{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.995713673381912,
  "eval_steps": 500,
  "global_step": 11660,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04,
      "learning_rate": 0.00019914236706689538,
      "loss": 2.573,
      "step": 50
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.00019828473413379075,
      "loss": 1.8291,
      "step": 100
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00019742710120068612,
      "loss": 1.7257,
      "step": 150
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00019656946826758148,
      "loss": 1.6338,
      "step": 200
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00019571183533447685,
      "loss": 1.6359,
      "step": 250
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00019485420240137222,
      "loss": 1.5935,
      "step": 300
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0001939965694682676,
      "loss": 1.5494,
      "step": 350
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00019313893653516296,
      "loss": 1.559,
      "step": 400
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00019228130360205833,
      "loss": 1.5403,
      "step": 450
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0001914236706689537,
      "loss": 1.528,
      "step": 500
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.00019056603773584906,
      "loss": 1.4899,
      "step": 550
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00018970840480274443,
      "loss": 1.4545,
      "step": 600
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0001888507718696398,
      "loss": 1.4658,
      "step": 650
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00018799313893653517,
      "loss": 1.4666,
      "step": 700
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00018713550600343054,
      "loss": 1.4603,
      "step": 750
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0001862778730703259,
      "loss": 1.4256,
      "step": 800
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00018542024013722128,
      "loss": 1.4422,
      "step": 850
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00018456260720411664,
      "loss": 1.4255,
      "step": 900
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.000183704974271012,
      "loss": 1.3872,
      "step": 950
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00018284734133790738,
      "loss": 1.3886,
      "step": 1000
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00018198970840480275,
      "loss": 1.4147,
      "step": 1050
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00018113207547169812,
      "loss": 1.405,
      "step": 1100
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0001802744425385935,
      "loss": 1.4027,
      "step": 1150
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00017941680960548886,
      "loss": 1.1787,
      "step": 1200
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.00017855917667238422,
      "loss": 1.058,
      "step": 1250
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00017770154373927962,
      "loss": 1.0577,
      "step": 1300
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00017684391080617496,
      "loss": 1.0675,
      "step": 1350
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00017598627787307033,
      "loss": 1.0459,
      "step": 1400
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0001751286449399657,
      "loss": 1.072,
      "step": 1450
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00017427101200686107,
      "loss": 1.0654,
      "step": 1500
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00017341337907375644,
      "loss": 1.0594,
      "step": 1550
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.00017255574614065183,
      "loss": 1.0594,
      "step": 1600
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00017169811320754717,
      "loss": 1.0851,
      "step": 1650
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00017084048027444254,
      "loss": 1.066,
