{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.4681318681318682,
  "eval_steps": 500,
  "global_step": 375,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006593406593406593,
      "grad_norm": 4.499257564544678,
      "learning_rate": 0.0,
      "loss": 5.3313,
      "step": 1
    },
    {
      "epoch": 0.013186813186813187,
      "grad_norm": 3.7660536766052246,
      "learning_rate": 4e-05,
      "loss": 4.7758,
      "step": 2
    },
    {
      "epoch": 0.01978021978021978,
      "grad_norm": 4.873054504394531,
      "learning_rate": 8e-05,
      "loss": 5.0889,
      "step": 3
    },
    {
      "epoch": 0.026373626373626374,
      "grad_norm": 5.257667064666748,
      "learning_rate": 0.00012,
      "loss": 4.6723,
      "step": 4
    },
    {
      "epoch": 0.03296703296703297,
      "grad_norm": 5.000584125518799,
      "learning_rate": 0.00016,
      "loss": 4.7013,
      "step": 5
    },
    {
      "epoch": 0.03956043956043956,
      "grad_norm": 5.8762288093566895,
      "learning_rate": 0.0002,
      "loss": 4.1236,
      "step": 6
    },
    {
      "epoch": 0.046153846153846156,
      "grad_norm": 6.645964622497559,
      "learning_rate": 0.00019955654101995565,
      "loss": 2.4379,
      "step": 7
    },
    {
      "epoch": 0.05274725274725275,
      "grad_norm": 2.519747734069824,
      "learning_rate": 0.00019911308203991133,
      "loss": 1.7819,
      "step": 8
    },
    {
      "epoch": 0.05934065934065934,
      "grad_norm": 1.9227187633514404,
      "learning_rate": 0.00019866962305986697,
      "loss": 1.5306,
      "step": 9
    },
    {
      "epoch": 0.06593406593406594,
      "grad_norm": 1.9636014699935913,
      "learning_rate": 0.00019822616407982261,
      "loss": 1.45,
      "step": 10
    },
    {
      "epoch": 0.07252747252747253,
      "grad_norm": 1.5986920595169067,
      "learning_rate": 0.00019778270509977829,
      "loss": 1.2691,
      "step": 11
    },
    {
      "epoch": 0.07912087912087912,
      "grad_norm": 1.2248263359069824,
      "learning_rate": 0.00019733924611973393,
      "loss": 0.9071,
      "step": 12
    },
    {
      "epoch": 0.08571428571428572,
      "grad_norm": 0.8552874326705933,
      "learning_rate": 0.0001968957871396896,
      "loss": 1.0657,
      "step": 13
    },
    {
      "epoch": 0.09230769230769231,
      "grad_norm": 0.7169967293739319,
      "learning_rate": 0.00019645232815964525,
      "loss": 1.0386,
      "step": 14
    },
    {
      "epoch": 0.0989010989010989,
      "grad_norm": 0.9130010008811951,
      "learning_rate": 0.00019600886917960092,
      "loss": 0.9252,
      "step": 15
    },
    {
      "epoch": 0.1054945054945055,
      "grad_norm": 0.7639235258102417,
      "learning_rate": 0.00019556541019955653,
      "loss": 0.869,
      "step": 16
    },
    {
      "epoch": 0.11208791208791209,
      "grad_norm": 0.6696504950523376,
      "learning_rate": 0.0001951219512195122,
      "loss": 0.8536,
      "step": 17
    },
    {
      "epoch": 0.11868131868131868,
      "grad_norm": 1.052955150604248,
      "learning_rate": 0.00019467849223946785,
      "loss": 0.9968,
      "step": 18
    },
    {
      "epoch": 0.12527472527472527,
      "grad_norm": 0.7062643766403198,
      "learning_rate": 0.00019423503325942352,
      "loss": 1.0926,
      "step": 19
    },
    {
      "epoch": 0.13186813186813187,
      "grad_norm": 0.631159245967865,
      "learning_rate": 0.00019379157427937917,
      "loss": 1.0521,
      "step": 20
    },
    {
      "epoch": 0.13846153846153847,
      "grad_norm": 0.8975602388381958,
      "learning_rate": 0.00019334811529933484,
      "loss": 1.0703,
      "step": 21
    },
    {
      "epoch": 0.14505494505494507,
      "grad_norm": 0.6034598350524902,
      "learning_rate": 0.00019290465631929045,
      "loss": 0.7942,
      "step": 22
    },
    {
      "epoch": 0.15164835164835164,
      "grad_norm": 0.6604612469673157,
      "learning_rate": 0.00019246119733924613,
      "loss": 1.1393,
      "step": 23
    },
    {
      "epoch": 0.15824175824175823,
      "grad_norm": 0.5436087846755981,
      "learning_rate": 0.00019201773835920177,
      "loss": 0.8635,
      "step": 24
    },
    {
      "epoch": 0.16483516483516483,
      "grad_norm": 0.629786491394043,
      "learning_rate": 0.00019157427937915744,
      "loss": 0.9028,
      "step": 25
    },
    {
      "epoch": 0.17142857142857143,
      "grad_norm": 0.6100546717643738,
      "learning_rate": 0.00019113082039911309,
      "loss": 1.021,
      "step": 26
    },
    {
      "epoch": 0.17802197802197803,
      "grad_norm": 0.580397367477417,
      "learning_rate": 0.00019068736141906876,
      "loss": 0.9918,
      "step": 27
    },
    {
      "epoch": 0.18461538461538463,
      "grad_norm": 0.514330267906189,
      "learning_rate": 0.0001902439024390244,
      "loss": 0.9234,
      "step": 28
    },
    {
      "epoch": 0.1912087912087912,
      "grad_norm": 0.48265182971954346,
      "learning_rate": 0.00018980044345898005,
      "loss": 0.8785,
      "step": 29
    },
    {
      "epoch": 0.1978021978021978,
      "grad_norm": 0.5611594319343567,
      "learning_rate": 0.00018935698447893572,
      "loss": 0.8495,
      "step": 30
    },
    {
      "epoch": 0.2043956043956044,
      "grad_norm": 0.5207082629203796,
      "learning_rate": 0.00018891352549889136,
      "loss": 1.0655,
      "step": 31
    },
    {
      "epoch": 0.210989010989011,
      "grad_norm": 0.593443751335144,
      "learning_rate": 0.00018847006651884703,
      "loss": 1.0377,
      "step": 32
    },
    {
      "epoch": 0.2175824175824176,
      "grad_norm": 0.49693596363067627,
      "learning_rate": 0.00018802660753880268,
      "loss": 0.8228,
      "step": 33
    },
    {
      "epoch": 0.22417582417582418,
      "grad_norm": 0.6294080018997192,
      "learning_rate": 0.00018758314855875832,
      "loss": 1.123,
      "step": 34
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 0.5960468053817749,
      "learning_rate": 0.00018713968957871397,
      "loss": 0.8843,
      "step": 35
    },
    {
      "epoch": 0.23736263736263735,
      "grad_norm": 0.5886806845664978,
      "learning_rate": 0.00018669623059866964,
      "loss": 0.7077,
      "step": 36
    },
    {
      "epoch": 0.24395604395604395,
      "grad_norm": 0.7527482509613037,
      "learning_rate": 0.00018625277161862528,
      "loss": 1.1688,
      "step": 37
    },
    {
      "epoch": 0.25054945054945055,
      "grad_norm": 0.5052762031555176,
      "learning_rate": 0.00018580931263858095,
      "loss": 0.9452,
      "step": 38
    },
    {
      "epoch": 0.2571428571428571,
      "grad_norm": 0.6011185646057129,
      "learning_rate": 0.0001853658536585366,
      "loss": 0.9889,
      "step": 39
    },
    {
      "epoch": 0.26373626373626374,
      "grad_norm": 0.4893234074115753,
      "learning_rate": 0.00018492239467849224,
      "loss": 0.9016,
      "step": 40
    },
    {
      "epoch": 0.2703296703296703,
      "grad_norm": 0.5956540703773499,
      "learning_rate": 0.0001844789356984479,
      "loss": 0.8951,
      "step": 41
    },
    {
      "epoch": 0.27692307692307694,
      "grad_norm": 0.6853311657905579,
      "learning_rate": 0.00018403547671840356,
      "loss": 0.8943,
      "step": 42
    },
    {
      "epoch": 0.2835164835164835,
      "grad_norm": 0.514132559299469,
      "learning_rate": 0.0001835920177383592,
      "loss": 1.014,
      "step": 43
    },
    {
      "epoch": 0.29010989010989013,
      "grad_norm": 0.7246399521827698,
      "learning_rate": 0.00018314855875831487,
      "loss": 1.124,
      "step": 44
    },
    {
      "epoch": 0.2967032967032967,
      "grad_norm": 0.534878671169281,
      "learning_rate": 0.00018270509977827052,
      "loss": 0.6777,
      "step": 45
    },
    {
      "epoch": 0.3032967032967033,
      "grad_norm": 0.46303069591522217,
      "learning_rate": 0.00018226164079822616,
      "loss": 0.8077,
      "step": 46
    },
    {
      "epoch": 0.3098901098901099,
      "grad_norm": 0.5512669086456299,
      "learning_rate": 0.00018181818181818183,
