{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 230,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008695652173913044,
      "grad_norm": 17.666513442993164,
      "learning_rate": 0.0,
      "loss": 4.5266,
      "step": 1
    },
    {
      "epoch": 0.017391304347826087,
      "grad_norm": 21.562305450439453,
      "learning_rate": 1e-05,
      "loss": 4.6683,
      "step": 2
    },
    {
      "epoch": 0.02608695652173913,
      "grad_norm": 17.93792152404785,
      "learning_rate": 2e-05,
      "loss": 4.4263,
      "step": 3
    },
    {
      "epoch": 0.034782608695652174,
      "grad_norm": 31.087865829467773,
      "learning_rate": 3e-05,
      "loss": 5.3455,
      "step": 4
    },
    {
      "epoch": 0.043478260869565216,
      "grad_norm": 16.84352684020996,
      "learning_rate": 4e-05,
      "loss": 4.7065,
      "step": 5
    },
    {
      "epoch": 0.05217391304347826,
      "grad_norm": 15.276086807250977,
      "learning_rate": 5e-05,
      "loss": 3.5936,
      "step": 6
    },
    {
      "epoch": 0.06086956521739131,
      "grad_norm": 15.930822372436523,
      "learning_rate": 6e-05,
      "loss": 3.7262,
      "step": 7
    },
    {
      "epoch": 0.06956521739130435,
      "grad_norm": 16.219757080078125,
      "learning_rate": 7e-05,
      "loss": 4.0023,
      "step": 8
    },
    {
      "epoch": 0.0782608695652174,
      "grad_norm": 13.747424125671387,
      "learning_rate": 8e-05,
      "loss": 3.8195,
      "step": 9
    },
    {
      "epoch": 0.08695652173913043,
      "grad_norm": 28.324934005737305,
      "learning_rate": 9e-05,
      "loss": 4.1877,
      "step": 10
    },
    {
      "epoch": 0.09565217391304348,
      "grad_norm": 11.464014053344727,
      "learning_rate": 0.0001,
      "loss": 2.5265,
      "step": 11
    },
    {
      "epoch": 0.10434782608695652,
      "grad_norm": 11.093965530395508,
      "learning_rate": 9.999490215047167e-05,
      "loss": 2.0965,
      "step": 12
    },
    {
      "epoch": 0.11304347826086956,
      "grad_norm": 13.068178176879883,
      "learning_rate": 9.997960964140947e-05,
      "loss": 2.4444,
      "step": 13
    },
    {
      "epoch": 0.12173913043478261,
      "grad_norm": 12.251531600952148,
      "learning_rate": 9.995412559116979e-05,
      "loss": 2.6647,
      "step": 14
    },
    {
      "epoch": 0.13043478260869565,
      "grad_norm": 15.61241340637207,
      "learning_rate": 9.991845519630678e-05,
      "loss": 2.9974,
      "step": 15
    },
    {
      "epoch": 0.1391304347826087,
      "grad_norm": 11.974143981933594,
      "learning_rate": 9.987260573051269e-05,
      "loss": 1.9998,
      "step": 16
    },
    {
      "epoch": 0.14782608695652175,
      "grad_norm": 11.015225410461426,
      "learning_rate": 9.981658654313457e-05,
      "loss": 2.1787,
      "step": 17
    },
    {
      "epoch": 0.1565217391304348,
      "grad_norm": 10.231270790100098,
      "learning_rate": 9.975040905726798e-05,
      "loss": 1.3827,
      "step": 18
    },
    {
      "epoch": 0.16521739130434782,
      "grad_norm": 10.304975509643555,
      "learning_rate": 9.967408676742751e-05,
      "loss": 1.8596,
      "step": 19
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 11.838497161865234,
      "learning_rate": 9.958763523679514e-05,
      "loss": 1.8492,
      "step": 20
    },
    {
      "epoch": 0.1826086956521739,
      "grad_norm": 14.471789360046387,
      "learning_rate": 9.949107209404665e-05,
      "loss": 1.4539,
      "step": 21
    },
    {
      "epoch": 0.19130434782608696,
      "grad_norm": 13.361359596252441,
      "learning_rate": 9.938441702975689e-05,
      "loss": 2.3912,
      "step": 22
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.12480640411377,
      "learning_rate": 9.926769179238466e-05,
      "loss": 1.188,
      "step": 23
    },
    {
      "epoch": 0.20869565217391303,
      "grad_norm": 11.180279731750488,
      "learning_rate": 9.914092018383778e-05,
      "loss": 1.6116,
      "step": 24
    },
    {
      "epoch": 0.21739130434782608,
      "grad_norm": 11.030242919921875,
      "learning_rate": 9.900412805461967e-05,
      "loss": 1.9637,
      "step": 25
    },
    {
      "epoch": 0.22608695652173913,
      "grad_norm": 9.22864818572998,
      "learning_rate": 9.885734329855798e-05,
      "loss": 1.8468,
      "step": 26
    },
    {
      "epoch": 0.23478260869565218,
      "grad_norm": 14.804508209228516,
      "learning_rate": 9.870059584711668e-05,
      "loss": 0.9744,
      "step": 27
    },
    {
      "epoch": 0.24347826086956523,
      "grad_norm": 11.706965446472168,
      "learning_rate": 9.853391766329263e-05,
      "loss": 1.9317,
      "step": 28
    },
    {
      "epoch": 0.25217391304347825,
      "grad_norm": 11.159095764160156,
      "learning_rate": 9.835734273509786e-05,
      "loss": 1.7246,
      "step": 29
    },
    {
      "epoch": 0.2608695652173913,
