{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 2226,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.013490725126475547,
      "grad_norm": 65.45332336425781,
      "learning_rate": 0.00019964061096136568,
      "loss": 22.1843,
      "step": 10
    },
    {
      "epoch": 0.026981450252951095,
      "grad_norm": 6.102770805358887,
      "learning_rate": 0.00019874213836477988,
      "loss": 7.6527,
      "step": 20
    },
    {
      "epoch": 0.04047217537942664,
      "grad_norm": 2.982379913330078,
      "learning_rate": 0.00019784366576819408,
      "loss": 3.4532,
      "step": 30
    },
    {
      "epoch": 0.05396290050590219,
      "grad_norm": 3.4628562927246094,
      "learning_rate": 0.00019694519317160826,
      "loss": 2.4467,
      "step": 40
    },
    {
      "epoch": 0.06745362563237774,
      "grad_norm": 3.0008349418640137,
      "learning_rate": 0.00019604672057502246,
      "loss": 1.6481,
      "step": 50
    },
    {
      "epoch": 0.08094435075885328,
      "grad_norm": 1.767865777015686,
      "learning_rate": 0.00019514824797843666,
      "loss": 1.2119,
      "step": 60
    },
    {
      "epoch": 0.09443507588532883,
      "grad_norm": 3.416968584060669,
      "learning_rate": 0.00019424977538185087,
      "loss": 0.967,
      "step": 70
    },
    {
      "epoch": 0.10792580101180438,
      "grad_norm": 3.718262195587158,
      "learning_rate": 0.00019335130278526507,
      "loss": 0.8783,
      "step": 80
    },
    {
      "epoch": 0.12141652613827993,
      "grad_norm": 3.689655065536499,
      "learning_rate": 0.00019245283018867927,
      "loss": 0.7619,
      "step": 90
    },
    {
      "epoch": 0.13490725126475547,
      "grad_norm": 6.728384017944336,
      "learning_rate": 0.00019155435759209347,
      "loss": 0.7244,
      "step": 100
    },
    {
      "epoch": 0.14839797639123103,
      "grad_norm": 4.880130290985107,
      "learning_rate": 0.00019065588499550765,
      "loss": 0.6931,
      "step": 110
    },
    {
      "epoch": 0.16188870151770657,
      "grad_norm": 10.14767074584961,
      "learning_rate": 0.00018975741239892185,
      "loss": 0.6995,
      "step": 120
    },
    {
      "epoch": 0.17537942664418213,
      "grad_norm": 10.897403717041016,
      "learning_rate": 0.00018885893980233605,
      "loss": 0.69,
      "step": 130
    },
    {
      "epoch": 0.18887015177065766,
      "grad_norm": 7.759915351867676,
      "learning_rate": 0.00018796046720575023,
      "loss": 0.6899,
      "step": 140
    },
    {
      "epoch": 0.20236087689713322,
      "grad_norm": 7.006256580352783,
      "learning_rate": 0.00018706199460916443,
      "loss": 0.6739,
      "step": 150
    },
    {
      "epoch": 0.21585160202360876,
      "grad_norm": 5.777213096618652,
      "learning_rate": 0.00018616352201257863,
      "loss": 0.6731,
      "step": 160
    },
    {
      "epoch": 0.22934232715008432,
      "grad_norm": 10.224111557006836,
      "learning_rate": 0.00018526504941599283,
      "loss": 0.6417,
      "step": 170
    },
    {
      "epoch": 0.24283305227655985,
      "grad_norm": 6.332294464111328,
      "learning_rate": 0.000184366576819407,
      "loss": 0.6452,
      "step": 180
    },
    {
      "epoch": 0.2563237774030354,
      "grad_norm": 3.9472005367279053,
      "learning_rate": 0.0001834681042228212,
      "loss": 0.6271,
      "step": 190
    },
    {
      "epoch": 0.26981450252951095,
      "grad_norm": 6.640835762023926,
      "learning_rate": 0.0001825696316262354,
      "loss": 0.6161,
      "step": 200
    },
    {
      "epoch": 0.28330522765598654,
      "grad_norm": 5.6504316329956055,
      "learning_rate": 0.00018167115902964959,
      "loss": 0.6452,
      "step": 210
    },
    {
      "epoch": 0.29679595278246207,
      "grad_norm": 3.4219472408294678,
      "learning_rate": 0.0001807726864330638,
      "loss": 0.6228,
      "step": 220
    },
    {
      "epoch": 0.3102866779089376,
      "grad_norm": 7.479050636291504,
      "learning_rate": 0.000179874213836478,
      "loss": 0.6035,
      "step": 230
    },
    {
      "epoch": 0.32377740303541314,
      "grad_norm": 6.84128475189209,
      "learning_rate": 0.0001789757412398922,
      "loss": 0.5763,
      "step": 240
    },
    {
      "epoch": 0.3372681281618887,
      "grad_norm": 3.407183885574341,
      "learning_rate": 0.0001780772686433064,
      "loss": 0.5981,
      "step": 250
    },
    {
      "epoch": 0.35075885328836426,
      "grad_norm": 6.634527683258057,
      "learning_rate": 0.0001771787960467206,
      "loss": 0.6313,
      "step": 260
    },
    {
      "epoch": 0.3642495784148398,
      "grad_norm": 6.274175643920898,
      "learning_rate": 0.0001762803234501348,
      "loss": 0.5971,
      "step": 270
    },
    {
      "epoch": 0.3777403035413153,
      "grad_norm": 5.464880466461182,