      "loss": 0.9223,
      "step": 47
    },
    {
      "epoch": 0.31648351648351647,
      "grad_norm": 0.48670482635498047,
      "learning_rate": 0.00018137472283813748,
      "loss": 0.9329,
      "step": 48
    },
    {
      "epoch": 0.3230769230769231,
      "grad_norm": 0.6091970205307007,
      "learning_rate": 0.00018093126385809312,
      "loss": 1.1785,
      "step": 49
    },
    {
      "epoch": 0.32967032967032966,
      "grad_norm": 0.5351585745811462,
      "learning_rate": 0.0001804878048780488,
      "loss": 1.1025,
      "step": 50
    },
    {
      "epoch": 0.3362637362637363,
      "grad_norm": 0.5463150143623352,
      "learning_rate": 0.00018004434589800444,
      "loss": 0.8758,
      "step": 51
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 0.4296164810657501,
      "learning_rate": 0.00017960088691796008,
      "loss": 0.5897,
      "step": 52
    },
    {
      "epoch": 0.34945054945054943,
      "grad_norm": 0.7088434100151062,
      "learning_rate": 0.00017915742793791575,
      "loss": 0.8874,
      "step": 53
    },
    {
      "epoch": 0.35604395604395606,
      "grad_norm": 0.5368345975875854,
      "learning_rate": 0.0001787139689578714,
      "loss": 0.8522,
      "step": 54
    },
    {
      "epoch": 0.3626373626373626,
      "grad_norm": 0.5056143999099731,
      "learning_rate": 0.00017827050997782707,
      "loss": 0.9154,
      "step": 55
    },
    {
      "epoch": 0.36923076923076925,
      "grad_norm": 0.48936212062835693,
      "learning_rate": 0.00017782705099778271,
      "loss": 0.8478,
      "step": 56
    },
    {
      "epoch": 0.3758241758241758,
      "grad_norm": 0.6096042990684509,
      "learning_rate": 0.00017738359201773839,
      "loss": 1.0707,
      "step": 57
    },
    {
      "epoch": 0.3824175824175824,
      "grad_norm": 0.5620200634002686,
      "learning_rate": 0.000176940133037694,
      "loss": 0.8305,
      "step": 58
    },
    {
      "epoch": 0.389010989010989,
      "grad_norm": 0.5068883895874023,
      "learning_rate": 0.00017649667405764967,
      "loss": 0.7459,
      "step": 59
    },
    {
      "epoch": 0.3956043956043956,
      "grad_norm": 0.609451949596405,
      "learning_rate": 0.00017605321507760532,
      "loss": 0.7703,
      "step": 60
    },
    {
      "epoch": 0.4021978021978022,
      "grad_norm": 0.6350243091583252,
      "learning_rate": 0.000175609756097561,
      "loss": 1.0613,
      "step": 61
    },
    {
      "epoch": 0.4087912087912088,
      "grad_norm": 0.6099206805229187,
      "learning_rate": 0.00017516629711751663,
      "loss": 0.9054,
      "step": 62
    },
    {
      "epoch": 0.4153846153846154,
      "grad_norm": 0.53519207239151,
      "learning_rate": 0.0001747228381374723,
      "loss": 0.7721,
      "step": 63
    },
    {
      "epoch": 0.421978021978022,
      "grad_norm": 0.565395712852478,
      "learning_rate": 0.00017427937915742792,
      "loss": 0.7332,
      "step": 64
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.4922870993614197,
      "learning_rate": 0.0001738359201773836,
      "loss": 0.7914,
      "step": 65
    },
    {
      "epoch": 0.4351648351648352,
      "grad_norm": 0.5346812605857849,
      "learning_rate": 0.00017339246119733924,
      "loss": 0.8879,
      "step": 66
    },
    {
      "epoch": 0.44175824175824174,
      "grad_norm": 0.5556306838989258,
      "learning_rate": 0.0001729490022172949,
      "loss": 0.9712,
      "step": 67
    },
    {
      "epoch": 0.44835164835164837,
      "grad_norm": 0.5152814388275146,
      "learning_rate": 0.00017250554323725056,
      "loss": 0.8266,
      "step": 68
    },
    {
      "epoch": 0.45494505494505494,
      "grad_norm": 0.5951769351959229,
      "learning_rate": 0.00017206208425720623,
      "loss": 0.7834,
      "step": 69
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 0.5551192760467529,
      "learning_rate": 0.00017161862527716187,
      "loss": 0.9461,
      "step": 70
    },
    {
      "epoch": 0.46813186813186813,
      "grad_norm": 0.4997961223125458,
      "learning_rate": 0.00017117516629711752,
      "loss": 0.9438,
      "step": 71
    },
    {
      "epoch": 0.4747252747252747,
      "grad_norm": 0.693430483341217,
      "learning_rate": 0.0001707317073170732,
      "loss": 1.0394,
      "step": 72
    },
    {
      "epoch": 0.48131868131868133,
      "grad_norm": 0.5007423162460327,
      "learning_rate": 0.00017028824833702883,
      "loss": 0.6293,
      "step": 73
    },
    {
      "epoch": 0.4879120879120879,
      "grad_norm": 0.5647709965705872,
      "learning_rate": 0.0001698447893569845,
      "loss": 0.8842,
      "step": 74
    },
    {
      "epoch": 0.4945054945054945,
      "grad_norm": 0.6222750544548035,
      "learning_rate": 0.00016940133037694015,
      "loss": 1.1177,
      "step": 75
    },
    {
      "epoch": 0.5010989010989011,
      "grad_norm": 0.5330225825309753,
      "learning_rate": 0.00016895787139689582,
      "loss": 0.8427,
      "step": 76
    },
    {
      "epoch": 0.5076923076923077,
      "grad_norm": 0.6527089476585388,
      "learning_rate": 0.00016851441241685144,
      "loss": 0.955,
      "step": 77
    },
    {
      "epoch": 0.5142857142857142,
      "grad_norm": 0.705427348613739,
      "learning_rate": 0.0001680709534368071,
      "loss": 1.0531,
      "step": 78
    },
    {
      "epoch": 0.5208791208791209,
      "grad_norm": 0.5138844847679138,
      "learning_rate": 0.00016762749445676275,
      "loss": 0.9054,
      "step": 79
    },
    {
      "epoch": 0.5274725274725275,
      "grad_norm": 0.632585346698761,
      "learning_rate": 0.00016718403547671842,
      "loss": 0.9244,
      "step": 80
    },
    {
      "epoch": 0.5340659340659341,
      "grad_norm": 0.5227172374725342,
      "learning_rate": 0.00016674057649667407,
      "loss": 0.9188,
      "step": 81
    },
    {
      "epoch": 0.5406593406593406,
      "grad_norm": 0.5184674263000488,
      "learning_rate": 0.00016629711751662974,
      "loss": 0.8139,
      "step": 82
    },
    {
      "epoch": 0.5472527472527473,
      "grad_norm": 0.5389203429222107,
      "learning_rate": 0.00016585365853658536,
      "loss": 1.0064,
      "step": 83
    },
    {
      "epoch": 0.5538461538461539,
      "grad_norm": 0.5201700329780579,
      "learning_rate": 0.00016541019955654103,
      "loss": 0.9715,
      "step": 84
    },
    {
      "epoch": 0.5604395604395604,
      "grad_norm": 0.4657474160194397,
      "learning_rate": 0.00016496674057649667,
      "loss": 0.6887,
      "step": 85
    },
    {
      "epoch": 0.567032967032967,
      "grad_norm": 0.48901787400245667,
      "learning_rate": 0.00016452328159645234,
      "loss": 0.9894,
      "step": 86
    },
    {
      "epoch": 0.5736263736263736,
      "grad_norm": 0.5055707693099976,
      "learning_rate": 0.000164079822616408,
      "loss": 0.7828,
      "step": 87
    },
    {
      "epoch": 0.5802197802197803,
      "grad_norm": 0.5926926136016846,
      "learning_rate": 0.00016363636363636366,
      "loss": 0.8319,
      "step": 88
    },
    {
      "epoch": 0.5868131868131868,
      "grad_norm": 0.5132814049720764,
      "learning_rate": 0.0001631929046563193,
      "loss": 0.651,
      "step": 89
    },
    {
      "epoch": 0.5934065934065934,
      "grad_norm": 0.5646933317184448,
      "learning_rate": 0.00016274944567627495,
      "loss": 0.8097,
      "step": 90
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.5312751531600952,
      "learning_rate": 0.0001623059866962306,
      "loss": 0.6701,
      "step": 91
    },
    {
      "epoch": 0.6065934065934065,
      "grad_norm": 0.8451480865478516,
      "learning_rate": 0.00016186252771618626,
      "loss": 0.8902,
      "step": 92
    },
    {
      "epoch": 0.6131868131868132,
      "grad_norm": 0.5425740480422974,
      "learning_rate": 0.0001614190687361419,
      "loss": 0.9255,
      "step": 93
    },
    {
      "epoch": 0.6197802197802198,
      "grad_norm": 0.7311372756958008,
      "learning_rate": 0.00016097560975609758,
      "loss": 1.0323,
      "step": 94
