{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.49952696310312206,
  "eval_steps": 500,
  "global_step": 99,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005045726900031536,
      "grad_norm": 0.16816571847556824,
      "learning_rate": 2.9999839160139495e-06,
      "loss": 0.7782,
      "step": 1
    },
    {
      "epoch": 0.010091453800063072,
      "grad_norm": 0.1469143977253523,
      "learning_rate": 2.9999356645057024e-06,
      "loss": 0.6817,
      "step": 2
    },
    {
      "epoch": 0.015137180700094607,
      "grad_norm": 0.07996774677933757,
      "learning_rate": 2.9998552468249567e-06,
      "loss": 0.6735,
      "step": 3
    },
    {
      "epoch": 0.020182907600126143,
      "grad_norm": 0.0800127664777818,
      "learning_rate": 2.999742665221167e-06,
      "loss": 0.6569,
      "step": 4
    },
    {
      "epoch": 0.02522863450015768,
      "grad_norm": 0.08070188267575489,
      "learning_rate": 2.999597922843484e-06,
      "loss": 0.6283,
      "step": 5
    },
    {
      "epoch": 0.030274361400189215,
      "grad_norm": 0.06839180655145351,
      "learning_rate": 2.999421023740663e-06,
      "loss": 0.6446,
      "step": 6
    },
    {
      "epoch": 0.03532008830022075,
      "grad_norm": 0.05534188923028301,
      "learning_rate": 2.9992119728609516e-06,
      "loss": 0.6371,
      "step": 7
    },
    {
      "epoch": 0.040365815200252286,
      "grad_norm": 0.07094943987370793,
      "learning_rate": 2.9989707760519526e-06,
      "loss": 0.6111,
      "step": 8
    },
    {
      "epoch": 0.04541154210028382,
      "grad_norm": 0.06436005389698786,
      "learning_rate": 2.9986974400604593e-06,
      "loss": 0.588,
      "step": 9
    },
    {
      "epoch": 0.05045726900031536,
      "grad_norm": 0.05699223365274786,
      "learning_rate": 2.9983919725322667e-06,
      "loss": 0.6101,
      "step": 10
    },
    {
      "epoch": 0.055502995900346894,
      "grad_norm": 0.058843386030182285,
      "learning_rate": 2.9980543820119585e-06,
      "loss": 0.6047,
      "step": 11
    },
    {
      "epoch": 0.06054872280037843,
      "grad_norm": 0.047228554008764044,
      "learning_rate": 2.997684677942667e-06,
      "loss": 0.5937,
      "step": 12
    },
    {
      "epoch": 0.06559444970040997,
      "grad_norm": 0.04830399085917525,
      "learning_rate": 2.9972828706658102e-06,
      "loss": 0.6448,
      "step": 13
    },
    {
      "epoch": 0.0706401766004415,
      "grad_norm": 0.04469640349332499,
      "learning_rate": 2.996848971420801e-06,
      "loss": 0.6145,
      "step": 14
    },
    {
      "epoch": 0.07568590350047304,
      "grad_norm": 0.048907003957727534,
      "learning_rate": 2.996382992344734e-06,
      "loss": 0.5755,
      "step": 15
    },
    {
      "epoch": 0.08073163040050457,
      "grad_norm": 0.04502223888969105,
      "learning_rate": 2.9958849464720457e-06,
      "loss": 0.5765,
      "step": 16
    },
    {
      "epoch": 0.08577735730053611,
      "grad_norm": 0.04485565875842678,
      "learning_rate": 2.9953548477341497e-06,
      "loss": 0.6364,
      "step": 17
    },
    {
      "epoch": 0.09082308420056764,
      "grad_norm": 0.04319237430058616,
      "learning_rate": 2.9947927109590477e-06,
      "loss": 0.568,
      "step": 18
    },
    {
      "epoch": 0.09586881110059918,
      "grad_norm": 0.042093297202993624,
      "learning_rate": 2.994198551870913e-06,
      "loss": 0.6184,
      "step": 19
    },
    {
      "epoch": 0.10091453800063072,
      "grad_norm": 0.04087623899598573,
