{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 3160,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007911392405063292,
      "grad_norm": 2.3738328422587087,
      "learning_rate": 1.5822784810126583e-06,
      "loss": 1.1261,
      "num_tokens": 5242880.0,
      "step": 5
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 1.7722440301570659,
      "learning_rate": 3.1645569620253167e-06,
      "loss": 1.1058,
      "num_tokens": 10470029.0,
      "step": 10
    },
    {
      "epoch": 0.023734177215189875,
      "grad_norm": 1.083276838182905,
      "learning_rate": 4.746835443037975e-06,
      "loss": 1.0806,
      "num_tokens": 15712909.0,
      "step": 15
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 0.8611689905530553,
      "learning_rate": 6.329113924050633e-06,
      "loss": 1.02,
      "num_tokens": 20951872.0,
      "step": 20
    },
    {
      "epoch": 0.03955696202531646,
      "grad_norm": 0.7250421722694049,
      "learning_rate": 7.911392405063292e-06,
      "loss": 0.9766,
      "num_tokens": 26185341.0,
      "step": 25
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.6382913905764741,
      "learning_rate": 9.49367088607595e-06,
      "loss": 0.9456,
      "num_tokens": 31428221.0,
      "step": 30
    },
    {
      "epoch": 0.055379746835443035,
      "grad_norm": 0.5436432996737252,
      "learning_rate": 1.1075949367088608e-05,
      "loss": 0.9187,
      "num_tokens": 36669330.0,
      "step": 35
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 0.45175884987072906,
      "learning_rate": 1.2658227848101267e-05,
      "loss": 0.8956,
      "num_tokens": 41910897.0,
      "step": 40
    },
    {
      "epoch": 0.07120253164556962,
      "grad_norm": 0.3906460936286765,
      "learning_rate": 1.4240506329113925e-05,
      "loss": 0.8787,
      "num_tokens": 47153777.0,
      "step": 45
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.38060702422443193,
      "learning_rate": 1.5822784810126583e-05,
      "loss": 0.8674,
      "num_tokens": 52391193.0,
      "step": 50
    },
    {
      "epoch": 0.08702531645569621,
      "grad_norm": 0.3976893756810291,
      "learning_rate": 1.7405063291139243e-05,
      "loss": 0.8589,
      "num_tokens": 57634073.0,
      "step": 55
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 0.39610453911616206,
      "learning_rate": 1.89873417721519e-05,
      "loss": 0.8577,
      "num_tokens": 62876953.0,
      "step": 60
    },
    {
      "epoch": 0.10284810126582279,
      "grad_norm": 0.3232899679188804,
      "learning_rate": 2.056962025316456e-05,
      "loss": 0.8456,
      "num_tokens": 68119833.0,
      "step": 65
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 0.3785345815272384,
      "learning_rate": 2.2151898734177217e-05,
      "loss": 0.8354,
      "num_tokens": 73362713.0,
      "step": 70
    },
    {
      "epoch": 0.11867088607594936,
      "grad_norm": 0.42958423811773794,
      "learning_rate": 2.3734177215189873e-05,
      "loss": 0.8277,
      "num_tokens": 78605593.0,
      "step": 75
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.3759983838903421,
      "learning_rate": 2.5316455696202533e-05,
      "loss": 0.8257,
      "num_tokens": 83848473.0,
      "step": 80
    },
    {
      "epoch": 0.13449367088607594,
      "grad_norm": 0.4694604862178893,
      "learning_rate": 2.689873417721519e-05,
      "loss": 0.8212,
      "num_tokens": 89091353.0,
      "step": 85
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.41926661815869387,
      "learning_rate": 2.848101265822785e-05,
      "loss": 0.8132,
      "num_tokens": 94334233.0,
      "step": 90
    },
    {
      "epoch": 0.15031645569620253,
      "grad_norm": 0.8644371279587081,
      "learning_rate": 3.0063291139240506e-05,
      "loss": 0.8356,
      "num_tokens": 99577113.0,
      "step": 95
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.7239620838593109,
      "learning_rate": 3.1645569620253167e-05,
      "loss": 0.8115,
      "num_tokens": 104819993.0,
      "step": 100
    },
    {
      "epoch": 0.1661392405063291,
      "grad_norm": 0.6759442627573765,
      "learning_rate": 3.322784810126582e-05,
      "loss": 0.8131,
      "num_tokens": 110062873.0,
      "step": 105
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 0.5445796635610345,
      "learning_rate": 3.4810126582278487e-05,
      "loss": 0.8051,
      "num_tokens": 115305753.0,
      "step": 110
    },
    {
      "epoch": 0.1819620253164557,
      "grad_norm": 0.6030674874140043,
      "learning_rate": 3.639240506329114e-05,
      "loss": 0.8069,
      "num_tokens": 120539230.0,
      "step": 115
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.5775646724321816,
      "learning_rate": 3.79746835443038e-05,
      "loss": 0.8001,
      "num_tokens": 125782110.0,
      "step": 120
    },
    {
      "epoch": 0.19778481012658228,
      "grad_norm": 1.1999946695600228,
      "learning_rate": 3.9556962025316456e-05,
      "loss": 0.7923,
      "num_tokens": 131014137.0,
      "step": 125
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.9396643672179282,
      "learning_rate": 4.113924050632912e-05,
      "loss": 0.8032,
      "num_tokens": 136257017.0,
      "step": 130
    },
    {
      "epoch": 0.21360759493670886,
      "grad_norm": 0.989562272615721,
      "learning_rate": 4.2721518987341776e-05,
      "loss": 0.7945,
      "num_tokens": 141499897.0,
      "step": 135
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 0.7938020654922184,
      "learning_rate": 4.430379746835443e-05,
      "loss": 0.7911,
      "num_tokens": 146712990.0,
      "step": 140
    },
    {
      "epoch": 0.22943037974683544,
      "grad_norm": 0.772547411213572,
      "learning_rate": 4.588607594936709e-05,
      "loss": 0.788,
      "num_tokens": 151927597.0,
      "step": 145
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 0.5427524570272351,
      "learning_rate": 4.7468354430379746e-05,
      "loss": 0.7931,
      "num_tokens": 157170477.0,
      "step": 150
    },
    {
      "epoch": 0.24525316455696203,
      "grad_norm": 0.520616997801594,
      "learning_rate": 4.905063291139241e-05,
      "loss": 0.7761,
      "num_tokens": 162397909.0,
      "step": 155
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 1.321171420646125,
      "learning_rate": 4.9999950717727614e-05,
      "loss": 0.8103,
      "num_tokens": 167639712.0,
      "step": 160
    },
    {
      "epoch": 0.2610759493670886,
      "grad_norm": 0.8179961728401967,
      "learning_rate": 4.999939629464255e-05,
      "loss": 0.7829,
      "num_tokens": 172882592.0,
      "step": 165
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 1.07950233433435,
      "learning_rate": 4.999822586086208e-05,
      "loss": 0.7815,
      "num_tokens": 178125472.0,
      "step": 170
    },
    {
      "epoch": 0.27689873417721517,
      "grad_norm": 0.764049579486957,
      "learning_rate": 4.9996439448431495e-05,
      "loss": 0.7936,
      "num_tokens": 183355239.0,
      "step": 175
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.6643682550526433,
      "learning_rate": 4.9994037106260965e-05,
      "loss": 0.777,
      "num_tokens": 188598119.0,
      "step": 180
    },
    {
      "epoch": 0.2927215189873418,
      "grad_norm": 0.4988533797702287,
      "learning_rate": 4.999101890012418e-05,
      "loss": 0.7755,
      "num_tokens": 193840999.0,
      "step": 185
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.49381791616490034,
      "learning_rate": 4.998738491265655e-05,
      "loss": 0.7646,
      "num_tokens": 199082063.0,
      "step": 190
    },
    {
      "epoch": 0.30854430379746833,
      "grad_norm": 0.5570710586204755,
      "learning_rate": 4.998313524335302e-05,
      "loss": 0.7698,
      "num_tokens": 204324943.0,
      "step": 195
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.4911267149039519,
      "learning_rate": 4.9978270008565184e-05,
      "loss": 0.7641,
      "num_tokens": 209561370.0,
      "step": 200
    },
    {
      "epoch": 0.32436708860759494,
      "grad_norm": 0.6409945115911554,
      "learning_rate": 4.997278934149828e-05,
      "loss": 0.7677,
      "num_tokens": 214804250.0,
      "step": 205
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.5401344460770882,
      "learning_rate": 4.996669339220741e-05,
      "loss": 0.7657,
      "num_tokens": 220047130.0,
      "step": 210
    },
    {
      "epoch": 0.3401898734177215,
      "grad_norm": 0.5361201681791574,
      "learning_rate": 4.995998232759349e-05,
      "loss": 0.7663,
      "num_tokens": 225290010.0,
      "step": 215
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.5803989669904568,
      "learning_rate": 4.995265633139869e-05,
      "loss": 0.7636,
      "num_tokens": 230532890.0,
      "step": 220
    },
    {
      "epoch": 0.3560126582278481,
      "grad_norm": 0.726211587240982,
      "learning_rate": 4.994471560420137e-05,
      "loss": 0.7631,
      "num_tokens": 235775770.0,
      "step": 225
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.698163213622165,
      "learning_rate": 4.9936160363410584e-05,
      "loss": 0.7545,
      "num_tokens": 241018650.0,
      "step": 230
    },
    {
      "epoch": 0.37183544303797467,
      "grad_norm": 0.5153045338635585,
      "learning_rate": 4.992699084326018e-05,
      "loss": 0.7541,
      "num_tokens": 246261530.0,
      "step": 235
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.6959186613242923,
      "learning_rate": 4.9917207294802346e-05,
      "loss": 0.7635,
      "num_tokens": 251504410.0,
      "step": 240
    },
    {
      "epoch": 0.3876582278481013,
      "grad_norm": 0.475983385570461,
      "learning_rate": 4.990680998590071e-05,
      "loss": 0.755,
      "num_tokens": 256747290.0,
      "step": 245
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.6721509783638105,
      "learning_rate": 4.989579920122307e-05,
      "loss": 0.7532,
      "num_tokens": 261990170.0,
      "step": 250
    },
    {
      "epoch": 0.40348101265822783,
      "grad_norm": 0.4813755853658302,
      "learning_rate": 4.9884175242233585e-05,
      "loss": 0.7552,
      "num_tokens": 267233050.0,
      "step": 255
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 0.46711611820253635,
      "learning_rate": 4.987193842718448e-05,
      "loss": 0.7513,
      "num_tokens": 272475930.0,
      "step": 260
    },
    {
      "epoch": 0.41930379746835444,
      "grad_norm": 0.5972032077244416,
      "learning_rate": 4.985908909110735e-05,
      "loss": 0.7443,
      "num_tokens": 277718810.0,
      "step": 265
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.5059990764921701,
      "learning_rate": 4.9845627585804025e-05,
      "loss": 0.7458,
      "num_tokens": 282942349.0,
      "step": 270
    },
    {
      "epoch": 0.435126582278481,
      "grad_norm": 0.5582209701112856,
      "learning_rate": 4.9831554279836884e-05,
      "loss": 0.7521,
      "num_tokens": 288185229.0,
      "step": 275
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.4595435461264603,
      "learning_rate": 4.9816869558518796e-05,
      "loss": 0.7517,
      "num_tokens": 293428109.0,
      "step": 280
    },
    {
      "epoch": 0.4509493670886076,
      "grad_norm": 0.4675296642906588,
      "learning_rate": 4.980157382390258e-05,
      "loss": 0.7427,
      "num_tokens": 298670989.0,
      "step": 285
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.4659606979020265,
      "learning_rate": 4.978566749476995e-05,
      "loss": 0.7397,
      "num_tokens": 303913869.0,
      "step": 290
    },
    {
      "epoch": 0.46677215189873417,
      "grad_norm": 0.5143550840646993,
      "learning_rate": 4.9769151006620115e-05,
      "loss": 0.7488,
      "num_tokens": 309151588.0,
      "step": 295
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.5527231062409742,
      "learning_rate": 4.975202481165778e-05,
      "loss": 0.7528,
      "num_tokens": 314394468.0,
      "step": 300
    },
    {
      "epoch": 0.4825949367088608,
      "grad_norm": 0.6582518727489046,
      "learning_rate": 4.973428937878085e-05,
      "loss": 0.7388,
      "num_tokens": 319622567.0,
      "step": 305
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.5202900734758532,
      "learning_rate": 4.9715945193567506e-05,
      "loss": 0.7332,
      "num_tokens": 324865447.0,
      "step": 310
    },
    {
      "epoch": 0.49841772151898733,
      "grad_norm": 0.38453517395197584,
      "learning_rate": 4.969699275826298e-05,
      "loss": 0.7364,
      "num_tokens": 330108327.0,
      "step": 315
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.5632153181654556,
      "learning_rate": 4.967743259176575e-05,
      "loss": 0.7387,
      "num_tokens": 335351207.0,
      "step": 320
    },
    {
      "epoch": 0.5142405063291139,
      "grad_norm": 0.44887504740519496,
      "learning_rate": 4.96572652296134e-05,
      "loss": 0.7354,
      "num_tokens": 340594087.0,
      "step": 325
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.48735207873608405,
      "learning_rate": 4.963649122396788e-05,
      "loss": 0.7376,
      "num_tokens": 345825159.0,
      "step": 330
    },
    {
      "epoch": 0.5300632911392406,
      "grad_norm": 0.4193108797718534,
      "learning_rate": 4.961511114360043e-05,
      "loss": 0.7361,
      "num_tokens": 351068039.0,
      "step": 335
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.5662379674502074,
      "learning_rate": 4.959312557387601e-05,
      "loss": 0.739,
      "num_tokens": 356304704.0,
      "step": 340
    },
    {
      "epoch": 0.5458860759493671,
      "grad_norm": 0.5775103716382664,
      "learning_rate": 4.957053511673728e-05,
      "loss": 0.739,
      "num_tokens": 361547584.0,
      "step": 345
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.4784805413121098,
      "learning_rate": 4.954734039068806e-05,
      "loss": 0.7369,
      "num_tokens": 366775047.0,
      "step": 350
    },
    {
      "epoch": 0.5617088607594937,
      "grad_norm": 0.3826615225742145,
      "learning_rate": 4.95235420307765e-05,
      "loss": 0.7371,
      "num_tokens": 372017927.0,
      "step": 355
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.5250012256720773,
      "learning_rate": 4.949914068857758e-05,
      "loss": 0.7401,
      "num_tokens": 377260807.0,
      "step": 360
    },
    {
      "epoch": 0.5775316455696202,
      "grad_norm": 0.5294573268848846,
      "learning_rate": 4.947413703217537e-05,
      "loss": 0.7303,
      "num_tokens": 382503687.0,
      "step": 365
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.4657426470359354,
      "learning_rate": 4.944853174614465e-05,
      "loss": 0.7276,
      "num_tokens": 387746567.0,
      "step": 370
    },
    {
      "epoch": 0.5933544303797469,
      "grad_norm": 0.6110352375166948,
      "learning_rate": 4.9422325531532254e-05,
      "loss": 0.7332,
      "num_tokens": 392989447.0,
      "step": 375
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.5173863425147587,
      "learning_rate": 4.9395519105837794e-05,
      "loss": 0.7345,
      "num_tokens": 398220435.0,
      "step": 380
    },
    {
      "epoch": 0.6091772151898734,
      "grad_norm": 0.44630851057828674,
      "learning_rate": 4.936811320299409e-05,
      "loss": 0.7344,
      "num_tokens": 403454091.0,
      "step": 385
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.5511830323483669,
      "learning_rate": 4.9340108573346994e-05,
      "loss": 0.7277,
      "num_tokens": 408695566.0,
      "step": 390
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.488590082197144,
      "learning_rate": 4.931150598363494e-05,
      "loss": 0.7262,
      "num_tokens": 413938446.0,
      "step": 395
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.5980267797290539,
      "learning_rate": 4.9282306216967866e-05,
      "loss": 0.7367,
      "num_tokens": 419181326.0,
      "step": 400
    },
    {
      "epoch": 0.6408227848101266,
      "grad_norm": 0.4042639446511123,
      "learning_rate": 4.925251007280584e-05,
      "loss": 0.7277,
      "num_tokens": 424410675.0,
      "step": 405
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.5873987854936606,
      "learning_rate": 4.922211836693711e-05,
      "loss": 0.7216,
