{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2012329805822042,
  "eval_steps": 500,
  "global_step": 6104,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010219892526291499,
      "grad_norm": 52.572940826416016,
      "learning_rate": 1.0157273918741808e-06,
      "loss": 9.6136,
      "step": 31
    },
    {
      "epoch": 0.0020439785052582997,
      "grad_norm": 26.900487899780273,
      "learning_rate": 2.0314547837483616e-06,
      "loss": 7.35,
      "step": 62
    },
    {
      "epoch": 0.003065967757887449,
      "grad_norm": 15.413382530212402,
      "learning_rate": 3.0471821756225426e-06,
      "loss": 5.872,
      "step": 93
    },
    {
      "epoch": 0.0040879570105165994,
      "grad_norm": 9.475220680236816,
      "learning_rate": 4.062909567496723e-06,
      "loss": 4.8148,
      "step": 124
    },
    {
      "epoch": 0.005109946263145749,
      "grad_norm": 16.33511734008789,
      "learning_rate": 5.078636959370905e-06,
      "loss": 4.2357,
      "step": 155
    },
    {
      "epoch": 0.006131935515774898,
      "grad_norm": 17.79788589477539,
      "learning_rate": 6.094364351245085e-06,
      "loss": 3.9005,
      "step": 186
    },
    {
      "epoch": 0.007153924768404049,
      "grad_norm": 7.385252952575684,
      "learning_rate": 7.110091743119267e-06,
      "loss": 3.6349,
      "step": 217
    },
    {
      "epoch": 0.008175914021033199,
      "grad_norm": 7.6574835777282715,
      "learning_rate": 8.125819134993446e-06,
      "loss": 3.4197,
      "step": 248
    },
    {
      "epoch": 0.009197903273662348,
      "grad_norm": 10.027671813964844,
      "learning_rate": 9.141546526867629e-06,
      "loss": 3.2883,
      "step": 279
    },
    {
      "epoch": 0.010219892526291498,
      "grad_norm": 7.447092056274414,
      "learning_rate": 1.015727391874181e-05,
      "loss": 3.1606,
      "step": 310
    },
    {
      "epoch": 0.011241881778920647,
      "grad_norm": 8.796594619750977,
      "learning_rate": 1.117300131061599e-05,
      "loss": 3.0724,
      "step": 341
    },
    {
      "epoch": 0.012263871031549797,
      "grad_norm": 8.47999382019043,
      "learning_rate": 1.218872870249017e-05,
      "loss": 3.0249,
      "step": 372
    },
    {
      "epoch": 0.013285860284178948,
      "grad_norm": 11.703062057495117,
      "learning_rate": 1.3204456094364351e-05,
      "loss": 2.9383,
      "step": 403
    },
    {
      "epoch": 0.014307849536808097,
      "grad_norm": 7.270478248596191,
      "learning_rate": 1.4220183486238533e-05,
      "loss": 2.8753,
      "step": 434
    },
    {
      "epoch": 0.015329838789437247,
      "grad_norm": 5.699134349822998,
      "learning_rate": 1.5235910878112714e-05,
      "loss": 2.8226,
      "step": 465
    },
    {
      "epoch": 0.016351828042066398,
      "grad_norm": 6.3912529945373535,
      "learning_rate": 1.6251638269986893e-05,
      "loss": 2.7754,
      "step": 496
    },
    {
      "epoch": 0.017373817294695545,
      "grad_norm": 8.240257263183594,
      "learning_rate": 1.7267365661861077e-05,
      "loss": 2.7351,
      "step": 527
    },
    {
      "epoch": 0.018395806547324697,
      "grad_norm": 9.017704010009766,
      "learning_rate": 1.8283093053735257e-05,
      "loss": 2.7084,
      "step": 558
    },
    {
      "epoch": 0.019417795799953844,
      "grad_norm": 6.017426013946533,
      "learning_rate": 1.9298820445609438e-05,
      "loss": 2.6741,
      "step": 589
    },
    {
      "epoch": 0.020439785052582995,
      "grad_norm": 3.789416790008545,
      "learning_rate": 2.031454783748362e-05,
      "loss": 2.6605,
      "step": 620
    },
    {
      "epoch": 0.021461774305212147,
      "grad_norm": 4.747366428375244,
      "learning_rate": 2.13302752293578e-05,
      "loss": 2.6245,
      "step": 651
    },
    {
      "epoch": 0.022483763557841294,
      "grad_norm": 3.8344268798828125,
      "learning_rate": 2.234600262123198e-05,
      "loss": 2.5791,
      "step": 682
    },
    {
      "epoch": 0.023505752810470446,
      "grad_norm": 4.5539116859436035,
      "learning_rate": 2.336173001310616e-05,
      "loss": 2.5648,
      "step": 713
    },
    {
      "epoch": 0.024527742063099593,
      "grad_norm": 4.216089248657227,
      "learning_rate": 2.437745740498034e-05,
      "loss": 2.5125,
      "step": 744
    },
