{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.986175115207373,
  "eval_steps": 500,
  "global_step": 216,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.013824884792626729,
      "grad_norm": 7.633559049394886,
      "learning_rate": 4.5454545454545457e-07,
      "loss": 0.8764,
      "step": 1
    },
    {
      "epoch": 0.027649769585253458,
      "grad_norm": 7.6469224678917,
      "learning_rate": 9.090909090909091e-07,
      "loss": 0.8712,
      "step": 2
    },
    {
      "epoch": 0.041474654377880185,
      "grad_norm": 7.5089460233893295,
      "learning_rate": 1.3636363636363636e-06,
      "loss": 0.8804,
      "step": 3
    },
    {
      "epoch": 0.055299539170506916,
      "grad_norm": 7.691760047071555,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 0.8775,
      "step": 4
    },
    {
      "epoch": 0.06912442396313365,
      "grad_norm": 7.294963661060089,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.8709,
      "step": 5
    },
    {
      "epoch": 0.08294930875576037,
      "grad_norm": 5.443182880842741,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 0.7907,
      "step": 6
    },
    {
      "epoch": 0.0967741935483871,
      "grad_norm": 5.047617387545844,
      "learning_rate": 3.181818181818182e-06,
      "loss": 0.7836,
      "step": 7
    },
    {
      "epoch": 0.11059907834101383,
      "grad_norm": 3.2253892674265594,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.7531,
      "step": 8
    },
    {
      "epoch": 0.12442396313364056,
      "grad_norm": 2.8131888920599657,
      "learning_rate": 4.0909090909090915e-06,
      "loss": 0.7417,
      "step": 9
    },
    {
      "epoch": 0.1382488479262673,
      "grad_norm": 3.6977429779435815,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.723,
      "step": 10
    },
    {
      "epoch": 0.15207373271889402,
      "grad_norm": 4.281498675575116,
      "learning_rate": 5e-06,
      "loss": 0.7246,
      "step": 11
    },
    {
      "epoch": 0.16589861751152074,
      "grad_norm": 3.6996387465448373,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 0.6595,
      "step": 12
    },
    {
      "epoch": 0.17972350230414746,
      "grad_norm": 3.78830344374056,
      "learning_rate": 5.90909090909091e-06,
      "loss": 0.6694,
      "step": 13
    },
    {
      "epoch": 0.1935483870967742,
      "grad_norm": 3.3490463351493713,
      "learning_rate": 6.363636363636364e-06,
      "loss": 0.6227,
      "step": 14
    },
    {
      "epoch": 0.2073732718894009,
      "grad_norm": 3.374405746489098,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.6273,
      "step": 15
    },
    {
      "epoch": 0.22119815668202766,
      "grad_norm": 2.4954014740266195,
      "learning_rate": 7.272727272727273e-06,
      "loss": 0.6062,
      "step": 16
    },
    {
      "epoch": 0.2350230414746544,
      "grad_norm": 1.8239008983436202,
      "learning_rate": 7.727272727272727e-06,
      "loss": 0.593,
      "step": 17
    },
    {
      "epoch": 0.2488479262672811,
      "grad_norm": 1.635875724467323,
      "learning_rate": 8.181818181818183e-06,
      "loss": 0.5959,
      "step": 18
    },
    {
      "epoch": 0.2626728110599078,
      "grad_norm": 2.0334716933077166,
      "learning_rate": 8.636363636363637e-06,
      "loss": 0.5672,
      "step": 19
    },
    {
      "epoch": 0.2764976958525346,
      "grad_norm": 1.9168703834196266,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.5729,
      "step": 20
    },
    {
      "epoch": 0.2903225806451613,
      "grad_norm": 1.4466966621915673,
      "learning_rate": 9.545454545454547e-06,
      "loss": 0.5534,
      "step": 21
    },
    {
      "epoch": 0.30414746543778803,
      "grad_norm": 1.0267896741626927,
      "learning_rate": 1e-05,
      "loss": 0.5628,
      "step": 22
    },
    {
      "epoch": 0.31797235023041476,
      "grad_norm": 1.262616482011545,
      "learning_rate": 9.999344418328161e-06,
      "loss": 0.5572,
      "step": 23
    },
    {
      "epoch": 0.3317972350230415,
      "grad_norm": 1.0906634918414906,
      "learning_rate": 9.997377845227577e-06,
      "loss": 0.5146,
      "step": 24
    },
    {
      "epoch": 0.3456221198156682,
      "grad_norm": 1.0415845965231307,
      "learning_rate": 9.994100796397954e-06,
      "loss": 0.4973,
      "step": 25
    },
    {
      "epoch": 0.35944700460829493,
      "grad_norm": 0.815227789570975,
      "learning_rate": 9.98951413118856e-06,
      "loss": 0.5218,
      "step": 26
    },
    {
      "epoch": 0.37327188940092165,
      "grad_norm": 0.8903243919241076,
      "learning_rate": 9.983619052372847e-06,
      "loss": 0.5167,
      "step": 27
    },
    {
