{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 335,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0029850746268656717,
      "grad_norm": 8.253611034452266,
      "learning_rate": 9.090909090909091e-07,
      "loss": 2.2247,
      "step": 1
    },
    {
      "epoch": 0.005970149253731343,
      "grad_norm": 7.395306126962346,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 2.1255,
      "step": 2
    },
    {
      "epoch": 0.008955223880597015,
      "grad_norm": 8.26748388260573,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 2.1023,
      "step": 3
    },
    {
      "epoch": 0.011940298507462687,
      "grad_norm": 7.1200500158632,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 2.0025,
      "step": 4
    },
    {
      "epoch": 0.014925373134328358,
      "grad_norm": 5.9633773670024315,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 1.9076,
      "step": 5
    },
    {
      "epoch": 0.01791044776119403,
      "grad_norm": 5.546547205735884,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 1.8587,
      "step": 6
    },
    {
      "epoch": 0.020895522388059702,
      "grad_norm": 5.042219157986929,
      "learning_rate": 6.363636363636364e-06,
      "loss": 1.8198,
      "step": 7
    },
    {
      "epoch": 0.023880597014925373,
      "grad_norm": 3.5185430665767887,
      "learning_rate": 7.272727272727273e-06,
      "loss": 1.5117,
      "step": 8
    },
    {
      "epoch": 0.026865671641791045,
      "grad_norm": 3.1354746744528867,
      "learning_rate": 8.181818181818183e-06,
      "loss": 1.4414,
      "step": 9
    },
    {
      "epoch": 0.029850746268656716,
      "grad_norm": 2.757132384211869,
      "learning_rate": 9.090909090909091e-06,
      "loss": 1.4059,
      "step": 10
    },
    {
      "epoch": 0.03283582089552239,
      "grad_norm": 5.307800727766051,
      "learning_rate": 1e-05,
      "loss": 1.4367,
      "step": 11
    },
    {
      "epoch": 0.03582089552238806,
      "grad_norm": 4.364662300190612,
      "learning_rate": 9.99976495753613e-06,
      "loss": 1.4175,
      "step": 12
    },
    {
      "epoch": 0.03880597014925373,
      "grad_norm": 3.136010913151125,
      "learning_rate": 9.999059852242508e-06,
      "loss": 1.297,
      "step": 13
    },
    {
      "epoch": 0.041791044776119404,
      "grad_norm": 2.6314270494566716,
      "learning_rate": 9.997884750411004e-06,
      "loss": 1.178,
      "step": 14
    },
    {
      "epoch": 0.04477611940298507,
      "grad_norm": 2.8108610676101105,
      "learning_rate": 9.996239762521152e-06,
      "loss": 1.311,
      "step": 15
    },
    {
      "epoch": 0.04776119402985075,
      "grad_norm": 2.646130122162001,
      "learning_rate": 9.994125043229753e-06,
      "loss": 1.1026,
      "step": 16
    },
    {
      "epoch": 0.050746268656716415,
      "grad_norm": 2.3026683820565506,
      "learning_rate": 9.991540791356342e-06,
      "loss": 1.0708,
      "step": 17
    },
    {
      "epoch": 0.05373134328358209,
      "grad_norm": 2.2060475680712157,
      "learning_rate": 9.98848724986449e-06,
      "loss": 1.1484,
      "step": 18
    },
    {
      "epoch": 0.056716417910447764,
      "grad_norm": 2.22608658907851,
      "learning_rate": 9.98496470583896e-06,
      "loss": 1.1819,
      "step": 19
    },
    {
      "epoch": 0.05970149253731343,
      "grad_norm": 2.012346588234655,
      "learning_rate": 9.980973490458728e-06,
      "loss": 1.1226,
      "step": 20
    },
    {
      "epoch": 0.0626865671641791,
      "grad_norm": 2.0805737688444155,
      "learning_rate": 9.976513978965829e-06,
      "loss": 1.0251,
      "step": 21
    },
    {
      "epoch": 0.06567164179104477,
      "grad_norm": 2.384555566832582,
      "learning_rate": 9.971586590630094e-06,
      "loss": 1.0278,
      "step": 22
    },
    {
      "epoch": 0.06865671641791045,
      "grad_norm": 2.1582387653704886,
      "learning_rate": 9.966191788709716e-06,
      "loss": 1.014,
      "step": 23
    },
    {
      "epoch": 0.07164179104477612,
      "grad_norm": 2.0561968840026026,
      "learning_rate": 9.960330080407712e-06,
      "loss": 1.0074,
      "step": 24
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 2.1634547485342828,
      "learning_rate": 9.954002016824226e-06,
      "loss": 1.0355,
      "step": 25
    },
    {
      "epoch": 0.07761194029850746,
      "grad_norm": 2.000183275885172,
      "learning_rate": 9.947208192904722e-06,
      "loss": 1.0241,
      "step": 26
    },
    {
      "epoch": 0.08059701492537313,
      "grad_norm": 1.9928462865297418,
      "learning_rate": 9.939949247384046e-06,
      "loss": 0.9917,
      "step": 27
    },
    {
      "epoch": 0.08358208955223881,
      "grad_norm": 2.1100381727065094,
      "learning_rate": 9.93222586272637e-06,
      "loss": 1.0393,
      "step": 28
    },
    {
      "epoch": 0.08656716417910448,
      "grad_norm": 1.9979189088773692,
      "learning_rate": 9.924038765061042e-06,
      "loss": 1.0134,
      "step": 29
    },
    {
      "epoch": 0.08955223880597014,
      "grad_norm": 2.1566867938814784,
      "learning_rate": 9.915388724114301e-06,
      "loss": 0.937,
      "step": 30
    },
    {
      "epoch": 0.09253731343283582,
      "grad_norm": 2.0548646219588886,
      "learning_rate": 9.906276553136924e-06,
      "loss": 1.0227,
      "step": 31
    },
    {
      "epoch": 0.0955223880597015,
      "grad_norm": 2.1094100666663493,
      "learning_rate": 9.896703108827758e-06,
      "loss": 0.9474,
      "step": 32
    },
    {
      "epoch": 0.09850746268656717,
      "grad_norm": 2.196878885778515,
      "learning_rate": 9.886669291253178e-06,
      "loss": 0.8954,
      "step": 33
    },
    {
      "epoch": 0.10149253731343283,
      "grad_norm": 1.9709535105103204,
      "learning_rate": 9.876176043762467e-06,
      "loss": 0.8858,
      "step": 34
    },
    {
      "epoch": 0.1044776119402985,
      "grad_norm": 2.4157359301325028,
      "learning_rate": 9.86522435289912e-06,
      "loss": 0.9492,
      "step": 35
    },
    {
      "epoch": 0.10746268656716418,
      "grad_norm": 2.0251547989723453,
      "learning_rate": 9.853815248308101e-06,
      "loss": 0.9819,
      "step": 36
    },
    {
      "epoch": 0.11044776119402985,
      "grad_norm": 2.043073108115636,
      "learning_rate": 9.841949802639031e-06,
      "loss": 0.9856,
      "step": 37
    },
    {
      "epoch": 0.11343283582089553,
      "grad_norm": 1.9051508783430386,
      "learning_rate": 9.829629131445342e-06,
      "loss": 0.8811,
      "step": 38
    },
    {
      "epoch": 0.11641791044776119,
      "grad_norm": 1.9720465747167384,
      "learning_rate": 9.816854393079402e-06,
      "loss": 0.8809,
      "step": 39
    },
    {
      "epoch": 0.11940298507462686,
      "grad_norm": 2.115788113435234,
      "learning_rate": 9.803626788583603e-06,
      "loss": 0.8754,
      "step": 40
    },
    {
      "epoch": 0.12238805970149254,
      "grad_norm": 2.0483615805890003,
      "learning_rate": 9.789947561577445e-06,
      "loss": 0.9111,
      "step": 41
    },
    {
      "epoch": 0.1253731343283582,
      "grad_norm": 2.0931565027152925,
      "learning_rate": 9.775817998140615e-06,
      "loss": 0.883,
      "step": 42
