{
  "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": 3.8321747022193033,
      "learning_rate": 4.5454545454545457e-07,
      "loss": 0.4904,
      "step": 1
    },
    {
      "epoch": 0.027649769585253458,
      "grad_norm": 4.166445470011081,
      "learning_rate": 9.090909090909091e-07,
      "loss": 0.4586,
      "step": 2
    },
    {
      "epoch": 0.041474654377880185,
      "grad_norm": 3.8454000093459273,
      "learning_rate": 1.3636363636363636e-06,
      "loss": 0.4491,
      "step": 3
    },
    {
      "epoch": 0.055299539170506916,
      "grad_norm": 3.9389330204801944,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 0.495,
      "step": 4
    },
    {
      "epoch": 0.06912442396313365,
      "grad_norm": 3.503496575080232,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.4518,
      "step": 5
    },
    {
      "epoch": 0.08294930875576037,
      "grad_norm": 2.5778889382942296,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 0.4245,
      "step": 6
    },
    {
      "epoch": 0.0967741935483871,
      "grad_norm": 2.6973870749604654,
      "learning_rate": 3.181818181818182e-06,
      "loss": 0.4485,
      "step": 7
    },
    {
      "epoch": 0.11059907834101383,
      "grad_norm": 2.087801488195196,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.4126,
      "step": 8
    },
    {
      "epoch": 0.12442396313364056,
      "grad_norm": 1.937230458954793,
      "learning_rate": 4.0909090909090915e-06,
      "loss": 0.3782,
      "step": 9
    },
    {
      "epoch": 0.1382488479262673,
      "grad_norm": 1.4844954637777827,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.4114,
      "step": 10
    },
    {
      "epoch": 0.15207373271889402,
      "grad_norm": 1.6047087443414494,
      "learning_rate": 5e-06,
      "loss": 0.4018,
      "step": 11
    },
    {
      "epoch": 0.16589861751152074,
      "grad_norm": 1.2519442814402522,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 0.3739,
      "step": 12
    },
    {
      "epoch": 0.17972350230414746,
      "grad_norm": 1.11650299203384,
      "learning_rate": 5.90909090909091e-06,
      "loss": 0.3516,
      "step": 13
    },
    {
      "epoch": 0.1935483870967742,
      "grad_norm": 1.3867124167107647,
      "learning_rate": 6.363636363636364e-06,
      "loss": 0.3873,
      "step": 14
    },
    {
      "epoch": 0.2073732718894009,
      "grad_norm": 1.6171625319611096,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.3635,
      "step": 15
    },
    {
      "epoch": 0.22119815668202766,
      "grad_norm": 1.4344025164842837,
      "learning_rate": 7.272727272727273e-06,
      "loss": 0.3538,
      "step": 16
    },
    {
      "epoch": 0.2350230414746544,
      "grad_norm": 1.2769447355987351,
      "learning_rate": 7.727272727272727e-06,
      "loss": 0.3322,
      "step": 17
    },
    {
      "epoch": 0.2488479262672811,
      "grad_norm": 1.1544799589527472,
      "learning_rate": 8.181818181818183e-06,
      "loss": 0.3175,
      "step": 18
    },
    {
      "epoch": 0.2626728110599078,
      "grad_norm": 0.9340143537679371,
      "learning_rate": 8.636363636363637e-06,
      "loss": 0.2943,
      "step": 19
    },
    {
      "epoch": 0.2764976958525346,
      "grad_norm": 0.8350563937099799,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.3269,
      "step": 20
    },
    {
      "epoch": 0.2903225806451613,
      "grad_norm": 0.9427953311385877,
      "learning_rate": 9.545454545454547e-06,
      "loss": 0.3665,
      "step": 21
    },
    {
      "epoch": 0.30414746543778803,
      "grad_norm": 0.8571337937972116,
      "learning_rate": 1e-05,
      "loss": 0.3081,
      "step": 22
    },
    {
      "epoch": 0.31797235023041476,
      "grad_norm": 0.9574970030840929,
      "learning_rate": 9.999344418328161e-06,
      "loss": 0.3216,
      "step": 23
    },
    {
      "epoch": 0.3317972350230415,
      "grad_norm": 0.8240000968027297,
      "learning_rate": 9.997377845227577e-06,
      "loss": 0.322,
      "step": 24
    },
    {
      "epoch": 0.3456221198156682,
      "grad_norm": 0.7442732006332352,
      "learning_rate": 9.994100796397954e-06,
      "loss": 0.3406,
      "step": 25
    },
    {
      "epoch": 0.35944700460829493,
      "grad_norm": 0.862489472956441,
      "learning_rate": 9.98951413118856e-06,
      "loss": 0.342,
      "step": 26
    },
    {
      "epoch": 0.37327188940092165,
      "grad_norm": 0.7897584822354033,
      "learning_rate": 9.983619052372847e-06,
      "loss": 0.3228,
      "step": 27
    },
    {
      "epoch": 0.3870967741935484,
      "grad_norm": 0.8247735964866483,
      "learning_rate": 9.97641710583307e-06,
