{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9995855781185247,
  "eval_steps": 500,
  "global_step": 1206,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008288437629506838,
      "grad_norm": 23.821738606267694,
      "learning_rate": 8.264462809917357e-08,
      "loss": 1.4077,
      "step": 1
    },
    {
      "epoch": 0.004144218814753419,
      "grad_norm": 22.834257268242467,
      "learning_rate": 4.132231404958678e-07,
      "loss": 1.4048,
      "step": 5
    },
    {
      "epoch": 0.008288437629506838,
      "grad_norm": 8.208014337309038,
      "learning_rate": 8.264462809917356e-07,
      "loss": 1.3206,
      "step": 10
    },
    {
      "epoch": 0.012432656444260257,
      "grad_norm": 10.88562979999232,
      "learning_rate": 1.2396694214876035e-06,
      "loss": 1.1572,
      "step": 15
    },
    {
      "epoch": 0.016576875259013676,
      "grad_norm": 3.02410216068558,
      "learning_rate": 1.6528925619834712e-06,
      "loss": 1.0108,
      "step": 20
    },
    {
      "epoch": 0.020721094073767096,
      "grad_norm": 2.9630819683565868,
      "learning_rate": 2.066115702479339e-06,
      "loss": 0.9599,
      "step": 25
    },
    {
      "epoch": 0.024865312888520515,
      "grad_norm": 2.450927822041959,
      "learning_rate": 2.479338842975207e-06,
      "loss": 0.9202,
      "step": 30
    },
    {
      "epoch": 0.029009531703273934,
      "grad_norm": 2.241659288093002,
      "learning_rate": 2.8925619834710743e-06,
      "loss": 0.9075,
      "step": 35
    },
    {
      "epoch": 0.03315375051802735,
      "grad_norm": 2.2795745210790708,
      "learning_rate": 3.3057851239669424e-06,
      "loss": 0.8904,
      "step": 40
    },
    {
      "epoch": 0.03729796933278077,
      "grad_norm": 2.2916121960785274,
      "learning_rate": 3.71900826446281e-06,
      "loss": 0.8778,
      "step": 45
    },
    {
      "epoch": 0.04144218814753419,
      "grad_norm": 3.0108486369173444,
      "learning_rate": 4.132231404958678e-06,
      "loss": 0.8811,
      "step": 50
    },
    {
      "epoch": 0.04558640696228761,
      "grad_norm": 2.34463490002114,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.8644,
      "step": 55
    },
    {
      "epoch": 0.04973062577704103,
      "grad_norm": 2.2048136988502836,
      "learning_rate": 4.958677685950414e-06,
      "loss": 0.8658,
      "step": 60
    },
    {
      "epoch": 0.053874844591794445,
      "grad_norm": 2.360407044698775,
      "learning_rate": 5.371900826446281e-06,
      "loss": 0.8483,
      "step": 65
    },
    {
      "epoch": 0.05801906340654787,
      "grad_norm": 2.3750627177827437,
      "learning_rate": 5.785123966942149e-06,
      "loss": 0.829,
      "step": 70
    },
    {
      "epoch": 0.06216328222130128,
      "grad_norm": 2.5351719045640424,
      "learning_rate": 6.198347107438017e-06,
      "loss": 0.8457,
      "step": 75
    },
    {
      "epoch": 0.0663075010360547,
      "grad_norm": 2.4658487406561163,
      "learning_rate": 6.611570247933885e-06,
      "loss": 0.8295,
      "step": 80
    },
    {
      "epoch": 0.07045171985080813,
      "grad_norm": 2.552123954872006,
      "learning_rate": 7.0247933884297525e-06,
      "loss": 0.8248,
      "step": 85
    },
    {
      "epoch": 0.07459593866556154,
      "grad_norm": 2.5523240708243984,
      "learning_rate": 7.43801652892562e-06,
      "loss": 0.8073,
      "step": 90
    },
    {
      "epoch": 0.07874015748031496,
      "grad_norm": 2.6557558002422343,
      "learning_rate": 7.851239669421489e-06,
      "loss": 0.8227,
      "step": 95
    },
    {
      "epoch": 0.08288437629506838,
      "grad_norm": 2.5132033700665164,
      "learning_rate": 8.264462809917356e-06,
      "loss": 0.803,
      "step": 100
    },
    {
      "epoch": 0.0870285951098218,
      "grad_norm": 2.720513367073119,
      "learning_rate": 8.677685950413224e-06,
      "loss": 0.8023,
      "step": 105
    },
    {
      "epoch": 0.09117281392457521,
      "grad_norm": 2.9809398097562543,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.7967,
      "step": 110
    },
    {
      "epoch": 0.09531703273932864,
      "grad_norm": 2.6852876988337564,
      "learning_rate": 9.50413223140496e-06,
      "loss": 0.8083,
      "step": 115
    },
    {
      "epoch": 0.09946125155408206,
      "grad_norm": 2.4634580540100686,
      "learning_rate": 9.917355371900828e-06,
      "loss": 0.7892,
      "step": 120
    },
    {
      "epoch": 0.10360547036883548,
      "grad_norm": 2.583731945309341,
      "learning_rate": 9.999664652243188e-06,
      "loss": 0.8053,
      "step": 125
    },
    {
      "epoch": 0.10774968918358889,
      "grad_norm": 2.6768428928051873,
      "learning_rate": 9.99830237907608e-06,
      "loss": 0.7899,
      "step": 130
    },
    {
      "epoch": 0.11189390799834231,
      "grad_norm": 2.550586632842143,
      "learning_rate": 9.995892506564461e-06,
      "loss": 0.7825,
      "step": 135
    },
    {
      "epoch": 0.11603812681309573,
      "grad_norm": 2.50256712970827,
      "learning_rate": 9.992435539796e-06,
      "loss": 0.7693,
      "step": 140
    },
    {
      "epoch": 0.12018234562784914,
      "grad_norm": 2.542375468259113,
      "learning_rate": 9.987932203319917e-06,
      "loss": 0.7716,
      "step": 145
    },
    {