      "num_tokens": 429653555.0,
      "step": 410
    },
    {
      "epoch": 0.6566455696202531,
      "grad_norm": 0.49135762238140734,
      "learning_rate": 4.9191131931455836e-05,
      "loss": 0.7249,
      "num_tokens": 434896435.0,
      "step": 415
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.5977070057438132,
      "learning_rate": 4.915955161473925e-05,
      "loss": 0.7213,
      "num_tokens": 440139315.0,
      "step": 420
    },
    {
      "epoch": 0.6724683544303798,
      "grad_norm": 0.5865545939218848,
      "learning_rate": 4.9127378281424454e-05,
      "loss": 0.7291,
      "num_tokens": 445382195.0,
      "step": 425
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.5141049741136753,
      "learning_rate": 4.909461281238475e-05,
      "loss": 0.7275,
      "num_tokens": 450625075.0,
      "step": 430
    },
    {
      "epoch": 0.6882911392405063,
      "grad_norm": 0.586173282775543,
      "learning_rate": 4.9061256104705514e-05,
      "loss": 0.7147,
      "num_tokens": 455867955.0,
      "step": 435
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.4884462198074093,
      "learning_rate": 4.902730907165962e-05,
      "loss": 0.7262,
      "num_tokens": 461110835.0,
      "step": 440
    },
    {
      "epoch": 0.7041139240506329,
      "grad_norm": 0.5155782485627577,
      "learning_rate": 4.899277264268249e-05,
      "loss": 0.726,
      "num_tokens": 466329951.0,
      "step": 445
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.5894348115226837,
      "learning_rate": 4.8957647763346575e-05,
      "loss": 0.7276,
      "num_tokens": 471572831.0,
      "step": 450
    },
    {
      "epoch": 0.7199367088607594,
      "grad_norm": 0.6178813430658108,
      "learning_rate": 4.892193539533553e-05,
      "loss": 0.7234,
      "num_tokens": 476815711.0,
      "step": 455
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.7787566109612348,
      "learning_rate": 4.888563651641784e-05,
      "loss": 0.7297,
      "num_tokens": 482044487.0,
      "step": 460
    },
    {
      "epoch": 0.7357594936708861,
      "grad_norm": 0.5460153094045689,
      "learning_rate": 4.884875212042005e-05,
      "loss": 0.7257,
      "num_tokens": 487287367.0,
      "step": 465
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.5019931926593302,
      "learning_rate": 4.88112832171996e-05,
      "loss": 0.7219,
      "num_tokens": 492530247.0,
      "step": 470
    },
    {
      "epoch": 0.7515822784810127,
      "grad_norm": 0.46922376523806203,
      "learning_rate": 4.877323083261713e-05,
      "loss": 0.7207,
      "num_tokens": 497773127.0,
      "step": 475
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.4491131631584376,
      "learning_rate": 4.873459600850841e-05,
      "loss": 0.718,
      "num_tokens": 503013415.0,
      "step": 480
    },
    {
      "epoch": 0.7674050632911392,
      "grad_norm": 0.4576902411009356,
      "learning_rate": 4.869537980265581e-05,
      "loss": 0.7119,
      "num_tokens": 508255611.0,
      "step": 485
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.6179813478813936,
      "learning_rate": 4.865558328875937e-05,
      "loss": 0.7086,
      "num_tokens": 513498491.0,
      "step": 490
    },
    {
      "epoch": 0.7832278481012658,
      "grad_norm": 0.5471403707624125,
      "learning_rate": 4.8615207556407324e-05,
      "loss": 0.7242,
      "num_tokens": 518723580.0,
      "step": 495
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.4589250795263245,
      "learning_rate": 4.857425371104638e-05,
      "loss": 0.7159,
      "num_tokens": 523966460.0,
      "step": 500
    },
    {
      "epoch": 0.7990506329113924,
      "grad_norm": 0.4660626316391112,
      "learning_rate": 4.853272287395135e-05,
      "loss": 0.7251,
      "num_tokens": 529209340.0,
      "step": 505
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.5382100486924232,
      "learning_rate": 4.8490616182194534e-05,
      "loss": 0.7187,
      "num_tokens": 534429844.0,
      "step": 510
    },
    {
      "epoch": 0.814873417721519,
      "grad_norm": 0.5238715230138492,
      "learning_rate": 4.8447934788614515e-05,
      "loss": 0.7187,
      "num_tokens": 539672724.0,
      "step": 515
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.5194663159593585,
      "learning_rate": 4.840467986178464e-05,
      "loss": 0.7168,
      "num_tokens": 544915604.0,
      "step": 520
    },
    {
      "epoch": 0.8306962025316456,
      "grad_norm": 0.4135374652648048,
      "learning_rate": 4.8360852585981034e-05,
      "loss": 0.7236,
      "num_tokens": 550158484.0,
      "step": 525
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.47687000377153205,
      "learning_rate": 4.831645416115014e-05,
      "loss": 0.7091,
      "num_tokens": 555401364.0,
      "step": 530
    },
    {
      "epoch": 0.8465189873417721,
      "grad_norm": 0.4659031703902228,
      "learning_rate": 4.827148580287588e-05,
      "loss": 0.7042,
      "num_tokens": 560644244.0,
      "step": 535
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.432869438663762,
      "learning_rate": 4.82259487423464e-05,
      "loss": 0.7119,
      "num_tokens": 565887124.0,
      "step": 540
    },
    {
      "epoch": 0.8623417721518988,
      "grad_norm": 0.3989063394441302,
      "learning_rate": 4.81798442263203e-05,
      "loss": 0.7164,
      "num_tokens": 571114334.0,
      "step": 545
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.5839119244408789,
      "learning_rate": 4.8133173517092575e-05,
      "loss": 0.7147,
      "num_tokens": 576351352.0,
      "step": 550
    },
    {
      "epoch": 0.8781645569620253,
      "grad_norm": 0.4989762196407589,
      "learning_rate": 4.808593789246e-05,
      "loss": 0.7116,
      "num_tokens": 581594232.0,
      "step": 555
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.6013182449359468,
      "learning_rate": 4.803813864568616e-05,
      "loss": 0.7101,
      "num_tokens": 586837112.0,
      "step": 560
    },
    {
      "epoch": 0.8939873417721519,
      "grad_norm": 0.47611561215232195,
      "learning_rate": 4.7989777085466054e-05,
      "loss": 0.7089,
      "num_tokens": 592079992.0,
      "step": 565
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.5649210203680972,
      "learning_rate": 4.794085453589022e-05,
      "loss": 0.7074,
      "num_tokens": 597322872.0,
      "step": 570
    },
    {
      "epoch": 0.9098101265822784,
      "grad_norm": 0.4954711968927705,
      "learning_rate": 4.789137233640858e-05,
      "loss": 0.7057,
      "num_tokens": 602545993.0,
      "step": 575
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.45694982466683615,
      "learning_rate": 4.7841331841793646e-05,
      "loss": 0.7099,
      "num_tokens": 607788873.0,
      "step": 580
    },
    {
      "epoch": 0.9256329113924051,
      "grad_norm": 0.4331078260141868,
      "learning_rate": 4.7790734422103503e-05,
      "loss": 0.7074,
      "num_tokens": 613031753.0,
      "step": 585
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.6029099335069901,
      "learning_rate": 4.77395814626443e-05,
      "loss": 0.7132,
      "num_tokens": 618274069.0,
      "step": 590
    },
    {
      "epoch": 0.9414556962025317,
      "grad_norm": 0.3964222914078214,
      "learning_rate": 4.7687874363932277e-05,
      "loss": 0.7107,
      "num_tokens": 623516949.0,
      "step": 595
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.40557860921547184,
      "learning_rate": 4.763561454165546e-05,
      "loss": 0.7114,
      "num_tokens": 628759829.0,
      "step": 600
    },
    {
      "epoch": 0.9572784810126582,
      "grad_norm": 0.3911037173148,
      "learning_rate": 4.758280342663488e-05,
      "loss": 0.7011,
      "num_tokens": 634002709.0,
      "step": 605
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.5205226454815663,
      "learning_rate": 4.7529442464785404e-05,
      "loss": 0.7035,
      "num_tokens": 639245589.0,
      "step": 610
    },
    {
      "epoch": 0.9731012658227848,
      "grad_norm": 0.6081952511481521,
      "learning_rate": 4.747553311707616e-05,
      "loss": 0.7109,
      "num_tokens": 644488469.0,
      "step": 615
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.5230701333389108,
      "learning_rate": 4.742107685949052e-05,
      "loss": 0.7054,
      "num_tokens": 649731349.0,
      "step": 620
    },
    {
      "epoch": 0.9889240506329114,
      "grad_norm": 0.45553268701488436,
      "learning_rate": 4.736607518298568e-05,
      "loss": 0.7084,
      "num_tokens": 654971226.0,
      "step": 625
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.4252255621189171,
      "learning_rate": 4.731052959345188e-05,
      "loss": 0.698,
      "num_tokens": 660214106.0,
      "step": 630
    },
    {
      "epoch": 1.004746835443038,
      "grad_norm": 0.5213909939853125,
      "learning_rate": 4.725444161167113e-05,
      "loss": 0.6984,
      "num_tokens": 665456986.0,
      "step": 635
    },
    {
      "epoch": 1.0126582278481013,
      "grad_norm": 0.6092766382306241,
      "learning_rate": 4.71978127732756e-05,
      "loss": 0.6834,
      "num_tokens": 670699866.0,
      "step": 640
    },
    {
      "epoch": 1.0205696202531644,
      "grad_norm": 0.6016879209190829,
      "learning_rate": 4.714064462870556e-05,
      "loss": 0.6799,
      "num_tokens": 675941341.0,
      "step": 645
    },
    {
      "epoch": 1.0284810126582278,
      "grad_norm": 0.46770946063736324,
      "learning_rate": 4.708293874316693e-05,
      "loss": 0.6718,
      "num_tokens": 681184221.0,
      "step": 650
    },
    {
      "epoch": 1.0363924050632911,
      "grad_norm": 0.5193625042396812,
      "learning_rate": 4.702469669658845e-05,
      "loss": 0.6824,
      "num_tokens": 686427101.0,
      "step": 655
    },
    {
      "epoch": 1.0443037974683544,
      "grad_norm": 0.4310707796867463,
      "learning_rate": 4.6965920083578406e-05,
      "loss": 0.6814,
      "num_tokens": 691643199.0,
      "step": 660
    },
    {
      "epoch": 1.0522151898734178,
      "grad_norm": 0.46509456309972086,
      "learning_rate": 4.690661051338096e-05,
      "loss": 0.6783,
      "num_tokens": 696885395.0,
      "step": 665
    },
    {
      "epoch": 1.0601265822784811,
      "grad_norm": 0.4290977849490662,
      "learning_rate": 4.6846769609832106e-05,
      "loss": 0.6745,
      "num_tokens": 702099431.0,
      "step": 670
    },
    {
      "epoch": 1.0680379746835442,
      "grad_norm": 0.44562761654453537,
      "learning_rate": 4.6786399011315215e-05,
      "loss": 0.6708,
      "num_tokens": 707335241.0,
      "step": 675
    },
    {
      "epoch": 1.0759493670886076,
      "grad_norm": 0.4027263571632704,
      "learning_rate": 4.672550037071616e-05,
      "loss": 0.6742,
      "num_tokens": 712578121.0,
      "step": 680
    },
    {
      "epoch": 1.0838607594936709,
      "grad_norm": 0.4509630202842952,
      "learning_rate": 4.6664075355378064e-05,
      "loss": 0.6787,
      "num_tokens": 717813308.0,
      "step": 685
    },
    {
      "epoch": 1.0917721518987342,
      "grad_norm": 0.579102128781878,
      "learning_rate": 4.660212564705569e-05,
      "loss": 0.6827,
      "num_tokens": 723056188.0,
      "step": 690
    },
    {
      "epoch": 1.0996835443037976,
      "grad_norm": 0.4750686074161037,
      "learning_rate": 4.653965294186933e-05,
      "loss": 0.6846,
      "num_tokens": 728299068.0,
      "step": 695
    },
    {
      "epoch": 1.1075949367088607,
      "grad_norm": 0.3720445468877555,
      "learning_rate": 4.647665895025842e-05,
      "loss": 0.68,
      "num_tokens": 733541948.0,
      "step": 700
    },
    {
      "epoch": 1.115506329113924,
      "grad_norm": 0.41896058368158273,
      "learning_rate": 4.6413145396934677e-05,
      "loss": 0.6727,
      "num_tokens": 738784828.0,
      "step": 705
    },
    {
      "epoch": 1.1234177215189873,
      "grad_norm": 0.41709918206048263,
      "learning_rate": 4.634911402083491e-05,
      "loss": 0.683,
      "num_tokens": 744027708.0,
      "step": 710
    },
    {
      "epoch": 1.1313291139240507,
      "grad_norm": 0.44679889502370607,
      "learning_rate": 4.628456657507336e-05,
      "loss": 0.6745,
      "num_tokens": 749270588.0,
      "step": 715
    },
    {
      "epoch": 1.139240506329114,
      "grad_norm": 0.4049444543836899,
      "learning_rate": 4.6219504826893774e-05,
      "loss": 0.6754,
      "num_tokens": 754513468.0,
      "step": 720
    },
    {
      "epoch": 1.1471518987341773,
      "grad_norm": 0.46788618407855487,
      "learning_rate": 4.615393055762095e-05,
      "loss": 0.6705,
      "num_tokens": 759755275.0,
      "step": 725
    },
    {
      "epoch": 1.1550632911392404,
      "grad_norm": 0.45889187811298193,
      "learning_rate": 4.6087845562612e-05,
      "loss": 0.6775,
      "num_tokens": 764993238.0,
      "step": 730
    },
    {
      "epoch": 1.1629746835443038,
      "grad_norm": 0.5053059074244526,
      "learning_rate": 4.602125165120721e-05,
      "loss": 0.6739,
      "num_tokens": 770236118.0,
      "step": 735
    },
    {
      "epoch": 1.1708860759493671,
      "grad_norm": 0.5163271050928327,
      "learning_rate": 4.595415064668044e-05,
      "loss": 0.6768,
      "num_tokens": 775478998.0,
      "step": 740
    },
    {
      "epoch": 1.1787974683544304,
      "grad_norm": 0.4282870421961532,
      "learning_rate": 4.5886544386189286e-05,
      "loss": 0.6681,
      "num_tokens": 780721878.0,
      "step": 745
    },
    {
      "epoch": 1.1867088607594938,
      "grad_norm": 0.46026364834735695,
      "learning_rate": 4.581843472072472e-05,
      "loss": 0.677,
      "num_tokens": 785964758.0,
      "step": 750
    },
    {
      "epoch": 1.1946202531645569,
      "grad_norm": 0.3647491363050659,
      "learning_rate": 4.574982351506044e-05,
      "loss": 0.673,
      "num_tokens": 791207638.0,
      "step": 755
    },
    {
      "epoch": 1.2025316455696202,
      "grad_norm": 0.40334584668377854,
      "learning_rate": 4.568071264770179e-05,
      "loss": 0.6733,
      "num_tokens": 796450518.0,
      "step": 760
    },
    {
      "epoch": 1.2104430379746836,
      "grad_norm": 0.6172425933341248,
      "learning_rate": 4.5611104010834384e-05,
      "loss": 0.6714,
      "num_tokens": 801691627.0,
      "step": 765
    },
    {
      "epoch": 1.2183544303797469,
      "grad_norm": 0.5660318449559009,
      "learning_rate": 4.554099951027223e-05,
      "loss": 0.6695,
      "num_tokens": 806934507.0,
      "step": 770
    },
    {
      "epoch": 1.2262658227848102,
      "grad_norm": 0.5013160792670336,
      "learning_rate": 4.5470401065405574e-05,
      "loss": 0.6733,
      "num_tokens": 812177387.0,
      "step": 775
    },
    {
      "epoch": 1.2341772151898733,
      "grad_norm": 0.48633536466547755,
      "learning_rate": 4.539931060914841e-05,
      "loss": 0.6775,
      "num_tokens": 817420267.0,
      "step": 780
    },
    {
      "epoch": 1.2420886075949367,
      "grad_norm": 0.5345960534136472,
      "learning_rate": 4.532773008788542e-05,
      "loss": 0.6742,
      "num_tokens": 822663147.0,
      "step": 785
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.46751637317248607,
      "learning_rate": 4.5255661461418854e-05,
      "loss": 0.6683,
      "num_tokens": 827902185.0,
      "step": 790
    },
    {
      "epoch": 1.2579113924050633,
      "grad_norm": 0.40210236140036404,
      "learning_rate": 4.5183106702914744e-05,
      "loss": 0.6684,
      "num_tokens": 833145065.0,
      "step": 795
    },
    {
      "epoch": 1.2658227848101267,
      "grad_norm": 0.40800636198331647,
      "learning_rate": 4.511006779884894e-05,
      "loss": 0.6717,
      "num_tokens": 838387945.0,
      "step": 800
    },
    {
      "epoch": 1.2737341772151898,
      "grad_norm": 0.39143626003055415,
      "learning_rate": 4.503654674895268e-05,
      "loss": 0.6729,
      "num_tokens": 843630825.0,
      "step": 805
    },
    {
      "epoch": 1.2816455696202531,
      "grad_norm": 0.4114368812178738,
      "learning_rate": 4.49625455661579e-05,
      "loss": 0.6662,
      "num_tokens": 848873705.0,
      "step": 810
    },
    {
      "epoch": 1.2895569620253164,
      "grad_norm": 0.3850979319103424,