      "loss": 0.34,
      "step": 28
    },
    {
      "epoch": 0.4009216589861751,
      "grad_norm": 0.6877859922485577,
      "learning_rate": 9.96791018015489e-06,
      "loss": 0.3624,
      "step": 29
    },
    {
      "epoch": 0.4147465437788018,
      "grad_norm": 0.6854976819111457,
      "learning_rate": 9.958100506132127e-06,
      "loss": 0.3362,
      "step": 30
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.6780848305353676,
      "learning_rate": 9.946990656181782e-06,
      "loss": 0.3306,
      "step": 31
    },
    {
      "epoch": 0.4423963133640553,
      "grad_norm": 0.8635579814520753,
      "learning_rate": 9.934583543669454e-06,
      "loss": 0.333,
      "step": 32
    },
    {
      "epoch": 0.45622119815668205,
      "grad_norm": 0.7095969477530761,
      "learning_rate": 9.920882422145372e-06,
      "loss": 0.3297,
      "step": 33
    },
    {
      "epoch": 0.4700460829493088,
      "grad_norm": 0.6554896662064854,
      "learning_rate": 9.905890884491196e-06,
      "loss": 0.3474,
      "step": 34
    },
    {
      "epoch": 0.4838709677419355,
      "grad_norm": 0.7233953981673558,
      "learning_rate": 9.889612861977855e-06,
      "loss": 0.3366,
      "step": 35
    },
    {
      "epoch": 0.4976958525345622,
      "grad_norm": 0.7808564766414073,
      "learning_rate": 9.872052623234632e-06,
      "loss": 0.3247,
      "step": 36
    },
    {
      "epoch": 0.511520737327189,
      "grad_norm": 0.7662284906743654,
      "learning_rate": 9.853214773129796e-06,
      "loss": 0.3035,
      "step": 37
    },
    {
      "epoch": 0.5253456221198156,
      "grad_norm": 0.7589864104505978,
      "learning_rate": 9.833104251563058e-06,
      "loss": 0.3222,
      "step": 38
    },
    {
      "epoch": 0.5391705069124424,
      "grad_norm": 0.7364643638285358,
      "learning_rate": 9.811726332170153e-06,
      "loss": 0.3557,
      "step": 39
    },
    {
      "epoch": 0.5529953917050692,
      "grad_norm": 0.7228711702399517,
      "learning_rate": 9.789086620939936e-06,
      "loss": 0.3392,
      "step": 40
    },
    {
      "epoch": 0.5668202764976958,
      "grad_norm": 0.6418796946575632,
      "learning_rate": 9.765191054744305e-06,
      "loss": 0.3305,
      "step": 41
    },
    {
      "epoch": 0.5806451612903226,
      "grad_norm": 0.7360519892248771,
      "learning_rate": 9.740045899781353e-06,
      "loss": 0.3387,
      "step": 42
    },
    {
      "epoch": 0.5944700460829493,
      "grad_norm": 0.7117878571151136,
      "learning_rate": 9.713657749932172e-06,
      "loss": 0.378,
      "step": 43
    },
    {
      "epoch": 0.6082949308755761,
      "grad_norm": 0.6562391589488417,
      "learning_rate": 9.68603352503172e-06,
      "loss": 0.3808,
      "step": 44
    },
    {
      "epoch": 0.6221198156682027,
      "grad_norm": 0.7011857576729418,
      "learning_rate": 9.657180469054213e-06,
      "loss": 0.3561,
      "step": 45
    },
    {
      "epoch": 0.6359447004608295,
      "grad_norm": 0.6588644804163393,
      "learning_rate": 9.627106148213521e-06,
      "loss": 0.3224,
      "step": 46
    },
    {
      "epoch": 0.6497695852534562,
      "grad_norm": 0.6413771673007567,
      "learning_rate": 9.595818448979061e-06,
      "loss": 0.3502,
      "step": 47
    },
    {
      "epoch": 0.663594470046083,
      "grad_norm": 0.6132130927693553,
      "learning_rate": 9.563325576007702e-06,
      "loss": 0.3172,
      "step": 48
    },
    {
      "epoch": 0.6774193548387096,
      "grad_norm": 0.7562554878178558,
      "learning_rate": 9.529636049992235e-06,
      "loss": 0.3078,
      "step": 49
    },
    {
      "epoch": 0.6912442396313364,
      "grad_norm": 0.7025632258958434,
      "learning_rate": 9.494758705426978e-06,
      "loss": 0.3378,
      "step": 50
    },
    {
      "epoch": 0.7050691244239631,
      "grad_norm": 0.6580901818465655,
      "learning_rate": 9.458702688291072e-06,
      "loss": 0.3749,
      "step": 51
    },
    {
      "epoch": 0.7188940092165899,
      "grad_norm": 0.837525662904706,
      "learning_rate": 9.421477453650118e-06,
      "loss": 0.3646,
      "step": 52
    },
    {
      "epoch": 0.7327188940092166,
      "grad_norm": 0.6988946242836359,
      "learning_rate": 9.38309276317674e-06,
      "loss": 0.2952,
      "step": 53
    },
    {
      "epoch": 0.7465437788018433,
      "grad_norm": 0.7088548576538783,
      "learning_rate": 9.343558682590757e-06,
      "loss": 0.3402,
      "step": 54
    },
    {
      "epoch": 0.7603686635944701,
      "grad_norm": 0.6031364330217948,
      "learning_rate": 9.302885579019626e-06,
      "loss": 0.3157,
      "step": 55
    },
    {
