{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 233,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004291845493562232,
      "grad_norm": 64.1849594116211,
      "learning_rate": 4.1666666666666667e-07,
      "loss": 1.9257,
      "step": 1
    },
    {
      "epoch": 0.008583690987124463,
      "grad_norm": 67.15699768066406,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.9675,
      "step": 2
    },
    {
      "epoch": 0.012875536480686695,
      "grad_norm": 63.228580474853516,
      "learning_rate": 1.25e-06,
      "loss": 1.9059,
      "step": 3
    },
    {
      "epoch": 0.017167381974248927,
      "grad_norm": 50.28240966796875,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.6307,
      "step": 4
    },
    {
      "epoch": 0.02145922746781116,
      "grad_norm": 56.101253509521484,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 1.174,
      "step": 5
    },
    {
      "epoch": 0.02575107296137339,
      "grad_norm": 34.66034698486328,
      "learning_rate": 2.5e-06,
      "loss": 0.5821,
      "step": 6
    },
    {
      "epoch": 0.030042918454935622,
      "grad_norm": 79.90634155273438,
      "learning_rate": 2.916666666666667e-06,
      "loss": 0.4952,
      "step": 7
    },
    {
      "epoch": 0.034334763948497854,
      "grad_norm": 31.585365295410156,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.4005,
      "step": 8
    },
    {
      "epoch": 0.03862660944206009,
      "grad_norm": 20.49032974243164,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.3842,
      "step": 9
    },
    {
      "epoch": 0.04291845493562232,
      "grad_norm": 12.812472343444824,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.3652,
      "step": 10
    },
    {
      "epoch": 0.04721030042918455,
      "grad_norm": 12.473087310791016,
      "learning_rate": 4.583333333333333e-06,
      "loss": 0.3918,
      "step": 11
    },
    {
      "epoch": 0.05150214592274678,
      "grad_norm": 12.261868476867676,
      "learning_rate": 5e-06,
      "loss": 0.3106,
      "step": 12
    },
    {
      "epoch": 0.055793991416309016,
      "grad_norm": 10.985730171203613,
      "learning_rate": 5.416666666666667e-06,
      "loss": 0.2936,
      "step": 13
    },
    {
      "epoch": 0.060085836909871244,
      "grad_norm": 6.1359944343566895,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.3058,
      "step": 14
    },
    {
      "epoch": 0.06437768240343347,
      "grad_norm": 19.714662551879883,
      "learning_rate": 6.25e-06,
      "loss": 0.3188,
      "step": 15
    },
    {
      "epoch": 0.06866952789699571,
      "grad_norm": 50.216400146484375,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.2829,
      "step": 16
    },
    {
      "epoch": 0.07296137339055794,
      "grad_norm": 36.995113372802734,
      "learning_rate": 7.083333333333335e-06,
      "loss": 0.2931,
      "step": 17
    },
    {
      "epoch": 0.07725321888412018,
      "grad_norm": 4.576523780822754,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3061,
      "step": 18
    },
    {
      "epoch": 0.0815450643776824,
      "grad_norm": 4.885584831237793,
      "learning_rate": 7.916666666666667e-06,
      "loss": 0.3042,
      "step": 19
    },
    {
      "epoch": 0.08583690987124463,
      "grad_norm": 4.763634204864502,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.2747,
      "step": 20
    },
    {
