{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 179,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00558659217877095,
      "grad_norm": 3.744102716445923,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 44.5125,
      "step": 1
    },
    {
      "epoch": 0.0111731843575419,
      "grad_norm": 4.090838432312012,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 44.5456,
      "step": 2
    },
    {
      "epoch": 0.01675977653631285,
      "grad_norm": 4.943282127380371,
      "learning_rate": 3e-06,
      "loss": 44.5655,
      "step": 3
    },
    {
      "epoch": 0.0223463687150838,
      "grad_norm": 5.5685224533081055,
      "learning_rate": 4.000000000000001e-06,
      "loss": 44.5459,
      "step": 4
    },
    {
      "epoch": 0.027932960893854747,
      "grad_norm": 6.192835807800293,
      "learning_rate": 5e-06,
      "loss": 44.5324,
      "step": 5
    },
    {
      "epoch": 0.0335195530726257,
      "grad_norm": 6.504510879516602,
      "learning_rate": 6e-06,
      "loss": 44.5719,
      "step": 6
    },
    {
      "epoch": 0.03910614525139665,
      "grad_norm": 7.036836624145508,
      "learning_rate": 7.000000000000001e-06,
      "loss": 44.5859,
      "step": 7
    },
    {
      "epoch": 0.0446927374301676,
      "grad_norm": 6.947369575500488,
      "learning_rate": 8.000000000000001e-06,
      "loss": 44.547,
      "step": 8
    },
    {
      "epoch": 0.05027932960893855,
      "grad_norm": 6.922020435333252,
      "learning_rate": 9e-06,
      "loss": 44.509,
      "step": 9
    },
    {
      "epoch": 0.055865921787709494,
      "grad_norm": 7.10311222076416,
      "learning_rate": 1e-05,
      "loss": 44.4877,
      "step": 10
    },
    {
      "epoch": 0.061452513966480445,
      "grad_norm": 7.27517032623291,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 44.5195,
      "step": 11
    },
    {
      "epoch": 0.0670391061452514,
      "grad_norm": 7.023418426513672,
      "learning_rate": 1.2e-05,
      "loss": 44.4954,
      "step": 12
    },
    {
      "epoch": 0.07262569832402235,
      "grad_norm": 7.948561191558838,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 44.4922,
      "step": 13
    },
    {
      "epoch": 0.0782122905027933,
      "grad_norm": 7.6504926681518555,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 44.459,
      "step": 14
    },
    {
      "epoch": 0.08379888268156424,
      "grad_norm": 8.089448928833008,
      "learning_rate": 1.5e-05,
      "loss": 44.4743,
      "step": 15
    },
    {
      "epoch": 0.0893854748603352,
      "grad_norm": 7.560296058654785,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 44.5191,
      "step": 16
    },
    {
      "epoch": 0.09497206703910614,
      "grad_norm": 7.827347755432129,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 44.4517,
      "step": 17
    },
    {
      "epoch": 0.1005586592178771,
      "grad_norm": 8.726825714111328,
      "learning_rate": 1.8e-05,
      "loss": 44.484,
      "step": 18
    },
    {
      "epoch": 0.10614525139664804,
      "grad_norm": 7.8799943923950195,
      "learning_rate": 1.9e-05,
      "loss": 44.4365,
      "step": 19
    },
    {
      "epoch": 0.11173184357541899,
      "grad_norm": 7.783633232116699,
      "learning_rate": 2e-05,
      "loss": 44.432,
      "step": 20
    },
    {
      "epoch": 0.11731843575418995,
      "grad_norm": 8.265217781066895,
      "learning_rate": 2.1e-05,
      "loss": 44.3541,
      "step": 21
    },
    {
      "epoch": 0.12290502793296089,
      "grad_norm": 8.179393768310547,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 44.4657,
      "step": 22
    },
    {
      "epoch": 0.12849162011173185,
      "grad_norm": 8.401519775390625,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 44.3664,