      "grad_norm": 8.815181732177734,
      "learning_rate": 9.817090706862895e-05,
      "loss": 1.2355,
      "step": 30
    },
    {
      "epoch": 0.26956521739130435,
      "grad_norm": 11.841676712036133,
      "learning_rate": 9.797464868072488e-05,
      "loss": 3.3564,
      "step": 31
    },
    {
      "epoch": 0.2782608695652174,
      "grad_norm": 15.580975532531738,
      "learning_rate": 9.776860759121484e-05,
      "loss": 2.0827,
      "step": 32
    },
    {
      "epoch": 0.28695652173913044,
      "grad_norm": 8.459304809570312,
      "learning_rate": 9.755282581475769e-05,
      "loss": 1.2298,
      "step": 33
    },
    {
      "epoch": 0.2956521739130435,
      "grad_norm": 9.495912551879883,
      "learning_rate": 9.73273473522745e-05,
      "loss": 1.3418,
      "step": 34
    },
    {
      "epoch": 0.30434782608695654,
      "grad_norm": 13.812179565429688,
      "learning_rate": 9.709221818197624e-05,
      "loss": 1.6258,
      "step": 35
    },
    {
      "epoch": 0.3130434782608696,
      "grad_norm": 9.413185119628906,
      "learning_rate": 9.68474862499881e-05,
      "loss": 1.168,
      "step": 36
    },
    {
      "epoch": 0.3217391304347826,
      "grad_norm": 13.937978744506836,
      "learning_rate": 9.659320146057262e-05,
      "loss": 1.4273,
      "step": 37
    },
    {
      "epoch": 0.33043478260869563,
      "grad_norm": 9.947927474975586,
      "learning_rate": 9.632941566595357e-05,
      "loss": 1.3857,
      "step": 38
    },
    {
      "epoch": 0.3391304347826087,
      "grad_norm": 10.392467498779297,
      "learning_rate": 9.60561826557425e-05,
      "loss": 0.7599,
      "step": 39
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 12.23806381225586,
      "learning_rate": 9.577355814597031e-05,
      "loss": 2.0868,
      "step": 40
    },
    {
      "epoch": 0.3565217391304348,
      "grad_norm": 6.938766002655029,
      "learning_rate": 9.548159976772592e-05,
      "loss": 0.4847,
      "step": 41
    },
    {
      "epoch": 0.3652173913043478,
      "grad_norm": 9.958048820495605,
      "learning_rate": 9.518036705540458e-05,
      "loss": 1.609,
      "step": 42
    },
    {
      "epoch": 0.3739130434782609,
      "grad_norm": 11.910273551940918,
      "learning_rate": 9.486992143456792e-05,
      "loss": 0.7905,
      "step": 43
    },
    {
      "epoch": 0.3826086956521739,
      "grad_norm": 8.610260963439941,
      "learning_rate": 9.45503262094184e-05,
      "loss": 1.4258,
      "step": 44
    },
    {
      "epoch": 0.391304347826087,
      "grad_norm": 7.331972122192383,
      "learning_rate": 9.422164654989072e-05,
      "loss": 0.9974,
      "step": 45
    },
    {
      "epoch": 0.4,
      "grad_norm": 8.997504234313965,
      "learning_rate": 9.388394947836279e-05,
      "loss": 1.1141,
      "step": 46
    },
    {
      "epoch": 0.40869565217391307,
      "grad_norm": 8.395123481750488,
      "learning_rate": 9.353730385598887e-05,
      "loss": 0.9686,
      "step": 47
    },
    {
      "epoch": 0.41739130434782606,
      "grad_norm": 8.04129695892334,
      "learning_rate": 9.318178036865785e-05,
      "loss": 0.3619,
      "step": 48
    },
    {
      "epoch": 0.4260869565217391,
      "grad_norm": 5.985868453979492,
      "learning_rate": 9.281745151257946e-05,
      "loss": 0.6158,
      "step": 49
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 12.913836479187012,
      "learning_rate": 9.244439157950114e-05,
      "loss": 1.6131,
      "step": 50
    },
    {
      "epoch": 0.4434782608695652,
      "grad_norm": 7.429251194000244,
      "learning_rate": 9.206267664155907e-05,
      "loss": 0.8433,
      "step": 51
    },
    {
      "epoch": 0.45217391304347826,
      "grad_norm": 8.468223571777344,
      "learning_rate": 9.167238453576589e-05,
      "loss": 0.8761,
      "step": 52
    },
    {
      "epoch": 0.4608695652173913,
      "grad_norm": 10.999699592590332,
      "learning_rate": 9.12735948481387e-05,
      "loss": 1.7577,
      "step": 53
    },
    {
      "epoch": 0.46956521739130436,
      "grad_norm": 8.069478988647461,
      "learning_rate": 9.086638889747035e-05,
      "loss": 1.1008,
      "step": 54
    },
    {
      "epoch": 0.4782608695652174,
      "grad_norm": 9.439266204833984,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.7977,
      "step": 55
    },
    {
      "epoch": 0.48695652173913045,
      "grad_norm": 8.834954261779785,
      "learning_rate": 9.002706204621803e-05,
      "loss": 1.2913,
      "step": 56
    },
    {
      "epoch": 0.4956521739130435,
      "grad_norm": 14.728219032287598,
      "learning_rate": 8.959511229611376e-05,
      "loss": 1.566,
      "step": 57
    },
    {
      "epoch": 0.5043478260869565,
      "grad_norm": 14.64725112915039,
      "learning_rate": 8.915508854902778e-05,
      "loss": 0.7095,
      "step": 58
    },
    {
      "epoch": 0.5130434782608696,