      "epoch": 0.09012875536480687,
      "grad_norm": 3.823352575302124,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.2904,
      "step": 21
    },
    {
      "epoch": 0.0944206008583691,
      "grad_norm": 3.134524345397949,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.2695,
      "step": 22
    },
    {
      "epoch": 0.09871244635193133,
      "grad_norm": 3.5232937335968018,
      "learning_rate": 9.583333333333335e-06,
      "loss": 0.2811,
      "step": 23
    },
    {
      "epoch": 0.10300429184549356,
      "grad_norm": 3.035285711288452,
      "learning_rate": 1e-05,
      "loss": 0.2878,
      "step": 24
    },
    {
      "epoch": 0.1072961373390558,
      "grad_norm": 5.4150614738464355,
      "learning_rate": 9.999435142363484e-06,
      "loss": 0.283,
      "step": 25
    },
    {
      "epoch": 0.11158798283261803,
      "grad_norm": 34.44544982910156,
      "learning_rate": 9.997740697079595e-06,
      "loss": 0.3557,
      "step": 26
    },
    {
      "epoch": 0.11587982832618025,
      "grad_norm": 17.197986602783203,
      "learning_rate": 9.994917046996472e-06,
      "loss": 0.4027,
      "step": 27
    },
    {
      "epoch": 0.12017167381974249,
      "grad_norm": 14.15243911743164,
      "learning_rate": 9.990964830098246e-06,
      "loss": 0.3703,
      "step": 28
    },
    {
      "epoch": 0.12446351931330472,
      "grad_norm": 14.733880043029785,
      "learning_rate": 9.985884939360873e-06,
      "loss": 0.3651,
      "step": 29
    },
    {
      "epoch": 0.12875536480686695,
      "grad_norm": 5.188531398773193,
      "learning_rate": 9.979678522550382e-06,
      "loss": 0.3329,
      "step": 30
    },
    {
      "epoch": 0.13304721030042918,
      "grad_norm": 10.757170677185059,
      "learning_rate": 9.972346981963546e-06,
      "loss": 0.3351,
      "step": 31
    },
    {
      "epoch": 0.13733905579399142,
      "grad_norm": 5.147034168243408,
      "learning_rate": 9.963891974111042e-06,
      "loss": 0.3193,
      "step": 32
    },
    {
      "epoch": 0.14163090128755365,
      "grad_norm": 8.824125289916992,
      "learning_rate": 9.95431540934317e-06,
      "loss": 0.3167,
      "step": 33
    },
    {
      "epoch": 0.1459227467811159,
      "grad_norm": 3.809666872024536,
      "learning_rate": 9.943619451418225e-06,
      "loss": 0.3238,
      "step": 34
    },
    {
      "epoch": 0.15021459227467812,
      "grad_norm": 6.210383415222168,
      "learning_rate": 9.931806517013612e-06,
      "loss": 0.2888,
      "step": 35
    },
    {
      "epoch": 0.15450643776824036,
      "grad_norm": 2.9546399116516113,
      "learning_rate": 9.918879275179819e-06,
      "loss": 0.2781,
      "step": 36
    },
    {
      "epoch": 0.15879828326180256,
      "grad_norm": 2.7553744316101074,
      "learning_rate": 9.904840646737346e-06,
      "loss": 0.2936,
      "step": 37
    },
    {
      "epoch": 0.1630901287553648,
      "grad_norm": 2.9384710788726807,
      "learning_rate": 9.889693803616793e-06,
      "loss": 0.3017,
      "step": 38
    },
    {
      "epoch": 0.16738197424892703,
      "grad_norm": 3.0700764656066895,
      "learning_rate": 9.873442168142158e-06,
      "loss": 0.2803,
      "step": 39
    },
    {
      "epoch": 0.17167381974248927,
      "grad_norm": 3.089503049850464,
      "learning_rate": 9.856089412257605e-06,
      "loss": 0.2679,
      "step": 40
    },
    {
