{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.998849252013809,
  "eval_steps": 500,
  "global_step": 217,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004602991944764097,
      "grad_norm": 1.9528553485870361,
      "learning_rate": 7.142857142857143e-07,
      "loss": 0.5319,
      "step": 1
    },
    {
      "epoch": 0.009205983889528193,
      "grad_norm": 1.8612775802612305,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 0.5107,
      "step": 2
    },
    {
      "epoch": 0.01380897583429229,
      "grad_norm": 1.9053367376327515,
      "learning_rate": 2.1428571428571427e-06,
      "loss": 0.5124,
      "step": 3
    },
    {
      "epoch": 0.018411967779056387,
      "grad_norm": 1.76083505153656,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.536,
      "step": 4
    },
    {
      "epoch": 0.023014959723820484,
      "grad_norm": 1.5051020383834839,
      "learning_rate": 3.5714285714285718e-06,
      "loss": 0.492,
      "step": 5
    },
    {
      "epoch": 0.02761795166858458,
      "grad_norm": 1.4977186918258667,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 0.5108,
      "step": 6
    },
    {
      "epoch": 0.03222094361334868,
      "grad_norm": 1.1002470254898071,
      "learning_rate": 5e-06,
      "loss": 0.4918,
      "step": 7
    },
    {
      "epoch": 0.03682393555811277,
      "grad_norm": 1.1534695625305176,
      "learning_rate": 4.999720254525684e-06,
      "loss": 0.4705,
      "step": 8
    },
    {
      "epoch": 0.04142692750287687,
      "grad_norm": 1.756179928779602,
      "learning_rate": 4.998881080708759e-06,
      "loss": 0.4622,
      "step": 9
    },
    {
      "epoch": 0.04602991944764097,
      "grad_norm": 1.7668858766555786,
      "learning_rate": 4.997482666353287e-06,
      "loss": 0.4538,
      "step": 10
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 1.736067295074463,
      "learning_rate": 4.995525324419338e-06,
      "loss": 0.4832,
      "step": 11
    },
    {
      "epoch": 0.05523590333716916,
      "grad_norm": 1.1255316734313965,
      "learning_rate": 4.993009492952951e-06,
      "loss": 0.443,
      "step": 12
    },
    {
      "epoch": 0.05983889528193326,
      "grad_norm": 0.9772641658782959,
      "learning_rate": 4.989935734988098e-06,
      "loss": 0.4184,
      "step": 13
    },
    {
      "epoch": 0.06444188722669736,
      "grad_norm": 0.7824466228485107,
      "learning_rate": 4.986304738420684e-06,
      "loss": 0.4166,
      "step": 14
    },
    {
      "epoch": 0.06904487917146145,
      "grad_norm": 0.7207390666007996,
      "learning_rate": 4.982117315854594e-06,
      "loss": 0.4344,
      "step": 15
    },
    {
      "epoch": 0.07364787111622555,
      "grad_norm": 0.7086092829704285,
      "learning_rate": 4.977374404419838e-06,
      "loss": 0.4281,
      "step": 16
    },
    {
      "epoch": 0.07825086306098965,
      "grad_norm": 0.791322648525238,
      "learning_rate": 4.9720770655628216e-06,
      "loss": 0.4428,
      "step": 17
    },
    {
      "epoch": 0.08285385500575373,
      "grad_norm": 0.674933671951294,
      "learning_rate": 4.966226484808804e-06,
      "loss": 0.4129,
      "step": 18
    },
    {
      "epoch": 0.08745684695051784,
      "grad_norm": 0.615633487701416,
      "learning_rate": 4.959823971496575e-06,
      "loss": 0.4362,
      "step": 19
    },
    {
      "epoch": 0.09205983889528194,
      "grad_norm": 0.656779944896698,
      "learning_rate": 4.9528709584854316e-06,
      "loss": 0.3944,
      "step": 20
    },
    {
      "epoch": 0.09666283084004602,
      "grad_norm": 0.6281305551528931,
      "learning_rate": 4.9453690018345144e-06,
      "loss": 0.4247,
      "step": 21
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.6116930842399597,
      "learning_rate": 4.937319780454559e-06,
      "loss": 0.452,
      "step": 22
    },
    {
      "epoch": 0.10586881472957423,
      "grad_norm": 0.5711112022399902,
      "learning_rate": 4.9287250957321685e-06,
      "loss": 0.4176,
      "step": 23
    },
    {
      "epoch": 0.11047180667433831,
      "grad_norm": 0.5996497273445129,
      "learning_rate": 4.919586871126667e-06,
      "loss": 0.4235,
      "step": 24
    },
    {
      "epoch": 0.11507479861910241,
      "grad_norm": 0.6450675129890442,
      "learning_rate": 4.909907151739634e-06,
      "loss": 0.4128,
      "step": 25
    },
    {
      "epoch": 0.11967779056386652,
      "grad_norm": 0.6167646050453186,
      "learning_rate": 4.899688103857223e-06,
      "loss": 0.3971,
      "step": 26
    },
    {
      "epoch": 0.12428078250863062,
      "grad_norm": 0.5306232571601868,
      "learning_rate": 4.8889320144653525e-06,
      "loss": 0.3675,
      "step": 27