      "epoch": 0.7741935483870968,
      "grad_norm": 0.6869976637076074,
      "learning_rate": 9.261084118279846e-06,
      "loss": 0.3724,
      "step": 56
    },
    {
      "epoch": 0.7880184331797235,
      "grad_norm": 0.6362516594505776,
      "learning_rate": 9.218165262080024e-06,
      "loss": 0.2915,
      "step": 57
    },
    {
      "epoch": 0.8018433179723502,
      "grad_norm": 0.6133826251571223,
      "learning_rate": 9.174140265146356e-06,
      "loss": 0.3374,
      "step": 58
    },
    {
      "epoch": 0.815668202764977,
      "grad_norm": 0.5552231441277784,
      "learning_rate": 9.129020672271283e-06,
      "loss": 0.3119,
      "step": 59
    },
    {
      "epoch": 0.8294930875576036,
      "grad_norm": 0.5344163684333119,
      "learning_rate": 9.082818315286054e-06,
      "loss": 0.3139,
      "step": 60
    },
    {
      "epoch": 0.8433179723502304,
      "grad_norm": 0.5908367481260651,
      "learning_rate": 9.035545309958048e-06,
      "loss": 0.3391,
      "step": 61
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.6049797134002246,
      "learning_rate": 8.987214052813605e-06,
      "loss": 0.3385,
      "step": 62
    },
    {
      "epoch": 0.8709677419354839,
      "grad_norm": 0.632151936446925,
      "learning_rate": 8.937837217887273e-06,
      "loss": 0.3216,
      "step": 63
    },
    {
      "epoch": 0.8847926267281107,
      "grad_norm": 0.573287253380456,
      "learning_rate": 8.887427753398249e-06,
      "loss": 0.2931,
      "step": 64
    },
    {
      "epoch": 0.8986175115207373,
      "grad_norm": 0.5852983805141764,
      "learning_rate": 8.83599887835493e-06,
      "loss": 0.3172,
      "step": 65
    },
    {
      "epoch": 0.9124423963133641,
      "grad_norm": 0.6363174322514683,
      "learning_rate": 8.783564079088478e-06,
      "loss": 0.3882,
      "step": 66
    },
    {
      "epoch": 0.9262672811059908,
      "grad_norm": 0.6351465289963182,
      "learning_rate": 8.730137105716231e-06,
      "loss": 0.2963,
      "step": 67
    },
    {
      "epoch": 0.9400921658986175,
      "grad_norm": 0.5424528987366075,
      "learning_rate": 8.675731968536004e-06,
      "loss": 0.2824,
      "step": 68
    },
    {
      "epoch": 0.9539170506912442,
      "grad_norm": 0.5850480108903799,
      "learning_rate": 8.620362934352109e-06,
      "loss": 0.3439,
      "step": 69
    },
    {
      "epoch": 0.967741935483871,
      "grad_norm": 0.5924509545505863,
      "learning_rate": 8.564044522734147e-06,
      "loss": 0.345,
      "step": 70
    },
    {
      "epoch": 0.9815668202764977,
      "grad_norm": 0.6356218187245817,
      "learning_rate": 8.506791502209497e-06,
      "loss": 0.3383,
      "step": 71
    },
    {
      "epoch": 0.9953917050691244,
      "grad_norm": 0.5255727075687244,
      "learning_rate": 8.448618886390523e-06,
      "loss": 0.31,
      "step": 72
    },
    {
      "epoch": 1.0092165898617511,
      "grad_norm": 1.075353564084177,
      "learning_rate": 8.389541930037516e-06,
      "loss": 0.4918,
      "step": 73
    },
    {
      "epoch": 1.023041474654378,
      "grad_norm": 0.5411851622437062,
      "learning_rate": 8.329576125058406e-06,
      "loss": 0.284,
      "step": 74
    },
    {
      "epoch": 1.0368663594470047,
      "grad_norm": 0.5438469424766922,
      "learning_rate": 8.268737196446264e-06,
      "loss": 0.2914,
      "step": 75
    },
    {
      "epoch": 1.0506912442396312,
      "grad_norm": 0.5420670331882803,
      "learning_rate": 8.207041098155701e-06,
      "loss": 0.2723,
      "step": 76
    },
    {
      "epoch": 1.064516129032258,
      "grad_norm": 0.6001856622683356,
      "learning_rate": 8.144504008919224e-06,
      "loss": 0.3119,
      "step": 77
    },
    {
      "epoch": 1.0783410138248848,
      "grad_norm": 0.5657652505048809,
      "learning_rate": 8.081142328004638e-06,
      "loss": 0.2598,
      "step": 78
    },
    {
      "epoch": 1.0921658986175116,
      "grad_norm": 0.5575714850260662,
      "learning_rate": 8.016972670914624e-06,
      "loss": 0.322,
      "step": 79
    },
    {
      "epoch": 1.1059907834101383,
      "grad_norm": 0.6083958309172656,
      "learning_rate": 7.952011865029614e-06,
      "loss": 0.2509,
      "step": 80
    },
    {
      "epoch": 1.119815668202765,
      "grad_norm": 0.5266548143398665,
      "learning_rate": 7.886276945195098e-06,
      "loss": 0.2206,
      "step": 81
    },
    {
      "epoch": 1.1336405529953917,
      "grad_norm": 0.5687307229026926,
      "learning_rate": 7.819785149254534e-06,
      "loss": 0.2954,
      "step": 82
    },
    {
      "epoch": 1.1474654377880185,
      "grad_norm": 0.5640940659138813,