    {
      "epoch": 0.025549731315728744,
      "grad_norm": 4.225554466247559,
      "learning_rate": 2.5393184796854525e-05,
      "loss": 2.4973,
      "step": 775
    },
    {
      "epoch": 0.026571720568357896,
      "grad_norm": 3.357351541519165,
      "learning_rate": 2.6408912188728702e-05,
      "loss": 2.4532,
      "step": 806
    },
    {
      "epoch": 0.027593709820987043,
      "grad_norm": 4.109170436859131,
      "learning_rate": 2.7424639580602886e-05,
      "loss": 2.4504,
      "step": 837
    },
    {
      "epoch": 0.028615699073616194,
      "grad_norm": 17.247940063476562,
      "learning_rate": 2.8440366972477066e-05,
      "loss": 2.4226,
      "step": 868
    },
    {
      "epoch": 0.029637688326245342,
      "grad_norm": 3.2851662635803223,
      "learning_rate": 2.9456094364351244e-05,
      "loss": 2.4053,
      "step": 899
    },
    {
      "epoch": 0.030659677578874493,
      "grad_norm": 3.3796141147613525,
      "learning_rate": 3.0471821756225428e-05,
      "loss": 2.3977,
      "step": 930
    },
    {
      "epoch": 0.03168166683150364,
      "grad_norm": 3.0984764099121094,
      "learning_rate": 3.148754914809961e-05,
      "loss": 2.3622,
      "step": 961
    },
    {
      "epoch": 0.032703656084132796,
      "grad_norm": 2.5464305877685547,
      "learning_rate": 3.2503276539973785e-05,
      "loss": 2.348,
      "step": 992
    },
    {
      "epoch": 0.03372564533676194,
      "grad_norm": 2.848860263824463,
      "learning_rate": 3.351900393184797e-05,
      "loss": 2.3208,
      "step": 1023
    },
    {
      "epoch": 0.03474763458939109,
      "grad_norm": 2.5870606899261475,
      "learning_rate": 3.453473132372215e-05,
      "loss": 2.3023,
      "step": 1054
    },
    {
      "epoch": 0.03576962384202024,
      "grad_norm": 2.6556804180145264,
      "learning_rate": 3.555045871559633e-05,
      "loss": 2.2698,
      "step": 1085
    },
    {
      "epoch": 0.03679161309464939,
      "grad_norm": 2.779650926589966,
      "learning_rate": 3.6566186107470514e-05,
      "loss": 2.2614,
      "step": 1116
    },
    {
      "epoch": 0.03781360234727854,
      "grad_norm": 2.624191999435425,
      "learning_rate": 3.7581913499344695e-05,
      "loss": 2.2422,
      "step": 1147
    },
    {
      "epoch": 0.03883559159990769,
      "grad_norm": 2.5255484580993652,
      "learning_rate": 3.8597640891218876e-05,
      "loss": 2.2419,
      "step": 1178
    },
    {
      "epoch": 0.03985758085253684,
      "grad_norm": 2.3195745944976807,
      "learning_rate": 3.9613368283093056e-05,
      "loss": 2.2468,
      "step": 1209
    },
    {
      "epoch": 0.04087957010516599,
      "grad_norm": 2.4712162017822266,
      "learning_rate": 4.062909567496724e-05,
      "loss": 2.2072,
      "step": 1240
    },
    {
      "epoch": 0.04190155935779514,
      "grad_norm": 4.451296806335449,
      "learning_rate": 4.164482306684142e-05,
      "loss": 2.2136,
      "step": 1271
    },
    {
      "epoch": 0.04292354861042429,
      "grad_norm": 2.4180150032043457,
      "learning_rate": 4.26605504587156e-05,
      "loss": 2.1843,
      "step": 1302
    },
    {
      "epoch": 0.04394553786305344,
      "grad_norm": 2.395840883255005,
      "learning_rate": 4.367627785058978e-05,
      "loss": 2.1447,
      "step": 1333
    },
    {
      "epoch": 0.04496752711568259,
      "grad_norm": 3.072429895401001,
      "learning_rate": 4.469200524246396e-05,
      "loss": 2.1405,
      "step": 1364
    },
    {
      "epoch": 0.045989516368311736,
      "grad_norm": 1.9566724300384521,
      "learning_rate": 4.570773263433814e-05,
      "loss": 2.1321,
      "step": 1395
    },
    {
      "epoch": 0.04701150562094089,
      "grad_norm": 23.60377311706543,
      "learning_rate": 4.672346002621232e-05,
      "loss": 2.1289,
      "step": 1426
    },
    {
      "epoch": 0.04803349487357004,
      "grad_norm": 2.117250919342041,
      "learning_rate": 4.77391874180865e-05,
      "loss": 2.1239,
      "step": 1457
    },
    {
      "epoch": 0.049055484126199186,
      "grad_norm": 2.361362934112549,
      "learning_rate": 4.875491480996068e-05,
      "loss": 2.0838,
      "step": 1488
    },
    {
      "epoch": 0.05007747337882834,
      "grad_norm": 2.1867992877960205,
      "learning_rate": 4.977064220183487e-05,