    },
    {
      "epoch": 0.12888377445339472,
      "grad_norm": 0.5794817209243774,
      "learning_rate": 4.8776412907378845e-06,
      "loss": 0.4189,
      "step": 28
    },
    {
      "epoch": 0.1334867663981588,
      "grad_norm": 0.5581746697425842,
      "learning_rate": 4.865818459497911e-06,
      "loss": 0.4309,
      "step": 29
    },
    {
      "epoch": 0.1380897583429229,
      "grad_norm": 0.5573146939277649,
      "learning_rate": 4.853466166652259e-06,
      "loss": 0.4159,
      "step": 30
    },
    {
      "epoch": 0.142692750287687,
      "grad_norm": 0.5848842859268188,
      "learning_rate": 4.8405871765993435e-06,
      "loss": 0.4043,
      "step": 31
    },
    {
      "epoch": 0.1472957422324511,
      "grad_norm": 0.5522124767303467,
      "learning_rate": 4.827184371610511e-06,
      "loss": 0.4331,
      "step": 32
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.542629063129425,
      "learning_rate": 4.813260751184992e-06,
      "loss": 0.4027,
      "step": 33
    },
    {
      "epoch": 0.1565017261219793,
      "grad_norm": 0.6027759313583374,
      "learning_rate": 4.7988194313786275e-06,
      "loss": 0.4286,
      "step": 34
    },
    {
      "epoch": 0.1611047180667434,
      "grad_norm": 0.5948846936225891,
      "learning_rate": 4.783863644106502e-06,
      "loss": 0.4334,
      "step": 35
    },
    {
      "epoch": 0.16570771001150747,
      "grad_norm": 0.5355780720710754,
      "learning_rate": 4.7683967364196624e-06,
      "loss": 0.4004,
      "step": 36
    },
    {
      "epoch": 0.17031070195627157,
      "grad_norm": 0.5100039839744568,
      "learning_rate": 4.752422169756048e-06,
      "loss": 0.4058,
      "step": 37
    },
    {
      "epoch": 0.17491369390103567,
      "grad_norm": 0.5452431440353394,
      "learning_rate": 4.735943519165843e-06,
      "loss": 0.4199,
      "step": 38
    },
    {
      "epoch": 0.17951668584579977,
      "grad_norm": 0.5305486917495728,
      "learning_rate": 4.718964472511386e-06,
      "loss": 0.3845,
      "step": 39
    },
    {
      "epoch": 0.18411967779056387,
      "grad_norm": 0.5316929817199707,
      "learning_rate": 4.701488829641845e-06,
      "loss": 0.385,
      "step": 40
    },
    {
      "epoch": 0.18872266973532797,
      "grad_norm": 0.5721151232719421,
      "learning_rate": 4.683520501542825e-06,
      "loss": 0.4199,
      "step": 41
    },
    {
      "epoch": 0.19332566168009205,
      "grad_norm": 0.5623852610588074,
      "learning_rate": 4.665063509461098e-06,
      "loss": 0.393,
      "step": 42
    },
    {
      "epoch": 0.19792865362485615,
      "grad_norm": 0.5426440238952637,
      "learning_rate": 4.646121984004666e-06,
      "loss": 0.4143,
      "step": 43
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.5393860936164856,
      "learning_rate": 4.626700164218349e-06,
      "loss": 0.419,
      "step": 44
    },
    {
      "epoch": 0.20713463751438435,
      "grad_norm": 0.5237355828285217,
      "learning_rate": 4.606802396635098e-06,
      "loss": 0.3802,
      "step": 45
    },
    {
      "epoch": 0.21173762945914845,
      "grad_norm": 0.5436042547225952,
      "learning_rate": 4.586433134303257e-06,
      "loss": 0.4001,
      "step": 46
    },
    {
      "epoch": 0.21634062140391255,
      "grad_norm": 0.5216776132583618,
      "learning_rate": 4.565596935789987e-06,
      "loss": 0.3838,
      "step": 47
    },
    {
      "epoch": 0.22094361334867663,
      "grad_norm": 0.5988544821739197,
      "learning_rate": 4.544298464161079e-06,
      "loss": 0.4193,
      "step": 48
    },
    {
      "epoch": 0.22554660529344073,
      "grad_norm": 0.5307754278182983,
      "learning_rate": 4.522542485937369e-06,
      "loss": 0.4008,
      "step": 49
    },
    {
      "epoch": 0.23014959723820483,
      "grad_norm": 0.532762885093689,
      "learning_rate": 4.500333870028017e-06,
      "loss": 0.4135,
      "step": 50
    },
    {
      "epoch": 0.23475258918296893,
      "grad_norm": 0.5370959639549255,
      "learning_rate": 4.477677586640854e-06,
      "loss": 0.4008,
      "step": 51
    },
    {
      "epoch": 0.23935558112773303,
      "grad_norm": 0.5499489307403564,
      "learning_rate": 4.454578706170075e-06,
      "loss": 0.423,
      "step": 52
    },
    {
      "epoch": 0.24395857307249713,
      "grad_norm": 0.5250299572944641,
      "learning_rate": 4.431042398061499e-06,
      "loss": 0.3918,
      "step": 53
    },
    {
      "epoch": 0.24856156501726123,
      "grad_norm": 0.5530039668083191,
      "learning_rate": 4.4070739296556665e-06,
      "loss": 0.4121,
      "step": 54
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.5588434934616089,