    },
    {
      "epoch": 0.12835820895522387,
      "grad_norm": 2.1619767559285896,
      "learning_rate": 9.761239426692077e-06,
      "loss": 0.8859,
      "step": 43
    },
    {
      "epoch": 0.13134328358208955,
      "grad_norm": 2.1803394559484515,
      "learning_rate": 9.74621321786517e-06,
      "loss": 0.9756,
      "step": 44
    },
    {
      "epoch": 0.13432835820895522,
      "grad_norm": 1.9473692197801435,
      "learning_rate": 9.730740784378755e-06,
      "loss": 0.8856,
      "step": 45
    },
    {
      "epoch": 0.1373134328358209,
      "grad_norm": 1.9440539475047278,
      "learning_rate": 9.71482358090438e-06,
      "loss": 0.8619,
      "step": 46
    },
    {
      "epoch": 0.14029850746268657,
      "grad_norm": 1.9268937369891654,
      "learning_rate": 9.698463103929542e-06,
      "loss": 0.8958,
      "step": 47
    },
    {
      "epoch": 0.14328358208955225,
      "grad_norm": 1.9576735974697994,
      "learning_rate": 9.681660891616967e-06,
      "loss": 0.9014,
      "step": 48
    },
    {
      "epoch": 0.14626865671641792,
      "grad_norm": 2.2502624703010787,
      "learning_rate": 9.664418523660004e-06,
      "loss": 0.9054,
      "step": 49
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 1.9625578640953845,
      "learning_rate": 9.646737621134112e-06,
      "loss": 0.9438,
      "step": 50
    },
    {
      "epoch": 0.15223880597014924,
      "grad_norm": 2.11803005668703,
      "learning_rate": 9.628619846344453e-06,
      "loss": 0.933,
      "step": 51
    },
    {
      "epoch": 0.15522388059701492,
      "grad_norm": 1.9405175828921,
      "learning_rate": 9.610066902669593e-06,
      "loss": 0.892,
      "step": 52
    },
    {
      "epoch": 0.1582089552238806,
      "grad_norm": 1.940879796450495,
      "learning_rate": 9.591080534401371e-06,
      "loss": 0.8912,
      "step": 53
    },
    {
      "epoch": 0.16119402985074627,
      "grad_norm": 2.1315443999032726,
      "learning_rate": 9.571662526580898e-06,
      "loss": 0.8837,
      "step": 54
    },
    {
      "epoch": 0.16417910447761194,
      "grad_norm": 1.9936535994324662,
      "learning_rate": 9.551814704830734e-06,
      "loss": 0.936,
      "step": 55
    },
    {
      "epoch": 0.16716417910447762,
      "grad_norm": 2.119381573598528,
      "learning_rate": 9.531538935183252e-06,
      "loss": 0.8668,
      "step": 56
    },
    {
      "epoch": 0.1701492537313433,
      "grad_norm": 2.0767671264680665,
      "learning_rate": 9.51083712390519e-06,
      "loss": 0.9144,
      "step": 57
    },
    {
      "epoch": 0.17313432835820897,
      "grad_norm": 1.7259669938296525,
      "learning_rate": 9.48971121731844e-06,
      "loss": 0.7978,
      "step": 58
    },
    {
      "epoch": 0.1761194029850746,
      "grad_norm": 2.2712488027683895,
      "learning_rate": 9.468163201617063e-06,
      "loss": 0.8566,
      "step": 59
    },
    {
      "epoch": 0.1791044776119403,
      "grad_norm": 2.1653559706182692,
      "learning_rate": 9.446195102680531e-06,
      "loss": 0.8772,
      "step": 60
    },
    {
      "epoch": 0.18208955223880596,
      "grad_norm": 2.019437924263334,
      "learning_rate": 9.423808985883289e-06,
      "loss": 0.8102,
      "step": 61
    },
    {
      "epoch": 0.18507462686567164,
      "grad_norm": 2.1016655793843895,
      "learning_rate": 9.401006955900555e-06,
      "loss": 0.8829,
      "step": 62
    },
    {
      "epoch": 0.1880597014925373,
      "grad_norm": 1.7789596254753794,
      "learning_rate": 9.377791156510456e-06,
      "loss": 0.8882,
      "step": 63
    },
    {
      "epoch": 0.191044776119403,
      "grad_norm": 3.956571817793095,
      "learning_rate": 9.35416377039246e-06,
      "loss": 0.9216,
      "step": 64
    },
    {
      "epoch": 0.19402985074626866,
      "grad_norm": 2.0814813313495764,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.8714,
      "step": 65
    },
    {
      "epoch": 0.19701492537313434,
      "grad_norm": 2.16440601583317,
      "learning_rate": 9.305683161962569e-06,
      "loss": 0.9033,
      "step": 66
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.121322858320113,
      "learning_rate": 9.280834497651334e-06,
      "loss": 0.8866,
      "step": 67
    },
    {
      "epoch": 0.20298507462686566,
      "grad_norm": 2.125810496652349,
      "learning_rate": 9.255583362184998e-06,
      "loss": 0.9084,
      "step": 68
    },
    {
      "epoch": 0.20597014925373133,
      "grad_norm": 2.2332151510392926,
      "learning_rate": 9.229932129599206e-06,
      "loss": 0.8995,
      "step": 69
    },
    {
      "epoch": 0.208955223880597,
      "grad_norm": 1.8844452106307579,
      "learning_rate": 9.203883211545517e-06,
      "loss": 0.8985,
      "step": 70
    },
    {
      "epoch": 0.21194029850746268,
      "grad_norm": 2.1170173520344253,
      "learning_rate": 9.177439057064684e-06,
      "loss": 0.8688,
      "step": 71
    },
    {
      "epoch": 0.21492537313432836,
      "grad_norm": 1.9805965194267483,
      "learning_rate": 9.150602152356394e-06,
      "loss": 0.8909,
      "step": 72
    },
    {
      "epoch": 0.21791044776119403,
      "grad_norm": 2.1059921689798116,
      "learning_rate": 9.123375020545534e-06,
      "loss": 0.9781,
      "step": 73
    },
    {
      "epoch": 0.2208955223880597,
      "grad_norm": 1.9571325614730988,
      "learning_rate": 9.09576022144496e-06,
      "loss": 0.8519,
      "step": 74
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 2.1730772437003867,
      "learning_rate": 9.067760351314838e-06,
      "loss": 0.8709,
      "step": 75
    },
    {
      "epoch": 0.22686567164179106,
      "grad_norm": 2.236623739502403,
      "learning_rate": 9.039378042618556e-06,
      "loss": 1.0019,
      "step": 76
    },
    {
      "epoch": 0.2298507462686567,
      "grad_norm": 2.049051045380969,
      "learning_rate": 9.01061596377522e-06,
      "loss": 0.8222,
      "step": 77
    },
    {
      "epoch": 0.23283582089552238,
      "grad_norm": 2.0708856760876833,
      "learning_rate": 8.981476818908778e-06,
      "loss": 0.9427,
      "step": 78
    },
    {
      "epoch": 0.23582089552238805,
      "grad_norm": 1.796129619411125,
      "learning_rate": 8.951963347593797e-06,
      "loss": 0.8325,
      "step": 79
    },
    {
      "epoch": 0.23880597014925373,
      "grad_norm": 2.1961650618323687,
      "learning_rate": 8.92207832459788e-06,
      "loss": 0.9022,
      "step": 80
    },
    {
      "epoch": 0.2417910447761194,
      "grad_norm": 2.1631152148881023,
      "learning_rate": 8.891824559620801e-06,
      "loss": 0.8278,
      "step": 81
    },
    {
      "epoch": 0.24477611940298508,
      "grad_norm": 2.025147970703757,
      "learning_rate": 8.861204897030346e-06,
      "loss": 0.7237,
      "step": 82
    },
    {
      "epoch": 0.24776119402985075,
      "grad_norm": 1.9375472063035821,
      "learning_rate": 8.83022221559489e-06,
      "loss": 0.7376,
      "step": 83
    },
    {
      "epoch": 0.2507462686567164,
      "grad_norm": 1.9368693802128247,
      "learning_rate": 8.798879428212748e-06,
      "loss": 0.8809,
      "step": 84
    },
    {
      "epoch": 0.2537313432835821,
      "grad_norm": 1.7919474666326674,