      "step": 23
    },
    {
      "epoch": 0.1340782122905028,
      "grad_norm": 8.343657493591309,
      "learning_rate": 2.4e-05,
      "loss": 44.356,
      "step": 24
    },
    {
      "epoch": 0.13966480446927373,
      "grad_norm": 8.581681251525879,
      "learning_rate": 2.5e-05,
      "loss": 44.3288,
      "step": 25
    },
    {
      "epoch": 0.1452513966480447,
      "grad_norm": 8.727663040161133,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 44.3571,
      "step": 26
    },
    {
      "epoch": 0.15083798882681565,
      "grad_norm": 8.828510284423828,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 44.2998,
      "step": 27
    },
    {
      "epoch": 0.1564245810055866,
      "grad_norm": 8.784181594848633,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 44.2794,
      "step": 28
    },
    {
      "epoch": 0.16201117318435754,
      "grad_norm": 8.831645965576172,
      "learning_rate": 2.9e-05,
      "loss": 44.2463,
      "step": 29
    },
    {
      "epoch": 0.16759776536312848,
      "grad_norm": 8.893630981445312,
      "learning_rate": 3e-05,
      "loss": 44.2231,
      "step": 30
    },
    {
      "epoch": 0.17318435754189945,
      "grad_norm": 9.3904390335083,
      "learning_rate": 3.1e-05,
      "loss": 44.1382,
      "step": 31
    },
    {
      "epoch": 0.1787709497206704,
      "grad_norm": 8.971839904785156,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 44.12,
      "step": 32
    },
    {
      "epoch": 0.18435754189944134,
      "grad_norm": 9.59831428527832,
      "learning_rate": 3.3e-05,
      "loss": 44.114,
      "step": 33
    },
    {
      "epoch": 0.18994413407821228,
      "grad_norm": 9.502549171447754,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 44.0284,
      "step": 34
    },
    {
      "epoch": 0.19553072625698323,
      "grad_norm": 9.720560073852539,
      "learning_rate": 3.5e-05,
      "loss": 43.9652,
      "step": 35
    },
    {
      "epoch": 0.2011173184357542,
      "grad_norm": 10.408364295959473,
      "learning_rate": 3.6e-05,
      "loss": 43.9892,
      "step": 36
    },
    {
      "epoch": 0.20670391061452514,
      "grad_norm": 9.865303993225098,
      "learning_rate": 3.7e-05,
      "loss": 43.9092,
      "step": 37
    },
    {
      "epoch": 0.2122905027932961,
      "grad_norm": 9.902585983276367,
      "learning_rate": 3.8e-05,
      "loss": 43.8354,
      "step": 38
    },
    {
      "epoch": 0.21787709497206703,
      "grad_norm": 9.79003620147705,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 43.7437,
      "step": 39
    },
    {
      "epoch": 0.22346368715083798,
      "grad_norm": 10.79495906829834,
      "learning_rate": 4e-05,
      "loss": 43.7073,
      "step": 40
    },
    {
      "epoch": 0.22905027932960895,
      "grad_norm": 10.677367210388184,
      "learning_rate": 4.1e-05,
      "loss": 43.6095,
      "step": 41
    },
    {
      "epoch": 0.2346368715083799,
      "grad_norm": 11.262158393859863,
      "learning_rate": 4.2e-05,
      "loss": 43.5185,
      "step": 42
    },
    {
      "epoch": 0.24022346368715083,
      "grad_norm": 11.903176307678223,
      "learning_rate": 4.3e-05,
      "loss": 43.3865,
      "step": 43
    },
    {
      "epoch": 0.24581005586592178,
      "grad_norm": 11.95860481262207,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 43.3,
      "step": 44
    },
    {
      "epoch": 0.25139664804469275,
      "grad_norm": 5.72437047958374,
      "learning_rate": 4.5e-05,
      "loss": 44.2749,
      "step": 45
    },
    {
      "epoch": 0.2569832402234637,
      "grad_norm": 4.2836833000183105,
      "learning_rate": 4.600000000000001e-05,
      "loss": 44.1268,
      "step": 46
    },
    {