      "learning_rate": 7.752553913529019e-06,
      "loss": 0.2628,
      "step": 83
    },
    {
      "epoch": 1.1612903225806452,
      "grad_norm": 0.6680289489599975,
      "learning_rate": 7.68460086824492e-06,
      "loss": 0.3206,
      "step": 84
    },
    {
      "epoch": 1.1751152073732718,
      "grad_norm": 0.5283769448858802,
      "learning_rate": 7.61594383291065e-06,
      "loss": 0.2363,
      "step": 85
    },
    {
      "epoch": 1.1889400921658986,
      "grad_norm": 0.6482385383082108,
      "learning_rate": 7.546600811643816e-06,
      "loss": 0.3353,
      "step": 86
    },
    {
      "epoch": 1.2027649769585254,
      "grad_norm": 0.5264348159834544,
      "learning_rate": 7.476589988449939e-06,
      "loss": 0.2618,
      "step": 87
    },
    {
      "epoch": 1.2165898617511521,
      "grad_norm": 0.5512488141503382,
      "learning_rate": 7.405929722454026e-06,
      "loss": 0.2639,
      "step": 88
    },
    {
      "epoch": 1.230414746543779,
      "grad_norm": 0.5093764966098362,
      "learning_rate": 7.334638543086203e-06,
      "loss": 0.2477,
      "step": 89
    },
    {
      "epoch": 1.2442396313364055,
      "grad_norm": 0.5160996147207031,
      "learning_rate": 7.262735145222696e-06,
      "loss": 0.2437,
      "step": 90
    },
    {
      "epoch": 1.2580645161290323,
      "grad_norm": 0.5507115707918743,
      "learning_rate": 7.190238384283413e-06,
      "loss": 0.2625,
      "step": 91
    },
    {
      "epoch": 1.271889400921659,
      "grad_norm": 0.5596845582953057,
      "learning_rate": 7.117167271287453e-06,
      "loss": 0.2838,
      "step": 92
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 0.5261309041679847,
      "learning_rate": 7.043540967867782e-06,
      "loss": 0.2596,
      "step": 93
    },
    {
      "epoch": 1.2995391705069124,
      "grad_norm": 0.606787789098388,
      "learning_rate": 6.969378781246436e-06,
      "loss": 0.295,
      "step": 94
    },
    {
      "epoch": 1.3133640552995391,
      "grad_norm": 0.5525493866624152,
      "learning_rate": 6.894700159171535e-06,
      "loss": 0.2478,
      "step": 95
    },
    {
      "epoch": 1.327188940092166,
      "grad_norm": 0.4948582007424209,
      "learning_rate": 6.819524684817439e-06,
      "loss": 0.2375,
      "step": 96
    },
    {
      "epoch": 1.3410138248847927,
      "grad_norm": 0.5910117531529238,
      "learning_rate": 6.743872071649411e-06,
      "loss": 0.293,
      "step": 97
    },
    {
      "epoch": 1.3548387096774195,
      "grad_norm": 0.5642544177081138,
      "learning_rate": 6.667762158254104e-06,
      "loss": 0.2759,
      "step": 98
    },
    {
      "epoch": 1.368663594470046,
      "grad_norm": 0.5640787140131098,
      "learning_rate": 6.591214903137221e-06,
      "loss": 0.2919,
      "step": 99
    },
    {
      "epoch": 1.3824884792626728,
      "grad_norm": 0.627883249036726,
      "learning_rate": 6.514250379489754e-06,
      "loss": 0.3236,
      "step": 100
    },
    {
      "epoch": 1.3963133640552996,
      "grad_norm": 0.5777799162176069,
      "learning_rate": 6.436888769924142e-06,
      "loss": 0.2599,
      "step": 101
    },
    {
      "epoch": 1.4101382488479262,
      "grad_norm": 0.5575789159801893,
      "learning_rate": 6.3591503611817155e-06,
      "loss": 0.2881,
      "step": 102
    },
    {
      "epoch": 1.423963133640553,
      "grad_norm": 0.5454339637824865,
      "learning_rate": 6.281055538812861e-06,
      "loss": 0.2528,
      "step": 103
    },
    {
      "epoch": 1.4377880184331797,
      "grad_norm": 0.5785302727121883,
      "learning_rate": 6.202624781831269e-06,
      "loss": 0.2586,
      "step": 104
    },
    {
      "epoch": 1.4516129032258065,
      "grad_norm": 0.5795551004814405,
      "learning_rate": 6.123878657343648e-06,
      "loss": 0.2933,
      "step": 105
    },
    {
      "epoch": 1.4654377880184333,
      "grad_norm": 0.5562868873733925,
      "learning_rate": 6.044837815156377e-06,
      "loss": 0.2682,
      "step": 106
    },
    {
      "epoch": 1.4792626728110598,
      "grad_norm": 0.5788310033089493,
      "learning_rate": 5.965522982360441e-06,
      "loss": 0.2574,
      "step": 107
    },
    {
      "epoch": 1.4930875576036866,
      "grad_norm": 0.5239597133970942,
      "learning_rate": 5.885954957896115e-06,
      "loss": 0.256,
      "step": 108
    },
    {
      "epoch": 1.5069124423963134,
      "grad_norm": 0.5873986776610259,
      "learning_rate": 5.806154607098799e-06,
      "loss": 0.3367,
      "step": 109
    },
    {
      "epoch": 1.52073732718894,
      "grad_norm": 0.573952411610799,
      "learning_rate": 5.726142856227453e-06,