      "learning_rate": 0.00017538185085354897,
      "loss": 0.629,
      "step": 280
    },
    {
      "epoch": 0.3912310286677909,
      "grad_norm": 7.9421234130859375,
      "learning_rate": 0.00017448337825696318,
      "loss": 0.5604,
      "step": 290
    },
    {
      "epoch": 0.40472175379426645,
      "grad_norm": 6.513714790344238,
      "learning_rate": 0.00017358490566037738,
      "loss": 0.5442,
      "step": 300
    },
    {
      "epoch": 0.418212478920742,
      "grad_norm": 4.856393337249756,
      "learning_rate": 0.00017268643306379155,
      "loss": 0.5444,
      "step": 310
    },
    {
      "epoch": 0.4317032040472175,
      "grad_norm": 7.252712249755859,
      "learning_rate": 0.00017178796046720576,
      "loss": 0.6071,
      "step": 320
    },
    {
      "epoch": 0.4451939291736931,
      "grad_norm": 6.8953328132629395,
      "learning_rate": 0.00017088948787061996,
      "loss": 0.5622,
      "step": 330
    },
    {
      "epoch": 0.45868465430016864,
      "grad_norm": 6.220505237579346,
      "learning_rate": 0.00016999101527403416,
      "loss": 0.6213,
      "step": 340
    },
    {
      "epoch": 0.47217537942664417,
      "grad_norm": 1.9960452318191528,
      "learning_rate": 0.00016909254267744833,
      "loss": 0.5752,
      "step": 350
    },
    {
      "epoch": 0.4856661045531197,
      "grad_norm": 4.8189778327941895,
      "learning_rate": 0.00016819407008086254,
      "loss": 0.5695,
      "step": 360
    },
    {
      "epoch": 0.4991568296795953,
      "grad_norm": 2.157482624053955,
      "learning_rate": 0.00016729559748427674,
      "loss": 0.5818,
      "step": 370
    },
    {
      "epoch": 0.5126475548060708,
      "grad_norm": 6.034278392791748,
      "learning_rate": 0.00016639712488769091,
      "loss": 0.5632,
      "step": 380
    },
    {
      "epoch": 0.5261382799325464,
      "grad_norm": 3.6643030643463135,
      "learning_rate": 0.00016549865229110512,
      "loss": 0.5149,
      "step": 390
    },
    {
      "epoch": 0.5396290050590219,
      "grad_norm": 3.4874682426452637,
      "learning_rate": 0.00016460017969451932,
      "loss": 0.5712,
      "step": 400
    },
    {
      "epoch": 0.5531197301854974,
      "grad_norm": 5.94505500793457,
      "learning_rate": 0.00016370170709793352,
      "loss": 0.5579,
      "step": 410
    },
    {
      "epoch": 0.5666104553119731,
      "grad_norm": 2.6858978271484375,
      "learning_rate": 0.00016280323450134772,
      "loss": 0.5601,
      "step": 420
    },
    {
      "epoch": 0.5801011804384486,
      "grad_norm": 0.5919632315635681,
      "learning_rate": 0.00016190476190476192,
      "loss": 0.5122,
      "step": 430
    },
    {
      "epoch": 0.5935919055649241,
      "grad_norm": 9.723355293273926,
      "learning_rate": 0.00016100628930817613,
      "loss": 0.5408,
      "step": 440
    },
    {
      "epoch": 0.6070826306913997,
      "grad_norm": 3.421534776687622,
      "learning_rate": 0.0001601078167115903,
      "loss": 0.5125,
      "step": 450
    },
    {
      "epoch": 0.6205733558178752,
      "grad_norm": 1.350819706916809,
      "learning_rate": 0.0001592093441150045,
      "loss": 0.5443,
      "step": 460
    },
    {
      "epoch": 0.6340640809443507,
      "grad_norm": 2.9618473052978516,
      "learning_rate": 0.0001583108715184187,
      "loss": 0.5792,
      "step": 470
    },
    {
      "epoch": 0.6475548060708263,
      "grad_norm": 11.16501522064209,
      "learning_rate": 0.00015741239892183288,
      "loss": 0.5499,
      "step": 480
    },
    {
      "epoch": 0.6610455311973018,
      "grad_norm": 2.7122786045074463,
      "learning_rate": 0.00015651392632524708,
      "loss": 0.589,
      "step": 490
    },
    {
      "epoch": 0.6745362563237775,
      "grad_norm": 6.2460408210754395,
      "learning_rate": 0.00015561545372866128,
      "loss": 0.5422,
      "step": 500
    },
    {
      "epoch": 0.688026981450253,
      "grad_norm": 12.943184852600098,
      "learning_rate": 0.0001547169811320755,
      "loss": 0.5431,
      "step": 510
    },
    {
      "epoch": 0.7015177065767285,
      "grad_norm": 2.4169883728027344,
      "learning_rate": 0.00015381850853548966,
      "loss": 0.4984,
      "step": 520
    },
    {
      "epoch": 0.715008431703204,
      "grad_norm": 5.108154773712158,
      "learning_rate": 0.00015292003593890386,
      "loss": 0.4795,
      "step": 530
    },
    {
      "epoch": 0.7284991568296796,
      "grad_norm": 3.268669605255127,
      "learning_rate": 0.00015202156334231807,
      "loss": 0.4961,
      "step": 540
    },
    {
      "epoch": 0.7419898819561551,
      "grad_norm": 1.7926982641220093,
      "learning_rate": 0.00015112309074573224,
      "loss": 0.5339,
      "step": 550
    },
    {
      "epoch": 0.7554806070826307,