      "learning_rate": 2.993572387089653e-06,
      "loss": 0.5822,
      "step": 20
    },
    {
      "epoch": 0.10596026490066225,
      "grad_norm": 0.042619877493329586,
      "learning_rate": 2.992914234130442e-06,
      "loss": 0.5983,
      "step": 21
    },
    {
      "epoch": 0.11100599180069379,
      "grad_norm": 0.04314774114784986,
      "learning_rate": 2.9922241114032345e-06,
      "loss": 0.6058,
      "step": 22
    },
    {
      "epoch": 0.11605171870072532,
      "grad_norm": 0.04125496902035363,
      "learning_rate": 2.9915020382122458e-06,
      "loss": 0.5741,
      "step": 23
    },
    {
      "epoch": 0.12109744560075686,
      "grad_norm": 0.03985368427853683,
      "learning_rate": 2.990748034755415e-06,
      "loss": 0.6002,
      "step": 24
    },
    {
      "epoch": 0.1261431725007884,
      "grad_norm": 0.04603566805698703,
      "learning_rate": 2.9899621221238394e-06,
      "loss": 0.5616,
      "step": 25
    },
    {
      "epoch": 0.13118889940081993,
      "grad_norm": 0.033944581121186666,
      "learning_rate": 2.989144322301186e-06,
      "loss": 0.591,
      "step": 26
    },
    {
      "epoch": 0.13623462630085148,
      "grad_norm": 0.0352127486146018,
      "learning_rate": 2.988294658163073e-06,
      "loss": 0.575,
      "step": 27
    },
    {
      "epoch": 0.141280353200883,
      "grad_norm": 0.04026082684896548,
      "learning_rate": 2.9874131534764325e-06,
      "loss": 0.5783,
      "step": 28
    },
    {
      "epoch": 0.14632608010091455,
      "grad_norm": 0.038584952671910096,
      "learning_rate": 2.9864998328988463e-06,
      "loss": 0.5814,
      "step": 29
    },
    {
      "epoch": 0.15137180700094607,
      "grad_norm": 0.03294755370363045,
      "learning_rate": 2.985554721977853e-06,
      "loss": 0.5688,
      "step": 30
    },
    {
      "epoch": 0.15641753390097762,
      "grad_norm": 0.035774614388450525,
      "learning_rate": 2.984577847150239e-06,
      "loss": 0.5914,
      "step": 31
    },
    {
      "epoch": 0.16146326080100915,
      "grad_norm": 0.04512017281393784,
      "learning_rate": 2.983569235741291e-06,
      "loss": 0.557,
      "step": 32
    },
    {
      "epoch": 0.1665089877010407,
      "grad_norm": 0.03447545680264101,
      "learning_rate": 2.9825289159640397e-06,
      "loss": 0.568,
      "step": 33
    },
    {
      "epoch": 0.17155471460107222,
      "grad_norm": 0.033658505681229516,
      "learning_rate": 2.9814569169184642e-06,
      "loss": 0.5868,
      "step": 34
    },
    {
      "epoch": 0.17660044150110377,
      "grad_norm": 0.03071546221735757,
      "learning_rate": 2.980353268590683e-06,
      "loss": 0.5487,
      "step": 35
    },
    {
      "epoch": 0.1816461684011353,
      "grad_norm": 0.07417860742940319,
      "learning_rate": 2.9792180018521128e-06,
      "loss": 0.6099,
      "step": 36
    },
    {
      "epoch": 0.18669189530116684,
      "grad_norm": 0.032180325133544276,
      "learning_rate": 2.978051148458604e-06,
      "loss": 0.5939,
      "step": 37
    },
    {
      "epoch": 0.19173762220119836,
      "grad_norm": 0.031347752245340116,
      "learning_rate": 2.976852741049554e-06,
      "loss": 0.5764,
      "step": 38
    },
    {
      "epoch": 0.1967833491012299,
      "grad_norm": 0.035873383222778825,
      "learning_rate": 2.975622813146996e-06,
      "loss": 0.57,
      "step": 39
    },
    {
      "epoch": 0.20182907600126143,
      "grad_norm": 0.03130302787258777,