      "loss": 2.0622,
      "step": 1519
    },
    {
      "epoch": 0.05109946263145749,
      "grad_norm": 1.9326456785202026,
      "learning_rate": 4.9999915451558777e-05,
      "loss": 2.0441,
      "step": 1550
    },
    {
      "epoch": 0.052121451884086636,
      "grad_norm": 2.076503038406372,
      "learning_rate": 4.999955597496219e-05,
      "loss": 2.0503,
      "step": 1581
    },
    {
      "epoch": 0.05314344113671579,
      "grad_norm": 3.051212787628174,
      "learning_rate": 4.9998914381774255e-05,
      "loss": 2.0357,
      "step": 1612
    },
    {
      "epoch": 0.05416543038934494,
      "grad_norm": 1.7924102544784546,
      "learning_rate": 4.999799067923527e-05,
      "loss": 2.0326,
      "step": 1643
    },
    {
      "epoch": 0.055187419641974086,
      "grad_norm": 1.8103365898132324,
      "learning_rate": 4.999678487776908e-05,
      "loss": 2.0023,
      "step": 1674
    },
    {
      "epoch": 0.056209408894603234,
      "grad_norm": 1.913725733757019,
      "learning_rate": 4.9995296990983006e-05,
      "loss": 2.0035,
      "step": 1705
    },
    {
      "epoch": 0.05723139814723239,
      "grad_norm": 1.7415859699249268,
      "learning_rate": 4.999352703566763e-05,
      "loss": 2.0063,
      "step": 1736
    },
    {
      "epoch": 0.058253387399861536,
      "grad_norm": 1.8145827054977417,
      "learning_rate": 4.999147503179668e-05,
      "loss": 1.9731,
      "step": 1767
    },
    {
      "epoch": 0.059275376652490684,
      "grad_norm": 1.8761731386184692,
      "learning_rate": 4.998914100252672e-05,
      "loss": 1.9883,
      "step": 1798
    },
    {
      "epoch": 0.06029736590511984,
      "grad_norm": 6.824073791503906,
      "learning_rate": 4.998652497419696e-05,
      "loss": 1.9713,
      "step": 1829
    },
    {
      "epoch": 0.061319355157748986,
      "grad_norm": 1.7231906652450562,
      "learning_rate": 4.9983626976328927e-05,
      "loss": 1.9492,
      "step": 1860
    },
    {
      "epoch": 0.062341344410378134,
      "grad_norm": 1.6888827085494995,
      "learning_rate": 4.998044704162613e-05,
      "loss": 1.9449,
      "step": 1891
    },
    {
      "epoch": 0.06336333366300728,
      "grad_norm": 1.5924124717712402,
      "learning_rate": 4.9976985205973705e-05,
      "loss": 1.938,
      "step": 1922
    },
    {
      "epoch": 0.06438532291563644,
      "grad_norm": 1.7029409408569336,
      "learning_rate": 4.997324150843799e-05,
      "loss": 1.9285,
      "step": 1953
    },
    {
      "epoch": 0.06540731216826559,
      "grad_norm": 1.5554330348968506,
      "learning_rate": 4.99692159912661e-05,
      "loss": 1.9091,
      "step": 1984
    },
    {
      "epoch": 0.06642930142089473,
      "grad_norm": 1.5639121532440186,
      "learning_rate": 4.996490869988546e-05,
      "loss": 1.9105,
      "step": 2015
    },
    {
      "epoch": 0.06745129067352389,
      "grad_norm": 1.6262747049331665,
      "learning_rate": 4.996031968290326e-05,
      "loss": 1.9164,
      "step": 2046
    },
    {
      "epoch": 0.06847327992615304,
      "grad_norm": 1.5408495664596558,
      "learning_rate": 4.995544899210594e-05,
      "loss": 1.8728,
      "step": 2077
    },
    {
      "epoch": 0.06949526917878218,
      "grad_norm": 1.5183970928192139,
      "learning_rate": 4.9950296682458583e-05,
      "loss": 1.8854,
      "step": 2108
    },
    {
      "epoch": 0.07051725843141134,
      "grad_norm": 1.9362810850143433,
      "learning_rate": 4.994486281210429e-05,
      "loss": 1.8811,
      "step": 2139
    },
    {
      "epoch": 0.07153924768404048,
      "grad_norm": 1.5646640062332153,
      "learning_rate": 4.9939147442363566e-05,
      "loss": 1.8744,
      "step": 2170
    },
    {
      "epoch": 0.07256123693666963,
      "grad_norm": 1.5929124355316162,
      "learning_rate": 4.9933150637733574e-05,
      "loss": 1.867,
      "step": 2201
    },
    {
      "epoch": 0.07358322618929879,
      "grad_norm": 1.7689995765686035,
      "learning_rate": 4.992687246588743e-05,
      "loss": 1.8659,
      "step": 2232
    },
    {
      "epoch": 0.07460521544192793,
      "grad_norm": 1.375406265258789,
      "learning_rate": 4.992031299767347e-05,
      "loss": 1.8477,
      "step": 2263
    },
    {
      "epoch": 0.07562720469455708,
      "grad_norm": 1.3596042394638062,