      "loss": 0.2801,
      "step": 110
    },
    {
      "epoch": 1.5345622119815667,
      "grad_norm": 0.5754764716263057,
      "learning_rate": 5.645940686977033e-06,
      "loss": 0.2512,
      "step": 111
    },
    {
      "epoch": 1.5483870967741935,
      "grad_norm": 0.5612364021450903,
      "learning_rate": 5.5655691309764225e-06,
      "loss": 0.2938,
      "step": 112
    },
    {
      "epoch": 1.5622119815668203,
      "grad_norm": 0.5214616228011532,
      "learning_rate": 5.485049264273241e-06,
      "loss": 0.2598,
      "step": 113
    },
    {
      "epoch": 1.576036866359447,
      "grad_norm": 0.5442190086345122,
      "learning_rate": 5.404402201807022e-06,
      "loss": 0.2587,
      "step": 114
    },
    {
      "epoch": 1.5898617511520738,
      "grad_norm": 0.6223597688467173,
      "learning_rate": 5.323649091872179e-06,
      "loss": 0.2917,
      "step": 115
    },
    {
      "epoch": 1.6036866359447006,
      "grad_norm": 0.5267088823599309,
      "learning_rate": 5.242811110572243e-06,
      "loss": 0.2357,
      "step": 116
    },
    {
      "epoch": 1.6175115207373272,
      "grad_norm": 0.5310849049848287,
      "learning_rate": 5.161909456266781e-06,
      "loss": 0.2259,
      "step": 117
    },
    {
      "epoch": 1.631336405529954,
      "grad_norm": 0.5327550485763282,
      "learning_rate": 5.080965344012509e-06,
      "loss": 0.2918,
      "step": 118
    },
    {
      "epoch": 1.6451612903225805,
      "grad_norm": 0.48433766437116366,
      "learning_rate": 5e-06,
      "loss": 0.2057,
      "step": 119
    },
    {
      "epoch": 1.6589861751152073,
      "grad_norm": 0.5201772106049566,
      "learning_rate": 4.919034655987493e-06,
      "loss": 0.2469,
      "step": 120
    },
    {
      "epoch": 1.672811059907834,
      "grad_norm": 0.5728059995080512,
      "learning_rate": 4.838090543733222e-06,
      "loss": 0.2977,
      "step": 121
    },
    {
      "epoch": 1.6866359447004609,
      "grad_norm": 0.5451785777689631,
      "learning_rate": 4.757188889427761e-06,
      "loss": 0.2664,
      "step": 122
    },
    {
      "epoch": 1.7004608294930876,
      "grad_norm": 0.55325878163931,
      "learning_rate": 4.6763509081278215e-06,
      "loss": 0.2938,
      "step": 123
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.5482322683176598,
      "learning_rate": 4.59559779819298e-06,
      "loss": 0.2918,
      "step": 124
    },
    {
      "epoch": 1.728110599078341,
      "grad_norm": 0.5317903391669024,
      "learning_rate": 4.51495073572676e-06,
      "loss": 0.2593,
      "step": 125
    },
    {
      "epoch": 1.7419354838709677,
      "grad_norm": 0.5947363067321887,
      "learning_rate": 4.434430869023579e-06,
      "loss": 0.2565,
      "step": 126
    },
    {
      "epoch": 1.7557603686635943,
      "grad_norm": 0.5166626884834731,
      "learning_rate": 4.3540593130229695e-06,
      "loss": 0.2542,
      "step": 127
    },
    {
      "epoch": 1.769585253456221,
      "grad_norm": 0.5716701172322802,
      "learning_rate": 4.27385714377255e-06,
      "loss": 0.2809,
      "step": 128
    },
    {
      "epoch": 1.7834101382488479,
      "grad_norm": 0.548457756610785,
      "learning_rate": 4.1938453929012014e-06,
      "loss": 0.2656,
      "step": 129
    },
    {
      "epoch": 1.7972350230414746,
      "grad_norm": 0.5101224008525175,
      "learning_rate": 4.1140450421038865e-06,
      "loss": 0.2602,
      "step": 130
    },
    {
      "epoch": 1.8110599078341014,
      "grad_norm": 0.5688581933379532,
      "learning_rate": 4.034477017639561e-06,
      "loss": 0.2708,
      "step": 131
    },
    {
      "epoch": 1.8248847926267282,
      "grad_norm": 0.5608572985897213,
      "learning_rate": 3.955162184843625e-06,
      "loss": 0.2728,
      "step": 132
    },
    {
      "epoch": 1.838709677419355,
      "grad_norm": 0.60913972327474,
      "learning_rate": 3.8761213426563546e-06,
      "loss": 0.2763,
      "step": 133
    },
    {
      "epoch": 1.8525345622119815,
      "grad_norm": 0.5498604689214065,
      "learning_rate": 3.7973752181687336e-06,
      "loss": 0.2574,
      "step": 134
    },
    {
      "epoch": 1.8663594470046083,
      "grad_norm": 0.6459792319529313,
      "learning_rate": 3.7189444611871383e-06,
      "loss": 0.2768,
      "step": 135
    },
    {
      "epoch": 1.8801843317972349,
      "grad_norm": 0.6375472108520018,
      "learning_rate": 3.6408496388182857e-06,
      "loss": 0.3172,
      "step": 136
    },
    {
      "epoch": 1.8940092165898617,
      "grad_norm": 0.5323139638568577,
      "learning_rate": 3.5631112300758595e-06,
      "loss": 0.2375,