      "grad_norm": 17.302316665649414,
      "learning_rate": 8.870708053195413e-05,
      "loss": 1.1349,
      "step": 59
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 13.19262981414795,
      "learning_rate": 8.825117959999116e-05,
      "loss": 1.0779,
      "step": 60
    },
    {
      "epoch": 0.5304347826086957,
      "grad_norm": 10.02542495727539,
      "learning_rate": 8.778747871771292e-05,
      "loss": 0.7204,
      "step": 61
    },
    {
      "epoch": 0.5391304347826087,
      "grad_norm": 2.038048505783081,
      "learning_rate": 8.731607244021236e-05,
      "loss": 0.0617,
      "step": 62
    },
    {
      "epoch": 0.5478260869565217,
      "grad_norm": 8.415672302246094,
      "learning_rate": 8.683705689382024e-05,
      "loss": 1.199,
      "step": 63
    },
    {
      "epoch": 0.5565217391304348,
      "grad_norm": 10.012858390808105,
      "learning_rate": 8.635052975650369e-05,
      "loss": 1.442,
      "step": 64
    },
    {
      "epoch": 0.5652173913043478,
      "grad_norm": 10.651287078857422,
      "learning_rate": 8.585659023794818e-05,
      "loss": 1.6201,
      "step": 65
    },
    {
      "epoch": 0.5739130434782609,
      "grad_norm": 8.589141845703125,
      "learning_rate": 8.535533905932738e-05,
      "loss": 1.1879,
      "step": 66
    },
    {
      "epoch": 0.5826086956521739,
      "grad_norm": 9.711124420166016,
      "learning_rate": 8.484687843276469e-05,
      "loss": 1.2133,
      "step": 67
    },
    {
      "epoch": 0.591304347826087,
      "grad_norm": 9.756671905517578,
      "learning_rate": 8.433131204049067e-05,
      "loss": 1.1111,
      "step": 68
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.514204025268555,
      "learning_rate": 8.380874501370097e-05,
      "loss": 0.4235,
      "step": 69
    },
    {
      "epoch": 0.6086956521739131,
      "grad_norm": 6.404147148132324,
      "learning_rate": 8.327928391111841e-05,
      "loss": 0.1837,
      "step": 70
    },
    {
      "epoch": 0.6173913043478261,
      "grad_norm": 12.888923645019531,
      "learning_rate": 8.274303669726426e-05,
      "loss": 1.2996,
      "step": 71
    },
    {
      "epoch": 0.6260869565217392,
      "grad_norm": 12.083588600158691,
      "learning_rate": 8.220011272044277e-05,
      "loss": 0.3065,
      "step": 72
    },
    {
      "epoch": 0.6347826086956522,
      "grad_norm": 7.7776665687561035,
      "learning_rate": 8.165062269044353e-05,
      "loss": 0.7543,
      "step": 73
    },
    {
      "epoch": 0.6434782608695652,
      "grad_norm": 8.809155464172363,
      "learning_rate": 8.109467865596612e-05,
      "loss": 1.0967,
      "step": 74
    },
    {
      "epoch": 0.6521739130434783,
      "grad_norm": 18.32766342163086,
      "learning_rate": 8.053239398177191e-05,
      "loss": 2.4623,
      "step": 75
    },
    {
      "epoch": 0.6608695652173913,
      "grad_norm": 12.604337692260742,
      "learning_rate": 7.996388332556735e-05,
      "loss": 1.7208,
      "step": 76
    },
    {
      "epoch": 0.6695652173913044,
      "grad_norm": 12.69120979309082,
      "learning_rate": 7.938926261462366e-05,
      "loss": 0.883,
      "step": 77
    },
    {
      "epoch": 0.6782608695652174,
      "grad_norm": 12.506765365600586,
      "learning_rate": 7.880864902213765e-05,
      "loss": 1.83,
      "step": 78
    },
    {
      "epoch": 0.6869565217391305,
      "grad_norm": 8.688498497009277,
      "learning_rate": 7.822216094333847e-05,
      "loss": 0.9735,
      "step": 79
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 14.368943214416504,
      "learning_rate": 7.762991797134514e-05,
      "loss": 2.193,
      "step": 80
    },
    {
      "epoch": 0.7043478260869566,
      "grad_norm": 6.735222339630127,
      "learning_rate": 7.703204087277988e-05,
      "loss": 0.6682,
      "step": 81
    },
    {
      "epoch": 0.7130434782608696,
      "grad_norm": 15.296740531921387,
      "learning_rate": 7.64286515631421e-05,
      "loss": 0.9763,
      "step": 82
    },
    {
      "epoch": 0.7217391304347827,
      "grad_norm": 16.399547576904297,
      "learning_rate": 7.58198730819481e-05,
      "loss": 2.4849,
      "step": 83
    },
    {
      "epoch": 0.7304347826086957,
      "grad_norm": 11.530505180358887,
      "learning_rate": 7.52058295676416e-05,
      "loss": 1.8405,
      "step": 84
    },
    {
      "epoch": 0.7391304347826086,
      "grad_norm": 7.331994533538818,
      "learning_rate": 7.45866462322802e-05,
      "loss": 0.9336,
      "step": 85
    },
    {
      "epoch": 0.7478260869565218,
      "grad_norm": 20.28060531616211,
      "learning_rate": 7.396244933600285e-05,
      "loss": 1.2999,
      "step": 86
    },
    {
      "epoch": 0.7565217391304347,
      "grad_norm": 8.961277961730957,
      "learning_rate": 7.333336616128369e-05,
      "loss": 1.0094,
      "step": 87
    },
    {
      "epoch": 0.7652173913043478,