      "learning_rate": 2.9743613991546548e-06,
      "loss": 0.5503,
      "step": 40
    },
    {
      "epoch": 0.20687480290129298,
      "grad_norm": 0.04111552220221803,
      "learning_rate": 2.9730685343569934e-06,
      "loss": 0.6028,
      "step": 41
    },
    {
      "epoch": 0.2119205298013245,
      "grad_norm": 0.031561335436647305,
      "learning_rate": 2.971744254918218e-06,
      "loss": 0.5682,
      "step": 42
    },
    {
      "epoch": 0.21696625670135605,
      "grad_norm": 0.03466870924962832,
      "learning_rate": 2.9703885978812726e-06,
      "loss": 0.55,
      "step": 43
    },
    {
      "epoch": 0.22201198360138757,
      "grad_norm": 0.03396258277418921,
      "learning_rate": 2.9690016011667974e-06,
      "loss": 0.5953,
      "step": 44
    },
    {
      "epoch": 0.22705771050141912,
      "grad_norm": 0.033463552224919146,
      "learning_rate": 2.967583303572073e-06,
      "loss": 0.6231,
      "step": 45
    },
    {
      "epoch": 0.23210343740145065,
      "grad_norm": 0.03747113039368738,
      "learning_rate": 2.9661337447699316e-06,
      "loss": 0.5742,
      "step": 46
    },
    {
      "epoch": 0.2371491643014822,
      "grad_norm": 0.04159182229285405,
      "learning_rate": 2.9646529653076493e-06,
      "loss": 0.5681,
      "step": 47
    },
    {
      "epoch": 0.24219489120151372,
      "grad_norm": 0.032311171301265075,
      "learning_rate": 2.9631410066058098e-06,
      "loss": 0.5464,
      "step": 48
    },
    {
      "epoch": 0.24724061810154527,
      "grad_norm": 0.035494911254562625,
      "learning_rate": 2.9615979109571493e-06,
      "loss": 0.5377,
      "step": 49
    },
    {
      "epoch": 0.2522863450015768,
      "grad_norm": 0.032401750473671755,
      "learning_rate": 2.9600237215253696e-06,
      "loss": 0.6043,
      "step": 50
    },
    {
      "epoch": 0.25733207190160834,
      "grad_norm": 0.03477400444401883,
      "learning_rate": 2.9584184823439337e-06,
      "loss": 0.6078,
      "step": 51
    },
    {
      "epoch": 0.26237779880163986,
      "grad_norm": 0.03586539534553979,
      "learning_rate": 2.9567822383148315e-06,
      "loss": 0.5857,
      "step": 52
    },
    {
      "epoch": 0.2674235257016714,
      "grad_norm": 0.034776366845092124,
      "learning_rate": 2.955115035207326e-06,
      "loss": 0.5652,
      "step": 53
    },
    {
      "epoch": 0.27246925260170296,
      "grad_norm": 0.047916672806890825,
      "learning_rate": 2.953416919656672e-06,
      "loss": 0.529,
      "step": 54
    },
    {
      "epoch": 0.2775149795017345,
      "grad_norm": 0.035512253032401846,
      "learning_rate": 2.9516879391628125e-06,
      "loss": 0.6018,
      "step": 55
    },
    {
      "epoch": 0.282560706401766,
      "grad_norm": 0.0669654595534551,
      "learning_rate": 2.9499281420890474e-06,
      "loss": 0.5832,
      "step": 56
    },
    {
      "epoch": 0.2876064333017975,
      "grad_norm": 0.04009377576904152,
      "learning_rate": 2.948137577660685e-06,
      "loss": 0.5376,
      "step": 57
    },
    {
      "epoch": 0.2926521602018291,
      "grad_norm": 0.05517678453435375,
      "learning_rate": 2.946316295963661e-06,
      "loss": 0.5725,
      "step": 58
    },
    {
      "epoch": 0.2976978871018606,
      "grad_norm": 0.040682905696082294,
      "learning_rate": 2.9444643479431393e-06,
      "loss": 0.5887,
      "step": 59
    },
    {
      "epoch": 0.30274361400189215,
      "grad_norm": 0.044774454426992336,
      "learning_rate": 2.9425817854020873e-06,
