{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2948547840188707,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.948547840188707e-05,
      "grad_norm": 4288.0,
      "learning_rate": 5.89622641509434e-10,
      "loss": 5.8964,
      "step": 1
    },
    {
      "epoch": 0.00014742739200943535,
      "grad_norm": 5152.0,
      "learning_rate": 2.9481132075471697e-09,
      "loss": 7.2418,
      "step": 5
    },
    {
      "epoch": 0.0002948547840188707,
      "grad_norm": 3872.0,
      "learning_rate": 5.896226415094339e-09,
      "loss": 6.7926,
      "step": 10
    },
    {
      "epoch": 0.0004422821760283061,
      "grad_norm": 2704.0,
      "learning_rate": 8.844339622641509e-09,
      "loss": 7.0144,
      "step": 15
    },
    {
      "epoch": 0.0005897095680377414,
      "grad_norm": 5824.0,
      "learning_rate": 1.1792452830188679e-08,
      "loss": 7.431,
      "step": 20
    },
    {
      "epoch": 0.0007371369600471767,
      "grad_norm": 3488.0,
      "learning_rate": 1.4740566037735849e-08,
      "loss": 7.3504,
      "step": 25
    },
    {
      "epoch": 0.0008845643520566122,
      "grad_norm": 4448.0,
      "learning_rate": 1.7688679245283017e-08,
      "loss": 6.8362,
      "step": 30
    },
    {
      "epoch": 0.0010319917440660474,
      "grad_norm": 3984.0,
      "learning_rate": 2.0636792452830187e-08,
      "loss": 7.2579,
      "step": 35
    },
    {
      "epoch": 0.0011794191360754828,
      "grad_norm": 6112.0,
      "learning_rate": 2.3584905660377358e-08,
      "loss": 7.0144,
      "step": 40
    },
    {
      "epoch": 0.0013268465280849183,
      "grad_norm": 3536.0,
      "learning_rate": 2.6533018867924528e-08,
      "loss": 6.6585,
      "step": 45
    },
    {
      "epoch": 0.0014742739200943535,
      "grad_norm": 5184.0,
      "learning_rate": 2.9481132075471698e-08,
      "loss": 7.1249,
      "step": 50
    },
    {
      "epoch": 0.001621701312103789,
      "grad_norm": 4736.0,
      "learning_rate": 3.242924528301887e-08,
      "loss": 7.0672,
      "step": 55
    },
    {
      "epoch": 0.0017691287041132243,
      "grad_norm": 3888.0,
      "learning_rate": 3.5377358490566035e-08,
      "loss": 6.9418,
      "step": 60
    },
    {
      "epoch": 0.0019165560961226596,
      "grad_norm": 4512.0,
      "learning_rate": 3.83254716981132e-08,
      "loss": 6.7915,
      "step": 65
    },
    {
      "epoch": 0.0020639834881320948,
      "grad_norm": 4544.0,
      "learning_rate": 4.1273584905660375e-08,
      "loss": 6.9867,
      "step": 70
    },
    {
      "epoch": 0.0022114108801415304,
      "grad_norm": 4160.0,
      "learning_rate": 4.422169811320754e-08,
      "loss": 6.9295,
      "step": 75
    },
    {
      "epoch": 0.0023588382721509656,
      "grad_norm": 5024.0,
      "learning_rate": 4.7169811320754715e-08,
      "loss": 7.1398,
      "step": 80
    },
    {
      "epoch": 0.002506265664160401,
      "grad_norm": 6688.0,
      "learning_rate": 5.011792452830189e-08,
      "loss": 6.8533,
      "step": 85
    },
    {
      "epoch": 0.0026536930561698365,
      "grad_norm": 4128.0,
      "learning_rate": 5.3066037735849055e-08,
      "loss": 6.8373,
      "step": 90
    },
    {
      "epoch": 0.0028011204481792717,
      "grad_norm": 4512.0,
      "learning_rate": 5.601415094339622e-08,
      "loss": 6.9857,
      "step": 95
    },
    {
      "epoch": 0.002948547840188707,
      "grad_norm": 4704.0,
      "learning_rate": 5.8962264150943396e-08,
      "loss": 7.3043,
      "step": 100
    },
    {
      "epoch": 0.0030959752321981426,
      "grad_norm": 5824.0,
      "learning_rate": 6.191037735849057e-08,
      "loss": 7.4446,
      "step": 105
    },
    {
      "epoch": 0.003243402624207578,
      "grad_norm": 5600.0,
      "learning_rate": 6.485849056603774e-08,
      "loss": 6.4237,
      "step": 110
    },
    {
      "epoch": 0.003390830016217013,
      "grad_norm": 3360.0,
      "learning_rate": 6.78066037735849e-08,
      "loss": 7.1134,
      "step": 115
    },
    {
      "epoch": 0.0035382574082264487,
      "grad_norm": 5120.0,
      "learning_rate": 7.075471698113207e-08,
      "loss": 6.541,
      "step": 120
    },
    {
      "epoch": 0.003685684800235884,
      "grad_norm": 4016.0,
      "learning_rate": 7.370283018867925e-08,
      "loss": 7.1745,
      "step": 125
    },
    {
      "epoch": 0.003833112192245319,
      "grad_norm": 4096.0,
      "learning_rate": 7.66509433962264e-08,
      "loss": 6.6107,
      "step": 130
    },
    {
      "epoch": 0.003980539584254755,
      "grad_norm": 6784.0,
      "learning_rate": 7.959905660377358e-08,
      "loss": 7.1709,
      "step": 135
    },
    {
      "epoch": 0.0041279669762641896,
      "grad_norm": 4800.0,
      "learning_rate": 8.254716981132075e-08,
      "loss": 6.7303,
      "step": 140
    },
    {
      "epoch": 0.004275394368273625,
      "grad_norm": 5472.0,
      "learning_rate": 8.549528301886792e-08,
      "loss": 6.6023,
      "step": 145
    },
    {
      "epoch": 0.004422821760283061,
      "grad_norm": 5952.0,
      "learning_rate": 8.844339622641508e-08,
      "loss": 7.3574,
      "step": 150
    },
    {
      "epoch": 0.004570249152292496,
      "grad_norm": 4480.0,
      "learning_rate": 9.139150943396226e-08,
      "loss": 6.4739,
      "step": 155
    },
    {
      "epoch": 0.004717676544301931,
      "grad_norm": 2928.0,
      "learning_rate": 9.433962264150943e-08,
      "loss": 6.5303,
      "step": 160
    },
    {
      "epoch": 0.004865103936311367,
      "grad_norm": 4064.0,
      "learning_rate": 9.72877358490566e-08,
      "loss": 6.7455,
      "step": 165
    },
    {
      "epoch": 0.005012531328320802,
      "grad_norm": 2928.0,
      "learning_rate": 1.0023584905660378e-07,
      "loss": 6.2466,
      "step": 170
    },
    {
      "epoch": 0.005159958720330237,
      "grad_norm": 3536.0,
      "learning_rate": 1.0318396226415093e-07,
      "loss": 6.493,
      "step": 175
    },
    {
      "epoch": 0.005307386112339673,
      "grad_norm": 3792.0,
      "learning_rate": 1.0613207547169811e-07,
      "loss": 6.1627,
      "step": 180
    },
    {
      "epoch": 0.005454813504349108,
      "grad_norm": 5568.0,
      "learning_rate": 1.0908018867924528e-07,
      "loss": 6.7054,
      "step": 185
    },
    {
      "epoch": 0.0056022408963585435,
      "grad_norm": 5312.0,
      "learning_rate": 1.1202830188679244e-07,
      "loss": 6.3761,
      "step": 190
    },
    {
      "epoch": 0.005749668288367979,
      "grad_norm": 3888.0,
      "learning_rate": 1.1497641509433961e-07,
      "loss": 6.4957,
      "step": 195
    },
    {
      "epoch": 0.005897095680377414,
      "grad_norm": 4016.0,
      "learning_rate": 1.1792452830188679e-07,
      "loss": 6.3387,
      "step": 200
    },
    {
      "epoch": 0.0060445230723868495,
      "grad_norm": 3712.0,
      "learning_rate": 1.2087264150943396e-07,
      "loss": 6.8239,
      "step": 205
    },
    {
      "epoch": 0.006191950464396285,
      "grad_norm": 2224.0,
      "learning_rate": 1.2382075471698114e-07,
      "loss": 6.4956,
      "step": 210
    },
    {
      "epoch": 0.00633937785640572,
      "grad_norm": 4736.0,
      "learning_rate": 1.267688679245283e-07,
      "loss": 6.2678,
      "step": 215
    },
    {
      "epoch": 0.006486805248415156,
      "grad_norm": 4320.0,
      "learning_rate": 1.2971698113207547e-07,
      "loss": 6.4021,
      "step": 220
    },
    {
      "epoch": 0.006634232640424591,
      "grad_norm": 4032.0,
      "learning_rate": 1.3266509433962265e-07,
      "loss": 5.8306,
      "step": 225
    },
    {
      "epoch": 0.006781660032434026,
      "grad_norm": 3664.0,
      "learning_rate": 1.356132075471698e-07,
      "loss": 6.7351,
      "step": 230
    },
    {
      "epoch": 0.006929087424443462,
      "grad_norm": 2032.0,
      "learning_rate": 1.3856132075471696e-07,
      "loss": 6.3457,
      "step": 235
    },
    {
      "epoch": 0.007076514816452897,
      "grad_norm": 2784.0,
      "learning_rate": 1.4150943396226414e-07,
      "loss": 6.1769,
      "step": 240
    },
    {
      "epoch": 0.007223942208462332,
      "grad_norm": 1464.0,
      "learning_rate": 1.4445754716981132e-07,
      "loss": 5.8652,
      "step": 245
    },
    {
      "epoch": 0.007371369600471768,
      "grad_norm": 2008.0,
      "learning_rate": 1.474056603773585e-07,
      "loss": 6.0493,
      "step": 250
    },
    {
      "epoch": 0.007518796992481203,
      "grad_norm": 1040.0,
      "learning_rate": 1.5035377358490565e-07,
      "loss": 5.9201,
      "step": 255
    },
    {
      "epoch": 0.007666224384490638,
      "grad_norm": 1816.0,
      "learning_rate": 1.533018867924528e-07,
      "loss": 6.2471,
      "step": 260
    },
    {
      "epoch": 0.007813651776500074,
      "grad_norm": 1432.0,
      "learning_rate": 1.5624999999999999e-07,
      "loss": 5.8522,
      "step": 265
    },
    {
      "epoch": 0.00796107916850951,
      "grad_norm": 1816.0,
      "learning_rate": 1.5919811320754717e-07,
      "loss": 5.9513,
      "step": 270
    },
    {
      "epoch": 0.008108506560518945,
      "grad_norm": 520.0,
      "learning_rate": 1.6214622641509435e-07,
      "loss": 5.784,
      "step": 275
    },
    {
      "epoch": 0.008255933952528379,
      "grad_norm": 1272.0,
      "learning_rate": 1.650943396226415e-07,
      "loss": 5.6907,
      "step": 280
    },
    {
      "epoch": 0.008403361344537815,
      "grad_norm": 852.0,
      "learning_rate": 1.6804245283018868e-07,
      "loss": 5.6086,
      "step": 285
    },
    {
      "epoch": 0.00855078873654725,
      "grad_norm": 266.0,
      "learning_rate": 1.7099056603773583e-07,
      "loss": 5.3121,
      "step": 290
    },
    {
      "epoch": 0.008698216128556686,
      "grad_norm": 756.0,
      "learning_rate": 1.7393867924528301e-07,
      "loss": 5.6198,
      "step": 295
    },
    {
      "epoch": 0.008845643520566122,
      "grad_norm": 568.0,
      "learning_rate": 1.7688679245283017e-07,
      "loss": 5.5308,
      "step": 300
    },
    {
      "epoch": 0.008993070912575557,
      "grad_norm": 644.0,
      "learning_rate": 1.7983490566037735e-07,
      "loss": 5.522,
      "step": 305
    },
    {
      "epoch": 0.009140498304584991,
      "grad_norm": 596.0,
      "learning_rate": 1.8278301886792453e-07,
      "loss": 5.6054,
      "step": 310
    },
    {
      "epoch": 0.009287925696594427,
      "grad_norm": 384.0,
      "learning_rate": 1.857311320754717e-07,
      "loss": 5.5085,
      "step": 315
    },
    {
      "epoch": 0.009435353088603863,
      "grad_norm": 504.0,
      "learning_rate": 1.8867924528301886e-07,
      "loss": 5.6657,
      "step": 320
    },
    {
      "epoch": 0.009582780480613298,
      "grad_norm": 460.0,
      "learning_rate": 1.9162735849056601e-07,
      "loss": 5.7185,
      "step": 325
    },
    {
      "epoch": 0.009730207872622734,
      "grad_norm": 141.0,
      "learning_rate": 1.945754716981132e-07,
      "loss": 5.5482,
      "step": 330
    },
    {
      "epoch": 0.009877635264632168,
      "grad_norm": 560.0,
      "learning_rate": 1.9752358490566037e-07,
      "loss": 5.378,
      "step": 335
    },
    {
      "epoch": 0.010025062656641603,
      "grad_norm": 512.0,
      "learning_rate": 2.0047169811320755e-07,
      "loss": 5.6247,
      "step": 340
    },
    {
      "epoch": 0.010172490048651039,
      "grad_norm": 528.0,
      "learning_rate": 2.034198113207547e-07,
      "loss": 5.1692,
      "step": 345
    },
    {
      "epoch": 0.010319917440660475,
      "grad_norm": 624.0,
      "learning_rate": 2.0636792452830186e-07,
      "loss": 5.43,
      "step": 350
    },
    {
      "epoch": 0.01046734483266991,
      "grad_norm": 592.0,
      "learning_rate": 2.0931603773584904e-07,
      "loss": 5.2031,
      "step": 355
    },
    {
      "epoch": 0.010614772224679346,
      "grad_norm": 396.0,
      "learning_rate": 2.1226415094339622e-07,
      "loss": 5.1502,
      "step": 360
    },
    {
      "epoch": 0.01076219961668878,
      "grad_norm": 516.0,
      "learning_rate": 2.1521226415094338e-07,
      "loss": 5.826,
      "step": 365
    },
    {
      "epoch": 0.010909627008698216,
      "grad_norm": 760.0,
      "learning_rate": 2.1816037735849056e-07,
      "loss": 5.5087,
      "step": 370
    },
    {
      "epoch": 0.011057054400707651,
      "grad_norm": 420.0,
      "learning_rate": 2.2110849056603774e-07,
      "loss": 5.146,
      "step": 375
    },
    {
      "epoch": 0.011204481792717087,
      "grad_norm": 192.0,
      "learning_rate": 2.240566037735849e-07,
      "loss": 5.3402,
      "step": 380
    },
    {
      "epoch": 0.011351909184726523,
      "grad_norm": 370.0,
      "learning_rate": 2.2700471698113207e-07,
      "loss": 5.3036,
      "step": 385
    },
    {
      "epoch": 0.011499336576735958,
      "grad_norm": 352.0,
      "learning_rate": 2.2995283018867922e-07,
      "loss": 5.1917,
      "step": 390
    },
    {
      "epoch": 0.011646763968745392,
      "grad_norm": 414.0,
      "learning_rate": 2.329009433962264e-07,
      "loss": 5.0797,
      "step": 395
    },
    {
      "epoch": 0.011794191360754828,
      "grad_norm": 484.0,
      "learning_rate": 2.3584905660377358e-07,
      "loss": 5.296,
      "step": 400
    },
    {
      "epoch": 0.011941618752764263,
      "grad_norm": 508.0,
      "learning_rate": 2.3879716981132076e-07,
      "loss": 5.026,
      "step": 405
    },
    {
      "epoch": 0.012089046144773699,
      "grad_norm": 450.0,
      "learning_rate": 2.417452830188679e-07,
      "loss": 4.9075,
      "step": 410
    },
    {
      "epoch": 0.012236473536783135,
      "grad_norm": 382.0,
      "learning_rate": 2.4469339622641507e-07,
      "loss": 5.0301,
      "step": 415
    },
    {
      "epoch": 0.01238390092879257,
      "grad_norm": 384.0,
      "learning_rate": 2.476415094339623e-07,
      "loss": 4.9027,
      "step": 420
    },
    {
      "epoch": 0.012531328320802004,
      "grad_norm": 188.0,
      "learning_rate": 2.505896226415094e-07,
      "loss": 5.0141,
      "step": 425
    },
    {
      "epoch": 0.01267875571281144,
      "grad_norm": 157.0,
      "learning_rate": 2.535377358490566e-07,
      "loss": 4.5994,
      "step": 430
    },
    {
      "epoch": 0.012826183104820876,
      "grad_norm": 386.0,
      "learning_rate": 2.5648584905660374e-07,
      "loss": 4.8006,
      "step": 435
    },
    {
      "epoch": 0.012973610496830311,
      "grad_norm": 270.0,
      "learning_rate": 2.5943396226415094e-07,
      "loss": 4.4348,
      "step": 440
    },
    {
      "epoch": 0.013121037888839747,
      "grad_norm": 161.0,
      "learning_rate": 2.623820754716981e-07,
      "loss": 4.7565,
      "step": 445
    },
    {
      "epoch": 0.013268465280849183,
      "grad_norm": 186.0,
      "learning_rate": 2.653301886792453e-07,
      "loss": 4.7097,
      "step": 450
    },
    {
      "epoch": 0.013415892672858616,
      "grad_norm": 115.0,
      "learning_rate": 2.6827830188679246e-07,
      "loss": 4.3228,
      "step": 455
    },
    {
      "epoch": 0.013563320064868052,
      "grad_norm": 83.0,
      "learning_rate": 2.712264150943396e-07,
      "loss": 4.52,
      "step": 460
    },
    {
      "epoch": 0.013710747456877488,
      "grad_norm": 142.0,
      "learning_rate": 2.7417452830188676e-07,
      "loss": 4.4627,
      "step": 465
    },
    {
      "epoch": 0.013858174848886923,
      "grad_norm": 247.0,
      "learning_rate": 2.771226415094339e-07,
      "loss": 4.2847,
      "step": 470
    },
    {
      "epoch": 0.014005602240896359,
      "grad_norm": 59.0,
      "learning_rate": 2.800707547169811e-07,
      "loss": 4.5636,
      "step": 475
    },
    {
      "epoch": 0.014153029632905795,
      "grad_norm": 168.0,
      "learning_rate": 2.830188679245283e-07,
      "loss": 4.3123,
      "step": 480
    },
    {
      "epoch": 0.014300457024915229,
      "grad_norm": 106.0,
      "learning_rate": 2.8596698113207543e-07,
      "loss": 4.3122,
      "step": 485
    },
    {
      "epoch": 0.014447884416924664,
      "grad_norm": 97.0,
      "learning_rate": 2.8891509433962264e-07,
      "loss": 4.2267,
      "step": 490
    },
    {
      "epoch": 0.0145953118089341,
      "grad_norm": 73.0,
      "learning_rate": 2.918632075471698e-07,
      "loss": 4.1379,
      "step": 495
    },
    {
      "epoch": 0.014742739200943536,
      "grad_norm": 60.25,
      "learning_rate": 2.94811320754717e-07,
      "loss": 4.0958,
      "step": 500
    },
    {
      "epoch": 0.014742739200943536,
      "eval_loss": 5.273393630981445,
      "eval_runtime": 4.7076,
      "eval_samples_per_second": 84.119,
      "eval_steps_per_second": 2.761,
      "step": 500
    },
    {
      "epoch": 0.014890166592952971,
      "grad_norm": 77.0,
      "learning_rate": 2.9775943396226415e-07,
      "loss": 3.849,
      "step": 505
    },
    {
      "epoch": 0.015037593984962405,
      "grad_norm": 83.0,
      "learning_rate": 3.007075471698113e-07,
      "loss": 4.2059,
      "step": 510
    },
    {
      "epoch": 0.01518502137697184,
      "grad_norm": 120.0,
      "learning_rate": 3.0365566037735846e-07,
      "loss": 4.1134,
      "step": 515
    },
    {
      "epoch": 0.015332448768981276,
      "grad_norm": 58.25,
      "learning_rate": 3.066037735849056e-07,
      "loss": 3.9189,
      "step": 520
    },
    {
      "epoch": 0.015479876160990712,
      "grad_norm": 77.0,
      "learning_rate": 3.095518867924528e-07,
      "loss": 4.1662,
      "step": 525
    },
    {
      "epoch": 0.015627303553000148,
      "grad_norm": 120.5,
      "learning_rate": 3.1249999999999997e-07,
      "loss": 4.0004,
      "step": 530
    },
    {
      "epoch": 0.01577473094500958,
      "grad_norm": 51.75,
      "learning_rate": 3.154481132075472e-07,
      "loss": 4.1284,
      "step": 535
    },
    {
      "epoch": 0.01592215833701902,
      "grad_norm": 56.0,
      "learning_rate": 3.1839622641509433e-07,
      "loss": 4.1131,
      "step": 540
    },
    {
      "epoch": 0.016069585729028453,
      "grad_norm": 87.0,
      "learning_rate": 3.2134433962264154e-07,
      "loss": 4.0497,
      "step": 545
    },
    {
      "epoch": 0.01621701312103789,
      "grad_norm": 53.25,
      "learning_rate": 3.242924528301887e-07,
      "loss": 3.8763,
      "step": 550
    },
    {
      "epoch": 0.016364440513047324,
      "grad_norm": 80.5,
      "learning_rate": 3.272405660377358e-07,
      "loss": 4.1176,
      "step": 555
    },
    {
      "epoch": 0.016511867905056758,
      "grad_norm": 72.0,
      "learning_rate": 3.30188679245283e-07,
      "loss": 4.0426,
      "step": 560
    },
    {
      "epoch": 0.016659295297066196,
      "grad_norm": 52.25,
      "learning_rate": 3.3313679245283015e-07,
      "loss": 3.7246,
      "step": 565
    },
    {
      "epoch": 0.01680672268907563,
      "grad_norm": 50.0,
      "learning_rate": 3.3608490566037736e-07,
      "loss": 3.914,
      "step": 570
    },
    {
      "epoch": 0.016954150081085067,
      "grad_norm": 48.25,
      "learning_rate": 3.390330188679245e-07,
      "loss": 4.1672,
      "step": 575
    },
    {
      "epoch": 0.0171015774730945,
      "grad_norm": 56.5,
      "learning_rate": 3.4198113207547167e-07,
      "loss": 3.7885,
      "step": 580
    },
    {
      "epoch": 0.017249004865103935,
      "grad_norm": 130.0,
      "learning_rate": 3.449292452830189e-07,
      "loss": 3.9511,
      "step": 585
    },
    {
      "epoch": 0.017396432257113372,
      "grad_norm": 69.5,
      "learning_rate": 3.4787735849056603e-07,
      "loss": 4.0633,
      "step": 590
    },
    {
      "epoch": 0.017543859649122806,
      "grad_norm": 49.5,
      "learning_rate": 3.5082547169811323e-07,
      "loss": 3.801,
      "step": 595
    },
    {
      "epoch": 0.017691287041132243,
      "grad_norm": 38.0,
      "learning_rate": 3.5377358490566033e-07,
      "loss": 3.6217,
      "step": 600
    },
    {
      "epoch": 0.017838714433141677,
      "grad_norm": 35.75,
      "learning_rate": 3.567216981132075e-07,
      "loss": 3.9652,
      "step": 605
    },
    {
      "epoch": 0.017986141825151115,
      "grad_norm": 52.0,
      "learning_rate": 3.596698113207547e-07,
      "loss": 3.7182,
      "step": 610
    },
    {
      "epoch": 0.01813356921716055,
      "grad_norm": 58.5,
      "learning_rate": 3.6261792452830185e-07,
      "loss": 3.6423,
      "step": 615
    },
    {
      "epoch": 0.018280996609169983,
      "grad_norm": 46.25,
      "learning_rate": 3.6556603773584905e-07,
      "loss": 3.7992,
      "step": 620
    },
    {
      "epoch": 0.01842842400117942,
      "grad_norm": 42.25,
      "learning_rate": 3.685141509433962e-07,
      "loss": 3.8397,
      "step": 625
    },
    {
      "epoch": 0.018575851393188854,
      "grad_norm": 36.0,
      "learning_rate": 3.714622641509434e-07,
      "loss": 3.6173,
      "step": 630
    },
    {
      "epoch": 0.01872327878519829,
      "grad_norm": 49.75,
      "learning_rate": 3.7441037735849057e-07,
      "loss": 4.0317,
      "step": 635
    },
    {
      "epoch": 0.018870706177207725,
      "grad_norm": 38.5,
      "learning_rate": 3.773584905660377e-07,
      "loss": 3.7633,
      "step": 640
    },
    {
      "epoch": 0.01901813356921716,
      "grad_norm": 41.75,
      "learning_rate": 3.803066037735849e-07,
      "loss": 3.8188,
      "step": 645
    },
    {
      "epoch": 0.019165560961226596,
      "grad_norm": 63.5,
      "learning_rate": 3.8325471698113203e-07,
      "loss": 3.7537,
      "step": 650
    },
    {
      "epoch": 0.01931298835323603,
      "grad_norm": 49.25,
      "learning_rate": 3.8620283018867924e-07,
      "loss": 3.6774,
      "step": 655
    },
    {
      "epoch": 0.019460415745245468,
      "grad_norm": 39.0,
      "learning_rate": 3.891509433962264e-07,
      "loss": 3.7567,
      "step": 660
    },
    {
      "epoch": 0.0196078431372549,
      "grad_norm": 39.0,
      "learning_rate": 3.920990566037736e-07,
      "loss": 3.756,
      "step": 665
    },
    {
      "epoch": 0.019755270529264336,
      "grad_norm": 69.0,
      "learning_rate": 3.9504716981132075e-07,
      "loss": 3.6765,
      "step": 670
    },
    {
      "epoch": 0.019902697921273773,
      "grad_norm": 60.75,
      "learning_rate": 3.979952830188679e-07,
      "loss": 3.949,
      "step": 675
    },
    {
      "epoch": 0.020050125313283207,
      "grad_norm": 47.25,
      "learning_rate": 4.009433962264151e-07,
      "loss": 3.7218,
      "step": 680
    },
    {
      "epoch": 0.020197552705292644,
      "grad_norm": 93.5,
      "learning_rate": 4.038915094339622e-07,
      "loss": 3.6829,
      "step": 685
    },
    {
      "epoch": 0.020344980097302078,
      "grad_norm": 39.0,
      "learning_rate": 4.068396226415094e-07,
      "loss": 3.5724,
      "step": 690
    },
    {
      "epoch": 0.020492407489311516,
      "grad_norm": 62.0,
      "learning_rate": 4.0978773584905657e-07,
      "loss": 3.6344,
      "step": 695
    },
    {
      "epoch": 0.02063983488132095,
      "grad_norm": 196.0,
      "learning_rate": 4.127358490566037e-07,
      "loss": 3.5317,
      "step": 700
    },
    {
      "epoch": 0.020787262273330383,
      "grad_norm": 41.0,
      "learning_rate": 4.1568396226415093e-07,
      "loss": 3.9269,
      "step": 705
    },
    {
      "epoch": 0.02093468966533982,
      "grad_norm": 49.0,
      "learning_rate": 4.186320754716981e-07,
      "loss": 3.7463,
      "step": 710
    },
    {
      "epoch": 0.021082117057349255,
      "grad_norm": 61.75,
      "learning_rate": 4.215801886792453e-07,
      "loss": 3.426,
      "step": 715
    },
    {
      "epoch": 0.021229544449358692,
      "grad_norm": 40.25,
      "learning_rate": 4.2452830188679244e-07,
      "loss": 3.5511,
      "step": 720
    },
    {
      "epoch": 0.021376971841368126,
      "grad_norm": 60.25,
      "learning_rate": 4.2747641509433965e-07,
      "loss": 3.8714,
      "step": 725
    },
    {
      "epoch": 0.02152439923337756,
      "grad_norm": 67.0,
      "learning_rate": 4.3042452830188675e-07,
      "loss": 3.756,
      "step": 730
    },
    {
      "epoch": 0.021671826625386997,
      "grad_norm": 47.0,
      "learning_rate": 4.333726415094339e-07,
      "loss": 3.548,
      "step": 735
    },
    {
      "epoch": 0.02181925401739643,
      "grad_norm": 32.25,
      "learning_rate": 4.363207547169811e-07,
      "loss": 3.6436,
      "step": 740
    },
    {
      "epoch": 0.02196668140940587,
      "grad_norm": 54.75,
      "learning_rate": 4.3926886792452826e-07,
      "loss": 3.4437,
      "step": 745
    },
    {
      "epoch": 0.022114108801415303,
      "grad_norm": 55.75,
      "learning_rate": 4.4221698113207547e-07,
      "loss": 3.6474,
      "step": 750
    },
    {
      "epoch": 0.02226153619342474,
      "grad_norm": 49.0,
      "learning_rate": 4.451650943396226e-07,
      "loss": 3.7664,
      "step": 755
    },
    {
      "epoch": 0.022408963585434174,
      "grad_norm": 37.0,
      "learning_rate": 4.481132075471698e-07,
      "loss": 3.4262,
      "step": 760
    },
    {
      "epoch": 0.022556390977443608,
      "grad_norm": 72.0,
      "learning_rate": 4.51061320754717e-07,
      "loss": 3.5507,
      "step": 765
    },
    {
      "epoch": 0.022703818369453045,
      "grad_norm": 34.75,
      "learning_rate": 4.5400943396226414e-07,
      "loss": 3.6294,
      "step": 770
    },
    {
      "epoch": 0.02285124576146248,
      "grad_norm": 71.0,
      "learning_rate": 4.569575471698113e-07,
      "loss": 3.3822,
      "step": 775
    },
    {
      "epoch": 0.022998673153471916,
      "grad_norm": 57.5,
      "learning_rate": 4.5990566037735845e-07,
      "loss": 3.6301,
      "step": 780
    },
    {
      "epoch": 0.02314610054548135,
      "grad_norm": 44.5,
      "learning_rate": 4.6285377358490565e-07,
      "loss": 3.5542,
      "step": 785
    },
    {
      "epoch": 0.023293527937490784,
      "grad_norm": 30.25,
      "learning_rate": 4.658018867924528e-07,
      "loss": 3.4738,
      "step": 790
    },
    {
      "epoch": 0.02344095532950022,
      "grad_norm": 36.25,
      "learning_rate": 4.6874999999999996e-07,
      "loss": 3.6581,
      "step": 795
    },
    {
      "epoch": 0.023588382721509656,
      "grad_norm": 27.5,
      "learning_rate": 4.7169811320754717e-07,
      "loss": 3.4342,
      "step": 800
    },
    {
      "epoch": 0.023735810113519093,
      "grad_norm": 36.5,
      "learning_rate": 4.746462264150943e-07,
      "loss": 3.6826,
      "step": 805
    },
    {
      "epoch": 0.023883237505528527,
      "grad_norm": 41.75,
      "learning_rate": 4.775943396226415e-07,
      "loss": 3.485,
      "step": 810
    },
    {
      "epoch": 0.024030664897537964,
      "grad_norm": 171.0,
      "learning_rate": 4.805424528301887e-07,
      "loss": 3.5775,
      "step": 815
    },
    {
      "epoch": 0.024178092289547398,
      "grad_norm": 50.25,
      "learning_rate": 4.834905660377358e-07,
      "loss": 3.6154,
      "step": 820
    },
    {
      "epoch": 0.024325519681556832,
      "grad_norm": 28.875,
      "learning_rate": 4.86438679245283e-07,
      "loss": 3.6943,
      "step": 825
    },
    {
      "epoch": 0.02447294707356627,
      "grad_norm": 35.75,
      "learning_rate": 4.893867924528301e-07,
      "loss": 3.7264,
      "step": 830
    },
    {
      "epoch": 0.024620374465575703,
      "grad_norm": 46.5,
      "learning_rate": 4.923349056603773e-07,
      "loss": 3.4934,
      "step": 835
    },
    {
      "epoch": 0.02476780185758514,
      "grad_norm": 42.0,
      "learning_rate": 4.952830188679246e-07,
      "loss": 3.5727,
      "step": 840
    },
    {
      "epoch": 0.024915229249594575,
      "grad_norm": 27.5,
      "learning_rate": 4.982311320754717e-07,
      "loss": 3.3648,
      "step": 845
    },
    {
      "epoch": 0.02506265664160401,
      "grad_norm": 22.75,
      "learning_rate": 5.011792452830188e-07,
      "loss": 3.5408,
      "step": 850
    },
    {
      "epoch": 0.025210084033613446,
      "grad_norm": 44.75,
      "learning_rate": 5.04127358490566e-07,
      "loss": 3.7565,
      "step": 855
    },
    {
      "epoch": 0.02535751142562288,
      "grad_norm": 33.75,
      "learning_rate": 5.070754716981132e-07,
      "loss": 3.5236,
      "step": 860
    },
    {
      "epoch": 0.025504938817632317,
      "grad_norm": 79.0,
      "learning_rate": 5.100235849056603e-07,
      "loss": 3.4544,
      "step": 865
    },
    {
      "epoch": 0.02565236620964175,
      "grad_norm": 54.75,
      "learning_rate": 5.129716981132075e-07,
      "loss": 3.3027,
      "step": 870
    },
    {
      "epoch": 0.025799793601651185,
      "grad_norm": 37.0,
      "learning_rate": 5.159198113207547e-07,
      "loss": 3.2541,
      "step": 875
    },
    {
      "epoch": 0.025947220993660623,
      "grad_norm": 33.0,
      "learning_rate": 5.188679245283019e-07,
      "loss": 3.5423,
      "step": 880
    },
    {
      "epoch": 0.026094648385670056,
      "grad_norm": 31.125,
      "learning_rate": 5.21816037735849e-07,
      "loss": 3.5534,
      "step": 885
    },
    {
      "epoch": 0.026242075777679494,
      "grad_norm": 30.25,
      "learning_rate": 5.247641509433962e-07,
      "loss": 3.6136,
      "step": 890
    },
    {
      "epoch": 0.026389503169688928,
      "grad_norm": 29.125,
      "learning_rate": 5.277122641509433e-07,
      "loss": 3.3698,
      "step": 895
    },
    {
      "epoch": 0.026536930561698365,
      "grad_norm": 30.5,
      "learning_rate": 5.306603773584906e-07,
      "loss": 3.4885,
      "step": 900
    },
    {
      "epoch": 0.0266843579537078,
      "grad_norm": 24.375,
      "learning_rate": 5.336084905660378e-07,
      "loss": 3.3895,
      "step": 905
    },
    {
      "epoch": 0.026831785345717233,
      "grad_norm": 34.5,
      "learning_rate": 5.365566037735849e-07,
      "loss": 3.5425,
      "step": 910
    },
    {
      "epoch": 0.02697921273772667,
      "grad_norm": 53.5,
      "learning_rate": 5.395047169811321e-07,
      "loss": 3.5077,
      "step": 915
    },
    {
      "epoch": 0.027126640129736104,
      "grad_norm": 25.375,
      "learning_rate": 5.424528301886792e-07,
      "loss": 3.3913,
      "step": 920
    },
    {
      "epoch": 0.02727406752174554,
      "grad_norm": 27.75,
      "learning_rate": 5.454009433962265e-07,
      "loss": 3.467,
      "step": 925
    },
    {
      "epoch": 0.027421494913754976,
      "grad_norm": 53.0,
      "learning_rate": 5.483490566037735e-07,
      "loss": 3.7339,
      "step": 930
    },
    {
      "epoch": 0.02756892230576441,
      "grad_norm": 34.0,
      "learning_rate": 5.512971698113207e-07,
      "loss": 3.4162,
      "step": 935
    },
    {
      "epoch": 0.027716349697773847,
      "grad_norm": 26.125,
      "learning_rate": 5.542452830188678e-07,
      "loss": 3.4891,
      "step": 940
    },
    {
      "epoch": 0.02786377708978328,
      "grad_norm": 30.625,
      "learning_rate": 5.57193396226415e-07,
      "loss": 3.6314,
      "step": 945
    },
    {
      "epoch": 0.028011204481792718,
      "grad_norm": 26.5,
      "learning_rate": 5.601415094339622e-07,
      "loss": 3.5523,
      "step": 950
    },
    {
      "epoch": 0.028158631873802152,
      "grad_norm": 33.75,
      "learning_rate": 5.630896226415094e-07,
      "loss": 3.3893,
      "step": 955
    },
    {
      "epoch": 0.02830605926581159,
      "grad_norm": 32.0,
      "learning_rate": 5.660377358490566e-07,
      "loss": 3.3843,
      "step": 960
    },
    {
      "epoch": 0.028453486657821023,
      "grad_norm": 29.5,
      "learning_rate": 5.689858490566037e-07,
      "loss": 3.5737,
      "step": 965
    },
    {
      "epoch": 0.028600914049830457,
      "grad_norm": 71.0,
      "learning_rate": 5.719339622641509e-07,
      "loss": 3.2828,
      "step": 970
    },
    {
      "epoch": 0.028748341441839895,
      "grad_norm": 42.25,
      "learning_rate": 5.748820754716981e-07,
      "loss": 3.414,
      "step": 975
    },
    {
      "epoch": 0.02889576883384933,
      "grad_norm": 30.0,
      "learning_rate": 5.778301886792453e-07,
      "loss": 3.3214,
      "step": 980
    },
    {
      "epoch": 0.029043196225858766,
      "grad_norm": 24.0,
      "learning_rate": 5.807783018867924e-07,
      "loss": 3.3956,
      "step": 985
    },
    {
      "epoch": 0.0291906236178682,
      "grad_norm": 29.0,
      "learning_rate": 5.837264150943396e-07,
      "loss": 3.4536,
      "step": 990
    },
    {
      "epoch": 0.029338051009877634,
      "grad_norm": 73.0,
      "learning_rate": 5.866745283018868e-07,
      "loss": 3.8505,
      "step": 995
    },
    {
      "epoch": 0.02948547840188707,
      "grad_norm": 28.875,
      "learning_rate": 5.89622641509434e-07,
      "loss": 3.4702,
      "step": 1000
    },
    {
      "epoch": 0.02948547840188707,
      "eval_loss": 4.123366832733154,
      "eval_runtime": 4.7173,
      "eval_samples_per_second": 83.946,
      "eval_steps_per_second": 2.756,
      "step": 1000
    },
    {
      "epoch": 0.029632905793896505,
      "grad_norm": 40.5,
      "learning_rate": 5.925707547169812e-07,
      "loss": 3.4618,
      "step": 1005
    },
    {
      "epoch": 0.029780333185905943,
      "grad_norm": 31.0,
      "learning_rate": 5.955188679245283e-07,
      "loss": 3.3067,
      "step": 1010
    },
    {
      "epoch": 0.029927760577915376,
      "grad_norm": 25.5,
      "learning_rate": 5.984669811320755e-07,
      "loss": 3.2988,
      "step": 1015
    },
    {
      "epoch": 0.03007518796992481,
      "grad_norm": 24.875,
      "learning_rate": 6.014150943396226e-07,
      "loss": 3.4574,
      "step": 1020
    },
    {
      "epoch": 0.030222615361934248,
      "grad_norm": 34.0,
      "learning_rate": 6.043632075471698e-07,
      "loss": 3.4777,
      "step": 1025
    },
    {
      "epoch": 0.03037004275394368,
      "grad_norm": 21.375,
      "learning_rate": 6.073113207547169e-07,
      "loss": 3.2491,
      "step": 1030
    },
    {
      "epoch": 0.03051747014595312,
      "grad_norm": 32.75,
      "learning_rate": 6.102594339622641e-07,
      "loss": 3.4012,
      "step": 1035
    },
    {
      "epoch": 0.030664897537962553,
      "grad_norm": 20.625,
      "learning_rate": 6.132075471698112e-07,
      "loss": 3.2219,
      "step": 1040
    },
    {
      "epoch": 0.03081232492997199,
      "grad_norm": 28.625,
      "learning_rate": 6.161556603773585e-07,
      "loss": 3.4702,
      "step": 1045
    },
    {
      "epoch": 0.030959752321981424,
      "grad_norm": 32.5,
      "learning_rate": 6.191037735849056e-07,
      "loss": 3.2885,
      "step": 1050
    },
    {
      "epoch": 0.031107179713990858,
      "grad_norm": 28.0,
      "learning_rate": 6.220518867924528e-07,
      "loss": 3.414,
      "step": 1055
    },
    {
      "epoch": 0.031254607106000296,
      "grad_norm": 24.125,
      "learning_rate": 6.249999999999999e-07,
      "loss": 3.35,
      "step": 1060
    },
    {
      "epoch": 0.03140203449800973,
      "grad_norm": 26.25,
      "learning_rate": 6.279481132075471e-07,
      "loss": 3.4552,
      "step": 1065
    },
    {
      "epoch": 0.03154946189001916,
      "grad_norm": 78.0,
      "learning_rate": 6.308962264150944e-07,
      "loss": 3.2664,
      "step": 1070
    },
    {
      "epoch": 0.0316968892820286,
      "grad_norm": 27.375,
      "learning_rate": 6.338443396226415e-07,
      "loss": 3.4228,
      "step": 1075
    },
    {
      "epoch": 0.03184431667403804,
      "grad_norm": 32.25,
      "learning_rate": 6.367924528301887e-07,
      "loss": 3.2762,
      "step": 1080
    },
    {
      "epoch": 0.03199174406604747,
      "grad_norm": 25.875,
      "learning_rate": 6.397405660377358e-07,
      "loss": 3.2858,
      "step": 1085
    },
    {
      "epoch": 0.032139171458056906,
      "grad_norm": 27.875,
      "learning_rate": 6.426886792452831e-07,
      "loss": 3.3251,
      "step": 1090
    },
    {
      "epoch": 0.03228659885006634,
      "grad_norm": 21.5,
      "learning_rate": 6.456367924528302e-07,
      "loss": 3.3962,
      "step": 1095
    },
    {
      "epoch": 0.03243402624207578,
      "grad_norm": 20.5,
      "learning_rate": 6.485849056603774e-07,
      "loss": 3.4683,
      "step": 1100
    },
    {
      "epoch": 0.03258145363408521,
      "grad_norm": 23.75,
      "learning_rate": 6.515330188679244e-07,
      "loss": 3.2019,
      "step": 1105
    },
    {
      "epoch": 0.03272888102609465,
      "grad_norm": 55.5,
      "learning_rate": 6.544811320754716e-07,
      "loss": 3.4679,
      "step": 1110
    },
    {
      "epoch": 0.032876308418104086,
      "grad_norm": 36.75,
      "learning_rate": 6.574292452830188e-07,
      "loss": 3.3551,
      "step": 1115
    },
    {
      "epoch": 0.033023735810113516,
      "grad_norm": 23.5,
      "learning_rate": 6.60377358490566e-07,
      "loss": 3.3352,
      "step": 1120
    },
    {
      "epoch": 0.033171163202122954,
      "grad_norm": 31.5,
      "learning_rate": 6.633254716981132e-07,
      "loss": 3.4598,
      "step": 1125
    },
    {
      "epoch": 0.03331859059413239,
      "grad_norm": 36.0,
      "learning_rate": 6.662735849056603e-07,
      "loss": 3.5186,
      "step": 1130
    },
    {
      "epoch": 0.03346601798614183,
      "grad_norm": 34.75,
      "learning_rate": 6.692216981132075e-07,
      "loss": 3.3776,
      "step": 1135
    },
    {
      "epoch": 0.03361344537815126,
      "grad_norm": 45.0,
      "learning_rate": 6.721698113207547e-07,
      "loss": 3.4942,
      "step": 1140
    },
    {
      "epoch": 0.033760872770160696,
      "grad_norm": 21.0,
      "learning_rate": 6.751179245283019e-07,
      "loss": 3.4764,
      "step": 1145
    },
    {
      "epoch": 0.033908300162170134,
      "grad_norm": 21.875,
      "learning_rate": 6.78066037735849e-07,
      "loss": 3.3914,
      "step": 1150
    },
    {
      "epoch": 0.034055727554179564,
      "grad_norm": 22.375,
      "learning_rate": 6.810141509433962e-07,
      "loss": 3.2916,
      "step": 1155
    },
    {
      "epoch": 0.034203154946189,
      "grad_norm": 24.875,
      "learning_rate": 6.839622641509433e-07,
      "loss": 3.3966,
      "step": 1160
    },
    {
      "epoch": 0.03435058233819844,
      "grad_norm": 32.25,
      "learning_rate": 6.869103773584906e-07,
      "loss": 3.2619,
      "step": 1165
    },
    {
      "epoch": 0.03449800973020787,
      "grad_norm": 24.875,
      "learning_rate": 6.898584905660377e-07,
      "loss": 3.2777,
      "step": 1170
    },
    {
      "epoch": 0.03464543712221731,
      "grad_norm": 20.875,
      "learning_rate": 6.928066037735849e-07,
      "loss": 3.228,
      "step": 1175
    },
    {
      "epoch": 0.034792864514226744,
      "grad_norm": 24.0,
      "learning_rate": 6.957547169811321e-07,
      "loss": 3.2484,
      "step": 1180
    },
    {
      "epoch": 0.03494029190623618,
      "grad_norm": 21.5,
      "learning_rate": 6.987028301886792e-07,
      "loss": 3.2644,
      "step": 1185
    },
    {
      "epoch": 0.03508771929824561,
      "grad_norm": 21.0,
      "learning_rate": 7.016509433962265e-07,
      "loss": 3.0695,
      "step": 1190
    },
    {
      "epoch": 0.03523514669025505,
      "grad_norm": 35.25,
      "learning_rate": 7.045990566037735e-07,
      "loss": 3.3176,
      "step": 1195
    },
    {
      "epoch": 0.03538257408226449,
      "grad_norm": 19.25,
      "learning_rate": 7.075471698113207e-07,
      "loss": 3.2097,
      "step": 1200
    },
    {
      "epoch": 0.03553000147427392,
      "grad_norm": 25.875,
      "learning_rate": 7.104952830188678e-07,
      "loss": 3.3202,
      "step": 1205
    },
    {
      "epoch": 0.035677428866283355,
      "grad_norm": 27.25,
      "learning_rate": 7.13443396226415e-07,
      "loss": 3.1499,
      "step": 1210
    },
    {
      "epoch": 0.03582485625829279,
      "grad_norm": 34.25,
      "learning_rate": 7.163915094339622e-07,
      "loss": 3.308,
      "step": 1215
    },
    {
      "epoch": 0.03597228365030223,
      "grad_norm": 23.875,
      "learning_rate": 7.193396226415094e-07,
      "loss": 3.0438,
      "step": 1220
    },
    {
      "epoch": 0.03611971104231166,
      "grad_norm": 26.5,
      "learning_rate": 7.222877358490565e-07,
      "loss": 3.2286,
      "step": 1225
    },
    {
      "epoch": 0.0362671384343211,
      "grad_norm": 25.0,
      "learning_rate": 7.252358490566037e-07,
      "loss": 3.1494,
      "step": 1230
    },
    {
      "epoch": 0.036414565826330535,
      "grad_norm": 33.5,
      "learning_rate": 7.28183962264151e-07,
      "loss": 3.3187,
      "step": 1235
    },
    {
      "epoch": 0.036561993218339965,
      "grad_norm": 20.625,
      "learning_rate": 7.311320754716981e-07,
      "loss": 3.2469,
      "step": 1240
    },
    {
      "epoch": 0.0367094206103494,
      "grad_norm": 20.75,
      "learning_rate": 7.340801886792453e-07,
      "loss": 3.2862,
      "step": 1245
    },
    {
      "epoch": 0.03685684800235884,
      "grad_norm": 27.875,
      "learning_rate": 7.370283018867924e-07,
      "loss": 3.2839,
      "step": 1250
    },
    {
      "epoch": 0.03700427539436827,