      "step": 1700
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0001699828473413379,
      "loss": 1.0591,
      "step": 1750
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.00016912521440823328,
      "loss": 1.0692,
      "step": 1800
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00016826758147512865,
      "loss": 1.0772,
      "step": 1850
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00016740994854202404,
      "loss": 1.075,
      "step": 1900
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.00016655231560891938,
      "loss": 1.0607,
      "step": 1950
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00016569468267581475,
      "loss": 1.0682,
      "step": 2000
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00016483704974271012,
      "loss": 1.0616,
      "step": 2050
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0001639794168096055,
      "loss": 1.0655,
      "step": 2100
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00016312178387650086,
      "loss": 1.0555,
      "step": 2150
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00016226415094339625,
      "loss": 1.0761,
      "step": 2200
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0001614065180102916,
      "loss": 1.0776,
      "step": 2250
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00016054888507718696,
      "loss": 1.0574,
      "step": 2300
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.00015969125214408233,
      "loss": 0.9443,
      "step": 2350
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0001588336192109777,
      "loss": 0.7435,
      "step": 2400
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.00015797598627787307,
      "loss": 0.7457,
      "step": 2450
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.00015711835334476847,
      "loss": 0.7553,
      "step": 2500
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0001562607204116638,
      "loss": 0.7452,
      "step": 2550
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.00015540308747855918,
      "loss": 0.7736,
      "step": 2600
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.00015454545454545454,
      "loss": 0.7649,
      "step": 2650
    },
    {
      "epoch": 2.31,
      "learning_rate": 0.0001536878216123499,
      "loss": 0.7557,
      "step": 2700
    },
    {
      "epoch": 2.36,
      "learning_rate": 0.0001528301886792453,
      "loss": 0.7556,
      "step": 2750
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.00015197255574614068,
      "loss": 0.7751,
      "step": 2800
    },
    {
      "epoch": 2.44,
      "learning_rate": 0.00015111492281303602,
      "loss": 0.7613,
      "step": 2850
    },
    {
      "epoch": 2.49,
      "learning_rate": 0.0001502572898799314,
      "loss": 0.7751,
      "step": 2900
    },
    {
      "epoch": 2.53,
      "learning_rate": 0.00014939965694682676,
      "loss": 0.774,
      "step": 2950
    },
    {
      "epoch": 2.57,
      "learning_rate": 0.00014854202401372212,
      "loss": 0.7803,
      "step": 3000
    },
    {
      "epoch": 2.61,
      "learning_rate": 0.00014768439108061752,
      "loss": 0.7691,
      "step": 3050
    },
    {
      "epoch": 2.66,
      "learning_rate": 0.0001468267581475129,
      "loss": 0.7684,
      "step": 3100
    },
    {
      "epoch": 2.7,
      "learning_rate": 0.00014596912521440823,
      "loss": 0.7811,
      "step": 3150
    },
    {
      "epoch": 2.74,
      "learning_rate": 0.0001451114922813036,
      "loss": 0.7631,
      "step": 3200
    },
    {
      "epoch": 2.79,
      "learning_rate": 0.00014425385934819897,
      "loss": 0.7782,
      "step": 3250
    },
    {
      "epoch": 2.83,
      "learning_rate": 0.00014339622641509434,
      "loss": 0.7686,
      "step": 3300
    },
    {
      "epoch": 2.87,
      "learning_rate": 0.00014253859348198973,
      "loss": 0.7919,
      "step": 3350
    },
    {
      "epoch": 2.91,
      "learning_rate": 0.0001416809605488851,
      "loss": 0.7768,
      "step": 3400