    },
    {
      "epoch": 0.6263736263736264,
      "grad_norm": 0.5717734098434448,
      "learning_rate": 0.00016053215077605322,
      "loss": 0.8789,
      "step": 95
    },
    {
      "epoch": 0.6329670329670329,
      "grad_norm": 0.6015205383300781,
      "learning_rate": 0.00016008869179600887,
      "loss": 0.9006,
      "step": 96
    },
    {
      "epoch": 0.6395604395604395,
      "grad_norm": 0.7425872683525085,
      "learning_rate": 0.00015964523281596454,
      "loss": 0.878,
      "step": 97
    },
    {
      "epoch": 0.6461538461538462,
      "grad_norm": 0.5961865186691284,
      "learning_rate": 0.00015920177383592018,
      "loss": 0.8861,
      "step": 98
    },
    {
      "epoch": 0.6527472527472528,
      "grad_norm": 0.5109598636627197,
      "learning_rate": 0.00015875831485587586,
      "loss": 0.6897,
      "step": 99
    },
    {
      "epoch": 0.6593406593406593,
      "grad_norm": 0.6388728618621826,
      "learning_rate": 0.0001583148558758315,
      "loss": 0.954,
      "step": 100
    },
    {
      "epoch": 0.6659340659340659,
      "grad_norm": 0.46394044160842896,
      "learning_rate": 0.00015787139689578714,
      "loss": 0.7564,
      "step": 101
    },
    {
      "epoch": 0.6725274725274726,
      "grad_norm": 0.5002279281616211,
      "learning_rate": 0.0001574279379157428,
      "loss": 0.8188,
      "step": 102
    },
    {
      "epoch": 0.6791208791208792,
      "grad_norm": 0.5562055110931396,
      "learning_rate": 0.00015698447893569846,
      "loss": 0.7793,
      "step": 103
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 0.5421922206878662,
      "learning_rate": 0.0001565410199556541,
      "loss": 0.7098,
      "step": 104
    },
    {
      "epoch": 0.6923076923076923,
      "grad_norm": 0.5745899677276611,
      "learning_rate": 0.00015609756097560978,
      "loss": 0.9227,
      "step": 105
    },
    {
      "epoch": 0.6989010989010989,
      "grad_norm": 0.5524622797966003,
      "learning_rate": 0.00015565410199556542,
      "loss": 0.8212,
      "step": 106
    },
    {
      "epoch": 0.7054945054945055,
      "grad_norm": 0.5849021077156067,
      "learning_rate": 0.00015521064301552106,
      "loss": 0.8768,
      "step": 107
    },
    {
      "epoch": 0.7120879120879121,
      "grad_norm": 0.5143813490867615,
      "learning_rate": 0.0001547671840354767,
      "loss": 0.8065,
      "step": 108
    },
    {
      "epoch": 0.7186813186813187,
      "grad_norm": 0.5505753755569458,
      "learning_rate": 0.00015432372505543238,
      "loss": 0.8493,
      "step": 109
    },
    {
      "epoch": 0.7252747252747253,
      "grad_norm": 0.5783767700195312,
      "learning_rate": 0.00015388026607538802,
      "loss": 0.9402,
      "step": 110
    },
    {
      "epoch": 0.7318681318681318,
      "grad_norm": 0.5197792053222656,
      "learning_rate": 0.0001534368070953437,
      "loss": 0.7935,
      "step": 111
    },
    {
      "epoch": 0.7384615384615385,
      "grad_norm": 0.6354763507843018,
      "learning_rate": 0.00015299334811529934,
      "loss": 0.9899,
      "step": 112
    },
    {
      "epoch": 0.7450549450549451,
      "grad_norm": 0.5677642226219177,
      "learning_rate": 0.00015254988913525498,
      "loss": 0.8933,
      "step": 113
    },
    {
      "epoch": 0.7516483516483516,
      "grad_norm": 0.5375794768333435,
      "learning_rate": 0.00015210643015521066,
      "loss": 0.772,
      "step": 114
    },
    {
      "epoch": 0.7582417582417582,
      "grad_norm": 0.5195283889770508,
      "learning_rate": 0.0001516629711751663,
      "loss": 0.9625,
      "step": 115
    },
    {
      "epoch": 0.7648351648351648,
      "grad_norm": 0.611674427986145,
      "learning_rate": 0.00015121951219512197,
      "loss": 0.7522,
      "step": 116
    },
    {
      "epoch": 0.7714285714285715,
      "grad_norm": 0.5556056499481201,
      "learning_rate": 0.00015077605321507762,
      "loss": 0.7993,
      "step": 117
    },
    {
      "epoch": 0.778021978021978,
      "grad_norm": 0.5370631814002991,
      "learning_rate": 0.0001503325942350333,
      "loss": 0.9481,
      "step": 118
    },
    {
      "epoch": 0.7846153846153846,
      "grad_norm": 0.6368606686592102,
      "learning_rate": 0.0001498891352549889,
      "loss": 0.9311,
      "step": 119
    },
    {
      "epoch": 0.7912087912087912,
      "grad_norm": 0.8189669251441956,
      "learning_rate": 0.00014944567627494458,
      "loss": 0.833,
      "step": 120
    },
    {
      "epoch": 0.7978021978021979,
      "grad_norm": 0.6427760124206543,
      "learning_rate": 0.00014900221729490022,
      "loss": 0.8768,
      "step": 121
    },
    {
      "epoch": 0.8043956043956044,
      "grad_norm": 0.5934387445449829,
      "learning_rate": 0.0001485587583148559,
      "loss": 0.9282,
      "step": 122
    },
    {
      "epoch": 0.810989010989011,
      "grad_norm": 0.5895951986312866,
      "learning_rate": 0.00014811529933481154,
      "loss": 0.8735,
      "step": 123
    },
    {
      "epoch": 0.8175824175824176,
      "grad_norm": 0.6617897152900696,
      "learning_rate": 0.0001476718403547672,
      "loss": 0.9052,
      "step": 124
    },
    {
      "epoch": 0.8241758241758241,
      "grad_norm": 0.529255211353302,
      "learning_rate": 0.00014722838137472282,
      "loss": 0.6584,
      "step": 125
    },
    {
      "epoch": 0.8307692307692308,
      "grad_norm": 0.728208601474762,
      "learning_rate": 0.0001467849223946785,
      "loss": 0.788,
      "step": 126
    },
    {
      "epoch": 0.8373626373626374,
      "grad_norm": 0.7506310939788818,
      "learning_rate": 0.00014634146341463414,
      "loss": 0.8579,
      "step": 127
    },
    {
      "epoch": 0.843956043956044,
      "grad_norm": 0.6317883133888245,
      "learning_rate": 0.0001458980044345898,
      "loss": 0.8235,
      "step": 128
    },
    {
      "epoch": 0.8505494505494505,
      "grad_norm": 0.5504070520401001,
      "learning_rate": 0.00014545454545454546,
      "loss": 0.8003,
      "step": 129
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.5831419825553894,
      "learning_rate": 0.00014501108647450113,
      "loss": 0.9854,
      "step": 130
    },
    {
      "epoch": 0.8637362637362638,
      "grad_norm": 0.5612383484840393,
      "learning_rate": 0.00014456762749445675,
      "loss": 0.8427,
      "step": 131
    },
    {
      "epoch": 0.8703296703296703,
      "grad_norm": 0.590522825717926,
      "learning_rate": 0.00014412416851441242,
      "loss": 0.8468,
      "step": 132
    },
    {
      "epoch": 0.8769230769230769,
      "grad_norm": 0.6120272278785706,
      "learning_rate": 0.00014368070953436806,
      "loss": 0.7596,
      "step": 133
    },
    {
      "epoch": 0.8835164835164835,
      "grad_norm": 0.5511749982833862,
      "learning_rate": 0.00014323725055432373,
      "loss": 0.8267,
      "step": 134
    },
    {
      "epoch": 0.8901098901098901,
      "grad_norm": 0.6696128249168396,
      "learning_rate": 0.00014279379157427938,
      "loss": 0.9554,
      "step": 135
    },
    {
      "epoch": 0.8967032967032967,
      "grad_norm": 0.5829819440841675,
      "learning_rate": 0.00014235033259423505,
      "loss": 1.0376,
      "step": 136
    },
    {
      "epoch": 0.9032967032967033,
      "grad_norm": 0.5929968953132629,
      "learning_rate": 0.0001419068736141907,
      "loss": 0.9404,
      "step": 137
    },
    {
      "epoch": 0.9098901098901099,
      "grad_norm": 0.5242624878883362,
      "learning_rate": 0.00014146341463414634,
      "loss": 0.6965,
      "step": 138
    },
    {
      "epoch": 0.9164835164835164,
      "grad_norm": 0.6169236302375793,
      "learning_rate": 0.000141019955654102,
      "loss": 0.7462,
      "step": 139
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.6347423195838928,
      "learning_rate": 0.00014057649667405765,
      "loss": 0.7877,
      "step": 140
    },
    {
      "epoch": 0.9296703296703297,
      "grad_norm": 0.5059258341789246,
      "learning_rate": 0.00014013303769401332,
      "loss": 0.7952,
      "step": 141