      "grad_norm": 20.75,
      "learning_rate": 7.399764150943396e-07,
      "loss": 3.3897,
      "step": 1255
    },
    {
      "epoch": 0.03715170278637771,
      "grad_norm": 27.125,
      "learning_rate": 7.429245283018868e-07,
      "loss": 3.1699,
      "step": 1260
    },
    {
      "epoch": 0.037299130178387145,
      "grad_norm": 21.75,
      "learning_rate": 7.45872641509434e-07,
      "loss": 3.163,
      "step": 1265
    },
    {
      "epoch": 0.03744655757039658,
      "grad_norm": 21.625,
      "learning_rate": 7.488207547169811e-07,
      "loss": 3.244,
      "step": 1270
    },
    {
      "epoch": 0.03759398496240601,
      "grad_norm": 27.75,
      "learning_rate": 7.517688679245283e-07,
      "loss": 3.2307,
      "step": 1275
    },
    {
      "epoch": 0.03774141235441545,
      "grad_norm": 40.75,
      "learning_rate": 7.547169811320754e-07,
      "loss": 3.2905,
      "step": 1280
    },
    {
      "epoch": 0.03788883974642489,
      "grad_norm": 25.25,
      "learning_rate": 7.576650943396226e-07,
      "loss": 3.3318,
      "step": 1285
    },
    {
      "epoch": 0.03803626713843432,
      "grad_norm": 32.5,
      "learning_rate": 7.606132075471698e-07,
      "loss": 3.2618,
      "step": 1290
    },
    {
      "epoch": 0.038183694530443756,
      "grad_norm": 33.5,
      "learning_rate": 7.635613207547169e-07,
      "loss": 3.3272,
      "step": 1295
    },
    {
      "epoch": 0.03833112192245319,
      "grad_norm": 28.125,
      "learning_rate": 7.665094339622641e-07,
      "loss": 3.1578,
      "step": 1300
    },
    {
      "epoch": 0.03847854931446263,
      "grad_norm": 21.375,
      "learning_rate": 7.694575471698112e-07,
      "loss": 3.0873,
      "step": 1305
    },
    {
      "epoch": 0.03862597670647206,
      "grad_norm": 23.75,
      "learning_rate": 7.724056603773585e-07,
      "loss": 3.3165,
      "step": 1310
    },
    {
      "epoch": 0.0387734040984815,
      "grad_norm": 21.5,
      "learning_rate": 7.753537735849056e-07,
      "loss": 3.1587,
      "step": 1315
    },
    {
      "epoch": 0.038920831490490936,
      "grad_norm": 20.5,
      "learning_rate": 7.783018867924528e-07,
      "loss": 3.3898,
      "step": 1320
    },
    {
      "epoch": 0.039068258882500366,
      "grad_norm": 34.0,
      "learning_rate": 7.812499999999999e-07,
      "loss": 3.1372,
      "step": 1325
    },
    {
      "epoch": 0.0392156862745098,
      "grad_norm": 18.875,
      "learning_rate": 7.841981132075472e-07,
      "loss": 3.2379,
      "step": 1330
    },
    {
      "epoch": 0.03936311366651924,
      "grad_norm": 21.125,
      "learning_rate": 7.871462264150943e-07,
      "loss": 3.3005,
      "step": 1335
    },
    {
      "epoch": 0.03951054105852867,
      "grad_norm": 23.25,
      "learning_rate": 7.900943396226415e-07,
      "loss": 3.1946,
      "step": 1340
    },
    {
      "epoch": 0.03965796845053811,
      "grad_norm": 23.375,
      "learning_rate": 7.930424528301887e-07,
      "loss": 3.0714,
      "step": 1345
    },
    {
      "epoch": 0.039805395842547546,
      "grad_norm": 28.375,
      "learning_rate": 7.959905660377358e-07,
      "loss": 3.3614,
      "step": 1350
    },
    {
      "epoch": 0.03995282323455698,
      "grad_norm": 24.25,
      "learning_rate": 7.989386792452831e-07,
      "loss": 3.1599,
      "step": 1355
    },
    {
      "epoch": 0.040100250626566414,
      "grad_norm": 20.375,
      "learning_rate": 8.018867924528302e-07,
      "loss": 3.1418,
      "step": 1360
    },
    {
      "epoch": 0.04024767801857585,
      "grad_norm": 23.5,
      "learning_rate": 8.048349056603774e-07,
      "loss": 2.9725,
      "step": 1365
    },
    {
      "epoch": 0.04039510541058529,
      "grad_norm": 29.0,
      "learning_rate": 8.077830188679244e-07,
      "loss": 3.0553,
      "step": 1370
    },
    {
      "epoch": 0.04054253280259472,
      "grad_norm": 29.0,
      "learning_rate": 8.107311320754716e-07,
      "loss": 3.1955,
      "step": 1375
    },
    {
      "epoch": 0.040689960194604156,
      "grad_norm": 21.75,
      "learning_rate": 8.136792452830188e-07,
      "loss": 3.1639,
      "step": 1380
    },
    {
      "epoch": 0.040837387586613594,
      "grad_norm": 33.0,
      "learning_rate": 8.16627358490566e-07,
      "loss": 3.1981,
      "step": 1385
    },
    {
      "epoch": 0.04098481497862303,
      "grad_norm": 25.625,
      "learning_rate": 8.195754716981131e-07,
      "loss": 3.081,
      "step": 1390
    },
    {
      "epoch": 0.04113224237063246,
      "grad_norm": 24.0,
      "learning_rate": 8.225235849056603e-07,
      "loss": 3.0706,
      "step": 1395
    },
    {
      "epoch": 0.0412796697626419,
      "grad_norm": 23.0,
      "learning_rate": 8.254716981132074e-07,
      "loss": 3.2926,
      "step": 1400
    },
    {
      "epoch": 0.041427097154651336,
      "grad_norm": 23.375,
      "learning_rate": 8.284198113207547e-07,
      "loss": 3.1699,
      "step": 1405
    },
    {
      "epoch": 0.04157452454666077,
      "grad_norm": 30.0,
      "learning_rate": 8.313679245283019e-07,
      "loss": 3.145,
      "step": 1410
    },
    {
      "epoch": 0.041721951938670204,
      "grad_norm": 29.25,
      "learning_rate": 8.34316037735849e-07,
      "loss": 3.2222,
      "step": 1415
    },
    {
      "epoch": 0.04186937933067964,
      "grad_norm": 36.5,
      "learning_rate": 8.372641509433962e-07,
      "loss": 3.2383,
      "step": 1420
    },
    {
      "epoch": 0.04201680672268908,
      "grad_norm": 19.875,
      "learning_rate": 8.402122641509433e-07,
      "loss": 2.9331,
      "step": 1425
    },
    {
      "epoch": 0.04216423411469851,
      "grad_norm": 24.375,
      "learning_rate": 8.431603773584906e-07,
      "loss": 3.076,
      "step": 1430
    },
    {
      "epoch": 0.04231166150670795,
      "grad_norm": 24.125,
      "learning_rate": 8.461084905660377e-07,
      "loss": 3.1641,
      "step": 1435
    },
    {
      "epoch": 0.042459088898717384,
      "grad_norm": 21.375,
      "learning_rate": 8.490566037735849e-07,
      "loss": 3.3244,
      "step": 1440
    },
    {
      "epoch": 0.042606516290726815,
      "grad_norm": 21.125,
      "learning_rate": 8.52004716981132e-07,
      "loss": 3.1178,
      "step": 1445
    },
    {
      "epoch": 0.04275394368273625,
      "grad_norm": 27.0,
      "learning_rate": 8.549528301886793e-07,
      "loss": 3.1739,
      "step": 1450
    },
    {
      "epoch": 0.04290137107474569,
      "grad_norm": 21.625,
      "learning_rate": 8.579009433962265e-07,
      "loss": 3.2333,
      "step": 1455
    },
    {
      "epoch": 0.04304879846675512,
      "grad_norm": 31.25,
      "learning_rate": 8.608490566037735e-07,
      "loss": 3.2602,
      "step": 1460
    },
    {
      "epoch": 0.04319622585876456,
      "grad_norm": 22.0,
      "learning_rate": 8.637971698113207e-07,
      "loss": 3.2519,
      "step": 1465
    },
    {
      "epoch": 0.043343653250773995,
      "grad_norm": 24.625,
      "learning_rate": 8.667452830188678e-07,
      "loss": 3.0988,
      "step": 1470
    },
    {
      "epoch": 0.04349108064278343,
      "grad_norm": 28.125,
      "learning_rate": 8.696933962264151e-07,
      "loss": 3.1193,
      "step": 1475
    },
    {
      "epoch": 0.04363850803479286,
      "grad_norm": 20.625,
      "learning_rate": 8.726415094339622e-07,
      "loss": 3.2805,
      "step": 1480
    },
    {
      "epoch": 0.0437859354268023,
      "grad_norm": 18.875,
      "learning_rate": 8.755896226415094e-07,
      "loss": 3.2113,
      "step": 1485
    },
    {
      "epoch": 0.04393336281881174,
      "grad_norm": 17.625,
      "learning_rate": 8.785377358490565e-07,
      "loss": 3.2982,
      "step": 1490
    },
    {
      "epoch": 0.04408079021082117,
      "grad_norm": 17.375,
      "learning_rate": 8.814858490566037e-07,
      "loss": 3.187,
      "step": 1495
    },
    {
      "epoch": 0.044228217602830605,
      "grad_norm": 21.75,
      "learning_rate": 8.844339622641509e-07,
      "loss": 3.1377,
      "step": 1500
    },
    {
      "epoch": 0.044228217602830605,
      "eval_loss": 3.5992095470428467,
      "eval_runtime": 4.6937,
      "eval_samples_per_second": 84.368,
      "eval_steps_per_second": 2.77,
      "step": 1500
    },
    {
      "epoch": 0.04437564499484004,
      "grad_norm": 188.0,
      "learning_rate": 8.873820754716981e-07,
      "loss": 3.1779,
      "step": 1505
    },
    {
      "epoch": 0.04452307238684948,
      "grad_norm": 45.5,
      "learning_rate": 8.903301886792452e-07,
      "loss": 3.2703,
      "step": 1510
    },
    {
      "epoch": 0.04467049977885891,
      "grad_norm": 20.875,
      "learning_rate": 8.932783018867924e-07,
      "loss": 3.1839,
      "step": 1515
    },
    {
      "epoch": 0.04481792717086835,
      "grad_norm": 19.375,
      "learning_rate": 8.962264150943396e-07,
      "loss": 3.0066,
      "step": 1520
    },
    {
      "epoch": 0.044965354562877785,
      "grad_norm": 29.5,
      "learning_rate": 8.991745283018868e-07,
      "loss": 3.2096,
      "step": 1525
    },
    {
      "epoch": 0.045112781954887216,
      "grad_norm": 26.875,
      "learning_rate": 9.02122641509434e-07,
      "loss": 3.0473,
      "step": 1530
    },
    {
      "epoch": 0.04526020934689665,
      "grad_norm": 19.625,
      "learning_rate": 9.050707547169811e-07,
      "loss": 3.0378,
      "step": 1535
    },
    {
      "epoch": 0.04540763673890609,
      "grad_norm": 17.5,
      "learning_rate": 9.080188679245283e-07,
      "loss": 3.2259,
      "step": 1540
    },
    {
      "epoch": 0.04555506413091552,
      "grad_norm": 22.625,
      "learning_rate": 9.109669811320755e-07,
      "loss": 3.0831,
      "step": 1545
    },
    {
      "epoch": 0.04570249152292496,
      "grad_norm": 19.125,
      "learning_rate": 9.139150943396226e-07,
      "loss": 3.1252,
      "step": 1550
    },
    {
      "epoch": 0.045849918914934396,
      "grad_norm": 23.25,
      "learning_rate": 9.168632075471697e-07,
      "loss": 3.2609,
      "step": 1555
    },
    {
      "epoch": 0.04599734630694383,
      "grad_norm": 27.625,
      "learning_rate": 9.198113207547169e-07,
      "loss": 3.1041,
      "step": 1560
    },
    {
      "epoch": 0.04614477369895326,
      "grad_norm": 22.75,
      "learning_rate": 9.22759433962264e-07,
      "loss": 3.289,
      "step": 1565
    },
    {
      "epoch": 0.0462922010909627,
      "grad_norm": 19.125,
      "learning_rate": 9.257075471698113e-07,
      "loss": 3.2432,
      "step": 1570
    },
    {
      "epoch": 0.04643962848297214,
      "grad_norm": 16.5,
      "learning_rate": 9.286556603773585e-07,
      "loss": 3.0007,
      "step": 1575
    },
    {
      "epoch": 0.04658705587498157,
      "grad_norm": 24.5,
      "learning_rate": 9.316037735849056e-07,
      "loss": 3.2563,
      "step": 1580
    },
    {
      "epoch": 0.046734483266991006,
      "grad_norm": 19.75,
      "learning_rate": 9.345518867924528e-07,
      "loss": 3.0233,
      "step": 1585
    },
    {
      "epoch": 0.04688191065900044,
      "grad_norm": 24.0,
      "learning_rate": 9.374999999999999e-07,
      "loss": 2.9052,
      "step": 1590
    },
    {
      "epoch": 0.04702933805100988,
      "grad_norm": 21.375,
      "learning_rate": 9.404481132075472e-07,
      "loss": 3.157,
      "step": 1595
    },
    {
      "epoch": 0.04717676544301931,
      "grad_norm": 37.0,
      "learning_rate": 9.433962264150943e-07,
      "loss": 2.9625,
      "step": 1600
    },
    {
      "epoch": 0.04732419283502875,
      "grad_norm": 16.125,
      "learning_rate": 9.463443396226415e-07,
      "loss": 3.1684,
      "step": 1605
    },
    {
      "epoch": 0.047471620227038186,
      "grad_norm": 17.0,
      "learning_rate": 9.492924528301886e-07,
      "loss": 3.1345,
      "step": 1610
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 24.125,
      "learning_rate": 9.522405660377358e-07,
      "loss": 3.2707,
      "step": 1615
    },
    {
      "epoch": 0.047766475011057054,
      "grad_norm": 19.375,
      "learning_rate": 9.55188679245283e-07,
      "loss": 3.1682,
      "step": 1620
    },
    {
      "epoch": 0.04791390240306649,
      "grad_norm": 20.125,
      "learning_rate": 9.581367924528302e-07,
      "loss": 3.1835,
      "step": 1625
    },
    {
      "epoch": 0.04806132979507593,
      "grad_norm": 19.75,
      "learning_rate": 9.610849056603774e-07,
      "loss": 3.0946,
      "step": 1630
    },
    {
      "epoch": 0.04820875718708536,
      "grad_norm": 27.0,
      "learning_rate": 9.640330188679245e-07,
      "loss": 2.9732,
      "step": 1635
    },
    {
      "epoch": 0.048356184579094796,
      "grad_norm": 28.0,
      "learning_rate": 9.669811320754717e-07,
      "loss": 3.0809,
      "step": 1640
    },
    {
      "epoch": 0.048503611971104234,
      "grad_norm": 16.625,
      "learning_rate": 9.699292452830188e-07,
      "loss": 2.8662,
      "step": 1645
    },
    {
      "epoch": 0.048651039363113664,
      "grad_norm": 25.0,
      "learning_rate": 9.72877358490566e-07,
      "loss": 3.1653,
      "step": 1650
    },
    {
      "epoch": 0.0487984667551231,
      "grad_norm": 20.875,
      "learning_rate": 9.758254716981131e-07,
      "loss": 2.9102,
      "step": 1655
    },
    {
      "epoch": 0.04894589414713254,
      "grad_norm": 15.625,
      "learning_rate": 9.787735849056603e-07,
      "loss": 3.0904,
      "step": 1660
    },
    {
      "epoch": 0.04909332153914197,
      "grad_norm": 26.5,
      "learning_rate": 9.817216981132074e-07,
      "loss": 3.2188,
      "step": 1665
    },
    {
      "epoch": 0.04924074893115141,
      "grad_norm": 20.0,
      "learning_rate": 9.846698113207546e-07,
      "loss": 3.0935,
      "step": 1670
    },
    {
      "epoch": 0.049388176323160844,
      "grad_norm": 22.625,
      "learning_rate": 9.876179245283017e-07,
      "loss": 3.0112,
      "step": 1675
    },
    {
      "epoch": 0.04953560371517028,
      "grad_norm": 28.875,
      "learning_rate": 9.90566037735849e-07,
      "loss": 3.0347,
      "step": 1680
    },
    {
      "epoch": 0.04968303110717971,
      "grad_norm": 28.0,
      "learning_rate": 9.935141509433963e-07,
      "loss": 3.1383,
      "step": 1685
    },
    {
      "epoch": 0.04983045849918915,
      "grad_norm": 22.625,
      "learning_rate": 9.964622641509434e-07,
      "loss": 3.195,
      "step": 1690
    },
    {
      "epoch": 0.04997788589119859,
      "grad_norm": 17.625,
      "learning_rate": 9.994103773584906e-07,
      "loss": 3.0536,
      "step": 1695
    },
    {
      "epoch": 0.05012531328320802,
      "grad_norm": 20.875,
      "learning_rate": 1.0023584905660375e-06,
      "loss": 3.1255,
      "step": 1700
    },
    {
      "epoch": 0.050272740675217455,
      "grad_norm": 23.625,
      "learning_rate": 1.0053066037735849e-06,
      "loss": 3.0029,
      "step": 1705
    },
    {
      "epoch": 0.05042016806722689,
      "grad_norm": 21.875,
      "learning_rate": 1.008254716981132e-06,
      "loss": 3.0566,
      "step": 1710
    },
    {
      "epoch": 0.05056759545923633,
      "grad_norm": 22.875,
      "learning_rate": 1.0112028301886792e-06,
      "loss": 3.235,
      "step": 1715
    },
    {
      "epoch": 0.05071502285124576,
      "grad_norm": 21.25,
      "learning_rate": 1.0141509433962263e-06,
      "loss": 3.0852,
      "step": 1720
    },
    {
      "epoch": 0.0508624502432552,
      "grad_norm": 20.875,
      "learning_rate": 1.0170990566037737e-06,
      "loss": 2.9238,
      "step": 1725
    },
    {
      "epoch": 0.051009877635264635,
      "grad_norm": 21.0,
      "learning_rate": 1.0200471698113206e-06,
      "loss": 2.9247,
      "step": 1730
    },
    {
      "epoch": 0.051157305027274065,
      "grad_norm": 22.375,
      "learning_rate": 1.022995283018868e-06,
      "loss": 2.9932,
      "step": 1735
    },
    {
      "epoch": 0.0513047324192835,
      "grad_norm": 21.625,
      "learning_rate": 1.025943396226415e-06,
      "loss": 3.1733,
      "step": 1740
    },
    {
      "epoch": 0.05145215981129294,
      "grad_norm": 18.75,
      "learning_rate": 1.0288915094339623e-06,
      "loss": 3.0096,
      "step": 1745
    },
    {
      "epoch": 0.05159958720330237,
      "grad_norm": 16.5,
      "learning_rate": 1.0318396226415095e-06,
      "loss": 3.0647,
      "step": 1750
    },
    {
      "epoch": 0.05174701459531181,
      "grad_norm": 20.75,
      "learning_rate": 1.0347877358490566e-06,
      "loss": 3.0325,
      "step": 1755
    },
    {
      "epoch": 0.051894441987321245,
      "grad_norm": 16.875,
      "learning_rate": 1.0377358490566038e-06,
      "loss": 3.0077,
      "step": 1760
    },
    {
      "epoch": 0.05204186937933068,
      "grad_norm": 18.625,
      "learning_rate": 1.040683962264151e-06,
      "loss": 2.9983,
      "step": 1765
    },
    {
      "epoch": 0.05218929677134011,
      "grad_norm": 21.75,
      "learning_rate": 1.043632075471698e-06,
      "loss": 3.103,
      "step": 1770
    },
    {
      "epoch": 0.05233672416334955,
      "grad_norm": 19.375,
      "learning_rate": 1.0465801886792452e-06,
      "loss": 3.0497,
      "step": 1775
    },
    {
      "epoch": 0.05248415155535899,
      "grad_norm": 27.125,
      "learning_rate": 1.0495283018867924e-06,
      "loss": 3.037,
      "step": 1780
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 14.0625,
      "learning_rate": 1.0524764150943395e-06,
      "loss": 2.7835,
      "step": 1785
    },
    {
      "epoch": 0.052779006339377856,
      "grad_norm": 20.75,
      "learning_rate": 1.0554245283018867e-06,
      "loss": 3.0535,
      "step": 1790
    },
    {
      "epoch": 0.05292643373138729,
      "grad_norm": 19.375,
      "learning_rate": 1.0583726415094338e-06,
      "loss": 3.0061,
      "step": 1795
    },
    {
      "epoch": 0.05307386112339673,
      "grad_norm": 18.375,
      "learning_rate": 1.0613207547169812e-06,
      "loss": 2.9059,
      "step": 1800
    },
    {
      "epoch": 0.05322128851540616,
      "grad_norm": 16.875,
      "learning_rate": 1.0642688679245282e-06,
      "loss": 3.0449,
      "step": 1805
    },
    {
      "epoch": 0.0533687159074156,
      "grad_norm": 19.875,
      "learning_rate": 1.0672169811320755e-06,
      "loss": 2.905,
      "step": 1810
    },
    {
      "epoch": 0.053516143299425036,
      "grad_norm": 20.25,
      "learning_rate": 1.0701650943396225e-06,
      "loss": 3.2346,
      "step": 1815
    },
    {
      "epoch": 0.053663570691434466,
      "grad_norm": 17.875,
      "learning_rate": 1.0731132075471698e-06,
      "loss": 2.9078,
      "step": 1820
    },
    {
      "epoch": 0.0538109980834439,
      "grad_norm": 20.125,
      "learning_rate": 1.076061320754717e-06,
      "loss": 3.1477,
      "step": 1825
    },
    {
      "epoch": 0.05395842547545334,
      "grad_norm": 24.0,
      "learning_rate": 1.0790094339622641e-06,
      "loss": 2.8281,
      "step": 1830
    },
    {
      "epoch": 0.05410585286746277,
      "grad_norm": 20.625,
      "learning_rate": 1.0819575471698113e-06,
      "loss": 3.0376,
      "step": 1835
    },
    {
      "epoch": 0.05425328025947221,
      "grad_norm": 15.6875,
      "learning_rate": 1.0849056603773584e-06,
      "loss": 2.8431,
      "step": 1840
    },
    {
      "epoch": 0.054400707651481646,
      "grad_norm": 18.875,
      "learning_rate": 1.0878537735849056e-06,
      "loss": 2.9487,
      "step": 1845
    },
    {
      "epoch": 0.05454813504349108,
      "grad_norm": 30.75,
      "learning_rate": 1.090801886792453e-06,
      "loss": 2.9563,
      "step": 1850
    },
    {
      "epoch": 0.054695562435500514,
      "grad_norm": 23.125,
      "learning_rate": 1.09375e-06,
      "loss": 2.871,
      "step": 1855
    },
    {
      "epoch": 0.05484298982750995,
      "grad_norm": 17.0,
      "learning_rate": 1.096698113207547e-06,
      "loss": 3.029,
      "step": 1860
    },
    {
      "epoch": 0.05499041721951939,
      "grad_norm": 14.625,
      "learning_rate": 1.0996462264150942e-06,
      "loss": 2.898,
      "step": 1865
    },
    {
      "epoch": 0.05513784461152882,
      "grad_norm": 21.25,
      "learning_rate": 1.1025943396226414e-06,
      "loss": 2.8212,
      "step": 1870
    },
    {
      "epoch": 0.055285272003538256,
      "grad_norm": 17.75,
      "learning_rate": 1.1055424528301887e-06,
      "loss": 2.9897,
      "step": 1875
    },
    {
      "epoch": 0.055432699395547694,
      "grad_norm": 17.125,
      "learning_rate": 1.1084905660377357e-06,
      "loss": 2.8868,
      "step": 1880
    },
    {
      "epoch": 0.05558012678755713,
      "grad_norm": 19.25,
      "learning_rate": 1.111438679245283e-06,
      "loss": 3.0397,
      "step": 1885
    },
    {
      "epoch": 0.05572755417956656,
      "grad_norm": 19.75,
      "learning_rate": 1.11438679245283e-06,
      "loss": 2.8232,
      "step": 1890
    },
    {
      "epoch": 0.055874981571576,
      "grad_norm": 36.75,
      "learning_rate": 1.1173349056603773e-06,
      "loss": 2.9727,
      "step": 1895
    },
    {
      "epoch": 0.056022408963585436,
      "grad_norm": 39.25,
      "learning_rate": 1.1202830188679245e-06,
      "loss": 3.3028,
      "step": 1900
    },
    {
      "epoch": 0.05616983635559487,
      "grad_norm": 19.25,
      "learning_rate": 1.1232311320754717e-06,
      "loss": 3.1002,
      "step": 1905
    },
    {
      "epoch": 0.056317263747604304,
      "grad_norm": 24.0,
      "learning_rate": 1.1261792452830188e-06,
      "loss": 3.163,
      "step": 1910
    },
    {
      "epoch": 0.05646469113961374,
      "grad_norm": 23.375,
      "learning_rate": 1.129127358490566e-06,
      "loss": 2.9574,
      "step": 1915
    },
    {
      "epoch": 0.05661211853162318,
      "grad_norm": 62.0,
      "learning_rate": 1.1320754716981131e-06,
      "loss": 2.9711,
      "step": 1920
    },
    {
      "epoch": 0.05675954592363261,
      "grad_norm": 20.875,
      "learning_rate": 1.1350235849056605e-06,
      "loss": 2.9719,
      "step": 1925
    },
    {
      "epoch": 0.05690697331564205,
      "grad_norm": 21.0,
      "learning_rate": 1.1379716981132074e-06,
      "loss": 2.9451,
      "step": 1930
    },
    {
      "epoch": 0.057054400707651484,
      "grad_norm": 19.5,
      "learning_rate": 1.1409198113207548e-06,
      "loss": 3.0775,
      "step": 1935
    },
    {
      "epoch": 0.057201828099660915,
      "grad_norm": 18.0,
      "learning_rate": 1.1438679245283017e-06,
      "loss": 2.9611,
      "step": 1940
    },
    {
      "epoch": 0.05734925549167035,
      "grad_norm": 19.5,
      "learning_rate": 1.1468160377358489e-06,
      "loss": 2.8335,
      "step": 1945
    },
    {
      "epoch": 0.05749668288367979,
      "grad_norm": 19.125,
      "learning_rate": 1.1497641509433962e-06,
      "loss": 2.9033,
      "step": 1950
    },
    {
      "epoch": 0.05764411027568922,
      "grad_norm": 19.75,
      "learning_rate": 1.1527122641509432e-06,
      "loss": 2.957,
      "step": 1955
    },
    {
      "epoch": 0.05779153766769866,
      "grad_norm": 21.375,
      "learning_rate": 1.1556603773584906e-06,
      "loss": 3.0282,
      "step": 1960
    },
    {
      "epoch": 0.057938965059708095,
      "grad_norm": 26.375,
      "learning_rate": 1.1586084905660375e-06,
      "loss": 2.9575,
      "step": 1965
    },
    {
      "epoch": 0.05808639245171753,
      "grad_norm": 16.375,
      "learning_rate": 1.1615566037735849e-06,
      "loss": 2.8847,
      "step": 1970
    },
    {
      "epoch": 0.05823381984372696,
      "grad_norm": 21.625,
      "learning_rate": 1.164504716981132e-06,
      "loss": 2.8438,
      "step": 1975
    },
    {
      "epoch": 0.0583812472357364,
      "grad_norm": 26.375,
      "learning_rate": 1.1674528301886792e-06,
      "loss": 3.0832,
      "step": 1980
    },
    {
      "epoch": 0.05852867462774584,
      "grad_norm": 17.5,
      "learning_rate": 1.1704009433962263e-06,
      "loss": 3.0941,
      "step": 1985
    },
    {
      "epoch": 0.05867610201975527,
      "grad_norm": 16.25,
      "learning_rate": 1.1733490566037737e-06,
      "loss": 2.716,
      "step": 1990
    },
    {
      "epoch": 0.058823529411764705,
      "grad_norm": 71.0,
      "learning_rate": 1.1762971698113206e-06,
      "loss": 2.9973,
      "step": 1995
    },
    {
      "epoch": 0.05897095680377414,
      "grad_norm": 24.625,
      "learning_rate": 1.179245283018868e-06,
      "loss": 2.9267,
      "step": 2000
    },
    {
      "epoch": 0.05897095680377414,
      "eval_loss": 3.27067494392395,
      "eval_runtime": 4.6982,
      "eval_samples_per_second": 84.287,
      "eval_steps_per_second": 2.767,
      "step": 2000
    },
    {
      "epoch": 0.05911838419578358,
      "grad_norm": 16.5,
      "learning_rate": 1.182193396226415e-06,
      "loss": 2.7164,
      "step": 2005
    },
    {
      "epoch": 0.05926581158779301,
      "grad_norm": 14.8125,
      "learning_rate": 1.1851415094339623e-06,
      "loss": 2.8899,
      "step": 2010
    },
    {
      "epoch": 0.05941323897980245,
      "grad_norm": 18.75,
      "learning_rate": 1.1880896226415095e-06,
      "loss": 2.926,
      "step": 2015
    },
    {
      "epoch": 0.059560666371811885,
      "grad_norm": 23.125,
      "learning_rate": 1.1910377358490566e-06,
      "loss": 2.853,
      "step": 2020
    },
    {
      "epoch": 0.059708093763821316,
      "grad_norm": 17.75,
      "learning_rate": 1.1939858490566038e-06,
      "loss": 2.8584,
      "step": 2025
    },
    {
      "epoch": 0.05985552115583075,
      "grad_norm": 20.75,
      "learning_rate": 1.196933962264151e-06,
      "loss": 2.8365,
      "step": 2030
    },
    {
      "epoch": 0.06000294854784019,
      "grad_norm": 19.0,
      "learning_rate": 1.199882075471698e-06,
      "loss": 3.0571,
      "step": 2035
    },
    {
      "epoch": 0.06015037593984962,
      "grad_norm": 15.6875,
      "learning_rate": 1.2028301886792452e-06,
      "loss": 2.7813,
      "step": 2040
    },
    {
      "epoch": 0.06029780333185906,
      "grad_norm": 16.125,
      "learning_rate": 1.2057783018867924e-06,
      "loss": 2.744,
      "step": 2045
    },
    {
      "epoch": 0.060445230723868495,
      "grad_norm": 18.125,
      "learning_rate": 1.2087264150943395e-06,
      "loss": 2.8005,
      "step": 2050
    },
    {
      "epoch": 0.06059265811587793,
      "grad_norm": 22.875,
      "learning_rate": 1.2116745283018867e-06,
      "loss": 2.9454,
      "step": 2055
    },
    {
      "epoch": 0.06074008550788736,
      "grad_norm": 16.875,
      "learning_rate": 1.2146226415094338e-06,
      "loss": 3.0735,
      "step": 2060
    },
    {
      "epoch": 0.0608875128998968,
      "grad_norm": 40.75,
      "learning_rate": 1.2175707547169812e-06,
      "loss": 2.9832,
      "step": 2065
    },
    {
      "epoch": 0.06103494029190624,
      "grad_norm": 22.375,
      "learning_rate": 1.2205188679245281e-06,
      "loss": 3.0261,
      "step": 2070
    },
    {
      "epoch": 0.06118236768391567,
      "grad_norm": 19.375,
      "learning_rate": 1.2234669811320755e-06,
      "loss": 3.0268,
      "step": 2075
    },
    {
      "epoch": 0.061329795075925106,
      "grad_norm": 22.5,
      "learning_rate": 1.2264150943396225e-06,
      "loss": 2.9181,
      "step": 2080
    },
    {
      "epoch": 0.06147722246793454,
      "grad_norm": 43.75,
      "learning_rate": 1.2293632075471698e-06,
      "loss": 2.7782,
      "step": 2085
    },
    {
      "epoch": 0.06162464985994398,
      "grad_norm": 15.125,
      "learning_rate": 1.232311320754717e-06,
      "loss": 2.8703,
      "step": 2090
    },
    {
      "epoch": 0.06177207725195341,
      "grad_norm": 21.75,
      "learning_rate": 1.2352594339622641e-06,
      "loss": 2.972,
      "step": 2095
    },
    {
      "epoch": 0.06191950464396285,
      "grad_norm": 16.75,
      "learning_rate": 1.2382075471698113e-06,
      "loss": 2.8651,
      "step": 2100
    },
    {
      "epoch": 0.062066932035972286,
      "grad_norm": 23.375,
      "learning_rate": 1.2411556603773584e-06,
      "loss": 3.1965,
      "step": 2105
    },
    {
      "epoch": 0.062214359427981716,
      "grad_norm": 21.125,
      "learning_rate": 1.2441037735849056e-06,
      "loss": 2.9804,
      "step": 2110
    },
    {
      "epoch": 0.062361786819991154,
      "grad_norm": 23.125,
      "learning_rate": 1.247051886792453e-06,
      "loss": 2.9106,
      "step": 2115
    },
    {
      "epoch": 0.06250921421200059,
      "grad_norm": 19.5,
      "learning_rate": 1.2499999999999999e-06,
      "loss": 2.9487,
      "step": 2120
    },
    {
      "epoch": 0.06265664160401002,
      "grad_norm": 43.5,
      "learning_rate": 1.252948113207547e-06,
      "loss": 2.9301,
      "step": 2125
    },
    {
      "epoch": 0.06280406899601947,
      "grad_norm": 17.75,
      "learning_rate": 1.2558962264150942e-06,
      "loss": 2.8569,
      "step": 2130
    },
    {
      "epoch": 0.0629514963880289,
      "grad_norm": 14.625,
      "learning_rate": 1.2588443396226414e-06,
      "loss": 2.851,
      "step": 2135
    },
    {
      "epoch": 0.06309892378003833,
      "grad_norm": 19.0,
      "learning_rate": 1.2617924528301887e-06,
      "loss": 2.959,
      "step": 2140
    },
    {
      "epoch": 0.06324635117204777,
      "grad_norm": 16.0,
      "learning_rate": 1.2647405660377357e-06,
      "loss": 3.0111,
      "step": 2145
    },
    {
      "epoch": 0.0633937785640572,
      "grad_norm": 21.25,
      "learning_rate": 1.267688679245283e-06,
      "loss": 2.9492,
      "step": 2150
    },
    {
      "epoch": 0.06354120595606663,
      "grad_norm": 27.875,
      "learning_rate": 1.27063679245283e-06,
      "loss": 2.8556,
      "step": 2155
    },
    {
      "epoch": 0.06368863334807608,
      "grad_norm": 292.0,
      "learning_rate": 1.2735849056603773e-06,
      "loss": 3.0161,
      "step": 2160
    },
    {
      "epoch": 0.0638360607400855,
      "grad_norm": 27.625,
      "learning_rate": 1.2765330188679245e-06,
      "loss": 2.8943,
      "step": 2165
    },
    {
      "epoch": 0.06398348813209494,
      "grad_norm": 20.5,
      "learning_rate": 1.2794811320754716e-06,
      "loss": 2.9569,
      "step": 2170
    },
    {
      "epoch": 0.06413091552410438,
      "grad_norm": 16.375,
      "learning_rate": 1.2824292452830188e-06,
      "loss": 2.8933,
      "step": 2175
    },
    {
      "epoch": 0.06427834291611381,
      "grad_norm": 21.75,
      "learning_rate": 1.2853773584905662e-06,
      "loss": 2.8857,
      "step": 2180
    },
    {
      "epoch": 0.06442577030812324,
      "grad_norm": 25.375,
      "learning_rate": 1.288325471698113e-06,
      "loss": 2.7119,
      "step": 2185
    },
    {
      "epoch": 0.06457319770013269,
      "grad_norm": 15.5,
      "learning_rate": 1.2912735849056605e-06,
      "loss": 2.9744,
      "step": 2190
    },
    {
      "epoch": 0.06472062509214212,
      "grad_norm": 28.0,
      "learning_rate": 1.2942216981132074e-06,
      "loss": 2.8828,
      "step": 2195
    },
    {
      "epoch": 0.06486805248415156,
      "grad_norm": 17.625,
      "learning_rate": 1.2971698113207548e-06,
      "loss": 2.9426,
      "step": 2200
    },
    {
      "epoch": 0.06501547987616099,
      "grad_norm": 16.75,
      "learning_rate": 1.300117924528302e-06,
      "loss": 2.9991,
      "step": 2205
    },
    {
      "epoch": 0.06516290726817042,
      "grad_norm": 32.5,
      "learning_rate": 1.3030660377358489e-06,
      "loss": 2.9827,
      "step": 2210
    },
    {
      "epoch": 0.06531033466017987,
      "grad_norm": 33.0,
      "learning_rate": 1.3060141509433962e-06,
      "loss": 2.7431,
      "step": 2215
    },
    {
      "epoch": 0.0654577620521893,
      "grad_norm": 18.125,
      "learning_rate": 1.3089622641509432e-06,
      "loss": 2.8863,
      "step": 2220
    },
    {
      "epoch": 0.06560518944419873,
      "grad_norm": 16.625,
      "learning_rate": 1.3119103773584905e-06,
      "loss": 2.8273,
      "step": 2225
    },
    {
      "epoch": 0.06575261683620817,
      "grad_norm": 17.5,
      "learning_rate": 1.3148584905660377e-06,
      "loss": 2.8624,
      "step": 2230
    },
    {
      "epoch": 0.0659000442282176,
      "grad_norm": 21.0,
      "learning_rate": 1.3178066037735848e-06,
      "loss": 2.8465,
      "step": 2235
    },
    {
      "epoch": 0.06604747162022703,
      "grad_norm": 25.375,
      "learning_rate": 1.320754716981132e-06,
      "loss": 2.8271,
      "step": 2240
    },
    {
      "epoch": 0.06619489901223648,
      "grad_norm": 16.375,
      "learning_rate": 1.3237028301886792e-06,
      "loss": 2.8661,
      "step": 2245
    },
    {
      "epoch": 0.06634232640424591,
      "grad_norm": 15.75,
      "learning_rate": 1.3266509433962263e-06,
      "loss": 2.8245,
      "step": 2250
    },
    {
      "epoch": 0.06648975379625534,
      "grad_norm": 16.25,
      "learning_rate": 1.3295990566037737e-06,
      "loss": 2.8016,
      "step": 2255
    },
    {
      "epoch": 0.06663718118826478,
      "grad_norm": 85.5,
      "learning_rate": 1.3325471698113206e-06,
      "loss": 3.0405,
      "step": 2260
    },
    {
      "epoch": 0.06678460858027421,
      "grad_norm": 16.5,
      "learning_rate": 1.335495283018868e-06,
      "loss": 2.941,
      "step": 2265
    },
    {
      "epoch": 0.06693203597228366,
      "grad_norm": 18.375,
      "learning_rate": 1.338443396226415e-06,
      "loss": 2.9167,
      "step": 2270
    },
    {
      "epoch": 0.06707946336429309,
      "grad_norm": 28.25,
      "learning_rate": 1.3413915094339623e-06,
      "loss": 2.9053,
      "step": 2275
    },
    {
      "epoch": 0.06722689075630252,
      "grad_norm": 20.125,
      "learning_rate": 1.3443396226415094e-06,
      "loss": 2.9119,
      "step": 2280
    },
    {
      "epoch": 0.06737431814831196,
      "grad_norm": 26.75,
      "learning_rate": 1.3472877358490566e-06,
      "loss": 2.8339,
      "step": 2285
    },
    {
      "epoch": 0.06752174554032139,
      "grad_norm": 16.75,
      "learning_rate": 1.3502358490566037e-06,
      "loss": 2.8336,
      "step": 2290
    },
    {
      "epoch": 0.06766917293233082,
      "grad_norm": 14.875,
      "learning_rate": 1.353183962264151e-06,
      "loss": 2.7363,
      "step": 2295
    },
    {
      "epoch": 0.06781660032434027,
      "grad_norm": 15.625,
      "learning_rate": 1.356132075471698e-06,
      "loss": 2.8677,
      "step": 2300
    },
    {
      "epoch": 0.0679640277163497,
      "grad_norm": 18.125,
      "learning_rate": 1.3590801886792452e-06,
      "loss": 2.8287,
      "step": 2305
    },
    {
      "epoch": 0.06811145510835913,
      "grad_norm": 17.25,
      "learning_rate": 1.3620283018867924e-06,
      "loss": 2.8503,
      "step": 2310
    },
    {
      "epoch": 0.06825888250036857,
      "grad_norm": 18.5,
      "learning_rate": 1.3649764150943395e-06,
      "loss": 2.8452,
      "step": 2315
    },
    {
      "epoch": 0.068406309892378,
      "grad_norm": 20.125,
      "learning_rate": 1.3679245283018867e-06,
      "loss": 2.8188,
      "step": 2320
    },
    {
      "epoch": 0.06855373728438743,
      "grad_norm": 21.875,
      "learning_rate": 1.3708726415094338e-06,
      "loss": 2.7956,
      "step": 2325
    },
    {
      "epoch": 0.06870116467639688,
      "grad_norm": 24.875,
      "learning_rate": 1.3738207547169812e-06,
      "loss": 2.8185,
      "step": 2330
    },
    {
      "epoch": 0.06884859206840631,
      "grad_norm": 18.25,
      "learning_rate": 1.3767688679245281e-06,
      "loss": 2.7711,
      "step": 2335
    },
    {
      "epoch": 0.06899601946041574,
      "grad_norm": 20.875,
      "learning_rate": 1.3797169811320755e-06,
      "loss": 2.9115,
      "step": 2340
    },
    {
      "epoch": 0.06914344685242518,
      "grad_norm": 18.125,
      "learning_rate": 1.3826650943396224e-06,
      "loss": 2.8137,
      "step": 2345
    },
    {
      "epoch": 0.06929087424443461,
      "grad_norm": 25.5,
      "learning_rate": 1.3856132075471698e-06,
      "loss": 2.9505,
      "step": 2350
    },
    {
      "epoch": 0.06943830163644406,
      "grad_norm": 24.375,
      "learning_rate": 1.388561320754717e-06,
      "loss": 2.8544,
      "step": 2355
    },
    {
      "epoch": 0.06958572902845349,
      "grad_norm": 24.875,
      "learning_rate": 1.3915094339622641e-06,
      "loss": 2.9782,
      "step": 2360
    },
    {
      "epoch": 0.06973315642046292,
      "grad_norm": 19.125,
      "learning_rate": 1.3944575471698113e-06,
      "loss": 3.0521,
      "step": 2365
    },
    {
      "epoch": 0.06988058381247236,
      "grad_norm": 16.625,
      "learning_rate": 1.3974056603773584e-06,
      "loss": 2.7751,
      "step": 2370
    },
    {
      "epoch": 0.0700280112044818,
      "grad_norm": 21.375,
      "learning_rate": 1.4003537735849056e-06,
      "loss": 2.7304,
      "step": 2375
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 32.25,
      "learning_rate": 1.403301886792453e-06,
      "loss": 2.8306,
      "step": 2380
    },
    {
      "epoch": 0.07032286598850067,
      "grad_norm": 20.375,
      "learning_rate": 1.4062499999999999e-06,
      "loss": 2.9704,
      "step": 2385
    },
    {
      "epoch": 0.0704702933805101,
      "grad_norm": 27.875,
      "learning_rate": 1.409198113207547e-06,
      "loss": 2.8213,
      "step": 2390
    },
    {
      "epoch": 0.07061772077251953,
      "grad_norm": 27.875,
      "learning_rate": 1.4121462264150942e-06,
      "loss": 2.8982,
      "step": 2395
    },
    {
      "epoch": 0.07076514816452897,
      "grad_norm": 29.375,
      "learning_rate": 1.4150943396226413e-06,
      "loss": 2.8757,
      "step": 2400
    },
    {
      "epoch": 0.0709125755565384,
      "grad_norm": 18.125,
      "learning_rate": 1.4180424528301887e-06,
      "loss": 2.7914,
      "step": 2405
    },
    {
      "epoch": 0.07106000294854783,
      "grad_norm": 63.0,
      "learning_rate": 1.4209905660377356e-06,
      "loss": 2.7504,
      "step": 2410
    },
    {
      "epoch": 0.07120743034055728,
      "grad_norm": 16.0,
      "learning_rate": 1.423938679245283e-06,
      "loss": 2.8089,
      "step": 2415
    },
    {
      "epoch": 0.07135485773256671,
      "grad_norm": 19.125,
      "learning_rate": 1.42688679245283e-06,
      "loss": 2.7972,
      "step": 2420
    },
    {
      "epoch": 0.07150228512457614,
      "grad_norm": 15.0,
      "learning_rate": 1.4298349056603773e-06,
      "loss": 2.7051,
      "step": 2425
    },
    {
      "epoch": 0.07164971251658558,
      "grad_norm": 15.1875,
      "learning_rate": 1.4327830188679245e-06,
      "loss": 2.8048,
      "step": 2430
    },
    {
      "epoch": 0.07179713990859501,
      "grad_norm": 18.375,
      "learning_rate": 1.4357311320754716e-06,
      "loss": 2.7196,
      "step": 2435
    },
    {
      "epoch": 0.07194456730060446,
      "grad_norm": 16.375,
      "learning_rate": 1.4386792452830188e-06,
      "loss": 2.695,
      "step": 2440
    },
    {
      "epoch": 0.07209199469261389,
      "grad_norm": 18.25,
      "learning_rate": 1.4416273584905661e-06,
      "loss": 2.9735,
      "step": 2445
    },
    {
      "epoch": 0.07223942208462332,
      "grad_norm": 18.25,
      "learning_rate": 1.444575471698113e-06,
      "loss": 2.8195,
      "step": 2450
    },
    {
      "epoch": 0.07238684947663276,
      "grad_norm": 17.875,
      "learning_rate": 1.4475235849056605e-06,
      "loss": 2.7522,
      "step": 2455
    },
    {
      "epoch": 0.0725342768686422,
      "grad_norm": 17.125,
      "learning_rate": 1.4504716981132074e-06,
      "loss": 2.8065,
      "step": 2460
    },
    {
      "epoch": 0.07268170426065163,
      "grad_norm": 15.625,
      "learning_rate": 1.4534198113207548e-06,
      "loss": 2.8917,
      "step": 2465
    },
    {
      "epoch": 0.07282913165266107,
      "grad_norm": 19.125,
      "learning_rate": 1.456367924528302e-06,
      "loss": 2.8464,
      "step": 2470
    },
    {
      "epoch": 0.0729765590446705,
      "grad_norm": 16.375,
      "learning_rate": 1.4593160377358489e-06,
      "loss": 2.9117,
      "step": 2475
    },
    {
      "epoch": 0.07312398643667993,
      "grad_norm": 16.375,
      "learning_rate": 1.4622641509433962e-06,
      "loss": 2.6589,
      "step": 2480
    },
    {
      "epoch": 0.07327141382868937,
      "grad_norm": 21.75,
      "learning_rate": 1.4652122641509432e-06,
      "loss": 2.915,
      "step": 2485
    },
    {
      "epoch": 0.0734188412206988,
      "grad_norm": 16.875,
      "learning_rate": 1.4681603773584905e-06,
      "loss": 2.7783,
      "step": 2490
    },
    {
      "epoch": 0.07356626861270824,
      "grad_norm": 16.875,
      "learning_rate": 1.4711084905660377e-06,
      "loss": 2.7156,
      "step": 2495
    },
    {
      "epoch": 0.07371369600471768,
      "grad_norm": 17.0,
      "learning_rate": 1.4740566037735848e-06,
      "loss": 2.6785,
      "step": 2500
    },
    {
      "epoch": 0.07371369600471768,
      "eval_loss": 3.0259971618652344,
      "eval_runtime": 4.7168,
      "eval_samples_per_second": 83.956,
      "eval_steps_per_second": 2.756,
      "step": 2500
    },
    {
      "epoch": 0.07386112339672711,