      "epoch": 0.12432656444260257,
      "grad_norm": 2.354650001491699,
      "learning_rate": 9.982383440995146e-06,
      "loss": 0.7789,
      "step": 150
    },
    {
      "epoch": 0.12847078325735597,
      "grad_norm": 2.680500650181213,
      "learning_rate": 9.975790415792497e-06,
      "loss": 0.772,
      "step": 155
    },
    {
      "epoch": 0.1326150020721094,
      "grad_norm": 2.8324037133090827,
      "learning_rate": 9.968154509550914e-06,
      "loss": 0.7574,
      "step": 160
    },
    {
      "epoch": 0.13675922088686282,
      "grad_norm": 3.314558520418501,
      "learning_rate": 9.959477322687852e-06,
      "loss": 0.7434,
      "step": 165
    },
    {
      "epoch": 0.14090343970161626,
      "grad_norm": 2.558536518840115,
      "learning_rate": 9.949760673863846e-06,
      "loss": 0.7454,
      "step": 170
    },
    {
      "epoch": 0.14504765851636967,
      "grad_norm": 2.6087877711239247,
      "learning_rate": 9.93900659960133e-06,
      "loss": 0.7488,
      "step": 175
    },
    {
      "epoch": 0.14919187733112307,
      "grad_norm": 2.4922551536546416,
      "learning_rate": 9.927217353857809e-06,
      "loss": 0.7453,
      "step": 180
    },
    {
      "epoch": 0.1533360961458765,
      "grad_norm": 2.4983517636117782,
      "learning_rate": 9.914395407553444e-06,
      "loss": 0.7441,
      "step": 185
    },
    {
      "epoch": 0.15748031496062992,
      "grad_norm": 2.493536518057786,
      "learning_rate": 9.900543448053164e-06,
      "loss": 0.7435,
      "step": 190
    },
    {
      "epoch": 0.16162453377538333,
      "grad_norm": 2.361368540133534,
      "learning_rate": 9.885664378603432e-06,
      "loss": 0.7216,
      "step": 195
    },
    {
      "epoch": 0.16576875259013676,
      "grad_norm": 2.2001360959731486,
      "learning_rate": 9.869761317723744e-06,
      "loss": 0.728,
      "step": 200
    },
    {
      "epoch": 0.16991297140489017,
      "grad_norm": 2.305194442427537,
      "learning_rate": 9.85283759855301e-06,
      "loss": 0.7263,
      "step": 205
    },
    {
      "epoch": 0.1740571902196436,
      "grad_norm": 2.208332719886559,
      "learning_rate": 9.834896768150963e-06,
      "loss": 0.7139,
      "step": 210
    },
    {
      "epoch": 0.17820140903439702,
      "grad_norm": 2.3264002622179887,
      "learning_rate": 9.81594258675473e-06,
      "loss": 0.7223,
      "step": 215
    },
    {
      "epoch": 0.18234562784915043,
      "grad_norm": 2.3609080234642983,
      "learning_rate": 9.795979026990717e-06,
      "loss": 0.7031,
      "step": 220
    },
    {
      "epoch": 0.18648984666390386,
      "grad_norm": 2.3331179014344055,
      "learning_rate": 9.775010273041975e-06,
      "loss": 0.715,
      "step": 225
    },
    {
      "epoch": 0.19063406547865727,
      "grad_norm": 2.4595571627938724,
      "learning_rate": 9.753040719771249e-06,
      "loss": 0.7084,
      "step": 230
    },
    {
      "epoch": 0.19477828429341068,
      "grad_norm": 2.5273928971647823,
      "learning_rate": 9.730074971799837e-06,
      "loss": 0.6928,
      "step": 235
    },
    {
      "epoch": 0.19892250310816412,
      "grad_norm": 2.4481147916191324,
      "learning_rate": 9.706117842542517e-06,
      "loss": 0.684,
      "step": 240
    },
    {
      "epoch": 0.20306672192291753,
      "grad_norm": 2.396714983793178,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.6942,
      "step": 245
    },
    {
      "epoch": 0.20721094073767096,
      "grad_norm": 2.23212792235539,
      "learning_rate": 9.655249731699973e-06,
      "loss": 0.6846,
      "step": 250
    },
    {
      "epoch": 0.21135515955242437,
      "grad_norm": 2.33058592821463,
      "learning_rate": 9.628349411614503e-06,
      "loss": 0.6819,
      "step": 255
    },
    {
      "epoch": 0.21549937836717778,
      "grad_norm": 2.3204394270788966,
      "learning_rate": 9.600479031008072e-06,
      "loss": 0.6773,
      "step": 260
    },
    {
      "epoch": 0.21964359718193122,
      "grad_norm": 2.2382725360662308,
      "learning_rate": 9.571644431262463e-06,
      "loss": 0.6649,
      "step": 265
    },
    {
      "epoch": 0.22378781599668462,
      "grad_norm": 2.2039133108908184,
      "learning_rate": 9.54185165585114e-06,
      "loss": 0.6537,
      "step": 270
    },
    {
      "epoch": 0.22793203481143803,
      "grad_norm": 2.4253359596794337,
      "learning_rate": 9.511106949072588e-06,
      "loss": 0.6621,
      "step": 275
    },
    {
      "epoch": 0.23207625362619147,
      "grad_norm": 2.4527370719290915,
      "learning_rate": 9.479416754741577e-06,
      "loss": 0.6694,
      "step": 280
    },
    {
      "epoch": 0.23622047244094488,
      "grad_norm": 2.3817761190945514,
      "learning_rate": 9.446787714838579e-06,
      "loss": 0.6594,
      "step": 285
    },
    {
      "epoch": 0.2403646912556983,
      "grad_norm": 2.2516467162015594,
      "learning_rate": 9.413226668117679e-06,
      "loss": 0.6535,
      "step": 290
    },
    {
      "epoch": 0.24450891007045172,
      "grad_norm": 2.516472366579631,
      "learning_rate": 9.37874064867323e-06,
      "loss": 0.6558,
      "step": 295
    },
    {
      "epoch": 0.24865312888520513,
      "grad_norm": 2.4095497698840336,
      "learning_rate": 9.343336884465577e-06,
      "loss": 0.6456,