    },
    {
      "epoch": 0.9362637362637363,
      "grad_norm": 0.5628349781036377,
      "learning_rate": 0.00013968957871396897,
      "loss": 0.8828,
      "step": 142
    },
    {
      "epoch": 0.9428571428571428,
      "grad_norm": 0.6259740591049194,
      "learning_rate": 0.00013924611973392464,
      "loss": 0.8689,
      "step": 143
    },
    {
      "epoch": 0.9494505494505494,
      "grad_norm": 0.6700652241706848,
      "learning_rate": 0.00013880266075388026,
      "loss": 0.938,
      "step": 144
    },
    {
      "epoch": 0.9560439560439561,
      "grad_norm": 0.601922333240509,
      "learning_rate": 0.00013835920177383593,
      "loss": 1.1164,
      "step": 145
    },
    {
      "epoch": 0.9626373626373627,
      "grad_norm": 0.558859646320343,
      "learning_rate": 0.00013791574279379157,
      "loss": 0.7786,
      "step": 146
    },
    {
      "epoch": 0.9692307692307692,
      "grad_norm": 0.5523409843444824,
      "learning_rate": 0.00013747228381374724,
      "loss": 0.8864,
      "step": 147
    },
    {
      "epoch": 0.9758241758241758,
      "grad_norm": 0.5421144962310791,
      "learning_rate": 0.0001370288248337029,
      "loss": 0.8224,
      "step": 148
    },
    {
      "epoch": 0.9824175824175824,
      "grad_norm": 0.634665310382843,
      "learning_rate": 0.00013658536585365856,
      "loss": 1.0744,
      "step": 149
    },
    {
      "epoch": 0.989010989010989,
      "grad_norm": 0.5394982099533081,
      "learning_rate": 0.00013614190687361418,
      "loss": 0.7321,
      "step": 150
    },
    {
      "epoch": 0.9956043956043956,
      "grad_norm": 0.6297814249992371,
      "learning_rate": 0.00013569844789356985,
      "loss": 0.8975,
      "step": 151
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.8643118739128113,
      "learning_rate": 0.0001352549889135255,
      "loss": 1.1409,
      "step": 152
    },
    {
      "epoch": 1.0065934065934066,
      "grad_norm": 0.518811821937561,
      "learning_rate": 0.00013481152993348116,
      "loss": 0.6006,
      "step": 153
    },
    {
      "epoch": 1.0131868131868131,
      "grad_norm": 0.5634834170341492,
      "learning_rate": 0.0001343680709534368,
      "loss": 0.6257,
      "step": 154
    },
    {
      "epoch": 1.0197802197802197,
      "grad_norm": 0.5955131649971008,
      "learning_rate": 0.00013392461197339248,
      "loss": 0.764,
      "step": 155
    },
    {
      "epoch": 1.0263736263736263,
      "grad_norm": 0.510903537273407,
      "learning_rate": 0.00013348115299334812,
      "loss": 0.6491,
      "step": 156
    },
    {
      "epoch": 1.032967032967033,
      "grad_norm": 0.7564712762832642,
      "learning_rate": 0.00013303769401330377,
      "loss": 0.9259,
      "step": 157
    },
    {
      "epoch": 1.0395604395604396,
      "grad_norm": 0.6564409136772156,
      "learning_rate": 0.00013259423503325944,
      "loss": 0.9303,
      "step": 158
    },
    {
      "epoch": 1.0461538461538462,
      "grad_norm": 0.574070394039154,
      "learning_rate": 0.00013215077605321509,
      "loss": 0.7309,
      "step": 159
    },
    {
      "epoch": 1.0527472527472528,
      "grad_norm": 0.49966731667518616,
      "learning_rate": 0.00013170731707317076,
      "loss": 0.6426,
      "step": 160
    },
    {
      "epoch": 1.0593406593406594,
      "grad_norm": 0.4854165315628052,
      "learning_rate": 0.0001312638580931264,
      "loss": 0.6247,
      "step": 161
    },
    {
      "epoch": 1.065934065934066,
      "grad_norm": 0.6286077499389648,
      "learning_rate": 0.00013082039911308205,
      "loss": 0.6843,
      "step": 162
    },
    {
      "epoch": 1.0725274725274725,
      "grad_norm": 0.6135156750679016,
      "learning_rate": 0.0001303769401330377,
      "loss": 0.6935,
      "step": 163
    },
    {
      "epoch": 1.079120879120879,
      "grad_norm": 0.6148243546485901,
      "learning_rate": 0.00012993348115299336,
      "loss": 0.7479,
      "step": 164
    },
    {
      "epoch": 1.0857142857142856,
      "grad_norm": 0.6616637110710144,
      "learning_rate": 0.000129490022172949,
      "loss": 0.7627,
      "step": 165
    },
    {
      "epoch": 1.0923076923076924,
      "grad_norm": 0.5575034618377686,
      "learning_rate": 0.00012904656319290468,
      "loss": 0.6285,
      "step": 166
    },
    {
      "epoch": 1.098901098901099,
      "grad_norm": 0.6019811630249023,
      "learning_rate": 0.00012860310421286032,
      "loss": 0.6534,
      "step": 167
    },
    {
      "epoch": 1.1054945054945056,
      "grad_norm": 0.55308997631073,
      "learning_rate": 0.00012815964523281597,
      "loss": 0.7245,
      "step": 168
    },
    {
      "epoch": 1.1120879120879121,
      "grad_norm": 0.5798131227493286,
      "learning_rate": 0.0001277161862527716,
      "loss": 0.6925,
      "step": 169
    },
    {
      "epoch": 1.1186813186813187,
      "grad_norm": 0.650526762008667,
      "learning_rate": 0.00012727272727272728,
      "loss": 0.7165,
      "step": 170
    },
    {
      "epoch": 1.1252747252747253,
      "grad_norm": 0.6464700102806091,
      "learning_rate": 0.00012682926829268293,
      "loss": 0.619,
      "step": 171
    },
    {
      "epoch": 1.1318681318681318,
      "grad_norm": 0.7274287939071655,
      "learning_rate": 0.0001263858093126386,
      "loss": 0.9021,
      "step": 172
    },
    {
      "epoch": 1.1384615384615384,
      "grad_norm": 0.6489081978797913,
      "learning_rate": 0.00012594235033259424,
      "loss": 0.6478,
      "step": 173
    },
    {
      "epoch": 1.145054945054945,
      "grad_norm": 0.545326292514801,
      "learning_rate": 0.00012549889135254989,
      "loss": 0.5022,
      "step": 174
    },
    {
      "epoch": 1.1516483516483516,
      "grad_norm": 0.5780823230743408,
      "learning_rate": 0.00012505543237250553,
      "loss": 0.7052,
      "step": 175
    },
    {
      "epoch": 1.1582417582417581,
      "grad_norm": 0.7771775722503662,
      "learning_rate": 0.0001246119733924612,
      "loss": 0.8831,
      "step": 176
    },
    {
      "epoch": 1.164835164835165,
      "grad_norm": 0.5964727997779846,
      "learning_rate": 0.00012416851441241685,
      "loss": 0.6699,
      "step": 177
    },
    {
      "epoch": 1.1714285714285715,
      "grad_norm": 0.672161877155304,
      "learning_rate": 0.00012372505543237252,
      "loss": 0.7591,
      "step": 178
    },
    {
      "epoch": 1.178021978021978,
      "grad_norm": 0.7593191266059875,
      "learning_rate": 0.00012328159645232816,
      "loss": 0.8816,
      "step": 179
    },
    {
      "epoch": 1.1846153846153846,
      "grad_norm": 0.6699647903442383,
      "learning_rate": 0.0001228381374722838,
      "loss": 0.5915,
      "step": 180
    },
    {
      "epoch": 1.1912087912087912,
      "grad_norm": 0.6623948812484741,
      "learning_rate": 0.00012239467849223948,
      "loss": 0.7442,
      "step": 181
    },
    {
      "epoch": 1.1978021978021978,
      "grad_norm": 0.631594717502594,
      "learning_rate": 0.00012195121951219512,
      "loss": 0.7496,
      "step": 182
    },
    {
      "epoch": 1.2043956043956043,
      "grad_norm": 0.767998993396759,
      "learning_rate": 0.00012150776053215078,
      "loss": 0.7835,
      "step": 183
    },
    {
      "epoch": 1.210989010989011,
      "grad_norm": 0.7424312233924866,
      "learning_rate": 0.00012106430155210644,
      "loss": 0.7279,
      "step": 184
    },
    {
      "epoch": 1.2175824175824177,
      "grad_norm": 0.664188027381897,
      "learning_rate": 0.0001206208425720621,
      "loss": 0.6582,
      "step": 185
    },
    {
      "epoch": 1.2241758241758243,
      "grad_norm": 0.7374634146690369,
      "learning_rate": 0.00012017738359201774,
      "loss": 1.0711,
      "step": 186
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 0.7106783390045166,
      "learning_rate": 0.0001197339246119734,
      "loss": 0.8675,
      "step": 187
    },
    {
      "epoch": 1.2373626373626374,
      "grad_norm": 0.7353464365005493,
      "learning_rate": 0.00011929046563192906,
      "loss": 0.7289,
      "step": 188