      "grad_norm": 11.6875,
      "learning_rate": 1.477004716981132e-06,
      "loss": 2.6697,
      "step": 2505
    },
    {
      "epoch": 0.07400855078873654,
      "grad_norm": 19.125,
      "learning_rate": 1.4799528301886791e-06,
      "loss": 2.9352,
      "step": 2510
    },
    {
      "epoch": 0.07415597818074599,
      "grad_norm": 24.5,
      "learning_rate": 1.4829009433962263e-06,
      "loss": 2.8369,
      "step": 2515
    },
    {
      "epoch": 0.07430340557275542,
      "grad_norm": 33.25,
      "learning_rate": 1.4858490566037737e-06,
      "loss": 2.7892,
      "step": 2520
    },
    {
      "epoch": 0.07445083296476486,
      "grad_norm": 23.0,
      "learning_rate": 1.4887971698113206e-06,
      "loss": 2.6221,
      "step": 2525
    },
    {
      "epoch": 0.07459826035677429,
      "grad_norm": 17.5,
      "learning_rate": 1.491745283018868e-06,
      "loss": 2.7536,
      "step": 2530
    },
    {
      "epoch": 0.07474568774878372,
      "grad_norm": 20.0,
      "learning_rate": 1.494693396226415e-06,
      "loss": 2.8727,
      "step": 2535
    },
    {
      "epoch": 0.07489311514079317,
      "grad_norm": 13.6875,
      "learning_rate": 1.4976415094339623e-06,
      "loss": 2.6381,
      "step": 2540
    },
    {
      "epoch": 0.0750405425328026,
      "grad_norm": 24.625,
      "learning_rate": 1.5005896226415094e-06,
      "loss": 2.8,
      "step": 2545
    },
    {
      "epoch": 0.07518796992481203,
      "grad_norm": 21.875,
      "learning_rate": 1.5035377358490566e-06,
      "loss": 2.6405,
      "step": 2550
    },
    {
      "epoch": 0.07533539731682147,
      "grad_norm": 13.75,
      "learning_rate": 1.5064858490566037e-06,
      "loss": 2.7266,
      "step": 2555
    },
    {
      "epoch": 0.0754828247088309,
      "grad_norm": 15.25,
      "learning_rate": 1.5094339622641509e-06,
      "loss": 2.7492,
      "step": 2560
    },
    {
      "epoch": 0.07563025210084033,
      "grad_norm": 19.5,
      "learning_rate": 1.512382075471698e-06,
      "loss": 2.7432,
      "step": 2565
    },
    {
      "epoch": 0.07577767949284978,
      "grad_norm": 21.0,
      "learning_rate": 1.5153301886792452e-06,
      "loss": 2.8012,
      "step": 2570
    },
    {
      "epoch": 0.0759251068848592,
      "grad_norm": 18.5,
      "learning_rate": 1.5182783018867923e-06,
      "loss": 2.8801,
      "step": 2575
    },
    {
      "epoch": 0.07607253427686864,
      "grad_norm": 15.9375,
      "learning_rate": 1.5212264150943395e-06,
      "loss": 2.6939,
      "step": 2580
    },
    {
      "epoch": 0.07621996166887808,
      "grad_norm": 16.375,
      "learning_rate": 1.5241745283018867e-06,
      "loss": 2.7358,
      "step": 2585
    },
    {
      "epoch": 0.07636738906088751,
      "grad_norm": 15.375,
      "learning_rate": 1.5271226415094338e-06,
      "loss": 2.7957,
      "step": 2590
    },
    {
      "epoch": 0.07651481645289694,
      "grad_norm": 18.375,
      "learning_rate": 1.5300707547169812e-06,
      "loss": 2.8,
      "step": 2595
    },
    {
      "epoch": 0.07666224384490639,
      "grad_norm": 20.625,
      "learning_rate": 1.5330188679245281e-06,
      "loss": 2.7041,
      "step": 2600
    },
    {
      "epoch": 0.07680967123691582,
      "grad_norm": 17.75,
      "learning_rate": 1.5359669811320755e-06,
      "loss": 2.8524,
      "step": 2605
    },
    {
      "epoch": 0.07695709862892526,
      "grad_norm": 20.625,
      "learning_rate": 1.5389150943396224e-06,
      "loss": 2.7402,
      "step": 2610
    },
    {
      "epoch": 0.07710452602093469,
      "grad_norm": 21.25,
      "learning_rate": 1.5418632075471698e-06,
      "loss": 2.8244,
      "step": 2615
    },
    {
      "epoch": 0.07725195341294412,
      "grad_norm": 24.125,
      "learning_rate": 1.544811320754717e-06,
      "loss": 2.7226,
      "step": 2620
    },
    {
      "epoch": 0.07739938080495357,
      "grad_norm": 19.625,
      "learning_rate": 1.547759433962264e-06,
      "loss": 2.637,
      "step": 2625
    },
    {
      "epoch": 0.077546808196963,
      "grad_norm": 22.375,
      "learning_rate": 1.5507075471698112e-06,
      "loss": 2.6678,
      "step": 2630
    },
    {
      "epoch": 0.07769423558897243,
      "grad_norm": 19.875,
      "learning_rate": 1.5536556603773586e-06,
      "loss": 2.5859,
      "step": 2635
    },
    {
      "epoch": 0.07784166298098187,
      "grad_norm": 19.75,
      "learning_rate": 1.5566037735849056e-06,
      "loss": 2.5884,
      "step": 2640
    },
    {
      "epoch": 0.0779890903729913,
      "grad_norm": 15.25,
      "learning_rate": 1.559551886792453e-06,
      "loss": 2.8667,
      "step": 2645
    },
    {
      "epoch": 0.07813651776500073,
      "grad_norm": 17.0,
      "learning_rate": 1.5624999999999999e-06,
      "loss": 2.6761,
      "step": 2650
    },
    {
      "epoch": 0.07828394515701018,
      "grad_norm": 21.75,
      "learning_rate": 1.565448113207547e-06,
      "loss": 2.6761,
      "step": 2655
    },
    {
      "epoch": 0.0784313725490196,
      "grad_norm": 32.25,
      "learning_rate": 1.5683962264150944e-06,
      "loss": 2.7592,
      "step": 2660
    },
    {
      "epoch": 0.07857879994102904,
      "grad_norm": 43.0,
      "learning_rate": 1.5713443396226413e-06,
      "loss": 2.7657,
      "step": 2665
    },
    {
      "epoch": 0.07872622733303848,
      "grad_norm": 22.875,
      "learning_rate": 1.5742924528301887e-06,
      "loss": 2.8494,
      "step": 2670
    },
    {
      "epoch": 0.07887365472504791,
      "grad_norm": 18.25,
      "learning_rate": 1.5772405660377356e-06,
      "loss": 2.7864,
      "step": 2675
    },
    {
      "epoch": 0.07902108211705734,
      "grad_norm": 24.5,
      "learning_rate": 1.580188679245283e-06,
      "loss": 2.598,
      "step": 2680
    },
    {
      "epoch": 0.07916850950906679,
      "grad_norm": 27.25,
      "learning_rate": 1.5831367924528301e-06,
      "loss": 2.751,
      "step": 2685
    },
    {
      "epoch": 0.07931593690107622,
      "grad_norm": 17.625,
      "learning_rate": 1.5860849056603773e-06,
      "loss": 2.7719,
      "step": 2690
    },
    {
      "epoch": 0.07946336429308566,
      "grad_norm": 20.75,
      "learning_rate": 1.5890330188679245e-06,
      "loss": 2.6851,
      "step": 2695
    },
    {
      "epoch": 0.07961079168509509,
      "grad_norm": 16.25,
      "learning_rate": 1.5919811320754716e-06,
      "loss": 2.7268,
      "step": 2700
    },
    {
      "epoch": 0.07975821907710452,
      "grad_norm": 20.75,
      "learning_rate": 1.5949292452830188e-06,
      "loss": 2.7603,
      "step": 2705
    },
    {
      "epoch": 0.07990564646911397,
      "grad_norm": 17.875,
      "learning_rate": 1.5978773584905661e-06,
      "loss": 2.8362,
      "step": 2710
    },
    {
      "epoch": 0.0800530738611234,
      "grad_norm": 20.375,
      "learning_rate": 1.600825471698113e-06,
      "loss": 2.7863,
      "step": 2715
    },
    {
      "epoch": 0.08020050125313283,
      "grad_norm": 16.5,
      "learning_rate": 1.6037735849056604e-06,
      "loss": 2.6775,
      "step": 2720
    },
    {
      "epoch": 0.08034792864514227,
      "grad_norm": 20.625,
      "learning_rate": 1.6067216981132074e-06,
      "loss": 2.7235,
      "step": 2725
    },
    {
      "epoch": 0.0804953560371517,
      "grad_norm": 15.5625,
      "learning_rate": 1.6096698113207547e-06,
      "loss": 2.655,
      "step": 2730
    },
    {
      "epoch": 0.08064278342916113,
      "grad_norm": 23.625,
      "learning_rate": 1.612617924528302e-06,
      "loss": 2.8043,
      "step": 2735
    },
    {
      "epoch": 0.08079021082117058,
      "grad_norm": 23.375,
      "learning_rate": 1.6155660377358488e-06,
      "loss": 2.6753,
      "step": 2740
    },
    {
      "epoch": 0.08093763821318001,
      "grad_norm": 16.75,
      "learning_rate": 1.6185141509433962e-06,
      "loss": 2.7216,
      "step": 2745
    },
    {
      "epoch": 0.08108506560518944,
      "grad_norm": 29.75,
      "learning_rate": 1.6214622641509431e-06,
      "loss": 2.893,
      "step": 2750
    },
    {
      "epoch": 0.08123249299719888,
      "grad_norm": 18.625,
      "learning_rate": 1.6244103773584905e-06,
      "loss": 2.7302,
      "step": 2755
    },
    {
      "epoch": 0.08137992038920831,
      "grad_norm": 17.0,
      "learning_rate": 1.6273584905660377e-06,
      "loss": 2.7501,
      "step": 2760
    },
    {
      "epoch": 0.08152734778121776,
      "grad_norm": 19.5,
      "learning_rate": 1.6303066037735848e-06,
      "loss": 2.7171,
      "step": 2765
    },
    {
      "epoch": 0.08167477517322719,
      "grad_norm": 14.9375,
      "learning_rate": 1.633254716981132e-06,
      "loss": 2.7143,
      "step": 2770
    },
    {
      "epoch": 0.08182220256523662,
      "grad_norm": 21.625,
      "learning_rate": 1.6362028301886791e-06,
      "loss": 2.632,
      "step": 2775
    },
    {
      "epoch": 0.08196962995724606,
      "grad_norm": 17.25,
      "learning_rate": 1.6391509433962263e-06,
      "loss": 2.7041,
      "step": 2780
    },
    {
      "epoch": 0.08211705734925549,
      "grad_norm": 27.25,
      "learning_rate": 1.6420990566037736e-06,
      "loss": 2.8002,
      "step": 2785
    },
    {
      "epoch": 0.08226448474126492,
      "grad_norm": 31.25,
      "learning_rate": 1.6450471698113206e-06,
      "loss": 2.5959,
      "step": 2790
    },
    {
      "epoch": 0.08241191213327437,
      "grad_norm": 19.875,
      "learning_rate": 1.647995283018868e-06,
      "loss": 2.7606,
      "step": 2795
    },
    {
      "epoch": 0.0825593395252838,
      "grad_norm": 18.5,
      "learning_rate": 1.6509433962264149e-06,
      "loss": 3.0455,
      "step": 2800
    },
    {
      "epoch": 0.08270676691729323,
      "grad_norm": 22.875,
      "learning_rate": 1.6538915094339623e-06,
      "loss": 2.7094,
      "step": 2805
    },
    {
      "epoch": 0.08285419430930267,
      "grad_norm": 17.375,
      "learning_rate": 1.6568396226415094e-06,
      "loss": 2.4643,
      "step": 2810
    },
    {
      "epoch": 0.0830016217013121,
      "grad_norm": 22.0,
      "learning_rate": 1.6597877358490566e-06,
      "loss": 2.6835,
      "step": 2815
    },
    {
      "epoch": 0.08314904909332153,
      "grad_norm": 18.125,
      "learning_rate": 1.6627358490566037e-06,
      "loss": 2.6437,
      "step": 2820
    },
    {
      "epoch": 0.08329647648533098,
      "grad_norm": 22.625,
      "learning_rate": 1.6656839622641509e-06,
      "loss": 2.6712,
      "step": 2825
    },
    {
      "epoch": 0.08344390387734041,
      "grad_norm": 17.0,
      "learning_rate": 1.668632075471698e-06,
      "loss": 2.7535,
      "step": 2830
    },
    {
      "epoch": 0.08359133126934984,
      "grad_norm": 24.375,
      "learning_rate": 1.6715801886792452e-06,
      "loss": 2.6405,
      "step": 2835
    },
    {
      "epoch": 0.08373875866135928,
      "grad_norm": 16.875,
      "learning_rate": 1.6745283018867923e-06,
      "loss": 2.5578,
      "step": 2840
    },
    {
      "epoch": 0.08388618605336871,
      "grad_norm": 17.375,
      "learning_rate": 1.6774764150943395e-06,
      "loss": 2.662,
      "step": 2845
    },
    {
      "epoch": 0.08403361344537816,
      "grad_norm": 18.0,
      "learning_rate": 1.6804245283018866e-06,
      "loss": 2.5877,
      "step": 2850
    },
    {
      "epoch": 0.08418104083738759,
      "grad_norm": 22.75,
      "learning_rate": 1.6833726415094338e-06,
      "loss": 2.7346,
      "step": 2855
    },
    {
      "epoch": 0.08432846822939702,
      "grad_norm": 25.0,
      "learning_rate": 1.6863207547169812e-06,
      "loss": 2.7402,
      "step": 2860
    },
    {
      "epoch": 0.08447589562140646,
      "grad_norm": 19.0,
      "learning_rate": 1.689268867924528e-06,
      "loss": 2.6026,
      "step": 2865
    },
    {
      "epoch": 0.0846233230134159,
      "grad_norm": 23.625,
      "learning_rate": 1.6922169811320755e-06,
      "loss": 2.6259,
      "step": 2870
    },
    {
      "epoch": 0.08477075040542532,
      "grad_norm": 16.625,
      "learning_rate": 1.6951650943396224e-06,
      "loss": 2.67,
      "step": 2875
    },
    {
      "epoch": 0.08491817779743477,
      "grad_norm": 22.875,
      "learning_rate": 1.6981132075471698e-06,
      "loss": 2.4129,
      "step": 2880
    },
    {
      "epoch": 0.0850656051894442,
      "grad_norm": 129.0,
      "learning_rate": 1.701061320754717e-06,
      "loss": 2.8064,
      "step": 2885
    },
    {
      "epoch": 0.08521303258145363,
      "grad_norm": 13.625,
      "learning_rate": 1.704009433962264e-06,
      "loss": 2.6248,
      "step": 2890
    },
    {
      "epoch": 0.08536045997346307,
      "grad_norm": 16.125,
      "learning_rate": 1.7069575471698112e-06,
      "loss": 2.6568,
      "step": 2895
    },
    {
      "epoch": 0.0855078873654725,
      "grad_norm": 15.75,
      "learning_rate": 1.7099056603773586e-06,
      "loss": 2.7321,
      "step": 2900
    },
    {
      "epoch": 0.08565531475748193,
      "grad_norm": 34.5,
      "learning_rate": 1.7128537735849055e-06,
      "loss": 2.6696,
      "step": 2905
    },
    {
      "epoch": 0.08580274214949138,
      "grad_norm": 18.375,
      "learning_rate": 1.715801886792453e-06,
      "loss": 2.7116,
      "step": 2910
    },
    {
      "epoch": 0.08595016954150081,
      "grad_norm": 16.125,
      "learning_rate": 1.7187499999999998e-06,
      "loss": 2.66,
      "step": 2915
    },
    {
      "epoch": 0.08609759693351024,
      "grad_norm": 18.25,
      "learning_rate": 1.721698113207547e-06,
      "loss": 2.7062,
      "step": 2920
    },
    {
      "epoch": 0.08624502432551968,
      "grad_norm": 19.75,
      "learning_rate": 1.7246462264150944e-06,
      "loss": 2.667,
      "step": 2925
    },
    {
      "epoch": 0.08639245171752911,
      "grad_norm": 13.125,
      "learning_rate": 1.7275943396226413e-06,
      "loss": 2.6131,
      "step": 2930
    },
    {
      "epoch": 0.08653987910953856,
      "grad_norm": 20.25,
      "learning_rate": 1.7305424528301887e-06,
      "loss": 2.6455,
      "step": 2935
    },
    {
      "epoch": 0.08668730650154799,
      "grad_norm": 18.625,
      "learning_rate": 1.7334905660377356e-06,
      "loss": 2.6537,
      "step": 2940
    },
    {
      "epoch": 0.08683473389355742,
      "grad_norm": 16.75,
      "learning_rate": 1.736438679245283e-06,
      "loss": 2.6189,
      "step": 2945
    },
    {
      "epoch": 0.08698216128556686,
      "grad_norm": 19.5,
      "learning_rate": 1.7393867924528301e-06,
      "loss": 2.6757,
      "step": 2950
    },
    {
      "epoch": 0.0871295886775763,
      "grad_norm": 20.5,
      "learning_rate": 1.7423349056603773e-06,
      "loss": 2.613,
      "step": 2955
    },
    {
      "epoch": 0.08727701606958572,
      "grad_norm": 23.125,
      "learning_rate": 1.7452830188679244e-06,
      "loss": 2.6377,
      "step": 2960
    },
    {
      "epoch": 0.08742444346159517,
      "grad_norm": 23.25,
      "learning_rate": 1.7482311320754716e-06,
      "loss": 2.7463,
      "step": 2965
    },
    {
      "epoch": 0.0875718708536046,
      "grad_norm": 69.5,
      "learning_rate": 1.7511792452830188e-06,
      "loss": 2.7704,
      "step": 2970
    },
    {
      "epoch": 0.08771929824561403,
      "grad_norm": 15.8125,
      "learning_rate": 1.7541273584905661e-06,
      "loss": 2.593,
      "step": 2975
    },
    {
      "epoch": 0.08786672563762347,
      "grad_norm": 53.25,
      "learning_rate": 1.757075471698113e-06,
      "loss": 2.5997,
      "step": 2980
    },
    {
      "epoch": 0.0880141530296329,
      "grad_norm": 33.75,
      "learning_rate": 1.7600235849056604e-06,
      "loss": 2.7404,
      "step": 2985
    },
    {
      "epoch": 0.08816158042164234,
      "grad_norm": 19.5,
      "learning_rate": 1.7629716981132074e-06,
      "loss": 2.5596,
      "step": 2990
    },
    {
      "epoch": 0.08830900781365178,
      "grad_norm": 24.625,
      "learning_rate": 1.7659198113207547e-06,
      "loss": 2.6808,
      "step": 2995
    },
    {
      "epoch": 0.08845643520566121,
      "grad_norm": 20.75,
      "learning_rate": 1.7688679245283019e-06,
      "loss": 2.534,
      "step": 3000
    },
    {
      "epoch": 0.08845643520566121,
      "eval_loss": 2.8255984783172607,
      "eval_runtime": 4.7091,
      "eval_samples_per_second": 84.093,
      "eval_steps_per_second": 2.761,
      "step": 3000
    },
    {
      "epoch": 0.08860386259767064,
      "grad_norm": 15.75,
      "learning_rate": 1.7718160377358488e-06,
      "loss": 2.6962,
      "step": 3005
    },
    {
      "epoch": 0.08875128998968008,
      "grad_norm": 19.125,
      "learning_rate": 1.7747641509433962e-06,
      "loss": 2.5769,
      "step": 3010
    },
    {
      "epoch": 0.08889871738168952,
      "grad_norm": 23.0,
      "learning_rate": 1.7777122641509431e-06,
      "loss": 2.5602,
      "step": 3015
    },
    {
      "epoch": 0.08904614477369896,
      "grad_norm": 16.0,
      "learning_rate": 1.7806603773584905e-06,
      "loss": 2.5782,
      "step": 3020
    },
    {
      "epoch": 0.08919357216570839,
      "grad_norm": 19.375,
      "learning_rate": 1.7836084905660377e-06,
      "loss": 2.6285,
      "step": 3025
    },
    {
      "epoch": 0.08934099955771782,
      "grad_norm": 21.25,
      "learning_rate": 1.7865566037735848e-06,
      "loss": 2.6416,
      "step": 3030
    },
    {
      "epoch": 0.08948842694972726,
      "grad_norm": 19.5,
      "learning_rate": 1.789504716981132e-06,
      "loss": 2.5563,
      "step": 3035
    },
    {
      "epoch": 0.0896358543417367,
      "grad_norm": 19.5,
      "learning_rate": 1.7924528301886791e-06,
      "loss": 2.6188,
      "step": 3040
    },
    {
      "epoch": 0.08978328173374613,
      "grad_norm": 15.75,
      "learning_rate": 1.7954009433962263e-06,
      "loss": 2.6018,
      "step": 3045
    },
    {
      "epoch": 0.08993070912575557,
      "grad_norm": 19.625,
      "learning_rate": 1.7983490566037736e-06,
      "loss": 2.6601,
      "step": 3050
    },
    {
      "epoch": 0.090078136517765,
      "grad_norm": 22.375,
      "learning_rate": 1.8012971698113206e-06,
      "loss": 2.6206,
      "step": 3055
    },
    {
      "epoch": 0.09022556390977443,
      "grad_norm": 25.625,
      "learning_rate": 1.804245283018868e-06,
      "loss": 2.7032,
      "step": 3060
    },
    {
      "epoch": 0.09037299130178388,
      "grad_norm": 16.75,
      "learning_rate": 1.8071933962264149e-06,
      "loss": 2.6075,
      "step": 3065
    },
    {
      "epoch": 0.0905204186937933,
      "grad_norm": 18.875,
      "learning_rate": 1.8101415094339622e-06,
      "loss": 2.7258,
      "step": 3070
    },
    {
      "epoch": 0.09066784608580274,
      "grad_norm": 23.375,
      "learning_rate": 1.8130896226415094e-06,
      "loss": 2.7048,
      "step": 3075
    },
    {
      "epoch": 0.09081527347781218,
      "grad_norm": 19.75,
      "learning_rate": 1.8160377358490566e-06,
      "loss": 2.4501,
      "step": 3080
    },
    {
      "epoch": 0.09096270086982161,
      "grad_norm": 19.125,
      "learning_rate": 1.8189858490566037e-06,
      "loss": 2.5144,
      "step": 3085
    },
    {
      "epoch": 0.09111012826183104,
      "grad_norm": 17.375,
      "learning_rate": 1.821933962264151e-06,
      "loss": 2.7287,
      "step": 3090
    },
    {
      "epoch": 0.09125755565384049,
      "grad_norm": 15.125,
      "learning_rate": 1.824882075471698e-06,
      "loss": 2.6906,
      "step": 3095
    },
    {
      "epoch": 0.09140498304584992,
      "grad_norm": 16.875,
      "learning_rate": 1.8278301886792452e-06,
      "loss": 2.6328,
      "step": 3100
    },
    {
      "epoch": 0.09155241043785936,
      "grad_norm": 28.375,
      "learning_rate": 1.8307783018867923e-06,
      "loss": 2.7216,
      "step": 3105
    },
    {
      "epoch": 0.09169983782986879,
      "grad_norm": 20.375,
      "learning_rate": 1.8337264150943395e-06,
      "loss": 2.6709,
      "step": 3110
    },
    {
      "epoch": 0.09184726522187822,
      "grad_norm": 17.625,
      "learning_rate": 1.8366745283018868e-06,
      "loss": 2.6247,
      "step": 3115
    },
    {
      "epoch": 0.09199469261388767,
      "grad_norm": 14.625,
      "learning_rate": 1.8396226415094338e-06,
      "loss": 2.704,
      "step": 3120
    },
    {
      "epoch": 0.0921421200058971,
      "grad_norm": 20.5,
      "learning_rate": 1.8425707547169811e-06,
      "loss": 2.6575,
      "step": 3125
    },
    {
      "epoch": 0.09228954739790653,
      "grad_norm": 16.75,
      "learning_rate": 1.845518867924528e-06,
      "loss": 2.6053,
      "step": 3130
    },
    {
      "epoch": 0.09243697478991597,
      "grad_norm": 18.875,
      "learning_rate": 1.8484669811320755e-06,
      "loss": 2.6402,
      "step": 3135
    },
    {
      "epoch": 0.0925844021819254,
      "grad_norm": 24.375,
      "learning_rate": 1.8514150943396226e-06,
      "loss": 2.6393,
      "step": 3140
    },
    {
      "epoch": 0.09273182957393483,
      "grad_norm": 23.125,
      "learning_rate": 1.8543632075471698e-06,
      "loss": 2.7078,
      "step": 3145
    },
    {
      "epoch": 0.09287925696594428,
      "grad_norm": 18.5,
      "learning_rate": 1.857311320754717e-06,
      "loss": 2.6088,
      "step": 3150
    },
    {
      "epoch": 0.0930266843579537,
      "grad_norm": 17.0,
      "learning_rate": 1.860259433962264e-06,
      "loss": 2.6411,
      "step": 3155
    },
    {
      "epoch": 0.09317411174996314,
      "grad_norm": 21.25,
      "learning_rate": 1.8632075471698112e-06,
      "loss": 2.5609,
      "step": 3160
    },
    {
      "epoch": 0.09332153914197258,
      "grad_norm": 19.0,
      "learning_rate": 1.8661556603773586e-06,
      "loss": 2.7937,
      "step": 3165
    },
    {
      "epoch": 0.09346896653398201,
      "grad_norm": 17.75,
      "learning_rate": 1.8691037735849055e-06,
      "loss": 2.5774,
      "step": 3170
    },
    {
      "epoch": 0.09361639392599144,
      "grad_norm": 29.125,
      "learning_rate": 1.8720518867924529e-06,
      "loss": 2.6118,
      "step": 3175
    },
    {
      "epoch": 0.09376382131800089,
      "grad_norm": 34.0,
      "learning_rate": 1.8749999999999998e-06,
      "loss": 2.7062,
      "step": 3180
    },
    {
      "epoch": 0.09391124871001032,
      "grad_norm": 21.375,
      "learning_rate": 1.877948113207547e-06,
      "loss": 2.4245,
      "step": 3185
    },
    {
      "epoch": 0.09405867610201976,
      "grad_norm": 21.0,
      "learning_rate": 1.8808962264150944e-06,
      "loss": 2.575,
      "step": 3190
    },
    {
      "epoch": 0.09420610349402919,
      "grad_norm": 19.75,
      "learning_rate": 1.8838443396226413e-06,
      "loss": 2.6233,
      "step": 3195
    },
    {
      "epoch": 0.09435353088603862,
      "grad_norm": 17.0,
      "learning_rate": 1.8867924528301887e-06,
      "loss": 2.63,
      "step": 3200
    },
    {
      "epoch": 0.09450095827804807,
      "grad_norm": 15.5,
      "learning_rate": 1.8897405660377356e-06,
      "loss": 2.4527,
      "step": 3205
    },
    {
      "epoch": 0.0946483856700575,
      "grad_norm": 28.125,
      "learning_rate": 1.892688679245283e-06,
      "loss": 2.6324,
      "step": 3210
    },
    {
      "epoch": 0.09479581306206693,
      "grad_norm": 13.5625,
      "learning_rate": 1.8956367924528301e-06,
      "loss": 2.5413,
      "step": 3215
    },
    {
      "epoch": 0.09494324045407637,
      "grad_norm": 20.625,
      "learning_rate": 1.8985849056603773e-06,
      "loss": 2.7314,
      "step": 3220
    },
    {
      "epoch": 0.0950906678460858,
      "grad_norm": 18.375,
      "learning_rate": 1.9015330188679244e-06,
      "loss": 2.5598,
      "step": 3225
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 25.75,
      "learning_rate": 1.9044811320754716e-06,
      "loss": 2.5823,
      "step": 3230
    },
    {
      "epoch": 0.09538552263010468,
      "grad_norm": 14.375,
      "learning_rate": 1.907429245283019e-06,
      "loss": 2.6562,
      "step": 3235
    },
    {
      "epoch": 0.09553295002211411,
      "grad_norm": 24.75,
      "learning_rate": 1.910377358490566e-06,
      "loss": 2.6288,
      "step": 3240
    },
    {
      "epoch": 0.09568037741412354,
      "grad_norm": 17.625,
      "learning_rate": 1.9133254716981133e-06,
      "loss": 2.5194,
      "step": 3245
    },
    {
      "epoch": 0.09582780480613298,
      "grad_norm": 15.4375,
      "learning_rate": 1.9162735849056604e-06,
      "loss": 2.5139,
      "step": 3250
    },
    {
      "epoch": 0.09597523219814241,
      "grad_norm": 20.375,
      "learning_rate": 1.9192216981132076e-06,
      "loss": 2.6683,
      "step": 3255
    },
    {
      "epoch": 0.09612265959015186,
      "grad_norm": 24.375,
      "learning_rate": 1.9221698113207547e-06,
      "loss": 2.4635,
      "step": 3260
    },
    {
      "epoch": 0.09627008698216129,
      "grad_norm": 29.75,
      "learning_rate": 1.925117924528302e-06,
      "loss": 2.6588,
      "step": 3265
    },
    {
      "epoch": 0.09641751437417072,
      "grad_norm": 15.375,
      "learning_rate": 1.928066037735849e-06,
      "loss": 2.4663,
      "step": 3270
    },
    {
      "epoch": 0.09656494176618016,
      "grad_norm": 17.25,
      "learning_rate": 1.931014150943396e-06,
      "loss": 2.5586,
      "step": 3275
    },
    {
      "epoch": 0.09671236915818959,
      "grad_norm": 16.75,
      "learning_rate": 1.9339622641509433e-06,
      "loss": 2.4494,
      "step": 3280
    },
    {
      "epoch": 0.09685979655019902,
      "grad_norm": 17.75,
      "learning_rate": 1.9369103773584905e-06,
      "loss": 2.513,
      "step": 3285
    },
    {
      "epoch": 0.09700722394220847,
      "grad_norm": 18.375,
      "learning_rate": 1.9398584905660376e-06,
      "loss": 2.5481,
      "step": 3290
    },
    {
      "epoch": 0.0971546513342179,
      "grad_norm": 15.1875,
      "learning_rate": 1.942806603773585e-06,
      "loss": 2.5519,
      "step": 3295
    },
    {
      "epoch": 0.09730207872622733,
      "grad_norm": 56.5,
      "learning_rate": 1.945754716981132e-06,
      "loss": 2.4955,
      "step": 3300
    },
    {
      "epoch": 0.09744950611823677,
      "grad_norm": 20.25,
      "learning_rate": 1.948702830188679e-06,
      "loss": 2.5694,
      "step": 3305
    },
    {
      "epoch": 0.0975969335102462,
      "grad_norm": 29.875,
      "learning_rate": 1.9516509433962263e-06,
      "loss": 2.5207,
      "step": 3310
    },
    {
      "epoch": 0.09774436090225563,
      "grad_norm": 17.0,
      "learning_rate": 1.9545990566037734e-06,
      "loss": 2.5684,
      "step": 3315
    },
    {
      "epoch": 0.09789178829426508,
      "grad_norm": 18.375,
      "learning_rate": 1.9575471698113206e-06,
      "loss": 2.5669,
      "step": 3320
    },
    {
      "epoch": 0.09803921568627451,
      "grad_norm": 18.875,
      "learning_rate": 1.9604952830188677e-06,
      "loss": 2.4171,
      "step": 3325
    },
    {
      "epoch": 0.09818664307828394,
      "grad_norm": 77.5,
      "learning_rate": 1.963443396226415e-06,
      "loss": 2.6574,
      "step": 3330
    },
    {
      "epoch": 0.09833407047029338,
      "grad_norm": 84.5,
      "learning_rate": 1.9663915094339624e-06,
      "loss": 2.574,
      "step": 3335
    },
    {
      "epoch": 0.09848149786230281,
      "grad_norm": 28.125,
      "learning_rate": 1.969339622641509e-06,
      "loss": 2.6212,
      "step": 3340
    },
    {
      "epoch": 0.09862892525431226,
      "grad_norm": 30.375,
      "learning_rate": 1.9722877358490568e-06,
      "loss": 2.421,
      "step": 3345
    },
    {
      "epoch": 0.09877635264632169,
      "grad_norm": 22.875,
      "learning_rate": 1.9752358490566035e-06,
      "loss": 2.5433,
      "step": 3350
    },
    {
      "epoch": 0.09892378003833112,
      "grad_norm": 24.0,
      "learning_rate": 1.978183962264151e-06,
      "loss": 2.733,
      "step": 3355
    },
    {
      "epoch": 0.09907120743034056,
      "grad_norm": 16.875,
      "learning_rate": 1.981132075471698e-06,
      "loss": 2.599,
      "step": 3360
    },
    {
      "epoch": 0.09921863482235,
      "grad_norm": 16.625,
      "learning_rate": 1.984080188679245e-06,
      "loss": 2.5051,
      "step": 3365
    },
    {
      "epoch": 0.09936606221435942,
      "grad_norm": 22.875,
      "learning_rate": 1.9870283018867925e-06,
      "loss": 2.4694,
      "step": 3370
    },
    {
      "epoch": 0.09951348960636887,
      "grad_norm": 19.625,
      "learning_rate": 1.9899764150943392e-06,
      "loss": 2.5643,
      "step": 3375
    },
    {
      "epoch": 0.0996609169983783,
      "grad_norm": 19.75,
      "learning_rate": 1.992924528301887e-06,
      "loss": 2.4453,
      "step": 3380
    },
    {
      "epoch": 0.09980834439038773,
      "grad_norm": 19.875,
      "learning_rate": 1.995872641509434e-06,
      "loss": 2.4782,
      "step": 3385
    },
    {
      "epoch": 0.09995577178239717,
      "grad_norm": 14.125,
      "learning_rate": 1.998820754716981e-06,
      "loss": 2.5011,
      "step": 3390
    },
    {
      "epoch": 0.1001031991744066,
      "grad_norm": 16.625,
      "learning_rate": 1.999999952328609e-06,
      "loss": 2.6024,
      "step": 3395
    },
    {
      "epoch": 0.10025062656641603,
      "grad_norm": 17.5,
      "learning_rate": 1.9999996610034596e-06,
      "loss": 2.5227,
      "step": 3400
    },
    {
      "epoch": 0.10039805395842548,
      "grad_norm": 17.0,
      "learning_rate": 1.9999991048373438e-06,
      "loss": 2.4102,
      "step": 3405
    },
    {
      "epoch": 0.10054548135043491,
      "grad_norm": 22.375,
      "learning_rate": 1.9999982838304092e-06,
      "loss": 2.5443,
      "step": 3410
    },
    {
      "epoch": 0.10069290874244434,
      "grad_norm": 19.25,
      "learning_rate": 1.9999971979828727e-06,
      "loss": 2.5334,
      "step": 3415
    },
    {
      "epoch": 0.10084033613445378,
      "grad_norm": 18.5,
      "learning_rate": 1.999995847295022e-06,
      "loss": 2.6654,
      "step": 3420
    },
    {
      "epoch": 0.10098776352646321,
      "grad_norm": 18.0,
      "learning_rate": 1.999994231767215e-06,
      "loss": 2.5571,
      "step": 3425
    },
    {
      "epoch": 0.10113519091847266,
      "grad_norm": 13.5,
      "learning_rate": 1.99999235139988e-06,
      "loss": 2.5254,
      "step": 3430
    },
    {
      "epoch": 0.10128261831048209,
      "grad_norm": 27.375,
      "learning_rate": 1.999990206193514e-06,
      "loss": 2.4695,
      "step": 3435
    },
    {
      "epoch": 0.10143004570249152,
      "grad_norm": 22.0,
      "learning_rate": 1.999987796148686e-06,
      "loss": 2.4044,
      "step": 3440
    },
    {
      "epoch": 0.10157747309450096,
      "grad_norm": 16.75,
      "learning_rate": 1.9999851212660336e-06,
      "loss": 2.4752,
      "step": 3445
    },
    {
      "epoch": 0.1017249004865104,
      "grad_norm": 18.75,
      "learning_rate": 1.9999821815462655e-06,
      "loss": 2.6967,
      "step": 3450
    },
    {
      "epoch": 0.10187232787851982,
      "grad_norm": 25.0,
      "learning_rate": 1.9999789769901606e-06,
      "loss": 2.5363,
      "step": 3455
    },
    {
      "epoch": 0.10201975527052927,
      "grad_norm": 26.375,
      "learning_rate": 1.9999755075985674e-06,
      "loss": 2.5621,
      "step": 3460
    },
    {
      "epoch": 0.1021671826625387,
      "grad_norm": 19.375,
      "learning_rate": 1.9999717733724043e-06,
      "loss": 2.5803,
      "step": 3465
    },
    {
      "epoch": 0.10231461005454813,
      "grad_norm": 22.625,
      "learning_rate": 1.9999677743126607e-06,
      "loss": 2.6209,
      "step": 3470
    },
    {
      "epoch": 0.10246203744655757,
      "grad_norm": 17.625,
      "learning_rate": 1.999963510420396e-06,
      "loss": 2.4312,
      "step": 3475
    },
    {
      "epoch": 0.102609464838567,
      "grad_norm": 41.0,
      "learning_rate": 1.999958981696739e-06,
      "loss": 2.5988,
      "step": 3480
    },
    {
      "epoch": 0.10275689223057644,
      "grad_norm": 21.375,
      "learning_rate": 1.999954188142889e-06,
      "loss": 2.5389,
      "step": 3485
    },
    {
      "epoch": 0.10290431962258588,
      "grad_norm": 16.0,
      "learning_rate": 1.9999491297601154e-06,
      "loss": 2.5496,
      "step": 3490
    },
    {
      "epoch": 0.10305174701459531,
      "grad_norm": 29.0,
      "learning_rate": 1.9999438065497587e-06,
      "loss": 2.5188,
      "step": 3495
    },
    {
      "epoch": 0.10319917440660474,
      "grad_norm": 18.125,
      "learning_rate": 1.999938218513228e-06,
      "loss": 2.4835,
      "step": 3500
    },
    {
      "epoch": 0.10319917440660474,
      "eval_loss": 2.641951322555542,
      "eval_runtime": 4.7225,
      "eval_samples_per_second": 83.855,
      "eval_steps_per_second": 2.753,
      "step": 3500
    },
    {
      "epoch": 0.10334660179861418,
      "grad_norm": 19.625,
      "learning_rate": 1.9999323656520037e-06,
      "loss": 2.4558,
      "step": 3505
    },
    {
      "epoch": 0.10349402919062362,
      "grad_norm": 17.375,
      "learning_rate": 1.999926247967635e-06,
      "loss": 2.4677,
      "step": 3510
    },
    {
      "epoch": 0.10364145658263306,
      "grad_norm": 18.0,
      "learning_rate": 1.999919865461743e-06,
      "loss": 2.4587,
      "step": 3515
    },
    {
      "epoch": 0.10378888397464249,
      "grad_norm": 20.875,
      "learning_rate": 1.999913218136018e-06,
      "loss": 2.5211,
      "step": 3520
    },
    {
      "epoch": 0.10393631136665192,
      "grad_norm": 16.25,
      "learning_rate": 1.99990630599222e-06,
      "loss": 2.552,
      "step": 3525
    },
    {
      "epoch": 0.10408373875866136,
      "grad_norm": 16.875,
      "learning_rate": 1.99989912903218e-06,
      "loss": 2.3828,
      "step": 3530
    },
    {
      "epoch": 0.1042311661506708,
      "grad_norm": 19.625,
      "learning_rate": 1.999891687257799e-06,
      "loss": 2.4554,
      "step": 3535
    },
    {
      "epoch": 0.10437859354268023,
      "grad_norm": 18.5,
      "learning_rate": 1.9998839806710466e-06,
      "loss": 2.5472,
      "step": 3540
    },
    {
      "epoch": 0.10452602093468967,
      "grad_norm": 17.5,
      "learning_rate": 1.9998760092739654e-06,
      "loss": 2.4378,
      "step": 3545
    },
    {
      "epoch": 0.1046734483266991,
      "grad_norm": 18.5,
      "learning_rate": 1.999867773068666e-06,
      "loss": 2.4833,
      "step": 3550
    },
    {
      "epoch": 0.10482087571870853,
      "grad_norm": 18.5,
      "learning_rate": 1.999859272057329e-06,
      "loss": 2.6006,
      "step": 3555
    },
    {
      "epoch": 0.10496830311071798,
      "grad_norm": 20.75,
      "learning_rate": 1.999850506242207e-06,
      "loss": 2.5066,
      "step": 3560
    },
    {
      "epoch": 0.1051157305027274,
      "grad_norm": 26.875,
      "learning_rate": 1.9998414756256208e-06,
      "loss": 2.5387,
      "step": 3565
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 34.0,
      "learning_rate": 1.9998321802099614e-06,
      "loss": 2.5366,
      "step": 3570
    },
    {
      "epoch": 0.10541058528674628,
      "grad_norm": 28.375,
      "learning_rate": 1.9998226199976925e-06,
      "loss": 2.6194,
      "step": 3575
    },
    {
      "epoch": 0.10555801267875571,
      "grad_norm": 18.0,
      "learning_rate": 1.9998127949913444e-06,
      "loss": 2.4925,
      "step": 3580
    },
    {
      "epoch": 0.10570544007076514,
      "grad_norm": 15.375,
      "learning_rate": 1.9998027051935198e-06,
      "loss": 2.4019,
      "step": 3585
    },
    {
      "epoch": 0.10585286746277459,
      "grad_norm": 17.25,
      "learning_rate": 1.999792350606891e-06,
      "loss": 2.4669,
      "step": 3590
    },
    {
      "epoch": 0.10600029485478402,
      "grad_norm": 17.375,
      "learning_rate": 1.9997817312342e-06,
      "loss": 2.6052,
      "step": 3595
    },
    {
      "epoch": 0.10614772224679346,
      "grad_norm": 17.25,
      "learning_rate": 1.9997708470782596e-06,
      "loss": 2.5472,
      "step": 3600
    },
    {
      "epoch": 0.10629514963880289,
      "grad_norm": 15.25,
      "learning_rate": 1.9997596981419517e-06,
      "loss": 2.4028,
      "step": 3605
    },
    {
      "epoch": 0.10644257703081232,
      "grad_norm": 16.5,
      "learning_rate": 1.99974828442823e-06,
      "loss": 2.5456,
      "step": 3610
    },
    {
      "epoch": 0.10659000442282177,
      "grad_norm": 16.375,
      "learning_rate": 1.9997366059401166e-06,
      "loss": 2.5217,
      "step": 3615
    },
    {
      "epoch": 0.1067374318148312,
      "grad_norm": 17.5,
      "learning_rate": 1.9997246626807045e-06,
      "loss": 2.5225,
      "step": 3620
    },
    {
      "epoch": 0.10688485920684063,
      "grad_norm": 18.25,
      "learning_rate": 1.9997124546531566e-06,
      "loss": 2.4429,
      "step": 3625
    },
    {
      "epoch": 0.10703228659885007,
      "grad_norm": 27.5,
      "learning_rate": 1.999699981860707e-06,
      "loss": 2.5007,
      "step": 3630
    },
    {
      "epoch": 0.1071797139908595,
      "grad_norm": 17.0,
      "learning_rate": 1.999687244306658e-06,
      "loss": 2.4672,
      "step": 3635
    },
    {
      "epoch": 0.10732714138286893,
      "grad_norm": 21.125,
      "learning_rate": 1.9996742419943834e-06,
      "loss": 2.4403,
      "step": 3640
    },
    {
      "epoch": 0.10747456877487838,
      "grad_norm": 20.625,
      "learning_rate": 1.9996609749273268e-06,
      "loss": 2.6091,
      "step": 3645
    },
    {
      "epoch": 0.1076219961668878,
      "grad_norm": 22.25,
      "learning_rate": 1.999647443109002e-06,
      "loss": 2.4657,
      "step": 3650
    },
    {
      "epoch": 0.10776942355889724,
      "grad_norm": 22.0,
      "learning_rate": 1.9996336465429923e-06,
      "loss": 2.4482,
      "step": 3655
    },
    {
      "epoch": 0.10791685095090668,
      "grad_norm": 18.0,
      "learning_rate": 1.999619585232952e-06,
      "loss": 2.4894,
      "step": 3660
    },
    {
      "epoch": 0.10806427834291611,
      "grad_norm": 13.25,
      "learning_rate": 1.999605259182605e-06,
      "loss": 2.4985,
      "step": 3665
    },
    {
      "epoch": 0.10821170573492554,
      "grad_norm": 19.375,
      "learning_rate": 1.999590668395745e-06,
      "loss": 2.5532,
      "step": 3670
    },
    {
      "epoch": 0.10835913312693499,
      "grad_norm": 17.0,
      "learning_rate": 1.9995758128762376e-06,
      "loss": 2.5648,
      "step": 3675
    },
    {
      "epoch": 0.10850656051894442,
      "grad_norm": 15.9375,
      "learning_rate": 1.9995606926280157e-06,
      "loss": 2.4608,
      "step": 3680
    },
    {
      "epoch": 0.10865398791095386,
      "grad_norm": 25.875,
      "learning_rate": 1.999545307655084e-06,
      "loss": 2.5516,
      "step": 3685
    },
    {
      "epoch": 0.10880141530296329,
      "grad_norm": 15.1875,
      "learning_rate": 1.999529657961518e-06,
      "loss": 2.5005,
      "step": 3690
    },
    {
      "epoch": 0.10894884269497272,
      "grad_norm": 19.125,
      "learning_rate": 1.999513743551461e-06,
      "loss": 2.6021,
      "step": 3695
    },
    {
      "epoch": 0.10909627008698217,
      "grad_norm": 18.875,
      "learning_rate": 1.999497564429129e-06,
      "loss": 2.5128,
      "step": 3700
    },
    {
      "epoch": 0.1092436974789916,
      "grad_norm": 19.25,
      "learning_rate": 1.9994811205988063e-06,
      "loss": 2.496,
      "step": 3705
    },
    {
      "epoch": 0.10939112487100103,
      "grad_norm": 20.5,
      "learning_rate": 1.999464412064848e-06,
      "loss": 2.3362,
      "step": 3710
    },
    {
      "epoch": 0.10953855226301047,
      "grad_norm": 21.25,
      "learning_rate": 1.9994474388316794e-06,
      "loss": 2.5267,
      "step": 3715
    },
    {
      "epoch": 0.1096859796550199,
      "grad_norm": 21.5,
      "learning_rate": 1.9994302009037957e-06,
      "loss": 2.4618,
      "step": 3720
    },
    {
      "epoch": 0.10983340704702933,
      "grad_norm": 17.625,
      "learning_rate": 1.9994126982857614e-06,
      "loss": 2.47,
      "step": 3725
    },
    {
      "epoch": 0.10998083443903878,
      "grad_norm": 16.0,
      "learning_rate": 1.999394930982213e-06,
      "loss": 2.5356,
      "step": 3730
    },
    {
      "epoch": 0.11012826183104821,
      "grad_norm": 18.625,
      "learning_rate": 1.9993768989978558e-06,
      "loss": 2.4532,
      "step": 3735
    },
    {
      "epoch": 0.11027568922305764,
      "grad_norm": 15.5625,
      "learning_rate": 1.9993586023374645e-06,
      "loss": 2.5099,
      "step": 3740
    },
    {
      "epoch": 0.11042311661506708,
      "grad_norm": 26.0,
      "learning_rate": 1.9993400410058864e-06,
      "loss": 2.4453,
      "step": 3745
    },
    {
      "epoch": 0.11057054400707651,
      "grad_norm": 17.0,
      "learning_rate": 1.999321215008036e-06,
      "loss": 2.5786,
      "step": 3750
    },
    {
      "epoch": 0.11071797139908596,
      "grad_norm": 15.3125,
      "learning_rate": 1.9993021243488994e-06,
      "loss": 2.5121,
      "step": 3755
    },
    {
      "epoch": 0.11086539879109539,
      "grad_norm": 14.3125,