      "grad_norm": 3.6382663249969482,
      "learning_rate": 0.00015022461814914644,
      "loss": 0.4855,
      "step": 560
    },
    {
      "epoch": 0.7689713322091062,
      "grad_norm": 8.025957107543945,
      "learning_rate": 0.00014932614555256067,
      "loss": 0.4901,
      "step": 570
    },
    {
      "epoch": 0.7824620573355818,
      "grad_norm": 4.220882892608643,
      "learning_rate": 0.00014842767295597485,
      "loss": 0.5448,
      "step": 580
    },
    {
      "epoch": 0.7959527824620574,
      "grad_norm": 6.42817497253418,
      "learning_rate": 0.00014752920035938905,
      "loss": 0.4969,
      "step": 590
    },
    {
      "epoch": 0.8094435075885329,
      "grad_norm": 4.182610988616943,
      "learning_rate": 0.00014663072776280325,
      "loss": 0.4863,
      "step": 600
    },
    {
      "epoch": 0.8229342327150084,
      "grad_norm": 3.687934637069702,
      "learning_rate": 0.00014573225516621745,
      "loss": 0.5162,
      "step": 610
    },
    {
      "epoch": 0.836424957841484,
      "grad_norm": 3.8240537643432617,
      "learning_rate": 0.00014483378256963163,
      "loss": 0.5089,
      "step": 620
    },
    {
      "epoch": 0.8499156829679595,
      "grad_norm": 3.3823423385620117,
      "learning_rate": 0.00014393530997304583,
      "loss": 0.5447,
      "step": 630
    },
    {
      "epoch": 0.863406408094435,
      "grad_norm": 4.591256618499756,
      "learning_rate": 0.00014303683737646003,
      "loss": 0.5067,
      "step": 640
    },
    {
      "epoch": 0.8768971332209107,
      "grad_norm": 4.414494514465332,
      "learning_rate": 0.0001421383647798742,
      "loss": 0.4858,
      "step": 650
    },
    {
      "epoch": 0.8903878583473862,
      "grad_norm": 3.4649970531463623,
      "learning_rate": 0.0001412398921832884,
      "loss": 0.5243,
      "step": 660
    },
    {
      "epoch": 0.9038785834738617,
      "grad_norm": 4.813696384429932,
      "learning_rate": 0.0001403414195867026,
      "loss": 0.51,
      "step": 670
    },
    {
      "epoch": 0.9173693086003373,
      "grad_norm": 4.597005844116211,
      "learning_rate": 0.00013944294699011681,
      "loss": 0.5079,
      "step": 680
    },
    {
      "epoch": 0.9308600337268128,
      "grad_norm": 4.291003704071045,
      "learning_rate": 0.000138544474393531,
      "loss": 0.4966,
      "step": 690
    },
    {
      "epoch": 0.9443507588532883,
      "grad_norm": 3.8949058055877686,
      "learning_rate": 0.0001376460017969452,
      "loss": 0.5086,
      "step": 700
    },
    {
      "epoch": 0.9578414839797639,
      "grad_norm": 4.685634613037109,
      "learning_rate": 0.0001367475292003594,
      "loss": 0.5005,
      "step": 710
    },
    {
      "epoch": 0.9713322091062394,
      "grad_norm": 3.9277801513671875,
      "learning_rate": 0.0001358490566037736,
      "loss": 0.4759,
      "step": 720
    },
    {
      "epoch": 0.984822934232715,
      "grad_norm": 4.3011345863342285,
      "learning_rate": 0.0001349505840071878,
      "loss": 0.4371,
      "step": 730
    },
    {
      "epoch": 0.9983136593591906,
      "grad_norm": 3.370706796646118,
      "learning_rate": 0.000134052111410602,
      "loss": 0.4811,
      "step": 740
    },
    {
      "epoch": 1.0107925801011803,
      "grad_norm": 5.607315540313721,
      "learning_rate": 0.00013315363881401617,
      "loss": 0.4648,
      "step": 750
    },
    {
      "epoch": 1.024283305227656,
      "grad_norm": 3.9047558307647705,
      "learning_rate": 0.00013225516621743038,
      "loss": 0.4813,
      "step": 760
    },
    {
      "epoch": 1.0377740303541316,
      "grad_norm": 3.686000108718872,
      "learning_rate": 0.00013135669362084458,
      "loss": 0.5189,
      "step": 770
    },
    {
      "epoch": 1.0512647554806072,
      "grad_norm": 3.8528940677642822,
      "learning_rate": 0.00013045822102425878,
      "loss": 0.4761,
      "step": 780
    },
    {
      "epoch": 1.0647554806070827,
      "grad_norm": 3.5714054107666016,
      "learning_rate": 0.00012955974842767296,
      "loss": 0.4845,
      "step": 790
    },
    {
      "epoch": 1.0782462057335582,
      "grad_norm": 3.3482840061187744,
      "learning_rate": 0.00012866127583108716,
      "loss": 0.487,
      "step": 800
    },
    {
      "epoch": 1.0917369308600338,
      "grad_norm": 4.3598127365112305,
      "learning_rate": 0.00012776280323450136,
      "loss": 0.4918,
      "step": 810
    },
    {
      "epoch": 1.1052276559865093,
      "grad_norm": 3.4356210231781006,
      "learning_rate": 0.00012686433063791554,
      "loss": 0.5042,
      "step": 820
    },
    {
      "epoch": 1.1187183811129848,
      "grad_norm": 3.015895366668701,
      "learning_rate": 0.00012596585804132974,
      "loss": 0.4911,
      "step": 830
    },
    {
      "epoch": 1.1322091062394604,