      "epoch": 0.3870967741935484,
      "grad_norm": 1.0991055385660755,
      "learning_rate": 9.97641710583307e-06,
      "loss": 0.5003,
      "step": 28
    },
    {
      "epoch": 0.4009216589861751,
      "grad_norm": 0.9348867416996314,
      "learning_rate": 9.96791018015489e-06,
      "loss": 0.5078,
      "step": 29
    },
    {
      "epoch": 0.4147465437788018,
      "grad_norm": 0.7138578554042776,
      "learning_rate": 9.958100506132127e-06,
      "loss": 0.5075,
      "step": 30
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.8349505675026924,
      "learning_rate": 9.946990656181782e-06,
      "loss": 0.511,
      "step": 31
    },
    {
      "epoch": 0.4423963133640553,
      "grad_norm": 1.052621697155694,
      "learning_rate": 9.934583543669454e-06,
      "loss": 0.4824,
      "step": 32
    },
    {
      "epoch": 0.45622119815668205,
      "grad_norm": 0.7618516629853732,
      "learning_rate": 9.920882422145372e-06,
      "loss": 0.4842,
      "step": 33
    },
    {
      "epoch": 0.4700460829493088,
      "grad_norm": 0.8523130797451687,
      "learning_rate": 9.905890884491196e-06,
      "loss": 0.4843,
      "step": 34
    },
    {
      "epoch": 0.4838709677419355,
      "grad_norm": 0.7854821313979271,
      "learning_rate": 9.889612861977855e-06,
      "loss": 0.4893,
      "step": 35
    },
    {
      "epoch": 0.4976958525345622,
      "grad_norm": 0.9098881831155701,
      "learning_rate": 9.872052623234632e-06,
      "loss": 0.5024,
      "step": 36
    },
    {
      "epoch": 0.511520737327189,
      "grad_norm": 0.9133145917535354,
      "learning_rate": 9.853214773129796e-06,
      "loss": 0.4828,
      "step": 37
    },
    {
      "epoch": 0.5253456221198156,
      "grad_norm": 0.7771673181178608,
      "learning_rate": 9.833104251563058e-06,
      "loss": 0.478,
      "step": 38
    },
    {
      "epoch": 0.5391705069124424,
      "grad_norm": 0.6686414859648492,
      "learning_rate": 9.811726332170153e-06,
      "loss": 0.4945,
      "step": 39
    },
    {
      "epoch": 0.5529953917050692,
      "grad_norm": 0.7697756564970056,
      "learning_rate": 9.789086620939936e-06,
      "loss": 0.4704,
      "step": 40
    },
    {
      "epoch": 0.5668202764976958,
      "grad_norm": 0.8065645041682494,
      "learning_rate": 9.765191054744305e-06,
      "loss": 0.4512,
      "step": 41
    },
    {
      "epoch": 0.5806451612903226,
      "grad_norm": 0.7428419868914332,
      "learning_rate": 9.740045899781353e-06,
      "loss": 0.4631,
      "step": 42
    },
    {
      "epoch": 0.5944700460829493,
      "grad_norm": 0.7370318617715542,
      "learning_rate": 9.713657749932172e-06,
      "loss": 0.4834,
      "step": 43
    },
    {
      "epoch": 0.6082949308755761,
      "grad_norm": 0.5647708496508699,
      "learning_rate": 9.68603352503172e-06,
      "loss": 0.4774,
      "step": 44
    },
    {
      "epoch": 0.6221198156682027,
      "grad_norm": 0.6790318592831586,
      "learning_rate": 9.657180469054213e-06,
      "loss": 0.4964,
      "step": 45
    },
    {
      "epoch": 0.6359447004608295,
      "grad_norm": 0.6728496708051098,
      "learning_rate": 9.627106148213521e-06,
      "loss": 0.5024,
      "step": 46
    },
    {
      "epoch": 0.6497695852534562,
      "grad_norm": 0.6158261913286127,
      "learning_rate": 9.595818448979061e-06,
      "loss": 0.4656,
      "step": 47
    },
    {
      "epoch": 0.663594470046083,
      "grad_norm": 0.6396258914140209,
      "learning_rate": 9.563325576007702e-06,
      "loss": 0.4593,
      "step": 48
    },
    {
      "epoch": 0.6774193548387096,
      "grad_norm": 0.55574623840064,
      "learning_rate": 9.529636049992235e-06,
      "loss": 0.4848,
      "step": 49
    },
    {
      "epoch": 0.6912442396313364,
      "grad_norm": 0.5960107153575481,
      "learning_rate": 9.494758705426978e-06,
      "loss": 0.4606,
      "step": 50
    },
    {
      "epoch": 0.7050691244239631,
      "grad_norm": 0.6187132785014103,
      "learning_rate": 9.458702688291072e-06,
      "loss": 0.4706,
      "step": 51
    },
    {
      "epoch": 0.7188940092165899,
      "grad_norm": 0.596391894724299,
      "learning_rate": 9.421477453650118e-06,
      "loss": 0.4552,
      "step": 52
    },
    {
      "epoch": 0.7327188940092166,
      "grad_norm": 0.5828525328792538,
      "learning_rate": 9.38309276317674e-06,
      "loss": 0.4419,
      "step": 53
    },
    {
      "epoch": 0.7465437788018433,
      "grad_norm": 0.5574488192167764,
      "learning_rate": 9.343558682590757e-06,
      "loss": 0.4482,
      "step": 54
    },
    {
      "epoch": 0.7603686635944701,
      "grad_norm": 0.5810226590826171,