    },
    {
      "epoch": 1.243956043956044,
      "grad_norm": 0.6719170212745667,
      "learning_rate": 0.00011884700665188471,
      "loss": 0.6848,
      "step": 189
    },
    {
      "epoch": 1.2505494505494505,
      "grad_norm": 0.7602320909500122,
      "learning_rate": 0.00011840354767184036,
      "loss": 0.6921,
      "step": 190
    },
    {
      "epoch": 1.2571428571428571,
      "grad_norm": 0.728472888469696,
      "learning_rate": 0.00011796008869179602,
      "loss": 0.6919,
      "step": 191
    },
    {
      "epoch": 1.2637362637362637,
      "grad_norm": 0.6524946093559265,
      "learning_rate": 0.00011751662971175166,
      "loss": 0.6286,
      "step": 192
    },
    {
      "epoch": 1.2703296703296703,
      "grad_norm": 0.7203388810157776,
      "learning_rate": 0.00011707317073170732,
      "loss": 0.7458,
      "step": 193
    },
    {
      "epoch": 1.2769230769230768,
      "grad_norm": 0.8589625954627991,
      "learning_rate": 0.00011662971175166298,
      "loss": 0.6636,
      "step": 194
    },
    {
      "epoch": 1.2835164835164834,
      "grad_norm": 0.6811608672142029,
      "learning_rate": 0.00011618625277161863,
      "loss": 0.6674,
      "step": 195
    },
    {
      "epoch": 1.2901098901098902,
      "grad_norm": 0.7866260409355164,
      "learning_rate": 0.00011574279379157429,
      "loss": 0.6771,
      "step": 196
    },
    {
      "epoch": 1.2967032967032968,
      "grad_norm": 0.7477112412452698,
      "learning_rate": 0.00011529933481152995,
      "loss": 0.7513,
      "step": 197
    },
    {
      "epoch": 1.3032967032967033,
      "grad_norm": 0.7092456817626953,
      "learning_rate": 0.00011485587583148558,
      "loss": 0.5706,
      "step": 198
    },
    {
      "epoch": 1.30989010989011,
      "grad_norm": 0.6812657117843628,
      "learning_rate": 0.00011441241685144124,
      "loss": 0.6004,
      "step": 199
    },
    {
      "epoch": 1.3164835164835165,
      "grad_norm": 0.7204579710960388,
      "learning_rate": 0.0001139689578713969,
      "loss": 0.5873,
      "step": 200
    },
    {
      "epoch": 1.323076923076923,
      "grad_norm": 0.7259777784347534,
      "learning_rate": 0.00011352549889135255,
      "loss": 0.5826,
      "step": 201
    },
    {
      "epoch": 1.3296703296703296,
      "grad_norm": 0.7753502726554871,
      "learning_rate": 0.00011308203991130821,
      "loss": 0.7349,
      "step": 202
    },
    {
      "epoch": 1.3362637362637364,
      "grad_norm": 0.6948927044868469,
      "learning_rate": 0.00011263858093126387,
      "loss": 0.6491,
      "step": 203
    },
    {
      "epoch": 1.342857142857143,
      "grad_norm": 0.7336272597312927,
      "learning_rate": 0.00011219512195121953,
      "loss": 0.5962,
      "step": 204
    },
    {
      "epoch": 1.3494505494505495,
      "grad_norm": 0.8620878458023071,
      "learning_rate": 0.00011175166297117516,
      "loss": 0.7701,
      "step": 205
    },
    {
      "epoch": 1.3560439560439561,
      "grad_norm": 0.7446387410163879,
      "learning_rate": 0.00011130820399113082,
      "loss": 0.6827,
      "step": 206
    },
    {
      "epoch": 1.3626373626373627,
      "grad_norm": 0.8848659992218018,
      "learning_rate": 0.00011086474501108647,
      "loss": 0.8417,
      "step": 207
    },
    {
      "epoch": 1.3692307692307693,
      "grad_norm": 0.7801833748817444,
      "learning_rate": 0.00011042128603104213,
      "loss": 0.8237,
      "step": 208
    },
    {
      "epoch": 1.3758241758241758,
      "grad_norm": 0.7675774097442627,
      "learning_rate": 0.00010997782705099779,
      "loss": 0.6546,
      "step": 209
    },
    {
      "epoch": 1.3824175824175824,
      "grad_norm": 0.8647592067718506,
      "learning_rate": 0.00010953436807095345,
      "loss": 0.5809,
      "step": 210
    },
    {
      "epoch": 1.389010989010989,
      "grad_norm": 0.763778805732727,
      "learning_rate": 0.00010909090909090909,
      "loss": 0.7383,
      "step": 211
    },
    {
      "epoch": 1.3956043956043955,
      "grad_norm": 0.839034914970398,
      "learning_rate": 0.00010864745011086475,
      "loss": 0.6358,
      "step": 212
    },
    {
      "epoch": 1.402197802197802,
      "grad_norm": 0.7566342949867249,
      "learning_rate": 0.00010820399113082041,
      "loss": 0.6745,
      "step": 213
    },
    {
      "epoch": 1.4087912087912087,
      "grad_norm": 0.9809067845344543,
      "learning_rate": 0.00010776053215077607,
      "loss": 0.7146,
      "step": 214
    },
    {
      "epoch": 1.4153846153846155,
      "grad_norm": 0.8211221694946289,
      "learning_rate": 0.00010731707317073172,
      "loss": 0.845,
      "step": 215
    },
    {
      "epoch": 1.421978021978022,
      "grad_norm": 0.7657312750816345,
      "learning_rate": 0.00010687361419068738,
      "loss": 0.6591,
      "step": 216
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.7452254295349121,
      "learning_rate": 0.00010643015521064301,
      "loss": 0.5856,
      "step": 217
    },
    {
      "epoch": 1.4351648351648352,
      "grad_norm": 0.7616993188858032,
      "learning_rate": 0.00010598669623059867,
      "loss": 0.7527,
      "step": 218
    },
    {
      "epoch": 1.4417582417582417,
      "grad_norm": 0.8793402910232544,
      "learning_rate": 0.00010554323725055433,
      "loss": 0.6939,
      "step": 219
    },
    {
      "epoch": 1.4483516483516483,
      "grad_norm": 0.8208599090576172,
      "learning_rate": 0.00010509977827050999,
      "loss": 0.7058,
      "step": 220
    },
    {
      "epoch": 1.4549450549450549,
      "grad_norm": 0.812629222869873,
      "learning_rate": 0.00010465631929046564,
      "loss": 0.6738,
      "step": 221
    },
    {
      "epoch": 1.4615384615384617,
      "grad_norm": 0.9596914649009705,
      "learning_rate": 0.0001042128603104213,
      "loss": 0.7589,
      "step": 222
    },
    {
      "epoch": 1.4681318681318682,
      "grad_norm": 0.7620672583580017,
      "learning_rate": 0.00010376940133037693,
      "loss": 0.5364,
      "step": 223
    },
    {
      "epoch": 1.4747252747252748,
      "grad_norm": 0.9214199185371399,
      "learning_rate": 0.00010332594235033259,
      "loss": 0.6894,
      "step": 224
    },
    {
      "epoch": 1.4813186813186814,
      "grad_norm": 0.8367658853530884,
      "learning_rate": 0.00010288248337028825,
      "loss": 0.5748,
      "step": 225
    },
    {
      "epoch": 1.487912087912088,
      "grad_norm": 0.8402177095413208,
      "learning_rate": 0.0001024390243902439,
      "loss": 0.7785,
      "step": 226
    },
    {
      "epoch": 1.4945054945054945,
      "grad_norm": 0.8243453502655029,
      "learning_rate": 0.00010199556541019956,
      "loss": 0.613,
      "step": 227
    },
    {
      "epoch": 1.501098901098901,
      "grad_norm": 0.9649449586868286,
      "learning_rate": 0.00010155210643015522,
      "loss": 0.8934,
      "step": 228
    },
    {
      "epoch": 1.5076923076923077,
      "grad_norm": 0.8533319234848022,
      "learning_rate": 0.00010110864745011087,
      "loss": 0.8142,
      "step": 229
    },
    {
      "epoch": 1.5142857142857142,
      "grad_norm": 0.8582454919815063,
      "learning_rate": 0.00010066518847006652,
      "loss": 0.911,
      "step": 230
    },
    {
      "epoch": 1.5208791208791208,
      "grad_norm": 0.9017567038536072,
      "learning_rate": 0.00010022172949002218,
      "loss": 0.6821,
      "step": 231
    },
    {
      "epoch": 1.5274725274725274,
      "grad_norm": 0.7732335329055786,
      "learning_rate": 9.977827050997783e-05,
      "loss": 0.6494,
      "step": 232
    },
    {
      "epoch": 1.534065934065934,
      "grad_norm": 0.8369095921516418,
      "learning_rate": 9.933481152993348e-05,
      "loss": 0.7336,
      "step": 233
    },
    {
      "epoch": 1.5406593406593405,
      "grad_norm": 0.8193228244781494,
      "learning_rate": 9.889135254988914e-05,
      "loss": 0.7975,
      "step": 234
    },
    {
      "epoch": 1.5472527472527473,
      "grad_norm": 0.9076023697853088,
      "learning_rate": 9.84478935698448e-05,