      "learning_rate": 8.767179481638303e-06,
      "loss": 0.827,
      "step": 85
    },
    {
      "epoch": 0.25671641791044775,
      "grad_norm": 1.8293764770314178,
      "learning_rate": 8.735125356204982e-06,
      "loss": 0.8348,
      "step": 86
    },
    {
      "epoch": 0.25970149253731345,
      "grad_norm": 2.2466601113818983,
      "learning_rate": 8.702720065545024e-06,
      "loss": 0.7715,
      "step": 87
    },
    {
      "epoch": 0.2626865671641791,
      "grad_norm": 2.007946173229526,
      "learning_rate": 8.669966656306176e-06,
      "loss": 0.7911,
      "step": 88
    },
    {
      "epoch": 0.2656716417910448,
      "grad_norm": 1.8905119743926408,
      "learning_rate": 8.636868207865244e-06,
      "loss": 0.8525,
      "step": 89
    },
    {
      "epoch": 0.26865671641791045,
      "grad_norm": 2.216132567769806,
      "learning_rate": 8.603427832038574e-06,
      "loss": 0.8743,
      "step": 90
    },
    {
      "epoch": 0.2716417910447761,
      "grad_norm": 2.0337720619135142,
      "learning_rate": 8.569648672789496e-06,
      "loss": 0.8652,
      "step": 91
    },
    {
      "epoch": 0.2746268656716418,
      "grad_norm": 2.117953862074952,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.8499,
      "step": 92
    },
    {
      "epoch": 0.27761194029850744,
      "grad_norm": 2.013179581976092,
      "learning_rate": 8.501086738835843e-06,
      "loss": 0.8635,
      "step": 93
    },
    {
      "epoch": 0.28059701492537314,
      "grad_norm": 2.148194857680393,
      "learning_rate": 8.466310410117622e-06,
      "loss": 0.8781,
      "step": 94
    },
    {
      "epoch": 0.2835820895522388,
      "grad_norm": 2.1704401998940024,
      "learning_rate": 8.43120818934367e-06,
      "loss": 0.8618,
      "step": 95
    },
    {
      "epoch": 0.2865671641791045,
      "grad_norm": 2.065149455076995,
      "learning_rate": 8.395783376718967e-06,
      "loss": 0.8133,
      "step": 96
    },
    {
      "epoch": 0.28955223880597014,
      "grad_norm": 2.0953263142220693,
      "learning_rate": 8.360039302777614e-06,
      "loss": 0.8189,
      "step": 97
    },
    {
      "epoch": 0.29253731343283584,
      "grad_norm": 1.8700626853391247,
      "learning_rate": 8.323979328069689e-06,
      "loss": 0.8338,
      "step": 98
    },
    {
      "epoch": 0.2955223880597015,
      "grad_norm": 1.886379993426298,
      "learning_rate": 8.28760684284532e-06,
      "loss": 0.8397,
      "step": 99
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 4.887769993107663,
      "learning_rate": 8.25092526673592e-06,
      "loss": 0.9114,
      "step": 100
    },
    {
      "epoch": 0.30149253731343284,
      "grad_norm": 2.1831478421086237,
      "learning_rate": 8.213938048432697e-06,
      "loss": 0.838,
      "step": 101
    },
    {
      "epoch": 0.3044776119402985,
      "grad_norm": 2.048417012849108,
      "learning_rate": 8.176648665362426e-06,
      "loss": 0.7323,
      "step": 102
    },
    {
      "epoch": 0.3074626865671642,
      "grad_norm": 2.245486182262689,
      "learning_rate": 8.139060623360494e-06,
      "loss": 0.8361,
      "step": 103
    },
    {
      "epoch": 0.31044776119402984,
      "grad_norm": 2.1131445651881937,
      "learning_rate": 8.101177456341301e-06,
      "loss": 0.8112,
      "step": 104
    },
    {
      "epoch": 0.31343283582089554,
      "grad_norm": 1.9438690743970823,
      "learning_rate": 8.063002725966014e-06,
      "loss": 0.8172,
      "step": 105
    },
    {
      "epoch": 0.3164179104477612,
      "grad_norm": 1.8330162019695133,
      "learning_rate": 8.024540021307709e-06,
      "loss": 0.7779,
      "step": 106
    },
    {
      "epoch": 0.3194029850746269,
      "grad_norm": 1.9694508107519921,
      "learning_rate": 7.985792958513932e-06,
      "loss": 0.8582,
      "step": 107
    },
    {
      "epoch": 0.32238805970149254,
      "grad_norm": 1.9336855933985957,
      "learning_rate": 7.946765180466725e-06,
      "loss": 0.7725,
      "step": 108
    },
    {
      "epoch": 0.3253731343283582,
      "grad_norm": 2.0505954912615127,
      "learning_rate": 7.907460356440133e-06,
      "loss": 0.8778,
      "step": 109
    },
    {
      "epoch": 0.3283582089552239,
      "grad_norm": 2.2263425751875183,
      "learning_rate": 7.86788218175523e-06,
      "loss": 0.8253,
      "step": 110
    },
    {
      "epoch": 0.33134328358208953,
      "grad_norm": 1.8159778540989733,
      "learning_rate": 7.828034377432694e-06,
      "loss": 0.8297,
      "step": 111
    },
    {
      "epoch": 0.33432835820895523,
      "grad_norm": 1.9962851923382634,
      "learning_rate": 7.787920689842965e-06,
      "loss": 0.8761,
      "step": 112
    },
    {
      "epoch": 0.3373134328358209,
      "grad_norm": 1.978659019282353,
      "learning_rate": 7.747544890354031e-06,
      "loss": 0.8364,
      "step": 113
    },
    {
      "epoch": 0.3402985074626866,
      "grad_norm": 1.8570435757482742,
      "learning_rate": 7.706910774976849e-06,
      "loss": 0.7402,
      "step": 114
    },
    {
      "epoch": 0.34328358208955223,
      "grad_norm": 1.9873552845505367,
      "learning_rate": 7.666022164008458e-06,
      "loss": 0.8159,
      "step": 115
    },
    {
      "epoch": 0.34626865671641793,
      "grad_norm": 1.8774777253882997,
      "learning_rate": 7.624882901672801e-06,
      "loss": 0.8035,
      "step": 116
    },
    {
      "epoch": 0.3492537313432836,
      "grad_norm": 2.042368536282115,
      "learning_rate": 7.5834968557593155e-06,
      "loss": 0.8852,
      "step": 117
    },
    {
      "epoch": 0.3522388059701492,
      "grad_norm": 1.9585431439709604,
      "learning_rate": 7.541867917259278e-06,
      "loss": 0.9395,
      "step": 118
    },
    {
      "epoch": 0.35522388059701493,
      "grad_norm": 1.9690363236772885,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.7769,
      "step": 119
    },
    {
      "epoch": 0.3582089552238806,
      "grad_norm": 1.9997251291932896,
      "learning_rate": 7.457897040276853e-06,
      "loss": 0.8867,
      "step": 120
    },
    {
      "epoch": 0.3611940298507463,
      "grad_norm": 1.9091614861364634,
      "learning_rate": 7.415562996483193e-06,
      "loss": 0.8203,
      "step": 121
    },
    {
      "epoch": 0.3641791044776119,
      "grad_norm": 2.0154748461484964,
      "learning_rate": 7.373001848738203e-06,
      "loss": 0.907,
      "step": 122
    },
    {
      "epoch": 0.36716417910447763,
      "grad_norm": 1.8850762145928235,
      "learning_rate": 7.330217598512696e-06,
      "loss": 0.7268,
      "step": 123
    },
    {
      "epoch": 0.3701492537313433,
      "grad_norm": 2.002378822284434,
      "learning_rate": 7.2872142682529045e-06,
      "loss": 0.75,
      "step": 124
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 1.8683346372861278,
      "learning_rate": 7.243995901002312e-06,
      "loss": 0.7998,
      "step": 125
    },
    {
      "epoch": 0.3761194029850746,
      "grad_norm": 1.8603961929214259,
      "learning_rate": 7.200566560021525e-06,
      "loss": 0.7736,
      "step": 126
    },
    {
      "epoch": 0.37910447761194027,
      "grad_norm": 2.086337061985904,
      "learning_rate": 7.156930328406268e-06,