      "learning_rate": 4.4888066276542076e-05,
      "loss": 0.6676,
      "num_tokens": 854116585.0,
      "step": 815
    },
    {
      "epoch": 1.2974683544303798,
      "grad_norm": 0.3942547463552602,
      "learning_rate": 4.481311091927278e-05,
      "loss": 0.6697,
      "num_tokens": 859340010.0,
      "step": 820
    },
    {
      "epoch": 1.3053797468354431,
      "grad_norm": 0.3827683736050918,
      "learning_rate": 4.47376815465518e-05,
      "loss": 0.6648,
      "num_tokens": 864560605.0,
      "step": 825
    },
    {
      "epoch": 1.3132911392405062,
      "grad_norm": 0.4613241812192372,
      "learning_rate": 4.466178022355902e-05,
      "loss": 0.6757,
      "num_tokens": 869803485.0,
      "step": 830
    },
    {
      "epoch": 1.3212025316455696,
      "grad_norm": 0.3407656518079187,
      "learning_rate": 4.458540902839582e-05,
      "loss": 0.6643,
      "num_tokens": 875046365.0,
      "step": 835
    },
    {
      "epoch": 1.3291139240506329,
      "grad_norm": 0.3433476064598642,
      "learning_rate": 4.450857005202823e-05,
      "loss": 0.6684,
      "num_tokens": 880272119.0,
      "step": 840
    },
    {
      "epoch": 1.3370253164556962,
      "grad_norm": 0.4072649317968132,
      "learning_rate": 4.443126539822962e-05,
      "loss": 0.6758,
      "num_tokens": 885514999.0,
      "step": 845
    },
    {
      "epoch": 1.3449367088607596,
      "grad_norm": 0.38388206068514286,
      "learning_rate": 4.435349718352319e-05,
      "loss": 0.6651,
      "num_tokens": 890751664.0,
      "step": 850
    },
    {
      "epoch": 1.3528481012658227,
      "grad_norm": 0.4906281471488268,
      "learning_rate": 4.427526753712392e-05,
      "loss": 0.6758,
      "num_tokens": 895978998.0,
      "step": 855
    },
    {
      "epoch": 1.360759493670886,
      "grad_norm": 0.5690161031589217,
      "learning_rate": 4.419657860088033e-05,
      "loss": 0.6661,
      "num_tokens": 901206430.0,
      "step": 860
    },
    {
      "epoch": 1.3686708860759493,
      "grad_norm": 0.5239364843573313,
      "learning_rate": 4.411743252921587e-05,
      "loss": 0.6711,
      "num_tokens": 906449310.0,
      "step": 865
    },
    {
      "epoch": 1.3765822784810127,
      "grad_norm": 0.4590626844046868,
      "learning_rate": 4.403783148906984e-05,
      "loss": 0.6742,
      "num_tokens": 911692190.0,
      "step": 870
    },
    {
      "epoch": 1.384493670886076,
      "grad_norm": 0.6369825573287001,
      "learning_rate": 4.3957777659838156e-05,
      "loss": 0.6691,
      "num_tokens": 916923938.0,
      "step": 875
    },
    {
      "epoch": 1.3924050632911391,
      "grad_norm": 0.5309325547971587,
      "learning_rate": 4.387727323331362e-05,
      "loss": 0.6671,
      "num_tokens": 922166818.0,
      "step": 880
    },
    {
      "epoch": 1.4003164556962027,
      "grad_norm": 0.4182221238855379,
      "learning_rate": 4.379632041362594e-05,
      "loss": 0.6673,
      "num_tokens": 927409698.0,
      "step": 885
    },
    {
      "epoch": 1.4082278481012658,
      "grad_norm": 0.37453984190952516,
      "learning_rate": 4.371492141718138e-05,
      "loss": 0.6669,
      "num_tokens": 932652578.0,
      "step": 890
    },
    {
      "epoch": 1.4161392405063291,
      "grad_norm": 0.4472314832066788,
      "learning_rate": 4.363307847260206e-05,
      "loss": 0.6609,
      "num_tokens": 937883810.0,
      "step": 895
    },
    {
      "epoch": 1.4240506329113924,
      "grad_norm": 0.40759290291866934,
      "learning_rate": 4.3550793820664965e-05,
      "loss": 0.6657,
      "num_tokens": 943113159.0,
      "step": 900
    },
    {
      "epoch": 1.4319620253164558,
      "grad_norm": 0.38119165043031206,
      "learning_rate": 4.3468069714240556e-05,
      "loss": 0.6674,
      "num_tokens": 948346815.0,
      "step": 905
    },
    {
      "epoch": 1.439873417721519,
      "grad_norm": 0.3408121507908347,
      "learning_rate": 4.3384908418231144e-05,
      "loss": 0.6649,
      "num_tokens": 953589695.0,
      "step": 910
    },
    {
      "epoch": 1.4477848101265822,
      "grad_norm": 0.3824255603451359,
      "learning_rate": 4.330131220950883e-05,
      "loss": 0.664,
      "num_tokens": 958831498.0,
      "step": 915
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 0.42035687659685655,
      "learning_rate": 4.321728337685318e-05,
      "loss": 0.667,
      "num_tokens": 964074378.0,
      "step": 920
    },
    {
      "epoch": 1.4636075949367089,
      "grad_norm": 0.45598858937412523,
      "learning_rate": 4.313282422088859e-05,
      "loss": 0.6685,
      "num_tokens": 969317258.0,
      "step": 925
    },
    {
      "epoch": 1.4715189873417722,
      "grad_norm": 0.44573187825703264,
      "learning_rate": 4.304793705402124e-05,
      "loss": 0.6613,
      "num_tokens": 974546034.0,
      "step": 930
    },
    {
      "epoch": 1.4794303797468356,
      "grad_norm": 0.3946235668573835,
      "learning_rate": 4.2962624200375835e-05,
      "loss": 0.6671,
      "num_tokens": 979788914.0,
      "step": 935
    },
    {
      "epoch": 1.4873417721518987,
      "grad_norm": 0.3228711597439979,
      "learning_rate": 4.287688799573195e-05,
      "loss": 0.6686,
      "num_tokens": 985031794.0,
      "step": 940
    },
    {
      "epoch": 1.495253164556962,
      "grad_norm": 0.3636228559548819,
      "learning_rate": 4.2790730787460096e-05,
      "loss": 0.6657,
      "num_tokens": 990255694.0,
      "step": 945
    },
    {
      "epoch": 1.5031645569620253,
      "grad_norm": 0.36277524743915257,
      "learning_rate": 4.270415493445739e-05,
      "loss": 0.6618,
      "num_tokens": 995498574.0,
      "step": 950
    },
    {
      "epoch": 1.5110759493670884,
      "grad_norm": 0.449377328776007,
      "learning_rate": 4.2617162807083084e-05,
      "loss": 0.664,
      "num_tokens": 1000741454.0,
      "step": 955
    },
    {
      "epoch": 1.518987341772152,
      "grad_norm": 0.5226848678404952,
      "learning_rate": 4.252975678709354e-05,
      "loss": 0.664,
      "num_tokens": 1005981331.0,
      "step": 960
    },
    {
      "epoch": 1.5268987341772151,
      "grad_norm": 0.426961497295756,
      "learning_rate": 4.244193926757713e-05,
      "loss": 0.6672,
      "num_tokens": 1011224211.0,
      "step": 965
    },
    {
      "epoch": 1.5348101265822784,
      "grad_norm": 0.3336409027569924,
      "learning_rate": 4.235371265288864e-05,
      "loss": 0.6681,
      "num_tokens": 1016467091.0,
      "step": 970
    },
    {
      "epoch": 1.5427215189873418,
      "grad_norm": 0.4036839066490488,
      "learning_rate": 4.2265079358583455e-05,
      "loss": 0.6643,
      "num_tokens": 1021695784.0,
      "step": 975
    },
    {
      "epoch": 1.5506329113924051,
      "grad_norm": 0.7140710792273346,
      "learning_rate": 4.217604181135147e-05,
      "loss": 0.6697,
      "num_tokens": 1026938664.0,
      "step": 980
    },
    {
      "epoch": 1.5585443037974684,
      "grad_norm": 0.5962932136354121,
      "learning_rate": 4.20866024489506e-05,
      "loss": 0.6636,
      "num_tokens": 1032181544.0,
      "step": 985
    },
    {
      "epoch": 1.5664556962025316,
      "grad_norm": 0.5399117523136845,
      "learning_rate": 4.199676372014004e-05,
      "loss": 0.6642,
      "num_tokens": 1037424424.0,
      "step": 990
    },
    {
      "epoch": 1.5743670886075949,
      "grad_norm": 0.5727471985317703,
      "learning_rate": 4.190652808461326e-05,
      "loss": 0.6673,
      "num_tokens": 1042667304.0,
      "step": 995
    },
    {
      "epoch": 1.5822784810126582,
      "grad_norm": 0.4218917245287635,
      "learning_rate": 4.1815898012930603e-05,
      "loss": 0.6597,
      "num_tokens": 1047910184.0,
      "step": 1000
    },
    {
      "epoch": 1.5901898734177216,
      "grad_norm": 0.3919982695655947,
      "learning_rate": 4.172487598645171e-05,
      "loss": 0.6587,
      "num_tokens": 1053153064.0,
      "step": 1005
    },
    {
      "epoch": 1.5981012658227849,
      "grad_norm": 0.35433285883968835,
      "learning_rate": 4.163346449726752e-05,
      "loss": 0.6658,
      "num_tokens": 1058395944.0,
      "step": 1010
    },
    {
      "epoch": 1.606012658227848,
      "grad_norm": 0.40624024878352494,
      "learning_rate": 4.154166604813206e-05,
      "loss": 0.661,
      "num_tokens": 1063638824.0,
      "step": 1015
    },
    {
      "epoch": 1.6139240506329116,
      "grad_norm": 0.45360696653600874,
      "learning_rate": 4.144948315239397e-05,
      "loss": 0.6636,
      "num_tokens": 1068880391.0,
      "step": 1020
    },
    {
      "epoch": 1.6218354430379747,
      "grad_norm": 0.528510428398564,
      "learning_rate": 4.13569183339276e-05,
      "loss": 0.6561,
      "num_tokens": 1074123271.0,
      "step": 1025
    },
    {
      "epoch": 1.629746835443038,
      "grad_norm": 0.435117205930865,
      "learning_rate": 4.1263974127063994e-05,
      "loss": 0.6572,
      "num_tokens": 1079366151.0,
      "step": 1030
    },
    {
      "epoch": 1.6376582278481013,
      "grad_norm": 0.35010239698525997,
      "learning_rate": 4.117065307652145e-05,
      "loss": 0.6675,
      "num_tokens": 1084603567.0,
      "step": 1035
    },
    {
      "epoch": 1.6455696202531644,
      "grad_norm": 0.34340761605514514,
      "learning_rate": 4.107695773733586e-05,
      "loss": 0.6634,
      "num_tokens": 1089846447.0,
      "step": 1040
    },
    {
      "epoch": 1.653481012658228,
      "grad_norm": 0.3800590366717773,
      "learning_rate": 4.098289067479077e-05,
      "loss": 0.6529,
      "num_tokens": 1095089327.0,
      "step": 1045
    },
    {
      "epoch": 1.6613924050632911,
      "grad_norm": 0.5301647765231198,
      "learning_rate": 4.0888454464347156e-05,
      "loss": 0.6608,
      "num_tokens": 1100328290.0,
      "step": 1050
    },
    {
      "epoch": 1.6693037974683544,
      "grad_norm": 0.40371596065925885,
      "learning_rate": 4.079365169157283e-05,
      "loss": 0.6596,
      "num_tokens": 1105565945.0,
      "step": 1055
    },
    {
      "epoch": 1.6772151898734178,
      "grad_norm": 0.44878584304658214,
      "learning_rate": 4.069848495207176e-05,
      "loss": 0.6584,
      "num_tokens": 1110808825.0,
      "step": 1060
    },
    {
      "epoch": 1.685126582278481,
      "grad_norm": 0.42760590310826146,
      "learning_rate": 4.060295685141295e-05,
      "loss": 0.6601,
      "num_tokens": 1116051705.0,
      "step": 1065
    },
    {
      "epoch": 1.6930379746835444,
      "grad_norm": 0.42094155594543997,
      "learning_rate": 4.0507070005059086e-05,
      "loss": 0.6636,
      "num_tokens": 1121294585.0,
      "step": 1070
    },
    {
      "epoch": 1.7009493670886076,
      "grad_norm": 0.366450602326517,
      "learning_rate": 4.0410827038294966e-05,
      "loss": 0.6565,
      "num_tokens": 1126537465.0,
      "step": 1075
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.3382264432193073,
      "learning_rate": 4.031423058615559e-05,
      "loss": 0.6637,
      "num_tokens": 1131778529.0,
      "step": 1080
    },
    {
      "epoch": 1.7167721518987342,
      "grad_norm": 0.4135053453091331,
      "learning_rate": 4.0217283293354044e-05,
      "loss": 0.6598,
      "num_tokens": 1137021409.0,
      "step": 1085
    },
    {
      "epoch": 1.7246835443037973,
      "grad_norm": 0.32789223717337174,
      "learning_rate": 4.011998781420907e-05,
      "loss": 0.66,
      "num_tokens": 1142264289.0,
      "step": 1090
    },
    {
      "epoch": 1.7325949367088609,
      "grad_norm": 0.37159495208450527,
      "learning_rate": 4.002234681257239e-05,
      "loss": 0.6634,
      "num_tokens": 1147488495.0,
      "step": 1095
    },
    {
      "epoch": 1.740506329113924,
      "grad_norm": 0.41706053572545077,
      "learning_rate": 3.992436296175581e-05,
      "loss": 0.6554,
      "num_tokens": 1152731375.0,
      "step": 1100
    },
    {
      "epoch": 1.7484177215189873,
      "grad_norm": 0.3410031339365452,
      "learning_rate": 3.982603894445796e-05,
      "loss": 0.6574,
      "num_tokens": 1157974255.0,
      "step": 1105
    },
    {
      "epoch": 1.7563291139240507,
      "grad_norm": 0.40387170849869736,
      "learning_rate": 3.97273774526909e-05,
      "loss": 0.6518,
      "num_tokens": 1163196277.0,
      "step": 1110
    },
    {
      "epoch": 1.7642405063291138,
      "grad_norm": 0.41494353564073877,
      "learning_rate": 3.962838118770643e-05,
      "loss": 0.6521,
      "num_tokens": 1168438593.0,
      "step": 1115
    },
    {
      "epoch": 1.7721518987341773,
      "grad_norm": 0.33845287541155555,
      "learning_rate": 3.952905285992206e-05,
      "loss": 0.6636,
      "num_tokens": 1173674462.0,
      "step": 1120
    },
    {
      "epoch": 1.7800632911392404,
      "grad_norm": 0.3616932835037642,
      "learning_rate": 3.942939518884686e-05,
      "loss": 0.6566,
      "num_tokens": 1178914750.0,
      "step": 1125
    },
    {
      "epoch": 1.7879746835443038,
      "grad_norm": 0.38849401008078355,
      "learning_rate": 3.932941090300699e-05,
      "loss": 0.6546,
      "num_tokens": 1184150094.0,
      "step": 1130
    },
    {
      "epoch": 1.7958860759493671,
      "grad_norm": 0.42281182227916236,
      "learning_rate": 3.922910273987098e-05,
      "loss": 0.6548,
      "num_tokens": 1189392974.0,
      "step": 1135
    },
    {
      "epoch": 1.8037974683544302,
      "grad_norm": 0.4210942734836147,
      "learning_rate": 3.912847344577481e-05,
      "loss": 0.6555,
      "num_tokens": 1194635854.0,
      "step": 1140
    },
    {
      "epoch": 1.8117088607594938,
      "grad_norm": 0.5474931691520276,
      "learning_rate": 3.9027525775846666e-05,
      "loss": 0.6533,
      "num_tokens": 1199878734.0,
      "step": 1145
    },
    {
      "epoch": 1.8196202531645569,
      "grad_norm": 0.3732844050319297,
      "learning_rate": 3.892626249393159e-05,
      "loss": 0.6533,
      "num_tokens": 1205121614.0,
      "step": 1150
    },
    {
      "epoch": 1.8275316455696202,
      "grad_norm": 0.45982960737179485,
      "learning_rate": 3.882468637251573e-05,
      "loss": 0.658,
      "num_tokens": 1210364494.0,
      "step": 1155
    },
    {
      "epoch": 1.8354430379746836,
      "grad_norm": 0.3489396715842175,
      "learning_rate": 3.872280019265046e-05,
      "loss": 0.6583,
      "num_tokens": 1215607374.0,
      "step": 1160
    },
    {
      "epoch": 1.8433544303797469,
      "grad_norm": 0.35751783832753387,
      "learning_rate": 3.8620606743876264e-05,
      "loss": 0.6522,
      "num_tokens": 1220843801.0,
      "step": 1165
    },
    {
      "epoch": 1.8512658227848102,
      "grad_norm": 0.4188675688211132,
      "learning_rate": 3.851810882414632e-05,
      "loss": 0.6497,
      "num_tokens": 1226086681.0,
      "step": 1170
    },
    {
      "epoch": 1.8591772151898733,
      "grad_norm": 0.3704256158964576,
      "learning_rate": 3.841530923974991e-05,
      "loss": 0.6542,
      "num_tokens": 1231329561.0,
      "step": 1175
    },
    {
      "epoch": 1.8670886075949367,
      "grad_norm": 0.35739760458738973,
      "learning_rate": 3.83122108052356e-05,
      "loss": 0.6558,
      "num_tokens": 1236572441.0,
      "step": 1180
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.4108797442075266,
      "learning_rate": 3.8208816343334156e-05,
      "loss": 0.6503,
      "num_tokens": 1241815321.0,
      "step": 1185
    },
    {
      "epoch": 1.8829113924050633,
      "grad_norm": 0.3989382930649431,
      "learning_rate": 3.810512868488129e-05,
      "loss": 0.6567,
      "num_tokens": 1247058201.0,
      "step": 1190
    },
    {
      "epoch": 1.8908227848101267,
      "grad_norm": 0.36123580918393156,
      "learning_rate": 3.800115066874014e-05,
      "loss": 0.6539,
      "num_tokens": 1252301081.0,
      "step": 1195
    },
    {
      "epoch": 1.8987341772151898,
      "grad_norm": 0.3372531720914223,
      "learning_rate": 3.789688514172353e-05,
      "loss": 0.6473,
      "num_tokens": 1257543961.0,
      "step": 1200
    },
    {
      "epoch": 1.9066455696202531,
      "grad_norm": 0.3512013159202198,
      "learning_rate": 3.779233495851604e-05,
      "loss": 0.6581,
      "num_tokens": 1262786841.0,
      "step": 1205
    },
    {
      "epoch": 1.9145569620253164,
      "grad_norm": 0.3542753269510345,