      "loss": 0.6208,
      "step": 235
    },
    {
      "epoch": 1.5538461538461539,
      "grad_norm": 0.9679794907569885,
      "learning_rate": 9.800443458980046e-05,
      "loss": 0.7791,
      "step": 236
    },
    {
      "epoch": 1.5604395604395604,
      "grad_norm": 0.9554039835929871,
      "learning_rate": 9.75609756097561e-05,
      "loss": 0.7881,
      "step": 237
    },
    {
      "epoch": 1.567032967032967,
      "grad_norm": 0.8737848997116089,
      "learning_rate": 9.711751662971176e-05,
      "loss": 0.6742,
      "step": 238
    },
    {
      "epoch": 1.5736263736263736,
      "grad_norm": 0.8482589721679688,
      "learning_rate": 9.667405764966742e-05,
      "loss": 0.6059,
      "step": 239
    },
    {
      "epoch": 1.5802197802197804,
      "grad_norm": 0.8119935393333435,
      "learning_rate": 9.623059866962306e-05,
      "loss": 0.7966,
      "step": 240
    },
    {
      "epoch": 1.586813186813187,
      "grad_norm": 0.8265219330787659,
      "learning_rate": 9.578713968957872e-05,
      "loss": 0.731,
      "step": 241
    },
    {
      "epoch": 1.5934065934065935,
      "grad_norm": 0.7965439558029175,
      "learning_rate": 9.534368070953438e-05,
      "loss": 0.6445,
      "step": 242
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.852805495262146,
      "learning_rate": 9.490022172949002e-05,
      "loss": 0.7913,
      "step": 243
    },
    {
      "epoch": 1.6065934065934067,
      "grad_norm": 0.7736412882804871,
      "learning_rate": 9.445676274944568e-05,
      "loss": 0.5741,
      "step": 244
    },
    {
      "epoch": 1.6131868131868132,
      "grad_norm": 0.8931556344032288,
      "learning_rate": 9.401330376940134e-05,
      "loss": 0.59,
      "step": 245
    },
    {
      "epoch": 1.6197802197802198,
      "grad_norm": 1.07955002784729,
      "learning_rate": 9.356984478935698e-05,
      "loss": 0.9203,
      "step": 246
    },
    {
      "epoch": 1.6263736263736264,
      "grad_norm": 0.9306849241256714,
      "learning_rate": 9.312638580931264e-05,
      "loss": 0.6126,
      "step": 247
    },
    {
      "epoch": 1.632967032967033,
      "grad_norm": 0.937964141368866,
      "learning_rate": 9.26829268292683e-05,
      "loss": 0.7288,
      "step": 248
    },
    {
      "epoch": 1.6395604395604395,
      "grad_norm": 0.8694849014282227,
      "learning_rate": 9.223946784922394e-05,
      "loss": 0.7572,
      "step": 249
    },
    {
      "epoch": 1.646153846153846,
      "grad_norm": 0.9174641370773315,
      "learning_rate": 9.17960088691796e-05,
      "loss": 0.6991,
      "step": 250
    },
    {
      "epoch": 1.6527472527472526,
      "grad_norm": 1.1306489706039429,
      "learning_rate": 9.135254988913526e-05,
      "loss": 0.9449,
      "step": 251
    },
    {
      "epoch": 1.6593406593406592,
      "grad_norm": 0.804764986038208,
      "learning_rate": 9.090909090909092e-05,
      "loss": 0.7083,
      "step": 252
    },
    {
      "epoch": 1.6659340659340658,
      "grad_norm": 0.8463274240493774,
      "learning_rate": 9.046563192904656e-05,
      "loss": 0.7701,
      "step": 253
    },
    {
      "epoch": 1.6725274725274726,
      "grad_norm": 0.9322347640991211,
      "learning_rate": 9.002217294900222e-05,
      "loss": 0.8077,
      "step": 254
    },
    {
      "epoch": 1.6791208791208792,
      "grad_norm": 0.871080756187439,
      "learning_rate": 8.957871396895788e-05,
      "loss": 0.7821,
      "step": 255
    },
    {
      "epoch": 1.6857142857142857,
      "grad_norm": 0.944723904132843,
      "learning_rate": 8.913525498891354e-05,
      "loss": 0.6424,
      "step": 256
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 0.8437178730964661,
      "learning_rate": 8.869179600886919e-05,
      "loss": 0.5616,
      "step": 257
    },
    {
      "epoch": 1.6989010989010989,
      "grad_norm": 0.9047809839248657,
      "learning_rate": 8.824833702882484e-05,
      "loss": 0.9232,
      "step": 258
    },
    {
      "epoch": 1.7054945054945057,
      "grad_norm": 1.0128016471862793,
      "learning_rate": 8.78048780487805e-05,
      "loss": 0.6997,
      "step": 259
    },
    {
      "epoch": 1.7120879120879122,
      "grad_norm": 0.8230351805686951,
      "learning_rate": 8.736141906873615e-05,
      "loss": 0.7098,
      "step": 260
    },
    {
      "epoch": 1.7186813186813188,
      "grad_norm": 0.874634861946106,
      "learning_rate": 8.69179600886918e-05,
      "loss": 0.6278,
      "step": 261
    },
    {
      "epoch": 1.7252747252747254,
      "grad_norm": 0.8664326667785645,
      "learning_rate": 8.647450110864746e-05,
      "loss": 0.6348,
      "step": 262
    },
    {
      "epoch": 1.731868131868132,
      "grad_norm": 1.0102225542068481,
      "learning_rate": 8.603104212860311e-05,
      "loss": 0.8396,
      "step": 263
    },
    {
      "epoch": 1.7384615384615385,
      "grad_norm": 0.8338132500648499,
      "learning_rate": 8.558758314855876e-05,
      "loss": 0.7125,
      "step": 264
    },
    {
      "epoch": 1.745054945054945,
      "grad_norm": 0.9371277093887329,
      "learning_rate": 8.514412416851442e-05,
      "loss": 0.7078,
      "step": 265
    },
    {
      "epoch": 1.7516483516483516,
      "grad_norm": 0.872138500213623,
      "learning_rate": 8.470066518847007e-05,
      "loss": 0.6301,
      "step": 266
    },
    {
      "epoch": 1.7582417582417582,
      "grad_norm": 0.8898120522499084,
      "learning_rate": 8.425720620842572e-05,
      "loss": 0.6915,
      "step": 267
    },
    {
      "epoch": 1.7648351648351648,
      "grad_norm": 0.8185475468635559,
      "learning_rate": 8.381374722838138e-05,
      "loss": 0.5814,
      "step": 268
    },
    {
      "epoch": 1.7714285714285714,
      "grad_norm": 0.862663745880127,
      "learning_rate": 8.337028824833703e-05,
      "loss": 0.6591,
      "step": 269
    },
    {
      "epoch": 1.778021978021978,
      "grad_norm": 0.8432035446166992,
      "learning_rate": 8.292682926829268e-05,
      "loss": 0.4927,
      "step": 270
    },
    {
      "epoch": 1.7846153846153845,
      "grad_norm": 0.891116201877594,
      "learning_rate": 8.248337028824834e-05,
      "loss": 0.6042,
      "step": 271
    },
    {
      "epoch": 1.791208791208791,
      "grad_norm": 0.9804494976997375,
      "learning_rate": 8.2039911308204e-05,
      "loss": 0.9575,
      "step": 272
    },
    {
      "epoch": 1.7978021978021979,
      "grad_norm": 0.9405154585838318,
      "learning_rate": 8.159645232815965e-05,
      "loss": 0.6518,
      "step": 273
    },
    {
      "epoch": 1.8043956043956044,
      "grad_norm": 0.8979936242103577,
      "learning_rate": 8.11529933481153e-05,
      "loss": 0.6592,
      "step": 274
    },
    {
      "epoch": 1.810989010989011,
      "grad_norm": 0.9056857824325562,
      "learning_rate": 8.070953436807095e-05,
      "loss": 0.6405,
      "step": 275
    },
    {
      "epoch": 1.8175824175824176,
      "grad_norm": 0.8889809250831604,
      "learning_rate": 8.026607538802661e-05,
      "loss": 0.7113,
      "step": 276
    },
    {
      "epoch": 1.8241758241758241,
      "grad_norm": 0.9943848848342896,
      "learning_rate": 7.982261640798227e-05,
      "loss": 0.7204,
      "step": 277
    },
    {
      "epoch": 1.830769230769231,
      "grad_norm": 1.01982843875885,
      "learning_rate": 7.937915742793793e-05,
      "loss": 0.8851,
      "step": 278
    },
    {
      "epoch": 1.8373626373626375,
      "grad_norm": 0.9351006746292114,
      "learning_rate": 7.893569844789357e-05,
      "loss": 0.5838,
      "step": 279
    },
    {
      "epoch": 1.843956043956044,
      "grad_norm": 1.039788842201233,
      "learning_rate": 7.849223946784923e-05,
      "loss": 0.6112,
      "step": 280
    },
    {
      "epoch": 1.8505494505494506,
      "grad_norm": 0.9234751462936401,
      "learning_rate": 7.804878048780489e-05,
      "loss": 0.5822,
      "step": 281
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 0.8766958117485046,
      "learning_rate": 7.760532150776053e-05,
      "loss": 0.5386,
      "step": 282
    },
    {