      "learning_rate": 4.9913472307114386e-05,
      "loss": 1.8498,
      "step": 2294
    },
    {
      "epoch": 0.07664919394718624,
      "grad_norm": 1.3918544054031372,
      "learning_rate": 4.9906350471406446e-05,
      "loss": 1.8538,
      "step": 2325
    },
    {
      "epoch": 0.07767118319981538,
      "grad_norm": 1.4614112377166748,
      "learning_rate": 4.989894757091861e-05,
      "loss": 1.8359,
      "step": 2356
    },
    {
      "epoch": 0.07869317245244453,
      "grad_norm": 1.563186764717102,
      "learning_rate": 4.989126368919158e-05,
      "loss": 1.821,
      "step": 2387
    },
    {
      "epoch": 0.07971516170507369,
      "grad_norm": 1.3217226266860962,
      "learning_rate": 4.988329891293693e-05,
      "loss": 1.8345,
      "step": 2418
    },
    {
      "epoch": 0.08073715095770283,
      "grad_norm": 1.3906276226043701,
      "learning_rate": 4.987505333203608e-05,
      "loss": 1.8313,
      "step": 2449
    },
    {
      "epoch": 0.08175914021033198,
      "grad_norm": 1.2736890316009521,
      "learning_rate": 4.9866527039539276e-05,
      "loss": 1.8234,
      "step": 2480
    },
    {
      "epoch": 0.08278112946296114,
      "grad_norm": 1.3050850629806519,
      "learning_rate": 4.9857720131664594e-05,
      "loss": 1.8181,
      "step": 2511
    },
    {
      "epoch": 0.08380311871559028,
      "grad_norm": 1.3847019672393799,
      "learning_rate": 4.9848632707796773e-05,
      "loss": 1.8128,
      "step": 2542
    },
    {
      "epoch": 0.08482510796821943,
      "grad_norm": 1.3893433809280396,
      "learning_rate": 4.9839264870486155e-05,
      "loss": 1.7871,
      "step": 2573
    },
    {
      "epoch": 0.08584709722084859,
      "grad_norm": 1.3469080924987793,
      "learning_rate": 4.9829616725447526e-05,
      "loss": 1.8212,
      "step": 2604
    },
    {
      "epoch": 0.08686908647347773,
      "grad_norm": 1.269865870475769,
      "learning_rate": 4.981968838155888e-05,
      "loss": 1.7868,
      "step": 2635
    },
    {
      "epoch": 0.08789107572610688,
      "grad_norm": 1.2972242832183838,
      "learning_rate": 4.980947995086024e-05,
      "loss": 1.7987,
      "step": 2666
    },
    {
      "epoch": 0.08891306497873604,
      "grad_norm": 1.3488340377807617,
      "learning_rate": 4.979899154855234e-05,
      "loss": 1.7859,
      "step": 2697
    },
    {
      "epoch": 0.08993505423136518,
      "grad_norm": 1.3487133979797363,
      "learning_rate": 4.9788223292995386e-05,
      "loss": 1.7851,
      "step": 2728
    },
    {
      "epoch": 0.09095704348399433,
      "grad_norm": 1.2412410974502563,
      "learning_rate": 4.977717530570768e-05,
      "loss": 1.7802,
      "step": 2759
    },
    {
      "epoch": 0.09197903273662347,
      "grad_norm": 1.3017562627792358,
      "learning_rate": 4.976584771136425e-05,
      "loss": 1.775,
      "step": 2790
    },
    {
      "epoch": 0.09300102198925263,
      "grad_norm": 1.2923076152801514,
      "learning_rate": 4.975424063779547e-05,
      "loss": 1.76,
      "step": 2821
    },
    {
      "epoch": 0.09402301124188178,
      "grad_norm": 1.2735223770141602,
      "learning_rate": 4.974235421598557e-05,
      "loss": 1.7604,
      "step": 2852
    },
    {
      "epoch": 0.09504500049451092,
      "grad_norm": 1.303673505783081,
      "learning_rate": 4.973018858007122e-05,
      "loss": 1.7685,
      "step": 2883
    },
    {
      "epoch": 0.09606698974714008,
      "grad_norm": 1.5336410999298096,
      "learning_rate": 4.9717743867339963e-05,
      "loss": 1.7682,
      "step": 2914
    },
    {
      "epoch": 0.09708897899976923,
      "grad_norm": 5.014227867126465,
      "learning_rate": 4.9705020218228695e-05,
      "loss": 1.7617,
      "step": 2945
    },
    {
      "epoch": 0.09811096825239837,
      "grad_norm": 1.2603938579559326,
      "learning_rate": 4.969201777632205e-05,
      "loss": 1.7621,
      "step": 2976
    },
    {
      "epoch": 0.09913295750502753,
      "grad_norm": 1.32491934299469,
      "learning_rate": 4.9678736688350846e-05,
      "loss": 1.7411,
      "step": 3007
    },
    {
      "epoch": 0.10015494675765668,
      "grad_norm": 1.206735372543335,
      "learning_rate": 4.966517710419033e-05,
      "loss": 1.7262,
      "step": 3038
    },
    {
      "epoch": 0.10117693601028582,