      "grad_norm": 3.4911601543426514,
      "learning_rate": 0.00012506738544474394,
      "loss": 0.507,
      "step": 840
    },
    {
      "epoch": 1.145699831365936,
      "grad_norm": 5.718015670776367,
      "learning_rate": 0.00012416891284815814,
      "loss": 0.4668,
      "step": 850
    },
    {
      "epoch": 1.1591905564924114,
      "grad_norm": 4.448844909667969,
      "learning_rate": 0.00012327044025157232,
      "loss": 0.465,
      "step": 860
    },
    {
      "epoch": 1.172681281618887,
      "grad_norm": 3.243500232696533,
      "learning_rate": 0.00012237196765498652,
      "loss": 0.4922,
      "step": 870
    },
    {
      "epoch": 1.1861720067453625,
      "grad_norm": 3.25435471534729,
      "learning_rate": 0.00012147349505840073,
      "loss": 0.4825,
      "step": 880
    },
    {
      "epoch": 1.199662731871838,
      "grad_norm": 3.0708346366882324,
      "learning_rate": 0.00012057502246181491,
      "loss": 0.4777,
      "step": 890
    },
    {
      "epoch": 1.2131534569983136,
      "grad_norm": 3.274693727493286,
      "learning_rate": 0.00011967654986522911,
      "loss": 0.4727,
      "step": 900
    },
    {
      "epoch": 1.2266441821247893,
      "grad_norm": 5.614912986755371,
      "learning_rate": 0.00011877807726864331,
      "loss": 0.4914,
      "step": 910
    },
    {
      "epoch": 1.2401349072512649,
      "grad_norm": 3.7809228897094727,
      "learning_rate": 0.0001178796046720575,
      "loss": 0.4727,
      "step": 920
    },
    {
      "epoch": 1.2536256323777404,
      "grad_norm": 3.46541428565979,
      "learning_rate": 0.0001169811320754717,
      "loss": 0.4799,
      "step": 930
    },
    {
      "epoch": 1.267116357504216,
      "grad_norm": 3.8525471687316895,
      "learning_rate": 0.0001160826594788859,
      "loss": 0.5079,
      "step": 940
    },
    {
      "epoch": 1.2806070826306915,
      "grad_norm": 3.959237575531006,
      "learning_rate": 0.00011518418688230011,
      "loss": 0.4689,
      "step": 950
    },
    {
      "epoch": 1.294097807757167,
      "grad_norm": 3.873572826385498,
      "learning_rate": 0.00011428571428571428,
      "loss": 0.4688,
      "step": 960
    },
    {
      "epoch": 1.3075885328836425,
      "grad_norm": 3.813100814819336,
      "learning_rate": 0.00011338724168912849,
      "loss": 0.4762,
      "step": 970
    },
    {
      "epoch": 1.321079258010118,
      "grad_norm": 3.517953634262085,
      "learning_rate": 0.00011248876909254269,
      "loss": 0.4595,
      "step": 980
    },
    {
      "epoch": 1.3345699831365936,
      "grad_norm": 3.5959513187408447,
      "learning_rate": 0.00011159029649595688,
      "loss": 0.4959,
      "step": 990
    },
    {
      "epoch": 1.3480607082630691,
      "grad_norm": 3.9473278522491455,
      "learning_rate": 0.00011069182389937108,
      "loss": 0.4552,
      "step": 1000
    },
    {
      "epoch": 1.3615514333895447,
      "grad_norm": 3.5634729862213135,
      "learning_rate": 0.00010979335130278528,
      "loss": 0.4765,
      "step": 1010
    },
    {
      "epoch": 1.3750421585160202,
      "grad_norm": 3.88102126121521,
      "learning_rate": 0.00010889487870619948,
      "loss": 0.4282,
      "step": 1020
    },
    {
      "epoch": 1.3885328836424957,
      "grad_norm": 4.591094970703125,
      "learning_rate": 0.00010799640610961366,
      "loss": 0.483,
      "step": 1030
    },
    {
      "epoch": 1.4020236087689713,
      "grad_norm": 2.726210355758667,
      "learning_rate": 0.00010709793351302786,
      "loss": 0.5491,
      "step": 1040
    },
    {
      "epoch": 1.4155143338954468,
      "grad_norm": 2.8899638652801514,
      "learning_rate": 0.00010619946091644206,
      "loss": 0.4494,
      "step": 1050
    },
    {
      "epoch": 1.4290050590219225,
      "grad_norm": 3.1624226570129395,
      "learning_rate": 0.00010530098831985624,
      "loss": 0.4861,
      "step": 1060
    },
    {
      "epoch": 1.4424957841483979,
      "grad_norm": 4.227782726287842,
      "learning_rate": 0.00010440251572327044,
      "loss": 0.4771,
      "step": 1070
    },
    {
      "epoch": 1.4559865092748736,
      "grad_norm": 3.1170289516448975,
      "learning_rate": 0.00010350404312668464,
      "loss": 0.4608,
      "step": 1080
    },
    {
      "epoch": 1.4694772344013491,
      "grad_norm": 3.7711753845214844,
      "learning_rate": 0.00010260557053009883,
      "loss": 0.4553,
      "step": 1090
    },
    {
      "epoch": 1.4829679595278247,
      "grad_norm": 3.723172187805176,
      "learning_rate": 0.00010170709793351303,
      "loss": 0.4347,
      "step": 1100
    },
    {
      "epoch": 1.4964586846543002,
      "grad_norm": 3.469057083129883,
      "learning_rate": 0.00010080862533692723,
      "loss": 0.4747,
      "step": 1110
    },
    {
      "epoch": 1.5099494097807757,