      "learning_rate": 4.382678665009028e-06,
      "loss": 0.4127,
      "step": 55
    },
    {
      "epoch": 0.25776754890678943,
      "grad_norm": 0.5205167531967163,
      "learning_rate": 4.357862063693486e-06,
      "loss": 0.4041,
      "step": 56
    },
    {
      "epoch": 0.26237054085155354,
      "grad_norm": 0.5239887833595276,
      "learning_rate": 4.332629679574566e-06,
      "loss": 0.3856,
      "step": 57
    },
    {
      "epoch": 0.2669735327963176,
      "grad_norm": 0.527583122253418,
      "learning_rate": 4.3069871595684795e-06,
      "loss": 0.3868,
      "step": 58
    },
    {
      "epoch": 0.2715765247410817,
      "grad_norm": 0.5135255455970764,
      "learning_rate": 4.280940242378363e-06,
      "loss": 0.3801,
      "step": 59
    },
    {
      "epoch": 0.2761795166858458,
      "grad_norm": 0.5258612036705017,
      "learning_rate": 4.2544947572099795e-06,
      "loss": 0.3916,
      "step": 60
    },
    {
      "epoch": 0.2807825086306099,
      "grad_norm": 0.5070046186447144,
      "learning_rate": 4.227656622467162e-06,
      "loss": 0.3729,
      "step": 61
    },
    {
      "epoch": 0.285385500575374,
      "grad_norm": 0.5091351866722107,
      "learning_rate": 4.200431844427299e-06,
      "loss": 0.3743,
      "step": 62
    },
    {
      "epoch": 0.2899884925201381,
      "grad_norm": 0.534995436668396,
      "learning_rate": 4.172826515897146e-06,
      "loss": 0.3639,
      "step": 63
    },
    {
      "epoch": 0.2945914844649022,
      "grad_norm": 0.5401779413223267,
      "learning_rate": 4.144846814849282e-06,
      "loss": 0.3813,
      "step": 64
    },
    {
      "epoch": 0.2991944764096663,
      "grad_norm": 0.5409132242202759,
      "learning_rate": 4.116499003039499e-06,
      "loss": 0.3792,
      "step": 65
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.595007061958313,
      "learning_rate": 4.087789424605447e-06,
      "loss": 0.4266,
      "step": 66
    },
    {
      "epoch": 0.3084004602991945,
      "grad_norm": 0.5434048771858215,
      "learning_rate": 4.058724504646834e-06,
      "loss": 0.3893,
      "step": 67
    },
    {
      "epoch": 0.3130034522439586,
      "grad_norm": 0.5291695594787598,
      "learning_rate": 4.029310747787516e-06,
      "loss": 0.3922,
      "step": 68
    },
    {
      "epoch": 0.3176064441887227,
      "grad_norm": 0.4903387129306793,
      "learning_rate": 3.999554736719785e-06,
      "loss": 0.3626,
      "step": 69
    },
    {
      "epoch": 0.3222094361334868,
      "grad_norm": 0.5239253044128418,
      "learning_rate": 3.969463130731183e-06,
      "loss": 0.3702,
      "step": 70
    },
    {
      "epoch": 0.32681242807825084,
      "grad_norm": 0.534202516078949,
      "learning_rate": 3.939042664214185e-06,
      "loss": 0.3826,
      "step": 71
    },
    {
      "epoch": 0.33141542002301494,
      "grad_norm": 0.5841035842895508,
      "learning_rate": 3.908300145159055e-06,
      "loss": 0.4251,
      "step": 72
    },
    {
      "epoch": 0.33601841196777904,
      "grad_norm": 0.5424667000770569,
      "learning_rate": 3.8772424536302565e-06,
      "loss": 0.391,
      "step": 73
    },
    {
      "epoch": 0.34062140391254314,
      "grad_norm": 0.5215575098991394,
      "learning_rate": 3.845876540226707e-06,
      "loss": 0.4103,
      "step": 74
    },
    {
      "epoch": 0.34522439585730724,
      "grad_norm": 0.5513401031494141,
      "learning_rate": 3.8142094245262617e-06,
      "loss": 0.3976,
      "step": 75
    },
    {
      "epoch": 0.34982738780207134,
      "grad_norm": 0.5887762308120728,
      "learning_rate": 3.782248193514766e-06,
      "loss": 0.407,
      "step": 76
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.533482015132904,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.3929,
      "step": 77
    },
    {
      "epoch": 0.35903337169159955,
      "grad_norm": 0.5082557201385498,
      "learning_rate": 3.7174720610109184e-06,
      "loss": 0.3859,
      "step": 78
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.538962721824646,
      "learning_rate": 3.684671656182497e-06,
      "loss": 0.3987,
      "step": 79
    },
    {
      "epoch": 0.36823935558112775,
      "grad_norm": 0.5121095180511475,
      "learning_rate": 3.6516061261265813e-06,
      "loss": 0.3847,
      "step": 80
    },
    {
      "epoch": 0.37284234752589185,
      "grad_norm": 0.5776516199111938,
      "learning_rate": 3.6182828707890816e-06,
      "loss": 0.3771,
      "step": 81
    },
    {
      "epoch": 0.37744533947065595,
      "grad_norm": 0.5664410591125488,
      "learning_rate": 3.5847093477938955e-06,
      "loss": 0.382,
      "step": 82
    },
    {
      "epoch": 0.38204833141542005,