    },
    {
      "epoch": 2.96,
      "learning_rate": 0.00014082332761578044,
      "loss": 0.7837,
      "step": 3450
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0001399656946826758,
      "loss": 0.7739,
      "step": 3500
    },
    {
      "epoch": 3.04,
      "learning_rate": 0.00013910806174957118,
      "loss": 0.4864,
      "step": 3550
    },
    {
      "epoch": 3.09,
      "learning_rate": 0.00013825042881646655,
      "loss": 0.4876,
      "step": 3600
    },
    {
      "epoch": 3.13,
      "learning_rate": 0.00013739279588336194,
      "loss": 0.4899,
      "step": 3650
    },
    {
      "epoch": 3.17,
      "learning_rate": 0.0001365351629502573,
      "loss": 0.4833,
      "step": 3700
    },
    {
      "epoch": 3.21,
      "learning_rate": 0.00013567753001715265,
      "loss": 0.5126,
      "step": 3750
    },
    {
      "epoch": 3.26,
      "learning_rate": 0.00013481989708404802,
      "loss": 0.4891,
      "step": 3800
    },
    {
      "epoch": 3.3,
      "learning_rate": 0.0001339622641509434,
      "loss": 0.5,
      "step": 3850
    },
    {
      "epoch": 3.34,
      "learning_rate": 0.00013310463121783879,
      "loss": 0.5078,
      "step": 3900
    },
    {
      "epoch": 3.39,
      "learning_rate": 0.00013224699828473415,
      "loss": 0.4931,
      "step": 3950
    },
    {
      "epoch": 3.43,
      "learning_rate": 0.00013138936535162952,
      "loss": 0.5062,
      "step": 4000
    },
    {
      "epoch": 3.47,
      "learning_rate": 0.00013053173241852486,
      "loss": 0.509,
      "step": 4050
    },
    {
      "epoch": 3.51,
      "learning_rate": 0.00012967409948542023,
      "loss": 0.5097,
      "step": 4100
    },
    {
      "epoch": 3.56,
      "learning_rate": 0.0001288164665523156,
      "loss": 0.5225,
      "step": 4150
    },
    {
      "epoch": 3.6,
      "learning_rate": 0.000127958833619211,
      "loss": 0.5133,
      "step": 4200
    },
    {
      "epoch": 3.64,
      "learning_rate": 0.00012710120068610637,
      "loss": 0.5183,
      "step": 4250
    },
    {
      "epoch": 3.69,
      "learning_rate": 0.00012624356775300173,
      "loss": 0.5259,
      "step": 4300
    },
    {
      "epoch": 3.73,
      "learning_rate": 0.00012538593481989708,
      "loss": 0.5172,
      "step": 4350
    },
    {
      "epoch": 3.77,
      "learning_rate": 0.00012452830188679244,
      "loss": 0.5135,
      "step": 4400
    },
    {
      "epoch": 3.81,
      "learning_rate": 0.0001236706689536878,
      "loss": 0.5218,
      "step": 4450
    },
    {
      "epoch": 3.86,
      "learning_rate": 0.0001228130360205832,
      "loss": 0.5252,
      "step": 4500
    },
    {
      "epoch": 3.9,
      "learning_rate": 0.00012195540308747856,
      "loss": 0.5139,
      "step": 4550
    },
    {
      "epoch": 3.94,
      "learning_rate": 0.00012109777015437393,
      "loss": 0.5267,
      "step": 4600
    },
    {
      "epoch": 3.99,
      "learning_rate": 0.0001202401372212693,
      "loss": 0.5193,
      "step": 4650
    },
    {
      "epoch": 4.03,
      "learning_rate": 0.00011938250428816467,
      "loss": 0.3797,
      "step": 4700
    },
    {
      "epoch": 4.07,
      "learning_rate": 0.00011852487135506002,
      "loss": 0.3031,
      "step": 4750
    },
    {
      "epoch": 4.11,
      "learning_rate": 0.00011766723842195542,
      "loss": 0.2867,
      "step": 4800
    },
    {
      "epoch": 4.16,
      "learning_rate": 0.00011680960548885078,
      "loss": 0.2986,
      "step": 4850
    },
    {
      "epoch": 4.2,
      "learning_rate": 0.00011595197255574614,
      "loss": 0.3021,
      "step": 4900
    },
    {
      "epoch": 4.24,
      "learning_rate": 0.00011509433962264151,
      "loss": 0.2983,
      "step": 4950
    },
    {
      "epoch": 4.29,
      "learning_rate": 0.00011423670668953688,
      "loss": 0.3014,
      "step": 5000
    },
    {
      "epoch": 4.33,
      "learning_rate": 0.00011337907375643224,
      "loss": 0.2966,
      "step": 5050
    },
    {
      "epoch": 4.37,