      "step": 137
    },
    {
      "epoch": 1.9078341013824884,
      "grad_norm": 0.5704026138530017,
      "learning_rate": 3.4857496205102475e-06,
      "loss": 0.2889,
      "step": 138
    },
    {
      "epoch": 1.9216589861751152,
      "grad_norm": 0.5751626829380007,
      "learning_rate": 3.4087850968627823e-06,
      "loss": 0.2668,
      "step": 139
    },
    {
      "epoch": 1.935483870967742,
      "grad_norm": 0.5176599089594356,
      "learning_rate": 3.3322378417458985e-06,
      "loss": 0.2675,
      "step": 140
    },
    {
      "epoch": 1.9493087557603688,
      "grad_norm": 0.605437544433559,
      "learning_rate": 3.2561279283505888e-06,
      "loss": 0.2711,
      "step": 141
    },
    {
      "epoch": 1.9631336405529956,
      "grad_norm": 0.5219033376075164,
      "learning_rate": 3.180475315182563e-06,
      "loss": 0.2665,
      "step": 142
    },
    {
      "epoch": 1.976958525345622,
      "grad_norm": 0.6188916507167108,
      "learning_rate": 3.1052998408284664e-06,
      "loss": 0.2708,
      "step": 143
    },
    {
      "epoch": 1.9907834101382489,
      "grad_norm": 0.5752086490165061,
      "learning_rate": 3.0306212187535653e-06,
      "loss": 0.2798,
      "step": 144
    },
    {
      "epoch": 2.0046082949308754,
      "grad_norm": 0.9737614239717218,
      "learning_rate": 2.9564590321322206e-06,
      "loss": 0.3732,
      "step": 145
    },
    {
      "epoch": 2.0184331797235022,
      "grad_norm": 0.6171348142735417,
      "learning_rate": 2.882832728712551e-06,
      "loss": 0.2203,
      "step": 146
    },
    {
      "epoch": 2.032258064516129,
      "grad_norm": 0.5478831551166238,
      "learning_rate": 2.8097616157165886e-06,
      "loss": 0.196,
      "step": 147
    },
    {
      "epoch": 2.046082949308756,
      "grad_norm": 0.5300984312964834,
      "learning_rate": 2.7372648547773063e-06,
      "loss": 0.2146,
      "step": 148
    },
    {
      "epoch": 2.0599078341013826,
      "grad_norm": 0.527691094200599,
      "learning_rate": 2.665361456913797e-06,
      "loss": 0.2404,
      "step": 149
    },
    {
      "epoch": 2.0737327188940093,
      "grad_norm": 0.5203329902273544,
      "learning_rate": 2.594070277545975e-06,
      "loss": 0.2051,
      "step": 150
    },
    {
      "epoch": 2.087557603686636,
      "grad_norm": 0.5500390166449323,
      "learning_rate": 2.5234100115500643e-06,
      "loss": 0.2648,
      "step": 151
    },
    {
      "epoch": 2.1013824884792625,
      "grad_norm": 0.4893779390529743,
      "learning_rate": 2.4533991883561868e-06,
      "loss": 0.2245,
      "step": 152
    },
    {
      "epoch": 2.1152073732718892,
      "grad_norm": 0.5441384122952133,
      "learning_rate": 2.38405616708935e-06,
      "loss": 0.2162,
      "step": 153
    },
    {
      "epoch": 2.129032258064516,
      "grad_norm": 0.6239869837559884,
      "learning_rate": 2.315399131755081e-06,
      "loss": 0.3197,
      "step": 154
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.5137901138193296,
      "learning_rate": 2.2474460864709825e-06,
      "loss": 0.2537,
      "step": 155
    },
    {
      "epoch": 2.1566820276497696,
      "grad_norm": 0.5666929612913303,
      "learning_rate": 2.1802148507454675e-06,
      "loss": 0.245,
      "step": 156
    },
    {
      "epoch": 2.1705069124423964,
      "grad_norm": 0.522249344639658,
      "learning_rate": 2.1137230548049042e-06,
      "loss": 0.2074,
      "step": 157
    },
    {
      "epoch": 2.184331797235023,
      "grad_norm": 0.5427924264706615,
      "learning_rate": 2.0479881349703885e-06,
      "loss": 0.1891,
      "step": 158
    },
    {
      "epoch": 2.19815668202765,
      "grad_norm": 0.5395578092838311,
      "learning_rate": 1.983027329085377e-06,
      "loss": 0.2141,
      "step": 159
    },
    {
      "epoch": 2.2119815668202767,
      "grad_norm": 0.6203214366115929,
      "learning_rate": 1.9188576719953635e-06,
      "loss": 0.2395,
      "step": 160
    },
    {
      "epoch": 2.225806451612903,
      "grad_norm": 0.5561057678969209,
      "learning_rate": 1.8554959910807773e-06,
      "loss": 0.2325,
      "step": 161
    },
    {
      "epoch": 2.23963133640553,
      "grad_norm": 0.4975297831721625,
      "learning_rate": 1.7929589018443016e-06,
      "loss": 0.2189,
      "step": 162
    },
    {
      "epoch": 2.2534562211981566,
      "grad_norm": 0.51442024085124,
      "learning_rate": 1.7312628035537388e-06,
      "loss": 0.1941,
      "step": 163
    },
    {
      "epoch": 2.2672811059907834,
      "grad_norm": 0.5689991394989393,
      "learning_rate": 1.6704238749415958e-06,
      "loss": 0.2274,
      "step": 164