      "epoch": 0.26256983240223464,
      "grad_norm": 5.577621936798096,
      "learning_rate": 4.7e-05,
      "loss": 43.9228,
      "step": 47
    },
    {
      "epoch": 0.2681564245810056,
      "grad_norm": 5.966395854949951,
      "learning_rate": 4.8e-05,
      "loss": 43.7729,
      "step": 48
    },
    {
      "epoch": 0.2737430167597765,
      "grad_norm": 6.50309419631958,
      "learning_rate": 4.9e-05,
      "loss": 43.671,
      "step": 49
    },
    {
      "epoch": 0.27932960893854747,
      "grad_norm": 6.988935947418213,
      "learning_rate": 5e-05,
      "loss": 43.5641,
      "step": 50
    },
    {
      "epoch": 0.2849162011173184,
      "grad_norm": 7.089043140411377,
      "learning_rate": 5.1000000000000006e-05,
      "loss": 43.5187,
      "step": 51
    },
    {
      "epoch": 0.2905027932960894,
      "grad_norm": 7.402230739593506,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 43.4116,
      "step": 52
    },
    {
      "epoch": 0.29608938547486036,
      "grad_norm": 7.716123580932617,
      "learning_rate": 5.300000000000001e-05,
      "loss": 43.2982,
      "step": 53
    },
    {
      "epoch": 0.3016759776536313,
      "grad_norm": 7.589552879333496,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 43.2493,
      "step": 54
    },
    {
      "epoch": 0.30726256983240224,
      "grad_norm": 8.715299606323242,
      "learning_rate": 5.500000000000001e-05,
      "loss": 43.0231,
      "step": 55
    },
    {
      "epoch": 0.3128491620111732,
      "grad_norm": 8.259060859680176,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 42.9955,
      "step": 56
    },
    {
      "epoch": 0.31843575418994413,
      "grad_norm": 7.728129863739014,
      "learning_rate": 5.6999999999999996e-05,
      "loss": 42.9847,
      "step": 57
    },
    {
      "epoch": 0.3240223463687151,
      "grad_norm": 8.82850170135498,
      "learning_rate": 5.8e-05,
      "loss": 42.7967,
      "step": 58
    },
    {
      "epoch": 0.329608938547486,
      "grad_norm": 7.893709659576416,
      "learning_rate": 5.9e-05,
      "loss": 42.8577,
      "step": 59
    },
    {
      "epoch": 0.33519553072625696,
      "grad_norm": 8.985579490661621,
      "learning_rate": 6e-05,
      "loss": 42.5203,
      "step": 60
    },
    {
      "epoch": 0.3407821229050279,
      "grad_norm": 8.47461223602295,
      "learning_rate": 6.1e-05,
      "loss": 42.5511,
      "step": 61
    },
    {
      "epoch": 0.3463687150837989,
      "grad_norm": 8.799412727355957,
      "learning_rate": 6.2e-05,
      "loss": 42.4044,
      "step": 62
    },
    {
      "epoch": 0.35195530726256985,
      "grad_norm": 8.567609786987305,
      "learning_rate": 6.3e-05,
      "loss": 42.4246,
      "step": 63
    },
    {
      "epoch": 0.3575418994413408,
      "grad_norm": 8.90191650390625,
      "learning_rate": 6.400000000000001e-05,
      "loss": 42.2267,
      "step": 64
    },
    {
      "epoch": 0.36312849162011174,
      "grad_norm": 8.651405334472656,
      "learning_rate": 6.500000000000001e-05,
      "loss": 42.2001,
      "step": 65
    },
    {
      "epoch": 0.3687150837988827,
      "grad_norm": 8.63864517211914,
      "learning_rate": 6.6e-05,
      "loss": 42.1931,
      "step": 66
    },
    {
      "epoch": 0.3743016759776536,
      "grad_norm": 9.15833854675293,
      "learning_rate": 6.7e-05,
      "loss": 41.9268,
      "step": 67
    },
    {
      "epoch": 0.37988826815642457,
      "grad_norm": 8.82404613494873,
      "learning_rate": 6.800000000000001e-05,
      "loss": 41.9039,
      "step": 68
    },
    {
      "epoch": 0.3854748603351955,
      "grad_norm": 9.213386535644531,
      "learning_rate": 6.9e-05,
      "loss": 41.7373,
      "step": 69
    },
    {
      "epoch": 0.39106145251396646,