      "learning_rate": 0.00011252144082332763,
      "loss": 0.3014,
      "step": 5100
    },
    {
      "epoch": 4.41,
      "learning_rate": 0.00011166380789022299,
      "loss": 0.3105,
      "step": 5150
    },
    {
      "epoch": 4.46,
      "learning_rate": 0.00011080617495711836,
      "loss": 0.3115,
      "step": 5200
    },
    {
      "epoch": 4.5,
      "learning_rate": 0.00010994854202401372,
      "loss": 0.3139,
      "step": 5250
    },
    {
      "epoch": 4.54,
      "learning_rate": 0.00010909090909090909,
      "loss": 0.3089,
      "step": 5300
    },
    {
      "epoch": 4.59,
      "learning_rate": 0.00010823327615780447,
      "loss": 0.3069,
      "step": 5350
    },
    {
      "epoch": 4.63,
      "learning_rate": 0.00010737564322469984,
      "loss": 0.3181,
      "step": 5400
    },
    {
      "epoch": 4.67,
      "learning_rate": 0.0001065180102915952,
      "loss": 0.3129,
      "step": 5450
    },
    {
      "epoch": 4.71,
      "learning_rate": 0.00010566037735849057,
      "loss": 0.3199,
      "step": 5500
    },
    {
      "epoch": 4.76,
      "learning_rate": 0.00010480274442538594,
      "loss": 0.3079,
      "step": 5550
    },
    {
      "epoch": 4.8,
      "learning_rate": 0.0001039451114922813,
      "loss": 0.32,
      "step": 5600
    },
    {
      "epoch": 4.84,
      "learning_rate": 0.00010308747855917669,
      "loss": 0.3157,
      "step": 5650
    },
    {
      "epoch": 4.89,
      "learning_rate": 0.00010222984562607205,
      "loss": 0.3181,
      "step": 5700
    },
    {
      "epoch": 4.93,
      "learning_rate": 0.00010137221269296742,
      "loss": 0.3217,
      "step": 5750
    },
    {
      "epoch": 4.97,
      "learning_rate": 0.00010051457975986278,
      "loss": 0.3132,
      "step": 5800
    },
    {
      "epoch": 5.02,
      "learning_rate": 9.965694682675816e-05,
      "loss": 0.2703,
      "step": 5850
    },
    {
      "epoch": 5.06,
      "learning_rate": 9.879931389365353e-05,
      "loss": 0.1618,
      "step": 5900
    },
    {
      "epoch": 5.1,
      "learning_rate": 9.794168096054888e-05,
      "loss": 0.1767,
      "step": 5950
    },
    {
      "epoch": 5.14,
      "learning_rate": 9.708404802744427e-05,
      "loss": 0.1647,
      "step": 6000
    },
    {
      "epoch": 5.19,
      "learning_rate": 9.622641509433963e-05,
      "loss": 0.1725,
      "step": 6050
    },
    {
      "epoch": 5.23,
      "learning_rate": 9.536878216123499e-05,
      "loss": 0.1732,
      "step": 6100
    },
    {
      "epoch": 5.27,
      "learning_rate": 9.451114922813037e-05,
      "loss": 0.168,
      "step": 6150
    },
    {
      "epoch": 5.32,
      "learning_rate": 9.365351629502574e-05,
      "loss": 0.1769,
      "step": 6200
    },
    {
      "epoch": 5.36,
      "learning_rate": 9.27958833619211e-05,
      "loss": 0.1718,
      "step": 6250
    },
    {
      "epoch": 5.4,
      "learning_rate": 9.193825042881648e-05,
      "loss": 0.1734,
      "step": 6300
    },
    {
      "epoch": 5.44,
      "learning_rate": 9.108061749571185e-05,
      "loss": 0.1815,
      "step": 6350
    },
    {
      "epoch": 5.49,
      "learning_rate": 9.02229845626072e-05,
      "loss": 0.1761,
      "step": 6400
    },
    {
      "epoch": 5.53,
      "learning_rate": 8.936535162950258e-05,
      "loss": 0.1783,
      "step": 6450
    },
    {
      "epoch": 5.57,
      "learning_rate": 8.850771869639795e-05,
      "loss": 0.1764,
      "step": 6500
    },
    {
      "epoch": 5.62,
      "learning_rate": 8.765008576329331e-05,
      "loss": 0.1763,
      "step": 6550
    },
    {
      "epoch": 5.66,
      "learning_rate": 8.679245283018869e-05,
      "loss": 0.1684,
      "step": 6600
    },
    {
      "epoch": 5.7,
      "learning_rate": 8.593481989708406e-05,
      "loss": 0.174,
      "step": 6650
    },
    {
      "epoch": 5.74,
      "learning_rate": 8.507718696397941e-05,
      "loss": 0.1724,
      "step": 6700
    },
    {
      "epoch": 5.79,
      "learning_rate": 8.42195540308748e-05,