      "learning_rate": 3.768750298159587e-05,
      "loss": 0.6445,
      "num_tokens": 1268029721.0,
      "step": 1210
    },
    {
      "epoch": 1.9224683544303798,
      "grad_norm": 0.4031550450820841,
      "learning_rate": 3.75823920811564e-05,
      "loss": 0.6508,
      "num_tokens": 1273258360.0,
      "step": 1215
    },
    {
      "epoch": 1.9303797468354431,
      "grad_norm": 0.4403211137259473,
      "learning_rate": 3.747700513502772e-05,
      "loss": 0.6573,
      "num_tokens": 1278501240.0,
      "step": 1220
    },
    {
      "epoch": 1.9382911392405062,
      "grad_norm": 0.4812191269085378,
      "learning_rate": 3.737134502859772e-05,
      "loss": 0.6536,
      "num_tokens": 1283744120.0,
      "step": 1225
    },
    {
      "epoch": 1.9462025316455698,
      "grad_norm": 0.39299909706187625,
      "learning_rate": 3.726541465473317e-05,
      "loss": 0.649,
      "num_tokens": 1288987000.0,
      "step": 1230
    },
    {
      "epoch": 1.9541139240506329,
      "grad_norm": 0.39824695051848524,
      "learning_rate": 3.7159216913700456e-05,
      "loss": 0.6478,
      "num_tokens": 1294229880.0,
      "step": 1235
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 0.42039758141791633,
      "learning_rate": 3.7052754713086246e-05,
      "loss": 0.6534,
      "num_tokens": 1299472760.0,
      "step": 1240
    },
    {
      "epoch": 1.9699367088607596,
      "grad_norm": 0.3960672982967281,
      "learning_rate": 3.694603096771781e-05,
      "loss": 0.6568,
      "num_tokens": 1304715640.0,
      "step": 1245
    },
    {
      "epoch": 1.9778481012658227,
      "grad_norm": 0.3747599520232236,
      "learning_rate": 3.683904859958329e-05,
      "loss": 0.651,
      "num_tokens": 1309942452.0,
      "step": 1250
    },
    {
      "epoch": 1.9857594936708862,
      "grad_norm": 0.34963602368732627,
      "learning_rate": 3.673181053775162e-05,
      "loss": 0.6591,
      "num_tokens": 1315185332.0,
      "step": 1255
    },
    {
      "epoch": 1.9936708860759493,
      "grad_norm": 0.3336233742912806,
      "learning_rate": 3.662431971829237e-05,
      "loss": 0.6567,
      "num_tokens": 1320428212.0,
      "step": 1260
    },
    {
      "epoch": 2.0015822784810124,
      "grad_norm": 0.34626223476299406,
      "learning_rate": 3.651657908419537e-05,
      "loss": 0.6528,
      "num_tokens": 1325656988.0,
      "step": 1265
    },
    {
      "epoch": 2.009493670886076,
      "grad_norm": 0.3739591180690111,
      "learning_rate": 3.640859158529014e-05,
      "loss": 0.6228,
      "num_tokens": 1330899868.0,
      "step": 1270
    },
    {
      "epoch": 2.017405063291139,
      "grad_norm": 0.36163026913569957,
      "learning_rate": 3.6300360178165065e-05,
      "loss": 0.6198,
      "num_tokens": 1336142748.0,
      "step": 1275
    },
    {
      "epoch": 2.0253164556962027,
      "grad_norm": 0.3413456124109847,
      "learning_rate": 3.619188782608653e-05,
      "loss": 0.6249,
      "num_tokens": 1341371387.0,
      "step": 1280
    },
    {
      "epoch": 2.0332278481012658,
      "grad_norm": 0.41603416587113357,
      "learning_rate": 3.6083177498917745e-05,
      "loss": 0.6137,
      "num_tokens": 1346614267.0,
      "step": 1285
    },
    {
      "epoch": 2.041139240506329,
      "grad_norm": 0.39514773706497214,
      "learning_rate": 3.5974232173037385e-05,
      "loss": 0.6292,
      "num_tokens": 1351857147.0,
      "step": 1290
    },
    {
      "epoch": 2.0490506329113924,
      "grad_norm": 0.3949477431561823,
      "learning_rate": 3.586505483125823e-05,
      "loss": 0.6204,
      "num_tokens": 1357100027.0,
      "step": 1295
    },
    {
      "epoch": 2.0569620253164556,
      "grad_norm": 0.324048846717354,
      "learning_rate": 3.5755648462745366e-05,
      "loss": 0.6263,
      "num_tokens": 1362342907.0,
      "step": 1300
    },
    {
      "epoch": 2.064873417721519,
      "grad_norm": 0.3478785681280589,
      "learning_rate": 3.5646016062934413e-05,
      "loss": 0.6216,
      "num_tokens": 1367576563.0,
      "step": 1305
    },
    {
      "epoch": 2.0727848101265822,
      "grad_norm": 0.38436773833943666,
      "learning_rate": 3.553616063344951e-05,
      "loss": 0.6255,
      "num_tokens": 1372800486.0,
      "step": 1310
    },
    {
      "epoch": 2.0806962025316458,
      "grad_norm": 0.38048821813178435,
      "learning_rate": 3.5426085182021114e-05,
      "loss": 0.6208,
      "num_tokens": 1378043366.0,
      "step": 1315
    },
    {
      "epoch": 2.088607594936709,
      "grad_norm": 0.3868214542706034,
      "learning_rate": 3.531579272240366e-05,
      "loss": 0.6235,
      "num_tokens": 1383286246.0,
      "step": 1320
    },
    {
      "epoch": 2.096518987341772,
      "grad_norm": 0.29924119376240554,
      "learning_rate": 3.520528627429304e-05,
      "loss": 0.616,
      "num_tokens": 1388529126.0,
      "step": 1325
    },
    {
      "epoch": 2.1044303797468356,
      "grad_norm": 0.30143694906012813,
      "learning_rate": 3.509456886324395e-05,
      "loss": 0.6228,
      "num_tokens": 1393772006.0,
      "step": 1330
    },
    {
      "epoch": 2.1123417721518987,
      "grad_norm": 0.3430834549488217,
      "learning_rate": 3.498364352058703e-05,
      "loss": 0.6231,
      "num_tokens": 1399014886.0,
      "step": 1335
    },
    {
      "epoch": 2.1202531645569622,
      "grad_norm": 0.3918758750528519,
      "learning_rate": 3.487251328334588e-05,
      "loss": 0.6184,
      "num_tokens": 1404254509.0,
      "step": 1340
    },
    {
      "epoch": 2.1281645569620253,
      "grad_norm": 0.3860740822731293,
      "learning_rate": 3.47611811941539e-05,
      "loss": 0.6223,
      "num_tokens": 1409497389.0,
      "step": 1345
    },
    {
      "epoch": 2.1360759493670884,
      "grad_norm": 0.30507806139472327,
      "learning_rate": 3.464965030117099e-05,
      "loss": 0.6227,
      "num_tokens": 1414740269.0,
      "step": 1350
    },
    {
      "epoch": 2.143987341772152,
      "grad_norm": 0.31931935671102873,
      "learning_rate": 3.453792365800011e-05,
      "loss": 0.6213,
      "num_tokens": 1419976138.0,
      "step": 1355
    },
    {
      "epoch": 2.151898734177215,
      "grad_norm": 0.4296009660896887,
      "learning_rate": 3.4426004323603655e-05,
      "loss": 0.6195,
      "num_tokens": 1425219018.0,
      "step": 1360
    },
    {
      "epoch": 2.1598101265822787,
      "grad_norm": 0.36080348698318637,
      "learning_rate": 3.4313895362219704e-05,
      "loss": 0.6321,
      "num_tokens": 1430461898.0,
      "step": 1365
    },
    {
      "epoch": 2.1677215189873418,
      "grad_norm": 0.31266622534218946,
      "learning_rate": 3.420159984327814e-05,
      "loss": 0.6329,
      "num_tokens": 1435704778.0,
      "step": 1370
    },
    {
      "epoch": 2.175632911392405,
      "grad_norm": 0.29368200019966956,
      "learning_rate": 3.40891208413166e-05,
      "loss": 0.622,
      "num_tokens": 1440941205.0,
      "step": 1375
    },
    {
      "epoch": 2.1835443037974684,
      "grad_norm": 0.3318857805197203,
      "learning_rate": 3.397646143589629e-05,
      "loss": 0.6218,
      "num_tokens": 1446182176.0,
      "step": 1380
    },
    {
      "epoch": 2.1914556962025316,
      "grad_norm": 0.3469231890169258,
      "learning_rate": 3.38636247115177e-05,
      "loss": 0.6241,
      "num_tokens": 1451425056.0,
      "step": 1385
    },
    {
      "epoch": 2.199367088607595,
      "grad_norm": 0.3555477367643954,
      "learning_rate": 3.3750613757536134e-05,
      "loss": 0.6237,
      "num_tokens": 1456667936.0,
      "step": 1390
    },
    {
      "epoch": 2.2072784810126582,
      "grad_norm": 0.4308050328882218,
      "learning_rate": 3.36374316680771e-05,
      "loss": 0.6269,
      "num_tokens": 1461909000.0,
      "step": 1395
    },
    {
      "epoch": 2.2151898734177213,
      "grad_norm": 0.35870973114272886,
      "learning_rate": 3.3524081541951644e-05,
      "loss": 0.6222,
      "num_tokens": 1467150803.0,
      "step": 1400
    },
    {
      "epoch": 2.223101265822785,
      "grad_norm": 0.3606193086420081,
      "learning_rate": 3.3410566482571484e-05,
      "loss": 0.625,
      "num_tokens": 1472393683.0,
      "step": 1405
    },
    {
      "epoch": 2.231012658227848,
      "grad_norm": 0.3173225301332037,
      "learning_rate": 3.329688959786403e-05,
      "loss": 0.6225,
      "num_tokens": 1477635158.0,
      "step": 1410
    },
    {
      "epoch": 2.2389240506329116,
      "grad_norm": 0.3691990634713518,
      "learning_rate": 3.318305400018732e-05,
      "loss": 0.6231,
      "num_tokens": 1482878038.0,
      "step": 1415
    },
    {
      "epoch": 2.2468354430379747,
      "grad_norm": 0.4172329830772149,
      "learning_rate": 3.306906280624479e-05,
      "loss": 0.6286,
      "num_tokens": 1488105187.0,
      "step": 1420
    },
    {
      "epoch": 2.254746835443038,
      "grad_norm": 0.3279503525329187,
      "learning_rate": 3.295491913699994e-05,
      "loss": 0.6233,
      "num_tokens": 1493346994.0,
      "step": 1425
    },
    {
      "epoch": 2.2626582278481013,
      "grad_norm": 0.3632060308412657,
      "learning_rate": 3.284062611759089e-05,
      "loss": 0.6226,
      "num_tokens": 1498589874.0,
      "step": 1430
    },
    {
      "epoch": 2.2705696202531644,
      "grad_norm": 0.46252170164225265,
      "learning_rate": 3.272618687724481e-05,
      "loss": 0.6225,
      "num_tokens": 1503832754.0,
      "step": 1435
    },
    {
      "epoch": 2.278481012658228,
      "grad_norm": 0.3317711593176921,
      "learning_rate": 3.261160454919228e-05,
      "loss": 0.6236,
      "num_tokens": 1509075634.0,
      "step": 1440
    },
    {
      "epoch": 2.286392405063291,
      "grad_norm": 0.3470857725220561,
      "learning_rate": 3.249688227058145e-05,
      "loss": 0.6273,
      "num_tokens": 1514318514.0,
      "step": 1445
    },
    {
      "epoch": 2.2943037974683547,
      "grad_norm": 0.29280649749694565,
      "learning_rate": 3.238202318239216e-05,
      "loss": 0.6304,
      "num_tokens": 1519551983.0,
      "step": 1450
    },
    {
      "epoch": 2.3022151898734178,
      "grad_norm": 0.37241388988066654,
      "learning_rate": 3.226703042935e-05,
      "loss": 0.6184,
      "num_tokens": 1524794863.0,
      "step": 1455
    },
    {
      "epoch": 2.310126582278481,
      "grad_norm": 0.33976317790538424,
      "learning_rate": 3.215190715984013e-05,
      "loss": 0.622,
      "num_tokens": 1530031528.0,
      "step": 1460
    },
    {
      "epoch": 2.3180379746835444,
      "grad_norm": 0.29066911940116874,
      "learning_rate": 3.2036656525821144e-05,
      "loss": 0.6231,
      "num_tokens": 1535274408.0,
      "step": 1465
    },
    {
      "epoch": 2.3259493670886076,
      "grad_norm": 0.28162811956261474,
      "learning_rate": 3.1921281682738734e-05,
      "loss": 0.6182,
      "num_tokens": 1540517288.0,
      "step": 1470
    },
    {
      "epoch": 2.333860759493671,
      "grad_norm": 0.3524371457891882,
      "learning_rate": 3.180578578943933e-05,
      "loss": 0.6294,
      "num_tokens": 1545748994.0,
      "step": 1475
    },
    {
      "epoch": 2.3417721518987342,
      "grad_norm": 0.3468858213633566,
      "learning_rate": 3.1690172008083554e-05,
      "loss": 0.619,
      "num_tokens": 1550991874.0,
      "step": 1480
    },
    {
      "epoch": 2.3496835443037973,
      "grad_norm": 0.4224520445380759,
      "learning_rate": 3.157444350405972e-05,
      "loss": 0.6257,
      "num_tokens": 1556223622.0,
      "step": 1485
    },
    {
      "epoch": 2.357594936708861,
      "grad_norm": 0.3754485859450013,
      "learning_rate": 3.1458603445897136e-05,
      "loss": 0.6255,
      "num_tokens": 1561439202.0,
      "step": 1490
    },
    {
      "epoch": 2.365506329113924,
      "grad_norm": 0.3607328127982716,
      "learning_rate": 3.134265500517932e-05,
      "loss": 0.6184,
      "num_tokens": 1566661187.0,
      "step": 1495
    },
    {
      "epoch": 2.3734177215189876,
      "grad_norm": 0.3016702669665355,
      "learning_rate": 3.122660135645721e-05,
      "loss": 0.6226,
      "num_tokens": 1571904067.0,
      "step": 1500
    },
    {
      "epoch": 2.3813291139240507,
      "grad_norm": 0.3397158516715143,
      "learning_rate": 3.111044567716223e-05,
      "loss": 0.6237,
      "num_tokens": 1577146947.0,
      "step": 1505
    },
    {
      "epoch": 2.3892405063291138,
      "grad_norm": 0.34056042132518993,
      "learning_rate": 3.0994191147519296e-05,
      "loss": 0.6205,
      "num_tokens": 1582389827.0,
      "step": 1510
    },
    {
      "epoch": 2.3971518987341773,
      "grad_norm": 0.34821134798667835,
      "learning_rate": 3.0877840950459763e-05,
      "loss": 0.6217,
      "num_tokens": 1587632707.0,
      "step": 1515
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 0.29071305472990666,
      "learning_rate": 3.076139827153425e-05,
      "loss": 0.6212,
      "num_tokens": 1592875587.0,
      "step": 1520
    },
    {
      "epoch": 2.412974683544304,
      "grad_norm": 0.3213307109034431,
      "learning_rate": 3.064486629882543e-05,
      "loss": 0.6261,
      "num_tokens": 1598118467.0,
      "step": 1525
    },
    {
      "epoch": 2.420886075949367,
      "grad_norm": 0.317409439920391,
      "learning_rate": 3.052824822286075e-05,
      "loss": 0.6192,
      "num_tokens": 1603361347.0,
      "step": 1530
    },
    {
      "epoch": 2.4287974683544302,
      "grad_norm": 0.31888869908174583,
      "learning_rate": 3.041154723652509e-05,
      "loss": 0.6208,
      "num_tokens": 1608604227.0,
      "step": 1535
    },
    {
      "epoch": 2.4367088607594938,
      "grad_norm": 0.2946434998719226,
      "learning_rate": 3.0294766534973335e-05,
      "loss": 0.6291,
      "num_tokens": 1613847107.0,
      "step": 1540
    },
    {
      "epoch": 2.444620253164557,
      "grad_norm": 0.34443852085875215,
      "learning_rate": 3.0177909315542862e-05,
      "loss": 0.6256,
      "num_tokens": 1619089987.0,
      "step": 1545
    },
    {
      "epoch": 2.4525316455696204,
      "grad_norm": 0.35106351147093817,
      "learning_rate": 3.0060978777666054e-05,
      "loss": 0.6217,
      "num_tokens": 1624332867.0,
      "step": 1550
    },
    {
      "epoch": 2.4604430379746836,
      "grad_norm": 0.31920875041898267,
      "learning_rate": 2.994397812278269e-05,
      "loss": 0.6229,
      "num_tokens": 1629575747.0,
      "step": 1555
    },
    {
      "epoch": 2.4683544303797467,
      "grad_norm": 0.3877394978298516,
      "learning_rate": 2.9826910554252253e-05,
      "loss": 0.6272,
      "num_tokens": 1634818627.0,
      "step": 1560
    },
    {
      "epoch": 2.4762658227848102,
      "grad_norm": 0.31347418712210595,
      "learning_rate": 2.9709779277266258e-05,
      "loss": 0.6276,
      "num_tokens": 1640061507.0,
      "step": 1565
    },
    {
      "epoch": 2.4841772151898733,
      "grad_norm": 0.3663502650533136,
      "learning_rate": 2.959258749876052e-05,
      "loss": 0.6228,
      "num_tokens": 1645292739.0,
      "step": 1570
    },
    {
      "epoch": 2.492088607594937,
      "grad_norm": 0.4369820023540263,
      "learning_rate": 2.9475338427327298e-05,
      "loss": 0.622,
      "num_tokens": 1650520838.0,
      "step": 1575
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.38676813237014673,
      "learning_rate": 2.9358035273127483e-05,
      "loss": 0.6269,
      "num_tokens": 1655763718.0,
      "step": 1580
    },
    {
      "epoch": 2.507911392405063,
      "grad_norm": 0.3206842281049778,
      "learning_rate": 2.924068124780269e-05,
      "loss": 0.625,
      "num_tokens": 1661006598.0,
      "step": 1585
    },
    {
      "epoch": 2.5158227848101267,
      "grad_norm": 0.3173236770874382,
      "learning_rate": 2.9123279564387355e-05,
      "loss": 0.6134,
      "num_tokens": 1666249478.0,
      "step": 1590
    },
    {
      "epoch": 2.5237341772151898,
      "grad_norm": 0.32043576387698464,
      "learning_rate": 2.900583343722072e-05,
      "loss": 0.6215,
      "num_tokens": 1671492358.0,
      "step": 1595
    },
    {
      "epoch": 2.5316455696202533,
      "grad_norm": 0.34279988720816984,
      "learning_rate": 2.8888346081858858e-05,
      "loss": 0.6176,
      "num_tokens": 1676735238.0,