      "epoch": 0.1759656652360515,
      "grad_norm": 3.7180137634277344,
      "learning_rate": 9.837639456697802e-06,
      "loss": 0.2983,
      "step": 41
    },
    {
      "epoch": 0.18025751072961374,
      "grad_norm": 2.5908358097076416,
      "learning_rate": 9.818096470102067e-06,
      "loss": 0.2669,
      "step": 42
    },
    {
      "epoch": 0.18454935622317598,
      "grad_norm": 3.1205332279205322,
      "learning_rate": 9.797464868072489e-06,
      "loss": 0.2712,
      "step": 43
    },
    {
      "epoch": 0.1888412017167382,
      "grad_norm": 2.6999526023864746,
      "learning_rate": 9.775749312176249e-06,
      "loss": 0.2755,
      "step": 44
    },
    {
      "epoch": 0.19313304721030042,
      "grad_norm": 3.2999775409698486,
      "learning_rate": 9.752954708892379e-06,
      "loss": 0.2785,
      "step": 45
    },
    {
      "epoch": 0.19742489270386265,
      "grad_norm": 3.793800115585327,
      "learning_rate": 9.729086208503174e-06,
      "loss": 0.2803,
      "step": 46
    },
    {
      "epoch": 0.2017167381974249,
      "grad_norm": 2.8122568130493164,
      "learning_rate": 9.704149203930522e-06,
      "loss": 0.2819,
      "step": 47
    },
    {
      "epoch": 0.20600858369098712,
      "grad_norm": 3.148585557937622,
      "learning_rate": 9.67814932951741e-06,
      "loss": 0.3123,
      "step": 48
    },
    {
      "epoch": 0.21030042918454936,
      "grad_norm": 3.0467166900634766,
      "learning_rate": 9.651092459754879e-06,
      "loss": 0.3074,
      "step": 49
    },
    {
      "epoch": 0.2145922746781116,
      "grad_norm": 2.846820831298828,
      "learning_rate": 9.622984707954732e-06,
      "loss": 0.2867,
      "step": 50
    },
    {
      "epoch": 0.21888412017167383,
      "grad_norm": 3.011904239654541,
      "learning_rate": 9.593832424868271e-06,
      "loss": 0.2857,
      "step": 51
    },
    {
      "epoch": 0.22317596566523606,
      "grad_norm": 3.032160758972168,
      "learning_rate": 9.563642197251382e-06,
      "loss": 0.312,
      "step": 52
    },
    {
      "epoch": 0.22746781115879827,
      "grad_norm": 3.2738237380981445,
      "learning_rate": 9.532420846376316e-06,
      "loss": 0.273,
      "step": 53
    },
    {
      "epoch": 0.2317596566523605,
      "grad_norm": 2.755352735519409,
      "learning_rate": 9.500175426490455e-06,
      "loss": 0.2749,
      "step": 54
    },
    {
      "epoch": 0.23605150214592274,
      "grad_norm": 2.5805509090423584,
      "learning_rate": 9.466913223222467e-06,
      "loss": 0.2828,
      "step": 55
    },
    {
      "epoch": 0.24034334763948498,
      "grad_norm": 2.1449105739593506,
      "learning_rate": 9.432641751936162e-06,
      "loss": 0.2776,
      "step": 56
    },
    {
      "epoch": 0.2446351931330472,
      "grad_norm": 2.747648000717163,
      "learning_rate": 9.397368756032445e-06,
      "loss": 0.2859,
      "step": 57
    },
    {
      "epoch": 0.24892703862660945,
      "grad_norm": 2.4565420150756836,
      "learning_rate": 9.361102205199762e-06,
      "loss": 0.2987,
      "step": 58
    },
    {
      "epoch": 0.2532188841201717,
      "grad_norm": 2.3403310775756836,
      "learning_rate": 9.32385029361338e-06,
      "loss": 0.2766,
      "step": 59
    },
    {
      "epoch": 0.2575107296137339,
      "grad_norm": 2.400320053100586,
      "learning_rate": 9.285621438083997e-06,
      "loss": 0.2477,