      "grad_norm": 3.3788669109344482,
      "learning_rate": 9.991015274034142e-05,
      "loss": 0.4788,
      "step": 1120
    },
    {
      "epoch": 1.5234401349072513,
      "grad_norm": 5.321389198303223,
      "learning_rate": 9.901168014375562e-05,
      "loss": 0.4424,
      "step": 1130
    },
    {
      "epoch": 1.5369308600337268,
      "grad_norm": 3.616950511932373,
      "learning_rate": 9.811320754716981e-05,
      "loss": 0.4765,
      "step": 1140
    },
    {
      "epoch": 1.5504215851602023,
      "grad_norm": 6.907515525817871,
      "learning_rate": 9.7214734950584e-05,
      "loss": 0.4797,
      "step": 1150
    },
    {
      "epoch": 1.5639123102866779,
      "grad_norm": 17.670085906982422,
      "learning_rate": 9.63162623539982e-05,
      "loss": 0.4531,
      "step": 1160
    },
    {
      "epoch": 1.5774030354131534,
      "grad_norm": 2.698268413543701,
      "learning_rate": 9.54177897574124e-05,
      "loss": 0.493,
      "step": 1170
    },
    {
      "epoch": 1.590893760539629,
      "grad_norm": 6.46640682220459,
      "learning_rate": 9.451931716082661e-05,
      "loss": 0.4616,
      "step": 1180
    },
    {
      "epoch": 1.6043844856661047,
      "grad_norm": 2.7934491634368896,
      "learning_rate": 9.36208445642408e-05,
      "loss": 0.4819,
      "step": 1190
    },
    {
      "epoch": 1.61787521079258,
      "grad_norm": 4.599141597747803,
      "learning_rate": 9.272237196765498e-05,
      "loss": 0.4506,
      "step": 1200
    },
    {
      "epoch": 1.6313659359190558,
      "grad_norm": 3.3172645568847656,
      "learning_rate": 9.182389937106919e-05,
      "loss": 0.477,
      "step": 1210
    },
    {
      "epoch": 1.644856661045531,
      "grad_norm": 3.431093215942383,
      "learning_rate": 9.092542677448338e-05,
      "loss": 0.4594,
      "step": 1220
    },
    {
      "epoch": 1.6583473861720068,
      "grad_norm": 3.2188451290130615,
      "learning_rate": 9.002695417789758e-05,
      "loss": 0.4311,
      "step": 1230
    },
    {
      "epoch": 1.6718381112984821,
      "grad_norm": 4.311905384063721,
      "learning_rate": 8.912848158131178e-05,
      "loss": 0.4608,
      "step": 1240
    },
    {
      "epoch": 1.685328836424958,
      "grad_norm": 2.9855377674102783,
      "learning_rate": 8.823000898472597e-05,
      "loss": 0.5016,
      "step": 1250
    },
    {
      "epoch": 1.6988195615514334,
      "grad_norm": 4.071963310241699,
      "learning_rate": 8.733153638814017e-05,
      "loss": 0.4684,
      "step": 1260
    },
    {
      "epoch": 1.712310286677909,
      "grad_norm": 3.417339563369751,
      "learning_rate": 8.643306379155436e-05,
      "loss": 0.4364,
      "step": 1270
    },
    {
      "epoch": 1.7258010118043845,
      "grad_norm": 3.9395434856414795,
      "learning_rate": 8.553459119496856e-05,
      "loss": 0.5122,
      "step": 1280
    },
    {
      "epoch": 1.73929173693086,
      "grad_norm": 2.934011220932007,
      "learning_rate": 8.463611859838275e-05,
      "loss": 0.4572,
      "step": 1290
    },
    {
      "epoch": 1.7527824620573356,
      "grad_norm": 3.452453851699829,
      "learning_rate": 8.373764600179695e-05,
      "loss": 0.4674,
      "step": 1300
    },
    {
      "epoch": 1.766273187183811,
      "grad_norm": 3.5324788093566895,
      "learning_rate": 8.283917340521114e-05,
      "loss": 0.4678,
      "step": 1310
    },
    {
      "epoch": 1.7797639123102866,
      "grad_norm": 3.4225003719329834,
      "learning_rate": 8.194070080862534e-05,
      "loss": 0.5027,
      "step": 1320
    },
    {
      "epoch": 1.7932546374367622,
      "grad_norm": 1.9840672016143799,
      "learning_rate": 8.104222821203954e-05,
      "loss": 0.4977,
      "step": 1330
    },
    {
      "epoch": 1.806745362563238,
      "grad_norm": 2.443516492843628,
      "learning_rate": 8.014375561545373e-05,
      "loss": 0.455,
      "step": 1340
    },
    {
      "epoch": 1.8202360876897132,
      "grad_norm": 2.689674139022827,
      "learning_rate": 7.924528301886794e-05,
      "loss": 0.4665,
      "step": 1350
    },
    {
      "epoch": 1.833726812816189,
      "grad_norm": 3.2184505462646484,
      "learning_rate": 7.834681042228212e-05,
      "loss": 0.475,
      "step": 1360
    },
    {
      "epoch": 1.8472175379426643,
      "grad_norm": 3.5544068813323975,
      "learning_rate": 7.744833782569631e-05,
      "loss": 0.4921,
      "step": 1370
    },
    {
      "epoch": 1.86070826306914,
      "grad_norm": 3.2145285606384277,
      "learning_rate": 7.654986522911051e-05,
      "loss": 0.4825,
      "step": 1380
    },
    {
      "epoch": 1.8741989881956154,
      "grad_norm": 3.448352813720703,
      "learning_rate": 7.56513926325247e-05,
      "loss": 0.4442,
      "step": 1390
    },
    {
      "epoch": 1.8876897133220911,