      "step": 300
    },
    {
      "epoch": 0.25279734769995854,
      "grad_norm": 2.4199182947381024,
      "learning_rate": 9.307022795806125e-06,
      "loss": 0.6406,
      "step": 305
    },
    {
      "epoch": 0.25694156651471195,
      "grad_norm": 2.282073481167692,
      "learning_rate": 9.26980599380213e-06,
      "loss": 0.6339,
      "step": 310
    },
    {
      "epoch": 0.2610857853294654,
      "grad_norm": 2.441801633470137,
      "learning_rate": 9.231694278761455e-06,
      "loss": 0.6326,
      "step": 315
    },
    {
      "epoch": 0.2652300041442188,
      "grad_norm": 12.39175458554138,
      "learning_rate": 9.192695638557723e-06,
      "loss": 0.6356,
      "step": 320
    },
    {
      "epoch": 0.26937422295897223,
      "grad_norm": 2.5618372097096525,
      "learning_rate": 9.1528182469561e-06,
      "loss": 0.6252,
      "step": 325
    },
    {
      "epoch": 0.27351844177372564,
      "grad_norm": 2.4608597563280967,
      "learning_rate": 9.112070461900178e-06,
      "loss": 0.6119,
      "step": 330
    },
    {
      "epoch": 0.27766266058847905,
      "grad_norm": 2.312033920022348,
      "learning_rate": 9.070460823760197e-06,
      "loss": 0.6216,
      "step": 335
    },
    {
      "epoch": 0.2818068794032325,
      "grad_norm": 2.2955010300704344,
      "learning_rate": 9.027998053543079e-06,
      "loss": 0.6216,
      "step": 340
    },
    {
      "epoch": 0.2859510982179859,
      "grad_norm": 2.2889198622068783,
      "learning_rate": 8.984691051064576e-06,
      "loss": 0.6104,
      "step": 345
    },
    {
      "epoch": 0.29009531703273933,
      "grad_norm": 2.2147528522580844,
      "learning_rate": 8.94054889308395e-06,
      "loss": 0.6096,
      "step": 350
    },
    {
      "epoch": 0.29423953584749274,
      "grad_norm": 2.3586059493212814,
      "learning_rate": 8.895580831401563e-06,
      "loss": 0.6126,
      "step": 355
    },
    {
      "epoch": 0.29838375466224615,
      "grad_norm": 2.7570148244223924,
      "learning_rate": 8.849796290919787e-06,
      "loss": 0.6102,
      "step": 360
    },
    {
      "epoch": 0.3025279734769996,
      "grad_norm": 2.2730196820740667,
      "learning_rate": 8.803204867667624e-06,
      "loss": 0.6037,
      "step": 365
    },
    {
      "epoch": 0.306672192291753,
      "grad_norm": 2.27164079566725,
      "learning_rate": 8.755816326789469e-06,
      "loss": 0.601,
      "step": 370
    },
    {
      "epoch": 0.31081641110650643,
      "grad_norm": 2.2763704686366557,
      "learning_rate": 8.70764060049842e-06,
      "loss": 0.5829,
      "step": 375
    },
    {
      "epoch": 0.31496062992125984,
      "grad_norm": 2.265603504905875,
      "learning_rate": 8.658687785994579e-06,
      "loss": 0.5957,
      "step": 380
    },
    {
      "epoch": 0.31910484873601325,
      "grad_norm": 2.4701408773246203,
      "learning_rate": 8.608968143348765e-06,
      "loss": 0.5846,
      "step": 385
    },
    {
      "epoch": 0.32324906755076666,
      "grad_norm": 2.2590604345543897,
      "learning_rate": 8.558492093352098e-06,
      "loss": 0.5948,
      "step": 390
    },
    {
      "epoch": 0.3273932863655201,
      "grad_norm": 2.112110217235179,
      "learning_rate": 8.50727021533189e-06,
      "loss": 0.5897,
      "step": 395
    },
    {
      "epoch": 0.33153750518027353,
      "grad_norm": 2.2220762245644297,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.5686,
      "step": 400
    },
    {
      "epoch": 0.33568172399502694,
      "grad_norm": 2.4588674796101087,
      "learning_rate": 8.402632071874348e-06,
      "loss": 0.5791,
      "step": 405
    },
    {
      "epoch": 0.33982594280978035,
      "grad_norm": 2.120301054281535,
      "learning_rate": 8.349237737653288e-06,
      "loss": 0.5761,
      "step": 410
    },
    {
      "epoch": 0.34397016162453375,
      "grad_norm": 2.27000349226283,
      "learning_rate": 8.29514143324466e-06,
      "loss": 0.5664,
      "step": 415
    },
    {
      "epoch": 0.3481143804392872,
      "grad_norm": 2.209196060426881,
      "learning_rate": 8.24035449674863e-06,
      "loss": 0.5598,
      "step": 420
    },
    {
      "epoch": 0.3522585992540406,
      "grad_norm": 2.765263815929266,
      "learning_rate": 8.184888411015655e-06,
      "loss": 0.5571,
      "step": 425
    },
    {
      "epoch": 0.35640281806879404,
      "grad_norm": 2.3926237887826405,
      "learning_rate": 8.128754801239781e-06,
      "loss": 0.556,
      "step": 430
    },
    {
      "epoch": 0.36054703688354744,
      "grad_norm": 2.4696018462025506,
      "learning_rate": 8.071965432522107e-06,
      "loss": 0.5614,
      "step": 435
    },
    {
      "epoch": 0.36469125569830085,
      "grad_norm": 2.2772583844160614,
      "learning_rate": 8.01453220740492e-06,
      "loss": 0.5521,
      "step": 440
    },
    {
      "epoch": 0.36883547451305426,
      "grad_norm": 2.374847612226747,
      "learning_rate": 7.956467163377037e-06,
      "loss": 0.5495,
      "step": 445
    },
    {
      "epoch": 0.3729796933278077,
      "grad_norm": 2.2043358974290386,
      "learning_rate": 7.89778247035085e-06,
      "loss": 0.5427,
      "step": 450
    },
    {
      "epoch": 0.37712391214256114,
      "grad_norm": 2.331068862386996,
      "learning_rate": 7.838490428111625e-06,