      "loss": 0.8291,
      "step": 127
    },
    {
      "epoch": 0.382089552238806,
      "grad_norm": 2.0488227138684203,
      "learning_rate": 7.113091308703498e-06,
      "loss": 0.8032,
      "step": 128
    },
    {
      "epoch": 0.3850746268656716,
      "grad_norm": 2.2156317114896504,
      "learning_rate": 7.069053622525697e-06,
      "loss": 0.9299,
      "step": 129
    },
    {
      "epoch": 0.3880597014925373,
      "grad_norm": 1.959090575118802,
      "learning_rate": 7.0248214101633685e-06,
      "loss": 0.7743,
      "step": 130
    },
    {
      "epoch": 0.39104477611940297,
      "grad_norm": 2.040790454179876,
      "learning_rate": 6.980398830195785e-06,
      "loss": 0.9517,
      "step": 131
    },
    {
      "epoch": 0.3940298507462687,
      "grad_norm": 2.0095877017198394,
      "learning_rate": 6.9357900591000034e-06,
      "loss": 0.8367,
      "step": 132
    },
    {
      "epoch": 0.3970149253731343,
      "grad_norm": 2.053685963143701,
      "learning_rate": 6.890999290858213e-06,
      "loss": 0.908,
      "step": 133
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.9552949066858203,
      "learning_rate": 6.8460307365634225e-06,
      "loss": 0.7769,
      "step": 134
    },
    {
      "epoch": 0.40298507462686567,
      "grad_norm": 1.9422682262777784,
      "learning_rate": 6.800888624023552e-06,
      "loss": 0.8613,
      "step": 135
    },
    {
      "epoch": 0.4059701492537313,
      "grad_norm": 2.0018278352562784,
      "learning_rate": 6.755577197363945e-06,
      "loss": 0.7314,
      "step": 136
    },
    {
      "epoch": 0.408955223880597,
      "grad_norm": 2.0748265555866023,
      "learning_rate": 6.710100716628345e-06,
      "loss": 0.7806,
      "step": 137
    },
    {
      "epoch": 0.41194029850746267,
      "grad_norm": 1.9674200834017632,
      "learning_rate": 6.6644634573783825e-06,
      "loss": 0.8074,
      "step": 138
    },
    {
      "epoch": 0.41492537313432837,
      "grad_norm": 1.9207512896621706,
      "learning_rate": 6.618669710291607e-06,
      "loss": 0.7148,
      "step": 139
    },
    {
      "epoch": 0.417910447761194,
      "grad_norm": 1.905193379967269,
      "learning_rate": 6.572723780758069e-06,
      "loss": 0.8142,
      "step": 140
    },
    {
      "epoch": 0.4208955223880597,
      "grad_norm": 2.1965158090202697,
      "learning_rate": 6.526629988475567e-06,
      "loss": 0.8123,
      "step": 141
    },
    {
      "epoch": 0.42388059701492536,
      "grad_norm": 1.8753650627785756,
      "learning_rate": 6.4803926670435e-06,
      "loss": 0.8388,
      "step": 142
    },
    {
      "epoch": 0.42686567164179107,
      "grad_norm": 2.0267686855160156,
      "learning_rate": 6.434016163555452e-06,
      "loss": 0.8065,
      "step": 143
    },
    {
      "epoch": 0.4298507462686567,
      "grad_norm": 1.9454127213119508,
      "learning_rate": 6.387504838190479e-06,
      "loss": 0.7268,
      "step": 144
    },
    {
      "epoch": 0.43283582089552236,
      "grad_norm": 1.9502519457009935,
      "learning_rate": 6.340863063803187e-06,
      "loss": 0.744,
      "step": 145
    },
    {
      "epoch": 0.43582089552238806,
      "grad_norm": 1.9053847115703428,
      "learning_rate": 6.294095225512604e-06,
      "loss": 0.8448,
      "step": 146
    },
    {
      "epoch": 0.4388059701492537,
      "grad_norm": 2.099027132432326,
      "learning_rate": 6.247205720289907e-06,
      "loss": 0.9298,
      "step": 147
    },
    {
      "epoch": 0.4417910447761194,
      "grad_norm": 2.0781658570138153,
      "learning_rate": 6.2001989565450305e-06,
      "loss": 0.7678,
      "step": 148
    },
    {
      "epoch": 0.44477611940298506,
      "grad_norm": 2.017544918765642,
      "learning_rate": 6.153079353712201e-06,
      "loss": 0.7815,
      "step": 149
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 2.0613800967577203,
      "learning_rate": 6.105851341834439e-06,
      "loss": 0.7604,
      "step": 150
    },
    {
      "epoch": 0.4507462686567164,
      "grad_norm": 1.8534108525240118,
      "learning_rate": 6.058519361147055e-06,
      "loss": 0.7744,
      "step": 151
    },
    {
      "epoch": 0.4537313432835821,
      "grad_norm": 2.0589449906784525,
      "learning_rate": 6.011087861660191e-06,
      "loss": 0.7458,
      "step": 152
    },
    {
      "epoch": 0.45671641791044776,
      "grad_norm": 1.7521074071886629,
      "learning_rate": 5.9635613027404495e-06,
      "loss": 0.7427,
      "step": 153
    },
    {
      "epoch": 0.4597014925373134,
      "grad_norm": 1.952094332549687,
      "learning_rate": 5.915944152691634e-06,
      "loss": 0.7228,
      "step": 154
    },
    {
      "epoch": 0.4626865671641791,
      "grad_norm": 2.0059139166291162,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.7888,
      "step": 155
    },
    {
      "epoch": 0.46567164179104475,
      "grad_norm": 1.9863550415015798,
      "learning_rate": 5.820455994586621e-06,
      "loss": 0.7431,
      "step": 156
    },
    {
      "epoch": 0.46865671641791046,
      "grad_norm": 2.186012143543713,
      "learning_rate": 5.772593964039203e-06,
      "loss": 0.9548,
      "step": 157
    },
    {
      "epoch": 0.4716417910447761,
      "grad_norm": 2.0569327383361524,
      "learning_rate": 5.724659296536234e-06,
      "loss": 0.7608,
      "step": 158
    },
    {
      "epoch": 0.4746268656716418,
      "grad_norm": 1.8427207627927604,
      "learning_rate": 5.6766564987506564e-06,
      "loss": 0.728,
      "step": 159
    },
    {
      "epoch": 0.47761194029850745,
      "grad_norm": 2.0454749932374874,
      "learning_rate": 5.628590083760815e-06,
      "loss": 0.7987,
      "step": 160
    },
    {
      "epoch": 0.48059701492537316,
      "grad_norm": 1.997088234967905,
      "learning_rate": 5.5804645706261515e-06,
      "loss": 0.87,
      "step": 161
    },
    {
      "epoch": 0.4835820895522388,
      "grad_norm": 2.0854705687983897,
      "learning_rate": 5.532284483962341e-06,
      "loss": 0.7816,
      "step": 162
    },
    {
      "epoch": 0.48656716417910445,
      "grad_norm": 2.063156543618824,
      "learning_rate": 5.484054353515896e-06,
      "loss": 0.7953,
      "step": 163
    },
    {
      "epoch": 0.48955223880597015,
      "grad_norm": 2.0601249949949025,
      "learning_rate": 5.435778713738292e-06,
      "loss": 0.8433,
      "step": 164
    },
    {
      "epoch": 0.4925373134328358,
      "grad_norm": 1.9082330488593775,
      "learning_rate": 5.387462103359655e-06,
      "loss": 0.78,
      "step": 165
    },
    {
      "epoch": 0.4955223880597015,
      "grad_norm": 1.861403808976405,
      "learning_rate": 5.339109064962047e-06,
      "loss": 0.6939,
      "step": 166
    },
    {
      "epoch": 0.49850746268656715,
      "grad_norm": 2.1165013908329393,
      "learning_rate": 5.290724144552379e-06,
      "loss": 0.6902,
      "step": 167
    },
    {
      "epoch": 0.5014925373134328,
      "grad_norm": 2.4674070182101544,
      "learning_rate": 5.242311891135016e-06,
      "loss": 0.8382,
      "step": 168
    },
    {
      "epoch": 0.5044776119402985,
      "grad_norm": 1.8675037579070943,
      "learning_rate": 5.193876856284085e-06,