    },
    {
      "epoch": 2.28110599078341,
      "grad_norm": 0.5545628564372191,
      "learning_rate": 1.6104580699624839e-06,
      "loss": 0.2582,
      "step": 165
    },
    {
      "epoch": 2.294930875576037,
      "grad_norm": 0.5532036253850108,
      "learning_rate": 1.5513811136094786e-06,
      "loss": 0.2518,
      "step": 166
    },
    {
      "epoch": 2.3087557603686637,
      "grad_norm": 0.4892665339324161,
      "learning_rate": 1.4932084977905043e-06,
      "loss": 0.1982,
      "step": 167
    },
    {
      "epoch": 2.3225806451612905,
      "grad_norm": 0.551947647990794,
      "learning_rate": 1.4359554772658551e-06,
      "loss": 0.2607,
      "step": 168
    },
    {
      "epoch": 2.3364055299539173,
      "grad_norm": 0.5151354916631968,
      "learning_rate": 1.3796370656478936e-06,
      "loss": 0.2072,
      "step": 169
    },
    {
      "epoch": 2.3502304147465436,
      "grad_norm": 0.5685893753938921,
      "learning_rate": 1.3242680314639995e-06,
      "loss": 0.2143,
      "step": 170
    },
    {
      "epoch": 2.3640552995391704,
      "grad_norm": 0.4880909328371241,
      "learning_rate": 1.2698628942837698e-06,
      "loss": 0.1718,
      "step": 171
    },
    {
      "epoch": 2.377880184331797,
      "grad_norm": 0.493208956793857,
      "learning_rate": 1.2164359209115235e-06,
      "loss": 0.1923,
      "step": 172
    },
    {
      "epoch": 2.391705069124424,
      "grad_norm": 0.5582991832891734,
      "learning_rate": 1.164001121645069e-06,
      "loss": 0.2524,
      "step": 173
    },
    {
      "epoch": 2.4055299539170507,
      "grad_norm": 0.505450721310959,
      "learning_rate": 1.1125722466017547e-06,
      "loss": 0.2125,
      "step": 174
    },
    {
      "epoch": 2.4193548387096775,
      "grad_norm": 0.5255011925428328,
      "learning_rate": 1.062162782112729e-06,
      "loss": 0.2456,
      "step": 175
    },
    {
      "epoch": 2.4331797235023043,
      "grad_norm": 0.4989323482397167,
      "learning_rate": 1.012785947186397e-06,
      "loss": 0.1834,
      "step": 176
    },
    {
      "epoch": 2.447004608294931,
      "grad_norm": 0.51435324430214,
      "learning_rate": 9.644546900419533e-07,
      "loss": 0.2056,
      "step": 177
    },
    {
      "epoch": 2.460829493087558,
      "grad_norm": 0.4941625397188085,
      "learning_rate": 9.171816847139447e-07,
      "loss": 0.2044,
      "step": 178
    },
    {
      "epoch": 2.474654377880184,
      "grad_norm": 0.5742862761852323,
      "learning_rate": 8.709793277287182e-07,
      "loss": 0.2334,
      "step": 179
    },
    {
      "epoch": 2.488479262672811,
      "grad_norm": 0.45964093874741846,
      "learning_rate": 8.258597348536452e-07,
      "loss": 0.1885,
      "step": 180
    },
    {
      "epoch": 2.5023041474654377,
      "grad_norm": 0.5736994982571209,
      "learning_rate": 7.818347379199781e-07,
      "loss": 0.2601,
      "step": 181
    },
    {
      "epoch": 2.5161290322580645,
      "grad_norm": 0.4946177722077173,
      "learning_rate": 7.389158817201541e-07,
      "loss": 0.1937,
      "step": 182
    },
    {
      "epoch": 2.5299539170506913,
      "grad_norm": 0.5484333491191469,
      "learning_rate": 6.971144209803738e-07,
      "loss": 0.2487,
      "step": 183
    },
    {
      "epoch": 2.543778801843318,
      "grad_norm": 0.5338015925288597,
      "learning_rate": 6.564413174092443e-07,
      "loss": 0.2134,
      "step": 184
    },
    {
      "epoch": 2.557603686635945,
      "grad_norm": 0.4987205048741818,
      "learning_rate": 6.16907236823262e-07,
      "loss": 0.2115,
      "step": 185
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.5116461576836463,
      "learning_rate": 5.785225463498828e-07,
      "loss": 0.2161,
      "step": 186
    },
    {
      "epoch": 2.5852534562211984,
      "grad_norm": 0.5549460042068343,
      "learning_rate": 5.412973117089288e-07,
      "loss": 0.249,
      "step": 187
    },
    {
      "epoch": 2.5990783410138247,
      "grad_norm": 0.4625486356955096,
      "learning_rate": 5.05241294573024e-07,
      "loss": 0.1794,
      "step": 188
    },
    {
      "epoch": 2.6129032258064515,
      "grad_norm": 0.5601735455591889,
      "learning_rate": 4.7036395000776556e-07,
      "loss": 0.2483,
      "step": 189
    },
    {
      "epoch": 2.6267281105990783,
      "grad_norm": 0.4873480362305034,
      "learning_rate": 4.3667442399229985e-07,
      "loss": 0.1891,
      "step": 190
    },
    {
      "epoch": 2.640552995391705,
      "grad_norm": 0.5036174624248448,
      "learning_rate": 4.041815510209396e-07,
      "loss": 0.2294,
      "step": 191
    },
    {