      "grad_norm": 0.5202717781066895,
      "learning_rate": 3.5508930707739143e-06,
      "loss": 0.3608,
      "step": 83
    },
    {
      "epoch": 0.3866513233601841,
      "grad_norm": 0.5437837243080139,
      "learning_rate": 3.516841607689501e-06,
      "loss": 0.4171,
      "step": 84
    },
    {
      "epoch": 0.3912543153049482,
      "grad_norm": 0.5397577881813049,
      "learning_rate": 3.4825625791348093e-06,
      "loss": 0.3906,
      "step": 85
    },
    {
      "epoch": 0.3958573072497123,
      "grad_norm": 0.5708394646644592,
      "learning_rate": 3.4480636566323215e-06,
      "loss": 0.3802,
      "step": 86
    },
    {
      "epoch": 0.4004602991944764,
      "grad_norm": 0.5787578821182251,
      "learning_rate": 3.4133525609159883e-06,
      "loss": 0.3996,
      "step": 87
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.596613347530365,
      "learning_rate": 3.3784370602033572e-06,
      "loss": 0.3938,
      "step": 88
    },
    {
      "epoch": 0.4096662830840046,
      "grad_norm": 0.5240587592124939,
      "learning_rate": 3.3433249684570757e-06,
      "loss": 0.3798,
      "step": 89
    },
    {
      "epoch": 0.4142692750287687,
      "grad_norm": 0.5174093842506409,
      "learning_rate": 3.3080241436361505e-06,
      "loss": 0.3804,
      "step": 90
    },
    {
      "epoch": 0.4188722669735328,
      "grad_norm": 0.5353110432624817,
      "learning_rate": 3.272542485937369e-06,
      "loss": 0.3529,
      "step": 91
    },
    {
      "epoch": 0.4234752589182969,
      "grad_norm": 0.5501390695571899,
      "learning_rate": 3.236887936027261e-06,
      "loss": 0.4038,
      "step": 92
    },
    {
      "epoch": 0.428078250863061,
      "grad_norm": 0.5099984407424927,
      "learning_rate": 3.201068473265007e-06,
      "loss": 0.3755,
      "step": 93
    },
    {
      "epoch": 0.4326812428078251,
      "grad_norm": 0.5678063035011292,
      "learning_rate": 3.165092113916688e-06,
      "loss": 0.386,
      "step": 94
    },
    {
      "epoch": 0.4372842347525892,
      "grad_norm": 0.5758556723594666,
      "learning_rate": 3.128966909361272e-06,
      "loss": 0.413,
      "step": 95
    },
    {
      "epoch": 0.44188722669735325,
      "grad_norm": 0.5315706729888916,
      "learning_rate": 3.092700944288744e-06,
      "loss": 0.3824,
      "step": 96
    },
    {
      "epoch": 0.44649021864211735,
      "grad_norm": 0.5057752728462219,
      "learning_rate": 3.056302334890786e-06,
      "loss": 0.3582,
      "step": 97
    },
    {
      "epoch": 0.45109321058688145,
      "grad_norm": 0.5494191646575928,
      "learning_rate": 3.019779227044398e-06,
      "loss": 0.389,
      "step": 98
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.5015504360198975,
      "learning_rate": 2.9831397944888833e-06,
      "loss": 0.4077,
      "step": 99
    },
    {
      "epoch": 0.46029919447640966,
      "grad_norm": 0.5424397587776184,
      "learning_rate": 2.946392236996592e-06,
      "loss": 0.4167,
      "step": 100
    },
    {
      "epoch": 0.46490218642117376,
      "grad_norm": 0.526132345199585,
      "learning_rate": 2.9095447785378446e-06,
      "loss": 0.3941,
      "step": 101
    },
    {
      "epoch": 0.46950517836593786,
      "grad_norm": 0.5486814379692078,
      "learning_rate": 2.872605665440436e-06,
      "loss": 0.3957,
      "step": 102
    },
    {
      "epoch": 0.47410817031070196,
      "grad_norm": 0.5627570152282715,
      "learning_rate": 2.835583164544139e-06,
      "loss": 0.3898,
      "step": 103
    },
    {
      "epoch": 0.47871116225546606,
      "grad_norm": 0.5162888765335083,
      "learning_rate": 2.7984855613506107e-06,
      "loss": 0.3785,
      "step": 104
    },
    {
      "epoch": 0.48331415420023016,
      "grad_norm": 0.5262818932533264,
      "learning_rate": 2.761321158169134e-06,
      "loss": 0.3767,
      "step": 105
    },
    {
      "epoch": 0.48791714614499426,
      "grad_norm": 0.5049371123313904,
      "learning_rate": 2.724098272258584e-06,
      "loss": 0.3817,
      "step": 106
    },
    {
      "epoch": 0.49252013808975836,
      "grad_norm": 0.5545706748962402,
      "learning_rate": 2.686825233966061e-06,
      "loss": 0.4088,
      "step": 107
    },
    {
      "epoch": 0.49712313003452246,
      "grad_norm": 0.5535576939582825,
      "learning_rate": 2.649510384862586e-06,
      "loss": 0.4014,
      "step": 108
    },
    {
      "epoch": 0.5017261219792866,
      "grad_norm": 0.500004231929779,
      "learning_rate": 2.6121620758762877e-06,
      "loss": 0.3753,
      "step": 109
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.5142732858657837,
      "learning_rate": 2.5747886654234967e-06,
      "loss": 0.3788,