      "loss": 0.7193,
      "step": 169
    },
    {
      "epoch": 0.5074626865671642,
      "grad_norm": 2.0252014042338415,
      "learning_rate": 5.145423593715558e-06,
      "loss": 0.8031,
      "step": 170
    },
    {
      "epoch": 0.5104477611940299,
      "grad_norm": 1.9108898103865828,
      "learning_rate": 5.096956658859122e-06,
      "loss": 0.7634,
      "step": 171
    },
    {
      "epoch": 0.5134328358208955,
      "grad_norm": 2.139726380565598,
      "learning_rate": 5.048480608429893e-06,
      "loss": 0.7139,
      "step": 172
    },
    {
      "epoch": 0.5164179104477612,
      "grad_norm": 1.9120901416594451,
      "learning_rate": 5e-06,
      "loss": 0.7729,
      "step": 173
    },
    {
      "epoch": 0.5194029850746269,
      "grad_norm": 1.766403288334722,
      "learning_rate": 4.951519391570108e-06,
      "loss": 0.7792,
      "step": 174
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 1.875054628847942,
      "learning_rate": 4.903043341140879e-06,
      "loss": 0.7026,
      "step": 175
    },
    {
      "epoch": 0.5253731343283582,
      "grad_norm": 2.0649412907897817,
      "learning_rate": 4.854576406284443e-06,
      "loss": 0.6678,
      "step": 176
    },
    {
      "epoch": 0.5283582089552239,
      "grad_norm": 1.9766676200083877,
      "learning_rate": 4.806123143715916e-06,
      "loss": 0.794,
      "step": 177
    },
    {
      "epoch": 0.5313432835820896,
      "grad_norm": 1.9628257930544168,
      "learning_rate": 4.7576881088649865e-06,
      "loss": 0.8011,
      "step": 178
    },
    {
      "epoch": 0.5343283582089552,
      "grad_norm": 1.993236983375051,
      "learning_rate": 4.7092758554476215e-06,
      "loss": 0.7871,
      "step": 179
    },
    {
      "epoch": 0.5373134328358209,
      "grad_norm": 1.8504281412049604,
      "learning_rate": 4.660890935037954e-06,
      "loss": 0.7736,
      "step": 180
    },
    {
      "epoch": 0.5402985074626866,
      "grad_norm": 1.9028018031804006,
      "learning_rate": 4.6125378966403465e-06,
      "loss": 0.7578,
      "step": 181
    },
    {
      "epoch": 0.5432835820895522,
      "grad_norm": 1.8659479024734795,
      "learning_rate": 4.564221286261709e-06,
      "loss": 0.7557,
      "step": 182
    },
    {
      "epoch": 0.5462686567164179,
      "grad_norm": 2.11712649211751,
      "learning_rate": 4.515945646484105e-06,
      "loss": 0.8088,
      "step": 183
    },
    {
      "epoch": 0.5492537313432836,
      "grad_norm": 1.789112706708441,
      "learning_rate": 4.467715516037659e-06,
      "loss": 0.8087,
      "step": 184
    },
    {
      "epoch": 0.5522388059701493,
      "grad_norm": 1.9892189797387887,
      "learning_rate": 4.4195354293738484e-06,
      "loss": 0.8395,
      "step": 185
    },
    {
      "epoch": 0.5552238805970149,
      "grad_norm": 2.1883581521829316,
      "learning_rate": 4.371409916239188e-06,
      "loss": 0.7684,
      "step": 186
    },
    {
      "epoch": 0.5582089552238806,
      "grad_norm": 2.1631251071113176,
      "learning_rate": 4.323343501249346e-06,
      "loss": 0.736,
      "step": 187
    },
    {
      "epoch": 0.5611940298507463,
      "grad_norm": 1.9289233899504126,
      "learning_rate": 4.275340703463767e-06,
      "loss": 0.6405,
      "step": 188
    },
    {
      "epoch": 0.564179104477612,
      "grad_norm": 1.9857433587359077,
      "learning_rate": 4.227406035960798e-06,
      "loss": 0.7959,
      "step": 189
    },
    {
      "epoch": 0.5671641791044776,
      "grad_norm": 1.8258458953238406,
      "learning_rate": 4.17954400541338e-06,
      "loss": 0.7396,
      "step": 190
    },
    {
      "epoch": 0.5701492537313433,
      "grad_norm": 1.9073211443331697,
      "learning_rate": 4.131759111665349e-06,
      "loss": 0.7894,
      "step": 191
    },
    {
      "epoch": 0.573134328358209,
      "grad_norm": 2.1144318330609955,
      "learning_rate": 4.084055847308367e-06,
      "loss": 0.7564,
      "step": 192
    },
    {
      "epoch": 0.5761194029850746,
      "grad_norm": 1.9704301567554263,
      "learning_rate": 4.036438697259551e-06,
      "loss": 0.768,
      "step": 193
    },
    {
      "epoch": 0.5791044776119403,
      "grad_norm": 2.641711220430738,
      "learning_rate": 3.988912138339812e-06,
      "loss": 0.7862,
      "step": 194
    },
    {
      "epoch": 0.582089552238806,
      "grad_norm": 2.1626033614930247,
      "learning_rate": 3.941480638852948e-06,
      "loss": 0.7856,
      "step": 195
    },
    {
      "epoch": 0.5850746268656717,
      "grad_norm": 1.9284923265396965,
      "learning_rate": 3.894148658165562e-06,
      "loss": 0.7674,
      "step": 196
    },
    {
      "epoch": 0.5880597014925373,
      "grad_norm": 2.058823111277626,
      "learning_rate": 3.8469206462878e-06,
      "loss": 0.7663,
      "step": 197
    },
    {
      "epoch": 0.591044776119403,
      "grad_norm": 2.1351322093158744,
      "learning_rate": 3.7998010434549716e-06,
      "loss": 0.8399,
      "step": 198
    },
    {
      "epoch": 0.5940298507462687,
      "grad_norm": 1.9380000299821818,
      "learning_rate": 3.752794279710094e-06,
      "loss": 0.7321,
      "step": 199
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 1.8441505158750409,
      "learning_rate": 3.705904774487396e-06,
      "loss": 0.774,
      "step": 200
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.0907330198564313,
      "learning_rate": 3.6591369361968127e-06,
      "loss": 0.7847,
      "step": 201
    },
    {
      "epoch": 0.6029850746268657,
      "grad_norm": 1.9371066770940577,
      "learning_rate": 3.6124951618095224e-06,
      "loss": 0.748,
      "step": 202
    },
    {
      "epoch": 0.6059701492537314,
      "grad_norm": 1.8374945672921557,
      "learning_rate": 3.5659838364445505e-06,
      "loss": 0.7537,
      "step": 203
    },
    {
      "epoch": 0.608955223880597,
      "grad_norm": 2.081262815281359,
      "learning_rate": 3.519607332956502e-06,
      "loss": 0.815,
      "step": 204
    },
    {
      "epoch": 0.6119402985074627,
      "grad_norm": 2.074788649262812,
      "learning_rate": 3.473370011524435e-06,
      "loss": 0.8674,
      "step": 205
    },
    {
      "epoch": 0.6149253731343284,
      "grad_norm": 1.8698000805877122,
      "learning_rate": 3.427276219241933e-06,
      "loss": 0.7202,
      "step": 206
    },
    {
      "epoch": 0.6179104477611941,
      "grad_norm": 1.9048492100127103,
      "learning_rate": 3.3813302897083955e-06,
      "loss": 0.8188,
      "step": 207
    },
    {
      "epoch": 0.6208955223880597,
      "grad_norm": 1.979184398237071,
      "learning_rate": 3.335536542621617e-06,
      "loss": 0.7503,
      "step": 208
    },
    {
      "epoch": 0.6238805970149254,
      "grad_norm": 1.9760275557890097,
      "learning_rate": 3.289899283371657e-06,
      "loss": 0.7997,
      "step": 209
    },
    {
      "epoch": 0.6268656716417911,
      "grad_norm": 2.1124769220840944,
      "learning_rate": 3.244422802636057e-06,
      "loss": 0.831,
      "step": 210
    },
    {
      "epoch": 0.6298507462686567,
      "grad_norm": 2.034333485591399,
      "learning_rate": 3.1991113759764493e-06,
      "loss": 0.7732,
      "step": 211
    },
