169,
      "grad_norm": 0.5820871330261026,
      "learning_rate": 3.522587850084197e-06,
      "loss": 0.4629,
      "step": 585
    },
    {
      "epoch": 6.629213483146067,
      "grad_norm": 0.49834548702881326,
      "learning_rate": 3.408228737357575e-06,
      "loss": 0.483,
      "step": 590
    },
    {
      "epoch": 6.685393258426966,
      "grad_norm": 0.5052799618260936,
      "learning_rate": 3.297968549286974e-06,
      "loss": 0.4808,
      "step": 595
    },
    {
      "epoch": 6.741573033707866,
      "grad_norm": 0.492591898344833,
      "learning_rate": 3.191866817197539e-06,
      "loss": 0.472,
      "step": 600
    },
    {
      "epoch": 6.797752808988764,
      "grad_norm": 0.49032082408314204,
      "learning_rate": 3.089980827194276e-06,
      "loss": 0.4611,
      "step": 605
    },
    {
      "epoch": 6.853932584269663,
      "grad_norm": 0.4959126643842184,
      "learning_rate": 2.9923655892323144e-06,
      "loss": 0.4726,
      "step": 610
    },
    {
      "epoch": 6.910112359550562,
      "grad_norm": 0.5335774103596451,
      "learning_rate": 2.8990738074161196e-06,
      "loss": 0.4792,
      "step": 615
    },
    {
      "epoch": 6.966292134831461,
      "grad_norm": 0.5201706543464556,
      "learning_rate": 2.8101558515436506e-06,
      "loss": 0.4711,
      "step": 620
    },
    {
      "epoch": 7.022471910112359,
      "grad_norm": 0.4623475019520009,
      "learning_rate": 2.725659729910878e-06,
      "loss": 0.4545,
      "step": 625
    },
    {
      "epoch": 7.078651685393258,
      "grad_norm": 0.4834608390252829,
      "learning_rate": 2.645631063391285e-06,
      "loss": 0.443,
      "step": 630
    },
    {
      "epoch": 7.134831460674158,
      "grad_norm": 0.5359592571307704,
      "learning_rate": 2.570113060804401e-06,
      "loss": 0.441,
      "step": 635
    },
    {
      "epoch": 7.191011235955056,
      "grad_norm": 0.5433362481606256,
      "learning_rate": 2.4991464955866314e-06,
      "loss": 0.4264,
      "step": 640
    },
    {
      "epoch": 7.247191011235955,
      "grad_norm": 0.5225953610583668,
      "learning_rate": 2.432769683776995e-06,
      "loss": 0.4145,
      "step": 645
    },
    {
      "epoch": 7.303370786516854,
      "grad_norm": 0.5205000317985969,
      "learning_rate": 2.371018463329651e-06,
      "loss": 0.435,
      "step": 650
    },
    {
      "epoch": 7.359550561797753,
      "grad_norm": 0.5007819526288158,
      "learning_rate": 2.3139261747643925e-06,
      "loss": 0.41,
      "step": 655
    },
    {
      "epoch": 7.415730337078652,
      "grad_norm": 0.518942310039327,
      "learning_rate": 2.261523643165532e-06,
      "loss": 0.4502,
      "step": 660
    },
    {
      "epoch": 7.47191011235955,
      "grad_norm": 0.49873700284836797,
      "learning_rate": 2.2138391615389306e-06,
      "loss": 0.4324,
      "step": 665
    },
    {
      "epoch": 7.52808988764045,
      "grad_norm": 0.5208414783528387,
      "learning_rate": 2.1708984755361205e-06,
      "loss": 0.4449,
      "step": 670
    },
    {
      "epoch": 7.584269662921348,
      "grad_norm": 0.5791900845785815,
      "learning_rate": 2.1327247695538015e-06,
      "loss": 0.4484,
      "step": 675
    },
    {
      "epoch": 7.640449438202247,
      "grad_norm": 0.5438940813210439,
      "learning_rate": 2.0993386542161944e-06,
      "loss": 0.4685,
      "step": 680
    },
    {
      "epoch": 7.696629213483146,
      "grad_norm": 0.5127534158556458,
      "learning_rate": 2.070758155247017e-06,
      "loss": 0.4339,
      "step": 685
    },
    {
      "epoch": 7.752808988764045,
      "grad_norm": 0.5415209164546988,
      "learning_rate": 2.0469987037371005e-06,
      "loss": 0.4103,
      "step": 690
    },
    {
      "epoch": 7.808988764044944,
      "grad_norm": 0.5136419675345572,
      "learning_rate": 2.028073127812876e-06,
      "loss": 0.4169,
      "step": 695
    },
    {
      "epoch": 7.865168539325842,
      "grad_norm": 0.5319921953080505,
      "learning_rate": 2.013991645710262e-06,
      "loss": 0.4286,
      "step": 700
    },
    {
      "epoch": 7.921348314606742,
      "grad_norm": 0.5733506841186511,
      "learning_rate": 2.0047618602576594e-06,
      "loss": 0.4505,
      "step": 705
    },
    {
      "epoch": 7.97752808988764,
      "grad_norm": 0.5142139821667071,
      "learning_rate": 2.0003887547710647e-06,
      "loss": 0.4202,
      "step": 710
    },
    {
      "epoch": 8.0,
      "step": 712,
      "total_flos": 40566502981632.0,
      "train_loss": 0.25444099340545995,
      "train_runtime": 2064.0536,
      "train_samples_per_second": 22.042,
      "train_steps_per_second": 0.345
    }
  ],
  "logging_steps": 5,
  "max_steps": 712,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 40566502981632.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}