      "grad_norm": 1.254231572151184,
      "learning_rate": 4.965133917685858e-05,
      "loss": 1.7454,
      "step": 3069
    },
    {
      "epoch": 0.10219892526291498,
      "grad_norm": 1.3085408210754395,
      "learning_rate": 4.9637223062514714e-05,
      "loss": 1.7517,
      "step": 3100
    },
    {
      "epoch": 0.10322091451554413,
      "grad_norm": 1.3184605836868286,
      "learning_rate": 4.962282892045718e-05,
      "loss": 1.7542,
      "step": 3131
    },
    {
      "epoch": 0.10424290376817327,
      "grad_norm": 1.2284983396530151,
      "learning_rate": 4.9608156913121904e-05,
      "loss": 1.738,
      "step": 3162
    },
    {
      "epoch": 0.10526489302080243,
      "grad_norm": 1.3870880603790283,
      "learning_rate": 4.959320720608049e-05,
      "loss": 1.7358,
      "step": 3193
    },
    {
      "epoch": 0.10628688227343158,
      "grad_norm": 1.1335322856903076,
      "learning_rate": 4.9577979968038354e-05,
      "loss": 1.742,
      "step": 3224
    },
    {
      "epoch": 0.10730887152606072,
      "grad_norm": 1.1794465780258179,
      "learning_rate": 4.956247537083282e-05,
      "loss": 1.7357,
      "step": 3255
    },
    {
      "epoch": 0.10833086077868988,
      "grad_norm": 1.1701149940490723,
      "learning_rate": 4.9546693589431145e-05,
      "loss": 1.7289,
      "step": 3286
    },
    {
      "epoch": 0.10935285003131903,
      "grad_norm": 1.169094443321228,
      "learning_rate": 4.9530634801928595e-05,
      "loss": 1.7238,
      "step": 3317
    },
    {
      "epoch": 0.11037483928394817,
      "grad_norm": 1.2735379934310913,
      "learning_rate": 4.9514299189546395e-05,
      "loss": 1.7275,
      "step": 3348
    },
    {
      "epoch": 0.11139682853657733,
      "grad_norm": 1.2806981801986694,
      "learning_rate": 4.949768693662973e-05,
      "loss": 1.7135,
      "step": 3379
    },
    {
      "epoch": 0.11241881778920647,
      "grad_norm": 1.208024263381958,
      "learning_rate": 4.948079823064559e-05,
      "loss": 1.7251,
      "step": 3410
    },
    {
      "epoch": 0.11344080704183562,
      "grad_norm": 1.275516152381897,
      "learning_rate": 4.946363326218074e-05,
      "loss": 1.7259,
      "step": 3441
    },
    {
      "epoch": 0.11446279629446478,
      "grad_norm": 1.1761465072631836,
      "learning_rate": 4.9446192224939525e-05,
      "loss": 1.7024,
      "step": 3472
    },
    {
      "epoch": 0.11548478554709392,
      "grad_norm": 1.1216075420379639,
      "learning_rate": 4.942847531574167e-05,
      "loss": 1.6849,
      "step": 3503
    },
    {
      "epoch": 0.11650677479972307,
      "grad_norm": 1.1286563873291016,
      "learning_rate": 4.941048273452008e-05,
      "loss": 1.7004,
      "step": 3534
    },
    {
      "epoch": 0.11752876405235223,
      "grad_norm": 1.1531614065170288,
      "learning_rate": 4.9392214684318605e-05,
      "loss": 1.7101,
      "step": 3565
    },
    {
      "epoch": 0.11855075330498137,
      "grad_norm": 1.143051266670227,
      "learning_rate": 4.93736713712897e-05,
      "loss": 1.7127,
      "step": 3596
    },
    {
      "epoch": 0.11957274255761052,
      "grad_norm": 1.1483272314071655,
      "learning_rate": 4.9354853004692124e-05,
      "loss": 1.6968,
      "step": 3627
    },
    {
      "epoch": 0.12059473181023968,
      "grad_norm": 1.1267555952072144,
      "learning_rate": 4.93357597968886e-05,
      "loss": 1.7023,
      "step": 3658
    },
    {
      "epoch": 0.12161672106286882,
      "grad_norm": 1.1552249193191528,
      "learning_rate": 4.931639196334338e-05,
      "loss": 1.7171,
      "step": 3689
    },
    {
      "epoch": 0.12263871031549797,
      "grad_norm": 1.1388061046600342,
      "learning_rate": 4.9296749722619826e-05,
      "loss": 1.7061,
      "step": 3720
    },
    {
      "epoch": 0.12366069956812713,
      "grad_norm": 1.127455234527588,
      "learning_rate": 4.9276833296377966e-05,
      "loss": 1.6879,
      "step": 3751
    },
    {
      "epoch": 0.12468268882075627,
      "grad_norm": 1.1534373760223389,
      "learning_rate": 4.925664290937196e-05,
      "loss": 1.7091,
      "step": 3782
    },
    {
      "epoch": 0.12570467807338542,
      "grad_norm": 1.2369154691696167,
      "learning_rate": 4.9236178789447576e-05,
      "loss": 1.6923,
      "step": 3813