      "loss": 0.177,
      "step": 6750
    },
    {
      "epoch": 5.83,
      "learning_rate": 8.336192109777016e-05,
      "loss": 0.1753,
      "step": 6800
    },
    {
      "epoch": 5.87,
      "learning_rate": 8.250428816466552e-05,
      "loss": 0.1765,
      "step": 6850
    },
    {
      "epoch": 5.92,
      "learning_rate": 8.16466552315609e-05,
      "loss": 0.1748,
      "step": 6900
    },
    {
      "epoch": 5.96,
      "learning_rate": 8.078902229845627e-05,
      "loss": 0.1781,
      "step": 6950
    },
    {
      "epoch": 6.0,
      "learning_rate": 7.993138936535164e-05,
      "loss": 0.1838,
      "step": 7000
    },
    {
      "epoch": 6.04,
      "learning_rate": 7.9073756432247e-05,
      "loss": 0.0912,
      "step": 7050
    },
    {
      "epoch": 6.09,
      "learning_rate": 7.821612349914237e-05,
      "loss": 0.0874,
      "step": 7100
    },
    {
      "epoch": 6.13,
      "learning_rate": 7.735849056603774e-05,
      "loss": 0.0879,
      "step": 7150
    },
    {
      "epoch": 6.17,
      "learning_rate": 7.650085763293311e-05,
      "loss": 0.0922,
      "step": 7200
    },
    {
      "epoch": 6.22,
      "learning_rate": 7.564322469982848e-05,
      "loss": 0.0979,
      "step": 7250
    },
    {
      "epoch": 6.26,
      "learning_rate": 7.478559176672385e-05,
      "loss": 0.0853,
      "step": 7300
    },
    {
      "epoch": 6.3,
      "learning_rate": 7.392795883361922e-05,
      "loss": 0.0866,
      "step": 7350
    },
    {
      "epoch": 6.34,
      "learning_rate": 7.307032590051459e-05,
      "loss": 0.0876,
      "step": 7400
    },
    {
      "epoch": 6.39,
      "learning_rate": 7.221269296740995e-05,
      "loss": 0.0895,
      "step": 7450
    },
    {
      "epoch": 6.43,
      "learning_rate": 7.135506003430532e-05,
      "loss": 0.0925,
      "step": 7500
    },
    {
      "epoch": 6.47,
      "learning_rate": 7.049742710120069e-05,
      "loss": 0.0915,
      "step": 7550
    },
    {
      "epoch": 6.52,
      "learning_rate": 6.963979416809606e-05,
      "loss": 0.0844,
      "step": 7600
    },
    {
      "epoch": 6.56,
      "learning_rate": 6.878216123499143e-05,
      "loss": 0.0949,
      "step": 7650
    },
    {
      "epoch": 6.6,
      "learning_rate": 6.79245283018868e-05,
      "loss": 0.0893,
      "step": 7700
    },
    {
      "epoch": 6.64,
      "learning_rate": 6.706689536878217e-05,
      "loss": 0.0921,
      "step": 7750
    },
    {
      "epoch": 6.69,
      "learning_rate": 6.620926243567753e-05,
      "loss": 0.0879,
      "step": 7800
    },
    {
      "epoch": 6.73,
      "learning_rate": 6.53516295025729e-05,
      "loss": 0.0886,
      "step": 7850
    },
    {
      "epoch": 6.77,
      "learning_rate": 6.449399656946827e-05,
      "loss": 0.0912,
      "step": 7900
    },
    {
      "epoch": 6.82,
      "learning_rate": 6.363636363636364e-05,
      "loss": 0.0898,
      "step": 7950
    },
    {
      "epoch": 6.86,
      "learning_rate": 6.277873070325901e-05,
      "loss": 0.0936,
      "step": 8000
    },
    {
      "epoch": 6.9,
      "learning_rate": 6.192109777015438e-05,
      "loss": 0.0903,
      "step": 8050
    },
    {
      "epoch": 6.94,
      "learning_rate": 6.106346483704975e-05,
      "loss": 0.0876,
      "step": 8100
    },
    {
      "epoch": 6.99,
      "learning_rate": 6.020583190394511e-05,
      "loss": 0.0887,
      "step": 8150
    },
    {
      "epoch": 7.03,
      "learning_rate": 5.9348198970840484e-05,
      "loss": 0.0559,
      "step": 8200
    },
    {
      "epoch": 7.07,
      "learning_rate": 5.849056603773585e-05,
      "loss": 0.042,
      "step": 8250
    },
    {
      "epoch": 7.12,
      "learning_rate": 5.763293310463123e-05,
      "loss": 0.0443,
      "step": 8300
    },
    {
      "epoch": 7.16,
      "learning_rate": 5.677530017152659e-05,
      "loss": 0.0419,
      "step": 8350
    },
    {
      "epoch": 7.2,
      "learning_rate": 5.591766723842196e-05,
      "loss": 0.0406,
      "step": 8400
    },