      "learning_rate": 9.302885579019626e-06,
      "loss": 0.4574,
      "step": 55
    },
    {
      "epoch": 0.7741935483870968,
      "grad_norm": 0.6373464198147619,
      "learning_rate": 9.261084118279846e-06,
      "loss": 0.5021,
      "step": 56
    },
    {
      "epoch": 0.7880184331797235,
      "grad_norm": 0.6473035563889259,
      "learning_rate": 9.218165262080024e-06,
      "loss": 0.4688,
      "step": 57
    },
    {
      "epoch": 0.8018433179723502,
      "grad_norm": 0.5496475133999553,
      "learning_rate": 9.174140265146356e-06,
      "loss": 0.4614,
      "step": 58
    },
    {
      "epoch": 0.815668202764977,
      "grad_norm": 0.5633672236656848,
      "learning_rate": 9.129020672271283e-06,
      "loss": 0.4631,
      "step": 59
    },
    {
      "epoch": 0.8294930875576036,
      "grad_norm": 0.5996514489934792,
      "learning_rate": 9.082818315286054e-06,
      "loss": 0.4709,
      "step": 60
    },
    {
      "epoch": 0.8433179723502304,
      "grad_norm": 0.5800489851592833,
      "learning_rate": 9.035545309958048e-06,
      "loss": 0.4529,
      "step": 61
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.5728565615351726,
      "learning_rate": 8.987214052813605e-06,
      "loss": 0.4689,
      "step": 62
    },
    {
      "epoch": 0.8709677419354839,
      "grad_norm": 0.635495629930791,
      "learning_rate": 8.937837217887273e-06,
      "loss": 0.4577,
      "step": 63
    },
    {
      "epoch": 0.8847926267281107,
      "grad_norm": 0.6832859023985155,
      "learning_rate": 8.887427753398249e-06,
      "loss": 0.4551,
      "step": 64
    },
    {
      "epoch": 0.8986175115207373,
      "grad_norm": 0.6250704218728608,
      "learning_rate": 8.83599887835493e-06,
      "loss": 0.4412,
      "step": 65
    },
    {
      "epoch": 0.9124423963133641,
      "grad_norm": 0.4941669349016631,
      "learning_rate": 8.783564079088478e-06,
      "loss": 0.452,
      "step": 66
    },
    {
      "epoch": 0.9262672811059908,
      "grad_norm": 0.6418218067831372,
      "learning_rate": 8.730137105716231e-06,
      "loss": 0.4374,
      "step": 67
    },
    {
      "epoch": 0.9400921658986175,
      "grad_norm": 0.4944995367016919,
      "learning_rate": 8.675731968536004e-06,
      "loss": 0.4538,
      "step": 68
    },
    {
      "epoch": 0.9539170506912442,
      "grad_norm": 0.5351112267428982,
      "learning_rate": 8.620362934352109e-06,
      "loss": 0.456,
      "step": 69
    },
    {
      "epoch": 0.967741935483871,
      "grad_norm": 0.550190483341697,
      "learning_rate": 8.564044522734147e-06,
      "loss": 0.4696,
      "step": 70
    },
    {
      "epoch": 0.9815668202764977,
      "grad_norm": 0.5416241347883733,
      "learning_rate": 8.506791502209497e-06,
      "loss": 0.4351,
      "step": 71
    },
    {
      "epoch": 0.9953917050691244,
      "grad_norm": 0.5646211850916225,
      "learning_rate": 8.448618886390523e-06,
      "loss": 0.4395,
      "step": 72
    },
    {
      "epoch": 1.0092165898617511,
      "grad_norm": 1.0497545314797017,
      "learning_rate": 8.389541930037516e-06,
      "loss": 0.7044,
      "step": 73
    },
    {
      "epoch": 1.023041474654378,
      "grad_norm": 0.6050451609110485,
      "learning_rate": 8.329576125058406e-06,
      "loss": 0.4019,
      "step": 74
    },
    {
      "epoch": 1.0368663594470047,
      "grad_norm": 0.668170300230099,
      "learning_rate": 8.268737196446264e-06,
      "loss": 0.4301,
      "step": 75
    },
    {
      "epoch": 1.0506912442396312,
      "grad_norm": 0.5614144997152238,
      "learning_rate": 8.207041098155701e-06,
      "loss": 0.3802,
      "step": 76
    },
    {
      "epoch": 1.064516129032258,
      "grad_norm": 0.6517654191953809,
      "learning_rate": 8.144504008919224e-06,
      "loss": 0.4079,
      "step": 77
    },
    {
      "epoch": 1.0783410138248848,
      "grad_norm": 0.5437003169084378,
      "learning_rate": 8.081142328004638e-06,
      "loss": 0.3725,
      "step": 78
    },
    {
      "epoch": 1.0921658986175116,
      "grad_norm": 0.6545122792430208,
      "learning_rate": 8.016972670914624e-06,
      "loss": 0.4282,
      "step": 79
    },
    {
      "epoch": 1.1059907834101383,
      "grad_norm": 0.6048618596815725,
      "learning_rate": 7.952011865029614e-06,
      "loss": 0.4037,
      "step": 80
    },
    {
      "epoch": 1.119815668202765,
      "grad_norm": 0.6691394124559549,
      "learning_rate": 7.886276945195098e-06,
      "loss": 0.3643,
      "step": 81
    },
    {
      "epoch": 1.1336405529953917,
      "grad_norm": 0.7270083026506138,
      "learning_rate": 7.819785149254534e-06,
      "loss": 0.4131,
      "step": 82
    },
    {