      "loss": 0.5449,
      "step": 455
    },
    {
      "epoch": 0.38126813095731454,
      "grad_norm": 2.2139576546877713,
      "learning_rate": 7.77860346373957e-06,
      "loss": 0.5561,
      "step": 460
    },
    {
      "epoch": 0.38541234977206795,
      "grad_norm": 2.1991958487671655,
      "learning_rate": 7.718134129005238e-06,
      "loss": 0.5419,
      "step": 465
    },
    {
      "epoch": 0.38955656858682136,
      "grad_norm": 2.1821797937738396,
      "learning_rate": 7.657095097738793e-06,
      "loss": 0.5451,
      "step": 470
    },
    {
      "epoch": 0.3937007874015748,
      "grad_norm": 2.2186951662903063,
      "learning_rate": 7.59549916317367e-06,
      "loss": 0.5347,
      "step": 475
    },
    {
      "epoch": 0.39784500621632823,
      "grad_norm": 2.3212871580558287,
      "learning_rate": 7.533359235265248e-06,
      "loss": 0.5382,
      "step": 480
    },
    {
      "epoch": 0.40198922503108164,
      "grad_norm": 2.1540998726161753,
      "learning_rate": 7.470688337985029e-06,
      "loss": 0.5212,
      "step": 485
    },
    {
      "epoch": 0.40613344384583505,
      "grad_norm": 2.1735121794127132,
      "learning_rate": 7.407499606590934e-06,
      "loss": 0.5312,
      "step": 490
    },
    {
      "epoch": 0.41027766266058846,
      "grad_norm": 2.173386732621735,
      "learning_rate": 7.343806284874268e-06,
      "loss": 0.5273,
      "step": 495
    },
    {
      "epoch": 0.4144218814753419,
      "grad_norm": 2.127262862741766,
      "learning_rate": 7.279621722383939e-06,
      "loss": 0.5145,
      "step": 500
    },
    {
      "epoch": 0.41856610029009533,
      "grad_norm": 2.146219494441019,
      "learning_rate": 7.214959371628522e-06,
      "loss": 0.5225,
      "step": 505
    },
    {
      "epoch": 0.42271031910484874,
      "grad_norm": 2.046689251344904,
      "learning_rate": 7.149832785256718e-06,
      "loss": 0.511,
      "step": 510
    },
    {
      "epoch": 0.42685453791960215,
      "grad_norm": 2.1558137201613423,
      "learning_rate": 7.084255613216855e-06,
      "loss": 0.5103,
      "step": 515
    },
    {
      "epoch": 0.43099875673435556,
      "grad_norm": 2.240438771177761,
      "learning_rate": 7.018241599895974e-06,
      "loss": 0.5187,
      "step": 520
    },
    {
      "epoch": 0.43514297554910897,
      "grad_norm": 2.1251451886790775,
      "learning_rate": 6.95180458123913e-06,
      "loss": 0.509,
      "step": 525
    },
    {
      "epoch": 0.43928719436386243,
      "grad_norm": 2.1265555528429743,
      "learning_rate": 6.8849584818494984e-06,
      "loss": 0.5133,
      "step": 530
    },
    {
      "epoch": 0.44343141317861584,
      "grad_norm": 2.1752821858955143,
      "learning_rate": 6.817717312069913e-06,
      "loss": 0.5102,
      "step": 535
    },
    {
      "epoch": 0.44757563199336925,
      "grad_norm": 1.9742904998884219,
      "learning_rate": 6.750095165046415e-06,
      "loss": 0.4997,
      "step": 540
    },
    {
      "epoch": 0.45171985080812266,
      "grad_norm": 2.131305839198853,
      "learning_rate": 6.682106213774459e-06,
      "loss": 0.5023,
      "step": 545
    },
    {
      "epoch": 0.45586406962287607,
      "grad_norm": 2.31125121725231,
      "learning_rate": 6.6137647081283776e-06,
      "loss": 0.4836,
      "step": 550
    },
    {
      "epoch": 0.46000828843762953,
      "grad_norm": 2.2445463529407705,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.5052,
      "step": 555
    },
    {
      "epoch": 0.46415250725238294,
      "grad_norm": 2.0590014501028078,
      "learning_rate": 6.476081399670212e-06,
      "loss": 0.4769,
      "step": 560
    },
    {
      "epoch": 0.46829672606713635,
      "grad_norm": 2.1235457676935554,
      "learning_rate": 6.406768454044581e-06,
      "loss": 0.482,
      "step": 565
    },
    {
      "epoch": 0.47244094488188976,
      "grad_norm": 2.245932227710593,
      "learning_rate": 6.337160662369519e-06,
      "loss": 0.4797,
      "step": 570
    },
    {
      "epoch": 0.47658516369664317,
      "grad_norm": 2.4691296526493414,
      "learning_rate": 6.267272613813789e-06,
      "loss": 0.48,
      "step": 575
    },
    {
      "epoch": 0.4807293825113966,
      "grad_norm": 2.0821407072798377,
      "learning_rate": 6.19711895628548e-06,
      "loss": 0.4775,
      "step": 580
    },
    {
      "epoch": 0.48487360132615004,
      "grad_norm": 2.2231462845429553,
      "learning_rate": 6.126714393361939e-06,
      "loss": 0.4791,
      "step": 585
    },
    {
      "epoch": 0.48901782014090345,
      "grad_norm": 2.2327431903192756,
      "learning_rate": 6.056073681208038e-06,
      "loss": 0.4687,
      "step": 590
    },
    {
      "epoch": 0.49316203895565686,
      "grad_norm": 2.171485152388387,
      "learning_rate": 5.985211625483403e-06,
      "loss": 0.4773,
      "step": 595
    },
    {
      "epoch": 0.49730625777041026,
      "grad_norm": 2.383751329793211,
      "learning_rate": 5.914143078239293e-06,
      "loss": 0.4841,
      "step": 600
    },
    {
      "epoch": 0.5014504765851637,
      "grad_norm": 2.1730844882589158,
      "learning_rate": 5.842882934805731e-06,
      "loss": 0.479,
      "step": 605
    },
    {
      "epoch": 0.5055946953999171,
      "grad_norm": 2.072456765223728,