    {
      "epoch": 0.6328358208955224,
      "grad_norm": 1.897048944733691,
      "learning_rate": 3.1539692634365788e-06,
      "loss": 0.754,
      "step": 212
    },
    {
      "epoch": 0.6358208955223881,
      "grad_norm": 1.9618035338024724,
      "learning_rate": 3.1090007091417884e-06,
      "loss": 0.6767,
      "step": 213
    },
    {
      "epoch": 0.6388059701492538,
      "grad_norm": 2.0148083710716764,
      "learning_rate": 3.0642099408999982e-06,
      "loss": 0.8271,
      "step": 214
    },
    {
      "epoch": 0.6417910447761194,
      "grad_norm": 2.283043013050116,
      "learning_rate": 3.019601169804216e-06,
      "loss": 0.7971,
      "step": 215
    },
    {
      "epoch": 0.6447761194029851,
      "grad_norm": 2.11854512509762,
      "learning_rate": 2.975178589836632e-06,
      "loss": 0.7609,
      "step": 216
    },
    {
      "epoch": 0.6477611940298508,
      "grad_norm": 1.8298105473763824,
      "learning_rate": 2.9309463774743047e-06,
      "loss": 0.8506,
      "step": 217
    },
    {
      "epoch": 0.6507462686567164,
      "grad_norm": 1.7780813510143145,
      "learning_rate": 2.886908691296504e-06,
      "loss": 0.7632,
      "step": 218
    },
    {
      "epoch": 0.6537313432835821,
      "grad_norm": 1.9126798345749187,
      "learning_rate": 2.843069671593734e-06,
      "loss": 0.7782,
      "step": 219
    },
    {
      "epoch": 0.6567164179104478,
      "grad_norm": 1.9350084312265137,
      "learning_rate": 2.7994334399784773e-06,
      "loss": 0.7372,
      "step": 220
    },
    {
      "epoch": 0.6597014925373135,
      "grad_norm": 1.9251939137126606,
      "learning_rate": 2.7560040989976894e-06,
      "loss": 0.7357,
      "step": 221
    },
    {
      "epoch": 0.6626865671641791,
      "grad_norm": 1.9256071124163803,
      "learning_rate": 2.7127857317470967e-06,
      "loss": 0.7855,
      "step": 222
    },
    {
      "epoch": 0.6656716417910448,
      "grad_norm": 2.0598971965285338,
      "learning_rate": 2.6697824014873076e-06,
      "loss": 0.7324,
      "step": 223
    },
    {
      "epoch": 0.6686567164179105,
      "grad_norm": 1.823285258131258,
      "learning_rate": 2.626998151261798e-06,
      "loss": 0.6943,
      "step": 224
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 2.7227864446392678,
      "learning_rate": 2.5844370035168077e-06,
      "loss": 0.7098,
      "step": 225
    },
    {
      "epoch": 0.6746268656716418,
      "grad_norm": 2.1471535886447035,
      "learning_rate": 2.5421029597231476e-06,
      "loss": 0.7776,
      "step": 226
    },
    {
      "epoch": 0.6776119402985075,
      "grad_norm": 2.429389297486461,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.7514,
      "step": 227
    },
    {
      "epoch": 0.6805970149253732,
      "grad_norm": 2.0726794837198725,
      "learning_rate": 2.458132082740724e-06,
      "loss": 0.7712,
      "step": 228
    },
    {
      "epoch": 0.6835820895522388,
      "grad_norm": 2.1998151305942395,
      "learning_rate": 2.4165031442406857e-06,
      "loss": 0.7435,
      "step": 229
    },
    {
      "epoch": 0.6865671641791045,
      "grad_norm": 2.02867494365551,
      "learning_rate": 2.3751170983272e-06,
      "loss": 0.8134,
      "step": 230
    },
    {
      "epoch": 0.6895522388059702,
      "grad_norm": 1.9535756316081252,
      "learning_rate": 2.333977835991545e-06,
      "loss": 0.7191,
      "step": 231
    },
    {
      "epoch": 0.6925373134328359,
      "grad_norm": 2.0484292921284477,
      "learning_rate": 2.293089225023152e-06,
      "loss": 0.8623,
      "step": 232
    },
    {
      "epoch": 0.6955223880597015,
      "grad_norm": 2.0146733780323505,
      "learning_rate": 2.2524551096459703e-06,
      "loss": 0.7654,
      "step": 233
    },
    {
      "epoch": 0.6985074626865672,
      "grad_norm": 2.007942017411499,
      "learning_rate": 2.2120793101570366e-06,
      "loss": 0.7197,
      "step": 234
    },
    {
      "epoch": 0.7014925373134329,
      "grad_norm": 1.9512262040861592,
      "learning_rate": 2.171965622567308e-06,
      "loss": 0.795,
      "step": 235
    },
    {
      "epoch": 0.7044776119402985,
      "grad_norm": 2.0223484864555608,
      "learning_rate": 2.132117818244771e-06,
      "loss": 0.73,
      "step": 236
    },
    {
      "epoch": 0.7074626865671642,
      "grad_norm": 1.9076970970748008,
      "learning_rate": 2.0925396435598665e-06,
      "loss": 0.7341,
      "step": 237
    },
    {
      "epoch": 0.7104477611940299,
      "grad_norm": 2.0099705479490084,
      "learning_rate": 2.053234819533276e-06,
      "loss": 0.706,
      "step": 238
    },
    {
      "epoch": 0.7134328358208956,
      "grad_norm": 2.088602599462484,
      "learning_rate": 2.0142070414860704e-06,
      "loss": 0.8061,
      "step": 239
    },
    {
      "epoch": 0.7164179104477612,
      "grad_norm": 2.1414530746534437,
      "learning_rate": 1.9754599786922913e-06,
      "loss": 0.7735,
      "step": 240
    },
    {
      "epoch": 0.7194029850746269,
      "grad_norm": 1.756460059490162,
      "learning_rate": 1.936997274033986e-06,
      "loss": 0.652,
      "step": 241
    },
    {
      "epoch": 0.7223880597014926,
      "grad_norm": 1.6835259072072402,
      "learning_rate": 1.8988225436587005e-06,
      "loss": 0.6575,
      "step": 242
    },
    {
      "epoch": 0.7253731343283583,
      "grad_norm": 1.9338570463688378,
      "learning_rate": 1.8609393766395083e-06,
      "loss": 0.8504,
      "step": 243
    },
    {
      "epoch": 0.7283582089552239,
      "grad_norm": 2.0279105685865884,
      "learning_rate": 1.823351334637576e-06,
      "loss": 0.8141,
      "step": 244
    },
    {
      "epoch": 0.7313432835820896,
      "grad_norm": 1.9475282359983863,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 0.7989,
      "step": 245
    },
    {
      "epoch": 0.7343283582089553,
      "grad_norm": 2.0357960960016093,
      "learning_rate": 1.7490747332640833e-06,
      "loss": 0.8093,
      "step": 246
    },
    {
      "epoch": 0.7373134328358208,
      "grad_norm": 2.0420189686519072,
      "learning_rate": 1.7123931571546826e-06,
      "loss": 0.8056,
      "step": 247
    },
    {
      "epoch": 0.7402985074626866,
      "grad_norm": 1.7925260574915485,
      "learning_rate": 1.6760206719303107e-06,
      "loss": 0.7222,
      "step": 248
    },
    {
      "epoch": 0.7432835820895523,
      "grad_norm": 1.9144780544797566,
      "learning_rate": 1.639960697222388e-06,
      "loss": 0.8697,
      "step": 249
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 1.9085137418012865,
      "learning_rate": 1.6042166232810346e-06,
      "loss": 0.7573,
      "step": 250
    },
    {
      "epoch": 0.7492537313432835,
      "grad_norm": 1.7594094715163107,
      "learning_rate": 1.5687918106563326e-06,
      "loss": 0.719,
      "step": 251
    },
    {
      "epoch": 0.7522388059701492,
      "grad_norm": 1.9721722301749733,
      "learning_rate": 1.5336895898823801e-06,
      "loss": 0.805,
      "step": 252
    },
    {
      "epoch": 0.755223880597015,
      "grad_norm": 2.2257145387956148,
      "learning_rate": 1.4989132611641576e-06,
      "loss": 0.7464,
      "step": 253
    },
    {