      "grad_norm": 7.146012306213379,
      "learning_rate": 7.269952498697734e-05,
      "loss": 0.8734,
      "step": 88
    },
    {
      "epoch": 0.7739130434782608,
      "grad_norm": 12.372821807861328,
      "learning_rate": 7.206105506216106e-05,
      "loss": 1.5859,
      "step": 89
    },
    {
      "epoch": 0.782608695652174,
      "grad_norm": 8.59324836730957,
      "learning_rate": 7.141808657977907e-05,
      "loss": 1.0735,
      "step": 90
    },
    {
      "epoch": 0.7913043478260869,
      "grad_norm": 6.564212799072266,
      "learning_rate": 7.077075065009433e-05,
      "loss": 0.8069,
      "step": 91
    },
    {
      "epoch": 0.8,
      "grad_norm": 10.416309356689453,
      "learning_rate": 7.01191792739534e-05,
      "loss": 1.0411,
      "step": 92
    },
    {
      "epoch": 0.808695652173913,
      "grad_norm": 11.143993377685547,
      "learning_rate": 6.946350531586959e-05,
      "loss": 0.8595,
      "step": 93
    },
    {
      "epoch": 0.8173913043478261,
      "grad_norm": 9.861275672912598,
      "learning_rate": 6.880386247692999e-05,
      "loss": 0.9368,
      "step": 94
    },
    {
      "epoch": 0.8260869565217391,
      "grad_norm": 10.142037391662598,
      "learning_rate": 6.814038526753205e-05,
      "loss": 1.5361,
      "step": 95
    },
    {
      "epoch": 0.8347826086956521,
      "grad_norm": 6.418692588806152,
      "learning_rate": 6.747320897995493e-05,
      "loss": 0.7167,
      "step": 96
    },
    {
      "epoch": 0.8434782608695652,
      "grad_norm": 4.113799571990967,
      "learning_rate": 6.680246966077151e-05,
      "loss": 0.3922,
      "step": 97
    },
    {
      "epoch": 0.8521739130434782,
      "grad_norm": 10.264501571655273,
      "learning_rate": 6.61283040831067e-05,
      "loss": 2.2257,
      "step": 98
    },
    {
      "epoch": 0.8608695652173913,
      "grad_norm": 15.571903228759766,
      "learning_rate": 6.545084971874738e-05,
      "loss": 2.4521,
      "step": 99
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 7.295588970184326,
      "learning_rate": 6.477024471011001e-05,
      "loss": 0.7105,
      "step": 100
    },
    {
      "epoch": 0.8782608695652174,
      "grad_norm": 11.271722793579102,
      "learning_rate": 6.408662784207149e-05,
      "loss": 2.0426,
      "step": 101
    },
    {
      "epoch": 0.8869565217391304,
      "grad_norm": 8.137569427490234,
      "learning_rate": 6.340013851366896e-05,
      "loss": 1.3701,
      "step": 102
    },
    {
      "epoch": 0.8956521739130435,
      "grad_norm": 10.493950843811035,
      "learning_rate": 6.271091670967436e-05,
      "loss": 1.1864,
      "step": 103
    },
    {
      "epoch": 0.9043478260869565,
      "grad_norm": 9.262134552001953,
      "learning_rate": 6.201910297204962e-05,
      "loss": 1.0413,
      "step": 104
    },
    {
      "epoch": 0.9130434782608695,
      "grad_norm": 10.564716339111328,
      "learning_rate": 6.132483837128823e-05,
      "loss": 2.303,
      "step": 105
    },
    {
      "epoch": 0.9217391304347826,
      "grad_norm": 4.90865421295166,
      "learning_rate": 6.062826447764883e-05,
      "loss": 0.3802,
      "step": 106
    },
    {
      "epoch": 0.9304347826086956,
      "grad_norm": 9.297486305236816,
      "learning_rate": 5.992952333228728e-05,
      "loss": 1.2512,
      "step": 107
    },
    {
      "epoch": 0.9391304347826087,
      "grad_norm": 8.179728507995605,
      "learning_rate": 5.9228757418292266e-05,
      "loss": 0.4771,
      "step": 108
    },
    {
      "epoch": 0.9478260869565217,
      "grad_norm": 18.19074249267578,
      "learning_rate": 5.85261096316312e-05,
      "loss": 2.0378,
      "step": 109
    },
    {
      "epoch": 0.9565217391304348,
      "grad_norm": 4.808218955993652,
      "learning_rate": 5.782172325201155e-05,
      "loss": 0.14,
      "step": 110
    },
    {
      "epoch": 0.9652173913043478,
      "grad_norm": 15.154193878173828,
      "learning_rate": 5.7115741913664264e-05,
      "loss": 1.3785,
      "step": 111
    },
    {
      "epoch": 0.9739130434782609,
      "grad_norm": 7.520651340484619,
      "learning_rate": 5.640830957605465e-05,
      "loss": 0.7942,
      "step": 112
    },
    {
      "epoch": 0.9826086956521739,
      "grad_norm": 11.108746528625488,
      "learning_rate": 5.569957049452703e-05,
      "loss": 2.0485,
      "step": 113
    },
    {
      "epoch": 0.991304347826087,
      "grad_norm": 8.641176223754883,
      "learning_rate": 5.4989669190889136e-05,
      "loss": 1.1212,
      "step": 114
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.526795387268066,
      "learning_rate": 5.427875042394199e-05,
      "loss": 0.7003,
      "step": 115
    },
    {
      "epoch": 1.008695652173913,
      "grad_norm": 7.168546676635742,
      "learning_rate": 5.3566959159961615e-05,
      "loss": 1.0922,
      "step": 116
    },
    {