      "epoch": 1.1474654377880185,
      "grad_norm": 0.5582455458120927,
      "learning_rate": 7.752553913529019e-06,
      "loss": 0.368,
      "step": 83
    },
    {
      "epoch": 1.1612903225806452,
      "grad_norm": 0.6782237172690554,
      "learning_rate": 7.68460086824492e-06,
      "loss": 0.4308,
      "step": 84
    },
    {
      "epoch": 1.1751152073732718,
      "grad_norm": 0.6173590398850847,
      "learning_rate": 7.61594383291065e-06,
      "loss": 0.3943,
      "step": 85
    },
    {
      "epoch": 1.1889400921658986,
      "grad_norm": 0.5909172461024365,
      "learning_rate": 7.546600811643816e-06,
      "loss": 0.4211,
      "step": 86
    },
    {
      "epoch": 1.2027649769585254,
      "grad_norm": 0.5670563053696445,
      "learning_rate": 7.476589988449939e-06,
      "loss": 0.3704,
      "step": 87
    },
    {
      "epoch": 1.2165898617511521,
      "grad_norm": 0.5698865538843494,
      "learning_rate": 7.405929722454026e-06,
      "loss": 0.3912,
      "step": 88
    },
    {
      "epoch": 1.230414746543779,
      "grad_norm": 0.5331376655462895,
      "learning_rate": 7.334638543086203e-06,
      "loss": 0.3836,
      "step": 89
    },
    {
      "epoch": 1.2442396313364055,
      "grad_norm": 0.6325980757371303,
      "learning_rate": 7.262735145222696e-06,
      "loss": 0.3915,
      "step": 90
    },
    {
      "epoch": 1.2580645161290323,
      "grad_norm": 0.5952012132281536,
      "learning_rate": 7.190238384283413e-06,
      "loss": 0.4005,
      "step": 91
    },
    {
      "epoch": 1.271889400921659,
      "grad_norm": 0.5581313424613799,
      "learning_rate": 7.117167271287453e-06,
      "loss": 0.3991,
      "step": 92
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 0.5043175284762973,
      "learning_rate": 7.043540967867782e-06,
      "loss": 0.3643,
      "step": 93
    },
    {
      "epoch": 1.2995391705069124,
      "grad_norm": 0.5598625221287372,
      "learning_rate": 6.969378781246436e-06,
      "loss": 0.4334,
      "step": 94
    },
    {
      "epoch": 1.3133640552995391,
      "grad_norm": 0.5503414372922475,
      "learning_rate": 6.894700159171535e-06,
      "loss": 0.415,
      "step": 95
    },
    {
      "epoch": 1.327188940092166,
      "grad_norm": 0.5592853069996093,
      "learning_rate": 6.819524684817439e-06,
      "loss": 0.407,
      "step": 96
    },
    {
      "epoch": 1.3410138248847927,
      "grad_norm": 0.5431739142566777,
      "learning_rate": 6.743872071649411e-06,
      "loss": 0.401,
      "step": 97
    },
    {
      "epoch": 1.3548387096774195,
      "grad_norm": 0.5168119415119301,
      "learning_rate": 6.667762158254104e-06,
      "loss": 0.3713,
      "step": 98
    },
    {
      "epoch": 1.368663594470046,
      "grad_norm": 0.6207479541491344,
      "learning_rate": 6.591214903137221e-06,
      "loss": 0.389,
      "step": 99
    },
    {
      "epoch": 1.3824884792626728,
      "grad_norm": 0.5364662180025893,
      "learning_rate": 6.514250379489754e-06,
      "loss": 0.3954,
      "step": 100
    },
    {
      "epoch": 1.3963133640552996,
      "grad_norm": 0.5066869500264701,
      "learning_rate": 6.436888769924142e-06,
      "loss": 0.3939,
      "step": 101
    },
    {
      "epoch": 1.4101382488479262,
      "grad_norm": 0.5780367268696035,
      "learning_rate": 6.3591503611817155e-06,
      "loss": 0.4076,
      "step": 102
    },
    {
      "epoch": 1.423963133640553,
      "grad_norm": 0.5070490336891309,
      "learning_rate": 6.281055538812861e-06,
      "loss": 0.3604,
      "step": 103
    },
    {
      "epoch": 1.4377880184331797,
      "grad_norm": 0.612079276266504,
      "learning_rate": 6.202624781831269e-06,
      "loss": 0.4067,
      "step": 104
    },
    {
      "epoch": 1.4516129032258065,
      "grad_norm": 0.5734702647572101,
      "learning_rate": 6.123878657343648e-06,
      "loss": 0.4255,
      "step": 105
    },
    {
      "epoch": 1.4654377880184333,
      "grad_norm": 0.5354043563850319,
      "learning_rate": 6.044837815156377e-06,
      "loss": 0.4074,
      "step": 106
    },
    {
      "epoch": 1.4792626728110598,
      "grad_norm": 0.5791707978034125,
      "learning_rate": 5.965522982360441e-06,
      "loss": 0.3895,
      "step": 107
    },
    {
      "epoch": 1.4930875576036866,
      "grad_norm": 0.5703922296796244,
      "learning_rate": 5.885954957896115e-06,
      "loss": 0.4086,
      "step": 108
    },
    {
      "epoch": 1.5069124423963134,
      "grad_norm": 0.5065784500919478,
      "learning_rate": 5.806154607098799e-06,
      "loss": 0.3693,
      "step": 109
    },
    {
      "epoch": 1.52073732718894,
      "grad_norm": 0.556653840063644,