      "learning_rate": 1.999282769033533e-06,
      "loss": 2.5239,
      "step": 3760
    },
    {
      "epoch": 0.11101282618310482,
      "grad_norm": 32.5,
      "learning_rate": 1.9992631490670623e-06,
      "loss": 2.3592,
      "step": 3765
    },
    {
      "epoch": 0.11116025357511426,
      "grad_norm": 15.6875,
      "learning_rate": 1.9992432644546836e-06,
      "loss": 2.4312,
      "step": 3770
    },
    {
      "epoch": 0.11130768096712369,
      "grad_norm": 38.5,
      "learning_rate": 1.999223115201664e-06,
      "loss": 2.5293,
      "step": 3775
    },
    {
      "epoch": 0.11145510835913312,
      "grad_norm": 17.375,
      "learning_rate": 1.9992027013133393e-06,
      "loss": 2.381,
      "step": 3780
    },
    {
      "epoch": 0.11160253575114257,
      "grad_norm": 19.125,
      "learning_rate": 1.999182022795116e-06,
      "loss": 2.4684,
      "step": 3785
    },
    {
      "epoch": 0.111749963143152,
      "grad_norm": 22.625,
      "learning_rate": 1.9991610796524697e-06,
      "loss": 2.367,
      "step": 3790
    },
    {
      "epoch": 0.11189739053516143,
      "grad_norm": 18.125,
      "learning_rate": 1.999139871890948e-06,
      "loss": 2.5411,
      "step": 3795
    },
    {
      "epoch": 0.11204481792717087,
      "grad_norm": 20.625,
      "learning_rate": 1.999118399516168e-06,
      "loss": 2.4369,
      "step": 3800
    },
    {
      "epoch": 0.1121922453191803,
      "grad_norm": 29.75,
      "learning_rate": 1.9990966625338154e-06,
      "loss": 2.588,
      "step": 3805
    },
    {
      "epoch": 0.11233967271118973,
      "grad_norm": 30.5,
      "learning_rate": 1.9990746609496476e-06,
      "loss": 2.4131,
      "step": 3810
    },
    {
      "epoch": 0.11248710010319918,
      "grad_norm": 12.3125,
      "learning_rate": 1.9990523947694917e-06,
      "loss": 2.3787,
      "step": 3815
    },
    {
      "epoch": 0.11263452749520861,
      "grad_norm": 27.125,
      "learning_rate": 1.999029863999244e-06,
      "loss": 2.4659,
      "step": 3820
    },
    {
      "epoch": 0.11278195488721804,
      "grad_norm": 16.875,
      "learning_rate": 1.9990070686448725e-06,
      "loss": 2.4582,
      "step": 3825
    },
    {
      "epoch": 0.11292938227922748,
      "grad_norm": 19.125,
      "learning_rate": 1.9989840087124134e-06,
      "loss": 2.3936,
      "step": 3830
    },
    {
      "epoch": 0.11307680967123691,
      "grad_norm": 20.375,
      "learning_rate": 1.9989606842079745e-06,
      "loss": 2.5457,
      "step": 3835
    },
    {
      "epoch": 0.11322423706324636,
      "grad_norm": 25.875,
      "learning_rate": 1.998937095137733e-06,
      "loss": 2.5145,
      "step": 3840
    },
    {
      "epoch": 0.11337166445525579,
      "grad_norm": 20.0,
      "learning_rate": 1.998913241507936e-06,
      "loss": 2.5409,
      "step": 3845
    },
    {
      "epoch": 0.11351909184726522,
      "grad_norm": 16.875,
      "learning_rate": 1.998889123324901e-06,
      "loss": 2.4614,
      "step": 3850
    },
    {
      "epoch": 0.11366651923927466,
      "grad_norm": 25.875,
      "learning_rate": 1.998864740595016e-06,
      "loss": 2.432,
      "step": 3855
    },
    {
      "epoch": 0.1138139466312841,
      "grad_norm": 19.375,
      "learning_rate": 1.998840093324738e-06,
      "loss": 2.5605,
      "step": 3860
    },
    {
      "epoch": 0.11396137402329352,
      "grad_norm": 24.625,
      "learning_rate": 1.998815181520595e-06,
      "loss": 2.411,
      "step": 3865
    },
    {
      "epoch": 0.11410880141530297,
      "grad_norm": 18.875,
      "learning_rate": 1.9987900051891843e-06,
      "loss": 2.5734,
      "step": 3870
    },
    {
      "epoch": 0.1142562288073124,
      "grad_norm": 15.75,
      "learning_rate": 1.9987645643371733e-06,
      "loss": 2.4893,
      "step": 3875
    },
    {
      "epoch": 0.11440365619932183,
      "grad_norm": 17.5,
      "learning_rate": 1.998738858971301e-06,
      "loss": 2.5551,
      "step": 3880
    },
    {
      "epoch": 0.11455108359133127,
      "grad_norm": 17.125,
      "learning_rate": 1.9987128890983736e-06,
      "loss": 2.5463,
      "step": 3885
    },
    {
      "epoch": 0.1146985109833407,
      "grad_norm": 16.25,
      "learning_rate": 1.9986866547252704e-06,
      "loss": 2.3747,
      "step": 3890
    },
    {
      "epoch": 0.11484593837535013,
      "grad_norm": 17.625,
      "learning_rate": 1.9986601558589393e-06,
      "loss": 2.5215,
      "step": 3895
    },
    {
      "epoch": 0.11499336576735958,
      "grad_norm": 15.3125,
      "learning_rate": 1.9986333925063968e-06,
      "loss": 2.4341,
      "step": 3900
    },
    {
      "epoch": 0.11514079315936901,
      "grad_norm": 15.25,
      "learning_rate": 1.9986063646747325e-06,
      "loss": 2.5542,
      "step": 3905
    },
    {
      "epoch": 0.11528822055137844,
      "grad_norm": 15.8125,
      "learning_rate": 1.998579072371104e-06,
      "loss": 2.5368,
      "step": 3910
    },
    {
      "epoch": 0.11543564794338788,
      "grad_norm": 23.625,
      "learning_rate": 1.998551515602739e-06,
      "loss": 2.4132,
      "step": 3915
    },
    {
      "epoch": 0.11558307533539731,
      "grad_norm": 17.625,
      "learning_rate": 1.9985236943769358e-06,
      "loss": 2.5002,
      "step": 3920
    },
    {
      "epoch": 0.11573050272740676,
      "grad_norm": 15.625,
      "learning_rate": 1.9984956087010635e-06,
      "loss": 2.2823,
      "step": 3925
    },
    {
      "epoch": 0.11587793011941619,
      "grad_norm": 18.25,
      "learning_rate": 1.9984672585825592e-06,
      "loss": 2.5808,
      "step": 3930
    },
    {
      "epoch": 0.11602535751142562,
      "grad_norm": 17.0,
      "learning_rate": 1.9984386440289315e-06,
      "loss": 2.5604,
      "step": 3935
    },
    {
      "epoch": 0.11617278490343506,
      "grad_norm": 18.5,
      "learning_rate": 1.998409765047759e-06,
      "loss": 2.482,
      "step": 3940
    },
    {
      "epoch": 0.1163202122954445,
      "grad_norm": 14.5,
      "learning_rate": 1.99838062164669e-06,
      "loss": 2.4246,
      "step": 3945
    },
    {
      "epoch": 0.11646763968745392,
      "grad_norm": 16.0,
      "learning_rate": 1.9983512138334425e-06,
      "loss": 2.4198,
      "step": 3950
    },
    {
      "epoch": 0.11661506707946337,
      "grad_norm": 16.25,
      "learning_rate": 1.998321541615805e-06,
      "loss": 2.434,
      "step": 3955
    },
    {
      "epoch": 0.1167624944714728,
      "grad_norm": 24.75,
      "learning_rate": 1.9982916050016364e-06,
      "loss": 2.4377,
      "step": 3960
    },
    {
      "epoch": 0.11690992186348223,
      "grad_norm": 21.625,
      "learning_rate": 1.9982614039988643e-06,
      "loss": 2.4218,
      "step": 3965
    },
    {
      "epoch": 0.11705734925549167,
      "grad_norm": 17.5,
      "learning_rate": 1.9982309386154884e-06,
      "loss": 2.4098,
      "step": 3970
    },
    {
      "epoch": 0.1172047766475011,
      "grad_norm": 16.5,
      "learning_rate": 1.998200208859576e-06,
      "loss": 2.3641,
      "step": 3975
    },
    {
      "epoch": 0.11735220403951054,
      "grad_norm": 17.125,
      "learning_rate": 1.9981692147392655e-06,
      "loss": 2.4643,
      "step": 3980
    },
    {
      "epoch": 0.11749963143151998,
      "grad_norm": 15.625,
      "learning_rate": 1.998137956262767e-06,
      "loss": 2.3954,
      "step": 3985
    },
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 44.75,
      "learning_rate": 1.9981064334383577e-06,
      "loss": 2.4361,
      "step": 3990
    },
    {
      "epoch": 0.11779448621553884,
      "grad_norm": 30.875,
      "learning_rate": 1.998074646274386e-06,
      "loss": 2.5092,
      "step": 3995
    },
    {
      "epoch": 0.11794191360754828,
      "grad_norm": 16.25,
      "learning_rate": 1.998042594779271e-06,
      "loss": 2.5085,
      "step": 4000
    },
    {
      "epoch": 0.11794191360754828,
      "eval_loss": 2.489173173904419,
      "eval_runtime": 4.7135,
      "eval_samples_per_second": 84.014,
      "eval_steps_per_second": 2.758,
      "step": 4000
    },
    {
      "epoch": 0.11808934099955772,
      "grad_norm": 17.875,
      "learning_rate": 1.9980102789615014e-06,
      "loss": 2.5217,
      "step": 4005
    },
    {
      "epoch": 0.11823676839156716,
      "grad_norm": 15.5625,
      "learning_rate": 1.9979776988296353e-06,
      "loss": 2.383,
      "step": 4010
    },
    {
      "epoch": 0.11838419578357659,
      "grad_norm": 66.5,
      "learning_rate": 1.997944854392301e-06,
      "loss": 2.5012,
      "step": 4015
    },
    {
      "epoch": 0.11853162317558602,
      "grad_norm": 18.125,
      "learning_rate": 1.997911745658198e-06,
      "loss": 2.4182,
      "step": 4020
    },
    {
      "epoch": 0.11867905056759546,
      "grad_norm": 22.75,
      "learning_rate": 1.9978783726360945e-06,
      "loss": 2.4518,
      "step": 4025
    },
    {
      "epoch": 0.1188264779596049,
      "grad_norm": 15.75,
      "learning_rate": 1.9978447353348287e-06,
      "loss": 2.3372,
      "step": 4030
    },
    {
      "epoch": 0.11897390535161433,
      "grad_norm": 15.875,
      "learning_rate": 1.9978108337633092e-06,
      "loss": 2.4263,
      "step": 4035
    },
    {
      "epoch": 0.11912133274362377,
      "grad_norm": 17.125,
      "learning_rate": 1.9977766679305143e-06,
      "loss": 2.547,
      "step": 4040
    },
    {
      "epoch": 0.1192687601356332,
      "grad_norm": 16.5,
      "learning_rate": 1.9977422378454936e-06,
      "loss": 2.3247,
      "step": 4045
    },
    {
      "epoch": 0.11941618752764263,
      "grad_norm": 15.5625,
      "learning_rate": 1.9977075435173646e-06,
      "loss": 2.3781,
      "step": 4050
    },
    {
      "epoch": 0.11956361491965208,
      "grad_norm": 23.75,
      "learning_rate": 1.997672584955316e-06,
      "loss": 2.3924,
      "step": 4055
    },
    {
      "epoch": 0.1197110423116615,
      "grad_norm": 21.75,
      "learning_rate": 1.997637362168606e-06,
      "loss": 2.4931,
      "step": 4060
    },
    {
      "epoch": 0.11985846970367094,
      "grad_norm": 20.375,
      "learning_rate": 1.997601875166564e-06,
      "loss": 2.4232,
      "step": 4065
    },
    {
      "epoch": 0.12000589709568038,
      "grad_norm": 13.6875,
      "learning_rate": 1.9975661239585874e-06,
      "loss": 2.3576,
      "step": 4070
    },
    {
      "epoch": 0.12015332448768981,
      "grad_norm": 16.875,
      "learning_rate": 1.997530108554145e-06,
      "loss": 2.439,
      "step": 4075
    },
    {
      "epoch": 0.12030075187969924,
      "grad_norm": 19.75,
      "learning_rate": 1.997493828962775e-06,
      "loss": 2.4518,
      "step": 4080
    },
    {
      "epoch": 0.12044817927170869,
      "grad_norm": 18.125,
      "learning_rate": 1.997457285194086e-06,
      "loss": 2.4058,
      "step": 4085
    },
    {
      "epoch": 0.12059560666371812,
      "grad_norm": 15.4375,
      "learning_rate": 1.9974204772577557e-06,
      "loss": 2.4125,
      "step": 4090
    },
    {
      "epoch": 0.12074303405572756,
      "grad_norm": 15.5,
      "learning_rate": 1.9973834051635332e-06,
      "loss": 2.4462,
      "step": 4095
    },
    {
      "epoch": 0.12089046144773699,
      "grad_norm": 14.25,
      "learning_rate": 1.9973460689212366e-06,
      "loss": 2.3354,
      "step": 4100
    },
    {
      "epoch": 0.12103788883974642,
      "grad_norm": 16.25,
      "learning_rate": 1.997308468540753e-06,
      "loss": 2.3519,
      "step": 4105
    },
    {
      "epoch": 0.12118531623175587,
      "grad_norm": 16.375,
      "learning_rate": 1.997270604032042e-06,
      "loss": 2.3724,
      "step": 4110
    },
    {
      "epoch": 0.1213327436237653,
      "grad_norm": 25.75,
      "learning_rate": 1.9972324754051306e-06,
      "loss": 2.5401,
      "step": 4115
    },
    {
      "epoch": 0.12148017101577473,
      "grad_norm": 17.0,
      "learning_rate": 1.9971940826701175e-06,
      "loss": 2.4093,
      "step": 4120
    },
    {
      "epoch": 0.12162759840778417,
      "grad_norm": 17.25,
      "learning_rate": 1.99715542583717e-06,
      "loss": 2.4163,
      "step": 4125
    },
    {
      "epoch": 0.1217750257997936,
      "grad_norm": 14.625,
      "learning_rate": 1.9971165049165266e-06,
      "loss": 2.397,
      "step": 4130
    },
    {
      "epoch": 0.12192245319180303,
      "grad_norm": 20.625,
      "learning_rate": 1.997077319918495e-06,
      "loss": 2.4253,
      "step": 4135
    },
    {
      "epoch": 0.12206988058381248,
      "grad_norm": 19.625,
      "learning_rate": 1.9970378708534527e-06,
      "loss": 2.3918,
      "step": 4140
    },
    {
      "epoch": 0.1222173079758219,
      "grad_norm": 13.1875,
      "learning_rate": 1.9969981577318476e-06,
      "loss": 2.4003,
      "step": 4145
    },
    {
      "epoch": 0.12236473536783134,
      "grad_norm": 16.375,
      "learning_rate": 1.9969581805641977e-06,
      "loss": 2.3088,
      "step": 4150
    },
    {
      "epoch": 0.12251216275984078,
      "grad_norm": 17.875,
      "learning_rate": 1.99691793936109e-06,
      "loss": 2.3631,
      "step": 4155
    },
    {
      "epoch": 0.12265959015185021,
      "grad_norm": 13.0625,
      "learning_rate": 1.9968774341331828e-06,
      "loss": 2.3745,
      "step": 4160
    },
    {
      "epoch": 0.12280701754385964,
      "grad_norm": 13.625,
      "learning_rate": 1.9968366648912024e-06,
      "loss": 2.3856,
      "step": 4165
    },
    {
      "epoch": 0.12295444493586909,
      "grad_norm": 22.5,
      "learning_rate": 1.9967956316459473e-06,
      "loss": 2.5149,
      "step": 4170
    },
    {
      "epoch": 0.12310187232787852,
      "grad_norm": 23.5,
      "learning_rate": 1.9967543344082845e-06,
      "loss": 2.3066,
      "step": 4175
    },
    {
      "epoch": 0.12324929971988796,
      "grad_norm": 14.125,
      "learning_rate": 1.996712773189151e-06,
      "loss": 2.3431,
      "step": 4180
    },
    {
      "epoch": 0.12339672711189739,
      "grad_norm": 19.625,
      "learning_rate": 1.996670947999554e-06,
      "loss": 2.3903,
      "step": 4185
    },
    {
      "epoch": 0.12354415450390682,
      "grad_norm": 17.625,
      "learning_rate": 1.9966288588505705e-06,
      "loss": 2.4746,
      "step": 4190
    },
    {
      "epoch": 0.12369158189591627,
      "grad_norm": 15.6875,
      "learning_rate": 1.9965865057533474e-06,
      "loss": 2.313,
      "step": 4195
    },
    {
      "epoch": 0.1238390092879257,
      "grad_norm": 17.375,
      "learning_rate": 1.996543888719101e-06,
      "loss": 2.5405,
      "step": 4200
    },
    {
      "epoch": 0.12398643667993513,
      "grad_norm": 16.0,
      "learning_rate": 1.996501007759119e-06,
      "loss": 2.3146,
      "step": 4205
    },
    {
      "epoch": 0.12413386407194457,
      "grad_norm": 17.125,
      "learning_rate": 1.996457862884758e-06,
      "loss": 2.2757,
      "step": 4210
    },
    {
      "epoch": 0.124281291463954,
      "grad_norm": 17.875,
      "learning_rate": 1.996414454107444e-06,
      "loss": 2.3661,
      "step": 4215
    },
    {
      "epoch": 0.12442871885596343,
      "grad_norm": 18.375,
      "learning_rate": 1.996370781438674e-06,
      "loss": 2.3062,
      "step": 4220
    },
    {
      "epoch": 0.12457614624797288,
      "grad_norm": 18.875,
      "learning_rate": 1.9963268448900133e-06,
      "loss": 2.4657,
      "step": 4225
    },
    {
      "epoch": 0.12472357363998231,
      "grad_norm": 16.375,
      "learning_rate": 1.9962826444730992e-06,
      "loss": 2.4002,
      "step": 4230
    },
    {
      "epoch": 0.12487100103199174,
      "grad_norm": 16.75,
      "learning_rate": 1.996238180199637e-06,
      "loss": 2.3283,
      "step": 4235
    },
    {
      "epoch": 0.12501842842400118,
      "grad_norm": 14.375,
      "learning_rate": 1.996193452081403e-06,
      "loss": 2.3856,
      "step": 4240
    },
    {
      "epoch": 0.1251658558160106,
      "grad_norm": 15.4375,
      "learning_rate": 1.996148460130243e-06,
      "loss": 2.3721,
      "step": 4245
    },
    {
      "epoch": 0.12531328320802004,
      "grad_norm": 18.625,
      "learning_rate": 1.9961032043580726e-06,
      "loss": 2.4488,
      "step": 4250
    },
    {
      "epoch": 0.12546071060002947,
      "grad_norm": 27.125,
      "learning_rate": 1.9960576847768784e-06,
      "loss": 2.3926,
      "step": 4255
    },
    {
      "epoch": 0.12560813799203893,
      "grad_norm": 17.625,
      "learning_rate": 1.996011901398714e-06,
      "loss": 2.3488,
      "step": 4260
    },
    {
      "epoch": 0.12575556538404836,
      "grad_norm": 16.375,
      "learning_rate": 1.995965854235706e-06,
      "loss": 2.441,
      "step": 4265
    },
    {
      "epoch": 0.1259029927760578,
      "grad_norm": 15.25,
      "learning_rate": 1.9959195433000496e-06,
      "loss": 2.4117,
      "step": 4270
    },
    {
      "epoch": 0.12605042016806722,
      "grad_norm": 14.75,
      "learning_rate": 1.995872968604009e-06,
      "loss": 2.3304,
      "step": 4275
    },
    {
      "epoch": 0.12619784756007665,
      "grad_norm": 14.3125,
      "learning_rate": 1.9958261301599195e-06,
      "loss": 2.4009,
      "step": 4280
    },
    {
      "epoch": 0.12634527495208608,
      "grad_norm": 17.25,
      "learning_rate": 1.995779027980187e-06,
      "loss": 2.4338,
      "step": 4285
    },
    {
      "epoch": 0.12649270234409554,
      "grad_norm": 14.4375,
      "learning_rate": 1.9957316620772842e-06,
      "loss": 2.4749,
      "step": 4290
    },
    {
      "epoch": 0.12664012973610497,
      "grad_norm": 18.375,
      "learning_rate": 1.9956840324637564e-06,
      "loss": 2.3095,
      "step": 4295
    },
    {
      "epoch": 0.1267875571281144,
      "grad_norm": 20.0,
      "learning_rate": 1.9956361391522177e-06,
      "loss": 2.3714,
      "step": 4300
    },
    {
      "epoch": 0.12693498452012383,
      "grad_norm": 22.125,
      "learning_rate": 1.995587982155353e-06,
      "loss": 2.4341,
      "step": 4305
    },
    {
      "epoch": 0.12708241191213326,
      "grad_norm": 16.625,
      "learning_rate": 1.995539561485915e-06,
      "loss": 2.3509,
      "step": 4310
    },
    {
      "epoch": 0.12722983930414272,
      "grad_norm": 16.0,
      "learning_rate": 1.9954908771567287e-06,
      "loss": 2.4267,
      "step": 4315
    },
    {
      "epoch": 0.12737726669615215,
      "grad_norm": 14.8125,
      "learning_rate": 1.9954419291806865e-06,
      "loss": 2.3689,
      "step": 4320
    },
    {
      "epoch": 0.12752469408816158,
      "grad_norm": 15.4375,
      "learning_rate": 1.995392717570753e-06,
      "loss": 2.3899,
      "step": 4325
    },
    {
      "epoch": 0.127672121480171,
      "grad_norm": 17.25,
      "learning_rate": 1.9953432423399606e-06,
      "loss": 2.2691,
      "step": 4330
    },
    {
      "epoch": 0.12781954887218044,
      "grad_norm": 13.5625,
      "learning_rate": 1.9952935035014126e-06,
      "loss": 2.4342,
      "step": 4335
    },
    {
      "epoch": 0.12796697626418987,
      "grad_norm": 22.0,
      "learning_rate": 1.995243501068282e-06,
      "loss": 2.2938,
      "step": 4340
    },
    {
      "epoch": 0.12811440365619933,
      "grad_norm": 22.75,
      "learning_rate": 1.9951932350538113e-06,
      "loss": 2.3834,
      "step": 4345
    },
    {
      "epoch": 0.12826183104820876,
      "grad_norm": 14.875,
      "learning_rate": 1.9951427054713137e-06,
      "loss": 2.5038,
      "step": 4350
    },
    {
      "epoch": 0.1284092584402182,
      "grad_norm": 16.875,
      "learning_rate": 1.9950919123341707e-06,
      "loss": 2.4017,
      "step": 4355
    },
    {
      "epoch": 0.12855668583222762,
      "grad_norm": 21.375,
      "learning_rate": 1.9950408556558344e-06,
      "loss": 2.3696,
      "step": 4360
    },
    {
      "epoch": 0.12870411322423705,
      "grad_norm": 13.875,
      "learning_rate": 1.9949895354498272e-06,
      "loss": 2.3681,
      "step": 4365
    },
    {
      "epoch": 0.12885154061624648,
      "grad_norm": 13.0,
      "learning_rate": 1.9949379517297404e-06,
      "loss": 2.4246,
      "step": 4370
    },
    {
      "epoch": 0.12899896800825594,
      "grad_norm": 14.5625,
      "learning_rate": 1.994886104509236e-06,
      "loss": 2.4722,
      "step": 4375
    },
    {
      "epoch": 0.12914639540026537,
      "grad_norm": 17.25,
      "learning_rate": 1.994833993802045e-06,
      "loss": 2.3309,
      "step": 4380
    },
    {
      "epoch": 0.1292938227922748,
      "grad_norm": 16.5,
      "learning_rate": 1.994781619621968e-06,
      "loss": 2.385,
      "step": 4385
    },
    {
      "epoch": 0.12944125018428423,
      "grad_norm": 13.75,
      "learning_rate": 1.9947289819828764e-06,
      "loss": 2.301,
      "step": 4390
    },
    {
      "epoch": 0.12958867757629366,
      "grad_norm": 17.0,
      "learning_rate": 1.9946760808987106e-06,
      "loss": 2.4786,
      "step": 4395
    },
    {
      "epoch": 0.12973610496830312,
      "grad_norm": 15.6875,
      "learning_rate": 1.994622916383481e-06,
      "loss": 2.2656,
      "step": 4400
    },
    {
      "epoch": 0.12988353236031255,
      "grad_norm": 14.5625,
      "learning_rate": 1.994569488451268e-06,
      "loss": 2.1932,
      "step": 4405
    },
    {
      "epoch": 0.13003095975232198,
      "grad_norm": 14.875,
      "learning_rate": 1.9945157971162207e-06,
      "loss": 2.3135,
      "step": 4410
    },
    {
      "epoch": 0.13017838714433141,
      "grad_norm": 22.625,
      "learning_rate": 1.99446184239256e-06,
      "loss": 2.3792,
      "step": 4415
    },
    {
      "epoch": 0.13032581453634084,
      "grad_norm": 20.625,
      "learning_rate": 1.9944076242945744e-06,
      "loss": 2.403,
      "step": 4420
    },
    {
      "epoch": 0.13047324192835028,
      "grad_norm": 15.5,
      "learning_rate": 1.9943531428366233e-06,
      "loss": 2.4211,
      "step": 4425
    },
    {
      "epoch": 0.13062066932035973,
      "grad_norm": 16.75,
      "learning_rate": 1.9942983980331355e-06,
      "loss": 2.3437,
      "step": 4430
    },
    {
      "epoch": 0.13076809671236916,
      "grad_norm": 12.0625,
      "learning_rate": 1.99424338989861e-06,
      "loss": 2.3165,
      "step": 4435
    },
    {
      "epoch": 0.1309155241043786,
      "grad_norm": 25.0,
      "learning_rate": 1.994188118447615e-06,
      "loss": 2.396,
      "step": 4440
    },
    {
      "epoch": 0.13106295149638802,
      "grad_norm": 19.125,
      "learning_rate": 1.9941325836947888e-06,
      "loss": 2.3832,
      "step": 4445
    },
    {
      "epoch": 0.13121037888839746,
      "grad_norm": 16.0,
      "learning_rate": 1.9940767856548395e-06,
      "loss": 2.3981,
      "step": 4450
    },
    {
      "epoch": 0.13135780628040689,
      "grad_norm": 17.5,
      "learning_rate": 1.994020724342544e-06,
      "loss": 2.3459,
      "step": 4455
    },
    {
      "epoch": 0.13150523367241634,
      "grad_norm": 15.625,
      "learning_rate": 1.99396439977275e-06,
      "loss": 2.4587,
      "step": 4460
    },
    {
      "epoch": 0.13165266106442577,
      "grad_norm": 25.125,
      "learning_rate": 1.9939078119603746e-06,
      "loss": 2.4407,
      "step": 4465
    },
    {
      "epoch": 0.1318000884564352,
      "grad_norm": 15.875,
      "learning_rate": 1.9938509609204047e-06,
      "loss": 2.3194,
      "step": 4470
    },
    {
      "epoch": 0.13194751584844464,
      "grad_norm": 18.0,
      "learning_rate": 1.9937938466678967e-06,
      "loss": 2.4082,
      "step": 4475
    },
    {
      "epoch": 0.13209494324045407,
      "grad_norm": 14.5625,
      "learning_rate": 1.9937364692179764e-06,
      "loss": 2.2631,
      "step": 4480
    },
    {
      "epoch": 0.13224237063246352,
      "grad_norm": 15.6875,
      "learning_rate": 1.993678828585841e-06,
      "loss": 2.2747,
      "step": 4485
    },
    {
      "epoch": 0.13238979802447295,
      "grad_norm": 15.75,
      "learning_rate": 1.9936209247867542e-06,
      "loss": 2.3008,
      "step": 4490
    },
    {
      "epoch": 0.13253722541648238,
      "grad_norm": 17.0,
      "learning_rate": 1.9935627578360526e-06,
      "loss": 2.3542,
      "step": 4495
    },
    {
      "epoch": 0.13268465280849182,
      "grad_norm": 15.0,
      "learning_rate": 1.9935043277491407e-06,
      "loss": 2.3161,
      "step": 4500
    },
    {
      "epoch": 0.13268465280849182,
      "eval_loss": 2.3908541202545166,
      "eval_runtime": 4.7136,
      "eval_samples_per_second": 84.013,
      "eval_steps_per_second": 2.758,
      "step": 4500
    },
    {
      "epoch": 0.13283208020050125,
      "grad_norm": 15.0625,
      "learning_rate": 1.9934456345414938e-06,
      "loss": 2.3238,
      "step": 4505
    },
    {
      "epoch": 0.13297950759251068,
      "grad_norm": 17.5,
      "learning_rate": 1.9933866782286553e-06,
      "loss": 2.2844,
      "step": 4510
    },
    {
      "epoch": 0.13312693498452013,
      "grad_norm": 15.1875,
      "learning_rate": 1.99332745882624e-06,
      "loss": 2.3055,
      "step": 4515
    },
    {
      "epoch": 0.13327436237652956,
      "grad_norm": 15.0,
      "learning_rate": 1.9932679763499313e-06,
      "loss": 2.4481,
      "step": 4520
    },
    {
      "epoch": 0.133421789768539,
      "grad_norm": 13.4375,
      "learning_rate": 1.9932082308154833e-06,
      "loss": 2.2749,
      "step": 4525
    },
    {
      "epoch": 0.13356921716054843,
      "grad_norm": 16.5,
      "learning_rate": 1.993148222238718e-06,
      "loss": 2.5095,
      "step": 4530
    },
    {
      "epoch": 0.13371664455255786,
      "grad_norm": 13.875,
      "learning_rate": 1.9930879506355285e-06,
      "loss": 2.4653,
      "step": 4535
    },
    {
      "epoch": 0.13386407194456731,
      "grad_norm": 16.375,
      "learning_rate": 1.9930274160218773e-06,
      "loss": 2.3229,
      "step": 4540
    },
    {
      "epoch": 0.13401149933657674,
      "grad_norm": 21.625,
      "learning_rate": 1.9929666184137964e-06,
      "loss": 2.3757,
      "step": 4545
    },
    {
      "epoch": 0.13415892672858618,
      "grad_norm": 16.75,
      "learning_rate": 1.992905557827388e-06,
      "loss": 2.4258,
      "step": 4550
    },
    {
      "epoch": 0.1343063541205956,
      "grad_norm": 13.75,
      "learning_rate": 1.992844234278823e-06,
      "loss": 2.3937,
      "step": 4555
    },
    {
      "epoch": 0.13445378151260504,
      "grad_norm": 19.375,
      "learning_rate": 1.9927826477843416e-06,
      "loss": 2.4173,
      "step": 4560
    },
    {
      "epoch": 0.13460120890461447,
      "grad_norm": 14.75,
      "learning_rate": 1.992720798360256e-06,
      "loss": 2.3872,
      "step": 4565
    },
    {
      "epoch": 0.13474863629662392,
      "grad_norm": 15.75,
      "learning_rate": 1.9926586860229455e-06,
      "loss": 2.3229,
      "step": 4570
    },
    {
      "epoch": 0.13489606368863336,
      "grad_norm": 13.1875,
      "learning_rate": 1.99259631078886e-06,
      "loss": 2.3331,
      "step": 4575
    },
    {
      "epoch": 0.13504349108064279,
      "grad_norm": 14.6875,
      "learning_rate": 1.9925336726745196e-06,
      "loss": 2.3182,
      "step": 4580
    },
    {
      "epoch": 0.13519091847265222,
      "grad_norm": 16.375,
      "learning_rate": 1.992470771696513e-06,
      "loss": 2.2982,
      "step": 4585
    },
    {
      "epoch": 0.13533834586466165,
      "grad_norm": 15.625,
      "learning_rate": 1.992407607871499e-06,
      "loss": 2.3287,
      "step": 4590
    },
    {
      "epoch": 0.13548577325667108,
      "grad_norm": 17.5,
      "learning_rate": 1.992344181216206e-06,
      "loss": 2.4427,
      "step": 4595
    },
    {
      "epoch": 0.13563320064868054,
      "grad_norm": 17.625,
      "learning_rate": 1.9922804917474316e-06,
      "loss": 2.3961,
      "step": 4600
    },
    {
      "epoch": 0.13578062804068997,
      "grad_norm": 16.625,
      "learning_rate": 1.9922165394820445e-06,
      "loss": 2.4671,
      "step": 4605
    },
    {
      "epoch": 0.1359280554326994,
      "grad_norm": 22.25,
      "learning_rate": 1.9921523244369805e-06,
      "loss": 2.4358,
      "step": 4610
    },
    {
      "epoch": 0.13607548282470883,
      "grad_norm": 16.125,
      "learning_rate": 1.9920878466292473e-06,
      "loss": 2.4312,
      "step": 4615
    },
    {
      "epoch": 0.13622291021671826,
      "grad_norm": 18.5,
      "learning_rate": 1.9920231060759207e-06,
      "loss": 2.5213,
      "step": 4620
    },
    {
      "epoch": 0.13637033760872772,
      "grad_norm": 14.8125,
      "learning_rate": 1.9919581027941476e-06,
      "loss": 2.2932,
      "step": 4625
    },
    {
      "epoch": 0.13651776500073715,
      "grad_norm": 20.875,
      "learning_rate": 1.9918928368011426e-06,
      "loss": 2.3134,
      "step": 4630
    },
    {
      "epoch": 0.13666519239274658,
      "grad_norm": 16.25,
      "learning_rate": 1.991827308114191e-06,
      "loss": 2.3506,
      "step": 4635
    },
    {
      "epoch": 0.136812619784756,
      "grad_norm": 14.5,
      "learning_rate": 1.9917615167506477e-06,
      "loss": 2.3828,
      "step": 4640
    },
    {
      "epoch": 0.13696004717676544,
      "grad_norm": 40.25,
      "learning_rate": 1.9916954627279373e-06,
      "loss": 2.3665,
      "step": 4645
    },
    {
      "epoch": 0.13710747456877487,
      "grad_norm": 21.75,
      "learning_rate": 1.9916291460635522e-06,
      "loss": 2.3163,
      "step": 4650
    },
    {
      "epoch": 0.13725490196078433,
      "grad_norm": 14.75,
      "learning_rate": 1.9915625667750577e-06,
      "loss": 2.3346,
      "step": 4655
    },
    {
      "epoch": 0.13740232935279376,
      "grad_norm": 18.125,
      "learning_rate": 1.991495724880085e-06,
      "loss": 2.3689,
      "step": 4660
    },
    {
      "epoch": 0.1375497567448032,
      "grad_norm": 16.0,
      "learning_rate": 1.991428620396338e-06,
      "loss": 2.3924,
      "step": 4665
    },
    {
      "epoch": 0.13769718413681262,
      "grad_norm": 15.5,
      "learning_rate": 1.9913612533415877e-06,
      "loss": 2.2315,
      "step": 4670
    },
    {
      "epoch": 0.13784461152882205,
      "grad_norm": 16.875,
      "learning_rate": 1.9912936237336764e-06,
      "loss": 2.3822,
      "step": 4675
    },
    {
      "epoch": 0.13799203892083148,
      "grad_norm": 21.5,
      "learning_rate": 1.9912257315905145e-06,
      "loss": 2.3552,
      "step": 4680
    },
    {
      "epoch": 0.13813946631284094,
      "grad_norm": 19.5,
      "learning_rate": 1.991157576930083e-06,
      "loss": 2.4338,
      "step": 4685
    },
    {
      "epoch": 0.13828689370485037,
      "grad_norm": 20.875,
      "learning_rate": 1.991089159770432e-06,
      "loss": 2.3953,
      "step": 4690
    },
    {
      "epoch": 0.1384343210968598,
      "grad_norm": 50.0,
      "learning_rate": 1.9910204801296814e-06,
      "loss": 2.3419,
      "step": 4695
    },
    {
      "epoch": 0.13858174848886923,
      "grad_norm": 15.25,
      "learning_rate": 1.99095153802602e-06,
      "loss": 2.342,
      "step": 4700
    },
    {
      "epoch": 0.13872917588087866,
      "grad_norm": 23.25,
      "learning_rate": 1.9908823334777067e-06,
      "loss": 2.4395,
      "step": 4705
    },
    {
      "epoch": 0.13887660327288812,
      "grad_norm": 17.0,
      "learning_rate": 1.9908128665030697e-06,
      "loss": 2.3248,
      "step": 4710
    },
    {
      "epoch": 0.13902403066489755,
      "grad_norm": 17.5,
      "learning_rate": 1.990743137120507e-06,
      "loss": 2.3639,
      "step": 4715
    },
    {
      "epoch": 0.13917145805690698,
      "grad_norm": 18.25,
      "learning_rate": 1.990673145348485e-06,
      "loss": 2.3686,
      "step": 4720
    },
    {
      "epoch": 0.1393188854489164,
      "grad_norm": 15.5,
      "learning_rate": 1.990602891205541e-06,
      "loss": 2.3286,
      "step": 4725
    },
    {
      "epoch": 0.13946631284092584,
      "grad_norm": 19.0,
      "learning_rate": 1.9905323747102813e-06,
      "loss": 2.3668,
      "step": 4730
    },
    {
      "epoch": 0.13961374023293527,
      "grad_norm": 14.125,
      "learning_rate": 1.9904615958813814e-06,
      "loss": 2.3007,
      "step": 4735
    },
    {
      "epoch": 0.13976116762494473,
      "grad_norm": 14.6875,
      "learning_rate": 1.990390554737586e-06,
      "loss": 2.3387,
      "step": 4740
    },
    {
      "epoch": 0.13990859501695416,
      "grad_norm": 17.625,
      "learning_rate": 1.9903192512977104e-06,
      "loss": 2.3229,
      "step": 4745
    },
    {
      "epoch": 0.1400560224089636,
      "grad_norm": 15.75,
      "learning_rate": 1.9902476855806382e-06,
      "loss": 2.3936,
      "step": 4750
    },
    {
      "epoch": 0.14020344980097302,
      "grad_norm": 16.375,
      "learning_rate": 1.990175857605323e-06,
      "loss": 2.4185,
      "step": 4755
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 13.125,
      "learning_rate": 1.9901037673907884e-06,
      "loss": 2.2939,
      "step": 4760
    },
    {
      "epoch": 0.14049830458499188,
      "grad_norm": 16.25,
      "learning_rate": 1.990031414956126e-06,
      "loss": 2.2509,
      "step": 4765
    },
    {
      "epoch": 0.14064573197700134,
      "grad_norm": 14.1875,
      "learning_rate": 1.989958800320498e-06,
      "loss": 2.255,
      "step": 4770
    },
    {
      "epoch": 0.14079315936901077,
      "grad_norm": 15.3125,
      "learning_rate": 1.989885923503136e-06,
      "loss": 2.3749,
      "step": 4775
    },
    {
      "epoch": 0.1409405867610202,
      "grad_norm": 14.0625,
      "learning_rate": 1.98981278452334e-06,
      "loss": 2.3896,
      "step": 4780
    },
    {
      "epoch": 0.14108801415302963,
      "grad_norm": 15.4375,
      "learning_rate": 1.989739383400481e-06,
      "loss": 2.1821,
      "step": 4785
    },
    {
      "epoch": 0.14123544154503906,
      "grad_norm": 17.5,
      "learning_rate": 1.989665720153999e-06,
      "loss": 2.3354,
      "step": 4790
    },
    {
      "epoch": 0.14138286893704852,
      "grad_norm": 14.9375,
      "learning_rate": 1.989591794803402e-06,
      "loss": 2.2849,
      "step": 4795
    },
    {
      "epoch": 0.14153029632905795,
      "grad_norm": 15.1875,
      "learning_rate": 1.9895176073682685e-06,
      "loss": 2.3225,
      "step": 4800
    },
    {
      "epoch": 0.14167772372106738,
      "grad_norm": 12.625,
      "learning_rate": 1.9894431578682474e-06,
      "loss": 2.3356,
      "step": 4805
    },
    {
      "epoch": 0.1418251511130768,
      "grad_norm": 25.0,
      "learning_rate": 1.989368446323055e-06,
      "loss": 2.3309,
      "step": 4810
    },
    {
      "epoch": 0.14197257850508624,
      "grad_norm": 17.875,
      "learning_rate": 1.989293472752479e-06,
      "loss": 2.2506,
      "step": 4815
    },
    {
      "epoch": 0.14212000589709567,
      "grad_norm": 13.875,
      "learning_rate": 1.989218237176374e-06,
      "loss": 2.2593,
      "step": 4820
    },
    {
      "epoch": 0.14226743328910513,
      "grad_norm": 17.0,
      "learning_rate": 1.989142739614667e-06,
      "loss": 2.3378,
      "step": 4825
    },
    {
      "epoch": 0.14241486068111456,
      "grad_norm": 14.1875,
      "learning_rate": 1.9890669800873518e-06,
      "loss": 2.2224,
      "step": 4830
    },
    {
      "epoch": 0.142562288073124,
      "grad_norm": 12.9375,
      "learning_rate": 1.9889909586144927e-06,
      "loss": 2.2454,
      "step": 4835
    },
    {
      "epoch": 0.14270971546513342,
      "grad_norm": 20.0,
      "learning_rate": 1.988914675216224e-06,
      "loss": 2.3326,
      "step": 4840
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 14.5,
      "learning_rate": 1.9888381299127484e-06,
      "loss": 2.289,
      "step": 4845
    },
    {
      "epoch": 0.14300457024915228,
      "grad_norm": 16.0,
      "learning_rate": 1.9887613227243377e-06,
      "loss": 2.3622,
      "step": 4850
    },
    {
      "epoch": 0.14315199764116174,
      "grad_norm": 20.0,
      "learning_rate": 1.9886842536713342e-06,
      "loss": 2.4091,
      "step": 4855
    },
    {
      "epoch": 0.14329942503317117,
      "grad_norm": 13.4375,
      "learning_rate": 1.988606922774149e-06,
      "loss": 2.3741,
      "step": 4860
    },
    {
      "epoch": 0.1434468524251806,
      "grad_norm": 16.375,
      "learning_rate": 1.9885293300532623e-06,
      "loss": 2.3865,
      "step": 4865
    },
    {
      "epoch": 0.14359427981719003,
      "grad_norm": 13.625,
      "learning_rate": 1.9884514755292236e-06,
      "loss": 2.297,
      "step": 4870
    },
    {
      "epoch": 0.14374170720919946,
      "grad_norm": 14.8125,
      "learning_rate": 1.988373359222652e-06,
      "loss": 2.3604,
      "step": 4875
    },
    {
      "epoch": 0.14388913460120892,
      "grad_norm": 17.375,
      "learning_rate": 1.9882949811542362e-06,
      "loss": 2.2758,
      "step": 4880
    },
    {
      "epoch": 0.14403656199321835,
      "grad_norm": 15.75,
      "learning_rate": 1.9882163413447337e-06,
      "loss": 2.4295,
      "step": 4885
    },
    {
      "epoch": 0.14418398938522778,
      "grad_norm": 16.625,
      "learning_rate": 1.9881374398149715e-06,
      "loss": 2.4267,
      "step": 4890
    },
    {
      "epoch": 0.1443314167772372,
      "grad_norm": 18.0,
      "learning_rate": 1.988058276585847e-06,
      "loss": 2.3796,
      "step": 4895
    },
    {
      "epoch": 0.14447884416924664,
      "grad_norm": 32.0,
      "learning_rate": 1.9879788516783242e-06,
      "loss": 2.2386,
      "step": 4900
    },
    {
      "epoch": 0.14462627156125607,
      "grad_norm": 18.0,
      "learning_rate": 1.9878991651134388e-06,
      "loss": 2.3415,
      "step": 4905
    },
    {
      "epoch": 0.14477369895326553,
      "grad_norm": 13.75,
      "learning_rate": 1.9878192169122957e-06,
      "loss": 2.2535,
      "step": 4910
    },
    {
      "epoch": 0.14492112634527496,
      "grad_norm": 17.75,
      "learning_rate": 1.9877390070960677e-06,
      "loss": 2.2624,
      "step": 4915
    },
    {
      "epoch": 0.1450685537372844,
      "grad_norm": 17.125,
      "learning_rate": 1.9876585356859977e-06,
      "loss": 2.3719,
      "step": 4920
    },
    {
      "epoch": 0.14521598112929382,
      "grad_norm": 25.125,
      "learning_rate": 1.987577802703398e-06,
      "loss": 2.3524,
      "step": 4925
    },
    {
      "epoch": 0.14536340852130325,
      "grad_norm": 18.75,
      "learning_rate": 1.98749680816965e-06,
      "loss": 2.3082,
      "step": 4930
    },
    {
      "epoch": 0.14551083591331268,
      "grad_norm": 29.125,
      "learning_rate": 1.9874155521062047e-06,
      "loss": 2.2639,
      "step": 4935
    },
    {
      "epoch": 0.14565826330532214,
      "grad_norm": 13.375,
      "learning_rate": 1.9873340345345816e-06,
      "loss": 2.2192,
      "step": 4940
    },
    {
      "epoch": 0.14580569069733157,
      "grad_norm": 15.875,
      "learning_rate": 1.9872522554763698e-06,
      "loss": 2.4222,
      "step": 4945
    },
    {
      "epoch": 0.145953118089341,
      "grad_norm": 17.625,
      "learning_rate": 1.987170214953228e-06,
      "loss": 2.4073,
      "step": 4950
    },
    {
      "epoch": 0.14610054548135043,
      "grad_norm": 16.0,
      "learning_rate": 1.9870879129868842e-06,
      "loss": 2.3876,
      "step": 4955
    },
    {
      "epoch": 0.14624797287335986,
      "grad_norm": 14.25,
      "learning_rate": 1.987005349599135e-06,
      "loss": 2.3663,
      "step": 4960
    },
    {
      "epoch": 0.14639540026536932,
      "grad_norm": 18.875,
      "learning_rate": 1.9869225248118463e-06,
      "loss": 2.5066,
      "step": 4965
    },
    {
      "epoch": 0.14654282765737875,
      "grad_norm": 12.5,
      "learning_rate": 1.9868394386469535e-06,
      "loss": 2.2031,
      "step": 4970
    },
    {
      "epoch": 0.14669025504938818,
      "grad_norm": 13.625,
      "learning_rate": 1.986756091126462e-06,
      "loss": 2.308,
      "step": 4975
    },
    {
      "epoch": 0.1468376824413976,
      "grad_norm": 17.75,
      "learning_rate": 1.986672482272445e-06,
      "loss": 2.3624,
      "step": 4980
    },
    {
      "epoch": 0.14698510983340704,
      "grad_norm": 16.875,
      "learning_rate": 1.9865886121070463e-06,
      "loss": 2.3513,
      "step": 4985
    },
    {
      "epoch": 0.14713253722541647,
      "grad_norm": 12.4375,
      "learning_rate": 1.986504480652477e-06,
      "loss": 2.2523,
      "step": 4990
    },
    {
      "epoch": 0.14727996461742593,
      "grad_norm": 13.6875,
      "learning_rate": 1.986420087931019e-06,
      "loss": 2.3941,
      "step": 4995
    },
    {
      "epoch": 0.14742739200943536,
      "grad_norm": 15.8125,
      "learning_rate": 1.9863354339650234e-06,
      "loss": 2.2635,
      "step": 5000
    },
    {
      "epoch": 0.14742739200943536,
      "eval_loss": 2.3236687183380127,
      "eval_runtime": 4.7176,
      "eval_samples_per_second": 83.94,
      "eval_steps_per_second": 2.756,
      "step": 5000
    },
    {
      "epoch": 0.1475748194014448,
      "grad_norm": 23.375,
      "learning_rate": 1.98625051877691e-06,
      "loss": 2.3127,
      "step": 5005
    },
    {
      "epoch": 0.14772224679345422,