      "loss": 0.5756,
      "step": 60
    },
    {
      "epoch": 0.30778934090192367,
      "grad_norm": 0.03263011910584542,
      "learning_rate": 2.940668660999826e-06,
      "loss": 0.5693,
      "step": 61
    },
    {
      "epoch": 0.31283506780195525,
      "grad_norm": 0.032496230797448664,
      "learning_rate": 2.9387250282505583e-06,
      "loss": 0.586,
      "step": 62
    },
    {
      "epoch": 0.31788079470198677,
      "grad_norm": 0.03310861754959189,
      "learning_rate": 2.9367509415218687e-06,
      "loss": 0.5548,
      "step": 63
    },
    {
      "epoch": 0.3229265216020183,
      "grad_norm": 0.031816229512850104,
      "learning_rate": 2.9347464560332084e-06,
      "loss": 0.6,
      "step": 64
    },
    {
      "epoch": 0.3279722485020498,
      "grad_norm": 0.036465675122600016,
      "learning_rate": 2.932711627854344e-06,
      "loss": 0.5613,
      "step": 65
    },
    {
      "epoch": 0.3330179754020814,
      "grad_norm": 0.03107217546123426,
      "learning_rate": 2.9306465139037947e-06,
      "loss": 0.5421,
      "step": 66
    },
    {
      "epoch": 0.3380637023021129,
      "grad_norm": 0.031633841290591734,
      "learning_rate": 2.9285511719472367e-06,
      "loss": 0.58,
      "step": 67
    },
    {
      "epoch": 0.34310942920214443,
      "grad_norm": 0.030853855883844275,
      "learning_rate": 2.9264256605958885e-06,
      "loss": 0.5496,
      "step": 68
    },
    {
      "epoch": 0.34815515610217596,
      "grad_norm": 0.036265638918391636,
      "learning_rate": 2.924270039304873e-06,
      "loss": 0.5939,
      "step": 69
    },
    {
      "epoch": 0.35320088300220753,
      "grad_norm": 0.03703293289195566,
      "learning_rate": 2.9220843683715497e-06,
      "loss": 0.5311,
      "step": 70
    },
    {
      "epoch": 0.35824660990223905,
      "grad_norm": 0.07682301940151573,
      "learning_rate": 2.9198687089338345e-06,
      "loss": 0.5655,
      "step": 71
    },
    {
      "epoch": 0.3632923368022706,
      "grad_norm": 0.03240731857642153,
      "learning_rate": 2.9176231229684835e-06,
      "loss": 0.5436,
      "step": 72
    },
    {
      "epoch": 0.3683380637023021,
      "grad_norm": 0.03550209155305971,
      "learning_rate": 2.9153476732893646e-06,
      "loss": 0.5529,
      "step": 73
    },
    {
      "epoch": 0.3733837906023337,
      "grad_norm": 0.03572110732287988,
      "learning_rate": 2.913042423545696e-06,
      "loss": 0.5601,
      "step": 74
    },
    {
      "epoch": 0.3784295175023652,
      "grad_norm": 0.030318267187340705,
      "learning_rate": 2.910707438220269e-06,
      "loss": 0.5827,
      "step": 75
    },
    {
      "epoch": 0.3834752444023967,
      "grad_norm": 0.030779462298936085,
      "learning_rate": 2.9083427826276414e-06,
      "loss": 0.5366,
      "step": 76
    },
    {
      "epoch": 0.38852097130242824,
      "grad_norm": 0.033078267956613755,
      "learning_rate": 2.905948522912315e-06,
      "loss": 0.5769,
      "step": 77
    },
    {
      "epoch": 0.3935666982024598,
      "grad_norm": 0.032022182515529865,
      "learning_rate": 2.90352472604688e-06,
      "loss": 0.6059,
      "step": 78
    },
    {
      "epoch": 0.39861242510249134,
      "grad_norm": 0.032572741826790486,
      "learning_rate": 2.901071459830145e-06,
      "loss": 0.5325,
      "step": 79
    },
    {
      "epoch": 0.40365815200252286,
      "grad_norm": 0.03322477222052267,
      "learning_rate": 2.89858879288524e-06,
      "loss": 0.6102,
      "step": 80
    },
    {