    {
      "epoch": 7.24,
      "learning_rate": 5.506003430531733e-05,
      "loss": 0.043,
      "step": 8450
    },
    {
      "epoch": 7.29,
      "learning_rate": 5.4202401372212695e-05,
      "loss": 0.042,
      "step": 8500
    },
    {
      "epoch": 7.33,
      "learning_rate": 5.3344768439108064e-05,
      "loss": 0.0404,
      "step": 8550
    },
    {
      "epoch": 7.37,
      "learning_rate": 5.248713550600344e-05,
      "loss": 0.0415,
      "step": 8600
    },
    {
      "epoch": 7.42,
      "learning_rate": 5.16295025728988e-05,
      "loss": 0.0443,
      "step": 8650
    },
    {
      "epoch": 7.46,
      "learning_rate": 5.077186963979417e-05,
      "loss": 0.0435,
      "step": 8700
    },
    {
      "epoch": 7.5,
      "learning_rate": 4.991423670668954e-05,
      "loss": 0.0393,
      "step": 8750
    },
    {
      "epoch": 7.54,
      "learning_rate": 4.9056603773584906e-05,
      "loss": 0.0424,
      "step": 8800
    },
    {
      "epoch": 7.59,
      "learning_rate": 4.819897084048028e-05,
      "loss": 0.0411,
      "step": 8850
    },
    {
      "epoch": 7.63,
      "learning_rate": 4.7341337907375644e-05,
      "loss": 0.0409,
      "step": 8900
    },
    {
      "epoch": 7.67,
      "learning_rate": 4.648370497427101e-05,
      "loss": 0.0431,
      "step": 8950
    },
    {
      "epoch": 7.72,
      "learning_rate": 4.562607204116639e-05,
      "loss": 0.0411,
      "step": 9000
    },
    {
      "epoch": 7.76,
      "learning_rate": 4.476843910806175e-05,
      "loss": 0.0378,
      "step": 9050
    },
    {
      "epoch": 7.8,
      "learning_rate": 4.391080617495712e-05,
      "loss": 0.0402,
      "step": 9100
    },
    {
      "epoch": 7.84,
      "learning_rate": 4.305317324185249e-05,
      "loss": 0.0396,
      "step": 9150
    },
    {
      "epoch": 7.89,
      "learning_rate": 4.219554030874786e-05,
      "loss": 0.0402,
      "step": 9200
    },
    {
      "epoch": 7.93,
      "learning_rate": 4.1337907375643224e-05,
      "loss": 0.0419,
      "step": 9250
    },
    {
      "epoch": 7.97,
      "learning_rate": 4.04802744425386e-05,
      "loss": 0.0403,
      "step": 9300
    },
    {
      "epoch": 8.02,
      "learning_rate": 3.962264150943397e-05,
      "loss": 0.0313,
      "step": 9350
    },
    {
      "epoch": 8.06,
      "learning_rate": 3.876500857632933e-05,
      "loss": 0.0147,
      "step": 9400
    },
    {
      "epoch": 8.1,
      "learning_rate": 3.7907375643224705e-05,
      "loss": 0.0174,
      "step": 9450
    },
    {
      "epoch": 8.14,
      "learning_rate": 3.704974271012007e-05,
      "loss": 0.0159,
      "step": 9500
    },
    {
      "epoch": 8.19,
      "learning_rate": 3.6192109777015435e-05,
      "loss": 0.0151,
      "step": 9550
    },
    {
      "epoch": 8.23,
      "learning_rate": 3.533447684391081e-05,
      "loss": 0.0165,
      "step": 9600
    },
    {
      "epoch": 8.27,
      "learning_rate": 3.447684391080618e-05,
      "loss": 0.0165,
      "step": 9650
    },
    {
      "epoch": 8.32,
      "learning_rate": 3.361921097770154e-05,
      "loss": 0.0154,
      "step": 9700
    },
    {
      "epoch": 8.36,
      "learning_rate": 3.2761578044596916e-05,
      "loss": 0.0158,
      "step": 9750
    },
    {
      "epoch": 8.4,
      "learning_rate": 3.1903945111492285e-05,
      "loss": 0.0162,
      "step": 9800
    },
    {
      "epoch": 8.44,
      "learning_rate": 3.1046312178387653e-05,
      "loss": 0.0142,
      "step": 9850
    },
    {
      "epoch": 8.49,
      "learning_rate": 3.018867924528302e-05,
      "loss": 0.0172,
      "step": 9900
    },
    {
      "epoch": 8.53,
      "learning_rate": 2.933104631217839e-05,
      "loss": 0.0178,
      "step": 9950
    },
    {
      "epoch": 8.57,
      "learning_rate": 2.847341337907376e-05,
      "loss": 0.0146,
      "step": 10000
    },
    {
      "epoch": 8.62,
      "learning_rate": 2.7615780445969124e-05,
      "loss": 0.0144,
      "step": 10050
    },
    {
      "epoch": 8.66,