      "step": 110
    },
    {
      "epoch": 0.5109321058688148,
      "grad_norm": 0.5009371042251587,
      "learning_rate": 2.5373985175381595e-06,
      "loss": 0.3713,
      "step": 111
    },
    {
      "epoch": 0.5155350978135789,
      "grad_norm": 0.5388939380645752,
      "learning_rate": 2.5e-06,
      "loss": 0.3991,
      "step": 112
    },
    {
      "epoch": 0.520138089758343,
      "grad_norm": 0.5310930609703064,
      "learning_rate": 2.4626014824618418e-06,
      "loss": 0.3934,
      "step": 113
    },
    {
      "epoch": 0.5247410817031071,
      "grad_norm": 0.522955596446991,
      "learning_rate": 2.4252113345765045e-06,
      "loss": 0.384,
      "step": 114
    },
    {
      "epoch": 0.5293440736478712,
      "grad_norm": 0.5394883751869202,
      "learning_rate": 2.3878379241237136e-06,
      "loss": 0.4068,
      "step": 115
    },
    {
      "epoch": 0.5339470655926352,
      "grad_norm": 0.5207949876785278,
      "learning_rate": 2.3504896151374145e-06,
      "loss": 0.3903,
      "step": 116
    },
    {
      "epoch": 0.5385500575373993,
      "grad_norm": 0.544926643371582,
      "learning_rate": 2.3131747660339396e-06,
      "loss": 0.3914,
      "step": 117
    },
    {
      "epoch": 0.5431530494821634,
      "grad_norm": 0.5148684978485107,
      "learning_rate": 2.2759017277414165e-06,
      "loss": 0.3669,
      "step": 118
    },
    {
      "epoch": 0.5477560414269275,
      "grad_norm": 0.54156893491745,
      "learning_rate": 2.238678841830867e-06,
      "loss": 0.404,
      "step": 119
    },
    {
      "epoch": 0.5523590333716916,
      "grad_norm": 0.5111318230628967,
      "learning_rate": 2.2015144386493898e-06,
      "loss": 0.3735,
      "step": 120
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.5289490818977356,
      "learning_rate": 2.1644168354558623e-06,
      "loss": 0.3711,
      "step": 121
    },
    {
      "epoch": 0.5615650172612198,
      "grad_norm": 0.5365771055221558,
      "learning_rate": 2.1273943345595637e-06,
      "loss": 0.4074,
      "step": 122
    },
    {
      "epoch": 0.5661680092059839,
      "grad_norm": 0.5566303730010986,
      "learning_rate": 2.090455221462156e-06,
      "loss": 0.4354,
      "step": 123
    },
    {
      "epoch": 0.570771001150748,
      "grad_norm": 0.5099798440933228,
      "learning_rate": 2.053607763003409e-06,
      "loss": 0.3841,
      "step": 124
    },
    {
      "epoch": 0.5753739930955121,
      "grad_norm": 0.5596857070922852,
      "learning_rate": 2.0168602055111175e-06,
      "loss": 0.3855,
      "step": 125
    },
    {
      "epoch": 0.5799769850402762,
      "grad_norm": 0.5068130493164062,
      "learning_rate": 1.9802207729556023e-06,
      "loss": 0.3629,
      "step": 126
    },
    {
      "epoch": 0.5845799769850403,
      "grad_norm": 0.5307357907295227,
      "learning_rate": 1.9436976651092143e-06,
      "loss": 0.3741,
      "step": 127
    },
    {
      "epoch": 0.5891829689298044,
      "grad_norm": 0.5442883372306824,
      "learning_rate": 1.9072990557112567e-06,
      "loss": 0.3913,
      "step": 128
    },
    {
      "epoch": 0.5937859608745685,
      "grad_norm": 0.5069390535354614,
      "learning_rate": 1.8710330906387288e-06,
      "loss": 0.3686,
      "step": 129
    },
    {
      "epoch": 0.5983889528193326,
      "grad_norm": 0.5523373484611511,
      "learning_rate": 1.8349078860833125e-06,
      "loss": 0.3757,
      "step": 130
    },
    {
      "epoch": 0.6029919447640967,
      "grad_norm": 0.5090730786323547,
      "learning_rate": 1.7989315267349936e-06,
      "loss": 0.3639,
      "step": 131
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.5404998660087585,
      "learning_rate": 1.7631120639727396e-06,
      "loss": 0.3849,
      "step": 132
    },
    {
      "epoch": 0.6121979286536249,
      "grad_norm": 0.5424637794494629,
      "learning_rate": 1.7274575140626318e-06,
      "loss": 0.3775,
      "step": 133
    },
    {
      "epoch": 0.616800920598389,
      "grad_norm": 0.5275416374206543,
      "learning_rate": 1.6919758563638506e-06,
      "loss": 0.3715,
      "step": 134
    },
    {
      "epoch": 0.6214039125431531,
      "grad_norm": 0.5071358680725098,
      "learning_rate": 1.6566750315429254e-06,
      "loss": 0.3685,
      "step": 135
    },
    {
      "epoch": 0.6260069044879172,
      "grad_norm": 0.4969387352466583,
      "learning_rate": 1.6215629397966432e-06,
      "loss": 0.3526,
      "step": 136
    },
    {
      "epoch": 0.6306098964326813,
      "grad_norm": 0.5176503658294678,
      "learning_rate": 1.5866474390840126e-06,
      "loss": 0.3862,
      "step": 137
    },
    {
      "epoch": 0.6352128883774454,