      "epoch": 1.017391304347826,
      "grad_norm": 8.669252395629883,
      "learning_rate": 5.2854440543138406e-05,
      "loss": 0.8577,
      "step": 117
    },
    {
      "epoch": 1.0260869565217392,
      "grad_norm": 7.540567398071289,
      "learning_rate": 5.2141339865980134e-05,
      "loss": 0.7922,
      "step": 118
    },
    {
      "epoch": 1.0347826086956522,
      "grad_norm": 7.116650581359863,
      "learning_rate": 5.142780253968481e-05,
      "loss": 0.7596,
      "step": 119
    },
    {
      "epoch": 1.0434782608695652,
      "grad_norm": 8.778299331665039,
      "learning_rate": 5.0713974064489367e-05,
      "loss": 2.5862,
      "step": 120
    },
    {
      "epoch": 1.0521739130434782,
      "grad_norm": 8.101160049438477,
      "learning_rate": 5e-05,
      "loss": 0.6208,
      "step": 121
    },
    {
      "epoch": 1.0608695652173914,
      "grad_norm": 5.799407958984375,
      "learning_rate": 4.928602593551065e-05,
      "loss": 0.5619,
      "step": 122
    },
    {
      "epoch": 1.0695652173913044,
      "grad_norm": 5.000042915344238,
      "learning_rate": 4.85721974603152e-05,
      "loss": 0.2951,
      "step": 123
    },
    {
      "epoch": 1.0782608695652174,
      "grad_norm": 8.474133491516113,
      "learning_rate": 4.7858660134019884e-05,
      "loss": 1.2103,
      "step": 124
    },
    {
      "epoch": 1.0869565217391304,
      "grad_norm": 10.887007713317871,
      "learning_rate": 4.71455594568616e-05,
      "loss": 1.7797,
      "step": 125
    },
    {
      "epoch": 1.0956521739130434,
      "grad_norm": 7.625034809112549,
      "learning_rate": 4.643304084003839e-05,
      "loss": 0.7967,
      "step": 126
    },
    {
      "epoch": 1.1043478260869566,
      "grad_norm": 8.578575134277344,
      "learning_rate": 4.5721249576058027e-05,
      "loss": 0.6249,
      "step": 127
    },
    {
      "epoch": 1.1130434782608696,
      "grad_norm": 6.662336349487305,
      "learning_rate": 4.501033080911086e-05,
      "loss": 0.637,
      "step": 128
    },
    {
      "epoch": 1.1217391304347826,
      "grad_norm": 8.214466094970703,
      "learning_rate": 4.4300429505472976e-05,
      "loss": 1.2037,
      "step": 129
    },
    {
      "epoch": 1.1304347826086956,
      "grad_norm": 9.812641143798828,
      "learning_rate": 4.359169042394536e-05,
      "loss": 1.0598,
      "step": 130
    },
    {
      "epoch": 1.1391304347826088,
      "grad_norm": 5.729506969451904,
      "learning_rate": 4.288425808633575e-05,
      "loss": 0.4854,
      "step": 131
    },
    {
      "epoch": 1.1478260869565218,
      "grad_norm": 7.675601959228516,
      "learning_rate": 4.2178276747988446e-05,
      "loss": 1.6833,
      "step": 132
    },
    {
      "epoch": 1.1565217391304348,
      "grad_norm": 5.756649494171143,
      "learning_rate": 4.147389036836881e-05,
      "loss": 0.586,
      "step": 133
    },
    {
      "epoch": 1.1652173913043478,
      "grad_norm": 4.710939407348633,
      "learning_rate": 4.077124258170774e-05,
      "loss": 0.1098,
      "step": 134
    },
    {
      "epoch": 1.1739130434782608,
      "grad_norm": 8.076876640319824,
      "learning_rate": 4.007047666771274e-05,
      "loss": 0.8247,
      "step": 135
    },
    {
      "epoch": 1.182608695652174,
      "grad_norm": 6.092112064361572,
      "learning_rate": 3.937173552235117e-05,
      "loss": 0.4976,
      "step": 136
    },
    {
      "epoch": 1.191304347826087,
      "grad_norm": 12.222753524780273,
      "learning_rate": 3.8675161628711776e-05,
      "loss": 0.7614,
      "step": 137
    },
    {
      "epoch": 1.2,
      "grad_norm": 9.492755889892578,
      "learning_rate": 3.798089702795038e-05,
      "loss": 0.7278,
      "step": 138
    },
    {
      "epoch": 1.208695652173913,
      "grad_norm": 8.09307861328125,
      "learning_rate": 3.728908329032567e-05,
      "loss": 1.0424,
      "step": 139
    },
    {
      "epoch": 1.2173913043478262,
      "grad_norm": 13.750467300415039,
      "learning_rate": 3.659986148633107e-05,
      "loss": 1.3154,
      "step": 140
    },
    {
      "epoch": 1.2260869565217392,
      "grad_norm": 5.85883903503418,
      "learning_rate": 3.591337215792852e-05,
      "loss": 0.151,
      "step": 141
    },
    {
      "epoch": 1.2347826086956522,
      "grad_norm": 7.29556131362915,
      "learning_rate": 3.522975528989e-05,
      "loss": 0.4567,
      "step": 142
    },
    {
      "epoch": 1.2434782608695651,
      "grad_norm": 10.308717727661133,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 1.3977,
      "step": 143
    },
    {
      "epoch": 1.2521739130434781,
      "grad_norm": 9.417724609375,
      "learning_rate": 3.3871695916893314e-05,
      "loss": 0.948,
      "step": 144
    },
    {
      "epoch": 1.2608695652173914,
      "grad_norm": 8.931994438171387,
      "learning_rate": 3.3197530339228487e-05,
      "loss": 0.9489,