      "epoch": 2.654377880184332,
      "grad_norm": 0.5288779116171689,
      "learning_rate": 3.728938517864794e-07,
      "loss": 0.2278,
      "step": 192
    },
    {
      "epoch": 2.6682027649769586,
      "grad_norm": 0.5090457286467068,
      "learning_rate": 3.4281953094578877e-07,
      "loss": 0.2144,
      "step": 193
    },
    {
      "epoch": 2.6820276497695854,
      "grad_norm": 0.5256240057085287,
      "learning_rate": 3.1396647496828245e-07,
      "loss": 0.2517,
      "step": 194
    },
    {
      "epoch": 2.6958525345622117,
      "grad_norm": 0.5629758036361966,
      "learning_rate": 2.8634225006782867e-07,
      "loss": 0.2611,
      "step": 195
    },
    {
      "epoch": 2.709677419354839,
      "grad_norm": 0.5114348887925635,
      "learning_rate": 2.599541002186479e-07,
      "loss": 0.2059,
      "step": 196
    },
    {
      "epoch": 2.7235023041474653,
      "grad_norm": 0.467691586903513,
      "learning_rate": 2.3480894525569564e-07,
      "loss": 0.1805,
      "step": 197
    },
    {
      "epoch": 2.737327188940092,
      "grad_norm": 0.5136153625365067,
      "learning_rate": 2.109133790600648e-07,
      "loss": 0.2479,
      "step": 198
    },
    {
      "epoch": 2.751152073732719,
      "grad_norm": 0.521711267048856,
      "learning_rate": 1.8827366782984913e-07,
      "loss": 0.2388,
      "step": 199
    },
    {
      "epoch": 2.7649769585253456,
      "grad_norm": 0.5264204516952466,
      "learning_rate": 1.6689574843694433e-07,
      "loss": 0.2238,
      "step": 200
    },
    {
      "epoch": 2.7788018433179724,
      "grad_norm": 0.5123667240282413,
      "learning_rate": 1.4678522687020414e-07,
      "loss": 0.1927,
      "step": 201
    },
    {
      "epoch": 2.792626728110599,
      "grad_norm": 0.5109045161700035,
      "learning_rate": 1.2794737676536993e-07,
      "loss": 0.219,
      "step": 202
    },
    {
      "epoch": 2.806451612903226,
      "grad_norm": 0.5412780752972016,
      "learning_rate": 1.1038713802214718e-07,
      "loss": 0.2258,
      "step": 203
    },
    {
      "epoch": 2.8202764976958523,
      "grad_norm": 0.5001657996677221,
      "learning_rate": 9.410911550880474e-08,
      "loss": 0.2272,
      "step": 204
    },
    {
      "epoch": 2.8341013824884795,
      "grad_norm": 0.4830345984951105,
      "learning_rate": 7.911757785462882e-08,
      "loss": 0.1845,
      "step": 205
    },
    {
      "epoch": 2.847926267281106,
      "grad_norm": 0.4754821970835569,
      "learning_rate": 6.54164563305465e-08,
      "loss": 0.1772,
      "step": 206
    },
    {
      "epoch": 2.8617511520737327,
      "grad_norm": 0.5180006446663228,
      "learning_rate": 5.3009343818219985e-08,
      "loss": 0.2429,
      "step": 207
    },
    {
      "epoch": 2.8755760368663594,
      "grad_norm": 0.4922784862533948,
      "learning_rate": 4.189949386787462e-08,
      "loss": 0.2,
      "step": 208
    },
    {
      "epoch": 2.889400921658986,
      "grad_norm": 0.5733544961657199,
      "learning_rate": 3.2089819845111946e-08,
      "loss": 0.2812,
      "step": 209
    },
    {
      "epoch": 2.903225806451613,
      "grad_norm": 0.5752601209401671,
      "learning_rate": 2.358289416693027e-08,
      "loss": 0.2579,
      "step": 210
    },
    {
      "epoch": 2.9170506912442398,
      "grad_norm": 0.4757128270112064,
      "learning_rate": 1.6380947627153143e-08,
      "loss": 0.2217,
      "step": 211
    },
    {
      "epoch": 2.9308755760368665,
      "grad_norm": 0.5139378693377304,
      "learning_rate": 1.0485868811441757e-08,
      "loss": 0.2273,
      "step": 212
    },
    {
      "epoch": 2.944700460829493,
      "grad_norm": 0.5102846639395582,
      "learning_rate": 5.899203602046655e-09,
      "loss": 0.2158,
      "step": 213
    },
    {
      "epoch": 2.9585253456221197,
      "grad_norm": 0.5057982261906225,
      "learning_rate": 2.6221547724253337e-09,
      "loss": 0.2063,
      "step": 214
    },
    {
      "epoch": 2.9723502304147464,
      "grad_norm": 0.5645707694001406,
      "learning_rate": 6.555816718389896e-10,
      "loss": 0.2588,
      "step": 215
    },
    {
      "epoch": 2.986175115207373,
      "grad_norm": 0.48505546530830695,
      "learning_rate": 0.0,
      "loss": 0.2179,
      "step": 216
    },
    {
      "epoch": 2.986175115207373,
      "step": 216,
      "total_flos": 32031101919232.0,
      "train_loss": 0.28372171659160544,
      "train_runtime": 1010.9103,
      "train_samples_per_second": 20.539,
      "train_steps_per_second": 0.214
    }
  ],
  "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": 32031101919232.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}