    },
    {
      "epoch": 0.12672666732601456,
      "grad_norm": 1.074436068534851,
      "learning_rate": 4.921544116753962e-05,
      "loss": 1.6805,
      "step": 3844
    },
    {
      "epoch": 0.12774865657864373,
      "grad_norm": 1.112226963043213,
      "learning_rate": 4.919443027766935e-05,
      "loss": 1.6527,
      "step": 3875
    },
    {
      "epoch": 0.12877064583127287,
      "grad_norm": 1.0999863147735596,
      "learning_rate": 4.91731463569418e-05,
      "loss": 1.6614,
      "step": 3906
    },
    {
      "epoch": 0.129792635083902,
      "grad_norm": 1.098036527633667,
      "learning_rate": 4.915158964554312e-05,
      "loss": 1.6826,
      "step": 3937
    },
    {
      "epoch": 0.13081462433653118,
      "grad_norm": 1.1108450889587402,
      "learning_rate": 4.912976038673786e-05,
      "loss": 1.6886,
      "step": 3968
    },
    {
      "epoch": 0.13183661358916032,
      "grad_norm": 1.0916872024536133,
      "learning_rate": 4.9107658826866254e-05,
      "loss": 1.6782,
      "step": 3999
    },
    {
      "epoch": 0.13285860284178946,
      "grad_norm": 1.0818581581115723,
      "learning_rate": 4.908528521534139e-05,
      "loss": 1.6796,
      "step": 4030
    },
    {
      "epoch": 0.13388059209441863,
      "grad_norm": 1.0908610820770264,
      "learning_rate": 4.906263980464644e-05,
      "loss": 1.6662,
      "step": 4061
    },
    {
      "epoch": 0.13490258134704777,
      "grad_norm": 1.0384143590927124,
      "learning_rate": 4.903972285033178e-05,
      "loss": 1.6811,
      "step": 4092
    },
    {
      "epoch": 0.1359245705996769,
      "grad_norm": 1.0998533964157104,
      "learning_rate": 4.901653461101213e-05,
      "loss": 1.6817,
      "step": 4123
    },
    {
      "epoch": 0.13694655985230608,
      "grad_norm": 1.1726231575012207,
      "learning_rate": 4.8993075348363626e-05,
      "loss": 1.6532,
      "step": 4154
    },
    {
      "epoch": 0.13796854910493522,
      "grad_norm": 1.075464129447937,
      "learning_rate": 4.896934532712084e-05,
      "loss": 1.671,
      "step": 4185
    },
    {
      "epoch": 0.13899053835756436,
      "grad_norm": 1.0557868480682373,
      "learning_rate": 4.8945344815073846e-05,
      "loss": 1.6548,
      "step": 4216
    },
    {
      "epoch": 0.14001252761019353,
      "grad_norm": 1.0531095266342163,
      "learning_rate": 4.892107408306516e-05,
      "loss": 1.6526,
      "step": 4247
    },
    {
      "epoch": 0.14103451686282267,
      "grad_norm": 1.138203501701355,
      "learning_rate": 4.889653340498669e-05,
      "loss": 1.678,
      "step": 4278
    },
    {
      "epoch": 0.1420565061154518,
      "grad_norm": 1.0668121576309204,
      "learning_rate": 4.8871723057776664e-05,
      "loss": 1.6552,
      "step": 4309
    },
    {
      "epoch": 0.14307849536808095,
      "grad_norm": 1.0312261581420898,
      "learning_rate": 4.8846643321416476e-05,
      "loss": 1.6731,
      "step": 4340
    },
    {
      "epoch": 0.14410048462071012,
      "grad_norm": 1.0868667364120483,
      "learning_rate": 4.882129447892753e-05,
      "loss": 1.6713,
      "step": 4371
    },
    {
      "epoch": 0.14512247387333926,
      "grad_norm": 1.0213130712509155,
      "learning_rate": 4.8795676816368076e-05,
      "loss": 1.6515,
      "step": 4402
    },
    {
      "epoch": 0.1461444631259684,
      "grad_norm": 1.190875768661499,
      "learning_rate": 4.876979062282995e-05,
      "loss": 1.651,
      "step": 4433
    },
    {
      "epoch": 0.14716645237859757,
      "grad_norm": 1.0551568269729614,
      "learning_rate": 4.8743636190435325e-05,
      "loss": 1.6751,
      "step": 4464
    },
    {
      "epoch": 0.1481884416312267,
      "grad_norm": 1.0654323101043701,
      "learning_rate": 4.871721381433344e-05,
      "loss": 1.6499,
      "step": 4495
    },
    {
      "epoch": 0.14921043088385585,
      "grad_norm": 1.04425048828125,
      "learning_rate": 4.869052379269719e-05,
      "loss": 1.6401,
      "step": 4526
    },
    {
      "epoch": 0.15023242013648502,
      "grad_norm": 1.3416290283203125,
      "learning_rate": 4.866356642671985e-05,
      "loss": 1.6553,
      "step": 4557
    },
    {
      "epoch": 0.15125440938911416,
      "grad_norm": 1.073529601097107,
      "learning_rate": 4.8636342020611634e-05,