      "step": 1600
    },
    {
      "epoch": 2.5395569620253164,
      "grad_norm": 0.3445942670476214,
      "learning_rate": 2.8770820714986647e-05,
      "loss": 0.6256,
      "num_tokens": 1681960978.0,
      "step": 1605
    },
    {
      "epoch": 2.5474683544303796,
      "grad_norm": 0.3249239294351048,
      "learning_rate": 2.8653260554329643e-05,
      "loss": 0.6229,
      "num_tokens": 1687203858.0,
      "step": 1610
    },
    {
      "epoch": 2.555379746835443,
      "grad_norm": 0.4195352553077635,
      "learning_rate": 2.8535668818566065e-05,
      "loss": 0.6209,
      "num_tokens": 1692446738.0,
      "step": 1615
    },
    {
      "epoch": 2.5632911392405062,
      "grad_norm": 0.3148180466101991,
      "learning_rate": 2.8418048727238585e-05,
      "loss": 0.6243,
      "num_tokens": 1697685701.0,
      "step": 1620
    },
    {
      "epoch": 2.5712025316455698,
      "grad_norm": 0.3145292653905644,
      "learning_rate": 2.8300403500666256e-05,
      "loss": 0.6118,
      "num_tokens": 1702928581.0,
      "step": 1625
    },
    {
      "epoch": 2.579113924050633,
      "grad_norm": 0.34661533020564694,
      "learning_rate": 2.81827363598563e-05,
      "loss": 0.6209,
      "num_tokens": 1708171461.0,
      "step": 1630
    },
    {
      "epoch": 2.587025316455696,
      "grad_norm": 0.32721458461346575,
      "learning_rate": 2.8065050526415916e-05,
      "loss": 0.6231,
      "num_tokens": 1713414341.0,
      "step": 1635
    },
    {
      "epoch": 2.5949367088607596,
      "grad_norm": 0.3293080910499085,
      "learning_rate": 2.794734922246408e-05,
      "loss": 0.6227,
      "num_tokens": 1718657221.0,
      "step": 1640
    },
    {
      "epoch": 2.6028481012658227,
      "grad_norm": 0.3006610821739955,
      "learning_rate": 2.7829635670543375e-05,
      "loss": 0.6198,
      "num_tokens": 1723897509.0,
      "step": 1645
    },
    {
      "epoch": 2.6107594936708862,
      "grad_norm": 0.3221265574540506,
      "learning_rate": 2.7711913093531682e-05,
      "loss": 0.6215,
      "num_tokens": 1729134925.0,
      "step": 1650
    },
    {
      "epoch": 2.6186708860759493,
      "grad_norm": 0.3629658305187627,
      "learning_rate": 2.7594184714554e-05,
      "loss": 0.6193,
      "num_tokens": 1734377805.0,
      "step": 1655
    },
    {
      "epoch": 2.6265822784810124,
      "grad_norm": 0.3305255086161955,
      "learning_rate": 2.7476453756894173e-05,
      "loss": 0.6163,
      "num_tokens": 1739620685.0,
      "step": 1660
    },
    {
      "epoch": 2.634493670886076,
      "grad_norm": 0.30385118713526643,
      "learning_rate": 2.7358723443906653e-05,
      "loss": 0.6168,
      "num_tokens": 1744863565.0,
      "step": 1665
    },
    {
      "epoch": 2.642405063291139,
      "grad_norm": 0.3729938271215438,
      "learning_rate": 2.724099699892822e-05,
      "loss": 0.6247,
      "num_tokens": 1750106445.0,
      "step": 1670
    },
    {
      "epoch": 2.6503164556962027,
      "grad_norm": 0.3363024823177763,
      "learning_rate": 2.712327764518979e-05,
      "loss": 0.6227,
      "num_tokens": 1755349325.0,
      "step": 1675
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.3055930853692746,
      "learning_rate": 2.7005568605728087e-05,
      "loss": 0.6229,
      "num_tokens": 1760592205.0,
      "step": 1680
    },
    {
      "epoch": 2.666139240506329,
      "grad_norm": 0.3298226038292838,
      "learning_rate": 2.6887873103297474e-05,
      "loss": 0.6214,
      "num_tokens": 1765835085.0,
      "step": 1685
    },
    {
      "epoch": 2.6740506329113924,
      "grad_norm": 0.2867154405096028,
      "learning_rate": 2.6770194360281653e-05,
      "loss": 0.6199,
      "num_tokens": 1771077965.0,
      "step": 1690
    },
    {
      "epoch": 2.6819620253164556,
      "grad_norm": 0.3195638173711784,
      "learning_rate": 2.6652535598605504e-05,
      "loss": 0.6191,
      "num_tokens": 1776320845.0,
      "step": 1695
    },
    {
      "epoch": 2.689873417721519,
      "grad_norm": 0.32434124204868575,
      "learning_rate": 2.6534900039646816e-05,
      "loss": 0.6181,
      "num_tokens": 1781532860.0,
      "step": 1700
    },
    {
      "epoch": 2.6977848101265822,
      "grad_norm": 0.2918992438055131,
      "learning_rate": 2.6417290904148116e-05,
      "loss": 0.6197,
      "num_tokens": 1786775740.0,
      "step": 1705
    },
    {
      "epoch": 2.7056962025316453,
      "grad_norm": 0.3055847287718812,
      "learning_rate": 2.629971141212848e-05,
      "loss": 0.6173,
      "num_tokens": 1792018620.0,
      "step": 1710
    },
    {
      "epoch": 2.713607594936709,
      "grad_norm": 0.2985953721458501,
      "learning_rate": 2.618216478279538e-05,
      "loss": 0.6165,
      "num_tokens": 1797261500.0,
      "step": 1715
    },
    {
      "epoch": 2.721518987341772,
      "grad_norm": 0.3212918053380019,
      "learning_rate": 2.6064654234456554e-05,
      "loss": 0.6214,
      "num_tokens": 1802504380.0,
      "step": 1720
    },
    {
      "epoch": 2.7294303797468356,
      "grad_norm": 0.2919099745172365,
      "learning_rate": 2.5947182984431827e-05,
      "loss": 0.6255,
      "num_tokens": 1807747260.0,
      "step": 1725
    },
    {
      "epoch": 2.7373417721518987,
      "grad_norm": 0.2860751140237954,
      "learning_rate": 2.5829754248965154e-05,
      "loss": 0.6189,
      "num_tokens": 1812990140.0,
      "step": 1730
    },
    {
      "epoch": 2.745253164556962,
      "grad_norm": 0.3019949199367483,
      "learning_rate": 2.571237124313643e-05,
      "loss": 0.6124,
      "num_tokens": 1818233020.0,
      "step": 1735
    },
    {
      "epoch": 2.7531645569620253,
      "grad_norm": 0.3187461502303022,
      "learning_rate": 2.5595037180773524e-05,
      "loss": 0.6196,
      "num_tokens": 1823475900.0,
      "step": 1740
    },
    {
      "epoch": 2.7610759493670884,
      "grad_norm": 0.31735001215836417,
      "learning_rate": 2.5477755274364308e-05,
      "loss": 0.623,
      "num_tokens": 1828718780.0,
      "step": 1745
    },
    {
      "epoch": 2.768987341772152,
      "grad_norm": 0.33876769709652277,
      "learning_rate": 2.5360528734968647e-05,
      "loss": 0.611,
      "num_tokens": 1833961660.0,
      "step": 1750
    },
    {
      "epoch": 2.776898734177215,
      "grad_norm": 0.3542233915904492,
      "learning_rate": 2.5243360772130532e-05,
      "loss": 0.625,
      "num_tokens": 1839200698.0,
      "step": 1755
    },
    {
      "epoch": 2.7848101265822782,
      "grad_norm": 0.3630368367495965,
      "learning_rate": 2.5126254593790182e-05,
      "loss": 0.6179,
      "num_tokens": 1844443578.0,
      "step": 1760
    },
    {
      "epoch": 2.7927215189873418,
      "grad_norm": 0.3826705315790488,
      "learning_rate": 2.500921340619621e-05,
      "loss": 0.6205,
      "num_tokens": 1849680596.0,
      "step": 1765
    },
    {
      "epoch": 2.8006329113924053,
      "grad_norm": 0.3644864833948713,
      "learning_rate": 2.489224041381787e-05,
      "loss": 0.6156,
      "num_tokens": 1854907408.0,
      "step": 1770
    },
    {
      "epoch": 2.8085443037974684,
      "grad_norm": 0.3386672412828899,
      "learning_rate": 2.477533881925727e-05,
      "loss": 0.6238,
      "num_tokens": 1860134742.0,
      "step": 1775
    },
    {
      "epoch": 2.8164556962025316,
      "grad_norm": 0.2921538507032039,
      "learning_rate": 2.465851182316176e-05,
      "loss": 0.6202,
      "num_tokens": 1865377622.0,
      "step": 1780
    },
    {
      "epoch": 2.8243670886075947,
      "grad_norm": 0.28494983081811437,
      "learning_rate": 2.4541762624136212e-05,
      "loss": 0.6176,
      "num_tokens": 1870620502.0,
      "step": 1785
    },
    {
      "epoch": 2.8322784810126582,
      "grad_norm": 0.29021546121305103,
      "learning_rate": 2.4425094418655526e-05,
      "loss": 0.6177,
      "num_tokens": 1875863382.0,
      "step": 1790
    },
    {
      "epoch": 2.8401898734177218,
      "grad_norm": 0.27953269634031763,
      "learning_rate": 2.430851040097707e-05,
      "loss": 0.6158,
      "num_tokens": 1881101588.0,
      "step": 1795
    },
    {
      "epoch": 2.848101265822785,
      "grad_norm": 0.28367003015841197,
      "learning_rate": 2.4192013763053222e-05,
      "loss": 0.6151,
      "num_tokens": 1886333356.0,
      "step": 1800
    },
    {
      "epoch": 2.856012658227848,
      "grad_norm": 0.30695047341331355,
      "learning_rate": 2.4075607694444014e-05,
      "loss": 0.615,
      "num_tokens": 1891576236.0,
      "step": 1805
    },
    {
      "epoch": 2.8639240506329116,
      "grad_norm": 0.3389129881606512,
      "learning_rate": 2.395929538222978e-05,
      "loss": 0.6194,
      "num_tokens": 1896816113.0,
      "step": 1810
    },
    {
      "epoch": 2.8718354430379747,
      "grad_norm": 0.2993951418009312,
      "learning_rate": 2.3843080010923873e-05,
      "loss": 0.6151,
      "num_tokens": 1902045462.0,
      "step": 1815
    },
    {
      "epoch": 2.879746835443038,
      "grad_norm": 0.2747247961261455,
      "learning_rate": 2.3726964762385546e-05,
      "loss": 0.6152,
      "num_tokens": 1907275468.0,
      "step": 1820
    },
    {
      "epoch": 2.8876582278481013,
      "grad_norm": 0.27656654436718103,
      "learning_rate": 2.361095281573274e-05,
      "loss": 0.6204,
      "num_tokens": 1912517035.0,
      "step": 1825
    },
    {
      "epoch": 2.8955696202531644,
      "grad_norm": 0.26450516299877097,
      "learning_rate": 2.3495047347255122e-05,
      "loss": 0.6141,
      "num_tokens": 1917752379.0,
      "step": 1830
    },
    {
      "epoch": 2.903481012658228,
      "grad_norm": 0.2680090640698674,
      "learning_rate": 2.3379251530327112e-05,
      "loss": 0.6098,
      "num_tokens": 1922995259.0,
      "step": 1835
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 0.3020470058551071,
      "learning_rate": 2.326356853532094e-05,
      "loss": 0.6121,
      "num_tokens": 1928238139.0,
      "step": 1840
    },
    {
      "epoch": 2.9193037974683547,
      "grad_norm": 0.35207295628986174,
      "learning_rate": 2.3148001529519918e-05,
      "loss": 0.6176,
      "num_tokens": 1933481019.0,
      "step": 1845
    },
    {
      "epoch": 2.9272151898734178,
      "grad_norm": 0.31909804688235877,
      "learning_rate": 2.303255367703169e-05,
      "loss": 0.6206,
      "num_tokens": 1938714322.0,
      "step": 1850
    },
    {
      "epoch": 2.935126582278481,
      "grad_norm": 0.30462726559520964,
      "learning_rate": 2.2917228138701596e-05,
      "loss": 0.6154,
      "num_tokens": 1943957202.0,
      "step": 1855
    },
    {
      "epoch": 2.9430379746835444,
      "grad_norm": 0.32103060565165314,
      "learning_rate": 2.2802028072026115e-05,
      "loss": 0.6089,
      "num_tokens": 1949191447.0,
      "step": 1860
    },
    {
      "epoch": 2.9509493670886076,
      "grad_norm": 0.3062127051035162,
      "learning_rate": 2.268695663106648e-05,
      "loss": 0.6127,
      "num_tokens": 1954434327.0,
      "step": 1865
    },
    {
      "epoch": 2.958860759493671,
      "grad_norm": 0.3239852849229511,
      "learning_rate": 2.257201696636225e-05,
      "loss": 0.615,
      "num_tokens": 1959677207.0,
      "step": 1870
    },
    {
      "epoch": 2.9667721518987342,
      "grad_norm": 0.2893785533195971,
      "learning_rate": 2.2457212224845115e-05,
      "loss": 0.6182,
      "num_tokens": 1964920087.0,
      "step": 1875
    },
    {
      "epoch": 2.9746835443037973,
      "grad_norm": 0.3565583813030718,
      "learning_rate": 2.2342545549752658e-05,
      "loss": 0.6187,
      "num_tokens": 1970162967.0,
      "step": 1880
    },
    {
      "epoch": 2.982594936708861,
      "grad_norm": 0.2901982983640066,
      "learning_rate": 2.2228020080542394e-05,
      "loss": 0.6155,
      "num_tokens": 1975405847.0,
      "step": 1885
    },
    {
      "epoch": 2.990506329113924,
      "grad_norm": 0.2552747365991554,
      "learning_rate": 2.211363895280573e-05,
      "loss": 0.617,
      "num_tokens": 1980648163.0,
      "step": 1890
    },
    {
      "epoch": 2.9984177215189876,
      "grad_norm": 0.2925757603065985,
      "learning_rate": 2.1999405298182176e-05,
      "loss": 0.6114,
      "num_tokens": 1985885198.0,
      "step": 1895
    },
    {
      "epoch": 3.0063291139240507,
      "grad_norm": 0.3169901185514265,
      "learning_rate": 2.1885322244273537e-05,
      "loss": 0.6001,
      "num_tokens": 1991128078.0,
      "step": 1900
    },
    {
      "epoch": 3.0142405063291138,
      "grad_norm": 0.2733791677119331,
      "learning_rate": 2.177139291455837e-05,
      "loss": 0.5909,
      "num_tokens": 1996370958.0,
      "step": 1905
    },
    {
      "epoch": 3.0221518987341773,
      "grad_norm": 0.2969117337975924,
      "learning_rate": 2.1657620428306386e-05,
      "loss": 0.5979,
      "num_tokens": 2001613838.0,
      "step": 1910
    },
    {
      "epoch": 3.0300632911392404,
      "grad_norm": 0.259222996285033,
      "learning_rate": 2.154400790049309e-05,
      "loss": 0.5935,
      "num_tokens": 2006856718.0,
      "step": 1915
    },
    {
      "epoch": 3.037974683544304,
      "grad_norm": 0.29751625615948685,
      "learning_rate": 2.1430558441714488e-05,
      "loss": 0.5892,
      "num_tokens": 2012099598.0,
      "step": 1920
    },
    {
      "epoch": 3.045886075949367,
      "grad_norm": 0.2808339111842118,
      "learning_rate": 2.131727515810193e-05,
      "loss": 0.5881,
      "num_tokens": 2017342478.0,
      "step": 1925
    },
    {
      "epoch": 3.0537974683544302,
      "grad_norm": 0.2665237486778169,
      "learning_rate": 2.1204161151237044e-05,
      "loss": 0.5869,
      "num_tokens": 2022585358.0,
      "step": 1930
    },
    {
      "epoch": 3.0617088607594938,
      "grad_norm": 0.28147532434254685,
      "learning_rate": 2.1091219518066846e-05,
      "loss": 0.5857,
      "num_tokens": 2027828238.0,
      "step": 1935
    },
    {
      "epoch": 3.069620253164557,
      "grad_norm": 0.27827204435611375,
      "learning_rate": 2.0978453350818923e-05,
      "loss": 0.5892,
      "num_tokens": 2033071118.0,
      "step": 1940
    },
    {
      "epoch": 3.0775316455696204,
      "grad_norm": 0.27504620542469616,
      "learning_rate": 2.0865865736916767e-05,
      "loss": 0.5927,
      "num_tokens": 2038313998.0,
      "step": 1945
    },
    {
      "epoch": 3.0854430379746836,
      "grad_norm": 0.2625667359010651,
      "learning_rate": 2.0753459758895305e-05,
      "loss": 0.5878,
      "num_tokens": 2043556878.0,
      "step": 1950
    },
    {
      "epoch": 3.0933544303797467,
      "grad_norm": 0.31412193685640266,
      "learning_rate": 2.0641238494316412e-05,
      "loss": 0.5954,
      "num_tokens": 2048795916.0,
      "step": 1955
    },
    {
      "epoch": 3.1012658227848102,
      "grad_norm": 0.2551742932124429,
      "learning_rate": 2.0529205015684733e-05,
      "loss": 0.5896,
      "num_tokens": 2054038796.0,
      "step": 1960
    },
    {
      "epoch": 3.1091772151898733,
      "grad_norm": 0.27107418641432374,
      "learning_rate": 2.0417362390363497e-05,
      "loss": 0.5949,
      "num_tokens": 2059281676.0,
      "step": 1965
    },
    {
      "epoch": 3.117088607594937,
      "grad_norm": 0.2910292647945281,
      "learning_rate": 2.030571368049058e-05,
      "loss": 0.593,
      "num_tokens": 2064524556.0,
      "step": 1970
    },
    {
      "epoch": 3.125,
      "grad_norm": 0.2702497087199772,
      "learning_rate": 2.0194261942894628e-05,
      "loss": 0.5913,
      "num_tokens": 2069763519.0,
      "step": 1975
    },
    {
      "epoch": 3.132911392405063,
      "grad_norm": 0.27879096153789196,
      "learning_rate": 2.0083010229011405e-05,
      "loss": 0.5871,
      "num_tokens": 2075006399.0,
      "step": 1980
    },
    {
      "epoch": 3.1408227848101267,
      "grad_norm": 0.2760119851379016,
      "learning_rate": 1.9971961584800218e-05,
      "loss": 0.5889,
      "num_tokens": 2080249279.0,
      "step": 1985
    },
    {
      "epoch": 3.1487341772151898,
      "grad_norm": 0.2980189742569214,
      "learning_rate": 1.986111905066055e-05,
      "loss": 0.5936,
      "num_tokens": 2085492159.0,
      "step": 1990
    },
    {
      "epoch": 3.1566455696202533,