      "epoch": 1.8637362637362638,
      "grad_norm": 0.8913479447364807,
      "learning_rate": 7.716186252771619e-05,
      "loss": 0.5775,
      "step": 283
    },
    {
      "epoch": 1.8703296703296703,
      "grad_norm": 0.9081965088844299,
      "learning_rate": 7.671840354767185e-05,
      "loss": 0.6348,
      "step": 284
    },
    {
      "epoch": 1.876923076923077,
      "grad_norm": 0.914999783039093,
      "learning_rate": 7.627494456762749e-05,
      "loss": 0.6788,
      "step": 285
    },
    {
      "epoch": 1.8835164835164835,
      "grad_norm": 0.9330105781555176,
      "learning_rate": 7.583148558758315e-05,
      "loss": 0.7427,
      "step": 286
    },
    {
      "epoch": 1.89010989010989,
      "grad_norm": 0.9035401344299316,
      "learning_rate": 7.538802660753881e-05,
      "loss": 0.6936,
      "step": 287
    },
    {
      "epoch": 1.8967032967032966,
      "grad_norm": 0.9681829810142517,
      "learning_rate": 7.494456762749445e-05,
      "loss": 0.6419,
      "step": 288
    },
    {
      "epoch": 1.9032967032967032,
      "grad_norm": 0.9247081279754639,
      "learning_rate": 7.450110864745011e-05,
      "loss": 0.6097,
      "step": 289
    },
    {
      "epoch": 1.9098901098901098,
      "grad_norm": 0.9844918251037598,
      "learning_rate": 7.405764966740577e-05,
      "loss": 0.7837,
      "step": 290
    },
    {
      "epoch": 1.9164835164835163,
      "grad_norm": 0.8996133208274841,
      "learning_rate": 7.361419068736141e-05,
      "loss": 0.6755,
      "step": 291
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 0.9681469798088074,
      "learning_rate": 7.317073170731707e-05,
      "loss": 0.7231,
      "step": 292
    },
    {
      "epoch": 1.9296703296703297,
      "grad_norm": 0.9642009139060974,
      "learning_rate": 7.272727272727273e-05,
      "loss": 0.6914,
      "step": 293
    },
    {
      "epoch": 1.9362637362637363,
      "grad_norm": 1.0456048250198364,
      "learning_rate": 7.228381374722837e-05,
      "loss": 0.8504,
      "step": 294
    },
    {
      "epoch": 1.9428571428571428,
      "grad_norm": 0.9584895372390747,
      "learning_rate": 7.184035476718403e-05,
      "loss": 0.6677,
      "step": 295
    },
    {
      "epoch": 1.9494505494505494,
      "grad_norm": 0.976585865020752,
      "learning_rate": 7.139689578713969e-05,
      "loss": 0.6948,
      "step": 296
    },
    {
      "epoch": 1.9560439560439562,
      "grad_norm": 1.0638996362686157,
      "learning_rate": 7.095343680709535e-05,
      "loss": 0.7012,
      "step": 297
    },
    {
      "epoch": 1.9626373626373628,
      "grad_norm": 0.7956065535545349,
      "learning_rate": 7.0509977827051e-05,
      "loss": 0.521,
      "step": 298
    },
    {
      "epoch": 1.9692307692307693,
      "grad_norm": 0.9544223546981812,
      "learning_rate": 7.006651884700666e-05,
      "loss": 0.5955,
      "step": 299
    },
    {
      "epoch": 1.975824175824176,
      "grad_norm": 0.9543131589889526,
      "learning_rate": 6.962305986696232e-05,
      "loss": 0.6752,
      "step": 300
    },
    {
      "epoch": 1.9824175824175825,
      "grad_norm": 1.0031977891921997,
      "learning_rate": 6.917960088691796e-05,
      "loss": 0.5938,
      "step": 301
    },
    {
      "epoch": 1.989010989010989,
      "grad_norm": 0.8765499591827393,
      "learning_rate": 6.873614190687362e-05,
      "loss": 0.7803,
      "step": 302
    },
    {
      "epoch": 1.9956043956043956,
      "grad_norm": 0.9553411602973938,
      "learning_rate": 6.829268292682928e-05,
      "loss": 0.739,
      "step": 303
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1459717750549316,
      "learning_rate": 6.784922394678492e-05,
      "loss": 0.6345,
      "step": 304
    },
    {
      "epoch": 2.0065934065934066,
      "grad_norm": 0.9005759358406067,
      "learning_rate": 6.740576496674058e-05,
      "loss": 0.5967,
      "step": 305
    },
    {
      "epoch": 2.013186813186813,
      "grad_norm": 0.8638750314712524,
      "learning_rate": 6.696230598669624e-05,
      "loss": 0.511,
      "step": 306
    },
    {
      "epoch": 2.0197802197802197,
      "grad_norm": 0.8762910962104797,
      "learning_rate": 6.651884700665188e-05,
      "loss": 0.5451,
      "step": 307
    },
    {
      "epoch": 2.0263736263736263,
      "grad_norm": 0.8487597703933716,
      "learning_rate": 6.607538802660754e-05,
      "loss": 0.5027,
      "step": 308
    },
    {
      "epoch": 2.032967032967033,
      "grad_norm": 0.8001495599746704,
      "learning_rate": 6.56319290465632e-05,
      "loss": 0.4955,
      "step": 309
    },
    {
      "epoch": 2.0395604395604394,
      "grad_norm": 0.7598649263381958,
      "learning_rate": 6.518847006651884e-05,
      "loss": 0.3704,
      "step": 310
    },
    {
      "epoch": 2.046153846153846,
      "grad_norm": 0.8900034427642822,
      "learning_rate": 6.47450110864745e-05,
      "loss": 0.4605,
      "step": 311
    },
    {
      "epoch": 2.0527472527472526,
      "grad_norm": 0.9368880391120911,
      "learning_rate": 6.430155210643016e-05,
      "loss": 0.4617,
      "step": 312
    },
    {
      "epoch": 2.059340659340659,
      "grad_norm": 0.9017435908317566,
      "learning_rate": 6.38580931263858e-05,
      "loss": 0.5245,
      "step": 313
    },
    {
      "epoch": 2.065934065934066,
      "grad_norm": 0.8987486362457275,
      "learning_rate": 6.341463414634146e-05,
      "loss": 0.5242,
      "step": 314
    },
    {
      "epoch": 2.0725274725274727,
      "grad_norm": 0.89276123046875,
      "learning_rate": 6.297117516629712e-05,
      "loss": 0.4719,
      "step": 315
    },
    {
      "epoch": 2.0791208791208793,
      "grad_norm": 0.9360655546188354,
      "learning_rate": 6.252771618625277e-05,
      "loss": 0.4863,
      "step": 316
    },
    {
      "epoch": 2.085714285714286,
      "grad_norm": 0.9246019124984741,
      "learning_rate": 6.208425720620842e-05,
      "loss": 0.4416,
      "step": 317
    },
    {
      "epoch": 2.0923076923076924,
      "grad_norm": 0.9819332361221313,
      "learning_rate": 6.164079822616408e-05,
      "loss": 0.7942,
      "step": 318
    },
    {
      "epoch": 2.098901098901099,
      "grad_norm": 1.025396466255188,
      "learning_rate": 6.119733924611974e-05,
      "loss": 0.6224,
      "step": 319
    },
    {
      "epoch": 2.1054945054945056,
      "grad_norm": 0.9883947968482971,
      "learning_rate": 6.075388026607539e-05,
      "loss": 0.5521,
      "step": 320
    },
    {
      "epoch": 2.112087912087912,
      "grad_norm": 1.186385989189148,
      "learning_rate": 6.031042128603105e-05,
      "loss": 0.4641,
      "step": 321
    },
    {
      "epoch": 2.1186813186813187,
      "grad_norm": 0.9412760138511658,
      "learning_rate": 5.98669623059867e-05,
      "loss": 0.4049,
      "step": 322
    },
    {
      "epoch": 2.1252747252747253,
      "grad_norm": 0.9762026071548462,
      "learning_rate": 5.942350332594236e-05,
      "loss": 0.4095,
      "step": 323
    },
    {
      "epoch": 2.131868131868132,
      "grad_norm": 1.0149884223937988,
      "learning_rate": 5.898004434589801e-05,
      "loss": 0.4509,
      "step": 324
    },
    {
      "epoch": 2.1384615384615384,
      "grad_norm": 0.9455697536468506,
      "learning_rate": 5.853658536585366e-05,
      "loss": 0.5114,
      "step": 325
    },
    {
      "epoch": 2.145054945054945,
      "grad_norm": 0.9814666509628296,
      "learning_rate": 5.809312638580932e-05,
      "loss": 0.5122,
      "step": 326
    },
    {
      "epoch": 2.1516483516483516,
      "grad_norm": 1.080235481262207,
      "learning_rate": 5.7649667405764975e-05,
      "loss": 0.4709,
      "step": 327
    },
    {
      "epoch": 2.158241758241758,
      "grad_norm": 1.0514662265777588,
      "learning_rate": 5.720620842572062e-05,
      "loss": 0.5737,
      "step": 328
    },
    {
      "epoch": 2.1648351648351647,
      "grad_norm": 1.1050742864608765,
      "learning_rate": 5.676274944567628e-05,
      "loss": 0.5696,
      "step": 329
    },
    {
      "epoch": 2.1714285714285713,