      "learning_rate": 5.771446130669589e-06,
      "loss": 0.4683,
      "step": 610
    },
    {
      "epoch": 0.5097389142146705,
      "grad_norm": 2.048482336533146,
      "learning_rate": 5.6998476383442345e-06,
      "loss": 0.4698,
      "step": 615
    },
    {
      "epoch": 0.5138831330294239,
      "grad_norm": 2.3268647107802702,
      "learning_rate": 5.628102464231429e-06,
      "loss": 0.4667,
      "step": 620
    },
    {
      "epoch": 0.5180273518441774,
      "grad_norm": 2.15935166681747,
      "learning_rate": 5.556225645476119e-06,
      "loss": 0.4719,
      "step": 625
    },
    {
      "epoch": 0.5221715706589308,
      "grad_norm": 2.2554049828411835,
      "learning_rate": 5.4842322468147926e-06,
      "loss": 0.4627,
      "step": 630
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 2.16547540035252,
      "learning_rate": 5.412137357418037e-06,
      "loss": 0.4542,
      "step": 635
    },
    {
      "epoch": 0.5304600082884376,
      "grad_norm": 2.205449973964332,
      "learning_rate": 5.339956087727985e-06,
      "loss": 0.4638,
      "step": 640
    },
    {
      "epoch": 0.534604227103191,
      "grad_norm": 2.046592350153715,
      "learning_rate": 5.2677035662913116e-06,
      "loss": 0.4504,
      "step": 645
    },
    {
      "epoch": 0.5387484459179445,
      "grad_norm": 2.1937309067407114,
      "learning_rate": 5.195394936588409e-06,
      "loss": 0.4401,
      "step": 650
    },
    {
      "epoch": 0.5428926647326979,
      "grad_norm": 2.2350791342967797,
      "learning_rate": 5.123045353859465e-06,
      "loss": 0.4508,
      "step": 655
    },
    {
      "epoch": 0.5470368835474513,
      "grad_norm": 2.1319372468329645,
      "learning_rate": 5.050669981928056e-06,
      "loss": 0.4521,
      "step": 660
    },
    {
      "epoch": 0.5511811023622047,
      "grad_norm": 2.130774233274741,
      "learning_rate": 4.978283990022936e-06,
      "loss": 0.4493,
      "step": 665
    },
    {
      "epoch": 0.5553253211769581,
      "grad_norm": 2.008484642041683,
      "learning_rate": 4.905902549598719e-06,
      "loss": 0.4325,
      "step": 670
    },
    {
      "epoch": 0.5594695399917116,
      "grad_norm": 2.0574926912311344,
      "learning_rate": 4.833540831156062e-06,
      "loss": 0.4453,
      "step": 675
    },
    {
      "epoch": 0.563613758806465,
      "grad_norm": 2.2090128918878174,
      "learning_rate": 4.761214001062079e-06,
      "loss": 0.4353,
      "step": 680
    },
    {
      "epoch": 0.5677579776212184,
      "grad_norm": 2.0771401122898197,
      "learning_rate": 4.688937218371592e-06,
      "loss": 0.4322,
      "step": 685
    },
    {
      "epoch": 0.5719021964359718,
      "grad_norm": 2.091649616042253,
      "learning_rate": 4.616725631649938e-06,
      "loss": 0.4353,
      "step": 690
    },
    {
      "epoch": 0.5760464152507252,
      "grad_norm": 2.140897800881621,
      "learning_rate": 4.544594375797969e-06,
      "loss": 0.4286,
      "step": 695
    },
    {
      "epoch": 0.5801906340654787,
      "grad_norm": 2.3556886559066887,
      "learning_rate": 4.472558568879901e-06,
      "loss": 0.4291,
      "step": 700
    },
    {
      "epoch": 0.5843348528802321,
      "grad_norm": 2.061915937329502,
      "learning_rate": 4.400633308954713e-06,
      "loss": 0.4261,
      "step": 705
    },
    {
      "epoch": 0.5884790716949855,
      "grad_norm": 2.1754455938746218,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.4263,
      "step": 710
    },
    {
      "epoch": 0.5926232905097389,
      "grad_norm": 2.047663516312869,
      "learning_rate": 4.257174703311032e-06,
      "loss": 0.4326,
      "step": 715
    },
    {
      "epoch": 0.5967675093244923,
      "grad_norm": 2.002919898924837,
      "learning_rate": 4.185671425229477e-06,
      "loss": 0.429,
      "step": 720
    },
    {
      "epoch": 0.6009117281392458,
      "grad_norm": 2.055858095540348,
      "learning_rate": 4.11433882311277e-06,
      "loss": 0.4227,
      "step": 725
    },
    {
      "epoch": 0.6050559469539992,
      "grad_norm": 2.098430357886563,
      "learning_rate": 4.043191847634469e-06,
      "loss": 0.4145,
      "step": 730
    },
    {
      "epoch": 0.6092001657687526,
      "grad_norm": 2.113187936891634,
      "learning_rate": 3.9722454105624545e-06,
      "loss": 0.421,
      "step": 735
    },
    {
      "epoch": 0.613344384583506,
      "grad_norm": 2.1821256779062397,
      "learning_rate": 3.901514381633555e-06,
      "loss": 0.4095,
      "step": 740
    },
    {
      "epoch": 0.6174886033982594,
      "grad_norm": 2.087497370021808,
      "learning_rate": 3.831013585436985e-06,
      "loss": 0.4209,
      "step": 745
    },
    {
      "epoch": 0.6216328222130129,
      "grad_norm": 2.1375456510365654,
      "learning_rate": 3.7607577983072486e-06,
      "loss": 0.4123,
      "step": 750
    },
    {
      "epoch": 0.6257770410277662,
      "grad_norm": 2.1081763007416034,
      "learning_rate": 3.6907617452271394e-06,
      "loss": 0.4049,
      "step": 755
    },
    {
      "epoch": 0.6299212598425197,
      "grad_norm": 2.1448026787250636,
      "learning_rate": 3.621040096741526e-06,
      "loss": 0.4159,
      "step": 760
    },
    {
      "epoch": 0.6340654786572731,
      "grad_norm": 2.1199166483261824,