      "grad_norm": 13.4375,
      "learning_rate": 1.9861653423891667e-06,
      "loss": 2.3383,
      "step": 5010
    },
    {
      "epoch": 0.14786967418546365,
      "grad_norm": 15.75,
      "learning_rate": 1.9860799048243535e-06,
      "loss": 2.3119,
      "step": 5015
    },
    {
      "epoch": 0.14801710157747308,
      "grad_norm": 17.125,
      "learning_rate": 1.9859942061050965e-06,
      "loss": 2.3144,
      "step": 5020
    },
    {
      "epoch": 0.14816452896948254,
      "grad_norm": 13.5625,
      "learning_rate": 1.985908246254093e-06,
      "loss": 2.166,
      "step": 5025
    },
    {
      "epoch": 0.14831195636149197,
      "grad_norm": 15.625,
      "learning_rate": 1.985822025294108e-06,
      "loss": 2.2456,
      "step": 5030
    },
    {
      "epoch": 0.1484593837535014,
      "grad_norm": 15.3125,
      "learning_rate": 1.9857355432479763e-06,
      "loss": 2.3008,
      "step": 5035
    },
    {
      "epoch": 0.14860681114551083,
      "grad_norm": 16.25,
      "learning_rate": 1.9856488001386026e-06,
      "loss": 2.3406,
      "step": 5040
    },
    {
      "epoch": 0.14875423853752026,
      "grad_norm": 15.8125,
      "learning_rate": 1.9855617959889598e-06,
      "loss": 2.3188,
      "step": 5045
    },
    {
      "epoch": 0.14890166592952972,
      "grad_norm": 19.25,
      "learning_rate": 1.9854745308220895e-06,
      "loss": 2.3037,
      "step": 5050
    },
    {
      "epoch": 0.14904909332153915,
      "grad_norm": 16.875,
      "learning_rate": 1.985387004661104e-06,
      "loss": 2.2972,
      "step": 5055
    },
    {
      "epoch": 0.14919652071354858,
      "grad_norm": 15.25,
      "learning_rate": 1.9852992175291837e-06,
      "loss": 2.3653,
      "step": 5060
    },
    {
      "epoch": 0.149343948105558,
      "grad_norm": 18.375,
      "learning_rate": 1.985211169449578e-06,
      "loss": 2.3043,
      "step": 5065
    },
    {
      "epoch": 0.14949137549756744,
      "grad_norm": 15.4375,
      "learning_rate": 1.9851228604456056e-06,
      "loss": 2.3394,
      "step": 5070
    },
    {
      "epoch": 0.14963880288957687,
      "grad_norm": 15.125,
      "learning_rate": 1.985034290540654e-06,
      "loss": 2.2842,
      "step": 5075
    },
    {
      "epoch": 0.14978623028158633,
      "grad_norm": 13.5,
      "learning_rate": 1.984945459758181e-06,
      "loss": 2.2723,
      "step": 5080
    },
    {
      "epoch": 0.14993365767359576,
      "grad_norm": 15.25,
      "learning_rate": 1.984856368121712e-06,
      "loss": 2.2761,
      "step": 5085
    },
    {
      "epoch": 0.1500810850656052,
      "grad_norm": 13.875,
      "learning_rate": 1.9847670156548424e-06,
      "loss": 2.2788,
      "step": 5090
    },
    {
      "epoch": 0.15022851245761462,
      "grad_norm": 15.0625,
      "learning_rate": 1.984677402381236e-06,
      "loss": 2.3517,
      "step": 5095
    },
    {
      "epoch": 0.15037593984962405,
      "grad_norm": 14.125,
      "learning_rate": 1.9845875283246267e-06,
      "loss": 2.2871,
      "step": 5100
    },
    {
      "epoch": 0.15052336724163348,
      "grad_norm": 15.6875,
      "learning_rate": 1.984497393508817e-06,
      "loss": 2.3584,
      "step": 5105
    },
    {
      "epoch": 0.15067079463364294,
      "grad_norm": 15.625,
      "learning_rate": 1.984406997957677e-06,
      "loss": 2.4279,
      "step": 5110
    },
    {
      "epoch": 0.15081822202565237,
      "grad_norm": 16.0,
      "learning_rate": 1.984316341695148e-06,
      "loss": 2.3569,
      "step": 5115
    },
    {
      "epoch": 0.1509656494176618,
      "grad_norm": 26.0,
      "learning_rate": 1.9842254247452402e-06,
      "loss": 2.2929,
      "step": 5120
    },
    {
      "epoch": 0.15111307680967123,
      "grad_norm": 12.625,
      "learning_rate": 1.984134247132031e-06,
      "loss": 2.2725,
      "step": 5125
    },
    {
      "epoch": 0.15126050420168066,
      "grad_norm": 15.1875,
      "learning_rate": 1.984042808879668e-06,
      "loss": 2.2815,
      "step": 5130
    },
    {
      "epoch": 0.15140793159369012,
      "grad_norm": 17.5,
      "learning_rate": 1.983951110012369e-06,
      "loss": 2.2957,
      "step": 5135
    },
    {
      "epoch": 0.15155535898569955,
      "grad_norm": 12.375,
      "learning_rate": 1.9838591505544182e-06,
      "loss": 2.2805,
      "step": 5140
    },
    {
      "epoch": 0.15170278637770898,
      "grad_norm": 16.75,
      "learning_rate": 1.983766930530171e-06,
      "loss": 2.2924,
      "step": 5145
    },
    {
      "epoch": 0.1518502137697184,
      "grad_norm": 14.9375,
      "learning_rate": 1.9836744499640515e-06,
      "loss": 2.2116,
      "step": 5150
    },
    {
      "epoch": 0.15199764116172784,
      "grad_norm": 15.125,
      "learning_rate": 1.983581708880551e-06,
      "loss": 2.3455,
      "step": 5155
    },
    {
      "epoch": 0.15214506855373727,
      "grad_norm": 17.5,
      "learning_rate": 1.983488707304232e-06,
      "loss": 2.35,
      "step": 5160
    },
    {
      "epoch": 0.15229249594574673,
      "grad_norm": 14.625,
      "learning_rate": 1.9833954452597255e-06,
      "loss": 2.2939,
      "step": 5165
    },
    {
      "epoch": 0.15243992333775616,
      "grad_norm": 14.75,
      "learning_rate": 1.9833019227717306e-06,
      "loss": 2.3036,
      "step": 5170
    },
    {
      "epoch": 0.1525873507297656,
      "grad_norm": 14.75,
      "learning_rate": 1.9832081398650158e-06,
      "loss": 2.353,
      "step": 5175
    },
    {
      "epoch": 0.15273477812177502,
      "grad_norm": 13.375,
      "learning_rate": 1.9831140965644187e-06,
      "loss": 2.1931,
      "step": 5180
    },
    {
      "epoch": 0.15288220551378445,
      "grad_norm": 16.0,
      "learning_rate": 1.9830197928948464e-06,
      "loss": 2.3679,
      "step": 5185
    },
    {
      "epoch": 0.15302963290579388,
      "grad_norm": 13.6875,
      "learning_rate": 1.9829252288812735e-06,
      "loss": 2.1672,
      "step": 5190
    },
    {
      "epoch": 0.15317706029780334,
      "grad_norm": 14.5,
      "learning_rate": 1.982830404548745e-06,
      "loss": 2.1524,
      "step": 5195
    },
    {
      "epoch": 0.15332448768981277,
      "grad_norm": 17.5,
      "learning_rate": 1.9827353199223744e-06,
      "loss": 2.4355,
      "step": 5200
    },
    {
      "epoch": 0.1534719150818222,
      "grad_norm": 15.0625,
      "learning_rate": 1.9826399750273432e-06,
      "loss": 2.3482,
      "step": 5205
    },
    {
      "epoch": 0.15361934247383163,
      "grad_norm": 16.875,
      "learning_rate": 1.9825443698889035e-06,
      "loss": 2.3487,
      "step": 5210
    },
    {
      "epoch": 0.15376676986584106,
      "grad_norm": 12.3125,
      "learning_rate": 1.982448504532375e-06,
      "loss": 2.2663,
      "step": 5215
    },
    {
      "epoch": 0.15391419725785052,
      "grad_norm": 14.6875,
      "learning_rate": 1.9823523789831474e-06,
      "loss": 2.3294,
      "step": 5220
    },
    {
      "epoch": 0.15406162464985995,
      "grad_norm": 16.25,
      "learning_rate": 1.982255993266678e-06,
      "loss": 2.2712,
      "step": 5225
    },
    {
      "epoch": 0.15420905204186938,
      "grad_norm": 14.9375,
      "learning_rate": 1.9821593474084938e-06,
      "loss": 2.3677,
      "step": 5230
    },
    {
      "epoch": 0.1543564794338788,
      "grad_norm": 14.5,
      "learning_rate": 1.982062441434191e-06,
      "loss": 2.2574,
      "step": 5235
    },
    {
      "epoch": 0.15450390682588824,
      "grad_norm": 49.0,
      "learning_rate": 1.9819652753694336e-06,
      "loss": 2.209,
      "step": 5240
    },
    {
      "epoch": 0.15465133421789767,
      "grad_norm": 12.25,
      "learning_rate": 1.9818678492399557e-06,
      "loss": 2.2998,
      "step": 5245
    },
    {
      "epoch": 0.15479876160990713,
      "grad_norm": 17.125,
      "learning_rate": 1.98177016307156e-06,
      "loss": 2.1183,
      "step": 5250
    },
    {
      "epoch": 0.15494618900191656,
      "grad_norm": 15.625,
      "learning_rate": 1.981672216890117e-06,
      "loss": 2.2239,
      "step": 5255
    },
    {
      "epoch": 0.155093616393926,
      "grad_norm": 16.375,
      "learning_rate": 1.9815740107215676e-06,
      "loss": 2.2356,
      "step": 5260
    },
    {
      "epoch": 0.15524104378593542,
      "grad_norm": 16.875,
      "learning_rate": 1.9814755445919204e-06,
      "loss": 2.3464,
      "step": 5265
    },
    {
      "epoch": 0.15538847117794485,
      "grad_norm": 14.6875,
      "learning_rate": 1.9813768185272536e-06,
      "loss": 2.3359,
      "step": 5270
    },
    {
      "epoch": 0.15553589856995428,
      "grad_norm": 15.25,
      "learning_rate": 1.981277832553713e-06,
      "loss": 2.3663,
      "step": 5275
    },
    {
      "epoch": 0.15568332596196374,
      "grad_norm": 15.9375,
      "learning_rate": 1.9811785866975153e-06,
      "loss": 2.2262,
      "step": 5280
    },
    {
      "epoch": 0.15583075335397317,
      "grad_norm": 15.375,
      "learning_rate": 1.9810790809849446e-06,
      "loss": 2.2454,
      "step": 5285
    },
    {
      "epoch": 0.1559781807459826,
      "grad_norm": 15.9375,
      "learning_rate": 1.980979315442354e-06,
      "loss": 2.289,
      "step": 5290
    },
    {
      "epoch": 0.15612560813799203,
      "grad_norm": 15.375,
      "learning_rate": 1.980879290096165e-06,
      "loss": 2.2416,
      "step": 5295
    },
    {
      "epoch": 0.15627303553000146,
      "grad_norm": 12.9375,
      "learning_rate": 1.9807790049728692e-06,
      "loss": 2.2994,
      "step": 5300
    },
    {
      "epoch": 0.15642046292201092,
      "grad_norm": 14.5625,
      "learning_rate": 1.9806784600990255e-06,
      "loss": 2.3597,
      "step": 5305
    },
    {
      "epoch": 0.15656789031402035,
      "grad_norm": 21.625,
      "learning_rate": 1.980577655501263e-06,
      "loss": 2.3614,
      "step": 5310
    },
    {
      "epoch": 0.15671531770602978,
      "grad_norm": 12.375,
      "learning_rate": 1.9804765912062786e-06,
      "loss": 2.2174,
      "step": 5315
    },
    {
      "epoch": 0.1568627450980392,
      "grad_norm": 13.4375,
      "learning_rate": 1.9803752672408385e-06,
      "loss": 2.2548,
      "step": 5320
    },
    {
      "epoch": 0.15701017249004864,
      "grad_norm": 15.5625,
      "learning_rate": 1.9802736836317767e-06,
      "loss": 2.3221,
      "step": 5325
    },
    {
      "epoch": 0.15715759988205807,
      "grad_norm": 19.25,
      "learning_rate": 1.9801718404059973e-06,
      "loss": 2.4205,
      "step": 5330
    },
    {
      "epoch": 0.15730502727406753,
      "grad_norm": 15.0625,
      "learning_rate": 1.9800697375904727e-06,
      "loss": 2.3176,
      "step": 5335
    },
    {
      "epoch": 0.15745245466607696,
      "grad_norm": 15.0,
      "learning_rate": 1.9799673752122436e-06,
      "loss": 2.3065,
      "step": 5340
    },
    {
      "epoch": 0.1575998820580864,
      "grad_norm": 37.5,
      "learning_rate": 1.9798647532984197e-06,
      "loss": 2.3744,
      "step": 5345
    },
    {
      "epoch": 0.15774730945009582,
      "grad_norm": 15.875,
      "learning_rate": 1.97976187187618e-06,
      "loss": 2.222,
      "step": 5350
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 22.0,
      "learning_rate": 1.979658730972771e-06,
      "loss": 2.2452,
      "step": 5355
    },
    {
      "epoch": 0.15804216423411468,
      "grad_norm": 16.5,
      "learning_rate": 1.9795553306155096e-06,
      "loss": 2.3767,
      "step": 5360
    },
    {
      "epoch": 0.15818959162612414,
      "grad_norm": 15.875,
      "learning_rate": 1.9794516708317792e-06,
      "loss": 2.2898,
      "step": 5365
    },
    {
      "epoch": 0.15833701901813357,
      "grad_norm": 15.1875,
      "learning_rate": 1.9793477516490343e-06,
      "loss": 2.3616,
      "step": 5370
    },
    {
      "epoch": 0.158484446410143,
      "grad_norm": 15.375,
      "learning_rate": 1.979243573094796e-06,
      "loss": 2.2351,
      "step": 5375
    },
    {
      "epoch": 0.15863187380215243,
      "grad_norm": 15.4375,
      "learning_rate": 1.979139135196656e-06,
      "loss": 2.3049,
      "step": 5380
    },
    {
      "epoch": 0.15877930119416186,
      "grad_norm": 16.625,
      "learning_rate": 1.9790344379822735e-06,
      "loss": 2.3058,
      "step": 5385
    },
    {
      "epoch": 0.15892672858617132,
      "grad_norm": 19.75,
      "learning_rate": 1.978929481479376e-06,
      "loss": 2.4101,
      "step": 5390
    },
    {
      "epoch": 0.15907415597818075,
      "grad_norm": 17.25,
      "learning_rate": 1.9788242657157613e-06,
      "loss": 2.2751,
      "step": 5395
    },
    {
      "epoch": 0.15922158337019018,
      "grad_norm": 16.0,
      "learning_rate": 1.9787187907192936e-06,
      "loss": 2.253,
      "step": 5400
    },
    {
      "epoch": 0.15936901076219961,
      "grad_norm": 15.5,
      "learning_rate": 1.978613056517908e-06,
      "loss": 2.3013,
      "step": 5405
    },
    {
      "epoch": 0.15951643815420904,
      "grad_norm": 15.4375,
      "learning_rate": 1.9785070631396072e-06,
      "loss": 2.2712,
      "step": 5410
    },
    {
      "epoch": 0.15966386554621848,
      "grad_norm": 31.0,
      "learning_rate": 1.978400810612462e-06,
      "loss": 2.3801,
      "step": 5415
    },
    {
      "epoch": 0.15981129293822793,
      "grad_norm": 12.6875,
      "learning_rate": 1.978294298964613e-06,
      "loss": 2.1809,
      "step": 5420
    },
    {
      "epoch": 0.15995872033023736,
      "grad_norm": 15.5625,
      "learning_rate": 1.978187528224269e-06,
      "loss": 2.1765,
      "step": 5425
    },
    {
      "epoch": 0.1601061477222468,
      "grad_norm": 13.875,
      "learning_rate": 1.978080498419706e-06,
      "loss": 2.2463,
      "step": 5430
    },
    {
      "epoch": 0.16025357511425622,
      "grad_norm": 16.75,
      "learning_rate": 1.9779732095792715e-06,
      "loss": 2.1921,
      "step": 5435
    },
    {
      "epoch": 0.16040100250626566,
      "grad_norm": 15.5625,
      "learning_rate": 1.977865661731379e-06,
      "loss": 2.3557,
      "step": 5440
    },
    {
      "epoch": 0.16054842989827509,
      "grad_norm": 12.0,
      "learning_rate": 1.977757854904512e-06,
      "loss": 2.2406,
      "step": 5445
    },
    {
      "epoch": 0.16069585729028454,
      "grad_norm": 16.5,
      "learning_rate": 1.9776497891272222e-06,
      "loss": 2.2514,
      "step": 5450
    },
    {
      "epoch": 0.16084328468229397,
      "grad_norm": 15.3125,
      "learning_rate": 1.9775414644281296e-06,
      "loss": 2.3495,
      "step": 5455
    },
    {
      "epoch": 0.1609907120743034,
      "grad_norm": 17.5,
      "learning_rate": 1.977432880835923e-06,
      "loss": 2.3667,
      "step": 5460
    },
    {
      "epoch": 0.16113813946631284,
      "grad_norm": 14.4375,
      "learning_rate": 1.97732403837936e-06,
      "loss": 2.2498,
      "step": 5465
    },
    {
      "epoch": 0.16128556685832227,
      "grad_norm": 15.5,
      "learning_rate": 1.9772149370872666e-06,
      "loss": 2.2269,
      "step": 5470
    },
    {
      "epoch": 0.16143299425033172,
      "grad_norm": 16.125,
      "learning_rate": 1.9771055769885366e-06,
      "loss": 2.3518,
      "step": 5475
    },
    {
      "epoch": 0.16158042164234115,
      "grad_norm": 14.625,
      "learning_rate": 1.9769959581121343e-06,
      "loss": 2.1878,
      "step": 5480
    },
    {
      "epoch": 0.16172784903435058,
      "grad_norm": 24.0,
      "learning_rate": 1.9768860804870905e-06,
      "loss": 2.3007,
      "step": 5485
    },
    {
      "epoch": 0.16187527642636002,
      "grad_norm": 15.0625,
      "learning_rate": 1.9767759441425055e-06,
      "loss": 2.3047,
      "step": 5490
    },
    {
      "epoch": 0.16202270381836945,
      "grad_norm": 25.25,
      "learning_rate": 1.9766655491075473e-06,
      "loss": 2.2359,
      "step": 5495
    },
    {
      "epoch": 0.16217013121037888,
      "grad_norm": 16.375,
      "learning_rate": 1.976554895411454e-06,
      "loss": 2.3287,
      "step": 5500
    },
    {
      "epoch": 0.16217013121037888,
      "eval_loss": 2.275634765625,
      "eval_runtime": 4.7203,
      "eval_samples_per_second": 83.894,
      "eval_steps_per_second": 2.754,
      "step": 5500
    },
    {
      "epoch": 0.16231755860238833,
      "grad_norm": 16.375,
      "learning_rate": 1.976443983083531e-06,
      "loss": 2.1762,
      "step": 5505
    },
    {
      "epoch": 0.16246498599439776,
      "grad_norm": 15.625,
      "learning_rate": 1.9763328121531517e-06,
      "loss": 2.3403,
      "step": 5510
    },
    {
      "epoch": 0.1626124133864072,
      "grad_norm": 29.125,
      "learning_rate": 1.9762213826497595e-06,
      "loss": 2.2859,
      "step": 5515
    },
    {
      "epoch": 0.16275984077841663,
      "grad_norm": 20.25,
      "learning_rate": 1.9761096946028654e-06,
      "loss": 2.2856,
      "step": 5520
    },
    {
      "epoch": 0.16290726817042606,
      "grad_norm": 14.0625,
      "learning_rate": 1.9759977480420485e-06,
      "loss": 2.282,
      "step": 5525
    },
    {
      "epoch": 0.16305469556243551,
      "grad_norm": 17.25,
      "learning_rate": 1.975885542996958e-06,
      "loss": 2.3487,
      "step": 5530
    },
    {
      "epoch": 0.16320212295444494,
      "grad_norm": 15.5625,
      "learning_rate": 1.9757730794973088e-06,
      "loss": 2.2834,
      "step": 5535
    },
    {
      "epoch": 0.16334955034645438,
      "grad_norm": 15.875,
      "learning_rate": 1.975660357572887e-06,
      "loss": 2.3061,
      "step": 5540
    },
    {
      "epoch": 0.1634969777384638,
      "grad_norm": 14.875,
      "learning_rate": 1.975547377253546e-06,
      "loss": 2.3091,
      "step": 5545
    },
    {
      "epoch": 0.16364440513047324,
      "grad_norm": 18.75,
      "learning_rate": 1.9754341385692067e-06,
      "loss": 2.3004,
      "step": 5550
    },
    {
      "epoch": 0.16379183252248267,
      "grad_norm": 15.5625,
      "learning_rate": 1.97532064154986e-06,
      "loss": 2.2878,
      "step": 5555
    },
    {
      "epoch": 0.16393925991449212,
      "grad_norm": 14.125,
      "learning_rate": 1.975206886225565e-06,
      "loss": 2.2775,
      "step": 5560
    },
    {
      "epoch": 0.16408668730650156,
      "grad_norm": 16.0,
      "learning_rate": 1.975092872626448e-06,
      "loss": 2.273,
      "step": 5565
    },
    {
      "epoch": 0.16423411469851099,
      "grad_norm": 14.75,
      "learning_rate": 1.9749786007827046e-06,
      "loss": 2.2756,
      "step": 5570
    },
    {
      "epoch": 0.16438154209052042,
      "grad_norm": 20.625,
      "learning_rate": 1.9748640707245992e-06,
      "loss": 2.2598,
      "step": 5575
    },
    {
      "epoch": 0.16452896948252985,
      "grad_norm": 24.25,
      "learning_rate": 1.9747492824824638e-06,
      "loss": 2.2965,
      "step": 5580
    },
    {
      "epoch": 0.16467639687453928,
      "grad_norm": 25.375,
      "learning_rate": 1.9746342360866987e-06,
      "loss": 2.2723,
      "step": 5585
    },
    {
      "epoch": 0.16482382426654874,
      "grad_norm": 17.625,
      "learning_rate": 1.974518931567773e-06,
      "loss": 2.4182,
      "step": 5590
    },
    {
      "epoch": 0.16497125165855817,
      "grad_norm": 13.25,
      "learning_rate": 1.9744033689562244e-06,
      "loss": 2.3571,
      "step": 5595
    },
    {
      "epoch": 0.1651186790505676,
      "grad_norm": 15.4375,
      "learning_rate": 1.9742875482826583e-06,
      "loss": 2.2016,
      "step": 5600
    },
    {
      "epoch": 0.16526610644257703,
      "grad_norm": 15.1875,
      "learning_rate": 1.974171469577749e-06,
      "loss": 2.1913,
      "step": 5605
    },
    {
      "epoch": 0.16541353383458646,
      "grad_norm": 15.125,
      "learning_rate": 1.9740551328722387e-06,
      "loss": 2.3939,
      "step": 5610
    },
    {
      "epoch": 0.16556096122659592,
      "grad_norm": 16.125,
      "learning_rate": 1.9739385381969387e-06,
      "loss": 2.31,
      "step": 5615
    },
    {
      "epoch": 0.16570838861860535,
      "grad_norm": 14.375,
      "learning_rate": 1.9738216855827276e-06,
      "loss": 2.3633,
      "step": 5620
    },
    {
      "epoch": 0.16585581601061478,
      "grad_norm": 15.5625,
      "learning_rate": 1.9737045750605523e-06,
      "loss": 2.297,
      "step": 5625
    },
    {
      "epoch": 0.1660032434026242,
      "grad_norm": 17.125,
      "learning_rate": 1.9735872066614294e-06,
      "loss": 2.3581,
      "step": 5630
    },
    {
      "epoch": 0.16615067079463364,
      "grad_norm": 17.0,
      "learning_rate": 1.973469580416442e-06,
      "loss": 2.2865,
      "step": 5635
    },
    {
      "epoch": 0.16629809818664307,
      "grad_norm": 13.375,
      "learning_rate": 1.9733516963567433e-06,
      "loss": 2.245,
      "step": 5640
    },
    {
      "epoch": 0.16644552557865253,
      "grad_norm": 17.625,
      "learning_rate": 1.973233554513553e-06,
      "loss": 2.3806,
      "step": 5645
    },
    {
      "epoch": 0.16659295297066196,
      "grad_norm": 15.4375,
      "learning_rate": 1.9731151549181603e-06,
      "loss": 2.204,
      "step": 5650
    },
    {
      "epoch": 0.1667403803626714,
      "grad_norm": 15.375,
      "learning_rate": 1.9729964976019223e-06,
      "loss": 2.1747,
      "step": 5655
    },
    {
      "epoch": 0.16688780775468082,
      "grad_norm": 15.25,
      "learning_rate": 1.972877582596264e-06,
      "loss": 2.134,
      "step": 5660
    },
    {
      "epoch": 0.16703523514669025,
      "grad_norm": 15.8125,
      "learning_rate": 1.9727584099326796e-06,
      "loss": 2.3692,
      "step": 5665
    },
    {
      "epoch": 0.16718266253869968,
      "grad_norm": 18.375,
      "learning_rate": 1.9726389796427303e-06,
      "loss": 2.3848,
      "step": 5670
    },
    {
      "epoch": 0.16733008993070914,
      "grad_norm": 17.875,
      "learning_rate": 1.9725192917580466e-06,
      "loss": 2.2684,
      "step": 5675
    },
    {
      "epoch": 0.16747751732271857,
      "grad_norm": 15.5625,
      "learning_rate": 1.9723993463103265e-06,
      "loss": 2.3341,
      "step": 5680
    },
    {
      "epoch": 0.167624944714728,
      "grad_norm": 15.25,
      "learning_rate": 1.9722791433313364e-06,
      "loss": 2.3528,
      "step": 5685
    },
    {
      "epoch": 0.16777237210673743,
      "grad_norm": 16.75,
      "learning_rate": 1.972158682852911e-06,
      "loss": 2.2617,
      "step": 5690
    },
    {
      "epoch": 0.16791979949874686,
      "grad_norm": 10.9375,
      "learning_rate": 1.9720379649069537e-06,
      "loss": 2.2348,
      "step": 5695
    },
    {
      "epoch": 0.16806722689075632,
      "grad_norm": 13.0625,
      "learning_rate": 1.9719169895254347e-06,
      "loss": 2.2609,
      "step": 5700
    },
    {
      "epoch": 0.16821465428276575,
      "grad_norm": 15.3125,
      "learning_rate": 1.971795756740394e-06,
      "loss": 2.3166,
      "step": 5705
    },
    {
      "epoch": 0.16836208167477518,
      "grad_norm": 13.9375,
      "learning_rate": 1.9716742665839387e-06,
      "loss": 2.1999,
      "step": 5710
    },
    {
      "epoch": 0.1685095090667846,
      "grad_norm": 15.0625,
      "learning_rate": 1.9715525190882444e-06,
      "loss": 2.2247,
      "step": 5715
    },
    {
      "epoch": 0.16865693645879404,
      "grad_norm": 15.625,
      "learning_rate": 1.9714305142855545e-06,
      "loss": 2.2155,
      "step": 5720
    },
    {
      "epoch": 0.16880436385080347,
      "grad_norm": 18.125,
      "learning_rate": 1.971308252208182e-06,
      "loss": 2.2774,
      "step": 5725
    },
    {
      "epoch": 0.16895179124281293,
      "grad_norm": 17.5,
      "learning_rate": 1.9711857328885056e-06,
      "loss": 2.3131,
      "step": 5730
    },
    {
      "epoch": 0.16909921863482236,
      "grad_norm": 14.75,
      "learning_rate": 1.971062956358974e-06,
      "loss": 2.1557,
      "step": 5735
    },
    {
      "epoch": 0.1692466460268318,
      "grad_norm": 19.0,
      "learning_rate": 1.9709399226521034e-06,
      "loss": 2.3339,
      "step": 5740
    },
    {
      "epoch": 0.16939407341884122,
      "grad_norm": 15.1875,
      "learning_rate": 1.9708166318004785e-06,
      "loss": 2.2882,
      "step": 5745
    },
    {
      "epoch": 0.16954150081085065,
      "grad_norm": 20.375,
      "learning_rate": 1.9706930838367513e-06,
      "loss": 2.2205,
      "step": 5750
    },
    {
      "epoch": 0.16968892820286008,
      "grad_norm": 13.625,
      "learning_rate": 1.9705692787936427e-06,
      "loss": 2.2539,
      "step": 5755
    },
    {
      "epoch": 0.16983635559486954,
      "grad_norm": 13.5625,
      "learning_rate": 1.970445216703941e-06,
      "loss": 2.2197,
      "step": 5760
    },
    {
      "epoch": 0.16998378298687897,
      "grad_norm": 15.375,
      "learning_rate": 1.9703208976005035e-06,
      "loss": 2.3474,
      "step": 5765
    },
    {
      "epoch": 0.1701312103788884,
      "grad_norm": 15.125,
      "learning_rate": 1.9701963215162546e-06,
      "loss": 2.2846,
      "step": 5770
    },
    {
      "epoch": 0.17027863777089783,
      "grad_norm": 13.875,
      "learning_rate": 1.9700714884841872e-06,
      "loss": 2.2496,
      "step": 5775
    },
    {
      "epoch": 0.17042606516290726,
      "grad_norm": 14.125,
      "learning_rate": 1.9699463985373623e-06,
      "loss": 2.2557,
      "step": 5780
    },
    {
      "epoch": 0.17057349255491672,
      "grad_norm": 15.8125,
      "learning_rate": 1.9698210517089085e-06,
      "loss": 2.3008,
      "step": 5785
    },
    {
      "epoch": 0.17072091994692615,
      "grad_norm": 17.125,
      "learning_rate": 1.9696954480320237e-06,
      "loss": 2.3636,
      "step": 5790
    },
    {
      "epoch": 0.17086834733893558,
      "grad_norm": 16.0,
      "learning_rate": 1.9695695875399717e-06,
      "loss": 2.3082,
      "step": 5795
    },
    {
      "epoch": 0.171015774730945,
      "grad_norm": 14.0625,
      "learning_rate": 1.9694434702660866e-06,
      "loss": 2.1362,
      "step": 5800
    },
    {
      "epoch": 0.17116320212295444,
      "grad_norm": 18.625,
      "learning_rate": 1.9693170962437686e-06,
      "loss": 2.196,
      "step": 5805
    },
    {
      "epoch": 0.17131062951496387,
      "grad_norm": 13.3125,
      "learning_rate": 1.9691904655064873e-06,
      "loss": 2.2226,
      "step": 5810
    },
    {
      "epoch": 0.17145805690697333,
      "grad_norm": 15.8125,
      "learning_rate": 1.9690635780877794e-06,
      "loss": 2.2897,
      "step": 5815
    },
    {
      "epoch": 0.17160548429898276,
      "grad_norm": 15.5,
      "learning_rate": 1.96893643402125e-06,
      "loss": 2.3065,
      "step": 5820
    },
    {
      "epoch": 0.1717529116909922,
      "grad_norm": 16.5,
      "learning_rate": 1.968809033340572e-06,
      "loss": 2.2296,
      "step": 5825
    },
    {
      "epoch": 0.17190033908300162,
      "grad_norm": 12.5625,
      "learning_rate": 1.9686813760794865e-06,
      "loss": 2.3175,
      "step": 5830
    },
    {
      "epoch": 0.17204776647501105,
      "grad_norm": 16.125,
      "learning_rate": 1.9685534622718023e-06,
      "loss": 2.2041,
      "step": 5835
    },
    {
      "epoch": 0.17219519386702048,
      "grad_norm": 18.5,
      "learning_rate": 1.9684252919513963e-06,
      "loss": 2.2254,
      "step": 5840
    },
    {
      "epoch": 0.17234262125902994,
      "grad_norm": 16.5,
      "learning_rate": 1.9682968651522133e-06,
      "loss": 2.3361,
      "step": 5845
    },
    {
      "epoch": 0.17249004865103937,
      "grad_norm": 13.625,
      "learning_rate": 1.9681681819082655e-06,
      "loss": 2.275,
      "step": 5850
    },
    {
      "epoch": 0.1726374760430488,
      "grad_norm": 13.9375,
      "learning_rate": 1.968039242253634e-06,
      "loss": 2.199,
      "step": 5855
    },
    {
      "epoch": 0.17278490343505823,
      "grad_norm": 20.5,
      "learning_rate": 1.9679100462224673e-06,
      "loss": 2.2255,
      "step": 5860
    },
    {
      "epoch": 0.17293233082706766,
      "grad_norm": 19.75,
      "learning_rate": 1.967780593848982e-06,
      "loss": 2.3403,
      "step": 5865
    },
    {
      "epoch": 0.17307975821907712,
      "grad_norm": 18.75,
      "learning_rate": 1.9676508851674616e-06,
      "loss": 2.2646,
      "step": 5870
    },
    {
      "epoch": 0.17322718561108655,
      "grad_norm": 15.625,
      "learning_rate": 1.9675209202122587e-06,
      "loss": 2.3766,
      "step": 5875
    },
    {
      "epoch": 0.17337461300309598,
      "grad_norm": 14.3125,
      "learning_rate": 1.967390699017794e-06,
      "loss": 2.266,
      "step": 5880
    },
    {
      "epoch": 0.1735220403951054,
      "grad_norm": 32.5,
      "learning_rate": 1.9672602216185545e-06,
      "loss": 2.1764,
      "step": 5885
    },
    {
      "epoch": 0.17366946778711484,
      "grad_norm": 15.4375,
      "learning_rate": 1.9671294880490966e-06,
      "loss": 2.1586,
      "step": 5890
    },
    {
      "epoch": 0.17381689517912427,
      "grad_norm": 15.0,
      "learning_rate": 1.9669984983440434e-06,
      "loss": 2.2782,
      "step": 5895
    },
    {
      "epoch": 0.17396432257113373,
      "grad_norm": 19.125,
      "learning_rate": 1.9668672525380865e-06,
      "loss": 2.1917,
      "step": 5900
    },
    {
      "epoch": 0.17411174996314316,
      "grad_norm": 14.625,
      "learning_rate": 1.9667357506659856e-06,
      "loss": 2.29,
      "step": 5905
    },
    {
      "epoch": 0.1742591773551526,
      "grad_norm": 13.9375,
      "learning_rate": 1.9666039927625673e-06,
      "loss": 2.2455,
      "step": 5910
    },
    {
      "epoch": 0.17440660474716202,
      "grad_norm": 16.5,
      "learning_rate": 1.9664719788627267e-06,
      "loss": 2.1826,
      "step": 5915
    },
    {
      "epoch": 0.17455403213917145,
      "grad_norm": 14.1875,
      "learning_rate": 1.9663397090014265e-06,
      "loss": 2.2362,
      "step": 5920
    },
    {
      "epoch": 0.17470145953118088,
      "grad_norm": 20.75,
      "learning_rate": 1.9662071832136973e-06,
      "loss": 2.2673,
      "step": 5925
    },
    {
      "epoch": 0.17484888692319034,
      "grad_norm": 17.625,
      "learning_rate": 1.966074401534637e-06,
      "loss": 2.2222,
      "step": 5930
    },
    {
      "epoch": 0.17499631431519977,
      "grad_norm": 18.875,
      "learning_rate": 1.9659413639994124e-06,
      "loss": 2.2023,
      "step": 5935
    },
    {
      "epoch": 0.1751437417072092,
      "grad_norm": 16.125,
      "learning_rate": 1.965808070643256e-06,
      "loss": 2.2542,
      "step": 5940
    },
    {
      "epoch": 0.17529116909921863,
      "grad_norm": 17.0,
      "learning_rate": 1.965674521501471e-06,
      "loss": 2.2869,
      "step": 5945
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 16.75,
      "learning_rate": 1.965540716609425e-06,
      "loss": 2.2464,
      "step": 5950
    },
    {
      "epoch": 0.17558602388323752,
      "grad_norm": 14.75,
      "learning_rate": 1.965406656002556e-06,
      "loss": 2.242,
      "step": 5955
    },
    {
      "epoch": 0.17573345127524695,
      "grad_norm": 12.5,
      "learning_rate": 1.965272339716369e-06,
      "loss": 2.2244,
      "step": 5960
    },
    {
      "epoch": 0.17588087866725638,
      "grad_norm": 17.75,
      "learning_rate": 1.965137767786436e-06,
      "loss": 2.2567,
      "step": 5965
    },
    {
      "epoch": 0.1760283060592658,
      "grad_norm": 15.0625,
      "learning_rate": 1.9650029402483974e-06,
      "loss": 2.258,
      "step": 5970
    },
    {
      "epoch": 0.17617573345127524,
      "grad_norm": 18.375,
      "learning_rate": 1.9648678571379603e-06,
      "loss": 2.268,
      "step": 5975
    },
    {
      "epoch": 0.17632316084328467,
      "grad_norm": 14.8125,
      "learning_rate": 1.9647325184909014e-06,
      "loss": 2.2409,
      "step": 5980
    },
    {
      "epoch": 0.17647058823529413,
      "grad_norm": 15.75,
      "learning_rate": 1.9645969243430632e-06,
      "loss": 2.3359,
      "step": 5985
    },
    {
      "epoch": 0.17661801562730356,
      "grad_norm": 17.625,
      "learning_rate": 1.9644610747303567e-06,
      "loss": 2.3641,
      "step": 5990
    },
    {
      "epoch": 0.176765443019313,
      "grad_norm": 17.125,
      "learning_rate": 1.9643249696887613e-06,
      "loss": 2.2575,
      "step": 5995
    },
    {
      "epoch": 0.17691287041132242,
      "grad_norm": 15.5625,
      "learning_rate": 1.9641886092543215e-06,
      "loss": 2.3008,
      "step": 6000
    },
    {
      "epoch": 0.17691287041132242,
      "eval_loss": 2.2403786182403564,
      "eval_runtime": 4.7252,
      "eval_samples_per_second": 83.807,
      "eval_steps_per_second": 2.751,
      "step": 6000
    },
    {
      "epoch": 0.17706029780333185,
      "grad_norm": 13.25,
      "learning_rate": 1.9640519934631527e-06,
      "loss": 2.1941,
      "step": 6005
    },
    {
      "epoch": 0.17720772519534128,
      "grad_norm": 14.375,
      "learning_rate": 1.9639151223514356e-06,
      "loss": 2.3091,
      "step": 6010
    },
    {
      "epoch": 0.17735515258735074,
      "grad_norm": 19.375,
      "learning_rate": 1.9637779959554193e-06,
      "loss": 2.3316,
      "step": 6015
    },
    {
      "epoch": 0.17750257997936017,
      "grad_norm": 14.75,
      "learning_rate": 1.963640614311421e-06,
      "loss": 2.3345,
      "step": 6020
    },
    {
      "epoch": 0.1776500073713696,
      "grad_norm": 17.125,
      "learning_rate": 1.9635029774558245e-06,
      "loss": 2.1962,
      "step": 6025
    },
    {
      "epoch": 0.17779743476337903,
      "grad_norm": 15.1875,
      "learning_rate": 1.9633650854250818e-06,
      "loss": 2.2806,
      "step": 6030
    },
    {
      "epoch": 0.17794486215538846,
      "grad_norm": 17.25,
      "learning_rate": 1.9632269382557123e-06,
      "loss": 2.2432,
      "step": 6035
    },
    {
      "epoch": 0.17809228954739792,
      "grad_norm": 14.5625,
      "learning_rate": 1.9630885359843034e-06,
      "loss": 2.2453,
      "step": 6040
    },
    {
      "epoch": 0.17823971693940735,
      "grad_norm": 13.25,
      "learning_rate": 1.9629498786475094e-06,
      "loss": 2.328,
      "step": 6045
    },
    {
      "epoch": 0.17838714433141678,
      "grad_norm": 17.0,
      "learning_rate": 1.9628109662820525e-06,
      "loss": 2.2208,
      "step": 6050
    },
    {
      "epoch": 0.1785345717234262,
      "grad_norm": 14.8125,
      "learning_rate": 1.9626717989247222e-06,
      "loss": 2.2077,
      "step": 6055
    },
    {
      "epoch": 0.17868199911543564,
      "grad_norm": 14.75,
      "learning_rate": 1.9625323766123764e-06,
      "loss": 2.2507,
      "step": 6060
    },
    {
      "epoch": 0.17882942650744507,
      "grad_norm": 14.3125,
      "learning_rate": 1.9623926993819394e-06,
      "loss": 2.2847,
      "step": 6065
    },
    {
      "epoch": 0.17897685389945453,
      "grad_norm": 16.0,
      "learning_rate": 1.962252767270403e-06,
      "loss": 2.2177,
      "step": 6070
    },
    {
      "epoch": 0.17912428129146396,
      "grad_norm": 17.875,
      "learning_rate": 1.9621125803148275e-06,
      "loss": 2.2478,
      "step": 6075
    },
    {
      "epoch": 0.1792717086834734,
      "grad_norm": 14.875,
      "learning_rate": 1.9619721385523404e-06,
      "loss": 2.1835,
      "step": 6080
    },
    {
      "epoch": 0.17941913607548282,
      "grad_norm": 18.125,
      "learning_rate": 1.961831442020136e-06,
      "loss": 2.3067,
      "step": 6085
    },
    {
      "epoch": 0.17956656346749225,
      "grad_norm": 17.125,
      "learning_rate": 1.961690490755477e-06,
      "loss": 2.2633,
      "step": 6090
    },
    {
      "epoch": 0.17971399085950168,
      "grad_norm": 14.1875,
      "learning_rate": 1.961549284795692e-06,
      "loss": 2.2044,
      "step": 6095
    },
    {
      "epoch": 0.17986141825151114,
      "grad_norm": 13.0,
      "learning_rate": 1.9614078241781797e-06,
      "loss": 2.1931,
      "step": 6100
    },
    {
      "epoch": 0.18000884564352057,
      "grad_norm": 17.125,
      "learning_rate": 1.961266108940403e-06,
      "loss": 2.2065,
      "step": 6105
    },
    {
      "epoch": 0.18015627303553,
      "grad_norm": 13.8125,
      "learning_rate": 1.9611241391198956e-06,
      "loss": 2.2238,
      "step": 6110
    },
    {
      "epoch": 0.18030370042753943,
      "grad_norm": 15.1875,
      "learning_rate": 1.9609819147542555e-06,
      "loss": 2.2079,
      "step": 6115
    },
    {
      "epoch": 0.18045112781954886,
      "grad_norm": 13.875,
      "learning_rate": 1.9608394358811505e-06,
      "loss": 2.176,
      "step": 6120
    },
    {
      "epoch": 0.18059855521155832,
      "grad_norm": 15.5,
      "learning_rate": 1.9606967025383147e-06,
      "loss": 2.3294,
      "step": 6125
    },
    {
      "epoch": 0.18074598260356775,
      "grad_norm": 13.375,
      "learning_rate": 1.9605537147635493e-06,
      "loss": 2.1568,
      "step": 6130
    },
    {
      "epoch": 0.18089340999557718,
      "grad_norm": 12.8125,
      "learning_rate": 1.960410472594723e-06,
      "loss": 2.1856,
      "step": 6135
    },
    {
      "epoch": 0.1810408373875866,
      "grad_norm": 14.6875,
      "learning_rate": 1.9602669760697737e-06,
      "loss": 2.3059,
      "step": 6140
    },
    {
      "epoch": 0.18118826477959604,
      "grad_norm": 15.875,
      "learning_rate": 1.960123225226704e-06,
      "loss": 2.305,
      "step": 6145
    },
    {
      "epoch": 0.18133569217160547,
      "grad_norm": 16.375,
      "learning_rate": 1.9599792201035852e-06,
      "loss": 2.3145,
      "step": 6150
    },
    {
      "epoch": 0.18148311956361493,
      "grad_norm": 14.8125,
      "learning_rate": 1.959834960738556e-06,
      "loss": 2.2752,
      "step": 6155
    },
    {
      "epoch": 0.18163054695562436,
      "grad_norm": 14.4375,
      "learning_rate": 1.9596904471698223e-06,
      "loss": 2.2703,
      "step": 6160
    },
    {
      "epoch": 0.1817779743476338,
      "grad_norm": 19.75,
      "learning_rate": 1.9595456794356564e-06,
      "loss": 2.1955,
      "step": 6165
    },
    {
      "epoch": 0.18192540173964322,
      "grad_norm": 19.5,
      "learning_rate": 1.9594006575743997e-06,
      "loss": 2.3101,
      "step": 6170
    },
    {
      "epoch": 0.18207282913165265,
      "grad_norm": 17.0,
      "learning_rate": 1.9592553816244596e-06,
      "loss": 2.153,
      "step": 6175
    },
    {
      "epoch": 0.18222025652366208,
      "grad_norm": 15.375,
      "learning_rate": 1.959109851624311e-06,
      "loss": 2.2548,
      "step": 6180
    },
    {
      "epoch": 0.18236768391567154,
      "grad_norm": 16.25,
      "learning_rate": 1.9589640676124963e-06,
      "loss": 2.1377,
      "step": 6185
    },
    {
      "epoch": 0.18251511130768097,
      "grad_norm": 15.875,
      "learning_rate": 1.9588180296276254e-06,
      "loss": 2.2501,
      "step": 6190
    },
    {
      "epoch": 0.1826625386996904,
      "grad_norm": 14.3125,
      "learning_rate": 1.9586717377083748e-06,
      "loss": 2.277,
      "step": 6195
    },
    {
      "epoch": 0.18280996609169983,
      "grad_norm": 12.0,
      "learning_rate": 1.9585251918934884e-06,
      "loss": 2.1835,
      "step": 6200
    },
    {
      "epoch": 0.18295739348370926,
      "grad_norm": 15.1875,
      "learning_rate": 1.958378392221778e-06,
      "loss": 2.0964,
      "step": 6205
    },
    {
      "epoch": 0.18310482087571872,
      "grad_norm": 15.5625,
      "learning_rate": 1.958231338732122e-06,
      "loss": 2.1945,
      "step": 6210
    },
    {
      "epoch": 0.18325224826772815,
      "grad_norm": 15.375,
      "learning_rate": 1.9580840314634665e-06,
      "loss": 2.2639,
      "step": 6215
    },
    {
      "epoch": 0.18339967565973758,
      "grad_norm": 15.6875,
      "learning_rate": 1.957936470454824e-06,
      "loss": 2.2378,
      "step": 6220
    },
    {
      "epoch": 0.183547103051747,
      "grad_norm": 14.0625,
      "learning_rate": 1.957788655745275e-06,
      "loss": 2.2162,
      "step": 6225
    },
    {
      "epoch": 0.18369453044375644,
      "grad_norm": 14.1875,
      "learning_rate": 1.9576405873739664e-06,
      "loss": 2.2687,
      "step": 6230
    },
    {
      "epoch": 0.18384195783576587,
      "grad_norm": 14.6875,
      "learning_rate": 1.9574922653801138e-06,
      "loss": 2.2551,
      "step": 6235
    },
    {
      "epoch": 0.18398938522777533,
      "grad_norm": 15.875,
      "learning_rate": 1.957343689802998e-06,
      "loss": 2.2938,
      "step": 6240
    },
    {
      "epoch": 0.18413681261978476,
      "grad_norm": 13.4375,
      "learning_rate": 1.9571948606819687e-06,
      "loss": 2.2451,
      "step": 6245
    },
    {
      "epoch": 0.1842842400117942,
      "grad_norm": 17.75,
      "learning_rate": 1.9570457780564415e-06,
      "loss": 2.3438,
      "step": 6250
    },
    {
      "epoch": 0.18443166740380362,
      "grad_norm": 14.375,
      "learning_rate": 1.956896441965899e-06,
      "loss": 2.2574,
      "step": 6255
    },
    {
      "epoch": 0.18457909479581305,
      "grad_norm": 18.625,
      "learning_rate": 1.956746852449893e-06,