      "grad_norm": 9.157903671264648,
      "learning_rate": 7e-05,
      "loss": 41.6397,
      "step": 70
    },
    {
      "epoch": 0.39664804469273746,
      "grad_norm": 9.073963165283203,
      "learning_rate": 7.1e-05,
      "loss": 41.5592,
      "step": 71
    },
    {
      "epoch": 0.4022346368715084,
      "grad_norm": 9.316044807434082,
      "learning_rate": 7.2e-05,
      "loss": 41.4418,
      "step": 72
    },
    {
      "epoch": 0.40782122905027934,
      "grad_norm": 9.458603858947754,
      "learning_rate": 7.3e-05,
      "loss": 41.2307,
      "step": 73
    },
    {
      "epoch": 0.4134078212290503,
      "grad_norm": 9.475081443786621,
      "learning_rate": 7.4e-05,
      "loss": 41.1623,
      "step": 74
    },
    {
      "epoch": 0.41899441340782123,
      "grad_norm": 9.649319648742676,
      "learning_rate": 7.500000000000001e-05,
      "loss": 41.0343,
      "step": 75
    },
    {
      "epoch": 0.4245810055865922,
      "grad_norm": 9.324378967285156,
      "learning_rate": 7.6e-05,
      "loss": 41.0046,
      "step": 76
    },
    {
      "epoch": 0.4301675977653631,
      "grad_norm": 9.610836029052734,
      "learning_rate": 7.7e-05,
      "loss": 40.7329,
      "step": 77
    },
    {
      "epoch": 0.43575418994413406,
      "grad_norm": 9.787153244018555,
      "learning_rate": 7.800000000000001e-05,
      "loss": 40.5896,
      "step": 78
    },
    {
      "epoch": 0.441340782122905,
      "grad_norm": 9.9530668258667,
      "learning_rate": 7.900000000000001e-05,
      "loss": 40.448,
      "step": 79
    },
    {
      "epoch": 0.44692737430167595,
      "grad_norm": 9.689455032348633,
      "learning_rate": 8e-05,
      "loss": 40.5078,
      "step": 80
    },
    {
      "epoch": 0.45251396648044695,
      "grad_norm": 9.322986602783203,
      "learning_rate": 8.1e-05,
      "loss": 40.4844,
      "step": 81
    },
    {
      "epoch": 0.4581005586592179,
      "grad_norm": 9.782176971435547,
      "learning_rate": 8.2e-05,
      "loss": 40.2081,
      "step": 82
    },
    {
      "epoch": 0.46368715083798884,
      "grad_norm": 9.894826889038086,
      "learning_rate": 8.3e-05,
      "loss": 40.0162,
      "step": 83
    },
    {
      "epoch": 0.4692737430167598,
      "grad_norm": 10.18328857421875,
      "learning_rate": 8.4e-05,
      "loss": 39.9261,
      "step": 84
    },
    {
      "epoch": 0.4748603351955307,
      "grad_norm": 10.649673461914062,
      "learning_rate": 8.5e-05,
      "loss": 39.692,
      "step": 85
    },
    {
      "epoch": 0.48044692737430167,
      "grad_norm": 10.699885368347168,
      "learning_rate": 8.6e-05,
      "loss": 39.5022,
      "step": 86
    },
    {
      "epoch": 0.4860335195530726,
      "grad_norm": 11.469558715820312,
      "learning_rate": 8.7e-05,
      "loss": 39.1245,
      "step": 87
    },
    {
      "epoch": 0.49162011173184356,
      "grad_norm": 10.971457481384277,
      "learning_rate": 8.800000000000001e-05,
      "loss": 39.1588,
      "step": 88
    },
    {
      "epoch": 0.4972067039106145,
      "grad_norm": 7.072690010070801,
      "learning_rate": 8.900000000000001e-05,
      "loss": 42.9127,
      "step": 89
    },
    {
      "epoch": 0.5027932960893855,
      "grad_norm": 5.59442663192749,
      "learning_rate": 9e-05,
      "loss": 42.0852,
      "step": 90
    },
    {
      "epoch": 0.5083798882681564,
      "grad_norm": 6.485333442687988,
      "learning_rate": 9.1e-05,
      "loss": 41.2355,
      "step": 91
    },
    {
      "epoch": 0.5139664804469274,
      "grad_norm": 7.4060797691345215,
      "learning_rate": 9.200000000000001e-05,
      "loss": 40.6051,
      "step": 92
    },
    {
      "epoch": 0.5195530726256983,
      "grad_norm": 7.867504596710205,