      "learning_rate": 5.726142856227453e-06,
      "loss": 0.4299,
      "step": 110
    },
    {
      "epoch": 1.5345622119815667,
      "grad_norm": 0.5474174976443692,
      "learning_rate": 5.645940686977033e-06,
      "loss": 0.4148,
      "step": 111
    },
    {
      "epoch": 1.5483870967741935,
      "grad_norm": 0.4806025731120033,
      "learning_rate": 5.5655691309764225e-06,
      "loss": 0.3924,
      "step": 112
    },
    {
      "epoch": 1.5622119815668203,
      "grad_norm": 0.5324709526786499,
      "learning_rate": 5.485049264273241e-06,
      "loss": 0.3789,
      "step": 113
    },
    {
      "epoch": 1.576036866359447,
      "grad_norm": 0.4659634896990746,
      "learning_rate": 5.404402201807022e-06,
      "loss": 0.3609,
      "step": 114
    },
    {
      "epoch": 1.5898617511520738,
      "grad_norm": 0.5808920912192916,
      "learning_rate": 5.323649091872179e-06,
      "loss": 0.4271,
      "step": 115
    },
    {
      "epoch": 1.6036866359447006,
      "grad_norm": 0.5993918356229727,
      "learning_rate": 5.242811110572243e-06,
      "loss": 0.3989,
      "step": 116
    },
    {
      "epoch": 1.6175115207373272,
      "grad_norm": 0.5109808802516896,
      "learning_rate": 5.161909456266781e-06,
      "loss": 0.3814,
      "step": 117
    },
    {
      "epoch": 1.631336405529954,
      "grad_norm": 0.5423585025364462,
      "learning_rate": 5.080965344012509e-06,
      "loss": 0.3985,
      "step": 118
    },
    {
      "epoch": 1.6451612903225805,
      "grad_norm": 0.5004581821839913,
      "learning_rate": 5e-06,
      "loss": 0.3523,
      "step": 119
    },
    {
      "epoch": 1.6589861751152073,
      "grad_norm": 0.555371258631358,
      "learning_rate": 4.919034655987493e-06,
      "loss": 0.3977,
      "step": 120
    },
    {
      "epoch": 1.672811059907834,
      "grad_norm": 0.6223674774033777,
      "learning_rate": 4.838090543733222e-06,
      "loss": 0.4122,
      "step": 121
    },
    {
      "epoch": 1.6866359447004609,
      "grad_norm": 0.49317572687267214,
      "learning_rate": 4.757188889427761e-06,
      "loss": 0.3936,
      "step": 122
    },
    {
      "epoch": 1.7004608294930876,
      "grad_norm": 0.5116997857144775,
      "learning_rate": 4.6763509081278215e-06,
      "loss": 0.3812,
      "step": 123
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.5050661891764184,
      "learning_rate": 4.59559779819298e-06,
      "loss": 0.3744,
      "step": 124
    },
    {
      "epoch": 1.728110599078341,
      "grad_norm": 0.5660294368372862,
      "learning_rate": 4.51495073572676e-06,
      "loss": 0.4133,
      "step": 125
    },
    {
      "epoch": 1.7419354838709677,
      "grad_norm": 0.5414753484738278,
      "learning_rate": 4.434430869023579e-06,
      "loss": 0.4059,
      "step": 126
    },
    {
      "epoch": 1.7557603686635943,
      "grad_norm": 0.5274523827704818,
      "learning_rate": 4.3540593130229695e-06,
      "loss": 0.4058,
      "step": 127
    },
    {
      "epoch": 1.769585253456221,
      "grad_norm": 0.48723665989341286,
      "learning_rate": 4.27385714377255e-06,
      "loss": 0.4261,
      "step": 128
    },
    {
      "epoch": 1.7834101382488479,
      "grad_norm": 0.5557410469617062,
      "learning_rate": 4.1938453929012014e-06,
      "loss": 0.3799,
      "step": 129
    },
    {
      "epoch": 1.7972350230414746,
      "grad_norm": 0.5628455615899114,
      "learning_rate": 4.1140450421038865e-06,
      "loss": 0.4027,
      "step": 130
    },
    {
      "epoch": 1.8110599078341014,
      "grad_norm": 0.5407292854180369,
      "learning_rate": 4.034477017639561e-06,
      "loss": 0.3414,
      "step": 131
    },
    {
      "epoch": 1.8248847926267282,
      "grad_norm": 0.5110854563369478,
      "learning_rate": 3.955162184843625e-06,
      "loss": 0.4054,
      "step": 132
    },
    {
      "epoch": 1.838709677419355,
      "grad_norm": 0.573230524328855,
      "learning_rate": 3.8761213426563546e-06,
      "loss": 0.4111,
      "step": 133
    },
    {
      "epoch": 1.8525345622119815,
      "grad_norm": 0.5266763198269163,
      "learning_rate": 3.7973752181687336e-06,
      "loss": 0.3725,
      "step": 134
    },
    {
      "epoch": 1.8663594470046083,
      "grad_norm": 0.5053510634901648,
      "learning_rate": 3.7189444611871383e-06,
      "loss": 0.3989,
      "step": 135
    },
    {
      "epoch": 1.8801843317972349,
      "grad_norm": 0.5767131796506803,
      "learning_rate": 3.6408496388182857e-06,
      "loss": 0.456,
      "step": 136
    },
    {
      "epoch": 1.8940092165898617,
      "grad_norm": 0.4875050911999597,
      "learning_rate": 3.5631112300758595e-06,
      "loss": 0.3442,