      "grad_norm": 0.27221033878781475,
      "learning_rate": 1.9750485661348762e-05,
      "loss": 0.5925,
      "num_tokens": 2090735039.0,
      "step": 1995
    },
    {
      "epoch": 3.1645569620253164,
      "grad_norm": 0.282708489024635,
      "learning_rate": 1.964006444589509e-05,
      "loss": 0.5906,
      "num_tokens": 2095977919.0,
      "step": 2000
    },
    {
      "epoch": 3.1724683544303796,
      "grad_norm": 0.2689336862752935,
      "learning_rate": 1.952985842752062e-05,
      "loss": 0.5908,
      "num_tokens": 2101206695.0,
      "step": 2005
    },
    {
      "epoch": 3.180379746835443,
      "grad_norm": 0.26260733949618775,
      "learning_rate": 1.941987062355458e-05,
      "loss": 0.5907,
      "num_tokens": 2106449575.0,
      "step": 2010
    },
    {
      "epoch": 3.1882911392405062,
      "grad_norm": 0.2829230017612488,
      "learning_rate": 1.9310104045351724e-05,
      "loss": 0.5869,
      "num_tokens": 2111692455.0,
      "step": 2015
    },
    {
      "epoch": 3.1962025316455698,
      "grad_norm": 0.2496685091023237,
      "learning_rate": 1.9200561698209828e-05,
      "loss": 0.5895,
      "num_tokens": 2116935335.0,
      "step": 2020
    },
    {
      "epoch": 3.204113924050633,
      "grad_norm": 0.30588754751321845,
      "learning_rate": 1.9091246581287487e-05,
      "loss": 0.5911,
      "num_tokens": 2122178215.0,
      "step": 2025
    },
    {
      "epoch": 3.212025316455696,
      "grad_norm": 0.3013756834853421,
      "learning_rate": 1.8982161687521938e-05,
      "loss": 0.5926,
      "num_tokens": 2127416178.0,
      "step": 2030
    },
    {
      "epoch": 3.2199367088607596,
      "grad_norm": 0.3260497257360817,
      "learning_rate": 1.8873310003547152e-05,
      "loss": 0.5957,
      "num_tokens": 2132659058.0,
      "step": 2035
    },
    {
      "epoch": 3.2278481012658227,
      "grad_norm": 18.37765842944507,
      "learning_rate": 1.8764694509612057e-05,
      "loss": 0.6045,
      "num_tokens": 2137900029.0,
      "step": 2040
    },
    {
      "epoch": 3.2357594936708862,
      "grad_norm": 0.2844360272669497,
      "learning_rate": 1.865631817949893e-05,
      "loss": 0.5961,
      "num_tokens": 2143139652.0,
      "step": 2045
    },
    {
      "epoch": 3.2436708860759493,
      "grad_norm": 0.25708276328077995,
      "learning_rate": 1.8548183980441985e-05,
      "loss": 0.5988,
      "num_tokens": 2148382532.0,
      "step": 2050
    },
    {
      "epoch": 3.2515822784810124,
      "grad_norm": 0.27434743791141486,
      "learning_rate": 1.8440294873046126e-05,
      "loss": 0.5904,
      "num_tokens": 2153625412.0,
      "step": 2055
    },
    {
      "epoch": 3.259493670886076,
      "grad_norm": 0.25975458422652054,
      "learning_rate": 1.8332653811205906e-05,
      "loss": 0.5928,
      "num_tokens": 2158854105.0,
      "step": 2060
    },
    {
      "epoch": 3.267405063291139,
      "grad_norm": 0.26925828595298595,
      "learning_rate": 1.8225263742024638e-05,
      "loss": 0.5833,
      "num_tokens": 2164096985.0,
      "step": 2065
    },
    {
      "epoch": 3.2753164556962027,
      "grad_norm": 0.282185762273433,
      "learning_rate": 1.8118127605733697e-05,
      "loss": 0.5981,
      "num_tokens": 2169337273.0,
      "step": 2070
    },
    {
      "epoch": 3.2832278481012658,
      "grad_norm": 0.25816071189825335,
      "learning_rate": 1.801124833561208e-05,
      "loss": 0.5961,
      "num_tokens": 2174564736.0,
      "step": 2075
    },
    {
      "epoch": 3.291139240506329,
      "grad_norm": 0.24619197376885332,
      "learning_rate": 1.7904628857905982e-05,
      "loss": 0.6011,
      "num_tokens": 2179797828.0,
      "step": 2080
    },
    {
      "epoch": 3.2990506329113924,
      "grad_norm": 0.24421414534208322,
      "learning_rate": 1.7798272091748785e-05,
      "loss": 0.5865,
      "num_tokens": 2185040708.0,
      "step": 2085
    },
    {
      "epoch": 3.3069620253164556,
      "grad_norm": 0.2524526157340073,
      "learning_rate": 1.7692180949081093e-05,
      "loss": 0.5959,
      "num_tokens": 2190283588.0,
      "step": 2090
    },
    {
      "epoch": 3.314873417721519,
      "grad_norm": 0.27130450376385096,
      "learning_rate": 1.758635833457099e-05,
      "loss": 0.5848,
      "num_tokens": 2195513594.0,
      "step": 2095
    },
    {
      "epoch": 3.3227848101265822,
      "grad_norm": 0.26771998653275103,
      "learning_rate": 1.748080714553455e-05,
      "loss": 0.5879,
      "num_tokens": 2200745342.0,
      "step": 2100
    },
    {
      "epoch": 3.3306962025316453,
      "grad_norm": 0.26138164229192307,
      "learning_rate": 1.7375530271856487e-05,
      "loss": 0.5922,
      "num_tokens": 2205988222.0,
      "step": 2105
    },
    {
      "epoch": 3.338607594936709,
      "grad_norm": 0.2715626154228517,
      "learning_rate": 1.727053059591101e-05,
      "loss": 0.5971,
      "num_tokens": 2211231102.0,
      "step": 2110
    },
    {
      "epoch": 3.346518987341772,
      "grad_norm": 0.2455227895466599,
      "learning_rate": 1.7165810992482994e-05,
      "loss": 0.5887,
      "num_tokens": 2216460869.0,
      "step": 2115
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 0.2777380961416323,
      "learning_rate": 1.706137432868917e-05,
      "loss": 0.5943,
      "num_tokens": 2221703749.0,
      "step": 2120
    },
    {
      "epoch": 3.3623417721518987,
      "grad_norm": 0.25563580174980555,
      "learning_rate": 1.6957223463899658e-05,
      "loss": 0.5897,
      "num_tokens": 2226946629.0,
      "step": 2125
    },
    {
      "epoch": 3.370253164556962,
      "grad_norm": 0.25164624412979186,
      "learning_rate": 1.6853361249659722e-05,
      "loss": 0.5929,
      "num_tokens": 2232189509.0,
      "step": 2130
    },
    {
      "epoch": 3.3781645569620253,
      "grad_norm": 0.2806108194410442,
      "learning_rate": 1.674979052961166e-05,
      "loss": 0.5975,
      "num_tokens": 2237432389.0,
      "step": 2135
    },
    {
      "epoch": 3.3860759493670884,
      "grad_norm": 0.2605837907585764,
      "learning_rate": 1.664651413941693e-05,
      "loss": 0.5907,
      "num_tokens": 2242665858.0,
      "step": 2140
    },
    {
      "epoch": 3.393987341772152,
      "grad_norm": 0.2598679967531081,
      "learning_rate": 1.6543534906678552e-05,
      "loss": 0.592,
      "num_tokens": 2247908738.0,
      "step": 2145
    },
    {
      "epoch": 3.401898734177215,
      "grad_norm": 0.25797499837806687,
      "learning_rate": 1.6440855650863695e-05,
      "loss": 0.5894,
      "num_tokens": 2253151618.0,
      "step": 2150
    },
    {
      "epoch": 3.4098101265822787,
      "grad_norm": 0.2427493258825347,
      "learning_rate": 1.6338479183226436e-05,
      "loss": 0.5924,
      "num_tokens": 2258394498.0,
      "step": 2155
    },
    {
      "epoch": 3.4177215189873418,
      "grad_norm": 0.2359671364270129,
      "learning_rate": 1.6236408306730828e-05,
      "loss": 0.5916,
      "num_tokens": 2263637378.0,
      "step": 2160
    },
    {
      "epoch": 3.425632911392405,
      "grad_norm": 0.23492984902262662,
      "learning_rate": 1.6134645815974153e-05,
      "loss": 0.588,
      "num_tokens": 2268880258.0,
      "step": 2165
    },
    {
      "epoch": 3.4335443037974684,
      "grad_norm": 0.2684129651198138,
      "learning_rate": 1.6033194497110387e-05,
      "loss": 0.5863,
      "num_tokens": 2274107070.0,
      "step": 2170
    },
    {
      "epoch": 3.4414556962025316,
      "grad_norm": 0.2551028754173729,
      "learning_rate": 1.5932057127773956e-05,
      "loss": 0.5912,
      "num_tokens": 2279335709.0,
      "step": 2175
    },
    {
      "epoch": 3.449367088607595,
      "grad_norm": 0.27552281820576113,
      "learning_rate": 1.5831236477003657e-05,
      "loss": 0.5901,
      "num_tokens": 2284578589.0,
      "step": 2180
    },
    {
      "epoch": 3.4572784810126582,
      "grad_norm": 0.245355717229245,
      "learning_rate": 1.573073530516683e-05,
      "loss": 0.5984,
      "num_tokens": 2289805799.0,
      "step": 2185
    },
    {
      "epoch": 3.4651898734177213,
      "grad_norm": 0.2711024543904534,
      "learning_rate": 1.563055636388385e-05,
      "loss": 0.5901,
      "num_tokens": 2295041668.0,
      "step": 2190
    },
    {
      "epoch": 3.473101265822785,
      "grad_norm": 0.25534456241862863,
      "learning_rate": 1.5530702395952726e-05,
      "loss": 0.5885,
      "num_tokens": 2300284548.0,
      "step": 2195
    },
    {
      "epoch": 3.481012658227848,
      "grad_norm": 0.28472473513191726,
      "learning_rate": 1.5431176135274004e-05,
      "loss": 0.5875,
      "num_tokens": 2305527428.0,
      "step": 2200
    },
    {
      "epoch": 3.4889240506329116,
      "grad_norm": 0.26401107803555085,
      "learning_rate": 1.533198030677596e-05,
      "loss": 0.587,
      "num_tokens": 2310770308.0,
      "step": 2205
    },
    {
      "epoch": 3.4968354430379747,
      "grad_norm": 0.23581382627455794,
      "learning_rate": 1.5233117626339988e-05,
      "loss": 0.5829,
      "num_tokens": 2316013188.0,
      "step": 2210
    },
    {
      "epoch": 3.504746835443038,
      "grad_norm": 0.2919647262861195,
      "learning_rate": 1.513459080072617e-05,
      "loss": 0.593,
      "num_tokens": 2321256068.0,
      "step": 2215
    },
    {
      "epoch": 3.5126582278481013,
      "grad_norm": 0.27795974121603223,
      "learning_rate": 1.503640252749928e-05,
      "loss": 0.5948,
      "num_tokens": 2326498948.0,
      "step": 2220
    },
    {
      "epoch": 3.5205696202531644,
      "grad_norm": 0.2513485774758495,
      "learning_rate": 1.4938555494954838e-05,
      "loss": 0.5928,
      "num_tokens": 2331740515.0,
      "step": 2225
    },
    {
      "epoch": 3.528481012658228,
      "grad_norm": 0.27468633702239875,
      "learning_rate": 1.484105238204555e-05,
      "loss": 0.592,
      "num_tokens": 2336983395.0,
      "step": 2230
    },
    {
      "epoch": 3.536392405063291,
      "grad_norm": 0.24973282593959645,
      "learning_rate": 1.4743895858307951e-05,
      "loss": 0.5858,
      "num_tokens": 2342226275.0,
      "step": 2235
    },
    {
      "epoch": 3.5443037974683547,
      "grad_norm": 0.236224838159416,
      "learning_rate": 1.4647088583789325e-05,
      "loss": 0.5943,
      "num_tokens": 2347461462.0,
      "step": 2240
    },
    {
      "epoch": 3.5522151898734178,
      "grad_norm": 0.25744574090772937,
      "learning_rate": 1.4550633208974832e-05,
      "loss": 0.5888,
      "num_tokens": 2352704342.0,
      "step": 2245
    },
    {
      "epoch": 3.560126582278481,
      "grad_norm": 0.24904139861794714,
      "learning_rate": 1.445453237471501e-05,
      "loss": 0.5937,
      "num_tokens": 2357947222.0,
      "step": 2250
    },
    {
      "epoch": 3.5680379746835444,
      "grad_norm": 0.26268941423266656,
      "learning_rate": 1.435878871215342e-05,
      "loss": 0.5903,
      "num_tokens": 2363190102.0,
      "step": 2255
    },
    {
      "epoch": 3.5759493670886076,
      "grad_norm": 0.24790895208789265,
      "learning_rate": 1.4263404842654638e-05,
      "loss": 0.5972,
      "num_tokens": 2368432982.0,
      "step": 2260
    },
    {
      "epoch": 3.583860759493671,
      "grad_norm": 0.2613242366100201,
      "learning_rate": 1.416838337773243e-05,
      "loss": 0.595,
      "num_tokens": 2373661081.0,
      "step": 2265
    },
    {
      "epoch": 3.5917721518987342,
      "grad_norm": 0.2922830460241977,
      "learning_rate": 1.4073726918978347e-05,
      "loss": 0.594,
      "num_tokens": 2378903961.0,
      "step": 2270
    },
    {
      "epoch": 3.5996835443037973,
      "grad_norm": 0.27101780547058585,
      "learning_rate": 1.3979438057990399e-05,
      "loss": 0.592,
      "num_tokens": 2384146841.0,
      "step": 2275
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.26512432923141993,
      "learning_rate": 1.3885519376302159e-05,
      "loss": 0.5878,
      "num_tokens": 2389389721.0,
      "step": 2280
    },
    {
      "epoch": 3.615506329113924,
      "grad_norm": 0.23952238274180007,
      "learning_rate": 1.3791973445312058e-05,
      "loss": 0.589,
      "num_tokens": 2394632601.0,
      "step": 2285
    },
    {
      "epoch": 3.6234177215189876,
      "grad_norm": 0.25423971405929885,
      "learning_rate": 1.3698802826213008e-05,
      "loss": 0.5908,
      "num_tokens": 2399860042.0,
      "step": 2290
    },
    {
      "epoch": 3.6313291139240507,
      "grad_norm": 0.24740192193842994,
      "learning_rate": 1.3606010069922232e-05,
      "loss": 0.5882,
      "num_tokens": 2405102922.0,
      "step": 2295
    },
    {
      "epoch": 3.6392405063291138,
      "grad_norm": 0.5035309107226975,
      "learning_rate": 1.351359771701147e-05,
      "loss": 0.5966,
      "num_tokens": 2410345802.0,
      "step": 2300
    },
    {
      "epoch": 3.6471518987341773,
      "grad_norm": 0.231935472046041,
      "learning_rate": 1.3421568297637404e-05,
      "loss": 0.5873,
      "num_tokens": 2415588682.0,
      "step": 2305
    },
    {
      "epoch": 3.6550632911392404,
      "grad_norm": 0.2584802973832586,
      "learning_rate": 1.3329924331472376e-05,
      "loss": 0.5925,
      "num_tokens": 2420806919.0,
      "step": 2310
    },
    {
      "epoch": 3.662974683544304,
      "grad_norm": 0.31314246833170745,
      "learning_rate": 1.3238668327635422e-05,
      "loss": 0.5922,
      "num_tokens": 2426049799.0,
      "step": 2315
    },
    {
      "epoch": 3.670886075949367,
      "grad_norm": 0.2663616787106021,
      "learning_rate": 1.3147802784623536e-05,
      "loss": 0.5962,
      "num_tokens": 2431292679.0,
      "step": 2320
    },
    {
      "epoch": 3.6787974683544302,
      "grad_norm": 0.23711780957648615,
      "learning_rate": 1.3057330190243317e-05,
      "loss": 0.5894,
      "num_tokens": 2436523336.0,
      "step": 2325
    },
    {
      "epoch": 3.6867088607594938,
      "grad_norm": 0.2549540114874716,
      "learning_rate": 1.2967253021542813e-05,
      "loss": 0.5957,
      "num_tokens": 2441766216.0,
      "step": 2330
    },
    {
      "epoch": 3.694620253164557,
      "grad_norm": 0.24052219589182647,
      "learning_rate": 1.2877573744743737e-05,
      "loss": 0.5935,
      "num_tokens": 2447009096.0,
      "step": 2335
    },
    {
      "epoch": 3.7025316455696204,
      "grad_norm": 0.23451782912483585,
      "learning_rate": 1.278829481517388e-05,
      "loss": 0.5907,
      "num_tokens": 2452248973.0,
      "step": 2340
    },
    {
      "epoch": 3.7104430379746836,
      "grad_norm": 0.2572531172651155,
      "learning_rate": 1.2699418677199992e-05,
      "loss": 0.5871,
      "num_tokens": 2457491853.0,
      "step": 2345
    },
    {
      "epoch": 3.7183544303797467,
      "grad_norm": 0.22205564327904748,
      "learning_rate": 1.2610947764160743e-05,
      "loss": 0.5911,
      "num_tokens": 2462732233.0,
      "step": 2350
    },
    {
      "epoch": 3.7262658227848102,
      "grad_norm": 0.25061314597163403,
      "learning_rate": 1.2522884498300185e-05,
      "loss": 0.589,
      "num_tokens": 2467975113.0,
      "step": 2355
    },
    {
      "epoch": 3.7341772151898733,
      "grad_norm": 0.24341650567160555,
      "learning_rate": 1.2435231290701398e-05,
      "loss": 0.5839,
      "num_tokens": 2473217993.0,
      "step": 2360
    },
    {
      "epoch": 3.742088607594937,
      "grad_norm": 0.23957280785298024,
      "learning_rate": 1.234799054122048e-05,
      "loss": 0.5856,
      "num_tokens": 2478460873.0,
      "step": 2365
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.23496985031896744,
      "learning_rate": 1.2261164638420832e-05,
      "loss": 0.5943,
      "num_tokens": 2483695118.0,
      "step": 2370
    },
    {
      "epoch": 3.757911392405063,
      "grad_norm": 0.2562958360759268,
      "learning_rate": 1.2174755959507785e-05,
      "loss": 0.5905,
      "num_tokens": 2488915999.0,
      "step": 2375
    },
    {
      "epoch": 3.7658227848101267,
      "grad_norm": 0.23818568022151865,
      "learning_rate": 1.2088766870263504e-05,
      "loss": 0.5833,
      "num_tokens": 2494158879.0,
      "step": 2380
    },
    {
      "epoch": 3.7737341772151898,
      "grad_norm": 0.271749751046968,
      "learning_rate": 1.2003199724982206e-05,