      "learning_rate": 9.300000000000001e-05,
      "loss": 40.2738,
      "step": 93
    },
    {
      "epoch": 0.5251396648044693,
      "grad_norm": 7.020578384399414,
      "learning_rate": 9.4e-05,
      "loss": 40.4952,
      "step": 94
    },
    {
      "epoch": 0.5307262569832403,
      "grad_norm": 7.6585516929626465,
      "learning_rate": 9.5e-05,
      "loss": 40.097,
      "step": 95
    },
    {
      "epoch": 0.5363128491620112,
      "grad_norm": 8.454778671264648,
      "learning_rate": 9.6e-05,
      "loss": 39.7017,
      "step": 96
    },
    {
      "epoch": 0.5418994413407822,
      "grad_norm": 8.670785903930664,
      "learning_rate": 9.7e-05,
      "loss": 39.4651,
      "step": 97
    },
    {
      "epoch": 0.547486033519553,
      "grad_norm": 8.416338920593262,
      "learning_rate": 9.8e-05,
      "loss": 39.519,
      "step": 98
    },
    {
      "epoch": 0.553072625698324,
      "grad_norm": 8.337173461914062,
      "learning_rate": 9.900000000000001e-05,
      "loss": 39.4007,
      "step": 99
    },
    {
      "epoch": 0.5586592178770949,
      "grad_norm": 8.42642879486084,
      "learning_rate": 0.0001,
      "loss": 39.1839,
      "step": 100
    },
    {
      "epoch": 0.5642458100558659,
      "grad_norm": 8.729036331176758,
      "learning_rate": 9.996046986136509e-05,
      "loss": 38.9884,
      "step": 101
    },
    {
      "epoch": 0.5698324022346368,
      "grad_norm": 8.876291275024414,
      "learning_rate": 9.98419419507348e-05,
      "loss": 38.7318,
      "step": 102
    },
    {
      "epoch": 0.5754189944134078,
      "grad_norm": 8.565130233764648,
      "learning_rate": 9.964460368509867e-05,
      "loss": 38.7833,
      "step": 103
    },
    {
      "epoch": 0.5810055865921788,
      "grad_norm": 8.649067878723145,
      "learning_rate": 9.936876709681668e-05,
      "loss": 38.6301,
      "step": 104
    },
    {
      "epoch": 0.5865921787709497,
      "grad_norm": 8.931990623474121,
      "learning_rate": 9.901486834023182e-05,
      "loss": 38.4478,
      "step": 105
    },
    {
      "epoch": 0.5921787709497207,
      "grad_norm": 9.383662223815918,
      "learning_rate": 9.85834670020205e-05,
      "loss": 38.209,
      "step": 106
    },
    {
      "epoch": 0.5977653631284916,
      "grad_norm": 8.624167442321777,
      "learning_rate": 9.807524521637102e-05,
      "loss": 38.3029,
      "step": 107
    },
    {
      "epoch": 0.6033519553072626,
      "grad_norm": 8.637563705444336,
      "learning_rate": 9.749100658638914e-05,
      "loss": 38.1992,
      "step": 108
    },
    {
      "epoch": 0.6089385474860335,
      "grad_norm": 9.110807418823242,
      "learning_rate": 9.68316749134364e-05,
      "loss": 37.9357,
      "step": 109
    },
    {
      "epoch": 0.6145251396648045,
      "grad_norm": 8.739825248718262,
      "learning_rate": 9.609829273641034e-05,
      "loss": 37.929,
      "step": 110
    },
    {
      "epoch": 0.6201117318435754,
      "grad_norm": 9.419713020324707,
      "learning_rate": 9.529201968327616e-05,
      "loss": 37.5423,
      "step": 111
    },
    {
      "epoch": 0.6256983240223464,
      "grad_norm": 9.36577320098877,
      "learning_rate": 9.44141306374566e-05,
      "loss": 37.4528,
      "step": 112
    },
    {
      "epoch": 0.6312849162011173,
      "grad_norm": 9.40024185180664,
      "learning_rate": 9.346601372197914e-05,
      "loss": 37.3334,
      "step": 113
    },
    {
      "epoch": 0.6368715083798883,
      "grad_norm": 9.359634399414062,
      "learning_rate": 9.244916810456821e-05,
      "loss": 37.3102,
      "step": 114
    },
    {
      "epoch": 0.6424581005586593,
      "grad_norm": 9.717304229736328,
      "learning_rate": 9.136520162715287e-05,
      "loss": 37.1064,