      "step": 137
    },
    {
      "epoch": 1.9078341013824884,
      "grad_norm": 0.5145298101038916,
      "learning_rate": 3.4857496205102475e-06,
      "loss": 0.4418,
      "step": 138
    },
    {
      "epoch": 1.9216589861751152,
      "grad_norm": 0.46196122478440654,
      "learning_rate": 3.4087850968627823e-06,
      "loss": 0.3882,
      "step": 139
    },
    {
      "epoch": 1.935483870967742,
      "grad_norm": 0.4906484173808876,
      "learning_rate": 3.3322378417458985e-06,
      "loss": 0.3947,
      "step": 140
    },
    {
      "epoch": 1.9493087557603688,
      "grad_norm": 0.4989650080684894,
      "learning_rate": 3.2561279283505888e-06,
      "loss": 0.3399,
      "step": 141
    },
    {
      "epoch": 1.9631336405529956,
      "grad_norm": 0.46237620183821815,
      "learning_rate": 3.180475315182563e-06,
      "loss": 0.405,
      "step": 142
    },
    {
      "epoch": 1.976958525345622,
      "grad_norm": 0.4637781556805509,
      "learning_rate": 3.1052998408284664e-06,
      "loss": 0.4145,
      "step": 143
    },
    {
      "epoch": 1.9907834101382489,
      "grad_norm": 0.4775120449168806,
      "learning_rate": 3.0306212187535653e-06,
      "loss": 0.3988,
      "step": 144
    },
    {
      "epoch": 2.0046082949308754,
      "grad_norm": 0.9379207114350527,
      "learning_rate": 2.9564590321322206e-06,
      "loss": 0.6161,
      "step": 145
    },
    {
      "epoch": 2.0184331797235022,
      "grad_norm": 0.4763284552212227,
      "learning_rate": 2.882832728712551e-06,
      "loss": 0.3328,
      "step": 146
    },
    {
      "epoch": 2.032258064516129,
      "grad_norm": 0.5068219788042055,
      "learning_rate": 2.8097616157165886e-06,
      "loss": 0.3514,
      "step": 147
    },
    {
      "epoch": 2.046082949308756,
      "grad_norm": 0.5004585471417805,
      "learning_rate": 2.7372648547773063e-06,
      "loss": 0.3663,
      "step": 148
    },
    {
      "epoch": 2.0599078341013826,
      "grad_norm": 0.48776506523697694,
      "learning_rate": 2.665361456913797e-06,
      "loss": 0.3557,
      "step": 149
    },
    {
      "epoch": 2.0737327188940093,
      "grad_norm": 0.4565469138842312,
      "learning_rate": 2.594070277545975e-06,
      "loss": 0.3325,
      "step": 150
    },
    {
      "epoch": 2.087557603686636,
      "grad_norm": 0.5109481857152307,
      "learning_rate": 2.5234100115500643e-06,
      "loss": 0.3723,
      "step": 151
    },
    {
      "epoch": 2.1013824884792625,
      "grad_norm": 0.503915544983321,
      "learning_rate": 2.4533991883561868e-06,
      "loss": 0.3454,
      "step": 152
    },
    {
      "epoch": 2.1152073732718892,
      "grad_norm": 0.4920495646406938,
      "learning_rate": 2.38405616708935e-06,
      "loss": 0.3491,
      "step": 153
    },
    {
      "epoch": 2.129032258064516,
      "grad_norm": 0.46249250593080354,
      "learning_rate": 2.315399131755081e-06,
      "loss": 0.3714,
      "step": 154
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.4756209372804288,
      "learning_rate": 2.2474460864709825e-06,
      "loss": 0.341,
      "step": 155
    },
    {
      "epoch": 2.1566820276497696,
      "grad_norm": 0.49448121204775364,
      "learning_rate": 2.1802148507454675e-06,
      "loss": 0.3362,
      "step": 156
    },
    {
      "epoch": 2.1705069124423964,
      "grad_norm": 0.49087318600689567,
      "learning_rate": 2.1137230548049042e-06,
      "loss": 0.3712,
      "step": 157
    },
    {
      "epoch": 2.184331797235023,
      "grad_norm": 0.4394130757578322,
      "learning_rate": 2.0479881349703885e-06,
      "loss": 0.3289,
      "step": 158
    },
    {
      "epoch": 2.19815668202765,
      "grad_norm": 0.49665520195966556,
      "learning_rate": 1.983027329085377e-06,
      "loss": 0.3648,
      "step": 159
    },
    {
      "epoch": 2.2119815668202767,
      "grad_norm": 0.4113878764187301,
      "learning_rate": 1.9188576719953635e-06,
      "loss": 0.2849,
      "step": 160
    },
    {
      "epoch": 2.225806451612903,
      "grad_norm": 0.4999778833865041,
      "learning_rate": 1.8554959910807773e-06,
      "loss": 0.3695,
      "step": 161
    },
    {
      "epoch": 2.23963133640553,
      "grad_norm": 0.5189860690853264,
      "learning_rate": 1.7929589018443016e-06,
      "loss": 0.3504,
      "step": 162
    },
    {
      "epoch": 2.2534562211981566,
      "grad_norm": 0.48043530531701206,
      "learning_rate": 1.7312628035537388e-06,
      "loss": 0.332,
      "step": 163
    },
    {
      "epoch": 2.2672811059907834,
      "grad_norm": 0.5324429864403901,
      "learning_rate": 1.6704238749415958e-06,
      "loss": 0.3387,
      "step": 164
    },