      "loss": 2.259,
      "step": 6260
    },
    {
      "epoch": 0.18472652218782248,
      "grad_norm": 16.125,
      "learning_rate": 1.95659700954804e-06,
      "loss": 2.2494,
      "step": 6265
    },
    {
      "epoch": 0.18487394957983194,
      "grad_norm": 21.375,
      "learning_rate": 1.9564469133000244e-06,
      "loss": 2.373,
      "step": 6270
    },
    {
      "epoch": 0.18502137697184137,
      "grad_norm": 13.4375,
      "learning_rate": 1.9562965637455984e-06,
      "loss": 2.2845,
      "step": 6275
    },
    {
      "epoch": 0.1851688043638508,
      "grad_norm": 19.125,
      "learning_rate": 1.95614596092458e-06,
      "loss": 2.1535,
      "step": 6280
    },
    {
      "epoch": 0.18531623175586023,
      "grad_norm": 15.5,
      "learning_rate": 1.955995104876856e-06,
      "loss": 2.2633,
      "step": 6285
    },
    {
      "epoch": 0.18546365914786966,
      "grad_norm": 15.5,
      "learning_rate": 1.9558439956423788e-06,
      "loss": 2.3352,
      "step": 6290
    },
    {
      "epoch": 0.18561108653987912,
      "grad_norm": 19.25,
      "learning_rate": 1.955692633261168e-06,
      "loss": 2.3492,
      "step": 6295
    },
    {
      "epoch": 0.18575851393188855,
      "grad_norm": 15.8125,
      "learning_rate": 1.9555410177733108e-06,
      "loss": 2.2918,
      "step": 6300
    },
    {
      "epoch": 0.18590594132389798,
      "grad_norm": 13.4375,
      "learning_rate": 1.9553891492189613e-06,
      "loss": 2.3091,
      "step": 6305
    },
    {
      "epoch": 0.1860533687159074,
      "grad_norm": 12.6875,
      "learning_rate": 1.9552370276383406e-06,
      "loss": 2.2345,
      "step": 6310
    },
    {
      "epoch": 0.18620079610791684,
      "grad_norm": 15.25,
      "learning_rate": 1.9550846530717368e-06,
      "loss": 2.1788,
      "step": 6315
    },
    {
      "epoch": 0.18634822349992627,
      "grad_norm": 15.8125,
      "learning_rate": 1.9549320255595044e-06,
      "loss": 2.2549,
      "step": 6320
    },
    {
      "epoch": 0.18649565089193573,
      "grad_norm": 13.75,
      "learning_rate": 1.954779145142066e-06,
      "loss": 2.2416,
      "step": 6325
    },
    {
      "epoch": 0.18664307828394516,
      "grad_norm": 12.8125,
      "learning_rate": 1.9546260118599103e-06,
      "loss": 2.1621,
      "step": 6330
    },
    {
      "epoch": 0.1867905056759546,
      "grad_norm": 21.0,
      "learning_rate": 1.9544726257535936e-06,
      "loss": 2.3335,
      "step": 6335
    },
    {
      "epoch": 0.18693793306796402,
      "grad_norm": 16.625,
      "learning_rate": 1.9543189868637383e-06,
      "loss": 2.2599,
      "step": 6340
    },
    {
      "epoch": 0.18708536045997345,
      "grad_norm": 15.25,
      "learning_rate": 1.954165095231035e-06,
      "loss": 2.1533,
      "step": 6345
    },
    {
      "epoch": 0.18723278785198288,
      "grad_norm": 13.9375,
      "learning_rate": 1.95401095089624e-06,
      "loss": 2.1882,
      "step": 6350
    },
    {
      "epoch": 0.18738021524399234,
      "grad_norm": 15.8125,
      "learning_rate": 1.9538565539001774e-06,
      "loss": 2.2502,
      "step": 6355
    },
    {
      "epoch": 0.18752764263600177,
      "grad_norm": 13.5625,
      "learning_rate": 1.953701904283737e-06,
      "loss": 2.2742,
      "step": 6360
    },
    {
      "epoch": 0.1876750700280112,
      "grad_norm": 14.375,
      "learning_rate": 1.9535470020878776e-06,
      "loss": 2.2428,
      "step": 6365
    },
    {
      "epoch": 0.18782249742002063,
      "grad_norm": 13.125,
      "learning_rate": 1.953391847353623e-06,
      "loss": 2.1909,
      "step": 6370
    },
    {
      "epoch": 0.18796992481203006,
      "grad_norm": 20.125,
      "learning_rate": 1.9532364401220645e-06,
      "loss": 2.3489,
      "step": 6375
    },
    {
      "epoch": 0.18811735220403952,
      "grad_norm": 18.0,
      "learning_rate": 1.9530807804343603e-06,
      "loss": 2.224,
      "step": 6380
    },
    {
      "epoch": 0.18826477959604895,
      "grad_norm": 15.0,
      "learning_rate": 1.952924868331736e-06,
      "loss": 2.2268,
      "step": 6385
    },
    {
      "epoch": 0.18841220698805838,
      "grad_norm": 13.625,
      "learning_rate": 1.9527687038554828e-06,
      "loss": 2.1472,
      "step": 6390
    },
    {
      "epoch": 0.18855963438006781,
      "grad_norm": 17.75,
      "learning_rate": 1.9526122870469603e-06,
      "loss": 2.3077,
      "step": 6395
    },
    {
      "epoch": 0.18870706177207724,
      "grad_norm": 14.0,
      "learning_rate": 1.952455617947593e-06,
      "loss": 2.3211,
      "step": 6400
    },
    {
      "epoch": 0.18885448916408668,
      "grad_norm": 16.375,
      "learning_rate": 1.952298696598874e-06,
      "loss": 2.2684,
      "step": 6405
    },
    {
      "epoch": 0.18900191655609613,
      "grad_norm": 18.75,
      "learning_rate": 1.952141523042363e-06,
      "loss": 2.3083,
      "step": 6410
    },
    {
      "epoch": 0.18914934394810556,
      "grad_norm": 15.4375,
      "learning_rate": 1.951984097319685e-06,
      "loss": 2.3116,
      "step": 6415
    },
    {
      "epoch": 0.189296771340115,
      "grad_norm": 13.5,
      "learning_rate": 1.9518264194725333e-06,
      "loss": 2.2399,
      "step": 6420
    },
    {
      "epoch": 0.18944419873212442,
      "grad_norm": 15.5,
      "learning_rate": 1.9516684895426676e-06,
      "loss": 2.2946,
      "step": 6425
    },
    {
      "epoch": 0.18959162612413386,
      "grad_norm": 14.0,
      "learning_rate": 1.9515103075719133e-06,
      "loss": 2.2819,
      "step": 6430
    },
    {
      "epoch": 0.18973905351614329,
      "grad_norm": 17.125,
      "learning_rate": 1.951351873602165e-06,
      "loss": 2.3172,
      "step": 6435
    },
    {
      "epoch": 0.18988648090815274,
      "grad_norm": 13.6875,
      "learning_rate": 1.9511931876753813e-06,
      "loss": 2.1765,
      "step": 6440
    },
    {
      "epoch": 0.19003390830016217,
      "grad_norm": 13.75,
      "learning_rate": 1.9510342498335893e-06,
      "loss": 2.1027,
      "step": 6445
    },
    {
      "epoch": 0.1901813356921716,
      "grad_norm": 13.875,
      "learning_rate": 1.9508750601188823e-06,
      "loss": 2.1639,
      "step": 6450
    },
    {
      "epoch": 0.19032876308418104,
      "grad_norm": 14.3125,
      "learning_rate": 1.95071561857342e-06,
      "loss": 2.1655,
      "step": 6455
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 18.375,
      "learning_rate": 1.9505559252394292e-06,
      "loss": 2.2215,
      "step": 6460
    },
    {
      "epoch": 0.19062361786819992,
      "grad_norm": 15.75,
      "learning_rate": 1.9503959801592035e-06,
      "loss": 2.3045,
      "step": 6465
    },
    {
      "epoch": 0.19077104526020935,
      "grad_norm": 22.25,
      "learning_rate": 1.950235783375102e-06,
      "loss": 2.3103,
      "step": 6470
    },
    {
      "epoch": 0.19091847265221878,
      "grad_norm": 15.8125,
      "learning_rate": 1.9500753349295524e-06,
      "loss": 2.2239,
      "step": 6475
    },
    {
      "epoch": 0.19106590004422822,
      "grad_norm": 13.4375,
      "learning_rate": 1.9499146348650477e-06,
      "loss": 2.2765,
      "step": 6480
    },
    {
      "epoch": 0.19121332743623765,
      "grad_norm": 13.25,
      "learning_rate": 1.949753683224148e-06,
      "loss": 2.1889,
      "step": 6485
    },
    {
      "epoch": 0.19136075482824708,
      "grad_norm": 16.375,
      "learning_rate": 1.9495924800494796e-06,
      "loss": 2.1884,
      "step": 6490
    },
    {
      "epoch": 0.19150818222025653,
      "grad_norm": 16.875,
      "learning_rate": 1.9494310253837357e-06,
      "loss": 2.2143,
      "step": 6495
    },
    {
      "epoch": 0.19165560961226596,
      "grad_norm": 15.625,
      "learning_rate": 1.9492693192696766e-06,
      "loss": 2.2323,
      "step": 6500
    },
    {
      "epoch": 0.19165560961226596,
      "eval_loss": 2.214259147644043,
      "eval_runtime": 4.7188,
      "eval_samples_per_second": 83.92,
      "eval_steps_per_second": 2.755,
      "step": 6500
    },
    {
      "epoch": 0.1918030370042754,
      "grad_norm": 19.75,
      "learning_rate": 1.949107361750128e-06,
      "loss": 2.2991,
      "step": 6505
    },
    {
      "epoch": 0.19195046439628483,
      "grad_norm": 16.125,
      "learning_rate": 1.948945152867984e-06,
      "loss": 2.3435,
      "step": 6510
    },
    {
      "epoch": 0.19209789178829426,
      "grad_norm": 13.4375,
      "learning_rate": 1.948782692666203e-06,
      "loss": 2.2187,
      "step": 6515
    },
    {
      "epoch": 0.19224531918030371,
      "grad_norm": 15.6875,
      "learning_rate": 1.9486199811878116e-06,
      "loss": 2.2188,
      "step": 6520
    },
    {
      "epoch": 0.19239274657231314,
      "grad_norm": 12.875,
      "learning_rate": 1.9484570184759027e-06,
      "loss": 2.2358,
      "step": 6525
    },
    {
      "epoch": 0.19254017396432258,
      "grad_norm": 18.0,
      "learning_rate": 1.9482938045736353e-06,
      "loss": 2.2281,
      "step": 6530
    },
    {
      "epoch": 0.192687601356332,
      "grad_norm": 13.875,
      "learning_rate": 1.948130339524235e-06,
      "loss": 2.2475,
      "step": 6535
    },
    {
      "epoch": 0.19283502874834144,
      "grad_norm": 15.5625,
      "learning_rate": 1.9479666233709945e-06,
      "loss": 2.1642,
      "step": 6540
    },
    {
      "epoch": 0.19298245614035087,
      "grad_norm": 16.125,
      "learning_rate": 1.947802656157272e-06,
      "loss": 2.2591,
      "step": 6545
    },
    {
      "epoch": 0.19312988353236032,
      "grad_norm": 17.375,
      "learning_rate": 1.9476384379264933e-06,
      "loss": 2.2818,
      "step": 6550
    },
    {
      "epoch": 0.19327731092436976,
      "grad_norm": 21.875,
      "learning_rate": 1.9474739687221494e-06,
      "loss": 2.1193,
      "step": 6555
    },
    {
      "epoch": 0.19342473831637919,
      "grad_norm": 14.4375,
      "learning_rate": 1.9473092485877994e-06,
      "loss": 2.1854,
      "step": 6560
    },
    {
      "epoch": 0.19357216570838862,
      "grad_norm": 14.625,
      "learning_rate": 1.9471442775670673e-06,
      "loss": 2.2582,
      "step": 6565
    },
    {
      "epoch": 0.19371959310039805,
      "grad_norm": 16.75,
      "learning_rate": 1.9469790557036443e-06,
      "loss": 2.2225,
      "step": 6570
    },
    {
      "epoch": 0.19386702049240748,
      "grad_norm": 17.0,
      "learning_rate": 1.9468135830412886e-06,
      "loss": 2.2606,
      "step": 6575
    },
    {
      "epoch": 0.19401444788441694,
      "grad_norm": 16.875,
      "learning_rate": 1.946647859623823e-06,
      "loss": 2.1635,
      "step": 6580
    },
    {
      "epoch": 0.19416187527642637,
      "grad_norm": 14.9375,
      "learning_rate": 1.9464818854951388e-06,
      "loss": 2.3414,
      "step": 6585
    },
    {
      "epoch": 0.1943093026684358,
      "grad_norm": 14.8125,
      "learning_rate": 1.9463156606991918e-06,
      "loss": 2.4052,
      "step": 6590
    },
    {
      "epoch": 0.19445673006044523,
      "grad_norm": 16.5,
      "learning_rate": 1.9461491852800065e-06,
      "loss": 2.129,
      "step": 6595
    },
    {
      "epoch": 0.19460415745245466,
      "grad_norm": 15.9375,
      "learning_rate": 1.9459824592816716e-06,
      "loss": 2.3136,
      "step": 6600
    },
    {
      "epoch": 0.19475158484446412,
      "grad_norm": 15.125,
      "learning_rate": 1.9458154827483427e-06,
      "loss": 2.1992,
      "step": 6605
    },
    {
      "epoch": 0.19489901223647355,
      "grad_norm": 17.375,
      "learning_rate": 1.9456482557242427e-06,
      "loss": 2.2089,
      "step": 6610
    },
    {
      "epoch": 0.19504643962848298,
      "grad_norm": 17.875,
      "learning_rate": 1.94548077825366e-06,
      "loss": 2.1713,
      "step": 6615
    },
    {
      "epoch": 0.1951938670204924,
      "grad_norm": 14.5,
      "learning_rate": 1.945313050380949e-06,
      "loss": 2.2417,
      "step": 6620
    },
    {
      "epoch": 0.19534129441250184,
      "grad_norm": 14.0625,
      "learning_rate": 1.945145072150532e-06,
      "loss": 2.3048,
      "step": 6625
    },
    {
      "epoch": 0.19548872180451127,
      "grad_norm": 16.75,
      "learning_rate": 1.9449768436068953e-06,
      "loss": 2.1964,
      "step": 6630
    },
    {
      "epoch": 0.19563614919652073,
      "grad_norm": 14.6875,
      "learning_rate": 1.944808364794594e-06,
      "loss": 2.2846,
      "step": 6635
    },
    {
      "epoch": 0.19578357658853016,
      "grad_norm": 17.75,
      "learning_rate": 1.944639635758247e-06,
      "loss": 2.2495,
      "step": 6640
    },
    {
      "epoch": 0.1959310039805396,
      "grad_norm": 15.1875,
      "learning_rate": 1.944470656542541e-06,
      "loss": 2.1223,
      "step": 6645
    },
    {
      "epoch": 0.19607843137254902,
      "grad_norm": 17.5,
      "learning_rate": 1.944301427192229e-06,
      "loss": 2.2621,
      "step": 6650
    },
    {
      "epoch": 0.19622585876455845,
      "grad_norm": 104.0,
      "learning_rate": 1.94413194775213e-06,
      "loss": 2.0909,
      "step": 6655
    },
    {
      "epoch": 0.19637328615656788,
      "grad_norm": 14.5625,
      "learning_rate": 1.9439622182671282e-06,
      "loss": 2.2395,
      "step": 6660
    },
    {
      "epoch": 0.19652071354857734,
      "grad_norm": 14.125,
      "learning_rate": 1.943792238782176e-06,
      "loss": 2.2649,
      "step": 6665
    },
    {
      "epoch": 0.19666814094058677,
      "grad_norm": 14.75,
      "learning_rate": 1.9436220093422907e-06,
      "loss": 2.3661,
      "step": 6670
    },
    {
      "epoch": 0.1968155683325962,
      "grad_norm": 16.375,
      "learning_rate": 1.9434515299925557e-06,
      "loss": 2.349,
      "step": 6675
    },
    {
      "epoch": 0.19696299572460563,
      "grad_norm": 16.625,
      "learning_rate": 1.943280800778121e-06,
      "loss": 2.2771,
      "step": 6680
    },
    {
      "epoch": 0.19711042311661506,
      "grad_norm": 14.9375,
      "learning_rate": 1.9431098217442027e-06,
      "loss": 2.2262,
      "step": 6685
    },
    {
      "epoch": 0.19725785050862452,
      "grad_norm": 14.375,
      "learning_rate": 1.942938592936083e-06,
      "loss": 2.2435,
      "step": 6690
    },
    {
      "epoch": 0.19740527790063395,
      "grad_norm": 13.4375,
      "learning_rate": 1.9427671143991103e-06,
      "loss": 2.1591,
      "step": 6695
    },
    {
      "epoch": 0.19755270529264338,
      "grad_norm": 14.75,
      "learning_rate": 1.9425953861787e-06,
      "loss": 2.296,
      "step": 6700
    },
    {
      "epoch": 0.1977001326846528,
      "grad_norm": 17.875,
      "learning_rate": 1.942423408320332e-06,
      "loss": 2.2397,
      "step": 6705
    },
    {
      "epoch": 0.19784756007666224,
      "grad_norm": 12.25,
      "learning_rate": 1.9422511808695525e-06,
      "loss": 2.2015,
      "step": 6710
    },
    {
      "epoch": 0.19799498746867167,
      "grad_norm": 15.25,
      "learning_rate": 1.942078703871976e-06,
      "loss": 2.2699,
      "step": 6715
    },
    {
      "epoch": 0.19814241486068113,
      "grad_norm": 14.8125,
      "learning_rate": 1.94190597737328e-06,
      "loss": 2.0769,
      "step": 6720
    },
    {
      "epoch": 0.19828984225269056,
      "grad_norm": 14.375,
      "learning_rate": 1.9417330014192103e-06,
      "loss": 2.0982,
      "step": 6725
    },
    {
      "epoch": 0.1984372696447,
      "grad_norm": 15.875,
      "learning_rate": 1.941559776055578e-06,
      "loss": 2.2998,
      "step": 6730
    },
    {
      "epoch": 0.19858469703670942,
      "grad_norm": 12.4375,
      "learning_rate": 1.94138630132826e-06,
      "loss": 2.1314,
      "step": 6735
    },
    {
      "epoch": 0.19873212442871885,
      "grad_norm": 18.5,
      "learning_rate": 1.9412125772832e-06,
      "loss": 2.2776,
      "step": 6740
    },
    {
      "epoch": 0.19887955182072828,
      "grad_norm": 13.1875,
      "learning_rate": 1.9410386039664067e-06,
      "loss": 2.1857,
      "step": 6745
    },
    {
      "epoch": 0.19902697921273774,
      "grad_norm": 14.4375,
      "learning_rate": 1.940864381423956e-06,
      "loss": 2.2383,
      "step": 6750
    },
    {
      "epoch": 0.19917440660474717,
      "grad_norm": 12.9375,
      "learning_rate": 1.9406899097019883e-06,
      "loss": 2.2486,
      "step": 6755
    },
    {
      "epoch": 0.1993218339967566,
      "grad_norm": 16.875,
      "learning_rate": 1.940515188846712e-06,
      "loss": 2.3064,
      "step": 6760
    },
    {
      "epoch": 0.19946926138876603,
      "grad_norm": 14.0,
      "learning_rate": 1.9403402189043994e-06,
      "loss": 2.1471,
      "step": 6765
    },
    {
      "epoch": 0.19961668878077546,
      "grad_norm": 14.0625,
      "learning_rate": 1.9401649999213904e-06,
      "loss": 2.1606,
      "step": 6770
    },
    {
      "epoch": 0.19976411617278492,
      "grad_norm": 19.0,
      "learning_rate": 1.9399895319440893e-06,
      "loss": 2.2425,
      "step": 6775
    },
    {
      "epoch": 0.19991154356479435,
      "grad_norm": 16.375,
      "learning_rate": 1.9398138150189683e-06,
      "loss": 2.1473,
      "step": 6780
    },
    {
      "epoch": 0.20005897095680378,
      "grad_norm": 13.375,
      "learning_rate": 1.9396378491925636e-06,
      "loss": 2.1876,
      "step": 6785
    },
    {
      "epoch": 0.2002063983488132,
      "grad_norm": 16.125,
      "learning_rate": 1.939461634511479e-06,
      "loss": 2.2899,
      "step": 6790
    },
    {
      "epoch": 0.20035382574082264,
      "grad_norm": 12.625,
      "learning_rate": 1.9392851710223823e-06,
      "loss": 2.1244,
      "step": 6795
    },
    {
      "epoch": 0.20050125313283207,
      "grad_norm": 11.9375,
      "learning_rate": 1.9391084587720093e-06,
      "loss": 2.257,
      "step": 6800
    },
    {
      "epoch": 0.20064868052484153,
      "grad_norm": 28.875,
      "learning_rate": 1.93893149780716e-06,
      "loss": 2.2221,
      "step": 6805
    },
    {
      "epoch": 0.20079610791685096,
      "grad_norm": 12.0625,
      "learning_rate": 1.9387542881747016e-06,
      "loss": 2.1064,
      "step": 6810
    },
    {
      "epoch": 0.2009435353088604,
      "grad_norm": 13.125,
      "learning_rate": 1.9385768299215656e-06,
      "loss": 2.1955,
      "step": 6815
    },
    {
      "epoch": 0.20109096270086982,
      "grad_norm": 15.8125,
      "learning_rate": 1.938399123094751e-06,
      "loss": 2.2424,
      "step": 6820
    },
    {
      "epoch": 0.20123839009287925,
      "grad_norm": 14.25,
      "learning_rate": 1.9382211677413213e-06,
      "loss": 2.1846,
      "step": 6825
    },
    {
      "epoch": 0.20138581748488868,
      "grad_norm": 14.375,
      "learning_rate": 1.938042963908407e-06,
      "loss": 2.2189,
      "step": 6830
    },
    {
      "epoch": 0.20153324487689814,
      "grad_norm": 13.1875,
      "learning_rate": 1.937864511643203e-06,
      "loss": 2.3278,
      "step": 6835
    },
    {
      "epoch": 0.20168067226890757,
      "grad_norm": 15.25,
      "learning_rate": 1.9376858109929713e-06,
      "loss": 2.2339,
      "step": 6840
    },
    {
      "epoch": 0.201828099660917,
      "grad_norm": 16.375,
      "learning_rate": 1.937506862005039e-06,
      "loss": 2.2715,
      "step": 6845
    },
    {
      "epoch": 0.20197552705292643,
      "grad_norm": 14.0625,
      "learning_rate": 1.9373276647267996e-06,
      "loss": 2.0827,
      "step": 6850
    },
    {
      "epoch": 0.20212295444493586,
      "grad_norm": 14.125,
      "learning_rate": 1.9371482192057114e-06,
      "loss": 2.2222,
      "step": 6855
    },
    {
      "epoch": 0.20227038183694532,
      "grad_norm": 14.5,
      "learning_rate": 1.936968525489299e-06,
      "loss": 2.2901,
      "step": 6860
    },
    {
      "epoch": 0.20241780922895475,
      "grad_norm": 12.0625,
      "learning_rate": 1.936788583625153e-06,
      "loss": 2.0957,
      "step": 6865
    },
    {
      "epoch": 0.20256523662096418,
      "grad_norm": 12.6875,
      "learning_rate": 1.936608393660929e-06,
      "loss": 2.1918,
      "step": 6870
    },
    {
      "epoch": 0.2027126640129736,
      "grad_norm": 19.625,
      "learning_rate": 1.9364279556443486e-06,
      "loss": 2.2613,
      "step": 6875
    },
    {
      "epoch": 0.20286009140498304,
      "grad_norm": 14.75,
      "learning_rate": 1.9362472696231994e-06,
      "loss": 2.2336,
      "step": 6880
    },
    {
      "epoch": 0.20300751879699247,
      "grad_norm": 12.75,
      "learning_rate": 1.9360663356453344e-06,
      "loss": 2.2267,
      "step": 6885
    },
    {
      "epoch": 0.20315494618900193,
      "grad_norm": 16.375,
      "learning_rate": 1.935885153758673e-06,
      "loss": 2.311,
      "step": 6890
    },
    {
      "epoch": 0.20330237358101136,
      "grad_norm": 14.625,
      "learning_rate": 1.9357037240111985e-06,
      "loss": 2.2432,
      "step": 6895
    },
    {
      "epoch": 0.2034498009730208,
      "grad_norm": 11.75,
      "learning_rate": 1.9355220464509617e-06,
      "loss": 2.17,
      "step": 6900
    },
    {
      "epoch": 0.20359722836503022,
      "grad_norm": 14.875,
      "learning_rate": 1.935340121126078e-06,
      "loss": 2.1693,
      "step": 6905
    },
    {
      "epoch": 0.20374465575703965,
      "grad_norm": 14.875,
      "learning_rate": 1.9351579480847288e-06,
      "loss": 2.2669,
      "step": 6910
    },
    {
      "epoch": 0.20389208314904908,
      "grad_norm": 17.25,
      "learning_rate": 1.934975527375161e-06,
      "loss": 2.2202,
      "step": 6915
    },
    {
      "epoch": 0.20403951054105854,
      "grad_norm": 13.4375,
      "learning_rate": 1.9347928590456874e-06,
      "loss": 2.2297,
      "step": 6920
    },
    {
      "epoch": 0.20418693793306797,
      "grad_norm": 15.625,
      "learning_rate": 1.9346099431446853e-06,
      "loss": 2.1914,
      "step": 6925
    },
    {
      "epoch": 0.2043343653250774,
      "grad_norm": 13.625,
      "learning_rate": 1.9344267797205988e-06,
      "loss": 2.2208,
      "step": 6930
    },
    {
      "epoch": 0.20448179271708683,
      "grad_norm": 13.25,
      "learning_rate": 1.934243368821937e-06,
      "loss": 2.1804,
      "step": 6935
    },
    {
      "epoch": 0.20462922010909626,
      "grad_norm": 17.0,
      "learning_rate": 1.934059710497275e-06,
      "loss": 2.2179,
      "step": 6940
    },
    {
      "epoch": 0.20477664750110572,
      "grad_norm": 15.1875,
      "learning_rate": 1.9338758047952527e-06,
      "loss": 2.1055,
      "step": 6945
    },
    {
      "epoch": 0.20492407489311515,
      "grad_norm": 14.625,
      "learning_rate": 1.9336916517645757e-06,
      "loss": 2.2436,
      "step": 6950
    },
    {
      "epoch": 0.20507150228512458,
      "grad_norm": 16.625,
      "learning_rate": 1.933507251454016e-06,
      "loss": 2.1477,
      "step": 6955
    },
    {
      "epoch": 0.205218929677134,
      "grad_norm": 15.5625,
      "learning_rate": 1.933322603912409e-06,
      "loss": 2.1234,
      "step": 6960
    },
    {
      "epoch": 0.20536635706914344,
      "grad_norm": 14.4375,
      "learning_rate": 1.933137709188659e-06,
      "loss": 2.257,
      "step": 6965
    },
    {
      "epoch": 0.20551378446115287,
      "grad_norm": 13.875,
      "learning_rate": 1.932952567331732e-06,
      "loss": 2.1056,
      "step": 6970
    },
    {
      "epoch": 0.20566121185316233,
      "grad_norm": 17.125,
      "learning_rate": 1.9327671783906614e-06,
      "loss": 2.161,
      "step": 6975
    },
    {
      "epoch": 0.20580863924517176,
      "grad_norm": 25.875,
      "learning_rate": 1.9325815424145465e-06,
      "loss": 2.1589,
      "step": 6980
    },
    {
      "epoch": 0.2059560666371812,
      "grad_norm": 16.625,
      "learning_rate": 1.9323956594525514e-06,
      "loss": 2.2228,
      "step": 6985
    },
    {
      "epoch": 0.20610349402919062,
      "grad_norm": 18.0,
      "learning_rate": 1.9322095295539045e-06,
      "loss": 2.2412,
      "step": 6990
    },
    {
      "epoch": 0.20625092142120005,
      "grad_norm": 15.125,
      "learning_rate": 1.9320231527679014e-06,
      "loss": 2.2868,
      "step": 6995
    },
    {
      "epoch": 0.20639834881320948,
      "grad_norm": 14.1875,
      "learning_rate": 1.931836529143902e-06,
      "loss": 2.1774,
      "step": 7000
    },
    {
      "epoch": 0.20639834881320948,
      "eval_loss": 2.193103075027466,
      "eval_runtime": 4.7184,
      "eval_samples_per_second": 83.927,
      "eval_steps_per_second": 2.755,
      "step": 7000
    },
    {
      "epoch": 0.20654577620521894,
      "grad_norm": 15.0625,
      "learning_rate": 1.9316496587313323e-06,
      "loss": 2.3368,
      "step": 7005
    },
    {
      "epoch": 0.20669320359722837,
      "grad_norm": 13.625,
      "learning_rate": 1.931462541579683e-06,
      "loss": 2.3225,
      "step": 7010
    },
    {
      "epoch": 0.2068406309892378,
      "grad_norm": 14.0625,
      "learning_rate": 1.9312751777385103e-06,
      "loss": 2.2313,
      "step": 7015
    },
    {
      "epoch": 0.20698805838124723,
      "grad_norm": 15.25,
      "learning_rate": 1.931087567257436e-06,
      "loss": 2.2718,
      "step": 7020
    },
    {
      "epoch": 0.20713548577325666,
      "grad_norm": 13.375,
      "learning_rate": 1.9308997101861474e-06,
      "loss": 2.1488,
      "step": 7025
    },
    {
      "epoch": 0.20728291316526612,
      "grad_norm": 14.25,
      "learning_rate": 1.930711606574396e-06,
      "loss": 2.2197,
      "step": 7030
    },
    {
      "epoch": 0.20743034055727555,
      "grad_norm": 16.125,
      "learning_rate": 1.930523256472e-06,
      "loss": 2.1245,
      "step": 7035
    },
    {
      "epoch": 0.20757776794928498,
      "grad_norm": 21.625,
      "learning_rate": 1.9303346599288415e-06,
      "loss": 2.2834,
      "step": 7040
    },
    {
      "epoch": 0.2077251953412944,
      "grad_norm": 13.1875,
      "learning_rate": 1.9301458169948695e-06,
      "loss": 2.145,
      "step": 7045
    },
    {
      "epoch": 0.20787262273330384,
      "grad_norm": 13.875,
      "learning_rate": 1.929956727720097e-06,
      "loss": 2.206,
      "step": 7050
    },
    {
      "epoch": 0.20802005012531327,
      "grad_norm": 13.4375,
      "learning_rate": 1.9297673921546026e-06,
      "loss": 2.1265,
      "step": 7055
    },
    {
      "epoch": 0.20816747751732273,
      "grad_norm": 15.3125,
      "learning_rate": 1.92957781034853e-06,
      "loss": 2.283,
      "step": 7060
    },
    {
      "epoch": 0.20831490490933216,
      "grad_norm": 15.125,
      "learning_rate": 1.929387982352088e-06,
      "loss": 2.2967,
      "step": 7065
    },
    {
      "epoch": 0.2084623323013416,
      "grad_norm": 17.0,
      "learning_rate": 1.9291979082155514e-06,
      "loss": 2.2077,
      "step": 7070
    },
    {
      "epoch": 0.20860975969335102,
      "grad_norm": 11.75,
      "learning_rate": 1.9290075879892593e-06,
      "loss": 2.1595,
      "step": 7075
    },
    {
      "epoch": 0.20875718708536045,
      "grad_norm": 15.0,
      "learning_rate": 1.9288170217236167e-06,
      "loss": 2.272,
      "step": 7080
    },
    {
      "epoch": 0.20890461447736988,
      "grad_norm": 13.9375,
      "learning_rate": 1.928626209469093e-06,
      "loss": 2.2495,
      "step": 7085
    },
    {
      "epoch": 0.20905204186937934,
      "grad_norm": 15.8125,
      "learning_rate": 1.9284351512762235e-06,
      "loss": 2.202,
      "step": 7090
    },
    {
      "epoch": 0.20919946926138877,
      "grad_norm": 17.5,
      "learning_rate": 1.9282438471956074e-06,
      "loss": 2.238,
      "step": 7095
    },
    {
      "epoch": 0.2093468966533982,
      "grad_norm": 19.5,
      "learning_rate": 1.9280522972779105e-06,
      "loss": 2.2096,
      "step": 7100
    },
    {
      "epoch": 0.20949432404540763,
      "grad_norm": 15.5,
      "learning_rate": 1.9278605015738635e-06,
      "loss": 2.3443,
      "step": 7105
    },
    {
      "epoch": 0.20964175143741706,
      "grad_norm": 15.3125,
      "learning_rate": 1.927668460134261e-06,
      "loss": 2.0769,
      "step": 7110
    },
    {
      "epoch": 0.20978917882942652,
      "grad_norm": 17.125,
      "learning_rate": 1.927476173009964e-06,
      "loss": 2.1903,
      "step": 7115
    },
    {
      "epoch": 0.20993660622143595,
      "grad_norm": 18.75,
      "learning_rate": 1.9272836402518975e-06,
      "loss": 2.3107,
      "step": 7120
    },
    {
      "epoch": 0.21008403361344538,
      "grad_norm": 15.6875,
      "learning_rate": 1.927090861911053e-06,
      "loss": 2.3182,
      "step": 7125
    },
    {
      "epoch": 0.2102314610054548,
      "grad_norm": 16.75,
      "learning_rate": 1.9268978380384846e-06,
      "loss": 2.2175,
      "step": 7130
    },
    {
      "epoch": 0.21037888839746424,
      "grad_norm": 15.5625,
      "learning_rate": 1.926704568685314e-06,
      "loss": 2.219,
      "step": 7135
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 15.1875,
      "learning_rate": 1.9265110539027273e-06,
      "loss": 2.1772,
      "step": 7140
    },
    {
      "epoch": 0.21067374318148313,
      "grad_norm": 16.5,
      "learning_rate": 1.9263172937419742e-06,
      "loss": 2.2305,
      "step": 7145
    },
    {
      "epoch": 0.21082117057349256,
      "grad_norm": 25.5,
      "learning_rate": 1.9261232882543706e-06,
      "loss": 2.3339,
      "step": 7150
    },
    {
      "epoch": 0.210968597965502,
      "grad_norm": 13.375,
      "learning_rate": 1.9259290374912976e-06,
      "loss": 2.0968,
      "step": 7155
    },
    {
      "epoch": 0.21111602535751142,
      "grad_norm": 12.8125,
      "learning_rate": 1.9257345415042e-06,
      "loss": 2.1461,
      "step": 7160
    },
    {
      "epoch": 0.21126345274952085,
      "grad_norm": 13.75,
      "learning_rate": 1.9255398003445887e-06,
      "loss": 2.2249,
      "step": 7165
    },
    {
      "epoch": 0.21141088014153028,
      "grad_norm": 16.75,
      "learning_rate": 1.9253448140640392e-06,
      "loss": 2.3032,
      "step": 7170
    },
    {
      "epoch": 0.21155830753353974,
      "grad_norm": 14.625,
      "learning_rate": 1.925149582714192e-06,
      "loss": 2.1457,
      "step": 7175
    },
    {
      "epoch": 0.21170573492554917,
      "grad_norm": 14.75,
      "learning_rate": 1.9249541063467524e-06,
      "loss": 2.1607,
      "step": 7180
    },
    {
      "epoch": 0.2118531623175586,
      "grad_norm": 16.375,
      "learning_rate": 1.92475838501349e-06,
      "loss": 2.1474,
      "step": 7185
    },
    {
      "epoch": 0.21200058970956803,
      "grad_norm": 14.875,
      "learning_rate": 1.92456241876624e-06,
      "loss": 2.2473,
      "step": 7190
    },
    {
      "epoch": 0.21214801710157746,
      "grad_norm": 13.75,
      "learning_rate": 1.9243662076569034e-06,
      "loss": 2.3273,
      "step": 7195
    },
    {
      "epoch": 0.21229544449358692,
      "grad_norm": 26.5,
      "learning_rate": 1.924169751737443e-06,
      "loss": 2.253,
      "step": 7200
    },
    {
      "epoch": 0.21244287188559635,
      "grad_norm": 14.6875,
      "learning_rate": 1.9239730510598906e-06,
      "loss": 2.1251,
      "step": 7205
    },
    {
      "epoch": 0.21259029927760578,
      "grad_norm": 14.25,
      "learning_rate": 1.923776105676339e-06,
      "loss": 2.1633,
      "step": 7210
    },
    {
      "epoch": 0.2127377266696152,
      "grad_norm": 14.875,
      "learning_rate": 1.923578915638948e-06,
      "loss": 2.2504,
      "step": 7215
    },
    {
      "epoch": 0.21288515406162464,
      "grad_norm": 14.375,
      "learning_rate": 1.9233814809999417e-06,
      "loss": 2.0792,
      "step": 7220
    },
    {
      "epoch": 0.21303258145363407,
      "grad_norm": 13.25,
      "learning_rate": 1.9231838018116084e-06,
      "loss": 2.0842,
      "step": 7225
    },
    {
      "epoch": 0.21318000884564353,
      "grad_norm": 15.875,
      "learning_rate": 1.922985878126302e-06,
      "loss": 2.373,
      "step": 7230
    },
    {
      "epoch": 0.21332743623765296,
      "grad_norm": 14.0,
      "learning_rate": 1.9227877099964413e-06,
      "loss": 2.2181,
      "step": 7235
    },
    {
      "epoch": 0.2134748636296624,
      "grad_norm": 14.5625,
      "learning_rate": 1.9225892974745083e-06,
      "loss": 2.2031,
      "step": 7240
    },
    {
      "epoch": 0.21362229102167182,
      "grad_norm": 13.6875,
      "learning_rate": 1.9223906406130515e-06,
      "loss": 2.0738,
      "step": 7245
    },
    {
      "epoch": 0.21376971841368125,
      "grad_norm": 13.375,
      "learning_rate": 1.9221917394646833e-06,
      "loss": 2.1385,
      "step": 7250
    },
    {
      "epoch": 0.21391714580569068,
      "grad_norm": 13.875,
      "learning_rate": 1.9219925940820813e-06,
      "loss": 2.2051,
      "step": 7255
    },
    {
      "epoch": 0.21406457319770014,
      "grad_norm": 13.875,
      "learning_rate": 1.9217932045179864e-06,
      "loss": 2.206,
      "step": 7260
    },
    {
      "epoch": 0.21421200058970957,
      "grad_norm": 13.3125,
      "learning_rate": 1.921593570825206e-06,
      "loss": 2.2147,
      "step": 7265
    },
    {
      "epoch": 0.214359427981719,
      "grad_norm": 16.125,
      "learning_rate": 1.921393693056611e-06,
      "loss": 2.1987,
      "step": 7270
    },
    {
      "epoch": 0.21450685537372843,
      "grad_norm": 19.375,
      "learning_rate": 1.921193571265137e-06,
      "loss": 2.268,
      "step": 7275
    },
    {
      "epoch": 0.21465428276573786,
      "grad_norm": 13.875,
      "learning_rate": 1.9209932055037844e-06,
      "loss": 2.2881,
      "step": 7280
    },
    {
      "epoch": 0.21480171015774732,
      "grad_norm": 16.125,
      "learning_rate": 1.920792595825619e-06,
      "loss": 2.1479,
      "step": 7285
    },
    {
      "epoch": 0.21494913754975675,
      "grad_norm": 16.0,
      "learning_rate": 1.92059174228377e-06,
      "loss": 2.2282,
      "step": 7290
    },
    {
      "epoch": 0.21509656494176618,
      "grad_norm": 14.5,
      "learning_rate": 1.9203906449314315e-06,
      "loss": 2.1761,
      "step": 7295
    },
    {
      "epoch": 0.2152439923337756,
      "grad_norm": 15.9375,
      "learning_rate": 1.920189303821862e-06,
      "loss": 2.2016,
      "step": 7300
    },
    {
      "epoch": 0.21539141972578504,
      "grad_norm": 17.25,
      "learning_rate": 1.9199877190083863e-06,
      "loss": 2.0975,
      "step": 7305
    },
    {
      "epoch": 0.21553884711779447,
      "grad_norm": 16.375,
      "learning_rate": 1.9197858905443916e-06,
      "loss": 2.2305,
      "step": 7310
    },
    {
      "epoch": 0.21568627450980393,
      "grad_norm": 13.1875,
      "learning_rate": 1.91958381848333e-06,
      "loss": 2.1979,
      "step": 7315
    },
    {
      "epoch": 0.21583370190181336,
      "grad_norm": 15.6875,
      "learning_rate": 1.919381502878718e-06,
      "loss": 2.2607,
      "step": 7320
    },
    {
      "epoch": 0.2159811292938228,
      "grad_norm": 11.9375,
      "learning_rate": 1.9191789437841384e-06,
      "loss": 2.1914,
      "step": 7325
    },
    {
      "epoch": 0.21612855668583222,
      "grad_norm": 15.1875,
      "learning_rate": 1.9189761412532365e-06,
      "loss": 2.2348,
      "step": 7330
    },
    {
      "epoch": 0.21627598407784165,
      "grad_norm": 14.0,
      "learning_rate": 1.9187730953397225e-06,
      "loss": 2.2635,
      "step": 7335
    },
    {
      "epoch": 0.21642341146985108,
      "grad_norm": 14.1875,
      "learning_rate": 1.918569806097372e-06,
      "loss": 2.2494,
      "step": 7340
    },
    {
      "epoch": 0.21657083886186054,
      "grad_norm": 16.375,
      "learning_rate": 1.9183662735800237e-06,
      "loss": 2.2639,
      "step": 7345
    },
    {
      "epoch": 0.21671826625386997,
      "grad_norm": 16.125,
      "learning_rate": 1.9181624978415814e-06,
      "loss": 2.2761,
      "step": 7350
    },
    {
      "epoch": 0.2168656936458794,
      "grad_norm": 15.5625,
      "learning_rate": 1.917958478936014e-06,
      "loss": 2.3698,
      "step": 7355
    },
    {
      "epoch": 0.21701312103788883,
      "grad_norm": 15.5,
      "learning_rate": 1.917754216917353e-06,
      "loss": 2.2325,
      "step": 7360
    },
    {
      "epoch": 0.21716054842989826,
      "grad_norm": 19.375,
      "learning_rate": 1.917549711839696e-06,
      "loss": 2.2518,
      "step": 7365
    },
    {
      "epoch": 0.21730797582190772,
      "grad_norm": 14.75,
      "learning_rate": 1.9173449637572042e-06,
      "loss": 2.2386,
      "step": 7370
    },
    {
      "epoch": 0.21745540321391715,
      "grad_norm": 13.25,
      "learning_rate": 1.9171399727241035e-06,
      "loss": 2.2323,
      "step": 7375
    },
    {
      "epoch": 0.21760283060592658,
      "grad_norm": 15.6875,
      "learning_rate": 1.9169347387946836e-06,
      "loss": 2.207,
      "step": 7380
    },
    {
      "epoch": 0.21775025799793601,
      "grad_norm": 19.0,
      "learning_rate": 1.916729262023299e-06,
      "loss": 2.2985,
      "step": 7385
    },
    {
      "epoch": 0.21789768538994544,
      "grad_norm": 15.875,
      "learning_rate": 1.916523542464369e-06,
      "loss": 2.156,
      "step": 7390
    },
    {
      "epoch": 0.21804511278195488,
      "grad_norm": 13.3125,
      "learning_rate": 1.916317580172376e-06,
      "loss": 2.1725,
      "step": 7395
    },
    {
      "epoch": 0.21819254017396433,
      "grad_norm": 14.0625,
      "learning_rate": 1.9161113752018666e-06,
      "loss": 2.2422,
      "step": 7400
    },
    {
      "epoch": 0.21833996756597376,
      "grad_norm": 15.4375,
      "learning_rate": 1.915904927607454e-06,
      "loss": 2.2404,
      "step": 7405
    },
    {
      "epoch": 0.2184873949579832,
      "grad_norm": 14.5625,
      "learning_rate": 1.915698237443812e-06,
      "loss": 2.3074,
      "step": 7410
    },
    {
      "epoch": 0.21863482234999262,
      "grad_norm": 15.25,
      "learning_rate": 1.915491304765682e-06,
      "loss": 2.1644,
      "step": 7415
    },
    {
      "epoch": 0.21878224974200206,
      "grad_norm": 20.625,
      "learning_rate": 1.9152841296278687e-06,
      "loss": 2.2966,
      "step": 7420
    },
    {
      "epoch": 0.21892967713401149,
      "grad_norm": 13.4375,
      "learning_rate": 1.915076712085239e-06,
      "loss": 2.223,
      "step": 7425
    },
    {
      "epoch": 0.21907710452602094,
      "grad_norm": 17.625,
      "learning_rate": 1.9148690521927267e-06,
      "loss": 2.2195,
      "step": 7430
    },
    {
      "epoch": 0.21922453191803037,
      "grad_norm": 15.0,
      "learning_rate": 1.9146611500053282e-06,
      "loss": 2.2043,
      "step": 7435
    },
    {
      "epoch": 0.2193719593100398,
      "grad_norm": 15.0,
      "learning_rate": 1.914453005578105e-06,
      "loss": 2.1418,
      "step": 7440
    },
    {
      "epoch": 0.21951938670204924,
      "grad_norm": 18.0,
      "learning_rate": 1.9142446189661818e-06,
      "loss": 2.1438,
      "step": 7445
    },
    {
      "epoch": 0.21966681409405867,
      "grad_norm": 15.25,
      "learning_rate": 1.9140359902247485e-06,
      "loss": 2.3102,
      "step": 7450
    },
    {
      "epoch": 0.21981424148606812,
      "grad_norm": 13.75,
      "learning_rate": 1.9138271194090576e-06,
      "loss": 2.1613,
      "step": 7455
    },
    {
      "epoch": 0.21996166887807755,
      "grad_norm": 16.5,
      "learning_rate": 1.9136180065744278e-06,
      "loss": 2.2573,
      "step": 7460
    },
    {
      "epoch": 0.22010909627008698,
      "grad_norm": 14.1875,
      "learning_rate": 1.91340865177624e-06,
      "loss": 2.1977,
      "step": 7465
    },
    {
      "epoch": 0.22025652366209642,
      "grad_norm": 13.75,
      "learning_rate": 1.91319905506994e-06,
      "loss": 2.2246,
      "step": 7470
    },
    {
      "epoch": 0.22040395105410585,
      "grad_norm": 14.125,
      "learning_rate": 1.9129892165110383e-06,
      "loss": 2.2511,
      "step": 7475
    },
    {
      "epoch": 0.22055137844611528,
      "grad_norm": 13.375,
      "learning_rate": 1.9127791361551077e-06,
      "loss": 2.1218,
      "step": 7480
    },
    {
      "epoch": 0.22069880583812473,
      "grad_norm": 14.0,
      "learning_rate": 1.912568814057787e-06,
      "loss": 2.1278,
      "step": 7485
    },
    {
      "epoch": 0.22084623323013416,
      "grad_norm": 14.875,
      "learning_rate": 1.9123582502747776e-06,
      "loss": 2.177,
      "step": 7490
    },
    {
      "epoch": 0.2209936606221436,
      "grad_norm": 18.0,
      "learning_rate": 1.9121474448618455e-06,
      "loss": 2.1431,
      "step": 7495
    },
    {
      "epoch": 0.22114108801415303,
      "grad_norm": 16.0,
      "learning_rate": 1.911936397874821e-06,
      "loss": 2.2285,
      "step": 7500
    },
    {
      "epoch": 0.22114108801415303,
      "eval_loss": 2.1763086318969727,
      "eval_runtime": 4.7087,
      "eval_samples_per_second": 84.099,
      "eval_steps_per_second": 2.761,
      "step": 7500
    },
    {
      "epoch": 0.22128851540616246,
      "grad_norm": 15.6875,
      "learning_rate": 1.911725109369598e-06,
      "loss": 2.2166,
      "step": 7505
    },
    {
      "epoch": 0.22143594279817191,