      "step": 145
    },
    {
      "epoch": 1.2695652173913043,
      "grad_norm": 14.605033874511719,
      "learning_rate": 3.2526791020045086e-05,
      "loss": 1.851,
      "step": 146
    },
    {
      "epoch": 1.2782608695652173,
      "grad_norm": 7.451152324676514,
      "learning_rate": 3.1859614732467954e-05,
      "loss": 0.4653,
      "step": 147
    },
    {
      "epoch": 1.2869565217391306,
      "grad_norm": 7.847959041595459,
      "learning_rate": 3.119613752307002e-05,
      "loss": 0.8261,
      "step": 148
    },
    {
      "epoch": 1.2956521739130435,
      "grad_norm": 6.851174831390381,
      "learning_rate": 3.053649468413043e-05,
      "loss": 0.5726,
      "step": 149
    },
    {
      "epoch": 1.3043478260869565,
      "grad_norm": 7.77811336517334,
      "learning_rate": 2.988082072604661e-05,
      "loss": 0.8105,
      "step": 150
    },
    {
      "epoch": 1.3130434782608695,
      "grad_norm": 9.027288436889648,
      "learning_rate": 2.9229249349905684e-05,
      "loss": 1.3896,
      "step": 151
    },
    {
      "epoch": 1.3217391304347825,
      "grad_norm": 6.307928085327148,
      "learning_rate": 2.858191342022095e-05,
      "loss": 0.4174,
      "step": 152
    },
    {
      "epoch": 1.3304347826086955,
      "grad_norm": 6.78994083404541,
      "learning_rate": 2.7938944937838923e-05,
      "loss": 0.5739,
      "step": 153
    },
    {
      "epoch": 1.3391304347826087,
      "grad_norm": 7.77897834777832,
      "learning_rate": 2.7300475013022663e-05,
      "loss": 0.6822,
      "step": 154
    },
    {
      "epoch": 1.3478260869565217,
      "grad_norm": 10.742020606994629,
      "learning_rate": 2.6666633838716314e-05,
      "loss": 1.1842,
      "step": 155
    },
    {
      "epoch": 1.3565217391304347,
      "grad_norm": 6.034131050109863,
      "learning_rate": 2.603755066399718e-05,
      "loss": 0.1571,
      "step": 156
    },
    {
      "epoch": 1.365217391304348,
      "grad_norm": 8.707799911499023,
      "learning_rate": 2.5413353767719805e-05,
      "loss": 0.9573,
      "step": 157
    },
    {
      "epoch": 1.373913043478261,
      "grad_norm": 5.427333354949951,
      "learning_rate": 2.4794170432358415e-05,
      "loss": 0.4321,
      "step": 158
    },
    {
      "epoch": 1.382608695652174,
      "grad_norm": 18.620357513427734,
      "learning_rate": 2.418012691805191e-05,
      "loss": 2.3272,
      "step": 159
    },
    {
      "epoch": 1.391304347826087,
      "grad_norm": 10.270234107971191,
      "learning_rate": 2.3571348436857904e-05,
      "loss": 0.7804,
      "step": 160
    },
    {
      "epoch": 1.4,
      "grad_norm": 4.281362056732178,
      "learning_rate": 2.296795912722014e-05,
      "loss": 0.1796,
      "step": 161
    },
    {
      "epoch": 1.4086956521739131,
      "grad_norm": 6.379110336303711,
      "learning_rate": 2.2370082028654866e-05,
      "loss": 0.5788,
      "step": 162
    },
    {
      "epoch": 1.4173913043478261,
      "grad_norm": 1.9616364240646362,
      "learning_rate": 2.1777839056661554e-05,
      "loss": 0.0406,
      "step": 163
    },
    {
      "epoch": 1.4260869565217391,
      "grad_norm": 8.34288215637207,
      "learning_rate": 2.119135097786236e-05,
      "loss": 0.914,
      "step": 164
    },
    {
      "epoch": 1.434782608695652,
      "grad_norm": 2.7565298080444336,
      "learning_rate": 2.061073738537635e-05,
      "loss": 0.0697,
      "step": 165
    },
    {
      "epoch": 1.4434782608695653,
      "grad_norm": 12.554067611694336,
      "learning_rate": 2.0036116674432654e-05,
      "loss": 1.2686,
      "step": 166
    },
    {
      "epoch": 1.4521739130434783,
      "grad_norm": 7.41255521774292,
      "learning_rate": 1.946760601822809e-05,
      "loss": 0.6045,
      "step": 167
    },
    {
      "epoch": 1.4608695652173913,
      "grad_norm": 8.661652565002441,
      "learning_rate": 1.8905321344033898e-05,
      "loss": 0.2404,
      "step": 168
    },
    {
      "epoch": 1.4695652173913043,
      "grad_norm": 8.020094871520996,
      "learning_rate": 1.8349377309556486e-05,
      "loss": 0.627,
      "step": 169
    },
    {
      "epoch": 1.4782608695652173,
      "grad_norm": 6.1133222579956055,
      "learning_rate": 1.7799887279557237e-05,
      "loss": 0.6232,
      "step": 170
    },
    {
      "epoch": 1.4869565217391305,
      "grad_norm": 11.473217010498047,
      "learning_rate": 1.725696330273575e-05,
      "loss": 0.8715,
      "step": 171
    },
    {
      "epoch": 1.4956521739130435,
      "grad_norm": 9.100458145141602,
      "learning_rate": 1.6720716088881594e-05,
      "loss": 1.0355,
      "step": 172
    },
    {
      "epoch": 1.5043478260869565,
      "grad_norm": 10.638351440429688,
      "learning_rate": 1.619125498629904e-05,
      "loss": 1.3327,
      "step": 173
    },
    {
      "epoch": 1.5130434782608697,
      "grad_norm": 9.494583129882812,