      "grad_norm": 0.5372660160064697,
      "learning_rate": 1.5519363433676794e-06,
      "loss": 0.3885,
      "step": 138
    },
    {
      "epoch": 0.6398158803222095,
      "grad_norm": 0.5269474387168884,
      "learning_rate": 1.5174374208651913e-06,
      "loss": 0.3759,
      "step": 139
    },
    {
      "epoch": 0.6444188722669736,
      "grad_norm": 0.5645244121551514,
      "learning_rate": 1.4831583923105e-06,
      "loss": 0.4084,
      "step": 140
    },
    {
      "epoch": 0.6490218642117376,
      "grad_norm": 0.5266773104667664,
      "learning_rate": 1.4491069292260867e-06,
      "loss": 0.3878,
      "step": 141
    },
    {
      "epoch": 0.6536248561565017,
      "grad_norm": 0.5369369983673096,
      "learning_rate": 1.415290652206105e-06,
      "loss": 0.4166,
      "step": 142
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.5402231216430664,
      "learning_rate": 1.3817171292109182e-06,
      "loss": 0.3803,
      "step": 143
    },
    {
      "epoch": 0.6628308400460299,
      "grad_norm": 0.5421996116638184,
      "learning_rate": 1.3483938738734197e-06,
      "loss": 0.3696,
      "step": 144
    },
    {
      "epoch": 0.667433831990794,
      "grad_norm": 0.5073113441467285,
      "learning_rate": 1.3153283438175036e-06,
      "loss": 0.3721,
      "step": 145
    },
    {
      "epoch": 0.6720368239355581,
      "grad_norm": 0.5329331755638123,
      "learning_rate": 1.2825279389890818e-06,
      "loss": 0.3657,
      "step": 146
    },
    {
      "epoch": 0.6766398158803222,
      "grad_norm": 0.5001095533370972,
      "learning_rate": 1.2500000000000007e-06,
      "loss": 0.3682,
      "step": 147
    },
    {
      "epoch": 0.6812428078250863,
      "grad_norm": 0.5310607552528381,
      "learning_rate": 1.2177518064852348e-06,
      "loss": 0.392,
      "step": 148
    },
    {
      "epoch": 0.6858457997698504,
      "grad_norm": 0.5475272536277771,
      "learning_rate": 1.185790575473738e-06,
      "loss": 0.4002,
      "step": 149
    },
    {
      "epoch": 0.6904487917146145,
      "grad_norm": 0.5163525342941284,
      "learning_rate": 1.1541234597732947e-06,
      "loss": 0.3925,
      "step": 150
    },
    {
      "epoch": 0.6950517836593786,
      "grad_norm": 0.4944920539855957,
      "learning_rate": 1.122757546369744e-06,
      "loss": 0.3788,
      "step": 151
    },
    {
      "epoch": 0.6996547756041427,
      "grad_norm": 0.5034787058830261,
      "learning_rate": 1.0916998548409449e-06,
      "loss": 0.3696,
      "step": 152
    },
    {
      "epoch": 0.7042577675489068,
      "grad_norm": 0.5238296985626221,
      "learning_rate": 1.0609573357858166e-06,
      "loss": 0.3842,
      "step": 153
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.5249406099319458,
      "learning_rate": 1.0305368692688175e-06,
      "loss": 0.3879,
      "step": 154
    },
    {
      "epoch": 0.713463751438435,
      "grad_norm": 0.5139413475990295,
      "learning_rate": 1.0004452632802158e-06,
      "loss": 0.3795,
      "step": 155
    },
    {
      "epoch": 0.7180667433831991,
      "grad_norm": 0.49963629245758057,
      "learning_rate": 9.70689252212484e-07,
      "loss": 0.3702,
      "step": 156
    },
    {
      "epoch": 0.7226697353279632,
      "grad_norm": 0.5272103548049927,
      "learning_rate": 9.412754953531664e-07,
      "loss": 0.3726,
      "step": 157
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.4988826513290405,
      "learning_rate": 9.122105753945532e-07,
      "loss": 0.3764,
      "step": 158
    },
    {
      "epoch": 0.7318757192174914,
      "grad_norm": 0.542015552520752,
      "learning_rate": 8.835009969605013e-07,
      "loss": 0.4064,
      "step": 159
    },
    {
      "epoch": 0.7364787111622555,
      "grad_norm": 0.5293359756469727,
      "learning_rate": 8.551531851507186e-07,
      "loss": 0.3943,
      "step": 160
    },
    {
      "epoch": 0.7410817031070196,
      "grad_norm": 0.4968125820159912,
      "learning_rate": 8.271734841028553e-07,
      "loss": 0.3727,
      "step": 161
    },
    {
      "epoch": 0.7456846950517837,
      "grad_norm": 0.5024369955062866,
      "learning_rate": 7.995681555727011e-07,
      "loss": 0.3721,
      "step": 162
    },
    {
      "epoch": 0.7502876869965478,
      "grad_norm": 0.5272569060325623,
      "learning_rate": 7.723433775328385e-07,
      "loss": 0.4127,
      "step": 163
    },
    {
      "epoch": 0.7548906789413119,
      "grad_norm": 0.5185663104057312,
      "learning_rate": 7.455052427900214e-07,
      "loss": 0.3832,
      "step": 164
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.5261924266815186,
      "learning_rate": 7.190597576216385e-07,
      "loss": 0.3804,