      "step": 115
    },
    {
      "epoch": 0.6480446927374302,
      "grad_norm": 9.580801963806152,
      "learning_rate": 9.021582826353824e-05,
      "loss": 37.0496,
      "step": 116
    },
    {
      "epoch": 0.6536312849162011,
      "grad_norm": 9.399163246154785,
      "learning_rate": 8.900286540926061e-05,
      "loss": 36.9901,
      "step": 117
    },
    {
      "epoch": 0.659217877094972,
      "grad_norm": 9.52145004272461,
      "learning_rate": 8.772823100791151e-05,
      "loss": 36.79,
      "step": 118
    },
    {
      "epoch": 0.664804469273743,
      "grad_norm": 9.533082962036133,
      "learning_rate": 8.639394051847472e-05,
      "loss": 36.6575,
      "step": 119
    },
    {
      "epoch": 0.6703910614525139,
      "grad_norm": 9.642265319824219,
      "learning_rate": 8.500210372847127e-05,
      "loss": 36.4436,
      "step": 120
    },
    {
      "epoch": 0.6759776536312849,
      "grad_norm": 10.193774223327637,
      "learning_rate": 8.355492141795185e-05,
      "loss": 36.1696,
      "step": 121
    },
    {
      "epoch": 0.6815642458100558,
      "grad_norm": 9.560378074645996,
      "learning_rate": 8.2054681879611e-05,
      "loss": 36.3306,
      "step": 122
    },
    {
      "epoch": 0.6871508379888268,
      "grad_norm": 9.826172828674316,
      "learning_rate": 8.050375730052621e-05,
      "loss": 36.1994,
      "step": 123
    },
    {
      "epoch": 0.6927374301675978,
      "grad_norm": 10.785990715026855,
      "learning_rate": 7.890460001124242e-05,
      "loss": 36.3292,
      "step": 124
    },
    {
      "epoch": 0.6983240223463687,
      "grad_norm": 10.091948509216309,
      "learning_rate": 7.725973860813338e-05,
      "loss": 35.8338,
      "step": 125
    },
    {
      "epoch": 0.7039106145251397,
      "grad_norm": 9.727133750915527,
      "learning_rate": 7.557177395517112e-05,
      "loss": 36.0643,
      "step": 126
    },
    {
      "epoch": 0.7094972067039106,
      "grad_norm": 10.128454208374023,
      "learning_rate": 7.384337507142531e-05,
      "loss": 35.8633,
      "step": 127
    },
    {
      "epoch": 0.7150837988826816,
      "grad_norm": 10.032463073730469,
      "learning_rate": 7.20772749107956e-05,
      "loss": 35.75,
      "step": 128
    },
    {
      "epoch": 0.7206703910614525,
      "grad_norm": 10.427178382873535,
      "learning_rate": 7.027626604064969e-05,
      "loss": 35.4646,
      "step": 129
    },
    {
      "epoch": 0.7262569832402235,
      "grad_norm": 10.725203514099121,
      "learning_rate": 6.844319622620039e-05,
      "loss": 35.4019,
      "step": 130
    },
    {
      "epoch": 0.7318435754189944,
      "grad_norm": 11.060110092163086,
      "learning_rate": 6.65809639276034e-05,
      "loss": 35.1042,
      "step": 131
    },
    {
      "epoch": 0.7374301675977654,
      "grad_norm": 11.478692054748535,
      "learning_rate": 6.469251371689606e-05,
      "loss": 34.8198,
      "step": 132
    },
    {
      "epoch": 0.7430167597765364,
      "grad_norm": 6.56369161605835,
      "learning_rate": 6.278083162202375e-05,
      "loss": 39.9774,
      "step": 133
    },
    {
      "epoch": 0.7486033519553073,
      "grad_norm": 6.60885763168335,
      "learning_rate": 6.08489404053159e-05,
      "loss": 38.8113,
      "step": 134
    },
    {
      "epoch": 0.7541899441340782,
      "grad_norm": 6.839470863342285,
      "learning_rate": 5.889989478387753e-05,
      "loss": 38.2958,
      "step": 135
    },
    {
      "epoch": 0.7597765363128491,
      "grad_norm": 7.189307689666748,
      "learning_rate": 5.6936776599453424e-05,
      "loss": 37.661,
      "step": 136
    },
    {
      "epoch": 0.7653631284916201,
      "grad_norm": 7.412992000579834,
      "learning_rate": 5.496268994540309e-05,