      "step": 60
    },
    {
      "epoch": 0.26180257510729615,
      "grad_norm": 2.79807710647583,
      "learning_rate": 9.246424276156008e-06,
      "loss": 0.2699,
      "step": 61
    },
    {
      "epoch": 0.26609442060085836,
      "grad_norm": 2.786733627319336,
      "learning_rate": 9.206267664155906e-06,
      "loss": 0.2822,
      "step": 62
    },
    {
      "epoch": 0.2703862660944206,
      "grad_norm": 2.586775302886963,
      "learning_rate": 9.165160675191272e-06,
      "loss": 0.2739,
      "step": 63
    },
    {
      "epoch": 0.27467811158798283,
      "grad_norm": 2.561934471130371,
      "learning_rate": 9.123112597100759e-06,
      "loss": 0.2809,
      "step": 64
    },
    {
      "epoch": 0.27896995708154504,
      "grad_norm": 2.8930823802948,
      "learning_rate": 9.080132930355567e-06,
      "loss": 0.2733,
      "step": 65
    },
    {
      "epoch": 0.2832618025751073,
      "grad_norm": 2.4183900356292725,
      "learning_rate": 9.03623138591289e-06,
      "loss": 0.2667,
      "step": 66
    },
    {
      "epoch": 0.2875536480686695,
      "grad_norm": 2.3127546310424805,
      "learning_rate": 8.99141788302178e-06,
      "loss": 0.2586,
      "step": 67
    },
    {
      "epoch": 0.2918454935622318,
      "grad_norm": 1.8758937120437622,
      "learning_rate": 8.94570254698197e-06,
      "loss": 0.2411,
      "step": 68
    },
    {
      "epoch": 0.296137339055794,
      "grad_norm": 2.9779796600341797,
      "learning_rate": 8.899095706856122e-06,
      "loss": 0.2645,
      "step": 69
    },
    {
      "epoch": 0.30042918454935624,
      "grad_norm": 2.948882579803467,
      "learning_rate": 8.851607893136065e-06,
      "loss": 0.2872,
      "step": 70
    },
    {
      "epoch": 0.30472103004291845,
      "grad_norm": 2.4292919635772705,
      "learning_rate": 8.803249835363486e-06,
      "loss": 0.247,
      "step": 71
    },
    {
      "epoch": 0.3090128755364807,
      "grad_norm": 2.62158465385437,
      "learning_rate": 8.754032459705672e-06,
      "loss": 0.2903,
      "step": 72
    },
    {
      "epoch": 0.3133047210300429,
      "grad_norm": 2.522172451019287,
      "learning_rate": 8.703966886486819e-06,
      "loss": 0.2734,
      "step": 73
    },
    {
      "epoch": 0.31759656652360513,
      "grad_norm": 2.3345656394958496,
      "learning_rate": 8.65306442767547e-06,
      "loss": 0.2551,
      "step": 74
    },
    {
      "epoch": 0.3218884120171674,
      "grad_norm": 2.487061023712158,
      "learning_rate": 8.601336584328659e-06,
      "loss": 0.2848,
      "step": 75
    },
    {
      "epoch": 0.3261802575107296,
      "grad_norm": 2.012143611907959,
      "learning_rate": 8.548795043993316e-06,
      "loss": 0.2493,
      "step": 76
    },
    {
      "epoch": 0.33047210300429186,
      "grad_norm": 2.668299436569214,
      "learning_rate": 8.495451678065563e-06,
      "loss": 0.284,
      "step": 77
    },
    {
      "epoch": 0.33476394849785407,
      "grad_norm": 2.1952147483825684,
      "learning_rate": 8.441318539108433e-06,
      "loss": 0.2868,
      "step": 78
    },
    {
      "epoch": 0.33905579399141633,
      "grad_norm": 2.1798348426818848,
      "learning_rate": 8.386407858128707e-06,
      "loss": 0.2732,
      "step": 79
    },
    {
      "epoch": 0.34334763948497854,
      "grad_norm": 2.0891451835632324,
      "learning_rate": 8.330732041813367e-06,