      "epoch": 0.7582089552238805,
      "grad_norm": 2.2389879292796944,
      "learning_rate": 1.4644660940672628e-06,
      "loss": 0.7688,
      "step": 254
    },
    {
      "epoch": 0.7611940298507462,
      "grad_norm": 2.2600400159677445,
      "learning_rate": 1.4303513272105057e-06,
      "loss": 0.6961,
      "step": 255
    },
    {
      "epoch": 0.764179104477612,
      "grad_norm": 1.9828908431355627,
      "learning_rate": 1.396572167961427e-06,
      "loss": 0.8139,
      "step": 256
    },
    {
      "epoch": 0.7671641791044777,
      "grad_norm": 2.123631920520936,
      "learning_rate": 1.3631317921347564e-06,
      "loss": 0.7817,
      "step": 257
    },
    {
      "epoch": 0.7701492537313432,
      "grad_norm": 1.91427428961712,
      "learning_rate": 1.330033343693824e-06,
      "loss": 0.7411,
      "step": 258
    },
    {
      "epoch": 0.7731343283582089,
      "grad_norm": 1.923869516924642,
      "learning_rate": 1.297279934454978e-06,
      "loss": 0.7464,
      "step": 259
    },
    {
      "epoch": 0.7761194029850746,
      "grad_norm": 2.097695064363166,
      "learning_rate": 1.264874643795021e-06,
      "loss": 0.7835,
      "step": 260
    },
    {
      "epoch": 0.7791044776119403,
      "grad_norm": 1.866847274620371,
      "learning_rate": 1.2328205183616964e-06,
      "loss": 0.7576,
      "step": 261
    },
    {
      "epoch": 0.7820895522388059,
      "grad_norm": 1.9518448868919547,
      "learning_rate": 1.2011205717872538e-06,
      "loss": 0.8182,
      "step": 262
    },
    {
      "epoch": 0.7850746268656716,
      "grad_norm": 2.0133337408527483,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 0.7536,
      "step": 263
    },
    {
      "epoch": 0.7880597014925373,
      "grad_norm": 1.9817193200479692,
      "learning_rate": 1.1387951029696543e-06,
      "loss": 0.7337,
      "step": 264
    },
    {
      "epoch": 0.7910447761194029,
      "grad_norm": 1.9081666878508032,
      "learning_rate": 1.1081754403792e-06,
      "loss": 0.7247,
      "step": 265
    },
    {
      "epoch": 0.7940298507462686,
      "grad_norm": 1.7424632653916463,
      "learning_rate": 1.0779216754021215e-06,
      "loss": 0.7659,
      "step": 266
    },
    {
      "epoch": 0.7970149253731343,
      "grad_norm": 2.1707469563008694,
      "learning_rate": 1.0480366524062041e-06,
      "loss": 0.7172,
      "step": 267
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.9358665287041594,
      "learning_rate": 1.0185231810912223e-06,
      "loss": 0.7672,
      "step": 268
    },
    {
      "epoch": 0.8029850746268656,
      "grad_norm": 1.9633165798810948,
      "learning_rate": 9.893840362247809e-07,
      "loss": 0.7251,
      "step": 269
    },
    {
      "epoch": 0.8059701492537313,
      "grad_norm": 1.811709330190077,
      "learning_rate": 9.606219573814447e-07,
      "loss": 0.7109,
      "step": 270
    },
    {
      "epoch": 0.808955223880597,
      "grad_norm": 2.113644270821526,
      "learning_rate": 9.322396486851626e-07,
      "loss": 0.7905,
      "step": 271
    },
    {
      "epoch": 0.8119402985074626,
      "grad_norm": 1.827142174127426,
      "learning_rate": 9.042397785550405e-07,
      "loss": 0.7433,
      "step": 272
    },
    {
      "epoch": 0.8149253731343283,
      "grad_norm": 2.014601055555163,
      "learning_rate": 8.766249794544662e-07,
      "loss": 0.7755,
      "step": 273
    },
    {
      "epoch": 0.817910447761194,
      "grad_norm": 2.091449213566336,
      "learning_rate": 8.49397847643606e-07,
      "loss": 0.8266,
      "step": 274
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 2.3792124584042407,
      "learning_rate": 8.225609429353187e-07,
      "loss": 0.8152,
      "step": 275
    },
    {
      "epoch": 0.8238805970149253,
      "grad_norm": 2.236371795361171,
      "learning_rate": 7.961167884544852e-07,
      "loss": 0.8004,
      "step": 276
    },
    {
      "epoch": 0.826865671641791,
      "grad_norm": 1.8675998248443555,
      "learning_rate": 7.700678704007947e-07,
      "loss": 0.7616,
      "step": 277
    },
    {
      "epoch": 0.8298507462686567,
      "grad_norm": 1.8934578283049441,
      "learning_rate": 7.444166378150014e-07,
      "loss": 0.7384,
      "step": 278
    },
    {
      "epoch": 0.8328358208955224,
      "grad_norm": 1.806959761192126,
      "learning_rate": 7.191655023486682e-07,
      "loss": 0.7859,
      "step": 279
    },
    {
      "epoch": 0.835820895522388,
      "grad_norm": 2.0920960597937883,
      "learning_rate": 6.94316838037431e-07,
      "loss": 0.6582,
      "step": 280
    },
    {
      "epoch": 0.8388059701492537,
      "grad_norm": 1.9049893572652696,
      "learning_rate": 6.698729810778065e-07,
      "loss": 0.7133,
      "step": 281
    },
    {
      "epoch": 0.8417910447761194,
      "grad_norm": 2.021598834791704,
      "learning_rate": 6.458362296075399e-07,
      "loss": 0.6986,
      "step": 282
    },
    {
      "epoch": 0.844776119402985,
      "grad_norm": 1.8785681628556798,
      "learning_rate": 6.222088434895462e-07,
      "loss": 0.7561,
      "step": 283
    },
    {
      "epoch": 0.8477611940298507,
      "grad_norm": 1.7981294150989415,
      "learning_rate": 5.989930440994451e-07,
      "loss": 0.6988,
      "step": 284
    },
    {
      "epoch": 0.8507462686567164,
      "grad_norm": 1.9804075333202549,
      "learning_rate": 5.76191014116711e-07,
      "loss": 0.7793,
      "step": 285
    },
    {
      "epoch": 0.8537313432835821,
      "grad_norm": 1.8946196230697367,
      "learning_rate": 5.538048973194699e-07,
      "loss": 0.7378,
      "step": 286
    },
    {
      "epoch": 0.8567164179104477,
      "grad_norm": 4.308097949167914,
      "learning_rate": 5.318367983829393e-07,
      "loss": 0.7545,
      "step": 287
    },
    {
      "epoch": 0.8597014925373134,
      "grad_norm": 1.7900383798917394,
      "learning_rate": 5.102887826815589e-07,
      "loss": 0.6725,
      "step": 288
    },
    {
      "epoch": 0.8626865671641791,
      "grad_norm": 1.8973145322914773,
      "learning_rate": 4.891628760948114e-07,
      "loss": 0.6943,
      "step": 289
    },
    {
      "epoch": 0.8656716417910447,
      "grad_norm": 2.0037685592372516,
      "learning_rate": 4.6846106481675035e-07,
      "loss": 0.7904,
      "step": 290
    },
    {
      "epoch": 0.8686567164179104,
      "grad_norm": 1.833347945444079,
      "learning_rate": 4.481852951692672e-07,
      "loss": 0.6899,
      "step": 291
    },
    {
      "epoch": 0.8716417910447761,
      "grad_norm": 2.016728151056703,
      "learning_rate": 4.283374734191037e-07,
      "loss": 0.8379,
      "step": 292
    },
    {
      "epoch": 0.8746268656716418,
      "grad_norm": 2.034748614996797,
      "learning_rate": 4.089194655986306e-07,
      "loss": 0.6884,
      "step": 293
    },
    {
      "epoch": 0.8776119402985074,
      "grad_norm": 2.103995265841144,
      "learning_rate": 3.899330973304083e-07,
      "loss": 0.8119,
      "step": 294
    },
    {
      "epoch": 0.8805970149253731,
      "grad_norm": 2.085775862013919,
      "learning_rate": 3.7138015365554834e-07,
      "loss": 0.7218,
      "step": 295
    },
    {
      "epoch": 0.8835820895522388,