      "loss": 37.3239,
      "step": 137
    },
    {
      "epoch": 0.770949720670391,
      "grad_norm": 7.726637363433838,
      "learning_rate": 5.2980756258490995e-05,
      "loss": 36.8866,
      "step": 138
    },
    {
      "epoch": 0.776536312849162,
      "grad_norm": 8.184843063354492,
      "learning_rate": 5.0994109383253506e-05,
      "loss": 36.6651,
      "step": 139
    },
    {
      "epoch": 0.7821229050279329,
      "grad_norm": 7.629161834716797,
      "learning_rate": 4.900589061674649e-05,
      "loss": 36.7795,
      "step": 140
    },
    {
      "epoch": 0.7877094972067039,
      "grad_norm": 8.492084503173828,
      "learning_rate": 4.701924374150901e-05,
      "loss": 36.3768,
      "step": 141
    },
    {
      "epoch": 0.7932960893854749,
      "grad_norm": 8.768118858337402,
      "learning_rate": 4.503731005459693e-05,
      "loss": 36.2209,
      "step": 142
    },
    {
      "epoch": 0.7988826815642458,
      "grad_norm": 9.186103820800781,
      "learning_rate": 4.3063223400546594e-05,
      "loss": 35.9594,
      "step": 143
    },
    {
      "epoch": 0.8044692737430168,
      "grad_norm": 7.938751220703125,
      "learning_rate": 4.11001052161225e-05,
      "loss": 36.3444,
      "step": 144
    },
    {
      "epoch": 0.8100558659217877,
      "grad_norm": 8.192068099975586,
      "learning_rate": 3.91510595946841e-05,
      "loss": 36.202,
      "step": 145
    },
    {
      "epoch": 0.8156424581005587,
      "grad_norm": 8.316092491149902,
      "learning_rate": 3.721916837797627e-05,
      "loss": 36.128,
      "step": 146
    },
    {
      "epoch": 0.8212290502793296,
      "grad_norm": 7.804288387298584,
      "learning_rate": 3.5307486283103966e-05,
      "loss": 36.2991,
      "step": 147
    },
    {
      "epoch": 0.8268156424581006,
      "grad_norm": 9.470741271972656,
      "learning_rate": 3.3419036072396616e-05,
      "loss": 35.5116,
      "step": 148
    },
    {
      "epoch": 0.8324022346368715,
      "grad_norm": 8.43538761138916,
      "learning_rate": 3.1556803773799614e-05,
      "loss": 35.9411,
      "step": 149
    },
    {
      "epoch": 0.8379888268156425,
      "grad_norm": 9.065338134765625,
      "learning_rate": 2.9723733959350307e-05,
      "loss": 35.6712,
      "step": 150
    },
    {
      "epoch": 0.8435754189944135,
      "grad_norm": 8.624382972717285,
      "learning_rate": 2.7922725089204426e-05,
      "loss": 35.8245,
      "step": 151
    },
    {
      "epoch": 0.8491620111731844,
      "grad_norm": 8.553651809692383,
      "learning_rate": 2.6156624928574707e-05,
      "loss": 35.8094,
      "step": 152
    },
    {
      "epoch": 0.8547486033519553,
      "grad_norm": 9.5851411819458,
      "learning_rate": 2.4428226044828896e-05,
      "loss": 35.4061,
      "step": 153
    },
    {
      "epoch": 0.8603351955307262,
      "grad_norm": 9.327908515930176,
      "learning_rate": 2.2740261391866637e-05,
      "loss": 35.4764,
      "step": 154
    },
    {
      "epoch": 0.8659217877094972,
      "grad_norm": 9.505815505981445,
      "learning_rate": 2.1095399988757574e-05,
      "loss": 35.3205,
      "step": 155
    },
    {
      "epoch": 0.8715083798882681,
      "grad_norm": 9.7329740524292,
      "learning_rate": 1.9496242699473783e-05,
      "loss": 35.1241,
      "step": 156
    },
    {
      "epoch": 0.8770949720670391,
      "grad_norm": 9.277899742126465,
      "learning_rate": 1.794531812038901e-05,
      "loss": 35.3066,
      "step": 157
    },
    {
      "epoch": 0.88268156424581,
      "grad_norm": 9.590785026550293,
      "learning_rate": 1.6445078582048155e-05,
      "loss": 35.1618,
      "step": 158
    },
    {
      "epoch": 0.888268156424581,
      "grad_norm": 9.559935569763184,