      "grad_norm": 3.329658269882202,
      "learning_rate": 7.47529200359389e-05,
      "loss": 0.4498,
      "step": 1400
    },
    {
      "epoch": 1.9011804384485667,
      "grad_norm": 2.8206887245178223,
      "learning_rate": 7.385444743935311e-05,
      "loss": 0.4525,
      "step": 1410
    },
    {
      "epoch": 1.9146711635750422,
      "grad_norm": 3.067368507385254,
      "learning_rate": 7.29559748427673e-05,
      "loss": 0.4652,
      "step": 1420
    },
    {
      "epoch": 1.9281618887015177,
      "grad_norm": 2.902029275894165,
      "learning_rate": 7.20575022461815e-05,
      "loss": 0.462,
      "step": 1430
    },
    {
      "epoch": 1.9416526138279933,
      "grad_norm": 2.9392433166503906,
      "learning_rate": 7.115902964959569e-05,
      "loss": 0.4505,
      "step": 1440
    },
    {
      "epoch": 1.9551433389544688,
      "grad_norm": 3.328568935394287,
      "learning_rate": 7.026055705300989e-05,
      "loss": 0.4688,
      "step": 1450
    },
    {
      "epoch": 1.9686340640809443,
      "grad_norm": 3.0431363582611084,
      "learning_rate": 6.936208445642408e-05,
      "loss": 0.4749,
      "step": 1460
    },
    {
      "epoch": 1.9821247892074199,
      "grad_norm": 3.3194100856781006,
      "learning_rate": 6.846361185983828e-05,
      "loss": 0.4364,
      "step": 1470
    },
    {
      "epoch": 1.9956155143338954,
      "grad_norm": 3.563108205795288,
      "learning_rate": 6.756513926325248e-05,
      "loss": 0.4635,
      "step": 1480
    },
    {
      "epoch": 2.0080944350758854,
      "grad_norm": 2.6294658184051514,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.4253,
      "step": 1490
    },
    {
      "epoch": 2.0215851602023607,
      "grad_norm": 2.6223340034484863,
      "learning_rate": 6.576819407008087e-05,
      "loss": 0.4534,
      "step": 1500
    },
    {
      "epoch": 2.0350758853288364,
      "grad_norm": 2.769744396209717,
      "learning_rate": 6.486972147349506e-05,
      "loss": 0.4168,
      "step": 1510
    },
    {
      "epoch": 2.048566610455312,
      "grad_norm": 2.9616827964782715,
      "learning_rate": 6.397124887690926e-05,
      "loss": 0.4464,
      "step": 1520
    },
    {
      "epoch": 2.0620573355817875,
      "grad_norm": 2.952134132385254,
      "learning_rate": 6.307277628032345e-05,
      "loss": 0.4814,
      "step": 1530
    },
    {
      "epoch": 2.0755480607082633,
      "grad_norm": 3.6748759746551514,
      "learning_rate": 6.217430368373764e-05,
      "loss": 0.4638,
      "step": 1540
    },
    {
      "epoch": 2.0890387858347386,
      "grad_norm": 1.927171230316162,
      "learning_rate": 6.127583108715184e-05,
      "loss": 0.4512,
      "step": 1550
    },
    {
      "epoch": 2.1025295109612143,
      "grad_norm": 2.6556429862976074,
      "learning_rate": 6.037735849056604e-05,
      "loss": 0.4752,
      "step": 1560
    },
    {
      "epoch": 2.1160202360876896,
      "grad_norm": 1.954484462738037,
      "learning_rate": 5.947888589398024e-05,
      "loss": 0.4347,
      "step": 1570
    },
    {
      "epoch": 2.1295109612141654,
      "grad_norm": 2.986450433731079,
      "learning_rate": 5.8580413297394435e-05,
      "loss": 0.4278,
      "step": 1580
    },
    {
      "epoch": 2.1430016863406407,
      "grad_norm": 2.7762913703918457,
      "learning_rate": 5.768194070080862e-05,
      "loss": 0.4496,
      "step": 1590
    },
    {
      "epoch": 2.1564924114671165,
      "grad_norm": 3.0169496536254883,
      "learning_rate": 5.6783468104222825e-05,
      "loss": 0.4359,
      "step": 1600
    },
    {
      "epoch": 2.1699831365935918,
      "grad_norm": 2.6418251991271973,
      "learning_rate": 5.5884995507637014e-05,
      "loss": 0.4512,
      "step": 1610
    },
    {
      "epoch": 2.1834738617200675,
      "grad_norm": 2.8345439434051514,
      "learning_rate": 5.4986522911051216e-05,
      "loss": 0.4339,
      "step": 1620
    },
    {
      "epoch": 2.196964586846543,
      "grad_norm": 2.8585574626922607,
      "learning_rate": 5.408805031446541e-05,
      "loss": 0.4072,
      "step": 1630
    },
    {
      "epoch": 2.2104553119730186,
      "grad_norm": 1.8751145601272583,
      "learning_rate": 5.31895777178796e-05,
      "loss": 0.4526,
      "step": 1640
    },
    {
      "epoch": 2.223946037099494,
      "grad_norm": 2.441622734069824,
      "learning_rate": 5.22911051212938e-05,
      "loss": 0.4758,
      "step": 1650
    },
    {
      "epoch": 2.2374367622259697,
      "grad_norm": 3.11625337600708,
      "learning_rate": 5.1392632524708e-05,
      "loss": 0.4705,
      "step": 1660
    },
    {
      "epoch": 2.2509274873524454,
      "grad_norm": 2.761462688446045,
      "learning_rate": 5.04941599281222e-05,
      "loss": 0.4439,
      "step": 1670
    },
    {
      "epoch": 2.2644182124789207,