      "loss": 1.6413,
      "step": 4588
    },
    {
      "epoch": 0.1522763986417433,
      "grad_norm": 1.0386462211608887,
      "learning_rate": 4.860885088159626e-05,
      "loss": 1.6595,
      "step": 4619
    },
    {
      "epoch": 0.15329838789437247,
      "grad_norm": 1.0544514656066895,
      "learning_rate": 4.858109331990751e-05,
      "loss": 1.6387,
      "step": 4650
    },
    {
      "epoch": 0.15432037714700161,
      "grad_norm": 1.088112473487854,
      "learning_rate": 4.855306964878567e-05,
      "loss": 1.628,
      "step": 4681
    },
    {
      "epoch": 0.15534236639963075,
      "grad_norm": 0.9930492639541626,
      "learning_rate": 4.8524780184474084e-05,
      "loss": 1.6554,
      "step": 4712
    },
    {
      "epoch": 0.15636435565225992,
      "grad_norm": 0.9864984154701233,
      "learning_rate": 4.8496225246215496e-05,
      "loss": 1.6558,
      "step": 4743
    },
    {
      "epoch": 0.15738634490488906,
      "grad_norm": 1.0170128345489502,
      "learning_rate": 4.8467405156248505e-05,
      "loss": 1.6289,
      "step": 4774
    },
    {
      "epoch": 0.1584083341575182,
      "grad_norm": 1.0360223054885864,
      "learning_rate": 4.843832023980392e-05,
      "loss": 1.6314,
      "step": 4805
    },
    {
      "epoch": 0.15943032341014737,
      "grad_norm": 1.0165129899978638,
      "learning_rate": 4.840897082510106e-05,
      "loss": 1.6294,
      "step": 4836
    },
    {
      "epoch": 0.16045231266277651,
      "grad_norm": 1.079991102218628,
      "learning_rate": 4.8379357243344084e-05,
      "loss": 1.6204,
      "step": 4867
    },
    {
      "epoch": 0.16147430191540565,
      "grad_norm": 1.0515645742416382,
      "learning_rate": 4.8349479828718236e-05,
      "loss": 1.6322,
      "step": 4898
    },
    {
      "epoch": 0.16249629116803482,
      "grad_norm": 1.0226655006408691,
      "learning_rate": 4.8319338918386075e-05,
      "loss": 1.6476,
      "step": 4929
    },
    {
      "epoch": 0.16351828042066396,
      "grad_norm": 1.0581114292144775,
      "learning_rate": 4.828893485248369e-05,
      "loss": 1.6302,
      "step": 4960
    },
    {
      "epoch": 0.1645402696732931,
      "grad_norm": 1.0156742334365845,
      "learning_rate": 4.825826797411682e-05,
      "loss": 1.6292,
      "step": 4991
    },
    {
      "epoch": 0.16556225892592227,
      "grad_norm": 1.0153559446334839,
      "learning_rate": 4.822733862935702e-05,
      "loss": 1.6367,
      "step": 5022
    },
    {
      "epoch": 0.16658424817855141,
      "grad_norm": 1.0488505363464355,
      "learning_rate": 4.819614716723775e-05,
      "loss": 1.644,
      "step": 5053
    },
    {
      "epoch": 0.16760623743118055,
      "grad_norm": 1.092781901359558,
      "learning_rate": 4.8164693939750425e-05,
      "loss": 1.6318,
      "step": 5084
    },
    {
      "epoch": 0.16862822668380972,
      "grad_norm": 0.9972744584083557,
      "learning_rate": 4.813297930184042e-05,
      "loss": 1.6206,
      "step": 5115
    },
    {
      "epoch": 0.16965021593643886,
      "grad_norm": 1.0455750226974487,
      "learning_rate": 4.810100361140314e-05,
      "loss": 1.6296,
      "step": 5146
    },
    {
      "epoch": 0.170672205189068,
      "grad_norm": 0.9624122977256775,
      "learning_rate": 4.8068767229279885e-05,
      "loss": 1.6307,
      "step": 5177
    },
    {
      "epoch": 0.17169419444169717,
      "grad_norm": 1.4828526973724365,
      "learning_rate": 4.8036270519253854e-05,
      "loss": 1.6377,
      "step": 5208
    },
    {
      "epoch": 0.17271618369432631,
      "grad_norm": 1.0732771158218384,
      "learning_rate": 4.8003513848046e-05,
      "loss": 1.6148,
      "step": 5239
    },
    {
      "epoch": 0.17373817294695545,
      "grad_norm": 1.0065757036209106,
      "learning_rate": 4.79704975853109e-05,
      "loss": 1.6192,
      "step": 5270
    },
    {
      "epoch": 0.17476016219958462,
      "grad_norm": 1.7071099281311035,
      "learning_rate": 4.793722210363262e-05,
      "loss": 1.6205,
      "step": 5301
    },
    {
      "epoch": 0.17578215145221376,
      "grad_norm": 1.009507417678833,
      "learning_rate": 4.7903687778520414e-05,
      "loss": 1.6261,
      "step": 5332
    },
    {
      "epoch": 0.1768041407048429,
      "grad_norm": 1.0593280792236328,