      "loss": 0.5898,
      "num_tokens": 2499392356.0,
      "step": 2385
    },
    {
      "epoch": 3.7816455696202533,
      "grad_norm": 0.2768085008184621,
      "learning_rate": 1.1918056866405717e-05,
      "loss": 0.5964,
      "num_tokens": 2504635236.0,
      "step": 2390
    },
    {
      "epoch": 3.7895569620253164,
      "grad_norm": 0.24209821208112645,
      "learning_rate": 1.1833340625659325e-05,
      "loss": 0.5907,
      "num_tokens": 2509878116.0,
      "step": 2395
    },
    {
      "epoch": 3.7974683544303796,
      "grad_norm": 0.2587091544484298,
      "learning_rate": 1.1749053322187933e-05,
      "loss": 0.5887,
      "num_tokens": 2515098012.0,
      "step": 2400
    },
    {
      "epoch": 3.805379746835443,
      "grad_norm": 0.241562313900529,
      "learning_rate": 1.1665197263692593e-05,
      "loss": 0.5914,
      "num_tokens": 2520340892.0,
      "step": 2405
    },
    {
      "epoch": 3.8132911392405062,
      "grad_norm": 0.21894193923034058,
      "learning_rate": 1.1581774746067315e-05,
      "loss": 0.5861,
      "num_tokens": 2525582367.0,
      "step": 2410
    },
    {
      "epoch": 3.8212025316455698,
      "grad_norm": 0.241635293617956,
      "learning_rate": 1.149878805333616e-05,
      "loss": 0.5875,
      "num_tokens": 2530819032.0,
      "step": 2415
    },
    {
      "epoch": 3.829113924050633,
      "grad_norm": 0.2377026625673294,
      "learning_rate": 1.1416239457590797e-05,
      "loss": 0.5895,
      "num_tokens": 2536061912.0,
      "step": 2420
    },
    {
      "epoch": 3.837025316455696,
      "grad_norm": 0.2500766928834093,
      "learning_rate": 1.1334131218928215e-05,
      "loss": 0.5955,
      "num_tokens": 2541294406.0,
      "step": 2425
    },
    {
      "epoch": 3.8449367088607596,
      "grad_norm": 0.24157809368048716,
      "learning_rate": 1.1252465585388861e-05,
      "loss": 0.593,
      "num_tokens": 2546537286.0,
      "step": 2430
    },
    {
      "epoch": 3.8528481012658227,
      "grad_norm": 0.23659644820965803,
      "learning_rate": 1.1171244792895122e-05,
      "loss": 0.5875,
      "num_tokens": 2551780166.0,
      "step": 2435
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 0.22665381320819925,
      "learning_rate": 1.1090471065190087e-05,
      "loss": 0.5895,
      "num_tokens": 2557023046.0,
      "step": 2440
    },
    {
      "epoch": 3.8686708860759493,
      "grad_norm": 0.2531032288030383,
      "learning_rate": 1.1010146613776646e-05,
      "loss": 0.5911,
      "num_tokens": 2562249118.0,
      "step": 2445
    },
    {
      "epoch": 3.8765822784810124,
      "grad_norm": 0.2375978750430171,
      "learning_rate": 1.0930273637856969e-05,
      "loss": 0.5913,
      "num_tokens": 2567491998.0,
      "step": 2450
    },
    {
      "epoch": 3.884493670886076,
      "grad_norm": 0.24286465990334033,
      "learning_rate": 1.0850854324272289e-05,
      "loss": 0.5877,
      "num_tokens": 2572734878.0,
      "step": 2455
    },
    {
      "epoch": 3.892405063291139,
      "grad_norm": 0.2575357182094295,
      "learning_rate": 1.0771890847443022e-05,
      "loss": 0.596,
      "num_tokens": 2577975987.0,
      "step": 2460
    },
    {
      "epoch": 3.9003164556962027,
      "grad_norm": 0.27638888789609667,
      "learning_rate": 1.069338536930923e-05,
      "loss": 0.5865,
      "num_tokens": 2583218867.0,
      "step": 2465
    },
    {
      "epoch": 3.9082278481012658,
      "grad_norm": 0.2344539051269508,
      "learning_rate": 1.0615340039271443e-05,
      "loss": 0.5892,
      "num_tokens": 2588461747.0,
      "step": 2470
    },
    {
      "epoch": 3.916139240506329,
      "grad_norm": 0.27080111151637076,
      "learning_rate": 1.0537756994131783e-05,
      "loss": 0.5895,
      "num_tokens": 2593704627.0,
      "step": 2475
    },
    {
      "epoch": 3.9240506329113924,
      "grad_norm": 0.2249920243150278,
      "learning_rate": 1.0460638358035496e-05,
      "loss": 0.5924,
      "num_tokens": 2598947507.0,
      "step": 2480
    },
    {
      "epoch": 3.9319620253164556,
      "grad_norm": 0.28780985267539305,
      "learning_rate": 1.0383986242412785e-05,
      "loss": 0.5871,
      "num_tokens": 2604190387.0,
      "step": 2485
    },
    {
      "epoch": 3.939873417721519,
      "grad_norm": 0.23427765653838825,
      "learning_rate": 1.0307802745920964e-05,
      "loss": 0.5839,
      "num_tokens": 2609433267.0,
      "step": 2490
    },
    {
      "epoch": 3.9477848101265822,
      "grad_norm": 0.2511572467034619,
      "learning_rate": 1.0232089954387068e-05,
      "loss": 0.591,
      "num_tokens": 2614676147.0,
      "step": 2495
    },
    {
      "epoch": 3.9556962025316453,
      "grad_norm": 0.24133671906166224,
      "learning_rate": 1.0156849940750688e-05,
      "loss": 0.5873,
      "num_tokens": 2619919027.0,
      "step": 2500
    },
    {
      "epoch": 3.963607594936709,
      "grad_norm": 0.24130871799729056,
      "learning_rate": 1.0082084765007226e-05,
      "loss": 0.5919,
      "num_tokens": 2625150643.0,
      "step": 2505
    },
    {
      "epoch": 3.971518987341772,
      "grad_norm": 0.23545792709918428,
      "learning_rate": 1.0007796474151514e-05,
      "loss": 0.5972,
      "num_tokens": 2630393523.0,
      "step": 2510
    },
    {
      "epoch": 3.9794303797468356,
      "grad_norm": 0.2506437876450052,
      "learning_rate": 9.933987102121764e-06,
      "loss": 0.5814,
      "num_tokens": 2635636403.0,
      "step": 2515
    },
    {
      "epoch": 3.9873417721518987,
      "grad_norm": 0.2284653998147006,
      "learning_rate": 9.860658669743861e-06,
      "loss": 0.5897,
      "num_tokens": 2640879283.0,
      "step": 2520
    },
    {
      "epoch": 3.995253164556962,
      "grad_norm": 0.22307394105006934,
      "learning_rate": 9.787813184676056e-06,
      "loss": 0.5861,
      "num_tokens": 2646099304.0,
      "step": 2525
    },
    {
      "epoch": 4.003164556962025,
      "grad_norm": 0.28089029948590827,
      "learning_rate": 9.715452641353992e-06,
      "loss": 0.5826,
      "num_tokens": 2651342184.0,
      "step": 2530
    },
    {
      "epoch": 4.011075949367089,
      "grad_norm": 0.23002963489023126,
      "learning_rate": 9.643579020936106e-06,
      "loss": 0.5662,
      "num_tokens": 2656585064.0,
      "step": 2535
    },
    {
      "epoch": 4.018987341772152,
      "grad_norm": 0.24076823558688779,
      "learning_rate": 9.572194291249362e-06,
      "loss": 0.5767,
      "num_tokens": 2661827944.0,
      "step": 2540
    },
    {
      "epoch": 4.026898734177215,
      "grad_norm": 0.22483920300780846,
      "learning_rate": 9.501300406735406e-06,
      "loss": 0.5743,
      "num_tokens": 2667070260.0,
      "step": 2545
    },
    {
      "epoch": 4.034810126582278,
      "grad_norm": 0.2425268841198098,
      "learning_rate": 9.430899308397024e-06,
      "loss": 0.5691,
      "num_tokens": 2672313140.0,
      "step": 2550
    },
    {
      "epoch": 4.042721518987341,
      "grad_norm": 0.22463870074019276,
      "learning_rate": 9.360992923745032e-06,
      "loss": 0.5727,
      "num_tokens": 2677556020.0,
      "step": 2555
    },
    {
      "epoch": 4.050632911392405,
      "grad_norm": 0.2257862359365115,
      "learning_rate": 9.29158316674548e-06,
      "loss": 0.5771,
      "num_tokens": 2682798900.0,
      "step": 2560
    },
    {
      "epoch": 4.0585443037974684,
      "grad_norm": 0.21855947078972188,
      "learning_rate": 9.222671937767247e-06,
      "loss": 0.5693,
      "num_tokens": 2688041780.0,
      "step": 2565
    },
    {
      "epoch": 4.0664556962025316,
      "grad_norm": 0.22655998729700472,
      "learning_rate": 9.154261123530024e-06,
      "loss": 0.5688,
      "num_tokens": 2693280743.0,
      "step": 2570
    },
    {
      "epoch": 4.074367088607595,
      "grad_norm": 0.2417803758293443,
      "learning_rate": 9.086352597052674e-06,
      "loss": 0.5699,
      "num_tokens": 2698523623.0,
      "step": 2575
    },
    {
      "epoch": 4.082278481012658,
      "grad_norm": 0.23491529993846316,
      "learning_rate": 9.018948217601894e-06,
      "loss": 0.5776,
      "num_tokens": 2703760050.0,
      "step": 2580
    },
    {
      "epoch": 4.090189873417722,
      "grad_norm": 0.21683310271118783,
      "learning_rate": 8.952049830641368e-06,
      "loss": 0.569,
      "num_tokens": 2709001853.0,
      "step": 2585
    },
    {
      "epoch": 4.098101265822785,
      "grad_norm": 0.22683758792329442,
      "learning_rate": 8.885659267781218e-06,
      "loss": 0.5723,
      "num_tokens": 2714244733.0,
      "step": 2590
    },
    {
      "epoch": 4.106012658227848,
      "grad_norm": 0.23440417433660363,
      "learning_rate": 8.819778346727839e-06,
      "loss": 0.571,
      "num_tokens": 2719487613.0,
      "step": 2595
    },
    {
      "epoch": 4.113924050632911,
      "grad_norm": 0.2192458553426772,
      "learning_rate": 8.754408871234168e-06,
      "loss": 0.5733,
      "num_tokens": 2724715076.0,
      "step": 2600
    },
    {
      "epoch": 4.121835443037975,
      "grad_norm": 0.2530876360164053,
      "learning_rate": 8.689552631050274e-06,
      "loss": 0.5714,
      "num_tokens": 2729957956.0,
      "step": 2605
    },
    {
      "epoch": 4.129746835443038,
      "grad_norm": 0.21492163566185005,
      "learning_rate": 8.625211401874362e-06,
      "loss": 0.5694,
      "num_tokens": 2735200836.0,
      "step": 2610
    },
    {
      "epoch": 4.137658227848101,
      "grad_norm": 0.22363359389581403,
      "learning_rate": 8.561386945304155e-06,
      "loss": 0.5727,
      "num_tokens": 2740443716.0,
      "step": 2615
    },
    {
      "epoch": 4.1455696202531644,
      "grad_norm": 0.2663268042449171,
      "learning_rate": 8.498081008788677e-06,
      "loss": 0.5689,
      "num_tokens": 2745686596.0,
      "step": 2620
    },
    {
      "epoch": 4.1534810126582276,
      "grad_norm": 0.2379773928525528,
      "learning_rate": 8.435295325580373e-06,
      "loss": 0.5693,
      "num_tokens": 2750929476.0,
      "step": 2625
    },
    {
      "epoch": 4.1613924050632916,
      "grad_norm": 0.2227866522568964,
      "learning_rate": 8.373031614687699e-06,
      "loss": 0.5698,
      "num_tokens": 2756172356.0,
      "step": 2630
    },
    {
      "epoch": 4.169303797468355,
      "grad_norm": 0.23722314775734296,
      "learning_rate": 8.311291580828034e-06,
      "loss": 0.5723,
      "num_tokens": 2761415236.0,
      "step": 2635
    },
    {
      "epoch": 4.177215189873418,
      "grad_norm": 0.2122410135791533,
      "learning_rate": 8.250076914381017e-06,
      "loss": 0.5661,
      "num_tokens": 2766658116.0,
      "step": 2640
    },
    {
      "epoch": 4.185126582278481,
      "grad_norm": 0.21872455435466664,
      "learning_rate": 8.189389291342229e-06,
      "loss": 0.5668,
      "num_tokens": 2771900996.0,
      "step": 2645
    },
    {
      "epoch": 4.193037974683544,
      "grad_norm": 0.22736670887055646,
      "learning_rate": 8.12923037327738e-06,
      "loss": 0.5723,
      "num_tokens": 2777143876.0,
      "step": 2650
    },
    {
      "epoch": 4.200949367088608,
      "grad_norm": 0.24119729706862408,
      "learning_rate": 8.06960180727674e-06,
      "loss": 0.5689,
      "num_tokens": 2782386756.0,
      "step": 2655
    },
    {
      "epoch": 4.208860759493671,
      "grad_norm": 0.22351050734609607,
      "learning_rate": 8.010505225910083e-06,
      "loss": 0.5677,
      "num_tokens": 2787629636.0,
      "step": 2660
    },
    {
      "epoch": 4.216772151898734,
      "grad_norm": 0.265494072724875,
      "learning_rate": 7.951942247181992e-06,
      "loss": 0.5718,
      "num_tokens": 2792872516.0,
      "step": 2665
    },
    {
      "epoch": 4.224683544303797,
      "grad_norm": 0.21820442695718212,
      "learning_rate": 7.89391447448755e-06,
      "loss": 0.5682,
      "num_tokens": 2798109181.0,
      "step": 2670
    },
    {
      "epoch": 4.2325949367088604,
      "grad_norm": 0.2420612432104863,
      "learning_rate": 7.836423496568418e-06,
      "loss": 0.5753,
      "num_tokens": 2803337280.0,
      "step": 2675
    },
    {
      "epoch": 4.2405063291139244,
      "grad_norm": 0.23517468870764166,
      "learning_rate": 7.779470887469387e-06,
      "loss": 0.5716,
      "num_tokens": 2808580160.0,
      "step": 2680
    },
    {
      "epoch": 4.2484177215189876,
      "grad_norm": 0.2267612123573776,
      "learning_rate": 7.723058206495242e-06,
      "loss": 0.5746,
      "num_tokens": 2813823040.0,
      "step": 2685
    },
    {
      "epoch": 4.256329113924051,
      "grad_norm": 0.22002454461146628,
      "learning_rate": 7.667186998168082e-06,
      "loss": 0.5684,
      "num_tokens": 2819065920.0,
      "step": 2690
    },
    {
      "epoch": 4.264240506329114,
      "grad_norm": 0.22912652065839756,
      "learning_rate": 7.611858792185038e-06,
      "loss": 0.5751,
      "num_tokens": 2824308800.0,
      "step": 2695
    },
    {
      "epoch": 4.272151898734177,
      "grad_norm": 0.22408651950567107,
      "learning_rate": 7.557075103376383e-06,
      "loss": 0.5724,
      "num_tokens": 2829542456.0,
      "step": 2700
    },
    {
      "epoch": 4.280063291139241,
      "grad_norm": 0.22342018834903565,
      "learning_rate": 7.502837431664059e-06,
      "loss": 0.5731,
      "num_tokens": 2834785336.0,
      "step": 2705
    },
    {
      "epoch": 4.287974683544304,
      "grad_norm": 0.22864174140723967,
      "learning_rate": 7.449147262020616e-06,
      "loss": 0.5679,
      "num_tokens": 2840015342.0,
      "step": 2710
    },
    {
      "epoch": 4.295886075949367,
      "grad_norm": 0.22865644771070281,
      "learning_rate": 7.396006064428554e-06,
      "loss": 0.5627,
      "num_tokens": 2845248811.0,
      "step": 2715
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 0.2674648148165605,
      "learning_rate": 7.34341529384006e-06,
      "loss": 0.5765,
      "num_tokens": 2850491691.0,
      "step": 2720
    },
    {
      "epoch": 4.311708860759493,
      "grad_norm": 0.22345156469242552,
      "learning_rate": 7.291376390137214e-06,
      "loss": 0.5805,
      "num_tokens": 2855734571.0,
      "step": 2725
    },
    {
      "epoch": 4.319620253164557,
      "grad_norm": 0.2122765810320669,
      "learning_rate": 7.239890778092509e-06,
      "loss": 0.5712,
      "num_tokens": 2860977451.0,
      "step": 2730
    },
    {
      "epoch": 4.3275316455696204,
      "grad_norm": 0.2477019061431497,
      "learning_rate": 7.188959867329893e-06,
      "loss": 0.5814,
      "num_tokens": 2866220331.0,
      "step": 2735
    },
    {
      "epoch": 4.3354430379746836,
      "grad_norm": 0.21235713827113287,
      "learning_rate": 7.138585052286142e-06,
      "loss": 0.5732,
      "num_tokens": 2871451563.0,
      "step": 2740
    },
    {
      "epoch": 4.343354430379747,
      "grad_norm": 0.2146641271693942,
      "learning_rate": 7.088767712172711e-06,
      "loss": 0.5717,
      "num_tokens": 2876694443.0,
      "step": 2745
    },
    {
      "epoch": 4.35126582278481,
      "grad_norm": 0.23309761275535962,
      "learning_rate": 7.039509210937932e-06,
      "loss": 0.5732,
      "num_tokens": 2881937323.0,
      "step": 2750
    },
    {
      "epoch": 4.359177215189874,
      "grad_norm": 0.2234172047939556,
      "learning_rate": 6.990810897229709e-06,
      "loss": 0.57,
      "num_tokens": 2887179519.0,
      "step": 2755
    },
    {
      "epoch": 4.367088607594937,
      "grad_norm": 0.22237797838934822,
      "learning_rate": 6.942674104358582e-06,
      "loss": 0.5738,
      "num_tokens": 2892422399.0,
      "step": 2760
    },
    {
      "epoch": 4.375,
      "grad_norm": 0.21593708745753062,
      "learning_rate": 6.8951001502612065e-06,
      "loss": 0.5658,
      "num_tokens": 2897636988.0,
      "step": 2765
    },
    {
      "epoch": 4.382911392405063,
      "grad_norm": 0.22987570961197548,
      "learning_rate": 6.848090337464294e-06,
      "loss": 0.5682,
      "num_tokens": 2902879868.0,
      "step": 2770
    },
    {
      "epoch": 4.390822784810126,
      "grad_norm": 0.22029484227034318,
      "learning_rate": 6.8016459530489335e-06,
      "loss": 0.5746,
      "num_tokens": 2908122748.0,
      "step": 2775
    },
    {
      "epoch": 4.39873417721519,