      "grad_norm": 2.8460707664489746,
      "learning_rate": 4.959568733153639e-05,
      "loss": 0.4595,
      "step": 1680
    },
    {
      "epoch": 2.2779089376053965,
      "grad_norm": 2.73833966255188,
      "learning_rate": 4.869721473495058e-05,
      "loss": 0.436,
      "step": 1690
    },
    {
      "epoch": 2.291399662731872,
      "grad_norm": 3.1314399242401123,
      "learning_rate": 4.7798742138364785e-05,
      "loss": 0.4302,
      "step": 1700
    },
    {
      "epoch": 2.3048903878583475,
      "grad_norm": 2.203801155090332,
      "learning_rate": 4.690026954177898e-05,
      "loss": 0.4353,
      "step": 1710
    },
    {
      "epoch": 2.318381112984823,
      "grad_norm": 1.5654208660125732,
      "learning_rate": 4.6001796945193176e-05,
      "loss": 0.4436,
      "step": 1720
    },
    {
      "epoch": 2.3318718381112986,
      "grad_norm": 1.8968826532363892,
      "learning_rate": 4.5103324348607365e-05,
      "loss": 0.4617,
      "step": 1730
    },
    {
      "epoch": 2.345362563237774,
      "grad_norm": 3.4319088459014893,
      "learning_rate": 4.420485175202157e-05,
      "loss": 0.4951,
      "step": 1740
    },
    {
      "epoch": 2.3588532883642497,
      "grad_norm": 0.9367527365684509,
      "learning_rate": 4.330637915543576e-05,
      "loss": 0.4469,
      "step": 1750
    },
    {
      "epoch": 2.372344013490725,
      "grad_norm": 1.3041300773620605,
      "learning_rate": 4.240790655884996e-05,
      "loss": 0.4696,
      "step": 1760
    },
    {
      "epoch": 2.3858347386172007,
      "grad_norm": 1.0405113697052002,
      "learning_rate": 4.150943396226415e-05,
      "loss": 0.4121,
      "step": 1770
    },
    {
      "epoch": 2.399325463743676,
      "grad_norm": 1.1513006687164307,
      "learning_rate": 4.061096136567835e-05,
      "loss": 0.4134,
      "step": 1780
    },
    {
      "epoch": 2.412816188870152,
      "grad_norm": 1.3874084949493408,
      "learning_rate": 3.971248876909254e-05,
      "loss": 0.4375,
      "step": 1790
    },
    {
      "epoch": 2.426306913996627,
      "grad_norm": 1.29698646068573,
      "learning_rate": 3.881401617250674e-05,
      "loss": 0.4752,
      "step": 1800
    },
    {
      "epoch": 2.439797639123103,
      "grad_norm": 0.8246452808380127,
      "learning_rate": 3.7915543575920934e-05,
      "loss": 0.4738,
      "step": 1810
    },
    {
      "epoch": 2.4532883642495786,
      "grad_norm": 1.246367335319519,
      "learning_rate": 3.7017070979335136e-05,
      "loss": 0.4583,
      "step": 1820
    },
    {
      "epoch": 2.466779089376054,
      "grad_norm": 2.288325071334839,
      "learning_rate": 3.611859838274933e-05,
      "loss": 0.4643,
      "step": 1830
    },
    {
      "epoch": 2.4802698145025297,
      "grad_norm": 2.6386337280273438,
      "learning_rate": 3.522012578616352e-05,
      "loss": 0.4748,
      "step": 1840
    },
    {
      "epoch": 2.493760539629005,
      "grad_norm": 2.788741111755371,
      "learning_rate": 3.4321653189577715e-05,
      "loss": 0.4623,
      "step": 1850
    },
    {
      "epoch": 2.5072512647554808,
      "grad_norm": 3.7216169834136963,
      "learning_rate": 3.342318059299192e-05,
      "loss": 0.441,
      "step": 1860
    },
    {
      "epoch": 2.520741989881956,
      "grad_norm": 2.007139205932617,
      "learning_rate": 3.252470799640611e-05,
      "loss": 0.4383,
      "step": 1870
    },
    {
      "epoch": 2.534232715008432,
      "grad_norm": 0.6199157238006592,
      "learning_rate": 3.162623539982031e-05,
      "loss": 0.4197,
      "step": 1880
    },
    {
      "epoch": 2.547723440134907,
      "grad_norm": 0.895648181438446,
      "learning_rate": 3.0727762803234503e-05,
      "loss": 0.4403,
      "step": 1890
    },
    {
      "epoch": 2.561214165261383,
      "grad_norm": 0.6603314280509949,
      "learning_rate": 2.9829290206648695e-05,
      "loss": 0.4794,
      "step": 1900
    },
    {
      "epoch": 2.574704890387858,
      "grad_norm": 1.2536957263946533,
      "learning_rate": 2.8930817610062894e-05,
      "loss": 0.4778,
      "step": 1910
    },
    {
      "epoch": 2.588195615514334,
      "grad_norm": 0.6780521869659424,
      "learning_rate": 2.803234501347709e-05,
      "loss": 0.4672,
      "step": 1920
    },
    {
      "epoch": 2.6016863406408093,
      "grad_norm": 0.6663461327552795,
      "learning_rate": 2.7133872416891288e-05,
      "loss": 0.4212,
      "step": 1930
    },
    {
      "epoch": 2.615177065767285,
      "grad_norm": 1.3189208507537842,
      "learning_rate": 2.6235399820305483e-05,
      "loss": 0.4366,
      "step": 1940
    },
    {
      "epoch": 2.6286677908937603,
      "grad_norm": 0.6126547455787659,
      "learning_rate": 2.5336927223719675e-05,
      "loss": 0.4467,
      "step": 1950
    },
    {
      "epoch": 2.642158516020236,