      "learning_rate": 4.7869894988404593e-05,
      "loss": 1.6286,
      "step": 5363
    },
    {
      "epoch": 0.17782612995747207,
      "grad_norm": 1.0053679943084717,
      "learning_rate": 4.783584411463221e-05,
      "loss": 1.6424,
      "step": 5394
    },
    {
      "epoch": 0.17884811921010121,
      "grad_norm": 0.983214795589447,
      "learning_rate": 4.780153554146274e-05,
      "loss": 1.6292,
      "step": 5425
    },
    {
      "epoch": 0.17987010846273035,
      "grad_norm": 0.9617491960525513,
      "learning_rate": 4.7766969656063766e-05,
      "loss": 1.6182,
      "step": 5456
    },
    {
      "epoch": 0.18089209771535952,
      "grad_norm": 1.0862797498703003,
      "learning_rate": 4.773214684850662e-05,
      "loss": 1.6213,
      "step": 5487
    },
    {
      "epoch": 0.18191408696798866,
      "grad_norm": 1.0468218326568604,
      "learning_rate": 4.769706751176193e-05,
      "loss": 1.6176,
      "step": 5518
    },
    {
      "epoch": 0.1829360762206178,
      "grad_norm": 0.9474911093711853,
      "learning_rate": 4.7661732041695264e-05,
      "loss": 1.5864,
      "step": 5549
    },
    {
      "epoch": 0.18395806547324695,
      "grad_norm": 0.9839109778404236,
      "learning_rate": 4.762614083706258e-05,
      "loss": 1.6177,
      "step": 5580
    },
    {
      "epoch": 0.18498005472587611,
      "grad_norm": 1.9305601119995117,
      "learning_rate": 4.759029429950581e-05,
      "loss": 1.6132,
      "step": 5611
    },
    {
      "epoch": 0.18600204397850525,
      "grad_norm": 0.9609850645065308,
      "learning_rate": 4.7554192833548235e-05,
      "loss": 1.6115,
      "step": 5642
    },
    {
      "epoch": 0.1870240332311344,
      "grad_norm": 1.0501559972763062,
      "learning_rate": 4.751783684659e-05,
      "loss": 1.6017,
      "step": 5673
    },
    {
      "epoch": 0.18804602248376356,
      "grad_norm": 0.9933464527130127,
      "learning_rate": 4.748122674890348e-05,
      "loss": 1.6136,
      "step": 5704
    },
    {
      "epoch": 0.1890680117363927,
      "grad_norm": 0.9463350772857666,
      "learning_rate": 4.7444362953628654e-05,
      "loss": 1.6102,
      "step": 5735
    },
    {
      "epoch": 0.19009000098902185,
      "grad_norm": 0.9974256753921509,
      "learning_rate": 4.7407245876768424e-05,
      "loss": 1.6101,
      "step": 5766
    },
    {
      "epoch": 0.19111199024165101,
      "grad_norm": 0.9747878313064575,
      "learning_rate": 4.736987593718397e-05,
      "loss": 1.5967,
      "step": 5797
    },
    {
      "epoch": 0.19213397949428015,
      "grad_norm": 0.9755719900131226,
      "learning_rate": 4.733225355658999e-05,
      "loss": 1.5987,
      "step": 5828
    },
    {
      "epoch": 0.1931559687469093,
      "grad_norm": 0.9605233669281006,
      "learning_rate": 4.7294379159549926e-05,
      "loss": 1.6238,
      "step": 5859
    },
    {
      "epoch": 0.19417795799953846,
      "grad_norm": 0.9777940511703491,
      "learning_rate": 4.725625317347119e-05,
      "loss": 1.5976,
      "step": 5890
    },
    {
      "epoch": 0.1951999472521676,
      "grad_norm": 0.9619265198707581,
      "learning_rate": 4.7217876028600374e-05,
      "loss": 1.5985,
      "step": 5921
    },
    {
      "epoch": 0.19622193650479675,
      "grad_norm": 0.970813512802124,
      "learning_rate": 4.717924815801832e-05,
      "loss": 1.6097,
      "step": 5952
    },
    {
      "epoch": 0.19724392575742591,
      "grad_norm": 1.1910721063613892,
      "learning_rate": 4.714036999763532e-05,
      "loss": 1.6156,
      "step": 5983
    },
    {
      "epoch": 0.19826591501005505,
      "grad_norm": 1.1106246709823608,
      "learning_rate": 4.7101241986186116e-05,
      "loss": 1.5761,
      "step": 6014
    },
    {
      "epoch": 0.1992879042626842,
      "grad_norm": 1.0065436363220215,
      "learning_rate": 4.7061864565225e-05,
      "loss": 1.5857,
      "step": 6045
    },
    {
      "epoch": 0.20030989351531336,
      "grad_norm": 0.9245477914810181,
      "learning_rate": 4.702223817912081e-05,
      "loss": 1.6099,
      "step": 6076
    }
  ],
  "logging_steps": 31,
  "max_steps": 30517,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 3052,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.263370839061168e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}