      "grad_norm": 0.2171355743200665,
      "learning_rate": 6.755768268615354e-06,
      "loss": 0.567,
      "num_tokens": 2913363036.0,
      "step": 2780
    },
    {
      "epoch": 4.406645569620253,
      "grad_norm": 0.2213627866085991,
      "learning_rate": 6.710458540248109e-06,
      "loss": 0.5725,
      "num_tokens": 2918605916.0,
      "step": 2785
    },
    {
      "epoch": 4.4145569620253164,
      "grad_norm": 0.21863302278329563,
      "learning_rate": 6.66571800848171e-06,
      "loss": 0.5733,
      "num_tokens": 2923835265.0,
      "step": 2790
    },
    {
      "epoch": 4.4224683544303796,
      "grad_norm": 0.2114144591578047,
      "learning_rate": 6.621547898266615e-06,
      "loss": 0.5731,
      "num_tokens": 2929076236.0,
      "step": 2795
    },
    {
      "epoch": 4.430379746835443,
      "grad_norm": 0.22099052084655507,
      "learning_rate": 6.577949418935732e-06,
      "loss": 0.5741,
      "num_tokens": 2934319116.0,
      "step": 2800
    },
    {
      "epoch": 4.438291139240507,
      "grad_norm": 0.23002587290313806,
      "learning_rate": 6.534923764171305e-06,
      "loss": 0.572,
      "num_tokens": 2939556532.0,
      "step": 2805
    },
    {
      "epoch": 4.44620253164557,
      "grad_norm": 0.2525898694118869,
      "learning_rate": 6.492472111972193e-06,
      "loss": 0.5758,
      "num_tokens": 2944799412.0,
      "step": 2810
    },
    {
      "epoch": 4.454113924050633,
      "grad_norm": 0.22052197863760098,
      "learning_rate": 6.450595624621672e-06,
      "loss": 0.5687,
      "num_tokens": 2950042292.0,
      "step": 2815
    },
    {
      "epoch": 4.462025316455696,
      "grad_norm": 0.22012961127429426,
      "learning_rate": 6.409295448655572e-06,
      "loss": 0.5701,
      "num_tokens": 2955275769.0,
      "step": 2820
    },
    {
      "epoch": 4.469936708860759,
      "grad_norm": 0.22397593983188469,
      "learning_rate": 6.3685727148309014e-06,
      "loss": 0.5708,
      "num_tokens": 2960518649.0,
      "step": 2825
    },
    {
      "epoch": 4.477848101265823,
      "grad_norm": 0.21586061078083138,
      "learning_rate": 6.328428538094895e-06,
      "loss": 0.5763,
      "num_tokens": 2965750136.0,
      "step": 2830
    },
    {
      "epoch": 4.485759493670886,
      "grad_norm": 0.24077949998848439,
      "learning_rate": 6.288864017554471e-06,
      "loss": 0.5733,
      "num_tokens": 2970993016.0,
      "step": 2835
    },
    {
      "epoch": 4.493670886075949,
      "grad_norm": 0.225684038542863,
      "learning_rate": 6.249880236446157e-06,
      "loss": 0.5733,
      "num_tokens": 2976232054.0,
      "step": 2840
    },
    {
      "epoch": 4.5015822784810124,
      "grad_norm": 0.212864190216861,
      "learning_rate": 6.211478262106416e-06,
      "loss": 0.5757,
      "num_tokens": 2981474934.0,
      "step": 2845
    },
    {
      "epoch": 4.509493670886076,
      "grad_norm": 0.2240303672598969,
      "learning_rate": 6.173659145942439e-06,
      "loss": 0.5715,
      "num_tokens": 2986716501.0,
      "step": 2850
    },
    {
      "epoch": 4.5174050632911396,
      "grad_norm": 0.21964401447515855,
      "learning_rate": 6.136423923403332e-06,
      "loss": 0.5729,
      "num_tokens": 2991948269.0,
      "step": 2855
    },
    {
      "epoch": 4.525316455696203,
      "grad_norm": 0.21745396683364604,
      "learning_rate": 6.099773613951805e-06,
      "loss": 0.5691,
      "num_tokens": 2997172161.0,
      "step": 2860
    },
    {
      "epoch": 4.533227848101266,
      "grad_norm": 0.2051009997691557,
      "learning_rate": 6.063709221036231e-06,
      "loss": 0.5729,
      "num_tokens": 3002415041.0,
      "step": 2865
    },
    {
      "epoch": 4.541139240506329,
      "grad_norm": 0.2082354488706409,
      "learning_rate": 6.028231732063179e-06,
      "loss": 0.5683,
      "num_tokens": 3007657921.0,
      "step": 2870
    },
    {
      "epoch": 4.549050632911392,
      "grad_norm": 0.20333733144092334,
      "learning_rate": 5.993342118370382e-06,
      "loss": 0.5728,
      "num_tokens": 3012900801.0,
      "step": 2875
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.22876886064683172,
      "learning_rate": 5.959041335200154e-06,
      "loss": 0.5713,
      "num_tokens": 3018143681.0,
      "step": 2880
    },
    {
      "epoch": 4.564873417721519,
      "grad_norm": 0.23107232316277548,
      "learning_rate": 5.925330321673209e-06,
      "loss": 0.5695,
      "num_tokens": 3023386561.0,
      "step": 2885
    },
    {
      "epoch": 4.572784810126582,
      "grad_norm": 0.2134849519354519,
      "learning_rate": 5.892210000762978e-06,
      "loss": 0.5669,
      "num_tokens": 3028629441.0,
      "step": 2890
    },
    {
      "epoch": 4.580696202531645,
      "grad_norm": 0.22796244574283725,
      "learning_rate": 5.859681279270323e-06,
      "loss": 0.5737,
      "num_tokens": 3033867096.0,
      "step": 2895
    },
    {
      "epoch": 4.588607594936709,
      "grad_norm": 0.2241995743982485,
      "learning_rate": 5.827745047798706e-06,
      "loss": 0.574,
      "num_tokens": 3039109976.0,
      "step": 2900
    },
    {
      "epoch": 4.5965189873417724,
      "grad_norm": 0.2432096355202257,
      "learning_rate": 5.796402180729816e-06,
      "loss": 0.5727,
      "num_tokens": 3044352856.0,
      "step": 2905
    },
    {
      "epoch": 4.6044303797468356,
      "grad_norm": 0.22578199424608314,
      "learning_rate": 5.7656535361996335e-06,
      "loss": 0.5732,
      "num_tokens": 3049595736.0,
      "step": 2910
    },
    {
      "epoch": 4.612341772151899,
      "grad_norm": 0.237878631640341,
      "learning_rate": 5.735499956074914e-06,
      "loss": 0.5694,
      "num_tokens": 3054838616.0,
      "step": 2915
    },
    {
      "epoch": 4.620253164556962,
      "grad_norm": 0.20655029863391322,
      "learning_rate": 5.705942265930159e-06,
      "loss": 0.568,
      "num_tokens": 3060076579.0,
      "step": 2920
    },
    {
      "epoch": 4.628164556962025,
      "grad_norm": 0.2102514850204724,
      "learning_rate": 5.676981275025011e-06,
      "loss": 0.5716,
      "num_tokens": 3065319459.0,
      "step": 2925
    },
    {
      "epoch": 4.636075949367089,
      "grad_norm": 0.21521201282981267,
      "learning_rate": 5.648617776282077e-06,
      "loss": 0.5727,
      "num_tokens": 3070562339.0,
      "step": 2930
    },
    {
      "epoch": 4.643987341772152,
      "grad_norm": 0.22220846495501412,
      "learning_rate": 5.620852546265245e-06,
      "loss": 0.5734,
      "num_tokens": 3075805219.0,
      "step": 2935
    },
    {
      "epoch": 4.651898734177215,
      "grad_norm": 0.20514030112007264,
      "learning_rate": 5.5936863451584056e-06,
      "loss": 0.5746,
      "num_tokens": 3081039464.0,
      "step": 2940
    },
    {
      "epoch": 4.659810126582278,
      "grad_norm": 0.1984930661211808,
      "learning_rate": 5.56711991674465e-06,
      "loss": 0.5775,
      "num_tokens": 3086282344.0,
      "step": 2945
    },
    {
      "epoch": 4.667721518987342,
      "grad_norm": 0.21848535912224024,
      "learning_rate": 5.54115398838589e-06,
      "loss": 0.5747,
      "num_tokens": 3091525224.0,
      "step": 2950
    },
    {
      "epoch": 4.675632911392405,
      "grad_norm": 0.21623565097615416,
      "learning_rate": 5.515789271002967e-06,
      "loss": 0.572,
      "num_tokens": 3096751361.0,
      "step": 2955
    },
    {
      "epoch": 4.6835443037974684,
      "grad_norm": 0.2067122434127961,
      "learning_rate": 5.491026459056163e-06,
      "loss": 0.5702,
      "num_tokens": 3101978695.0,
      "step": 2960
    },
    {
      "epoch": 4.6914556962025316,
      "grad_norm": 0.23059832785126186,
      "learning_rate": 5.466866230526202e-06,
      "loss": 0.5763,
      "num_tokens": 3107221575.0,
      "step": 2965
    },
    {
      "epoch": 4.699367088607595,
      "grad_norm": 0.21776141057424536,
      "learning_rate": 5.443309246895688e-06,
      "loss": 0.5724,
      "num_tokens": 3112464455.0,
      "step": 2970
    },
    {
      "epoch": 4.707278481012658,
      "grad_norm": 0.2086382988283168,
      "learning_rate": 5.420356153130986e-06,
      "loss": 0.5756,
      "num_tokens": 3117704332.0,
      "step": 2975
    },
    {
      "epoch": 4.715189873417722,
      "grad_norm": 0.22426323834508063,
      "learning_rate": 5.398007577664566e-06,
      "loss": 0.5766,
      "num_tokens": 3122940201.0,
      "step": 2980
    },
    {
      "epoch": 4.723101265822785,
      "grad_norm": 0.2134984091805231,
      "learning_rate": 5.376264132377806e-06,
      "loss": 0.5715,
      "num_tokens": 3128183081.0,
      "step": 2985
    },
    {
      "epoch": 4.731012658227848,
      "grad_norm": 0.2078865591432323,
      "learning_rate": 5.355126412584226e-06,
      "loss": 0.5796,
      "num_tokens": 3133425961.0,
      "step": 2990
    },
    {
      "epoch": 4.738924050632911,
      "grad_norm": 0.21033825281525734,
      "learning_rate": 5.334594997013194e-06,
      "loss": 0.5722,
      "num_tokens": 3138668841.0,
      "step": 2995
    },
    {
      "epoch": 4.746835443037975,
      "grad_norm": 0.20218242486649357,
      "learning_rate": 5.314670447794094e-06,
      "loss": 0.5697,
      "num_tokens": 3143911721.0,
      "step": 3000
    },
    {
      "epoch": 4.754746835443038,
      "grad_norm": 0.20528211055221243,
      "learning_rate": 5.295353310440908e-06,
      "loss": 0.5775,
      "num_tokens": 3149154601.0,
      "step": 3005
    },
    {
      "epoch": 4.762658227848101,
      "grad_norm": 0.20386433324381634,
      "learning_rate": 5.276644113837305e-06,
      "loss": 0.5703,
      "num_tokens": 3154397481.0,
      "step": 3010
    },
    {
      "epoch": 4.7705696202531644,
      "grad_norm": 0.21879527292522297,
      "learning_rate": 5.258543370222146e-06,
      "loss": 0.5739,
      "num_tokens": 3159640361.0,
      "step": 3015
    },
    {
      "epoch": 4.7784810126582276,
      "grad_norm": 0.238651659867707,
      "learning_rate": 5.241051575175476e-06,
      "loss": 0.567,
      "num_tokens": 3164869054.0,
      "step": 3020
    },
    {
      "epoch": 4.786392405063291,
      "grad_norm": 0.21930826191732458,
      "learning_rate": 5.2241692076049294e-06,
      "loss": 0.5713,
      "num_tokens": 3170097693.0,
      "step": 3025
    },
    {
      "epoch": 4.794303797468355,
      "grad_norm": 0.24549638164106027,
      "learning_rate": 5.207896729732644e-06,
      "loss": 0.5784,
      "num_tokens": 3175338802.0,
      "step": 3030
    },
    {
      "epoch": 4.802215189873418,
      "grad_norm": 0.20497121752386122,
      "learning_rate": 5.1922345870825936e-06,
      "loss": 0.5718,
      "num_tokens": 3180581682.0,
      "step": 3035
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 0.21556589673067728,
      "learning_rate": 5.1771832084683874e-06,
      "loss": 0.5755,
      "num_tokens": 3185824562.0,
      "step": 3040
    },
    {
      "epoch": 4.818037974683544,
      "grad_norm": 0.2177509045329286,
      "learning_rate": 5.162743005981538e-06,
      "loss": 0.5693,
      "num_tokens": 3191054329.0,
      "step": 3045
    },
    {
      "epoch": 4.825949367088608,
      "grad_norm": 0.22258143026660726,
      "learning_rate": 5.148914374980175e-06,
      "loss": 0.5697,
      "num_tokens": 3196297209.0,
      "step": 3050
    },
    {
      "epoch": 4.833860759493671,
      "grad_norm": 0.21877472366506615,
      "learning_rate": 5.135697694078209e-06,
      "loss": 0.5732,
      "num_tokens": 3201522963.0,
      "step": 3055
    },
    {
      "epoch": 4.841772151898734,
      "grad_norm": 0.2194924983288656,
      "learning_rate": 5.123093325134992e-06,
      "loss": 0.577,
      "num_tokens": 3206765843.0,
      "step": 3060
    },
    {
      "epoch": 4.849683544303797,
      "grad_norm": 0.24236174132097704,
      "learning_rate": 5.111101613245384e-06,
      "loss": 0.5734,
      "num_tokens": 3212008723.0,
      "step": 3065
    },
    {
      "epoch": 4.8575949367088604,
      "grad_norm": 0.22136019033385568,
      "learning_rate": 5.099722886730315e-06,
      "loss": 0.5738,
      "num_tokens": 3217244067.0,
      "step": 3070
    },
    {
      "epoch": 4.865506329113924,
      "grad_norm": 0.21908628318619908,
      "learning_rate": 5.088957457127804e-06,
      "loss": 0.5657,
      "num_tokens": 3222479254.0,
      "step": 3075
    },
    {
      "epoch": 4.8734177215189876,
      "grad_norm": 0.23149848485508415,
      "learning_rate": 5.078805619184415e-06,
      "loss": 0.5762,
      "num_tokens": 3227722134.0,
      "step": 3080
    },
    {
      "epoch": 4.881329113924051,
      "grad_norm": 0.21576735952327258,
      "learning_rate": 5.069267650847191e-06,
      "loss": 0.5687,
      "num_tokens": 3232960340.0,
      "step": 3085
    },
    {
      "epoch": 4.889240506329114,
      "grad_norm": 0.21146979612133376,
      "learning_rate": 5.060343813256054e-06,
      "loss": 0.5749,
      "num_tokens": 3238203220.0,
      "step": 3090
    },
    {
      "epoch": 4.897151898734177,
      "grad_norm": 0.22343231609158218,
      "learning_rate": 5.052034350736642e-06,
      "loss": 0.577,
      "num_tokens": 3243446100.0,
      "step": 3095
    },
    {
      "epoch": 4.905063291139241,
      "grad_norm": 0.21750437307740444,
      "learning_rate": 5.044339490793628e-06,
      "loss": 0.5746,
      "num_tokens": 3248682414.0,
      "step": 3100
    },
    {
      "epoch": 4.912974683544304,
      "grad_norm": 0.2040207080561066,
      "learning_rate": 5.037259444104488e-06,
      "loss": 0.5752,
      "num_tokens": 3253925294.0,
      "step": 3105
    },
    {
      "epoch": 4.920886075949367,
      "grad_norm": 0.21492843271025994,
      "learning_rate": 5.030794404513734e-06,
      "loss": 0.5723,
      "num_tokens": 3259168174.0,
      "step": 3110
    },
    {
      "epoch": 4.92879746835443,
      "grad_norm": 0.22472996856568292,
      "learning_rate": 5.024944549027605e-06,
      "loss": 0.5757,
      "num_tokens": 3264411054.0,
      "step": 3115
    },
    {
      "epoch": 4.936708860759493,
      "grad_norm": 0.2105235983069695,
      "learning_rate": 5.019710037809223e-06,
      "loss": 0.5764,
      "num_tokens": 3269653934.0,
      "step": 3120
    },
    {
      "epoch": 4.944620253164557,
      "grad_norm": 0.20160858018191585,
      "learning_rate": 5.015091014174209e-06,
      "loss": 0.5676,
      "num_tokens": 3274885682.0,
      "step": 3125
    },
    {
      "epoch": 4.9525316455696204,
      "grad_norm": 0.2079284452125249,
      "learning_rate": 5.0110876045867496e-06,
      "loss": 0.5662,
      "num_tokens": 3280128562.0,
      "step": 3130
    },
    {
      "epoch": 4.9604430379746836,
      "grad_norm": 0.21834231623167524,
      "learning_rate": 5.007699918656152e-06,
      "loss": 0.5719,
      "num_tokens": 3285371442.0,
      "step": 3135
    },
    {
      "epoch": 4.968354430379747,
      "grad_norm": 0.21543773946569159,
      "learning_rate": 5.004928049133823e-06,
      "loss": 0.5694,
      "num_tokens": 3290614322.0,
      "step": 3140
    },
    {
      "epoch": 4.97626582278481,
      "grad_norm": 0.2150336576277292,
      "learning_rate": 5.002772071910747e-06,
      "loss": 0.5747,
      "num_tokens": 3295843098.0,
      "step": 3145
    },
    {
      "epoch": 4.984177215189874,
      "grad_norm": 0.20873534752474804,
      "learning_rate": 5.0012320460153974e-06,
      "loss": 0.5696,
      "num_tokens": 3301070530.0,
      "step": 3150
    },
    {
      "epoch": 4.992088607594937,
      "grad_norm": 0.21825458015635657,
      "learning_rate": 5.000308013612126e-06,
      "loss": 0.5762,
      "num_tokens": 3306313410.0,
      "step": 3155
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.2206973186589871,
      "learning_rate": 5e-06,
      "loss": 0.5717,
      "num_tokens": 3311556290.0,
      "step": 3160
    },
    {
      "epoch": 5.0,
      "step": 3160,
      "total_flos": 2880677515100160.0,
      "train_loss": 0.6431828314744974,
      "train_runtime": 46582.8084,
      "train_samples_per_second": 4.34,
      "train_steps_per_second": 0.068
    }
  ],
  "logging_steps": 5,
  "max_steps": 3160,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2880677515100160.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}