      "learning_rate": 3.55160746588254e-06,
      "loss": 0.4071,
      "step": 765
    },
    {
      "epoch": 0.6382096974720265,
      "grad_norm": 2.023385824501021,
      "learning_rate": 3.482478405106803e-06,
      "loss": 0.409,
      "step": 770
    },
    {
      "epoch": 0.64235391628678,
      "grad_norm": 2.049301320892124,
      "learning_rate": 3.4136674032453787e-06,
      "loss": 0.4114,
      "step": 775
    },
    {
      "epoch": 0.6464981351015333,
      "grad_norm": 2.049238108716136,
      "learning_rate": 3.34518888246703e-06,
      "loss": 0.4051,
      "step": 780
    },
    {
      "epoch": 0.6506423539162868,
      "grad_norm": 2.1220065162587036,
      "learning_rate": 3.2770571952554674e-06,
      "loss": 0.4012,
      "step": 785
    },
    {
      "epoch": 0.6547865727310402,
      "grad_norm": 2.036047531360177,
      "learning_rate": 3.2092866214011984e-06,
      "loss": 0.4126,
      "step": 790
    },
    {
      "epoch": 0.6589307915457936,
      "grad_norm": 2.1291206967434593,
      "learning_rate": 3.141891365008609e-06,
      "loss": 0.4119,
      "step": 795
    },
    {
      "epoch": 0.6630750103605471,
      "grad_norm": 2.07461368161324,
      "learning_rate": 3.0748855515189104e-06,
      "loss": 0.3945,
      "step": 800
    },
    {
      "epoch": 0.6672192291753004,
      "grad_norm": 1.9714918496227842,
      "learning_rate": 3.00828322474958e-06,
      "loss": 0.3954,
      "step": 805
    },
    {
      "epoch": 0.6713634479900539,
      "grad_norm": 3.9519332717986444,
      "learning_rate": 2.942098343950891e-06,
      "loss": 0.4031,
      "step": 810
    },
    {
      "epoch": 0.6755076668048073,
      "grad_norm": 2.1077325411543293,
      "learning_rate": 2.8763447808801914e-06,
      "loss": 0.3884,
      "step": 815
    },
    {
      "epoch": 0.6796518856195607,
      "grad_norm": 2.025521267691112,
      "learning_rate": 2.8110363168944976e-06,
      "loss": 0.394,
      "step": 820
    },
    {
      "epoch": 0.6837961044343142,
      "grad_norm": 2.130847414216846,
      "learning_rate": 2.7461866400620506e-06,
      "loss": 0.3857,
      "step": 825
    },
    {
      "epoch": 0.6879403232490675,
      "grad_norm": 2.0051322597023566,
      "learning_rate": 2.6818093422934254e-06,
      "loss": 0.3906,
      "step": 830
    },
    {
      "epoch": 0.692084542063821,
      "grad_norm": 2.001609177757284,
      "learning_rate": 2.617917916492776e-06,
      "loss": 0.3921,
      "step": 835
    },
    {
      "epoch": 0.6962287608785744,
      "grad_norm": 2.072211567891342,
      "learning_rate": 2.5545257537298497e-06,
      "loss": 0.3982,
      "step": 840
    },
    {
      "epoch": 0.7003729796933278,
      "grad_norm": 1.9935034929673958,
      "learning_rate": 2.491646140433346e-06,
      "loss": 0.3798,
      "step": 845
    },
    {
      "epoch": 0.7045171985080813,
      "grad_norm": 1.9339328870220762,
      "learning_rate": 2.4292922556061877e-06,
      "loss": 0.3853,
      "step": 850
    },
    {
      "epoch": 0.7086614173228346,
      "grad_norm": 2.1038621746542767,
      "learning_rate": 2.367477168063326e-06,
      "loss": 0.3817,
      "step": 855
    },
    {
      "epoch": 0.7128056361375881,
      "grad_norm": 2.0504707093576777,
      "learning_rate": 2.3062138336926406e-06,
      "loss": 0.3828,
      "step": 860
    },
    {
      "epoch": 0.7169498549523415,
      "grad_norm": 2.128613755893242,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.3854,
      "step": 865
    },
    {
      "epoch": 0.7210940737670949,
      "grad_norm": 2.021286433659744,
      "learning_rate": 2.185393667115513e-06,
      "loss": 0.3767,
      "step": 870
    },
    {
      "epoch": 0.7252382925818484,
      "grad_norm": 2.175790029610662,
      "learning_rate": 2.125862157732245e-06,
      "loss": 0.3752,
      "step": 875
    },
    {
      "epoch": 0.7293825113966017,
      "grad_norm": 2.055231327953937,
      "learning_rate": 2.066933041860059e-06,
      "loss": 0.3797,
      "step": 880
    },
    {
      "epoch": 0.7335267302113552,
      "grad_norm": 2.0609885500688745,
      "learning_rate": 2.0086186705130545e-06,
      "loss": 0.3798,
      "step": 885
    },
    {
      "epoch": 0.7376709490261085,
      "grad_norm": 1.9460493213850336,
      "learning_rate": 1.9509312658603954e-06,
      "loss": 0.3716,
      "step": 890
    },
    {
      "epoch": 0.741815167840862,
      "grad_norm": 2.107550534612032,
      "learning_rate": 1.8938829186646484e-06,
      "loss": 0.3812,
      "step": 895
    },
    {
      "epoch": 0.7459593866556155,
      "grad_norm": 1.9693264223542786,
      "learning_rate": 1.8374855857476687e-06,
      "loss": 0.371,
      "step": 900
    },
    {
      "epoch": 0.7501036054703688,
      "grad_norm": 2.0116271854033188,
      "learning_rate": 1.7817510874845585e-06,
      "loss": 0.3705,
      "step": 905
    },
    {
      "epoch": 0.7542478242851223,
      "grad_norm": 1.9693053868651615,
      "learning_rate": 1.7266911053262196e-06,
      "loss": 0.3676,
      "step": 910
    },
    {
      "epoch": 0.7583920430998756,
      "grad_norm": 2.1020529435678186,
      "learning_rate": 1.6723171793510363e-06,
      "loss": 0.3737,
      "step": 915
    },
    {
      "epoch": 0.7625362619146291,