      "grad_norm": 0.8297685980796814,
      "learning_rate": 2.4438454627133874e-05,
      "loss": 0.4203,
      "step": 1960
    },
    {
      "epoch": 2.655649241146712,
      "grad_norm": 0.81744784116745,
      "learning_rate": 2.353998203054807e-05,
      "loss": 0.4714,
      "step": 1970
    },
    {
      "epoch": 2.669139966273187,
      "grad_norm": 1.0496429204940796,
      "learning_rate": 2.2641509433962265e-05,
      "loss": 0.4517,
      "step": 1980
    },
    {
      "epoch": 2.6826306913996625,
      "grad_norm": 0.843941330909729,
      "learning_rate": 2.174303683737646e-05,
      "loss": 0.4361,
      "step": 1990
    },
    {
      "epoch": 2.6961214165261382,
      "grad_norm": 0.7860977649688721,
      "learning_rate": 2.0844564240790655e-05,
      "loss": 0.4061,
      "step": 2000
    },
    {
      "epoch": 2.709612141652614,
      "grad_norm": 1.3650184869766235,
      "learning_rate": 1.9946091644204854e-05,
      "loss": 0.4641,
      "step": 2010
    },
    {
      "epoch": 2.7231028667790893,
      "grad_norm": 0.623533308506012,
      "learning_rate": 1.9047619047619046e-05,
      "loss": 0.4296,
      "step": 2020
    },
    {
      "epoch": 2.736593591905565,
      "grad_norm": 1.0236366987228394,
      "learning_rate": 1.8149146451033245e-05,
      "loss": 0.449,
      "step": 2030
    },
    {
      "epoch": 2.7500843170320404,
      "grad_norm": 0.6305617690086365,
      "learning_rate": 1.725067385444744e-05,
      "loss": 0.4888,
      "step": 2040
    },
    {
      "epoch": 2.763575042158516,
      "grad_norm": 0.5004448294639587,
      "learning_rate": 1.6352201257861635e-05,
      "loss": 0.4227,
      "step": 2050
    },
    {
      "epoch": 2.7770657672849914,
      "grad_norm": 0.6496240496635437,
      "learning_rate": 1.545372866127583e-05,
      "loss": 0.4499,
      "step": 2060
    },
    {
      "epoch": 2.790556492411467,
      "grad_norm": 0.9616618752479553,
      "learning_rate": 1.455525606469003e-05,
      "loss": 0.4368,
      "step": 2070
    },
    {
      "epoch": 2.8040472175379425,
      "grad_norm": 0.5831645727157593,
      "learning_rate": 1.3656783468104223e-05,
      "loss": 0.4452,
      "step": 2080
    },
    {
      "epoch": 2.8175379426644183,
      "grad_norm": 0.8784683346748352,
      "learning_rate": 1.275831087151842e-05,
      "loss": 0.4698,
      "step": 2090
    },
    {
      "epoch": 2.8310286677908936,
      "grad_norm": 1.3440258502960205,
      "learning_rate": 1.1859838274932616e-05,
      "loss": 0.4195,
      "step": 2100
    },
    {
      "epoch": 2.8445193929173693,
      "grad_norm": 1.2359696626663208,
      "learning_rate": 1.0961365678346811e-05,
      "loss": 0.4305,
      "step": 2110
    },
    {
      "epoch": 2.858010118043845,
      "grad_norm": 1.1054702997207642,
      "learning_rate": 1.0062893081761008e-05,
      "loss": 0.4649,
      "step": 2120
    },
    {
      "epoch": 2.8715008431703204,
      "grad_norm": 1.379691481590271,
      "learning_rate": 9.164420485175203e-06,
      "loss": 0.4876,
      "step": 2130
    },
    {
      "epoch": 2.8849915682967957,
      "grad_norm": 1.6011934280395508,
      "learning_rate": 8.265947888589399e-06,
      "loss": 0.4759,
      "step": 2140
    },
    {
      "epoch": 2.8984822934232715,
      "grad_norm": 1.1230287551879883,
      "learning_rate": 7.367475292003594e-06,
      "loss": 0.4569,
      "step": 2150
    },
    {
      "epoch": 2.911973018549747,
      "grad_norm": 0.5642672181129456,
      "learning_rate": 6.46900269541779e-06,
      "loss": 0.4302,
      "step": 2160
    },
    {
      "epoch": 2.9254637436762225,
      "grad_norm": 0.5524741411209106,
      "learning_rate": 5.570530098831986e-06,
      "loss": 0.4062,
      "step": 2170
    },
    {
      "epoch": 2.9389544688026983,
      "grad_norm": 0.5504957437515259,
      "learning_rate": 4.6720575022461816e-06,
      "loss": 0.4336,
      "step": 2180
    },
    {
      "epoch": 2.9524451939291736,
      "grad_norm": 0.9138615131378174,
      "learning_rate": 3.7735849056603773e-06,
      "loss": 0.4463,
      "step": 2190
    },
    {
      "epoch": 2.9659359190556494,
      "grad_norm": 0.5804033875465393,
      "learning_rate": 2.8751123090745735e-06,
      "loss": 0.4323,
      "step": 2200
    },
    {
      "epoch": 2.9794266441821247,
      "grad_norm": 0.8893300890922546,
      "learning_rate": 1.9766397124887693e-06,
      "loss": 0.4244,
      "step": 2210
    },
    {
      "epoch": 2.9929173693086004,
      "grad_norm": 0.6132921576499939,
      "learning_rate": 1.078167115902965e-06,
      "loss": 0.4562,
      "step": 2220
    }
  ],
  "logging_steps": 10,
  "max_steps": 2226,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 1.107103129778258e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}