    {
      "epoch": 2.28110599078341,
      "grad_norm": 0.5029628675101937,
      "learning_rate": 1.6104580699624839e-06,
      "loss": 0.3592,
      "step": 165
    },
    {
      "epoch": 2.294930875576037,
      "grad_norm": 0.45804507081769336,
      "learning_rate": 1.5513811136094786e-06,
      "loss": 0.3063,
      "step": 166
    },
    {
      "epoch": 2.3087557603686637,
      "grad_norm": 0.4647295250952741,
      "learning_rate": 1.4932084977905043e-06,
      "loss": 0.3254,
      "step": 167
    },
    {
      "epoch": 2.3225806451612905,
      "grad_norm": 0.46453643921934695,
      "learning_rate": 1.4359554772658551e-06,
      "loss": 0.3774,
      "step": 168
    },
    {
      "epoch": 2.3364055299539173,
      "grad_norm": 0.45343468728467234,
      "learning_rate": 1.3796370656478936e-06,
      "loss": 0.3325,
      "step": 169
    },
    {
      "epoch": 2.3502304147465436,
      "grad_norm": 0.4732479977172983,
      "learning_rate": 1.3242680314639995e-06,
      "loss": 0.3373,
      "step": 170
    },
    {
      "epoch": 2.3640552995391704,
      "grad_norm": 0.47208675924851334,
      "learning_rate": 1.2698628942837698e-06,
      "loss": 0.3427,
      "step": 171
    },
    {
      "epoch": 2.377880184331797,
      "grad_norm": 0.4454283369984702,
      "learning_rate": 1.2164359209115235e-06,
      "loss": 0.3257,
      "step": 172
    },
    {
      "epoch": 2.391705069124424,
      "grad_norm": 0.49273795207826715,
      "learning_rate": 1.164001121645069e-06,
      "loss": 0.3772,
      "step": 173
    },
    {
      "epoch": 2.4055299539170507,
      "grad_norm": 0.46302965066920987,
      "learning_rate": 1.1125722466017547e-06,
      "loss": 0.326,
      "step": 174
    },
    {
      "epoch": 2.4193548387096775,
      "grad_norm": 0.44068332388009435,
      "learning_rate": 1.062162782112729e-06,
      "loss": 0.3947,
      "step": 175
    },
    {
      "epoch": 2.4331797235023043,
      "grad_norm": 0.41223830858735033,
      "learning_rate": 1.012785947186397e-06,
      "loss": 0.3027,
      "step": 176
    },
    {
      "epoch": 2.447004608294931,
      "grad_norm": 0.44346734726903453,
      "learning_rate": 9.644546900419533e-07,
      "loss": 0.3306,
      "step": 177
    },
    {
      "epoch": 2.460829493087558,
      "grad_norm": 0.45992313059173967,
      "learning_rate": 9.171816847139447e-07,
      "loss": 0.3596,
      "step": 178
    },
    {
      "epoch": 2.474654377880184,
      "grad_norm": 0.4558549067269223,
      "learning_rate": 8.709793277287182e-07,
      "loss": 0.3684,
      "step": 179
    },
    {
      "epoch": 2.488479262672811,
      "grad_norm": 0.41803380966992537,
      "learning_rate": 8.258597348536452e-07,
      "loss": 0.32,
      "step": 180
    },
    {
      "epoch": 2.5023041474654377,
      "grad_norm": 0.42403429172582247,
      "learning_rate": 7.818347379199781e-07,
      "loss": 0.3273,
      "step": 181
    },
    {
      "epoch": 2.5161290322580645,
      "grad_norm": 0.46169239465826656,
      "learning_rate": 7.389158817201541e-07,
      "loss": 0.3313,
      "step": 182
    },
    {
      "epoch": 2.5299539170506913,
      "grad_norm": 0.49336488496655645,
      "learning_rate": 6.971144209803738e-07,
      "loss": 0.369,
      "step": 183
    },
    {
      "epoch": 2.543778801843318,
      "grad_norm": 0.47648081907983714,
      "learning_rate": 6.564413174092443e-07,
      "loss": 0.3464,
      "step": 184
    },
    {
      "epoch": 2.557603686635945,
      "grad_norm": 0.4151223124023637,
      "learning_rate": 6.16907236823262e-07,
      "loss": 0.3334,
      "step": 185
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.4214581402580324,
      "learning_rate": 5.785225463498828e-07,
      "loss": 0.3336,
      "step": 186
    },
    {
      "epoch": 2.5852534562211984,
      "grad_norm": 0.4404488936158401,
      "learning_rate": 5.412973117089288e-07,
      "loss": 0.357,
      "step": 187
    },
    {
      "epoch": 2.5990783410138247,
      "grad_norm": 0.4694197953797554,
      "learning_rate": 5.05241294573024e-07,
      "loss": 0.3389,
      "step": 188
    },
    {
      "epoch": 2.6129032258064515,
      "grad_norm": 0.43224485913576827,
      "learning_rate": 4.7036395000776556e-07,
      "loss": 0.3816,
      "step": 189
    },
    {
      "epoch": 2.6267281105990783,
      "grad_norm": 0.46915895400007485,
      "learning_rate": 4.3667442399229985e-07,
      "loss": 0.3212,
      "step": 190
    },
    {
      "epoch": 2.640552995391705,
      "grad_norm": 0.41455265232398864,
      "learning_rate": 4.041815510209396e-07,
      "loss": 0.349,
      "step": 191
    },
    {
      "epoch": 2.654377880184332,