      "grad_norm": 2.093530940289316,
      "learning_rate": 1.6186407058461622e-06,
      "loss": 0.3797,
      "step": 920
    },
    {
      "epoch": 0.7666804807293826,
      "grad_norm": 2.0472605324094326,
      "learning_rate": 1.5656729349189742e-06,
      "loss": 0.3716,
      "step": 925
    },
    {
      "epoch": 0.7708246995441359,
      "grad_norm": 2.0335898500157925,
      "learning_rate": 1.5134249681391416e-06,
      "loss": 0.3672,
      "step": 930
    },
    {
      "epoch": 0.7749689183588894,
      "grad_norm": 1.9454457763257478,
      "learning_rate": 1.4619077562118477e-06,
      "loss": 0.368,
      "step": 935
    },
    {
      "epoch": 0.7791131371736427,
      "grad_norm": 2.057587525250662,
      "learning_rate": 1.411132096682606e-06,
      "loss": 0.3694,
      "step": 940
    },
    {
      "epoch": 0.7832573559883962,
      "grad_norm": 2.1292730946235934,
      "learning_rate": 1.3611086316742057e-06,
      "loss": 0.3595,
      "step": 945
    },
    {
      "epoch": 0.7874015748031497,
      "grad_norm": 2.022478659877928,
      "learning_rate": 1.3118478456562073e-06,
      "loss": 0.3635,
      "step": 950
    },
    {
      "epoch": 0.791545793617903,
      "grad_norm": 2.0367686920587857,
      "learning_rate": 1.2633600632474962e-06,
      "loss": 0.3593,
      "step": 955
    },
    {
      "epoch": 0.7956900124326565,
      "grad_norm": 2.1462356201632065,
      "learning_rate": 1.2156554470523364e-06,
      "loss": 0.3589,
      "step": 960
    },
    {
      "epoch": 0.7998342312474098,
      "grad_norm": 2.068674533072379,
      "learning_rate": 1.1687439955303764e-06,
      "loss": 0.3645,
      "step": 965
    },
    {
      "epoch": 0.8039784500621633,
      "grad_norm": 2.107218769695033,
      "learning_rate": 1.1226355409010686e-06,
      "loss": 0.3629,
      "step": 970
    },
    {
      "epoch": 0.8081226688769167,
      "grad_norm": 2.1600739342990427,
      "learning_rate": 1.0773397470829145e-06,
      "loss": 0.3509,
      "step": 975
    },
    {
      "epoch": 0.8122668876916701,
      "grad_norm": 2.056346466443948,
      "learning_rate": 1.032866107667999e-06,
      "loss": 0.3642,
      "step": 980
    },
    {
      "epoch": 0.8164111065064236,
      "grad_norm": 2.055999415717534,
      "learning_rate": 9.892239439322243e-07,
      "loss": 0.3608,
      "step": 985
    },
    {
      "epoch": 0.8205553253211769,
      "grad_norm": 1.9865625625009857,
      "learning_rate": 9.464224028816427e-07,
      "loss": 0.352,
      "step": 990
    },
    {
      "epoch": 0.8246995441359304,
      "grad_norm": 2.090284485836539,
      "learning_rate": 9.044704553353323e-07,
      "loss": 0.3552,
      "step": 995
    },
    {
      "epoch": 0.8288437629506838,
      "grad_norm": 2.0868083838688585,
      "learning_rate": 8.633768940451981e-07,
      "loss": 0.3494,
      "step": 1000
    },
    {
      "epoch": 0.8329879817654372,
      "grad_norm": 2.0849345095616507,
      "learning_rate": 8.231503318530814e-07,
      "loss": 0.3503,
      "step": 1005
    },
    {
      "epoch": 0.8371322005801907,
      "grad_norm": 2.036555088205375,
      "learning_rate": 7.837991998855899e-07,
      "loss": 0.3532,
      "step": 1010
    },
    {
      "epoch": 0.841276419394944,
      "grad_norm": 1.9647600324287684,
      "learning_rate": 7.453317457870096e-07,
      "loss": 0.3526,
      "step": 1015
    },
    {
      "epoch": 0.8454206382096975,
      "grad_norm": 1.9946404926738106,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.3507,
      "step": 1020
    },
    {
      "epoch": 0.8495648570244508,
      "grad_norm": 2.0929585255558805,
      "learning_rate": 6.710799340291341e-07,
      "loss": 0.3565,
      "step": 1025
    },
    {
      "epoch": 0.8537090758392043,
      "grad_norm": 1.9862050269807172,
      "learning_rate": 6.353111388835564e-07,
      "loss": 0.3504,
      "step": 1030
    },
    {
      "epoch": 0.8578532946539578,
      "grad_norm": 2.027382864636772,
      "learning_rate": 6.00457143372557e-07,
      "loss": 0.352,
      "step": 1035
    },
    {
      "epoch": 0.8619975134687111,
      "grad_norm": 2.090716583561947,
      "learning_rate": 5.665252525809583e-07,
      "loss": 0.3425,
      "step": 1040
    },
    {
      "epoch": 0.8661417322834646,
      "grad_norm": 2.06905352203202,
      "learning_rate": 5.335225783287051e-07,
      "loss": 0.3407,
      "step": 1045
    },
    {
      "epoch": 0.8702859510982179,
      "grad_norm": 2.125696851804249,
      "learning_rate": 5.014560376802913e-07,
      "loss": 0.3487,
      "step": 1050
    },
    {
      "epoch": 0.8744301699129714,
      "grad_norm": 1.9923666150875952,
      "learning_rate": 4.703323514950042e-07,
      "loss": 0.3409,
      "step": 1055
    },
    {
      "epoch": 0.8785743887277249,
      "grad_norm": 2.0327181763907736,
      "learning_rate": 4.401580430182928e-07,
      "loss": 0.3506,
      "step": 1060
    },
    {
      "epoch": 0.8827186075424782,
      "grad_norm": 2.255853441666556,
      "learning_rate": 4.1093943651455305e-07,
      "loss": 0.3459,
      "step": 1065
    },
    {
      "epoch": 0.8868628263572317,
      "grad_norm": 2.0691292772221663,
      "learning_rate": 3.826826559416219e-07,
      "loss": 0.3483,
      "step": 1070
    },
    {