      "grad_norm": 27.875,
      "learning_rate": 1.9115135794021336e-06,
      "loss": 2.2402,
      "step": 7510
    },
    {
      "epoch": 0.22158337019018134,
      "grad_norm": 14.4375,
      "learning_rate": 1.91130180802845e-06,
      "loss": 2.204,
      "step": 7515
    },
    {
      "epoch": 0.22173079758219078,
      "grad_norm": 14.1875,
      "learning_rate": 1.911089795304634e-06,
      "loss": 2.1249,
      "step": 7520
    },
    {
      "epoch": 0.2218782249742002,
      "grad_norm": 13.8125,
      "learning_rate": 1.9108775412868333e-06,
      "loss": 2.1547,
      "step": 7525
    },
    {
      "epoch": 0.22202565236620964,
      "grad_norm": 15.5625,
      "learning_rate": 1.9106650460312634e-06,
      "loss": 2.2753,
      "step": 7530
    },
    {
      "epoch": 0.22217307975821907,
      "grad_norm": 14.75,
      "learning_rate": 1.9104523095942003e-06,
      "loss": 2.1348,
      "step": 7535
    },
    {
      "epoch": 0.22232050715022852,
      "grad_norm": 15.25,
      "learning_rate": 1.910239332031986e-06,
      "loss": 2.2014,
      "step": 7540
    },
    {
      "epoch": 0.22246793454223796,
      "grad_norm": 15.4375,
      "learning_rate": 1.910026113401026e-06,
      "loss": 2.2529,
      "step": 7545
    },
    {
      "epoch": 0.22261536193424739,
      "grad_norm": 25.5,
      "learning_rate": 1.909812653757789e-06,
      "loss": 2.2115,
      "step": 7550
    },
    {
      "epoch": 0.22276278932625682,
      "grad_norm": 14.9375,
      "learning_rate": 1.909598953158808e-06,
      "loss": 2.228,
      "step": 7555
    },
    {
      "epoch": 0.22291021671826625,
      "grad_norm": 14.3125,
      "learning_rate": 1.909385011660679e-06,
      "loss": 2.2131,
      "step": 7560
    },
    {
      "epoch": 0.22305764411027568,
      "grad_norm": 14.25,
      "learning_rate": 1.9091708293200635e-06,
      "loss": 2.1549,
      "step": 7565
    },
    {
      "epoch": 0.22320507150228514,
      "grad_norm": 13.875,
      "learning_rate": 1.908956406193685e-06,
      "loss": 2.1435,
      "step": 7570
    },
    {
      "epoch": 0.22335249889429457,
      "grad_norm": 15.0625,
      "learning_rate": 1.908741742338332e-06,
      "loss": 2.2743,
      "step": 7575
    },
    {
      "epoch": 0.223499926286304,
      "grad_norm": 23.625,
      "learning_rate": 1.908526837810857e-06,
      "loss": 2.2482,
      "step": 7580
    },
    {
      "epoch": 0.22364735367831343,
      "grad_norm": 15.375,
      "learning_rate": 1.9083116926681735e-06,
      "loss": 2.348,
      "step": 7585
    },
    {
      "epoch": 0.22379478107032286,
      "grad_norm": 15.625,
      "learning_rate": 1.908096306967263e-06,
      "loss": 2.1831,
      "step": 7590
    },
    {
      "epoch": 0.22394220846233232,
      "grad_norm": 14.0,
      "learning_rate": 1.907880680765167e-06,
      "loss": 2.2471,
      "step": 7595
    },
    {
      "epoch": 0.22408963585434175,
      "grad_norm": 15.5,
      "learning_rate": 1.9076648141189925e-06,
      "loss": 2.1452,
      "step": 7600
    },
    {
      "epoch": 0.22423706324635118,
      "grad_norm": 14.375,
      "learning_rate": 1.9074487070859102e-06,
      "loss": 2.2139,
      "step": 7605
    },
    {
      "epoch": 0.2243844906383606,
      "grad_norm": 13.375,
      "learning_rate": 1.907232359723154e-06,
      "loss": 2.0523,
      "step": 7610
    },
    {
      "epoch": 0.22453191803037004,
      "grad_norm": 14.1875,
      "learning_rate": 1.9070157720880213e-06,
      "loss": 2.2727,
      "step": 7615
    },
    {
      "epoch": 0.22467934542237947,
      "grad_norm": 15.625,
      "learning_rate": 1.9067989442378738e-06,
      "loss": 2.2582,
      "step": 7620
    },
    {
      "epoch": 0.22482677281438893,
      "grad_norm": 14.8125,
      "learning_rate": 1.9065818762301363e-06,
      "loss": 2.1734,
      "step": 7625
    },
    {
      "epoch": 0.22497420020639836,
      "grad_norm": 14.4375,
      "learning_rate": 1.906364568122297e-06,
      "loss": 2.1759,
      "step": 7630
    },
    {
      "epoch": 0.2251216275984078,
      "grad_norm": 14.625,
      "learning_rate": 1.9061470199719083e-06,
      "loss": 2.1728,
      "step": 7635
    },
    {
      "epoch": 0.22526905499041722,
      "grad_norm": 16.625,
      "learning_rate": 1.9059292318365855e-06,
      "loss": 2.3187,
      "step": 7640
    },
    {
      "epoch": 0.22541648238242665,
      "grad_norm": 15.0,
      "learning_rate": 1.9057112037740084e-06,
      "loss": 2.1401,
      "step": 7645
    },
    {
      "epoch": 0.22556390977443608,
      "grad_norm": 15.0,
      "learning_rate": 1.9054929358419195e-06,
      "loss": 2.1785,
      "step": 7650
    },
    {
      "epoch": 0.22571133716644554,
      "grad_norm": 15.875,
      "learning_rate": 1.9052744280981251e-06,
      "loss": 2.3024,
      "step": 7655
    },
    {
      "epoch": 0.22585876455845497,
      "grad_norm": 15.3125,
      "learning_rate": 1.9050556806004955e-06,
      "loss": 2.2835,
      "step": 7660
    },
    {
      "epoch": 0.2260061919504644,
      "grad_norm": 14.4375,
      "learning_rate": 1.904836693406963e-06,
      "loss": 2.2379,
      "step": 7665
    },
    {
      "epoch": 0.22615361934247383,
      "grad_norm": 13.5,
      "learning_rate": 1.9046174665755252e-06,
      "loss": 2.1806,
      "step": 7670
    },
    {
      "epoch": 0.22630104673448326,
      "grad_norm": 15.0,
      "learning_rate": 1.904398000164242e-06,
      "loss": 2.2357,
      "step": 7675
    },
    {
      "epoch": 0.22644847412649272,
      "grad_norm": 19.125,
      "learning_rate": 1.9041782942312374e-06,
      "loss": 2.1981,
      "step": 7680
    },
    {
      "epoch": 0.22659590151850215,
      "grad_norm": 15.0,
      "learning_rate": 1.9039583488346987e-06,
      "loss": 2.2131,
      "step": 7685
    },
    {
      "epoch": 0.22674332891051158,
      "grad_norm": 15.375,
      "learning_rate": 1.9037381640328757e-06,
      "loss": 2.1719,
      "step": 7690
    },
    {
      "epoch": 0.226890756302521,
      "grad_norm": 15.0625,
      "learning_rate": 1.9035177398840832e-06,
      "loss": 2.308,
      "step": 7695
    },
    {
      "epoch": 0.22703818369453044,
      "grad_norm": 10.875,
      "learning_rate": 1.9032970764466981e-06,
      "loss": 2.139,
      "step": 7700
    },
    {
      "epoch": 0.22718561108653987,
      "grad_norm": 16.625,
      "learning_rate": 1.9030761737791612e-06,
      "loss": 2.1771,
      "step": 7705
    },
    {
      "epoch": 0.22733303847854933,
      "grad_norm": 16.0,
      "learning_rate": 1.9028550319399765e-06,
      "loss": 2.1578,
      "step": 7710
    },
    {
      "epoch": 0.22748046587055876,
      "grad_norm": 15.9375,
      "learning_rate": 1.9026336509877119e-06,
      "loss": 2.174,
      "step": 7715
    },
    {
      "epoch": 0.2276278932625682,
      "grad_norm": 13.625,
      "learning_rate": 1.9024120309809978e-06,
      "loss": 2.1622,
      "step": 7720
    },
    {
      "epoch": 0.22777532065457762,
      "grad_norm": 19.375,
      "learning_rate": 1.9021901719785282e-06,
      "loss": 2.2129,
      "step": 7725
    },
    {
      "epoch": 0.22792274804658705,
      "grad_norm": 17.0,
      "learning_rate": 1.9019680740390607e-06,
      "loss": 2.2368,
      "step": 7730
    },
    {
      "epoch": 0.22807017543859648,
      "grad_norm": 13.25,
      "learning_rate": 1.9017457372214158e-06,
      "loss": 2.1988,
      "step": 7735
    },
    {
      "epoch": 0.22821760283060594,
      "grad_norm": 12.625,
      "learning_rate": 1.9015231615844773e-06,
      "loss": 2.1546,
      "step": 7740
    },
    {
      "epoch": 0.22836503022261537,
      "grad_norm": 16.5,
      "learning_rate": 1.9013003471871932e-06,
      "loss": 2.2133,
      "step": 7745
    },
    {
      "epoch": 0.2285124576146248,
      "grad_norm": 13.875,
      "learning_rate": 1.9010772940885727e-06,
      "loss": 2.1367,
      "step": 7750
    },
    {
      "epoch": 0.22865988500663423,
      "grad_norm": 14.3125,
      "learning_rate": 1.90085400234769e-06,
      "loss": 2.2353,
      "step": 7755
    },
    {
      "epoch": 0.22880731239864366,
      "grad_norm": 14.1875,
      "learning_rate": 1.9006304720236818e-06,
      "loss": 2.1319,
      "step": 7760
    },
    {
      "epoch": 0.22895473979065312,
      "grad_norm": 15.625,
      "learning_rate": 1.9004067031757484e-06,
      "loss": 2.1314,
      "step": 7765
    },
    {
      "epoch": 0.22910216718266255,
      "grad_norm": 18.5,
      "learning_rate": 1.9001826958631529e-06,
      "loss": 2.1489,
      "step": 7770
    },
    {
      "epoch": 0.22924959457467198,
      "grad_norm": 15.5,
      "learning_rate": 1.8999584501452213e-06,
      "loss": 2.2059,
      "step": 7775
    },
    {
      "epoch": 0.2293970219666814,
      "grad_norm": 13.8125,
      "learning_rate": 1.8997339660813433e-06,
      "loss": 2.0898,
      "step": 7780
    },
    {
      "epoch": 0.22954444935869084,
      "grad_norm": 13.375,
      "learning_rate": 1.8995092437309714e-06,
      "loss": 2.1954,
      "step": 7785
    },
    {
      "epoch": 0.22969187675070027,
      "grad_norm": 15.1875,
      "learning_rate": 1.8992842831536215e-06,
      "loss": 2.2219,
      "step": 7790
    },
    {
      "epoch": 0.22983930414270973,
      "grad_norm": 20.5,
      "learning_rate": 1.8990590844088723e-06,
      "loss": 2.333,
      "step": 7795
    },
    {
      "epoch": 0.22998673153471916,
      "grad_norm": 31.5,
      "learning_rate": 1.8988336475563654e-06,
      "loss": 2.0667,
      "step": 7800
    },
    {
      "epoch": 0.2301341589267286,
      "grad_norm": 14.8125,
      "learning_rate": 1.8986079726558064e-06,
      "loss": 2.2898,
      "step": 7805
    },
    {
      "epoch": 0.23028158631873802,
      "grad_norm": 15.6875,
      "learning_rate": 1.8983820597669626e-06,
      "loss": 2.2644,
      "step": 7810
    },
    {
      "epoch": 0.23042901371074745,
      "grad_norm": 13.625,
      "learning_rate": 1.8981559089496652e-06,
      "loss": 2.2789,
      "step": 7815
    },
    {
      "epoch": 0.23057644110275688,
      "grad_norm": 14.0625,
      "learning_rate": 1.8979295202638086e-06,
      "loss": 2.1742,
      "step": 7820
    },
    {
      "epoch": 0.23072386849476634,
      "grad_norm": 13.9375,
      "learning_rate": 1.8977028937693493e-06,
      "loss": 2.2069,
      "step": 7825
    },
    {
      "epoch": 0.23087129588677577,
      "grad_norm": 15.625,
      "learning_rate": 1.8974760295263075e-06,
      "loss": 2.1841,
      "step": 7830
    },
    {
      "epoch": 0.2310187232787852,
      "grad_norm": 17.875,
      "learning_rate": 1.8972489275947662e-06,
      "loss": 2.2387,
      "step": 7835
    },
    {
      "epoch": 0.23116615067079463,
      "grad_norm": 13.8125,
      "learning_rate": 1.8970215880348713e-06,
      "loss": 2.2129,
      "step": 7840
    },
    {
      "epoch": 0.23131357806280406,
      "grad_norm": 17.75,
      "learning_rate": 1.8967940109068316e-06,
      "loss": 2.1558,
      "step": 7845
    },
    {
      "epoch": 0.23146100545481352,
      "grad_norm": 14.625,
      "learning_rate": 1.896566196270919e-06,
      "loss": 2.2171,
      "step": 7850
    },
    {
      "epoch": 0.23160843284682295,
      "grad_norm": 15.875,
      "learning_rate": 1.896338144187468e-06,
      "loss": 2.2006,
      "step": 7855
    },
    {
      "epoch": 0.23175586023883238,
      "grad_norm": 16.125,
      "learning_rate": 1.896109854716876e-06,
      "loss": 2.1893,
      "step": 7860
    },
    {
      "epoch": 0.2319032876308418,
      "grad_norm": 16.125,
      "learning_rate": 1.895881327919604e-06,
      "loss": 2.1611,
      "step": 7865
    },
    {
      "epoch": 0.23205071502285124,
      "grad_norm": 15.3125,
      "learning_rate": 1.8956525638561749e-06,
      "loss": 2.1714,
      "step": 7870
    },
    {
      "epoch": 0.23219814241486067,
      "grad_norm": 16.25,
      "learning_rate": 1.8954235625871747e-06,
      "loss": 2.0627,
      "step": 7875
    },
    {
      "epoch": 0.23234556980687013,
      "grad_norm": 16.625,
      "learning_rate": 1.8951943241732526e-06,
      "loss": 2.1204,
      "step": 7880
    },
    {
      "epoch": 0.23249299719887956,
      "grad_norm": 13.5,
      "learning_rate": 1.8949648486751202e-06,
      "loss": 2.2194,
      "step": 7885
    },
    {
      "epoch": 0.232640424590889,
      "grad_norm": 13.5,
      "learning_rate": 1.8947351361535523e-06,
      "loss": 2.1443,
      "step": 7890
    },
    {
      "epoch": 0.23278785198289842,
      "grad_norm": 13.875,
      "learning_rate": 1.8945051866693856e-06,
      "loss": 2.244,
      "step": 7895
    },
    {
      "epoch": 0.23293527937490785,
      "grad_norm": 15.9375,
      "learning_rate": 1.8942750002835205e-06,
      "loss": 2.1937,
      "step": 7900
    },
    {
      "epoch": 0.23308270676691728,
      "grad_norm": 17.375,
      "learning_rate": 1.89404457705692e-06,
      "loss": 2.3146,
      "step": 7905
    },
    {
      "epoch": 0.23323013415892674,
      "grad_norm": 15.0625,
      "learning_rate": 1.8938139170506095e-06,
      "loss": 2.1182,
      "step": 7910
    },
    {
      "epoch": 0.23337756155093617,
      "grad_norm": 16.625,
      "learning_rate": 1.8935830203256772e-06,
      "loss": 2.272,
      "step": 7915
    },
    {
      "epoch": 0.2335249889429456,
      "grad_norm": 16.25,
      "learning_rate": 1.893351886943274e-06,
      "loss": 2.1908,
      "step": 7920
    },
    {
      "epoch": 0.23367241633495503,
      "grad_norm": 15.4375,
      "learning_rate": 1.8931205169646136e-06,
      "loss": 2.2338,
      "step": 7925
    },
    {
      "epoch": 0.23381984372696446,
      "grad_norm": 12.0625,
      "learning_rate": 1.8928889104509721e-06,
      "loss": 2.1454,
      "step": 7930
    },
    {
      "epoch": 0.23396727111897392,
      "grad_norm": 14.5625,
      "learning_rate": 1.8926570674636888e-06,
      "loss": 2.2695,
      "step": 7935
    },
    {
      "epoch": 0.23411469851098335,
      "grad_norm": 22.25,
      "learning_rate": 1.8924249880641647e-06,
      "loss": 2.1593,
      "step": 7940
    },
    {
      "epoch": 0.23426212590299278,
      "grad_norm": 14.125,
      "learning_rate": 1.8921926723138644e-06,
      "loss": 2.1392,
      "step": 7945
    },
    {
      "epoch": 0.2344095532950022,
      "grad_norm": 22.875,
      "learning_rate": 1.8919601202743146e-06,
      "loss": 2.1933,
      "step": 7950
    },
    {
      "epoch": 0.23455698068701164,
      "grad_norm": 14.8125,
      "learning_rate": 1.8917273320071044e-06,
      "loss": 2.1975,
      "step": 7955
    },
    {
      "epoch": 0.23470440807902107,
      "grad_norm": 16.0,
      "learning_rate": 1.8914943075738856e-06,
      "loss": 2.1972,
      "step": 7960
    },
    {
      "epoch": 0.23485183547103053,
      "grad_norm": 15.75,
      "learning_rate": 1.891261047036373e-06,
      "loss": 2.2886,
      "step": 7965
    },
    {
      "epoch": 0.23499926286303996,
      "grad_norm": 14.5625,
      "learning_rate": 1.8910275504563434e-06,
      "loss": 2.2055,
      "step": 7970
    },
    {
      "epoch": 0.2351466902550494,
      "grad_norm": 16.625,
      "learning_rate": 1.890793817895636e-06,
      "loss": 2.1969,
      "step": 7975
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 16.0,
      "learning_rate": 1.8905598494161535e-06,
      "loss": 2.1731,
      "step": 7980
    },
    {
      "epoch": 0.23544154503906825,
      "grad_norm": 14.625,
      "learning_rate": 1.8903256450798594e-06,
      "loss": 2.2552,
      "step": 7985
    },
    {
      "epoch": 0.23558897243107768,
      "grad_norm": 15.75,
      "learning_rate": 1.890091204948782e-06,
      "loss": 2.2014,
      "step": 7990
    },
    {
      "epoch": 0.23573639982308714,
      "grad_norm": 16.375,
      "learning_rate": 1.889856529085009e-06,
      "loss": 2.2227,
      "step": 7995
    },
    {
      "epoch": 0.23588382721509657,
      "grad_norm": 13.875,
      "learning_rate": 1.8896216175506932e-06,
      "loss": 2.1592,
      "step": 8000
    },
    {
      "epoch": 0.23588382721509657,
      "eval_loss": 2.1633877754211426,
      "eval_runtime": 4.717,
      "eval_samples_per_second": 83.952,
      "eval_steps_per_second": 2.756,
      "step": 8000
    },
    {
      "epoch": 0.236031254607106,
      "grad_norm": 11.625,
      "learning_rate": 1.889386470408049e-06,
      "loss": 2.1509,
      "step": 8005
    },
    {
      "epoch": 0.23617868199911543,
      "grad_norm": 13.3125,
      "learning_rate": 1.8891510877193522e-06,
      "loss": 2.2325,
      "step": 8010
    },
    {
      "epoch": 0.23632610939112486,
      "grad_norm": 14.25,
      "learning_rate": 1.8889154695469424e-06,
      "loss": 2.2458,
      "step": 8015
    },
    {
      "epoch": 0.23647353678313432,
      "grad_norm": 15.5625,
      "learning_rate": 1.888679615953221e-06,
      "loss": 2.2179,
      "step": 8020
    },
    {
      "epoch": 0.23662096417514375,
      "grad_norm": 13.5,
      "learning_rate": 1.8884435270006516e-06,
      "loss": 2.1758,
      "step": 8025
    },
    {
      "epoch": 0.23676839156715318,
      "grad_norm": 15.875,
      "learning_rate": 1.8882072027517598e-06,
      "loss": 2.1682,
      "step": 8030
    },
    {
      "epoch": 0.2369158189591626,
      "grad_norm": 14.5625,
      "learning_rate": 1.8879706432691344e-06,
      "loss": 2.1355,
      "step": 8035
    },
    {
      "epoch": 0.23706324635117204,
      "grad_norm": 18.875,
      "learning_rate": 1.8877338486154263e-06,
      "loss": 2.1882,
      "step": 8040
    },
    {
      "epoch": 0.23721067374318147,
      "grad_norm": 17.125,
      "learning_rate": 1.8874968188533482e-06,
      "loss": 2.2153,
      "step": 8045
    },
    {
      "epoch": 0.23735810113519093,
      "grad_norm": 14.625,
      "learning_rate": 1.887259554045675e-06,
      "loss": 2.1607,
      "step": 8050
    },
    {
      "epoch": 0.23750552852720036,
      "grad_norm": 15.5625,
      "learning_rate": 1.8870220542552445e-06,
      "loss": 2.1877,
      "step": 8055
    },
    {
      "epoch": 0.2376529559192098,
      "grad_norm": 53.5,
      "learning_rate": 1.8867843195449563e-06,
      "loss": 2.1403,
      "step": 8060
    },
    {
      "epoch": 0.23780038331121922,
      "grad_norm": 12.125,
      "learning_rate": 1.8865463499777724e-06,
      "loss": 2.1845,
      "step": 8065
    },
    {
      "epoch": 0.23794781070322865,
      "grad_norm": 14.75,
      "learning_rate": 1.886308145616717e-06,
      "loss": 2.1896,
      "step": 8070
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 16.25,
      "learning_rate": 1.8860697065248759e-06,
      "loss": 2.2754,
      "step": 8075
    },
    {
      "epoch": 0.23824266548724754,
      "grad_norm": 15.8125,
      "learning_rate": 1.8858310327653982e-06,
      "loss": 2.178,
      "step": 8080
    },
    {
      "epoch": 0.23839009287925697,
      "grad_norm": 14.625,
      "learning_rate": 1.885592124401494e-06,
      "loss": 2.2494,
      "step": 8085
    },
    {
      "epoch": 0.2385375202712664,
      "grad_norm": 16.0,
      "learning_rate": 1.8853529814964365e-06,
      "loss": 2.1538,
      "step": 8090
    },
    {
      "epoch": 0.23868494766327583,
      "grad_norm": 15.9375,
      "learning_rate": 1.88511360411356e-06,
      "loss": 2.2637,
      "step": 8095
    },
    {
      "epoch": 0.23883237505528526,
      "grad_norm": 15.6875,
      "learning_rate": 1.884873992316262e-06,
      "loss": 2.2347,
      "step": 8100
    },
    {
      "epoch": 0.23897980244729472,
      "grad_norm": 15.8125,
      "learning_rate": 1.8846341461680013e-06,
      "loss": 2.2894,
      "step": 8105
    },
    {
      "epoch": 0.23912722983930415,
      "grad_norm": 15.375,
      "learning_rate": 1.884394065732299e-06,
      "loss": 2.2071,
      "step": 8110
    },
    {
      "epoch": 0.23927465723131358,
      "grad_norm": 16.625,
      "learning_rate": 1.8841537510727383e-06,
      "loss": 2.2701,
      "step": 8115
    },
    {
      "epoch": 0.239422084623323,
      "grad_norm": 27.75,
      "learning_rate": 1.8839132022529642e-06,
      "loss": 2.1806,
      "step": 8120
    },
    {
      "epoch": 0.23956951201533244,
      "grad_norm": 15.25,
      "learning_rate": 1.8836724193366844e-06,
      "loss": 2.1222,
      "step": 8125
    },
    {
      "epoch": 0.23971693940734187,
      "grad_norm": 21.0,
      "learning_rate": 1.8834314023876675e-06,
      "loss": 2.2144,
      "step": 8130
    },
    {
      "epoch": 0.23986436679935133,
      "grad_norm": 16.5,
      "learning_rate": 1.8831901514697452e-06,
      "loss": 2.1489,
      "step": 8135
    },
    {
      "epoch": 0.24001179419136076,
      "grad_norm": 13.75,
      "learning_rate": 1.8829486666468104e-06,
      "loss": 2.1968,
      "step": 8140
    },
    {
      "epoch": 0.2401592215833702,
      "grad_norm": 16.0,
      "learning_rate": 1.882706947982818e-06,
      "loss": 2.148,
      "step": 8145
    },
    {
      "epoch": 0.24030664897537962,
      "grad_norm": 15.625,
      "learning_rate": 1.8824649955417853e-06,
      "loss": 2.0118,
      "step": 8150
    },
    {
      "epoch": 0.24045407636738905,
      "grad_norm": 14.5,
      "learning_rate": 1.8822228093877912e-06,
      "loss": 2.2747,
      "step": 8155
    },
    {
      "epoch": 0.24060150375939848,
      "grad_norm": 15.6875,
      "learning_rate": 1.8819803895849767e-06,
      "loss": 2.2399,
      "step": 8160
    },
    {
      "epoch": 0.24074893115140794,
      "grad_norm": 16.75,
      "learning_rate": 1.881737736197544e-06,
      "loss": 2.2886,
      "step": 8165
    },
    {
      "epoch": 0.24089635854341737,
      "grad_norm": 22.625,
      "learning_rate": 1.881494849289758e-06,
      "loss": 2.1599,
      "step": 8170
    },
    {
      "epoch": 0.2410437859354268,
      "grad_norm": 13.875,
      "learning_rate": 1.8812517289259454e-06,
      "loss": 2.1261,
      "step": 8175
    },
    {
      "epoch": 0.24119121332743623,
      "grad_norm": 14.75,
      "learning_rate": 1.881008375170494e-06,
      "loss": 2.2204,
      "step": 8180
    },
    {
      "epoch": 0.24133864071944566,
      "grad_norm": 14.0,
      "learning_rate": 1.880764788087854e-06,
      "loss": 2.1926,
      "step": 8185
    },
    {
      "epoch": 0.24148606811145512,
      "grad_norm": 15.75,
      "learning_rate": 1.8805209677425374e-06,
      "loss": 2.1834,
      "step": 8190
    },
    {
      "epoch": 0.24163349550346455,
      "grad_norm": 15.125,
      "learning_rate": 1.8802769141991177e-06,
      "loss": 2.2266,
      "step": 8195
    },
    {
      "epoch": 0.24178092289547398,
      "grad_norm": 15.625,
      "learning_rate": 1.8800326275222304e-06,
      "loss": 2.3082,
      "step": 8200
    },
    {
      "epoch": 0.2419283502874834,
      "grad_norm": 15.5,
      "learning_rate": 1.8797881077765724e-06,
      "loss": 2.2098,
      "step": 8205
    },
    {
      "epoch": 0.24207577767949284,
      "grad_norm": 14.5,
      "learning_rate": 1.8795433550269028e-06,
      "loss": 2.139,
      "step": 8210
    },
    {
      "epoch": 0.24222320507150227,
      "grad_norm": 25.625,
      "learning_rate": 1.8792983693380424e-06,
      "loss": 2.2472,
      "step": 8215
    },
    {
      "epoch": 0.24237063246351173,
      "grad_norm": 15.5,
      "learning_rate": 1.879053150774873e-06,
      "loss": 2.2395,
      "step": 8220
    },
    {
      "epoch": 0.24251805985552116,
      "grad_norm": 14.25,
      "learning_rate": 1.8788076994023387e-06,
      "loss": 2.1807,
      "step": 8225
    },
    {
      "epoch": 0.2426654872475306,
      "grad_norm": 14.625,
      "learning_rate": 1.8785620152854453e-06,
      "loss": 2.2557,
      "step": 8230
    },
    {
      "epoch": 0.24281291463954002,
      "grad_norm": 14.8125,
      "learning_rate": 1.8783160984892598e-06,
      "loss": 2.1958,
      "step": 8235
    },
    {
      "epoch": 0.24296034203154945,
      "grad_norm": 14.5,
      "learning_rate": 1.878069949078911e-06,
      "loss": 2.2767,
      "step": 8240
    },
    {
      "epoch": 0.24310776942355888,
      "grad_norm": 14.25,
      "learning_rate": 1.8778235671195897e-06,
      "loss": 2.3034,
      "step": 8245
    },
    {
      "epoch": 0.24325519681556834,
      "grad_norm": 15.375,
      "learning_rate": 1.877576952676548e-06,
      "loss": 2.2439,
      "step": 8250
    },
    {
      "epoch": 0.24340262420757777,
      "grad_norm": 16.5,
      "learning_rate": 1.877330105815099e-06,
      "loss": 2.1794,
      "step": 8255
    },
    {
      "epoch": 0.2435500515995872,
      "grad_norm": 16.0,
      "learning_rate": 1.8770830266006182e-06,
      "loss": 2.212,
      "step": 8260
    },
    {
      "epoch": 0.24369747899159663,
      "grad_norm": 15.5,
      "learning_rate": 1.8768357150985424e-06,
      "loss": 2.0937,
      "step": 8265
    },
    {
      "epoch": 0.24384490638360606,
      "grad_norm": 16.375,
      "learning_rate": 1.8765881713743696e-06,
      "loss": 2.2065,
      "step": 8270
    },
    {
      "epoch": 0.24399233377561552,
      "grad_norm": 13.625,
      "learning_rate": 1.87634039549366e-06,
      "loss": 2.3326,
      "step": 8275
    },
    {
      "epoch": 0.24413976116762495,
      "grad_norm": 14.6875,
      "learning_rate": 1.876092387522034e-06,
      "loss": 2.1759,
      "step": 8280
    },
    {
      "epoch": 0.24428718855963438,
      "grad_norm": 17.0,
      "learning_rate": 1.875844147525175e-06,
      "loss": 2.2042,
      "step": 8285
    },
    {
      "epoch": 0.2444346159516438,
      "grad_norm": 25.375,
      "learning_rate": 1.875595675568827e-06,
      "loss": 2.2385,
      "step": 8290
    },
    {
      "epoch": 0.24458204334365324,
      "grad_norm": 15.125,
      "learning_rate": 1.8753469717187956e-06,
      "loss": 2.1559,
      "step": 8295
    },
    {
      "epoch": 0.24472947073566267,
      "grad_norm": 15.125,
      "learning_rate": 1.8750980360409478e-06,
      "loss": 2.1629,
      "step": 8300
    },
    {
      "epoch": 0.24487689812767213,
      "grad_norm": 13.8125,
      "learning_rate": 1.8748488686012118e-06,
      "loss": 2.136,
      "step": 8305
    },
    {
      "epoch": 0.24502432551968156,
      "grad_norm": 16.0,
      "learning_rate": 1.8745994694655775e-06,
      "loss": 2.1955,
      "step": 8310
    },
    {
      "epoch": 0.245171752911691,
      "grad_norm": 13.375,
      "learning_rate": 1.874349838700096e-06,
      "loss": 2.2115,
      "step": 8315
    },
    {
      "epoch": 0.24531918030370042,
      "grad_norm": 16.5,
      "learning_rate": 1.8740999763708798e-06,
      "loss": 2.1912,
      "step": 8320
    },
    {
      "epoch": 0.24546660769570985,
      "grad_norm": 14.4375,
      "learning_rate": 1.8738498825441025e-06,
      "loss": 2.247,
      "step": 8325
    },
    {
      "epoch": 0.24561403508771928,
      "grad_norm": 37.75,
      "learning_rate": 1.8735995572859995e-06,
      "loss": 2.2053,
      "step": 8330
    },
    {
      "epoch": 0.24576146247972874,
      "grad_norm": 15.5,
      "learning_rate": 1.8733490006628672e-06,
      "loss": 2.1562,
      "step": 8335
    },
    {
      "epoch": 0.24590888987173817,
      "grad_norm": 13.0,
      "learning_rate": 1.873098212741063e-06,
      "loss": 2.1676,
      "step": 8340
    },
    {
      "epoch": 0.2460563172637476,
      "grad_norm": 13.75,
      "learning_rate": 1.8728471935870062e-06,
      "loss": 2.1155,
      "step": 8345
    },
    {
      "epoch": 0.24620374465575703,
      "grad_norm": 11.8125,
      "learning_rate": 1.8725959432671767e-06,
      "loss": 2.1968,
      "step": 8350
    },
    {
      "epoch": 0.24635117204776646,
      "grad_norm": 14.1875,
      "learning_rate": 1.8723444618481162e-06,
      "loss": 2.2811,
      "step": 8355
    },
    {
      "epoch": 0.24649859943977592,
      "grad_norm": 14.375,
      "learning_rate": 1.872092749396427e-06,
      "loss": 2.177,
      "step": 8360
    },
    {
      "epoch": 0.24664602683178535,
      "grad_norm": 14.9375,
      "learning_rate": 1.8718408059787727e-06,
      "loss": 2.1862,
      "step": 8365
    },
    {
      "epoch": 0.24679345422379478,
      "grad_norm": 14.3125,
      "learning_rate": 1.8715886316618787e-06,
      "loss": 2.2133,
      "step": 8370
    },
    {
      "epoch": 0.24694088161580421,
      "grad_norm": 11.5625,
      "learning_rate": 1.8713362265125313e-06,
      "loss": 2.0847,
      "step": 8375
    },
    {
      "epoch": 0.24708830900781364,
      "grad_norm": 16.75,
      "learning_rate": 1.871083590597577e-06,
      "loss": 2.2498,
      "step": 8380
    },
    {
      "epoch": 0.24723573639982308,
      "grad_norm": 18.375,
      "learning_rate": 1.8708307239839248e-06,
      "loss": 2.2175,
      "step": 8385
    },
    {
      "epoch": 0.24738316379183253,
      "grad_norm": 20.5,
      "learning_rate": 1.8705776267385436e-06,
      "loss": 2.2745,
      "step": 8390
    },
    {
      "epoch": 0.24753059118384196,
      "grad_norm": 15.0625,
      "learning_rate": 1.8703242989284647e-06,
      "loss": 2.0961,
      "step": 8395
    },
    {
      "epoch": 0.2476780185758514,
      "grad_norm": 17.875,
      "learning_rate": 1.8700707406207793e-06,
      "loss": 2.2088,
      "step": 8400
    },
    {
      "epoch": 0.24782544596786082,
      "grad_norm": 14.4375,
      "learning_rate": 1.8698169518826397e-06,
      "loss": 2.1112,
      "step": 8405
    },
    {
      "epoch": 0.24797287335987026,
      "grad_norm": 16.75,
      "learning_rate": 1.86956293278126e-06,
      "loss": 2.2312,
      "step": 8410
    },
    {
      "epoch": 0.24812030075187969,
      "grad_norm": 15.8125,
      "learning_rate": 1.8693086833839148e-06,
      "loss": 2.1858,
      "step": 8415
    },
    {
      "epoch": 0.24826772814388914,
      "grad_norm": 14.75,
      "learning_rate": 1.86905420375794e-06,
      "loss": 2.2127,
      "step": 8420
    },
    {
      "epoch": 0.24841515553589857,
      "grad_norm": 14.75,
      "learning_rate": 1.868799493970732e-06,
      "loss": 2.1444,
      "step": 8425
    },
    {
      "epoch": 0.248562582927908,
      "grad_norm": 14.625,
      "learning_rate": 1.8685445540897483e-06,
      "loss": 2.2396,
      "step": 8430
    },
    {
      "epoch": 0.24871001031991744,
      "grad_norm": 31.125,
      "learning_rate": 1.8682893841825074e-06,
      "loss": 2.1866,
      "step": 8435
    },
    {
      "epoch": 0.24885743771192687,
      "grad_norm": 14.5,
      "learning_rate": 1.8680339843165891e-06,
      "loss": 2.1701,
      "step": 8440
    },
    {
      "epoch": 0.24900486510393632,
      "grad_norm": 14.5,
      "learning_rate": 1.8677783545596338e-06,
      "loss": 2.1813,
      "step": 8445
    },
    {
      "epoch": 0.24915229249594575,
      "grad_norm": 15.0625,
      "learning_rate": 1.8675224949793424e-06,
      "loss": 2.2112,
      "step": 8450
    },
    {
      "epoch": 0.24929971988795518,
      "grad_norm": 15.8125,
      "learning_rate": 1.8672664056434773e-06,
      "loss": 2.2627,
      "step": 8455
    },
    {
      "epoch": 0.24944714727996462,
      "grad_norm": 14.75,
      "learning_rate": 1.8670100866198613e-06,
      "loss": 2.2006,
      "step": 8460
    },
    {
      "epoch": 0.24959457467197405,
      "grad_norm": 15.0,
      "learning_rate": 1.866753537976378e-06,
      "loss": 2.1875,
      "step": 8465
    },
    {
      "epoch": 0.24974200206398348,
      "grad_norm": 13.625,
      "learning_rate": 1.8664967597809729e-06,
      "loss": 2.3218,
      "step": 8470
    },
    {
      "epoch": 0.24988942945599293,
      "grad_norm": 15.1875,
      "learning_rate": 1.8662397521016503e-06,
      "loss": 2.2054,
      "step": 8475
    },
    {
      "epoch": 0.25003685684800236,
      "grad_norm": 17.375,
      "learning_rate": 1.8659825150064773e-06,
      "loss": 2.2152,
      "step": 8480
    },
    {
      "epoch": 0.25018428424001177,
      "grad_norm": 16.25,
      "learning_rate": 1.86572504856358e-06,
      "loss": 2.0906,
      "step": 8485
    },
    {
      "epoch": 0.2503317116320212,
      "grad_norm": 14.875,
      "learning_rate": 1.8654673528411466e-06,
      "loss": 2.1785,
      "step": 8490
    },
    {
      "epoch": 0.2504791390240307,
      "grad_norm": 17.375,
      "learning_rate": 1.8652094279074255e-06,
      "loss": 2.3136,
      "step": 8495
    },
    {
      "epoch": 0.2506265664160401,
      "grad_norm": 14.0625,
      "learning_rate": 1.8649512738307258e-06,
      "loss": 2.1993,
      "step": 8500
    },
    {
      "epoch": 0.2506265664160401,
      "eval_loss": 2.151705265045166,
      "eval_runtime": 4.7212,
      "eval_samples_per_second": 83.877,
      "eval_steps_per_second": 2.754,
      "step": 8500
    },
    {
      "epoch": 0.25077399380804954,
      "grad_norm": 17.125,
      "learning_rate": 1.864692890679417e-06,
      "loss": 2.1964,
      "step": 8505
    },
    {
      "epoch": 0.25092142120005895,
      "grad_norm": 16.75,
      "learning_rate": 1.86443427852193e-06,
      "loss": 2.2787,
      "step": 8510
    },
    {
      "epoch": 0.2510688485920684,
      "grad_norm": 14.625,
      "learning_rate": 1.8641754374267558e-06,
      "loss": 2.1101,
      "step": 8515
    },
    {
      "epoch": 0.25121627598407786,
      "grad_norm": 14.75,
      "learning_rate": 1.863916367462446e-06,
      "loss": 2.26,
      "step": 8520
    },
    {
      "epoch": 0.25136370337608727,
      "grad_norm": 14.125,
      "learning_rate": 1.8636570686976127e-06,
      "loss": 2.1509,
      "step": 8525
    },
    {
      "epoch": 0.2515111307680967,
      "grad_norm": 16.125,
      "learning_rate": 1.8633975412009294e-06,
      "loss": 2.2034,
      "step": 8530
    },
    {
      "epoch": 0.2516585581601061,
      "grad_norm": 15.125,
      "learning_rate": 1.8631377850411293e-06,
      "loss": 2.1605,
      "step": 8535
    },
    {
      "epoch": 0.2518059855521156,
      "grad_norm": 15.75,
      "learning_rate": 1.8628778002870069e-06,
      "loss": 2.1062,
      "step": 8540
    },
    {
      "epoch": 0.25195341294412504,
      "grad_norm": 17.0,
      "learning_rate": 1.8626175870074165e-06,
      "loss": 2.1868,
      "step": 8545
    },
    {
      "epoch": 0.25210084033613445,
      "grad_norm": 17.125,
      "learning_rate": 1.862357145271273e-06,
      "loss": 2.2456,
      "step": 8550
    },
    {
      "epoch": 0.2522482677281439,
      "grad_norm": 15.5625,
      "learning_rate": 1.8620964751475525e-06,
      "loss": 2.2008,
      "step": 8555
    },
    {
      "epoch": 0.2523956951201533,
      "grad_norm": 21.625,
      "learning_rate": 1.861835576705291e-06,
      "loss": 2.1544,
      "step": 8560
    },
    {
      "epoch": 0.25254312251216277,
      "grad_norm": 14.625,
      "learning_rate": 1.8615744500135855e-06,
      "loss": 2.18,
      "step": 8565
    },
    {
      "epoch": 0.25269054990417217,
      "grad_norm": 13.5625,
      "learning_rate": 1.8613130951415924e-06,
      "loss": 2.1628,
      "step": 8570
    },
    {
      "epoch": 0.2528379772961816,
      "grad_norm": 14.625,
      "learning_rate": 1.8610515121585296e-06,
      "loss": 2.1433,
      "step": 8575
    },
    {
      "epoch": 0.2529854046881911,
      "grad_norm": 12.875,
      "learning_rate": 1.8607897011336746e-06,
      "loss": 2.1122,
      "step": 8580
    },
    {
      "epoch": 0.2531328320802005,
      "grad_norm": 14.25,
      "learning_rate": 1.8605276621363664e-06,
      "loss": 2.1748,
      "step": 8585
    },
    {
      "epoch": 0.25328025947220995,
      "grad_norm": 19.625,
      "learning_rate": 1.8602653952360032e-06,
      "loss": 2.1691,
      "step": 8590
    },
    {
      "epoch": 0.25342768686421935,
      "grad_norm": 14.8125,
      "learning_rate": 1.8600029005020442e-06,
      "loss": 2.2135,
      "step": 8595
    },
    {
      "epoch": 0.2535751142562288,
      "grad_norm": 14.5625,
      "learning_rate": 1.8597401780040088e-06,
      "loss": 2.2858,
      "step": 8600
    },
    {
      "epoch": 0.25372254164823826,
      "grad_norm": 21.0,
      "learning_rate": 1.8594772278114764e-06,
      "loss": 2.2727,
      "step": 8605
    },
    {
      "epoch": 0.25386996904024767,
      "grad_norm": 16.125,
      "learning_rate": 1.8592140499940876e-06,
      "loss": 2.1415,
      "step": 8610
    },
    {
      "epoch": 0.2540173964322571,
      "grad_norm": 12.5625,
      "learning_rate": 1.858950644621542e-06,
      "loss": 2.1434,
      "step": 8615
    },
    {
      "epoch": 0.25416482382426653,
      "grad_norm": 13.375,
      "learning_rate": 1.858687011763601e-06,
      "loss": 2.2154,
      "step": 8620
    },
    {
      "epoch": 0.254312251216276,
      "grad_norm": 15.5,
      "learning_rate": 1.8584231514900842e-06,
      "loss": 2.2386,
      "step": 8625
    },
    {
      "epoch": 0.25445967860828544,
      "grad_norm": 16.0,
      "learning_rate": 1.8581590638708739e-06,
      "loss": 2.2126,
      "step": 8630
    },
    {
      "epoch": 0.25460710600029485,
      "grad_norm": 19.125,
      "learning_rate": 1.8578947489759105e-06,
      "loss": 2.2209,
      "step": 8635
    },
    {
      "epoch": 0.2547545333923043,
      "grad_norm": 14.9375,
      "learning_rate": 1.8576302068751958e-06,
      "loss": 2.2278,
      "step": 8640
    },
    {
      "epoch": 0.2549019607843137,
      "grad_norm": 14.6875,
      "learning_rate": 1.8573654376387915e-06,
      "loss": 2.0973,
      "step": 8645
    },
    {
      "epoch": 0.25504938817632317,
      "grad_norm": 13.8125,
      "learning_rate": 1.857100441336819e-06,
      "loss": 2.1817,
      "step": 8650
    },
    {
      "epoch": 0.25519681556833257,
      "grad_norm": 17.125,
      "learning_rate": 1.8568352180394603e-06,
      "loss": 2.161,
      "step": 8655
    },
    {
      "epoch": 0.255344242960342,
      "grad_norm": 14.9375,
      "learning_rate": 1.8565697678169578e-06,
      "loss": 2.2521,
      "step": 8660
    },
    {
      "epoch": 0.2554916703523515,
      "grad_norm": 14.1875,
      "learning_rate": 1.8563040907396132e-06,
      "loss": 2.1843,
      "step": 8665
    },
    {
      "epoch": 0.2556390977443609,
      "grad_norm": 16.75,
      "learning_rate": 1.8560381868777886e-06,
      "loss": 2.1212,
      "step": 8670
    },
    {
      "epoch": 0.25578652513637035,
      "grad_norm": 15.5625,
      "learning_rate": 1.8557720563019068e-06,
      "loss": 2.2056,
      "step": 8675
    },
    {
      "epoch": 0.25593395252837975,
      "grad_norm": 12.0625,
      "learning_rate": 1.8555056990824496e-06,
      "loss": 2.0907,
      "step": 8680
    },
    {
      "epoch": 0.2560813799203892,
      "grad_norm": 16.75,
      "learning_rate": 1.8552391152899599e-06,
      "loss": 2.3035,
      "step": 8685
    },
    {
      "epoch": 0.25622880731239867,
      "grad_norm": 15.375,
      "learning_rate": 1.854972304995039e-06,
      "loss": 2.2537,
      "step": 8690
    },
    {
      "epoch": 0.25637623470440807,
      "grad_norm": 17.25,
      "learning_rate": 1.8547052682683505e-06,
      "loss": 2.2827,
      "step": 8695
    },
    {
      "epoch": 0.2565236620964175,
      "grad_norm": 14.9375,
      "learning_rate": 1.8544380051806157e-06,
      "loss": 2.1187,
      "step": 8700
    },
    {
      "epoch": 0.25667108948842693,
      "grad_norm": 14.9375,
      "learning_rate": 1.8541705158026173e-06,
      "loss": 2.1568,
      "step": 8705
    },
    {
      "epoch": 0.2568185168804364,
      "grad_norm": 15.875,
      "learning_rate": 1.8539028002051973e-06,
      "loss": 2.1486,
      "step": 8710
    },
    {
      "epoch": 0.25696594427244585,
      "grad_norm": 15.125,
      "learning_rate": 1.8536348584592578e-06,
      "loss": 2.2145,
      "step": 8715
    },
    {
      "epoch": 0.25711337166445525,
      "grad_norm": 13.0625,
      "learning_rate": 1.8533666906357607e-06,
      "loss": 2.1246,
      "step": 8720
    },
    {
      "epoch": 0.2572607990564647,
      "grad_norm": 14.8125,
      "learning_rate": 1.8530982968057285e-06,
      "loss": 2.2059,
      "step": 8725
    },
    {
      "epoch": 0.2574082264484741,
      "grad_norm": 14.125,
      "learning_rate": 1.852829677040242e-06,
      "loss": 2.2232,
      "step": 8730
    },
    {
      "epoch": 0.25755565384048357,
      "grad_norm": 14.3125,
      "learning_rate": 1.8525608314104431e-06,
      "loss": 2.2328,
      "step": 8735
    },
    {
      "epoch": 0.25770308123249297,
      "grad_norm": 15.5,
      "learning_rate": 1.8522917599875334e-06,
      "loss": 2.2464,
      "step": 8740
    },
    {
      "epoch": 0.25785050862450243,
      "grad_norm": 15.375,
      "learning_rate": 1.8520224628427736e-06,
      "loss": 2.2532,
      "step": 8745
    },
    {
      "epoch": 0.2579979360165119,
      "grad_norm": 16.125,
      "learning_rate": 1.8517529400474848e-06,
      "loss": 2.1911,
      "step": 8750
    },
    {
      "epoch": 0.2581453634085213,
      "grad_norm": 14.6875,
      "learning_rate": 1.8514831916730482e-06,
      "loss": 2.2427,
      "step": 8755
    },
    {
      "epoch": 0.25829279080053075,
      "grad_norm": 13.625,
      "learning_rate": 1.8512132177909034e-06,