      "learning_rate": 1.566868795950932e-05,
      "loss": 0.3127,
      "step": 174
    },
    {
      "epoch": 1.5217391304347827,
      "grad_norm": 9.716391563415527,
      "learning_rate": 1.5153121567235335e-05,
      "loss": 0.7739,
      "step": 175
    },
    {
      "epoch": 1.5304347826086957,
      "grad_norm": 8.240947723388672,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.7155,
      "step": 176
    },
    {
      "epoch": 1.5391304347826087,
      "grad_norm": 8.80859661102295,
      "learning_rate": 1.414340976205183e-05,
      "loss": 0.8378,
      "step": 177
    },
    {
      "epoch": 1.5478260869565217,
      "grad_norm": 7.551889896392822,
      "learning_rate": 1.3649470243496326e-05,
      "loss": 0.5837,
      "step": 178
    },
    {
      "epoch": 1.5565217391304347,
      "grad_norm": 7.766860008239746,
      "learning_rate": 1.3162943106179749e-05,
      "loss": 0.6585,
      "step": 179
    },
    {
      "epoch": 1.5652173913043477,
      "grad_norm": 7.3335347175598145,
      "learning_rate": 1.2683927559787655e-05,
      "loss": 0.5556,
      "step": 180
    },
    {
      "epoch": 1.5739130434782609,
      "grad_norm": 3.552537441253662,
      "learning_rate": 1.2212521282287092e-05,
      "loss": 0.0628,
      "step": 181
    },
    {
      "epoch": 1.5826086956521739,
      "grad_norm": 7.300377368927002,
      "learning_rate": 1.1748820400008843e-05,
      "loss": 0.6523,
      "step": 182
    },
    {
      "epoch": 1.591304347826087,
      "grad_norm": 10.045459747314453,
      "learning_rate": 1.1292919468045877e-05,
      "loss": 0.7011,
      "step": 183
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.843255043029785,
      "learning_rate": 1.0844911450972229e-05,
      "loss": 0.3824,
      "step": 184
    },
    {
      "epoch": 1.608695652173913,
      "grad_norm": 7.702304840087891,
      "learning_rate": 1.0404887703886251e-05,
      "loss": 0.782,
      "step": 185
    },
    {
      "epoch": 1.617391304347826,
      "grad_norm": 6.250712871551514,
      "learning_rate": 9.972937953781986e-06,
      "loss": 0.3592,
      "step": 186
    },
    {
      "epoch": 1.626086956521739,
      "grad_norm": 8.084029197692871,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.5398,
      "step": 187
    },
    {
      "epoch": 1.634782608695652,
      "grad_norm": 3.528081178665161,
      "learning_rate": 9.133611102529654e-06,
      "loss": 0.0699,
      "step": 188
    },
    {
      "epoch": 1.643478260869565,
      "grad_norm": 3.4632437229156494,
      "learning_rate": 8.7264051518613e-06,
      "loss": 0.0686,
      "step": 189
    },
    {
      "epoch": 1.6521739130434783,
      "grad_norm": 7.657890319824219,
      "learning_rate": 8.327615464234129e-06,
      "loss": 0.6129,
      "step": 190
    },
    {
      "epoch": 1.6608695652173913,
      "grad_norm": 9.106538772583008,
      "learning_rate": 7.937323358440935e-06,
      "loss": 0.8445,
      "step": 191
    },
    {
      "epoch": 1.6695652173913045,
      "grad_norm": 6.055113315582275,
      "learning_rate": 7.555608420498872e-06,
      "loss": 0.6026,
      "step": 192
    },
    {
      "epoch": 1.6782608695652175,
      "grad_norm": 17.939279556274414,
      "learning_rate": 7.182548487420554e-06,
      "loss": 1.5311,
      "step": 193
    },
    {
      "epoch": 1.6869565217391305,
      "grad_norm": 5.84937858581543,
      "learning_rate": 6.818219631342149e-06,
      "loss": 0.1457,
      "step": 194
    },
    {
      "epoch": 1.6956521739130435,
      "grad_norm": 10.647073745727539,
      "learning_rate": 6.462696144011149e-06,
      "loss": 1.0994,
      "step": 195
    },
    {
      "epoch": 1.7043478260869565,
      "grad_norm": 6.252867221832275,
      "learning_rate": 6.116050521637218e-06,
      "loss": 0.3648,
      "step": 196
    },
    {
      "epoch": 1.7130434782608694,
      "grad_norm": 6.860467910766602,
      "learning_rate": 5.778353450109286e-06,
      "loss": 0.6171,
      "step": 197
    },
    {
      "epoch": 1.7217391304347827,
      "grad_norm": 6.052587032318115,
      "learning_rate": 5.449673790581611e-06,
      "loss": 0.3541,
      "step": 198
    },
    {
      "epoch": 1.7304347826086957,
      "grad_norm": 4.472194194793701,
      "learning_rate": 5.13007856543209e-06,
      "loss": 0.0866,
      "step": 199
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 9.050908088684082,
      "learning_rate": 4.819632944595415e-06,
      "loss": 0.7291,
      "step": 200
    },
    {
      "epoch": 1.7478260869565219,
      "grad_norm": 15.068604469299316,
      "learning_rate": 4.5184002322740785e-06,
      "loss": 0.7179,
      "step": 201
    },
    {
      "epoch": 1.7565217391304349,
      "grad_norm": 9.076845169067383,
      "learning_rate": 4.2264418540297e-06,
      "loss": 1.4375,
      "step": 202
    },
    {
      "epoch": 1.7652173913043478,