      "epoch": 0.4087038789025544,
      "grad_norm": 0.03290381540817977,
      "learning_rate": 2.896076794657696e-06,
      "loss": 0.5297,
      "step": 81
    },
    {
      "epoch": 0.41374960580258596,
      "grad_norm": 0.02986308712893659,
      "learning_rate": 2.893535535413504e-06,
      "loss": 0.6016,
      "step": 82
    },
    {
      "epoch": 0.4187953327026175,
      "grad_norm": 0.03708195442407903,
      "learning_rate": 2.8909650862371465e-06,
      "loss": 0.5644,
      "step": 83
    },
    {
      "epoch": 0.423841059602649,
      "grad_norm": 0.05585756602200335,
      "learning_rate": 2.888365519029615e-06,
      "loss": 0.5645,
      "step": 84
    },
    {
      "epoch": 0.42888678650268053,
      "grad_norm": 0.03232081336561299,
      "learning_rate": 2.8857369065063893e-06,
      "loss": 0.5492,
      "step": 85
    },
    {
      "epoch": 0.4339325134027121,
      "grad_norm": 0.03807439954613977,
      "learning_rate": 2.883079322195415e-06,
      "loss": 0.5694,
      "step": 86
    },
    {
      "epoch": 0.4389782403027436,
      "grad_norm": 0.03718669059491054,
      "learning_rate": 2.880392840435036e-06,
      "loss": 0.5603,
      "step": 87
    },
    {
      "epoch": 0.44402396720277515,
      "grad_norm": 0.02993361885707706,
      "learning_rate": 2.8776775363719244e-06,
      "loss": 0.5193,
      "step": 88
    },
    {
      "epoch": 0.44906969410280667,
      "grad_norm": 0.03471859873605538,
      "learning_rate": 2.8749334859589696e-06,
      "loss": 0.5195,
      "step": 89
    },
    {
      "epoch": 0.45411542100283825,
      "grad_norm": 0.03468783264087511,
      "learning_rate": 2.872160765953162e-06,
      "loss": 0.5685,
      "step": 90
    },
    {
      "epoch": 0.45916114790286977,
      "grad_norm": 0.06028818099354049,
      "learning_rate": 2.86935945391344e-06,
      "loss": 0.5875,
      "step": 91
    },
    {
      "epoch": 0.4642068748029013,
      "grad_norm": 0.03189592063110031,
      "learning_rate": 2.8665296281985232e-06,
      "loss": 0.5627,
      "step": 92
    },
    {
      "epoch": 0.4692526017029328,
      "grad_norm": 0.032489964455779306,
      "learning_rate": 2.8636713679647195e-06,
      "loss": 0.5398,
      "step": 93
    },
    {
      "epoch": 0.4742983286029644,
      "grad_norm": 0.03300509062363307,
      "learning_rate": 2.8607847531637127e-06,
      "loss": 0.5675,
      "step": 94
    },
    {
      "epoch": 0.4793440555029959,
      "grad_norm": 0.034805575232998785,
      "learning_rate": 2.857869864540323e-06,
      "loss": 0.5526,
      "step": 95
    },
    {
      "epoch": 0.48438978240302744,
      "grad_norm": 0.03304213276713402,
      "learning_rate": 2.854926783630253e-06,
      "loss": 0.5475,
      "step": 96
    },
    {
      "epoch": 0.48943550930305896,
      "grad_norm": 0.03753659611183512,
      "learning_rate": 2.851955592757801e-06,
      "loss": 0.5511,
      "step": 97
    },
    {
      "epoch": 0.49448123620309054,
      "grad_norm": 0.033892234979303396,
      "learning_rate": 2.848956375033562e-06,
      "loss": 0.5232,
      "step": 98
    },
    {
      "epoch": 0.49952696310312206,
      "grad_norm": 0.037074509268233,
      "learning_rate": 2.845929214352105e-06,
      "loss": 0.5655,
      "step": 99
    }
  ],
  "logging_steps": 1,
  "max_steps": 594,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 99,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 410485593735168.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}