      "epoch": 0.891007045171985,
      "grad_norm": 2.0290765748725605,
      "learning_rate": 3.5539362366724784e-07,
      "loss": 0.3501,
      "step": 1075
    },
    {
      "epoch": 0.8951512639867385,
      "grad_norm": 2.1150371568126975,
      "learning_rate": 3.290780592278148e-07,
      "loss": 0.3442,
      "step": 1080
    },
    {
      "epoch": 0.899295482801492,
      "grad_norm": 1.9671376281528878,
      "learning_rate": 3.0374147812958387e-07,
      "loss": 0.3431,
      "step": 1085
    },
    {
      "epoch": 0.9034397016162453,
      "grad_norm": 2.046538088407744,
      "learning_rate": 2.7938919069268654e-07,
      "loss": 0.337,
      "step": 1090
    },
    {
      "epoch": 0.9075839204309988,
      "grad_norm": 2.0585547681661756,
      "learning_rate": 2.5602630093813253e-07,
      "loss": 0.3452,
      "step": 1095
    },
    {
      "epoch": 0.9117281392457521,
      "grad_norm": 2.080490788044654,
      "learning_rate": 2.3365770551805223e-07,
      "loss": 0.3399,
      "step": 1100
    },
    {
      "epoch": 0.9158723580605056,
      "grad_norm": 2.040411049775766,
      "learning_rate": 2.1228809268940164e-07,
      "loss": 0.3393,
      "step": 1105
    },
    {
      "epoch": 0.9200165768752591,
      "grad_norm": 2.0655064123833498,
      "learning_rate": 1.919219413313478e-07,
      "loss": 0.3395,
      "step": 1110
    },
    {
      "epoch": 0.9241607956900124,
      "grad_norm": 2.140596373973613,
      "learning_rate": 1.725635200065323e-07,
      "loss": 0.3405,
      "step": 1115
    },
    {
      "epoch": 0.9283050145047659,
      "grad_norm": 1.986374862328418,
      "learning_rate": 1.5421688606642392e-07,
      "loss": 0.3428,
      "step": 1120
    },
    {
      "epoch": 0.9324492333195192,
      "grad_norm": 2.040129619292376,
      "learning_rate": 1.3688588480092913e-07,
      "loss": 0.3446,
      "step": 1125
    },
    {
      "epoch": 0.9365934521342727,
      "grad_norm": 2.01389490087078,
      "learning_rate": 1.205741486324552e-07,
      "loss": 0.3397,
      "step": 1130
    },
    {
      "epoch": 0.9407376709490262,
      "grad_norm": 2.1289409520473326,
      "learning_rate": 1.0528509635458873e-07,
      "loss": 0.3401,
      "step": 1135
    },
    {
      "epoch": 0.9448818897637795,
      "grad_norm": 2.0835955722730697,
      "learning_rate": 9.102193241554757e-08,
      "loss": 0.3415,
      "step": 1140
    },
    {
      "epoch": 0.949026108578533,
      "grad_norm": 2.122449489250868,
      "learning_rate": 7.778764624655433e-08,
      "loss": 0.3437,
      "step": 1145
    },
    {
      "epoch": 0.9531703273932863,
      "grad_norm": 2.0577943752263836,
      "learning_rate": 6.558501163527964e-08,
      "loss": 0.3388,
      "step": 1150
    },
    {
      "epoch": 0.9573145462080398,
      "grad_norm": 2.126069664831503,
      "learning_rate": 5.44165861444812e-08,
      "loss": 0.337,
      "step": 1155
    },
    {
      "epoch": 0.9614587650227931,
      "grad_norm": 1.998215049010736,
      "learning_rate": 4.428471057596362e-08,
      "loss": 0.3378,
      "step": 1160
    },
    {
      "epoch": 0.9656029838375466,
      "grad_norm": 2.0929533756003904,
      "learning_rate": 3.519150847996422e-08,
      "loss": 0.3379,
      "step": 1165
    },
    {
      "epoch": 0.9697472026523001,
      "grad_norm": 2.0561899209040186,
      "learning_rate": 2.713888571007739e-08,
      "loss": 0.3388,
      "step": 1170
    },
    {
      "epoch": 0.9738914214670534,
      "grad_norm": 1.9664762265851157,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.3396,
      "step": 1175
    },
    {
      "epoch": 0.9780356402818069,
      "grad_norm": 1.999931793720276,
      "learning_rate": 1.4161910728816009e-08,
      "loss": 0.3408,
      "step": 1180
    },
    {
      "epoch": 0.9821798590965602,
      "grad_norm": 2.099348366631809,
      "learning_rate": 9.240278374995637e-09,
      "loss": 0.3438,
      "step": 1185
    },
    {
      "epoch": 0.9863240779113137,
      "grad_norm": 2.0961585662402937,
      "learning_rate": 5.364664492337746e-09,
      "loss": 0.3382,
      "step": 1190
    },
    {
      "epoch": 0.9904682967260672,
      "grad_norm": 2.0713909816710956,
      "learning_rate": 2.5358813747500266e-09,
      "loss": 0.3392,
      "step": 1195
    },
    {
      "epoch": 0.9946125155408205,
      "grad_norm": 2.0497170104333793,
      "learning_rate": 7.545219097987444e-10,
      "loss": 0.333,
      "step": 1200
    },
    {
      "epoch": 0.998756734355574,
      "grad_norm": 2.039086392116031,
      "learning_rate": 2.0959454449243076e-11,
      "loss": 0.3409,
      "step": 1205
    },
    {
      "epoch": 0.9995855781185247,
      "eval_loss": 0.3220805823802948,
      "eval_runtime": 1.1842,
      "eval_samples_per_second": 2.533,
      "eval_steps_per_second": 0.844,
      "step": 1206
    },
    {
      "epoch": 0.9995855781185247,
      "step": 1206,
      "total_flos": 252407174922240.0,
      "train_loss": 0.5270945630461027,
      "train_runtime": 23848.9709,
      "train_samples_per_second": 1.619,
      "train_steps_per_second": 0.051
    }
  ],
  "logging_steps": 5,
  "max_steps": 1206,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 252407174922240.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}