      "grad_norm": 7.426385879516602,
      "learning_rate": 3.9438173442575e-06,
      "loss": 0.7565,
      "step": 203
    },
    {
      "epoch": 1.7739130434782608,
      "grad_norm": 7.295407772064209,
      "learning_rate": 3.6705843340464286e-06,
      "loss": 0.4621,
      "step": 204
    },
    {
      "epoch": 1.7826086956521738,
      "grad_norm": 5.500656604766846,
      "learning_rate": 3.406798539427386e-06,
      "loss": 0.0947,
      "step": 205
    },
    {
      "epoch": 1.7913043478260868,
      "grad_norm": 7.827192783355713,
      "learning_rate": 3.1525137500119207e-06,
      "loss": 0.5042,
      "step": 206
    },
    {
      "epoch": 1.8,
      "grad_norm": 10.173295021057129,
      "learning_rate": 2.9077818180237693e-06,
      "loss": 0.7344,
      "step": 207
    },
    {
      "epoch": 1.808695652173913,
      "grad_norm": 8.067317962646484,
      "learning_rate": 2.6726526477254987e-06,
      "loss": 0.8057,
      "step": 208
    },
    {
      "epoch": 1.8173913043478263,
      "grad_norm": 6.457295894622803,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 0.5597,
      "step": 209
    },
    {
      "epoch": 1.8260869565217392,
      "grad_norm": 6.850070953369141,
      "learning_rate": 2.2313924087851656e-06,
      "loss": 0.556,
      "step": 210
    },
    {
      "epoch": 1.8347826086956522,
      "grad_norm": 6.4678144454956055,
      "learning_rate": 2.0253513192751373e-06,
      "loss": 0.351,
      "step": 211
    },
    {
      "epoch": 1.8434782608695652,
      "grad_norm": 6.126864433288574,
      "learning_rate": 1.8290929313710513e-06,
      "loss": 0.3537,
      "step": 212
    },
    {
      "epoch": 1.8521739130434782,
      "grad_norm": 6.90460729598999,
      "learning_rate": 1.6426572649021476e-06,
      "loss": 0.1795,
      "step": 213
    },
    {
      "epoch": 1.8608695652173912,
      "grad_norm": 8.008868217468262,
      "learning_rate": 1.4660823367073751e-06,
      "loss": 0.5865,
      "step": 214
    },
    {
      "epoch": 1.8695652173913042,
      "grad_norm": 9.392267227172852,
      "learning_rate": 1.2994041528833266e-06,
      "loss": 0.8658,
      "step": 215
    },
    {
      "epoch": 1.8782608695652174,
      "grad_norm": 8.422052383422852,
      "learning_rate": 1.1426567014420297e-06,
      "loss": 0.636,
      "step": 216
    },
    {
      "epoch": 1.8869565217391304,
      "grad_norm": 12.24099063873291,
      "learning_rate": 9.958719453803278e-07,
      "loss": 1.9606,
      "step": 217
    },
    {
      "epoch": 1.8956521739130436,
      "grad_norm": 6.257460117340088,
      "learning_rate": 8.590798161622227e-07,
      "loss": 0.4456,
      "step": 218
    },
    {
      "epoch": 1.9043478260869566,
      "grad_norm": 12.042879104614258,
      "learning_rate": 7.323082076153509e-07,
      "loss": 1.299,
      "step": 219
    },
    {
      "epoch": 1.9130434782608696,
      "grad_norm": 16.869014739990234,
      "learning_rate": 6.15582970243117e-07,
      "loss": 1.6858,
      "step": 220
    },
    {
      "epoch": 1.9217391304347826,
      "grad_norm": 10.271821975708008,
      "learning_rate": 5.089279059533658e-07,
      "loss": 0.8549,
      "step": 221
    },
    {
      "epoch": 1.9304347826086956,
      "grad_norm": 6.406334400177002,
      "learning_rate": 4.123647632048644e-07,
      "loss": 0.5526,
      "step": 222
    },
    {
      "epoch": 1.9391304347826086,
      "grad_norm": 7.156313896179199,
      "learning_rate": 3.2591323257248893e-07,
      "loss": 0.2008,
      "step": 223
    },
    {
      "epoch": 1.9478260869565216,
      "grad_norm": 8.381826400756836,
      "learning_rate": 2.4959094273201977e-07,
      "loss": 0.4191,
      "step": 224
    },
    {
      "epoch": 1.9565217391304348,
      "grad_norm": 7.376916408538818,
      "learning_rate": 1.8341345686543332e-07,
      "loss": 0.4707,
      "step": 225
    },
    {
      "epoch": 1.9652173913043478,
      "grad_norm": 9.728565216064453,
      "learning_rate": 1.2739426948732424e-07,
      "loss": 1.1065,
      "step": 226
    },
    {
      "epoch": 1.973913043478261,
      "grad_norm": 7.97186279296875,
      "learning_rate": 8.15448036932176e-08,
      "loss": 0.879,
      "step": 227
    },
    {
      "epoch": 1.982608695652174,
      "grad_norm": 4.87440824508667,
      "learning_rate": 4.5874408830215434e-08,
      "loss": 0.1072,
      "step": 228
    },
    {
      "epoch": 1.991304347826087,
      "grad_norm": 9.028919219970703,
      "learning_rate": 2.0390358590538504e-08,
      "loss": 0.6967,
      "step": 229
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.545851230621338,
      "learning_rate": 5.097849528334919e-09,
      "loss": 0.6529,
      "step": 230
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 230,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 33878645932032.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}