      "learning_rate": 2.6758147512864496e-05,
      "loss": 0.0141,
      "step": 10100
    },
    {
      "epoch": 8.7,
      "learning_rate": 2.5900514579759865e-05,
      "loss": 0.0143,
      "step": 10150
    },
    {
      "epoch": 8.74,
      "learning_rate": 2.504288164665523e-05,
      "loss": 0.0164,
      "step": 10200
    },
    {
      "epoch": 8.79,
      "learning_rate": 2.4185248713550602e-05,
      "loss": 0.017,
      "step": 10250
    },
    {
      "epoch": 8.83,
      "learning_rate": 2.332761578044597e-05,
      "loss": 0.0156,
      "step": 10300
    },
    {
      "epoch": 8.87,
      "learning_rate": 2.246998284734134e-05,
      "loss": 0.0139,
      "step": 10350
    },
    {
      "epoch": 8.92,
      "learning_rate": 2.1612349914236708e-05,
      "loss": 0.0147,
      "step": 10400
    },
    {
      "epoch": 8.96,
      "learning_rate": 2.0754716981132076e-05,
      "loss": 0.0151,
      "step": 10450
    },
    {
      "epoch": 9.0,
      "learning_rate": 1.9897084048027445e-05,
      "loss": 0.015,
      "step": 10500
    },
    {
      "epoch": 9.04,
      "learning_rate": 1.9039451114922813e-05,
      "loss": 0.006,
      "step": 10550
    },
    {
      "epoch": 9.09,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.0064,
      "step": 10600
    },
    {
      "epoch": 9.13,
      "learning_rate": 1.732418524871355e-05,
      "loss": 0.007,
      "step": 10650
    },
    {
      "epoch": 9.17,
      "learning_rate": 1.6466552315608923e-05,
      "loss": 0.0062,
      "step": 10700
    },
    {
      "epoch": 9.22,
      "learning_rate": 1.5608919382504288e-05,
      "loss": 0.0055,
      "step": 10750
    },
    {
      "epoch": 9.26,
      "learning_rate": 1.4751286449399656e-05,
      "loss": 0.0058,
      "step": 10800
    },
    {
      "epoch": 9.3,
      "learning_rate": 1.3893653516295027e-05,
      "loss": 0.0062,
      "step": 10850
    },
    {
      "epoch": 9.34,
      "learning_rate": 1.3036020583190395e-05,
      "loss": 0.0057,
      "step": 10900
    },
    {
      "epoch": 9.39,
      "learning_rate": 1.2178387650085764e-05,
      "loss": 0.0064,
      "step": 10950
    },
    {
      "epoch": 9.43,
      "learning_rate": 1.1320754716981132e-05,
      "loss": 0.0062,
      "step": 11000
    },
    {
      "epoch": 9.47,
      "learning_rate": 1.0463121783876503e-05,
      "loss": 0.006,
      "step": 11050
    },
    {
      "epoch": 9.52,
      "learning_rate": 9.60548885077187e-06,
      "loss": 0.0064,
      "step": 11100
    },
    {
      "epoch": 9.56,
      "learning_rate": 8.747855917667238e-06,
      "loss": 0.0061,
      "step": 11150
    },
    {
      "epoch": 9.6,
      "learning_rate": 7.890222984562608e-06,
      "loss": 0.0056,
      "step": 11200
    },
    {
      "epoch": 9.64,
      "learning_rate": 7.032590051457977e-06,
      "loss": 0.0056,
      "step": 11250
    },
    {
      "epoch": 9.69,
      "learning_rate": 6.1749571183533456e-06,
      "loss": 0.0059,
      "step": 11300
    },
    {
      "epoch": 9.73,
      "learning_rate": 5.317324185248713e-06,
      "loss": 0.0062,
      "step": 11350
    },
    {
      "epoch": 9.77,
      "learning_rate": 4.459691252144083e-06,
      "loss": 0.0056,
      "step": 11400
    },
    {
      "epoch": 9.82,
      "learning_rate": 3.6020583190394517e-06,
      "loss": 0.0061,
      "step": 11450
    },
    {
      "epoch": 9.86,
      "learning_rate": 2.74442538593482e-06,
      "loss": 0.0061,
      "step": 11500
    },
    {
      "epoch": 9.9,
      "learning_rate": 1.8867924528301887e-06,
      "loss": 0.0056,
      "step": 11550
    },
    {
      "epoch": 9.94,
      "learning_rate": 1.0291595197255577e-06,
      "loss": 0.0058,
      "step": 11600
    },
    {
      "epoch": 9.99,
      "learning_rate": 1.7152658662092624e-07,
      "loss": 0.0064,
      "step": 11650
    }
  ],
  "logging_steps": 50,
  "max_steps": 11660,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 2.652879536528646e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}