      "grad_norm": 1.859225098794404,
      "learning_rate": 3.5326237886588734e-07,
      "loss": 0.761,
      "step": 296
    },
    {
      "epoch": 0.8865671641791045,
      "grad_norm": 1.9937848736813224,
      "learning_rate": 3.355814763399973e-07,
      "loss": 0.7734,
      "step": 297
    },
    {
      "epoch": 0.8895522388059701,
      "grad_norm": 1.956060370495559,
      "learning_rate": 3.183391083830345e-07,
      "loss": 0.7522,
      "step": 298
    },
    {
      "epoch": 0.8925373134328358,
      "grad_norm": 2.1663087142629953,
      "learning_rate": 3.015368960704584e-07,
      "loss": 0.7592,
      "step": 299
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 1.8473510328793137,
      "learning_rate": 2.8517641909562075e-07,
      "loss": 0.7569,
      "step": 300
    },
    {
      "epoch": 0.8985074626865671,
      "grad_norm": 1.9818158957078058,
      "learning_rate": 2.6925921562124867e-07,
      "loss": 0.637,
      "step": 301
    },
    {
      "epoch": 0.9014925373134328,
      "grad_norm": 1.9920228777916036,
      "learning_rate": 2.5378678213483057e-07,
      "loss": 0.766,
      "step": 302
    },
    {
      "epoch": 0.9044776119402985,
      "grad_norm": 1.960285579935006,
      "learning_rate": 2.3876057330792344e-07,
      "loss": 0.7499,
      "step": 303
    },
    {
      "epoch": 0.9074626865671642,
      "grad_norm": 1.94458918165242,
      "learning_rate": 2.2418200185938488e-07,
      "loss": 0.7358,
      "step": 304
    },
    {
      "epoch": 0.9104477611940298,
      "grad_norm": 2.081159928237178,
      "learning_rate": 2.1005243842255552e-07,
      "loss": 0.882,
      "step": 305
    },
    {
      "epoch": 0.9134328358208955,
      "grad_norm": 1.9550064790232975,
      "learning_rate": 1.9637321141639743e-07,
      "loss": 0.6894,
      "step": 306
    },
    {
      "epoch": 0.9164179104477612,
      "grad_norm": 2.1049771274583486,
      "learning_rate": 1.8314560692059836e-07,
      "loss": 0.7991,
      "step": 307
    },
    {
      "epoch": 0.9194029850746268,
      "grad_norm": 2.0857228086478448,
      "learning_rate": 1.7037086855465902e-07,
      "loss": 0.7768,
      "step": 308
    },
    {
      "epoch": 0.9223880597014925,
      "grad_norm": 1.935207733713808,
      "learning_rate": 1.5805019736097105e-07,
      "loss": 0.7564,
      "step": 309
    },
    {
      "epoch": 0.9253731343283582,
      "grad_norm": 2.072194118865995,
      "learning_rate": 1.4618475169190017e-07,
      "loss": 0.7604,
      "step": 310
    },
    {
      "epoch": 0.9283582089552239,
      "grad_norm": 2.1552417858238493,
      "learning_rate": 1.3477564710088097e-07,
      "loss": 0.793,
      "step": 311
    },
    {
      "epoch": 0.9313432835820895,
      "grad_norm": 1.951097534718573,
      "learning_rate": 1.2382395623753484e-07,
      "loss": 0.7764,
      "step": 312
    },
    {
      "epoch": 0.9343283582089552,
      "grad_norm": 1.9504459470286735,
      "learning_rate": 1.1333070874682217e-07,
      "loss": 0.7777,
      "step": 313
    },
    {
      "epoch": 0.9373134328358209,
      "grad_norm": 2.175522821443901,
      "learning_rate": 1.0329689117224262e-07,
      "loss": 0.8003,
      "step": 314
    },
    {
      "epoch": 0.9402985074626866,
      "grad_norm": 1.896018319659114,
      "learning_rate": 9.372344686307655e-08,
      "loss": 0.6819,
      "step": 315
    },
    {
      "epoch": 0.9432835820895522,
      "grad_norm": 1.8626576145737677,
      "learning_rate": 8.461127588570039e-08,
      "loss": 0.7407,
      "step": 316
    },
    {
      "epoch": 0.9462686567164179,
      "grad_norm": 2.255720786589443,
      "learning_rate": 7.59612349389599e-08,
      "loss": 0.8032,
      "step": 317
    },
    {
      "epoch": 0.9492537313432836,
      "grad_norm": 1.988888031528982,
      "learning_rate": 6.777413727363069e-08,
      "loss": 0.6934,
      "step": 318
    },
    {
      "epoch": 0.9522388059701492,
      "grad_norm": 1.936424618352781,
      "learning_rate": 6.005075261595495e-08,
      "loss": 0.7447,
      "step": 319
    },
    {
      "epoch": 0.9552238805970149,
      "grad_norm": 2.166079975131979,
      "learning_rate": 5.279180709527765e-08,
      "loss": 0.6995,
      "step": 320
    },
    {
      "epoch": 0.9582089552238806,
      "grad_norm": 1.910155391887948,
      "learning_rate": 4.599798317577342e-08,
      "loss": 0.7723,
      "step": 321
    },
    {
      "epoch": 0.9611940298507463,
      "grad_norm": 2.015438544350679,
      "learning_rate": 3.9669919592288385e-08,
      "loss": 0.7478,
      "step": 322
    },
    {
      "epoch": 0.9641791044776119,
      "grad_norm": 1.7894327871602922,
      "learning_rate": 3.3808211290284886e-08,
      "loss": 0.6971,
      "step": 323
    },
    {
      "epoch": 0.9671641791044776,
      "grad_norm": 2.243535756266807,
      "learning_rate": 2.8413409369907887e-08,
      "loss": 0.8014,
      "step": 324
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 2.104094911680065,
      "learning_rate": 2.3486021034170857e-08,
      "loss": 0.7373,
      "step": 325
    },
    {
      "epoch": 0.9731343283582089,
      "grad_norm": 1.9746672549193203,
      "learning_rate": 1.9026509541272276e-08,
      "loss": 0.7462,
      "step": 326
    },
    {
      "epoch": 0.9761194029850746,
      "grad_norm": 1.8638633602470398,
      "learning_rate": 1.5035294161039882e-08,
      "loss": 0.7276,
      "step": 327
    },
    {
      "epoch": 0.9791044776119403,
      "grad_norm": 1.9221365215727475,
      "learning_rate": 1.1512750135511674e-08,
      "loss": 0.6948,
      "step": 328
    },
    {
      "epoch": 0.982089552238806,
      "grad_norm": 1.9696506075193372,
      "learning_rate": 8.459208643659122e-09,
      "loss": 0.8016,
      "step": 329
    },
    {
      "epoch": 0.9850746268656716,
      "grad_norm": 1.9260689325158074,
      "learning_rate": 5.874956770248186e-09,
      "loss": 0.772,
      "step": 330
    },
    {
      "epoch": 0.9880597014925373,
      "grad_norm": 1.9022494846934377,
      "learning_rate": 3.760237478849793e-09,
      "loss": 0.639,
      "step": 331
    },
    {
      "epoch": 0.991044776119403,
      "grad_norm": 2.1198036949572523,
      "learning_rate": 2.1152495889970035e-09,
      "loss": 0.7524,
      "step": 332
    },
    {
      "epoch": 0.9940298507462687,
      "grad_norm": 2.520497927670345,
      "learning_rate": 9.401477574932927e-10,
      "loss": 0.7804,
      "step": 333
    },
    {
      "epoch": 0.9970149253731343,
      "grad_norm": 1.9463118318794423,
      "learning_rate": 2.3504246386918394e-10,
      "loss": 0.6975,
      "step": 334
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.0407770920954924,
      "learning_rate": 0.0,
      "loss": 0.7463,
      "step": 335
    },
    {
      "epoch": 1.0,
      "step": 335,
      "total_flos": 27958090317824.0,
      "train_loss": 0.8481638431549072,
      "train_runtime": 15355.884,
      "train_samples_per_second": 0.175,
      "train_steps_per_second": 0.022
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 335,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2060,
  "total_flos": 27958090317824.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}