      "step": 165
    },
    {
      "epoch": 0.7640966628308401,
      "grad_norm": 0.5177868008613586,
      "learning_rate": 6.930128404315214e-07,
      "loss": 0.3873,
      "step": 166
    },
    {
      "epoch": 0.7686996547756041,
      "grad_norm": 0.5173397660255432,
      "learning_rate": 6.673703204254348e-07,
      "loss": 0.367,
      "step": 167
    },
    {
      "epoch": 0.7733026467203682,
      "grad_norm": 0.5059581995010376,
      "learning_rate": 6.421379363065142e-07,
      "loss": 0.3738,
      "step": 168
    },
    {
      "epoch": 0.7779056386651323,
      "grad_norm": 0.6127777695655823,
      "learning_rate": 6.17321334990973e-07,
      "loss": 0.3857,
      "step": 169
    },
    {
      "epoch": 0.7825086306098964,
      "grad_norm": 0.5311411023139954,
      "learning_rate": 5.929260703443337e-07,
      "loss": 0.3903,
      "step": 170
    },
    {
      "epoch": 0.7871116225546605,
      "grad_norm": 0.5173105597496033,
      "learning_rate": 5.689576019385015e-07,
      "loss": 0.3747,
      "step": 171
    },
    {
      "epoch": 0.7917146144994246,
      "grad_norm": 0.5185012817382812,
      "learning_rate": 5.454212938299256e-07,
      "loss": 0.3778,
      "step": 172
    },
    {
      "epoch": 0.7963176064441887,
      "grad_norm": 0.5253241658210754,
      "learning_rate": 5.223224133591475e-07,
      "loss": 0.3962,
      "step": 173
    },
    {
      "epoch": 0.8009205983889528,
      "grad_norm": 0.5264940857887268,
      "learning_rate": 4.996661299719846e-07,
      "loss": 0.3891,
      "step": 174
    },
    {
      "epoch": 0.8055235903337169,
      "grad_norm": 0.5051758289337158,
      "learning_rate": 4.774575140626317e-07,
      "loss": 0.3573,
      "step": 175
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.5200861096382141,
      "learning_rate": 4.5570153583892165e-07,
      "loss": 0.3924,
      "step": 176
    },
    {
      "epoch": 0.8147295742232451,
      "grad_norm": 0.5002381205558777,
      "learning_rate": 4.344030642100133e-07,
      "loss": 0.3797,
      "step": 177
    },
    {
      "epoch": 0.8193325661680092,
      "grad_norm": 0.5222682952880859,
      "learning_rate": 4.1356686569674344e-07,
      "loss": 0.3688,
      "step": 178
    },
    {
      "epoch": 0.8239355581127733,
      "grad_norm": 0.5002561807632446,
      "learning_rate": 3.931976033649021e-07,
      "loss": 0.3858,
      "step": 179
    },
    {
      "epoch": 0.8285385500575374,
      "grad_norm": 0.5189615488052368,
      "learning_rate": 3.732998357816514e-07,
      "loss": 0.3782,
      "step": 180
    },
    {
      "epoch": 0.8331415420023015,
      "grad_norm": 0.5137009620666504,
      "learning_rate": 3.538780159953348e-07,
      "loss": 0.3729,
      "step": 181
    },
    {
      "epoch": 0.8377445339470656,
      "grad_norm": 0.533768892288208,
      "learning_rate": 3.3493649053890325e-07,
      "loss": 0.3887,
      "step": 182
    },
    {
      "epoch": 0.8423475258918297,
      "grad_norm": 0.5303562879562378,
      "learning_rate": 3.164794984571759e-07,
      "loss": 0.3734,
      "step": 183
    },
    {
      "epoch": 0.8469505178365938,
      "grad_norm": 0.5147203803062439,
      "learning_rate": 2.98511170358155e-07,
      "loss": 0.3757,
      "step": 184
    },
    {
      "epoch": 0.8515535097813579,
      "grad_norm": 0.487350732088089,
      "learning_rate": 2.810355274886148e-07,
      "loss": 0.3692,
      "step": 185
    },
    {
      "epoch": 0.856156501726122,
      "grad_norm": 0.4838736057281494,
      "learning_rate": 2.6405648083415835e-07,
      "loss": 0.3469,
      "step": 186
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.5008881092071533,
      "learning_rate": 2.4757783024395244e-07,
      "loss": 0.3714,
      "step": 187
    },
    {
      "epoch": 0.8653624856156502,
      "grad_norm": 0.4887051582336426,
      "learning_rate": 2.316032635803378e-07,
      "loss": 0.3681,
      "step": 188
    },
    {
      "epoch": 0.8699654775604143,
      "grad_norm": 0.521045446395874,
      "learning_rate": 2.1613635589349756e-07,
      "loss": 0.3828,
      "step": 189
    },
    {
      "epoch": 0.8745684695051784,
      "grad_norm": 0.5309224724769592,
      "learning_rate": 2.0118056862137358e-07,
      "loss": 0.3827,
      "step": 190
    },
    {
      "epoch": 0.8791714614499425,
      "grad_norm": 0.5095206499099731,
      "learning_rate": 1.8673924881500826e-07,
      "loss": 0.3707,
      "step": 191
    },
    {
      "epoch": 0.8837744533947065,
      "grad_norm": 0.49568819999694824,
      "learning_rate": 1.7281562838948968e-07,
      "loss": 0.3698,
      "step": 192
    },
    {
      "epoch": 0.8883774453394706,