      "loss": 2.1896,
      "step": 8760
    },
    {
      "epoch": 0.25844021819254015,
      "grad_norm": 13.0625,
      "learning_rate": 1.8509430184725513e-06,
      "loss": 2.0831,
      "step": 8765
    },
    {
      "epoch": 0.2585876455845496,
      "grad_norm": 12.9375,
      "learning_rate": 1.8506725937895515e-06,
      "loss": 2.1604,
      "step": 8770
    },
    {
      "epoch": 0.25873507297655907,
      "grad_norm": 28.875,
      "learning_rate": 1.8504019438135235e-06,
      "loss": 2.2493,
      "step": 8775
    },
    {
      "epoch": 0.25888250036856847,
      "grad_norm": 15.1875,
      "learning_rate": 1.8501310686161463e-06,
      "loss": 2.1755,
      "step": 8780
    },
    {
      "epoch": 0.2590299277605779,
      "grad_norm": 14.3125,
      "learning_rate": 1.8498599682691592e-06,
      "loss": 2.2261,
      "step": 8785
    },
    {
      "epoch": 0.25917735515258733,
      "grad_norm": 18.625,
      "learning_rate": 1.8495886428443605e-06,
      "loss": 2.1731,
      "step": 8790
    },
    {
      "epoch": 0.2593247825445968,
      "grad_norm": 15.125,
      "learning_rate": 1.8493170924136083e-06,
      "loss": 2.3196,
      "step": 8795
    },
    {
      "epoch": 0.25947220993660625,
      "grad_norm": 16.75,
      "learning_rate": 1.8490453170488202e-06,
      "loss": 2.2446,
      "step": 8800
    },
    {
      "epoch": 0.25961963732861565,
      "grad_norm": 14.875,
      "learning_rate": 1.8487733168219739e-06,
      "loss": 2.1648,
      "step": 8805
    },
    {
      "epoch": 0.2597670647206251,
      "grad_norm": 14.3125,
      "learning_rate": 1.8485010918051059e-06,
      "loss": 2.2215,
      "step": 8810
    },
    {
      "epoch": 0.2599144921126345,
      "grad_norm": 17.5,
      "learning_rate": 1.8482286420703125e-06,
      "loss": 2.3307,
      "step": 8815
    },
    {
      "epoch": 0.26006191950464397,
      "grad_norm": 14.375,
      "learning_rate": 1.84795596768975e-06,
      "loss": 2.1618,
      "step": 8820
    },
    {
      "epoch": 0.26020934689665337,
      "grad_norm": 14.9375,
      "learning_rate": 1.847683068735633e-06,
      "loss": 2.2691,
      "step": 8825
    },
    {
      "epoch": 0.26035677428866283,
      "grad_norm": 11.0625,
      "learning_rate": 1.8474099452802369e-06,
      "loss": 2.117,
      "step": 8830
    },
    {
      "epoch": 0.2605042016806723,
      "grad_norm": 14.3125,
      "learning_rate": 1.8471365973958962e-06,
      "loss": 2.0912,
      "step": 8835
    },
    {
      "epoch": 0.2606516290726817,
      "grad_norm": 14.8125,
      "learning_rate": 1.846863025155004e-06,
      "loss": 2.1718,
      "step": 8840
    },
    {
      "epoch": 0.26079905646469115,
      "grad_norm": 17.0,
      "learning_rate": 1.8465892286300137e-06,
      "loss": 2.0818,
      "step": 8845
    },
    {
      "epoch": 0.26094648385670055,
      "grad_norm": 16.875,
      "learning_rate": 1.8463152078934383e-06,
      "loss": 2.2438,
      "step": 8850
    },
    {
      "epoch": 0.26109391124871,
      "grad_norm": 13.9375,
      "learning_rate": 1.8460409630178493e-06,
      "loss": 2.2129,
      "step": 8855
    },
    {
      "epoch": 0.26124133864071947,
      "grad_norm": 15.3125,
      "learning_rate": 1.8457664940758782e-06,
      "loss": 2.1616,
      "step": 8860
    },
    {
      "epoch": 0.26138876603272887,
      "grad_norm": 14.5,
      "learning_rate": 1.8454918011402155e-06,
      "loss": 2.2635,
      "step": 8865
    },
    {
      "epoch": 0.26153619342473833,
      "grad_norm": 14.0625,
      "learning_rate": 1.8452168842836114e-06,
      "loss": 2.2376,
      "step": 8870
    },
    {
      "epoch": 0.26168362081674773,
      "grad_norm": 13.5,
      "learning_rate": 1.8449417435788748e-06,
      "loss": 2.1005,
      "step": 8875
    },
    {
      "epoch": 0.2618310482087572,
      "grad_norm": 17.5,
      "learning_rate": 1.8446663790988742e-06,
      "loss": 2.164,
      "step": 8880
    },
    {
      "epoch": 0.26197847560076665,
      "grad_norm": 15.5,
      "learning_rate": 1.844390790916538e-06,
      "loss": 2.1707,
      "step": 8885
    },
    {
      "epoch": 0.26212590299277605,
      "grad_norm": 20.25,
      "learning_rate": 1.844114979104853e-06,
      "loss": 2.2746,
      "step": 8890
    },
    {
      "epoch": 0.2622733303847855,
      "grad_norm": 13.8125,
      "learning_rate": 1.843838943736865e-06,
      "loss": 2.1086,
      "step": 8895
    },
    {
      "epoch": 0.2624207577767949,
      "grad_norm": 16.125,
      "learning_rate": 1.8435626848856805e-06,
      "loss": 2.3454,
      "step": 8900
    },
    {
      "epoch": 0.26256818516880437,
      "grad_norm": 12.5625,
      "learning_rate": 1.8432862026244633e-06,
      "loss": 2.188,
      "step": 8905
    },
    {
      "epoch": 0.26271561256081377,
      "grad_norm": 14.8125,
      "learning_rate": 1.8430094970264374e-06,
      "loss": 2.1443,
      "step": 8910
    },
    {
      "epoch": 0.26286303995282323,
      "grad_norm": 18.375,
      "learning_rate": 1.8427325681648861e-06,
      "loss": 2.1986,
      "step": 8915
    },
    {
      "epoch": 0.2630104673448327,
      "grad_norm": 15.5,
      "learning_rate": 1.8424554161131515e-06,
      "loss": 2.1675,
      "step": 8920
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 13.4375,
      "learning_rate": 1.8421780409446347e-06,
      "loss": 2.2025,
      "step": 8925
    },
    {
      "epoch": 0.26330532212885155,
      "grad_norm": 14.125,
      "learning_rate": 1.841900442732796e-06,
      "loss": 2.1276,
      "step": 8930
    },
    {
      "epoch": 0.26345274952086095,
      "grad_norm": 17.75,
      "learning_rate": 1.841622621551155e-06,
      "loss": 2.1602,
      "step": 8935
    },
    {
      "epoch": 0.2636001769128704,
      "grad_norm": 13.875,
      "learning_rate": 1.8413445774732901e-06,
      "loss": 2.2945,
      "step": 8940
    },
    {
      "epoch": 0.26374760430487987,
      "grad_norm": 15.375,
      "learning_rate": 1.8410663105728387e-06,
      "loss": 2.0937,
      "step": 8945
    },
    {
      "epoch": 0.26389503169688927,
      "grad_norm": 17.75,
      "learning_rate": 1.840787820923497e-06,
      "loss": 2.2,
      "step": 8950
    },
    {
      "epoch": 0.26404245908889873,
      "grad_norm": 14.625,
      "learning_rate": 1.8405091085990213e-06,
      "loss": 2.1271,
      "step": 8955
    },
    {
      "epoch": 0.26418988648090813,
      "grad_norm": 25.75,
      "learning_rate": 1.840230173673225e-06,
      "loss": 2.1916,
      "step": 8960
    },
    {
      "epoch": 0.2643373138729176,
      "grad_norm": 17.25,
      "learning_rate": 1.8399510162199824e-06,
      "loss": 2.2481,
      "step": 8965
    },
    {
      "epoch": 0.26448474126492705,
      "grad_norm": 15.8125,
      "learning_rate": 1.8396716363132255e-06,
      "loss": 2.1464,
      "step": 8970
    },
    {
      "epoch": 0.26463216865693645,
      "grad_norm": 17.0,
      "learning_rate": 1.8393920340269458e-06,
      "loss": 2.1157,
      "step": 8975
    },
    {
      "epoch": 0.2647795960489459,
      "grad_norm": 14.3125,
      "learning_rate": 1.8391122094351933e-06,
      "loss": 2.0538,
      "step": 8980
    },
    {
      "epoch": 0.2649270234409553,
      "grad_norm": 12.6875,
      "learning_rate": 1.8388321626120769e-06,
      "loss": 2.2269,
      "step": 8985
    },
    {
      "epoch": 0.26507445083296477,
      "grad_norm": 16.75,
      "learning_rate": 1.8385518936317645e-06,
      "loss": 2.1711,
      "step": 8990
    },
    {
      "epoch": 0.26522187822497423,
      "grad_norm": 13.75,
      "learning_rate": 1.838271402568483e-06,
      "loss": 2.1552,
      "step": 8995
    },
    {
      "epoch": 0.26536930561698363,
      "grad_norm": 14.3125,
      "learning_rate": 1.837990689496518e-06,
      "loss": 2.1522,
      "step": 9000
    },
    {
      "epoch": 0.26536930561698363,
      "eval_loss": 2.142622232437134,
      "eval_runtime": 4.7105,
      "eval_samples_per_second": 84.067,
      "eval_steps_per_second": 2.76,
      "step": 9000
    },
    {
      "epoch": 0.2655167330089931,
      "grad_norm": 17.5,
      "learning_rate": 1.837709754490214e-06,
      "loss": 2.1945,
      "step": 9005
    },
    {
      "epoch": 0.2656641604010025,
      "grad_norm": 16.125,
      "learning_rate": 1.8374285976239734e-06,
      "loss": 2.2451,
      "step": 9010
    },
    {
      "epoch": 0.26581158779301195,
      "grad_norm": 15.625,
      "learning_rate": 1.8371472189722589e-06,
      "loss": 2.2319,
      "step": 9015
    },
    {
      "epoch": 0.26595901518502135,
      "grad_norm": 19.25,
      "learning_rate": 1.8368656186095905e-06,
      "loss": 2.2219,
      "step": 9020
    },
    {
      "epoch": 0.2661064425770308,
      "grad_norm": 22.125,
      "learning_rate": 1.8365837966105481e-06,
      "loss": 2.1642,
      "step": 9025
    },
    {
      "epoch": 0.26625386996904027,
      "grad_norm": 15.6875,
      "learning_rate": 1.8363017530497693e-06,
      "loss": 2.2582,
      "step": 9030
    },
    {
      "epoch": 0.26640129736104967,
      "grad_norm": 14.0625,
      "learning_rate": 1.836019488001951e-06,
      "loss": 2.1031,
      "step": 9035
    },
    {
      "epoch": 0.26654872475305913,
      "grad_norm": 15.6875,
      "learning_rate": 1.8357370015418488e-06,
      "loss": 2.225,
      "step": 9040
    },
    {
      "epoch": 0.26669615214506853,
      "grad_norm": 15.0,
      "learning_rate": 1.835454293744276e-06,
      "loss": 2.254,
      "step": 9045
    },
    {
      "epoch": 0.266843579537078,
      "grad_norm": 14.9375,
      "learning_rate": 1.8351713646841059e-06,
      "loss": 2.2191,
      "step": 9050
    },
    {
      "epoch": 0.26699100692908745,
      "grad_norm": 14.6875,
      "learning_rate": 1.8348882144362694e-06,
      "loss": 2.2358,
      "step": 9055
    },
    {
      "epoch": 0.26713843432109685,
      "grad_norm": 12.8125,
      "learning_rate": 1.8346048430757566e-06,
      "loss": 2.1893,
      "step": 9060
    },
    {
      "epoch": 0.2672858617131063,
      "grad_norm": 15.9375,
      "learning_rate": 1.8343212506776155e-06,
      "loss": 2.1395,
      "step": 9065
    },
    {
      "epoch": 0.2674332891051157,
      "grad_norm": 16.375,
      "learning_rate": 1.8340374373169533e-06,
      "loss": 2.1742,
      "step": 9070
    },
    {
      "epoch": 0.26758071649712517,
      "grad_norm": 18.125,
      "learning_rate": 1.8337534030689355e-06,
      "loss": 2.1369,
      "step": 9075
    },
    {
      "epoch": 0.26772814388913463,
      "grad_norm": 22.75,
      "learning_rate": 1.8334691480087856e-06,
      "loss": 2.1244,
      "step": 9080
    },
    {
      "epoch": 0.26787557128114403,
      "grad_norm": 13.6875,
      "learning_rate": 1.8331846722117864e-06,
      "loss": 2.227,
      "step": 9085
    },
    {
      "epoch": 0.2680229986731535,
      "grad_norm": 15.75,
      "learning_rate": 1.8328999757532788e-06,
      "loss": 2.1094,
      "step": 9090
    },
    {
      "epoch": 0.2681704260651629,
      "grad_norm": 13.75,
      "learning_rate": 1.8326150587086617e-06,
      "loss": 2.1565,
      "step": 9095
    },
    {
      "epoch": 0.26831785345717235,
      "grad_norm": 14.375,
      "learning_rate": 1.832329921153393e-06,
      "loss": 2.272,
      "step": 9100
    },
    {
      "epoch": 0.26846528084918175,
      "grad_norm": 17.375,
      "learning_rate": 1.8320445631629892e-06,
      "loss": 2.2073,
      "step": 9105
    },
    {
      "epoch": 0.2686127082411912,
      "grad_norm": 10.8125,
      "learning_rate": 1.8317589848130246e-06,
      "loss": 2.052,
      "step": 9110
    },
    {
      "epoch": 0.26876013563320067,
      "grad_norm": 14.0625,
      "learning_rate": 1.831473186179132e-06,
      "loss": 2.1517,
      "step": 9115
    },
    {
      "epoch": 0.2689075630252101,
      "grad_norm": 12.75,
      "learning_rate": 1.8311871673370023e-06,
      "loss": 2.0941,
      "step": 9120
    },
    {
      "epoch": 0.26905499041721953,
      "grad_norm": 17.125,
      "learning_rate": 1.8309009283623854e-06,
      "loss": 2.2009,
      "step": 9125
    },
    {
      "epoch": 0.26920241780922893,
      "grad_norm": 16.75,
      "learning_rate": 1.8306144693310893e-06,
      "loss": 2.0812,
      "step": 9130
    },
    {
      "epoch": 0.2693498452012384,
      "grad_norm": 12.75,
      "learning_rate": 1.8303277903189798e-06,
      "loss": 2.1815,
      "step": 9135
    },
    {
      "epoch": 0.26949727259324785,
      "grad_norm": 14.3125,
      "learning_rate": 1.8300408914019813e-06,
      "loss": 2.1776,
      "step": 9140
    },
    {
      "epoch": 0.26964469998525725,
      "grad_norm": 16.125,
      "learning_rate": 1.8297537726560766e-06,
      "loss": 2.2445,
      "step": 9145
    },
    {
      "epoch": 0.2697921273772667,
      "grad_norm": 15.875,
      "learning_rate": 1.8294664341573063e-06,
      "loss": 2.2909,
      "step": 9150
    },
    {
      "epoch": 0.2699395547692761,
      "grad_norm": 16.625,
      "learning_rate": 1.8291788759817695e-06,
      "loss": 2.2258,
      "step": 9155
    },
    {
      "epoch": 0.27008698216128557,
      "grad_norm": 12.0625,
      "learning_rate": 1.8288910982056237e-06,
      "loss": 2.1185,
      "step": 9160
    },
    {
      "epoch": 0.27023440955329503,
      "grad_norm": 14.5625,
      "learning_rate": 1.8286031009050837e-06,
      "loss": 2.1964,
      "step": 9165
    },
    {
      "epoch": 0.27038183694530443,
      "grad_norm": 11.8125,
      "learning_rate": 1.8283148841564234e-06,
      "loss": 2.1557,
      "step": 9170
    },
    {
      "epoch": 0.2705292643373139,
      "grad_norm": 14.6875,
      "learning_rate": 1.8280264480359747e-06,
      "loss": 2.1727,
      "step": 9175
    },
    {
      "epoch": 0.2706766917293233,
      "grad_norm": 14.25,
      "learning_rate": 1.8277377926201268e-06,
      "loss": 2.1334,
      "step": 9180
    },
    {
      "epoch": 0.27082411912133275,
      "grad_norm": 14.9375,
      "learning_rate": 1.8274489179853273e-06,
      "loss": 2.0905,
      "step": 9185
    },
    {
      "epoch": 0.27097154651334215,
      "grad_norm": 17.5,
      "learning_rate": 1.827159824208083e-06,
      "loss": 2.2571,
      "step": 9190
    },
    {
      "epoch": 0.2711189739053516,
      "grad_norm": 15.125,
      "learning_rate": 1.826870511364957e-06,
      "loss": 2.1973,
      "step": 9195
    },
    {
      "epoch": 0.27126640129736107,
      "grad_norm": 12.4375,
      "learning_rate": 1.8265809795325713e-06,
      "loss": 2.0869,
      "step": 9200
    },
    {
      "epoch": 0.2714138286893705,
      "grad_norm": 27.0,
      "learning_rate": 1.8262912287876065e-06,
      "loss": 2.1648,
      "step": 9205
    },
    {
      "epoch": 0.27156125608137993,
      "grad_norm": 20.5,
      "learning_rate": 1.8260012592067995e-06,
      "loss": 2.1417,
      "step": 9210
    },
    {
      "epoch": 0.27170868347338933,
      "grad_norm": 16.0,
      "learning_rate": 1.825711070866947e-06,
      "loss": 2.224,
      "step": 9215
    },
    {
      "epoch": 0.2718561108653988,
      "grad_norm": 15.625,
      "learning_rate": 1.825420663844902e-06,
      "loss": 2.1834,
      "step": 9220
    },
    {
      "epoch": 0.27200353825740825,
      "grad_norm": 15.25,
      "learning_rate": 1.8251300382175767e-06,
      "loss": 2.0962,
      "step": 9225
    },
    {
      "epoch": 0.27215096564941765,
      "grad_norm": 13.8125,
      "learning_rate": 1.824839194061941e-06,
      "loss": 2.1178,
      "step": 9230
    },
    {
      "epoch": 0.2722983930414271,
      "grad_norm": 21.625,
      "learning_rate": 1.824548131455022e-06,
      "loss": 2.2477,
      "step": 9235
    },
    {
      "epoch": 0.2724458204334365,
      "grad_norm": 17.5,
      "learning_rate": 1.8242568504739046e-06,
      "loss": 2.1384,
      "step": 9240
    },
    {
      "epoch": 0.272593247825446,
      "grad_norm": 17.125,
      "learning_rate": 1.8239653511957326e-06,
      "loss": 2.2245,
      "step": 9245
    },
    {
      "epoch": 0.27274067521745543,
      "grad_norm": 15.1875,
      "learning_rate": 1.8236736336977065e-06,
      "loss": 2.2292,
      "step": 9250
    },
    {
      "epoch": 0.27288810260946483,
      "grad_norm": 16.125,
      "learning_rate": 1.8233816980570857e-06,
      "loss": 2.2241,
      "step": 9255
    },
    {
      "epoch": 0.2730355300014743,
      "grad_norm": 16.875,
      "learning_rate": 1.8230895443511861e-06,
      "loss": 2.2429,
      "step": 9260
    },
    {
      "epoch": 0.2731829573934837,
      "grad_norm": 12.25,
      "learning_rate": 1.8227971726573825e-06,
      "loss": 2.2361,
      "step": 9265
    },
    {
      "epoch": 0.27333038478549315,
      "grad_norm": 13.75,
      "learning_rate": 1.8225045830531068e-06,
      "loss": 2.2299,
      "step": 9270
    },
    {
      "epoch": 0.27347781217750256,
      "grad_norm": 11.8125,
      "learning_rate": 1.8222117756158486e-06,
      "loss": 2.1748,
      "step": 9275
    },
    {
      "epoch": 0.273625239569512,
      "grad_norm": 14.4375,
      "learning_rate": 1.8219187504231553e-06,
      "loss": 2.1427,
      "step": 9280
    },
    {
      "epoch": 0.27377266696152147,
      "grad_norm": 16.5,
      "learning_rate": 1.821625507552632e-06,
      "loss": 2.2049,
      "step": 9285
    },
    {
      "epoch": 0.2739200943535309,
      "grad_norm": 13.6875,
      "learning_rate": 1.8213320470819413e-06,
      "loss": 2.1321,
      "step": 9290
    },
    {
      "epoch": 0.27406752174554033,
      "grad_norm": 13.125,
      "learning_rate": 1.821038369088804e-06,
      "loss": 2.057,
      "step": 9295
    },
    {
      "epoch": 0.27421494913754973,
      "grad_norm": 15.0625,
      "learning_rate": 1.820744473650998e-06,
      "loss": 2.0797,
      "step": 9300
    },
    {
      "epoch": 0.2743623765295592,
      "grad_norm": 15.3125,
      "learning_rate": 1.8204503608463586e-06,
      "loss": 2.1928,
      "step": 9305
    },
    {
      "epoch": 0.27450980392156865,
      "grad_norm": 15.4375,
      "learning_rate": 1.8201560307527793e-06,
      "loss": 2.2871,
      "step": 9310
    },
    {
      "epoch": 0.27465723131357805,
      "grad_norm": 15.4375,
      "learning_rate": 1.8198614834482107e-06,
      "loss": 2.2493,
      "step": 9315
    },
    {
      "epoch": 0.2748046587055875,
      "grad_norm": 14.9375,
      "learning_rate": 1.8195667190106607e-06,
      "loss": 2.1325,
      "step": 9320
    },
    {
      "epoch": 0.2749520860975969,
      "grad_norm": 18.5,
      "learning_rate": 1.8192717375181954e-06,
      "loss": 2.2074,
      "step": 9325
    },
    {
      "epoch": 0.2750995134896064,
      "grad_norm": 14.75,
      "learning_rate": 1.8189765390489375e-06,
      "loss": 2.1769,
      "step": 9330
    },
    {
      "epoch": 0.27524694088161583,
      "grad_norm": 14.75,
      "learning_rate": 1.8186811236810686e-06,
      "loss": 2.3256,
      "step": 9335
    },
    {
      "epoch": 0.27539436827362523,
      "grad_norm": 13.875,
      "learning_rate": 1.818385491492826e-06,
      "loss": 2.1801,
      "step": 9340
    },
    {
      "epoch": 0.2755417956656347,
      "grad_norm": 15.1875,
      "learning_rate": 1.8180896425625054e-06,
      "loss": 2.1626,
      "step": 9345
    },
    {
      "epoch": 0.2756892230576441,
      "grad_norm": 17.25,
      "learning_rate": 1.81779357696846e-06,
      "loss": 2.126,
      "step": 9350
    },
    {
      "epoch": 0.27583665044965355,
      "grad_norm": 14.875,
      "learning_rate": 1.8174972947890998e-06,
      "loss": 2.0943,
      "step": 9355
    },
    {
      "epoch": 0.27598407784166296,
      "grad_norm": 18.75,
      "learning_rate": 1.8172007961028928e-06,
      "loss": 2.2155,
      "step": 9360
    },
    {
      "epoch": 0.2761315052336724,
      "grad_norm": 13.5,
      "learning_rate": 1.8169040809883639e-06,
      "loss": 2.1628,
      "step": 9365
    },
    {
      "epoch": 0.2762789326256819,
      "grad_norm": 15.0625,
      "learning_rate": 1.8166071495240952e-06,
      "loss": 2.1454,
      "step": 9370
    },
    {
      "epoch": 0.2764263600176913,
      "grad_norm": 14.125,
      "learning_rate": 1.816310001788727e-06,
      "loss": 2.1551,
      "step": 9375
    },
    {
      "epoch": 0.27657378740970073,
      "grad_norm": 14.6875,
      "learning_rate": 1.8160126378609553e-06,
      "loss": 2.1669,
      "step": 9380
    },
    {
      "epoch": 0.27672121480171014,
      "grad_norm": 15.6875,
      "learning_rate": 1.8157150578195348e-06,
      "loss": 2.1653,
      "step": 9385
    },
    {
      "epoch": 0.2768686421937196,
      "grad_norm": 16.875,
      "learning_rate": 1.8154172617432772e-06,
      "loss": 2.2091,
      "step": 9390
    },
    {
      "epoch": 0.27701606958572905,
      "grad_norm": 14.5,
      "learning_rate": 1.8151192497110505e-06,
      "loss": 2.2048,
      "step": 9395
    },
    {
      "epoch": 0.27716349697773845,
      "grad_norm": 17.25,
      "learning_rate": 1.814821021801781e-06,
      "loss": 2.2176,
      "step": 9400
    },
    {
      "epoch": 0.2773109243697479,
      "grad_norm": 14.125,
      "learning_rate": 1.8145225780944515e-06,
      "loss": 2.2279,
      "step": 9405
    },
    {
      "epoch": 0.2774583517617573,
      "grad_norm": 15.3125,
      "learning_rate": 1.8142239186681022e-06,
      "loss": 2.1464,
      "step": 9410
    },
    {
      "epoch": 0.2776057791537668,
      "grad_norm": 13.4375,
      "learning_rate": 1.8139250436018303e-06,
      "loss": 2.2263,
      "step": 9415
    },
    {
      "epoch": 0.27775320654577623,
      "grad_norm": 16.125,
      "learning_rate": 1.81362595297479e-06,
      "loss": 2.1134,
      "step": 9420
    },
    {
      "epoch": 0.27790063393778563,
      "grad_norm": 15.25,
      "learning_rate": 1.8133266468661934e-06,
      "loss": 2.148,
      "step": 9425
    },
    {
      "epoch": 0.2780480613297951,
      "grad_norm": 14.625,
      "learning_rate": 1.8130271253553084e-06,
      "loss": 2.1851,
      "step": 9430
    },
    {
      "epoch": 0.2781954887218045,
      "grad_norm": 16.25,
      "learning_rate": 1.8127273885214609e-06,
      "loss": 2.2678,
      "step": 9435
    },
    {
      "epoch": 0.27834291611381395,
      "grad_norm": 16.375,
      "learning_rate": 1.8124274364440337e-06,
      "loss": 2.1842,
      "step": 9440
    },
    {
      "epoch": 0.27849034350582336,
      "grad_norm": 14.6875,
      "learning_rate": 1.8121272692024658e-06,
      "loss": 2.2537,
      "step": 9445
    },
    {
      "epoch": 0.2786377708978328,
      "grad_norm": 14.75,
      "learning_rate": 1.8118268868762546e-06,
      "loss": 2.1291,
      "step": 9450
    },
    {
      "epoch": 0.2787851982898423,
      "grad_norm": 15.0,
      "learning_rate": 1.811526289544953e-06,
      "loss": 2.1512,
      "step": 9455
    },
    {
      "epoch": 0.2789326256818517,
      "grad_norm": 14.5625,
      "learning_rate": 1.8112254772881717e-06,
      "loss": 2.1319,
      "step": 9460
    },
    {
      "epoch": 0.27908005307386113,
      "grad_norm": 14.8125,
      "learning_rate": 1.8109244501855782e-06,
      "loss": 2.1444,
      "step": 9465
    },
    {
      "epoch": 0.27922748046587054,
      "grad_norm": 13.875,
      "learning_rate": 1.810623208316897e-06,
      "loss": 2.1868,
      "step": 9470
    },
    {
      "epoch": 0.27937490785788,
      "grad_norm": 15.0625,
      "learning_rate": 1.8103217517619094e-06,
      "loss": 2.2162,
      "step": 9475
    },
    {
      "epoch": 0.27952233524988945,
      "grad_norm": 16.125,
      "learning_rate": 1.810020080600453e-06,
      "loss": 2.2668,
      "step": 9480
    },
    {
      "epoch": 0.27966976264189886,
      "grad_norm": 16.0,
      "learning_rate": 1.809718194912423e-06,
      "loss": 2.2179,
      "step": 9485
    },
    {
      "epoch": 0.2798171900339083,
      "grad_norm": 14.875,
      "learning_rate": 1.809416094777771e-06,
      "loss": 2.1202,
      "step": 9490
    },
    {
      "epoch": 0.2799646174259177,
      "grad_norm": 10.75,
      "learning_rate": 1.8091137802765058e-06,
      "loss": 2.0973,
      "step": 9495
    },
    {
      "epoch": 0.2801120448179272,
      "grad_norm": 12.0625,
      "learning_rate": 1.8088112514886923e-06,
      "loss": 2.0952,
      "step": 9500
    },
    {
      "epoch": 0.2801120448179272,
      "eval_loss": 2.1350300312042236,
      "eval_runtime": 4.7104,
      "eval_samples_per_second": 84.069,
      "eval_steps_per_second": 2.76,
      "step": 9500
    },
    {
      "epoch": 0.28025947220993663,
      "grad_norm": 16.5,
      "learning_rate": 1.8085085084944526e-06,
      "loss": 2.1947,
      "step": 9505
    },
    {
      "epoch": 0.28040689960194604,
      "grad_norm": 16.875,
      "learning_rate": 1.808205551373966e-06,
      "loss": 2.2008,
      "step": 9510
    },
    {
      "epoch": 0.2805543269939555,
      "grad_norm": 16.125,
      "learning_rate": 1.8079023802074674e-06,
      "loss": 2.244,
      "step": 9515
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 14.9375,
      "learning_rate": 1.807598995075249e-06,
      "loss": 2.2662,
      "step": 9520
    },
    {
      "epoch": 0.28084918177797435,
      "grad_norm": 15.75,
      "learning_rate": 1.80729539605766e-06,
      "loss": 2.1337,
      "step": 9525
    },
    {
      "epoch": 0.28099660916998376,
      "grad_norm": 14.375,
      "learning_rate": 1.8069915832351057e-06,
      "loss": 2.1507,
      "step": 9530
    },
    {
      "epoch": 0.2811440365619932,
      "grad_norm": 13.25,
      "learning_rate": 1.8066875566880482e-06,
      "loss": 2.1803,
      "step": 9535
    },
    {
      "epoch": 0.2812914639540027,
      "grad_norm": 11.25,
      "learning_rate": 1.8063833164970061e-06,
      "loss": 2.0851,
      "step": 9540
    },
    {
      "epoch": 0.2814388913460121,
      "grad_norm": 13.5,
      "learning_rate": 1.8060788627425548e-06,
      "loss": 2.2345,
      "step": 9545
    },
    {
      "epoch": 0.28158631873802153,
      "grad_norm": 15.875,
      "learning_rate": 1.8057741955053261e-06,
      "loss": 2.3918,
      "step": 9550
    },
    {
      "epoch": 0.28173374613003094,
      "grad_norm": 14.6875,
      "learning_rate": 1.8054693148660088e-06,
      "loss": 2.1558,
      "step": 9555
    },
    {
      "epoch": 0.2818811735220404,
      "grad_norm": 15.0625,
      "learning_rate": 1.805164220905347e-06,
      "loss": 2.1571,
      "step": 9560
    },
    {
      "epoch": 0.28202860091404985,
      "grad_norm": 15.0,
      "learning_rate": 1.8048589137041427e-06,
      "loss": 2.0988,
      "step": 9565
    },
    {
      "epoch": 0.28217602830605926,
      "grad_norm": 15.3125,
      "learning_rate": 1.8045533933432538e-06,
      "loss": 2.1293,
      "step": 9570
    },
    {
      "epoch": 0.2823234556980687,
      "grad_norm": 24.25,
      "learning_rate": 1.8042476599035944e-06,
      "loss": 2.1763,
      "step": 9575
    },
    {
      "epoch": 0.2824708830900781,
      "grad_norm": 14.5,
      "learning_rate": 1.8039417134661354e-06,
      "loss": 2.1462,
      "step": 9580
    },
    {
      "epoch": 0.2826183104820876,
      "grad_norm": 12.125,
      "learning_rate": 1.8036355541119038e-06,
      "loss": 2.2079,
      "step": 9585
    },
    {
      "epoch": 0.28276573787409703,
      "grad_norm": 14.5,
      "learning_rate": 1.8033291819219833e-06,
      "loss": 2.1601,
      "step": 9590
    },
    {
      "epoch": 0.28291316526610644,
      "grad_norm": 14.375,
      "learning_rate": 1.803022596977514e-06,
      "loss": 2.2201,
      "step": 9595
    },
    {
      "epoch": 0.2830605926581159,
      "grad_norm": 59.25,
      "learning_rate": 1.802715799359692e-06,
      "loss": 2.1743,
      "step": 9600
    },
    {
      "epoch": 0.2832080200501253,
      "grad_norm": 15.0625,
      "learning_rate": 1.8024087891497697e-06,
      "loss": 2.2809,
      "step": 9605
    },
    {
      "epoch": 0.28335544744213476,
      "grad_norm": 18.0,
      "learning_rate": 1.8021015664290563e-06,
      "loss": 2.0981,
      "step": 9610
    },
    {
      "epoch": 0.28350287483414416,
      "grad_norm": 14.25,
      "learning_rate": 1.8017941312789172e-06,
      "loss": 2.3601,
      "step": 9615
    },
    {
      "epoch": 0.2836503022261536,
      "grad_norm": 12.8125,
      "learning_rate": 1.8014864837807732e-06,
      "loss": 2.0582,
      "step": 9620
    },
    {
      "epoch": 0.2837977296181631,
      "grad_norm": 15.4375,
      "learning_rate": 1.8011786240161025e-06,
      "loss": 2.1897,
      "step": 9625
    },
    {
      "epoch": 0.2839451570101725,
      "grad_norm": 14.9375,
      "learning_rate": 1.8008705520664388e-06,
      "loss": 2.1527,
      "step": 9630
    },
    {
      "epoch": 0.28409258440218194,
      "grad_norm": 17.25,
      "learning_rate": 1.8005622680133722e-06,
      "loss": 2.1402,
      "step": 9635
    },
    {
      "epoch": 0.28424001179419134,
      "grad_norm": 15.6875,
      "learning_rate": 1.8002537719385492e-06,
      "loss": 2.2136,
      "step": 9640
    },
    {
      "epoch": 0.2843874391862008,
      "grad_norm": 15.375,
      "learning_rate": 1.7999450639236716e-06,
      "loss": 2.2145,
      "step": 9645
    },
    {
      "epoch": 0.28453486657821025,
      "grad_norm": 14.0625,
      "learning_rate": 1.799636144050499e-06,
      "loss": 2.2849,
      "step": 9650
    },
    {
      "epoch": 0.28468229397021966,
      "grad_norm": 17.25,
      "learning_rate": 1.7993270124008454e-06,
      "loss": 2.1804,
      "step": 9655
    },
    {
      "epoch": 0.2848297213622291,
      "grad_norm": 14.0625,
      "learning_rate": 1.7990176690565815e-06,
      "loss": 2.1777,
      "step": 9660
    },
    {
      "epoch": 0.2849771487542385,
      "grad_norm": 14.5625,
      "learning_rate": 1.798708114099634e-06,
      "loss": 2.1158,
      "step": 9665
    },
    {
      "epoch": 0.285124576146248,
      "grad_norm": 14.0,
      "learning_rate": 1.7983983476119864e-06,
      "loss": 2.1636,
      "step": 9670
    },
    {
      "epoch": 0.28527200353825743,
      "grad_norm": 13.6875,
      "learning_rate": 1.798088369675677e-06,
      "loss": 2.1837,
      "step": 9675
    },
    {
      "epoch": 0.28541943093026684,
      "grad_norm": 15.25,
      "learning_rate": 1.797778180372801e-06,
      "loss": 2.2862,
      "step": 9680
    },
    {
      "epoch": 0.2855668583222763,
      "grad_norm": 17.0,
      "learning_rate": 1.7974677797855092e-06,
      "loss": 2.2452,
      "step": 9685
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 15.875,
      "learning_rate": 1.7971571679960081e-06,
      "loss": 2.1044,
      "step": 9690
    },
    {
      "epoch": 0.28586171310629516,
      "grad_norm": 13.5625,
      "learning_rate": 1.7968463450865608e-06,
      "loss": 2.1527,
      "step": 9695
    },
    {
      "epoch": 0.28600914049830456,
      "grad_norm": 16.125,
      "learning_rate": 1.796535311139486e-06,
      "loss": 2.1591,
      "step": 9700
    },
    {
      "epoch": 0.286156567890314,
      "grad_norm": 12.6875,
      "learning_rate": 1.796224066237158e-06,
      "loss": 2.2229,
      "step": 9705
    },
    {
      "epoch": 0.2863039952823235,
      "grad_norm": 16.125,
      "learning_rate": 1.7959126104620074e-06,
      "loss": 2.2106,
      "step": 9710
    },
    {
      "epoch": 0.2864514226743329,
      "grad_norm": 16.625,
      "learning_rate": 1.7956009438965204e-06,
      "loss": 2.1492,
      "step": 9715
    },
    {
      "epoch": 0.28659885006634234,
      "grad_norm": 13.9375,
      "learning_rate": 1.7952890666232391e-06,
      "loss": 2.1762,
      "step": 9720
    },
    {
      "epoch": 0.28674627745835174,
      "grad_norm": 15.0,
      "learning_rate": 1.7949769787247617e-06,
      "loss": 2.1311,
      "step": 9725
    },
    {
      "epoch": 0.2868937048503612,
      "grad_norm": 14.625,
      "learning_rate": 1.7946646802837416e-06,
      "loss": 2.0461,
      "step": 9730
    },
    {
      "epoch": 0.28704113224237066,
      "grad_norm": 18.5,
      "learning_rate": 1.7943521713828883e-06,
      "loss": 2.113,
      "step": 9735
    },
    {
      "epoch": 0.28718855963438006,
      "grad_norm": 17.25,
      "learning_rate": 1.7940394521049667e-06,
      "loss": 2.2751,
      "step": 9740
    },
    {
      "epoch": 0.2873359870263895,
      "grad_norm": 13.25,
      "learning_rate": 1.7937265225327983e-06,
      "loss": 2.1439,
      "step": 9745
    },
    {
      "epoch": 0.2874834144183989,
      "grad_norm": 18.125,
      "learning_rate": 1.793413382749259e-06,
      "loss": 2.2289,
      "step": 9750
    },
    {
      "epoch": 0.2876308418104084,
      "grad_norm": 15.4375,
      "learning_rate": 1.7931000328372818e-06,
      "loss": 2.3197,
      "step": 9755
    },
    {
      "epoch": 0.28777826920241784,
      "grad_norm": 15.0,
      "learning_rate": 1.7927864728798543e-06,
      "loss": 2.1508,
      "step": 9760
    },
    {
      "epoch": 0.28792569659442724,
      "grad_norm": 12.0,
      "learning_rate": 1.7924727029600198e-06,
      "loss": 2.1825,
      "step": 9765
    },
    {
      "epoch": 0.2880731239864367,
      "grad_norm": 13.125,
      "learning_rate": 1.7921587231608777e-06,
      "loss": 2.1113,
      "step": 9770
    },
    {
      "epoch": 0.2882205513784461,
      "grad_norm": 13.9375,
      "learning_rate": 1.791844533565583e-06,
      "loss": 2.1209,
      "step": 9775
    },
    {
      "epoch": 0.28836797877045556,
      "grad_norm": 13.6875,
      "learning_rate": 1.7915301342573455e-06,
      "loss": 2.2531,
      "step": 9780
    },
    {
      "epoch": 0.28851540616246496,
      "grad_norm": 13.25,
      "learning_rate": 1.7912155253194311e-06,
      "loss": 2.1772,
      "step": 9785
    },
    {
      "epoch": 0.2886628335544744,
      "grad_norm": 13.375,
      "learning_rate": 1.7909007068351617e-06,
      "loss": 1.9749,
      "step": 9790
    },
    {
      "epoch": 0.2888102609464839,
      "grad_norm": 14.6875,
      "learning_rate": 1.7905856788879135e-06,
      "loss": 2.2297,
      "step": 9795
    },
    {
      "epoch": 0.2889576883384933,
      "grad_norm": 21.125,
      "learning_rate": 1.7902704415611194e-06,
      "loss": 2.1577,
      "step": 9800
    },
    {
      "epoch": 0.28910511573050274,
      "grad_norm": 15.4375,
      "learning_rate": 1.7899549949382667e-06,
      "loss": 2.1224,
      "step": 9805
    },
    {
      "epoch": 0.28925254312251214,
      "grad_norm": 15.375,
      "learning_rate": 1.789639339102899e-06,
      "loss": 2.1734,
      "step": 9810
    },
    {
      "epoch": 0.2893999705145216,
      "grad_norm": 14.5625,
      "learning_rate": 1.7893234741386148e-06,
      "loss": 1.9936,
      "step": 9815
    },
    {
      "epoch": 0.28954739790653106,
      "grad_norm": 15.625,
      "learning_rate": 1.7890074001290678e-06,
      "loss": 2.1629,
      "step": 9820
    },
    {
      "epoch": 0.28969482529854046,
      "grad_norm": 15.125,
      "learning_rate": 1.7886911171579678e-06,
      "loss": 2.1329,
      "step": 9825
    },
    {
      "epoch": 0.2898422526905499,
      "grad_norm": 13.5625,
      "learning_rate": 1.7883746253090792e-06,
      "loss": 2.2002,
      "step": 9830
    },
    {
      "epoch": 0.2899896800825593,
      "grad_norm": 18.5,
      "learning_rate": 1.7880579246662225e-06,
      "loss": 2.1516,
      "step": 9835
    },
    {
      "epoch": 0.2901371074745688,
      "grad_norm": 17.0,
      "learning_rate": 1.7877410153132727e-06,
      "loss": 2.3072,
      "step": 9840
    },
    {
      "epoch": 0.29028453486657824,
      "grad_norm": 15.4375,
      "learning_rate": 1.7874238973341602e-06,
      "loss": 2.094,
      "step": 9845
    },
    {
      "epoch": 0.29043196225858764,
      "grad_norm": 14.0,
      "learning_rate": 1.7871065708128712e-06,
      "loss": 2.1111,
      "step": 9850
    },
    {
      "epoch": 0.2905793896505971,
      "grad_norm": 15.625,
      "learning_rate": 1.7867890358334465e-06,
      "loss": 2.2635,
      "step": 9855
    },
    {
      "epoch": 0.2907268170426065,
      "grad_norm": 17.625,
      "learning_rate": 1.7864712924799829e-06,
      "loss": 2.0714,
      "step": 9860
    },
    {
      "epoch": 0.29087424443461596,
      "grad_norm": 14.875,
      "learning_rate": 1.7861533408366315e-06,
      "loss": 2.1918,
      "step": 9865
    },
    {
      "epoch": 0.29102167182662536,
      "grad_norm": 15.125,
      "learning_rate": 1.7858351809875992e-06,
      "loss": 2.1891,
      "step": 9870
    },
    {
      "epoch": 0.2911690992186348,
      "grad_norm": 13.5625,
      "learning_rate": 1.7855168130171471e-06,
      "loss": 2.2214,
      "step": 9875
    },
    {
      "epoch": 0.2913165266106443,
      "grad_norm": 13.875,
      "learning_rate": 1.785198237009593e-06,
      "loss": 2.1657,
      "step": 9880
    },
    {
      "epoch": 0.2914639540026537,
      "grad_norm": 15.3125,
      "learning_rate": 1.7848794530493083e-06,
      "loss": 2.2377,
      "step": 9885
    },
    {
      "epoch": 0.29161138139466314,
      "grad_norm": 14.3125,
      "learning_rate": 1.7845604612207206e-06,
      "loss": 2.1852,
      "step": 9890
    },
    {
      "epoch": 0.29175880878667254,
      "grad_norm": 15.0625,
      "learning_rate": 1.7842412616083116e-06,
      "loss": 2.2394,
      "step": 9895
    },
    {
      "epoch": 0.291906236178682,
      "grad_norm": 16.5,
      "learning_rate": 1.7839218542966185e-06,
      "loss": 2.2163,
      "step": 9900
    },
    {
      "epoch": 0.29205366357069146,
      "grad_norm": 14.0625,
      "learning_rate": 1.7836022393702334e-06,
      "loss": 2.1405,
      "step": 9905
    },
    {
      "epoch": 0.29220109096270086,
      "grad_norm": 18.625,
      "learning_rate": 1.783282416913804e-06,
      "loss": 2.2301,
      "step": 9910
    },
    {
      "epoch": 0.2923485183547103,
      "grad_norm": 12.5625,
      "learning_rate": 1.782962387012032e-06,
      "loss": 2.0984,
      "step": 9915
    },
    {
      "epoch": 0.2924959457467197,
      "grad_norm": 14.0,
      "learning_rate": 1.7826421497496743e-06,
      "loss": 2.1405,
      "step": 9920
    },
    {
      "epoch": 0.2926433731387292,
      "grad_norm": 15.5,
      "learning_rate": 1.7823217052115434e-06,
      "loss": 2.2241,
      "step": 9925
    },
    {
      "epoch": 0.29279080053073864,
      "grad_norm": 15.4375,
      "learning_rate": 1.7820010534825057e-06,
      "loss": 2.1421,
      "step": 9930
    },
    {
      "epoch": 0.29293822792274804,
      "grad_norm": 16.625,
      "learning_rate": 1.7816801946474832e-06,
      "loss": 2.1865,
      "step": 9935
    },
    {
      "epoch": 0.2930856553147575,
      "grad_norm": 11.625,
      "learning_rate": 1.781359128791452e-06,
      "loss": 2.1026,
      "step": 9940
    },
    {
      "epoch": 0.2932330827067669,
      "grad_norm": 12.9375,
      "learning_rate": 1.7810378559994442e-06,
      "loss": 2.1224,
      "step": 9945
    },
    {
      "epoch": 0.29338051009877636,
      "grad_norm": 13.8125,
      "learning_rate": 1.7807163763565457e-06,
      "loss": 2.1026,
      "step": 9950
    },
    {
      "epoch": 0.29352793749078576,
      "grad_norm": 17.75,
      "learning_rate": 1.7803946899478972e-06,
      "loss": 2.1495,
      "step": 9955
    },
    {
      "epoch": 0.2936753648827952,
      "grad_norm": 14.5,
      "learning_rate": 1.7800727968586952e-06,
      "loss": 2.2219,
      "step": 9960
    },
    {
      "epoch": 0.2938227922748047,
      "grad_norm": 14.5,
      "learning_rate": 1.7797506971741899e-06,
      "loss": 2.1837,
      "step": 9965
    },
    {
      "epoch": 0.2939702196668141,
      "grad_norm": 12.1875,
      "learning_rate": 1.779428390979686e-06,
      "loss": 2.0926,
      "step": 9970
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 24.625,
      "learning_rate": 1.7791058783605442e-06,
      "loss": 2.192,
      "step": 9975
    },
    {
      "epoch": 0.29426507445083294,
      "grad_norm": 13.75,
      "learning_rate": 1.7787831594021787e-06,
      "loss": 2.2558,
      "step": 9980
    },
    {
      "epoch": 0.2944125018428424,
      "grad_norm": 15.125,
      "learning_rate": 1.7784602341900585e-06,
      "loss": 2.2184,
      "step": 9985
    },
    {
      "epoch": 0.29455992923485186,
      "grad_norm": 26.125,
      "learning_rate": 1.7781371028097079e-06,
      "loss": 2.2509,
      "step": 9990
    },
    {
      "epoch": 0.29470735662686126,
      "grad_norm": 14.9375,
      "learning_rate": 1.7778137653467052e-06,
      "loss": 2.2607,
      "step": 9995
    },
    {
      "epoch": 0.2948547840188707,
      "grad_norm": 13.875,
      "learning_rate": 1.7774902218866833e-06,
      "loss": 2.1507,
      "step": 10000
    },
    {
      "epoch": 0.2948547840188707,
      "eval_loss": 2.1286072731018066,
      "eval_runtime": 4.7121,
      "eval_samples_per_second": 84.038,
      "eval_steps_per_second": 2.759,
      "step": 10000
    }
  ],
  "logging_steps": 5,
  "max_steps": 33915,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.7972623782104793e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}