      "grad_norm": 0.4138588542457796,
      "learning_rate": 3.728938517864794e-07,
      "loss": 0.3155,
      "step": 192
    },
    {
      "epoch": 2.6682027649769586,
      "grad_norm": 0.44949477525662346,
      "learning_rate": 3.4281953094578877e-07,
      "loss": 0.3608,
      "step": 193
    },
    {
      "epoch": 2.6820276497695854,
      "grad_norm": 0.45077480983651,
      "learning_rate": 3.1396647496828245e-07,
      "loss": 0.366,
      "step": 194
    },
    {
      "epoch": 2.6958525345622117,
      "grad_norm": 0.43977571760655004,
      "learning_rate": 2.8634225006782867e-07,
      "loss": 0.3447,
      "step": 195
    },
    {
      "epoch": 2.709677419354839,
      "grad_norm": 0.4531801768047782,
      "learning_rate": 2.599541002186479e-07,
      "loss": 0.3791,
      "step": 196
    },
    {
      "epoch": 2.7235023041474653,
      "grad_norm": 0.40108250009956564,
      "learning_rate": 2.3480894525569564e-07,
      "loss": 0.303,
      "step": 197
    },
    {
      "epoch": 2.737327188940092,
      "grad_norm": 0.40980648885083565,
      "learning_rate": 2.109133790600648e-07,
      "loss": 0.3427,
      "step": 198
    },
    {
      "epoch": 2.751152073732719,
      "grad_norm": 0.41297927531361595,
      "learning_rate": 1.8827366782984913e-07,
      "loss": 0.325,
      "step": 199
    },
    {
      "epoch": 2.7649769585253456,
      "grad_norm": 0.429935038739435,
      "learning_rate": 1.6689574843694433e-07,
      "loss": 0.3852,
      "step": 200
    },
    {
      "epoch": 2.7788018433179724,
      "grad_norm": 0.4020104101913433,
      "learning_rate": 1.4678522687020414e-07,
      "loss": 0.3513,
      "step": 201
    },
    {
      "epoch": 2.792626728110599,
      "grad_norm": 0.39219309859175616,
      "learning_rate": 1.2794737676536993e-07,
      "loss": 0.331,
      "step": 202
    },
    {
      "epoch": 2.806451612903226,
      "grad_norm": 0.4097391288745776,
      "learning_rate": 1.1038713802214718e-07,
      "loss": 0.3411,
      "step": 203
    },
    {
      "epoch": 2.8202764976958523,
      "grad_norm": 0.4309769162405172,
      "learning_rate": 9.410911550880474e-08,
      "loss": 0.3244,
      "step": 204
    },
    {
      "epoch": 2.8341013824884795,
      "grad_norm": 0.4449894760218027,
      "learning_rate": 7.911757785462882e-08,
      "loss": 0.3417,
      "step": 205
    },
    {
      "epoch": 2.847926267281106,
      "grad_norm": 0.4303644153909974,
      "learning_rate": 6.54164563305465e-08,
      "loss": 0.3391,
      "step": 206
    },
    {
      "epoch": 2.8617511520737327,
      "grad_norm": 0.42194746880922407,
      "learning_rate": 5.3009343818219985e-08,
      "loss": 0.3707,
      "step": 207
    },
    {
      "epoch": 2.8755760368663594,
      "grad_norm": 0.4196594211589896,
      "learning_rate": 4.189949386787462e-08,
      "loss": 0.351,
      "step": 208
    },
    {
      "epoch": 2.889400921658986,
      "grad_norm": 0.40539018280598516,
      "learning_rate": 3.2089819845111946e-08,
      "loss": 0.3168,
      "step": 209
    },
    {
      "epoch": 2.903225806451613,
      "grad_norm": 0.4658694331536401,
      "learning_rate": 2.358289416693027e-08,
      "loss": 0.4198,
      "step": 210
    },
    {
      "epoch": 2.9170506912442398,
      "grad_norm": 0.41736837046748076,
      "learning_rate": 1.6380947627153143e-08,
      "loss": 0.3174,
      "step": 211
    },
    {
      "epoch": 2.9308755760368665,
      "grad_norm": 0.42675226026620394,
      "learning_rate": 1.0485868811441757e-08,
      "loss": 0.3538,
      "step": 212
    },
    {
      "epoch": 2.944700460829493,
      "grad_norm": 0.4160647077774905,
      "learning_rate": 5.899203602046655e-09,
      "loss": 0.3474,
      "step": 213
    },
    {
      "epoch": 2.9585253456221197,
      "grad_norm": 0.40018027901405295,
      "learning_rate": 2.6221547724253337e-09,
      "loss": 0.319,
      "step": 214
    },
    {
      "epoch": 2.9723502304147464,
      "grad_norm": 0.4334075522903773,
      "learning_rate": 6.555816718389896e-10,
      "loss": 0.3884,
      "step": 215
    },
    {
      "epoch": 2.986175115207373,
      "grad_norm": 0.4031378328194389,
      "learning_rate": 0.0,
      "loss": 0.331,
      "step": 216
    },
    {
      "epoch": 2.986175115207373,
      "step": 216,
      "total_flos": 130743839866880.0,
      "train_loss": 0.4320466773653472,
      "train_runtime": 2638.4795,
      "train_samples_per_second": 7.869,
      "train_steps_per_second": 0.082
    }
  ],
  "logging_steps": 1,
  "max_steps": 216,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 130743839866880.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}