      "grad_norm": 0.9380549192428589,
      "learning_rate": 5.6319290465631935e-05,
      "loss": 0.3554,
      "step": 330
    },
    {
      "epoch": 2.178021978021978,
      "grad_norm": 1.1426416635513306,
      "learning_rate": 5.587583148558758e-05,
      "loss": 0.6584,
      "step": 331
    },
    {
      "epoch": 2.184615384615385,
      "grad_norm": 1.0272771120071411,
      "learning_rate": 5.543237250554324e-05,
      "loss": 0.4778,
      "step": 332
    },
    {
      "epoch": 2.1912087912087914,
      "grad_norm": 1.1341707706451416,
      "learning_rate": 5.4988913525498895e-05,
      "loss": 0.5255,
      "step": 333
    },
    {
      "epoch": 2.197802197802198,
      "grad_norm": 1.0804990530014038,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 0.4938,
      "step": 334
    },
    {
      "epoch": 2.2043956043956046,
      "grad_norm": 1.0827864408493042,
      "learning_rate": 5.4101995565410204e-05,
      "loss": 0.5047,
      "step": 335
    },
    {
      "epoch": 2.210989010989011,
      "grad_norm": 1.1237510442733765,
      "learning_rate": 5.365853658536586e-05,
      "loss": 0.5028,
      "step": 336
    },
    {
      "epoch": 2.2175824175824177,
      "grad_norm": 1.0361775159835815,
      "learning_rate": 5.3215077605321506e-05,
      "loss": 0.533,
      "step": 337
    },
    {
      "epoch": 2.2241758241758243,
      "grad_norm": 1.0317890644073486,
      "learning_rate": 5.2771618625277164e-05,
      "loss": 0.4797,
      "step": 338
    },
    {
      "epoch": 2.230769230769231,
      "grad_norm": 1.2635154724121094,
      "learning_rate": 5.232815964523282e-05,
      "loss": 0.5563,
      "step": 339
    },
    {
      "epoch": 2.2373626373626374,
      "grad_norm": 1.0624594688415527,
      "learning_rate": 5.1884700665188466e-05,
      "loss": 0.4421,
      "step": 340
    },
    {
      "epoch": 2.243956043956044,
      "grad_norm": 1.240909218788147,
      "learning_rate": 5.1441241685144124e-05,
      "loss": 0.4886,
      "step": 341
    },
    {
      "epoch": 2.2505494505494505,
      "grad_norm": 1.031750202178955,
      "learning_rate": 5.099778270509978e-05,
      "loss": 0.4334,
      "step": 342
    },
    {
      "epoch": 2.257142857142857,
      "grad_norm": 1.0513916015625,
      "learning_rate": 5.0554323725055433e-05,
      "loss": 0.404,
      "step": 343
    },
    {
      "epoch": 2.2637362637362637,
      "grad_norm": 1.4357242584228516,
      "learning_rate": 5.011086474501109e-05,
      "loss": 0.5123,
      "step": 344
    },
    {
      "epoch": 2.2703296703296703,
      "grad_norm": 1.1289379596710205,
      "learning_rate": 4.966740576496674e-05,
      "loss": 0.5256,
      "step": 345
    },
    {
      "epoch": 2.276923076923077,
      "grad_norm": 1.2562658786773682,
      "learning_rate": 4.92239467849224e-05,
      "loss": 0.4841,
      "step": 346
    },
    {
      "epoch": 2.2835164835164834,
      "grad_norm": 0.9977976679801941,
      "learning_rate": 4.878048780487805e-05,
      "loss": 0.3359,
      "step": 347
    },
    {
      "epoch": 2.29010989010989,
      "grad_norm": 1.2712724208831787,
      "learning_rate": 4.833702882483371e-05,
      "loss": 0.5365,
      "step": 348
    },
    {
      "epoch": 2.2967032967032965,
      "grad_norm": 1.1830925941467285,
      "learning_rate": 4.789356984478936e-05,
      "loss": 0.5171,
      "step": 349
    },
    {
      "epoch": 2.303296703296703,
      "grad_norm": 1.27028226852417,
      "learning_rate": 4.745011086474501e-05,
      "loss": 0.4854,
      "step": 350
    },
    {
      "epoch": 2.3098901098901097,
      "grad_norm": 1.1081290245056152,
      "learning_rate": 4.700665188470067e-05,
      "loss": 0.418,
      "step": 351
    },
    {
      "epoch": 2.3164835164835162,
      "grad_norm": 1.3142579793930054,
      "learning_rate": 4.656319290465632e-05,
      "loss": 0.5387,
      "step": 352
    },
    {
      "epoch": 2.3230769230769233,
      "grad_norm": 1.2025885581970215,
      "learning_rate": 4.611973392461197e-05,
      "loss": 0.4615,
      "step": 353
    },
    {
      "epoch": 2.32967032967033,
      "grad_norm": 1.235427975654602,
      "learning_rate": 4.567627494456763e-05,
      "loss": 0.7564,
      "step": 354
    },
    {
      "epoch": 2.3362637362637364,
      "grad_norm": 1.1488170623779297,
      "learning_rate": 4.523281596452328e-05,
      "loss": 0.3861,
      "step": 355
    },
    {
      "epoch": 2.342857142857143,
      "grad_norm": 0.9209251403808594,
      "learning_rate": 4.478935698447894e-05,
      "loss": 0.3882,
      "step": 356
    },
    {
      "epoch": 2.3494505494505495,
      "grad_norm": 1.2640239000320435,
      "learning_rate": 4.4345898004434597e-05,
      "loss": 0.5635,
      "step": 357
    },
    {
      "epoch": 2.356043956043956,
      "grad_norm": 1.2771700620651245,
      "learning_rate": 4.390243902439025e-05,
      "loss": 0.5304,
      "step": 358
    },
    {
      "epoch": 2.3626373626373627,
      "grad_norm": 1.0825189352035522,
      "learning_rate": 4.34589800443459e-05,
      "loss": 0.3598,
      "step": 359
    },
    {
      "epoch": 2.3692307692307693,
      "grad_norm": 1.1906499862670898,
      "learning_rate": 4.301552106430156e-05,
      "loss": 0.5769,
      "step": 360
    },
    {
      "epoch": 2.375824175824176,
      "grad_norm": 1.0458546876907349,
      "learning_rate": 4.257206208425721e-05,
      "loss": 0.4245,
      "step": 361
    },
    {
      "epoch": 2.3824175824175824,
      "grad_norm": 1.178634524345398,
      "learning_rate": 4.212860310421286e-05,
      "loss": 0.5836,
      "step": 362
    },
    {
      "epoch": 2.389010989010989,
      "grad_norm": 1.4680427312850952,
      "learning_rate": 4.168514412416852e-05,
      "loss": 0.5803,
      "step": 363
    },
    {
      "epoch": 2.3956043956043955,
      "grad_norm": 1.1693741083145142,
      "learning_rate": 4.124168514412417e-05,
      "loss": 0.5112,
      "step": 364
    },
    {
      "epoch": 2.402197802197802,
      "grad_norm": 1.0590739250183105,
      "learning_rate": 4.0798226164079826e-05,
      "loss": 0.5013,
      "step": 365
    },
    {
      "epoch": 2.4087912087912087,
      "grad_norm": 1.1774996519088745,
      "learning_rate": 4.035476718403548e-05,
      "loss": 0.5876,
      "step": 366
    },
    {
      "epoch": 2.4153846153846152,
      "grad_norm": 1.1976975202560425,
      "learning_rate": 3.9911308203991135e-05,
      "loss": 0.4861,
      "step": 367
    },
    {
      "epoch": 2.421978021978022,
      "grad_norm": 1.096479892730713,
      "learning_rate": 3.9467849223946786e-05,
      "loss": 0.4427,
      "step": 368
    },
    {
      "epoch": 2.4285714285714284,
      "grad_norm": 1.3175643682479858,
      "learning_rate": 3.9024390243902444e-05,
      "loss": 0.5493,
      "step": 369
    },
    {
      "epoch": 2.4351648351648354,
      "grad_norm": 1.1405586004257202,
      "learning_rate": 3.8580931263858095e-05,
      "loss": 0.4199,
      "step": 370
    },
    {
      "epoch": 2.441758241758242,
      "grad_norm": 1.1859890222549438,
      "learning_rate": 3.8137472283813746e-05,
      "loss": 0.4247,
      "step": 371
    },
    {
      "epoch": 2.4483516483516485,
      "grad_norm": 1.0968153476715088,
      "learning_rate": 3.7694013303769404e-05,
      "loss": 0.3636,
      "step": 372
    },
    {
      "epoch": 2.454945054945055,
      "grad_norm": 1.3394428491592407,
      "learning_rate": 3.7250554323725055e-05,
      "loss": 0.5288,
      "step": 373
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 1.2634552717208862,
      "learning_rate": 3.6807095343680706e-05,
      "loss": 0.6161,
      "step": 374
    },
    {
      "epoch": 2.4681318681318682,
      "grad_norm": 1.2109339237213135,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.5383,
      "step": 375
    }
  ],
  "logging_steps": 1,
  "max_steps": 456,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 15,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.915871761707827e+16,
  "train_batch_size": 22,
  "trial_name": null,
  "trial_params": null
}