      "learning_rate": 1.4997896271528739e-05,
      "loss": 35.1343,
      "step": 159
    },
    {
      "epoch": 0.8938547486033519,
      "grad_norm": 9.499947547912598,
      "learning_rate": 1.3606059481525296e-05,
      "loss": 35.1343,
      "step": 160
    },
    {
      "epoch": 0.8994413407821229,
      "grad_norm": 9.608393669128418,
      "learning_rate": 1.2271768992088489e-05,
      "loss": 35.2073,
      "step": 161
    },
    {
      "epoch": 0.9050279329608939,
      "grad_norm": 9.510586738586426,
      "learning_rate": 1.09971345907394e-05,
      "loss": 35.1469,
      "step": 162
    },
    {
      "epoch": 0.9106145251396648,
      "grad_norm": 9.812493324279785,
      "learning_rate": 9.784171736461762e-06,
      "loss": 34.9144,
      "step": 163
    },
    {
      "epoch": 0.9162011173184358,
      "grad_norm": 10.067626953125,
      "learning_rate": 8.634798372847148e-06,
      "loss": 34.9231,
      "step": 164
    },
    {
      "epoch": 0.9217877094972067,
      "grad_norm": 9.783865928649902,
      "learning_rate": 7.550831895431798e-06,
      "loss": 34.8879,
      "step": 165
    },
    {
      "epoch": 0.9273743016759777,
      "grad_norm": 10.413803100585938,
      "learning_rate": 6.533986278020876e-06,
      "loss": 34.6783,
      "step": 166
    },
    {
      "epoch": 0.9329608938547486,
      "grad_norm": 9.892301559448242,
      "learning_rate": 5.585869362543416e-06,
      "loss": 34.9329,
      "step": 167
    },
    {
      "epoch": 0.9385474860335196,
      "grad_norm": 10.211995124816895,
      "learning_rate": 4.707980316723837e-06,
      "loss": 34.9967,
      "step": 168
    },
    {
      "epoch": 0.9441340782122905,
      "grad_norm": 10.002083778381348,
      "learning_rate": 3.901707263589671e-06,
      "loss": 34.7943,
      "step": 169
    },
    {
      "epoch": 0.9497206703910615,
      "grad_norm": 9.93572998046875,
      "learning_rate": 3.1683250865636114e-06,
      "loss": 34.842,
      "step": 170
    },
    {
      "epoch": 0.9553072625698324,
      "grad_norm": 9.871206283569336,
      "learning_rate": 2.5089934136108664e-06,
      "loss": 34.8242,
      "step": 171
    },
    {
      "epoch": 0.9608938547486033,
      "grad_norm": 10.331404685974121,
      "learning_rate": 1.9247547836289793e-06,
      "loss": 34.9185,
      "step": 172
    },
    {
      "epoch": 0.9664804469273743,
      "grad_norm": 10.266081809997559,
      "learning_rate": 1.4165329979794973e-06,
      "loss": 34.6464,
      "step": 173
    },
    {
      "epoch": 0.9720670391061452,
      "grad_norm": 11.090232849121094,
      "learning_rate": 9.851316597681958e-07,
      "loss": 34.4733,
      "step": 174
    },
    {
      "epoch": 0.9776536312849162,
      "grad_norm": 11.902629852294922,
      "learning_rate": 6.312329031833319e-07,
      "loss": 34.1088,
      "step": 175
    },
    {
      "epoch": 0.9832402234636871,
      "grad_norm": 10.5414400100708,
      "learning_rate": 3.553963149013295e-07,
      "loss": 34.5576,
      "step": 176
    },
    {
      "epoch": 0.9888268156424581,
      "grad_norm": 6.398901462554932,
      "learning_rate": 1.580580492652084e-07,
      "loss": 37.1557,
      "step": 177
    },
    {
      "epoch": 0.994413407821229,
      "grad_norm": 9.017801284790039,
      "learning_rate": 3.953013863490784e-08,
      "loss": 35.3366,
      "step": 178
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.074228286743164,
      "learning_rate": 0.0,
      "loss": 34.6733,
      "step": 179
    }
  ],
  "logging_steps": 1,
  "max_steps": 179,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 239,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 154875372503040.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}