      "grad_norm": 0.48604169487953186,
      "learning_rate": 1.59412823400657e-07,
      "loss": 0.3597,
      "step": 193
    },
    {
      "epoch": 0.8929804372842347,
      "grad_norm": 0.4996773898601532,
      "learning_rate": 1.465338333477423e-07,
      "loss": 0.3825,
      "step": 194
    },
    {
      "epoch": 0.8975834292289988,
      "grad_norm": 0.4933512508869171,
      "learning_rate": 1.3418154050208937e-07,
      "loss": 0.3477,
      "step": 195
    },
    {
      "epoch": 0.9021864211737629,
      "grad_norm": 0.5096411108970642,
      "learning_rate": 1.223587092621162e-07,
      "loss": 0.3609,
      "step": 196
    },
    {
      "epoch": 0.906789413118527,
      "grad_norm": 0.5132354497909546,
      "learning_rate": 1.1106798553464804e-07,
      "loss": 0.3804,
      "step": 197
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.5068391561508179,
      "learning_rate": 1.0031189614277765e-07,
      "loss": 0.372,
      "step": 198
    },
    {
      "epoch": 0.9159953970080552,
      "grad_norm": 0.4996238350868225,
      "learning_rate": 9.00928482603669e-08,
      "loss": 0.3798,
      "step": 199
    },
    {
      "epoch": 0.9205983889528193,
      "grad_norm": 0.5047898888587952,
      "learning_rate": 8.041312887333396e-08,
      "loss": 0.372,
      "step": 200
    },
    {
      "epoch": 0.9252013808975834,
      "grad_norm": 0.500909686088562,
      "learning_rate": 7.127490426783124e-08,
      "loss": 0.3909,
      "step": 201
    },
    {
      "epoch": 0.9298043728423475,
      "grad_norm": 0.5062527060508728,
      "learning_rate": 6.268021954544095e-08,
      "loss": 0.3921,
      "step": 202
    },
    {
      "epoch": 0.9344073647871116,
      "grad_norm": 0.5464354753494263,
      "learning_rate": 5.463099816548578e-08,
      "loss": 0.4175,
      "step": 203
    },
    {
      "epoch": 0.9390103567318757,
      "grad_norm": 0.4842507839202881,
      "learning_rate": 4.712904151456865e-08,
      "loss": 0.3589,
      "step": 204
    },
    {
      "epoch": 0.9436133486766398,
      "grad_norm": 0.5234150290489197,
      "learning_rate": 4.017602850342584e-08,
      "loss": 0.3839,
      "step": 205
    },
    {
      "epoch": 0.9482163406214039,
      "grad_norm": 0.5067636370658875,
      "learning_rate": 3.377351519119665e-08,
      "loss": 0.3901,
      "step": 206
    },
    {
      "epoch": 0.952819332566168,
      "grad_norm": 0.48845621943473816,
      "learning_rate": 2.7922934437178695e-08,
      "loss": 0.3694,
      "step": 207
    },
    {
      "epoch": 0.9574223245109321,
      "grad_norm": 0.5254659652709961,
      "learning_rate": 2.262559558016325e-08,
      "loss": 0.3935,
      "step": 208
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.5244017839431763,
      "learning_rate": 1.7882684145406616e-08,
      "loss": 0.3797,
      "step": 209
    },
    {
      "epoch": 0.9666283084004603,
      "grad_norm": 0.5155873894691467,
      "learning_rate": 1.3695261579316776e-08,
      "loss": 0.371,
      "step": 210
    },
    {
      "epoch": 0.9712313003452244,
      "grad_norm": 0.4954945147037506,
      "learning_rate": 1.006426501190233e-08,
      "loss": 0.3671,
      "step": 211
    },
    {
      "epoch": 0.9758342922899885,
      "grad_norm": 0.5454512238502502,
      "learning_rate": 6.990507047049677e-09,
      "loss": 0.3847,
      "step": 212
    },
    {
      "epoch": 0.9804372842347526,
      "grad_norm": 0.5421791076660156,
      "learning_rate": 4.474675580662113e-09,
      "loss": 0.3857,
      "step": 213
    },
    {
      "epoch": 0.9850402761795167,
      "grad_norm": 0.5033833980560303,
      "learning_rate": 2.5173336467135266e-09,
      "loss": 0.3734,
      "step": 214
    },
    {
      "epoch": 0.9896432681242808,
      "grad_norm": 0.5605362057685852,
      "learning_rate": 1.1189192912416935e-09,
      "loss": 0.3936,
      "step": 215
    },
    {
      "epoch": 0.9942462600690449,
      "grad_norm": 0.5034700632095337,
      "learning_rate": 2.797454743164174e-10,
      "loss": 0.3813,
      "step": 216
    },
    {
      "epoch": 0.998849252013809,
      "grad_norm": 0.5011968016624451,
      "learning_rate": 0.0,
      "loss": 0.4095,
      "step": 217
    },
    {
      "epoch": 0.998849252013809,
      "step": 217,
      "total_flos": 2.9080852357930025e+18,
      "train_loss": 0.39522024055230454,
      "train_runtime": 4203.0638,
      "train_samples_per_second": 13.229,
      "train_steps_per_second": 0.052
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 217,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 400,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.9080852357930025e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}