      "loss": 0.2509,
      "step": 80
    },
    {
      "epoch": 0.34763948497854075,
      "grad_norm": 2.444913148880005,
      "learning_rate": 8.274303669726427e-06,
      "loss": 0.2836,
      "step": 81
    },
    {
      "epoch": 0.351931330472103,
      "grad_norm": 2.4908130168914795,
      "learning_rate": 8.217135491466636e-06,
      "loss": 0.2891,
      "step": 82
    },
    {
      "epoch": 0.3562231759656652,
      "grad_norm": 2.058657169342041,
      "learning_rate": 8.15924042378682e-06,
      "loss": 0.2482,
      "step": 83
    },
    {
      "epoch": 0.3605150214592275,
      "grad_norm": 1.9691925048828125,
      "learning_rate": 8.100631547675417e-06,
      "loss": 0.2429,
      "step": 84
    },
    {
      "epoch": 0.3648068669527897,
      "grad_norm": 2.1856772899627686,
      "learning_rate": 8.041322105400923e-06,
      "loss": 0.2823,
      "step": 85
    },
    {
      "epoch": 0.36909871244635195,
      "grad_norm": 2.004932165145874,
      "learning_rate": 7.981325497519892e-06,
      "loss": 0.2552,
      "step": 86
    },
    {
      "epoch": 0.37339055793991416,
      "grad_norm": 2.2658703327178955,
      "learning_rate": 7.920655279849173e-06,
      "loss": 0.264,
      "step": 87
    },
    {
      "epoch": 0.3776824034334764,
      "grad_norm": 2.124732255935669,
      "learning_rate": 7.859325160403073e-06,
      "loss": 0.2842,
      "step": 88
    },
    {
      "epoch": 0.38197424892703863,
      "grad_norm": 2.223417043685913,
      "learning_rate": 7.797348996296116e-06,
      "loss": 0.2552,
      "step": 89
    },
    {
      "epoch": 0.38626609442060084,
      "grad_norm": 2.6624255180358887,
      "learning_rate": 7.734740790612137e-06,
      "loss": 0.2935,
      "step": 90
    },
    {
      "epoch": 0.3905579399141631,
      "grad_norm": 3.1369097232818604,
      "learning_rate": 7.671514689240366e-06,
      "loss": 0.2568,
      "step": 91
    },
    {
      "epoch": 0.3948497854077253,
      "grad_norm": 2.0869433879852295,
      "learning_rate": 7.607684977679284e-06,
      "loss": 0.2414,
      "step": 92
    },
    {
      "epoch": 0.39914163090128757,
      "grad_norm": 2.3208043575286865,
      "learning_rate": 7.543266077808893e-06,
      "loss": 0.2272,
      "step": 93
    },
    {
      "epoch": 0.4034334763948498,
      "grad_norm": 2.5530765056610107,
      "learning_rate": 7.478272544632204e-06,
      "loss": 0.2636,
      "step": 94
    },
    {
      "epoch": 0.40772532188841204,
      "grad_norm": 2.2211928367614746,
      "learning_rate": 7.412719062986632e-06,
      "loss": 0.282,
      "step": 95
    },
    {
      "epoch": 0.41201716738197425,
      "grad_norm": 1.9540337324142456,
      "learning_rate": 7.3466204442260605e-06,
      "loss": 0.2303,
      "step": 96
    },
    {
      "epoch": 0.41630901287553645,
      "grad_norm": 2.4863698482513428,
      "learning_rate": 7.279991622874319e-06,
      "loss": 0.2969,
      "step": 97
    },
    {
      "epoch": 0.4206008583690987,
      "grad_norm": 1.811273217201233,
      "learning_rate": 7.212847653250828e-06,
      "loss": 0.2439,
      "step": 98
    },
    {
      "epoch": 0.4248927038626609,
      "grad_norm": 1.7187399864196777,
      "learning_rate": 7.145203706069183e-06,
      "loss": 0.2353,
      "step": 99
    },
    {
      "epoch": 0.4291845493562232,
      "grad_norm": 2.329214572906494,
      "learning_rate": 7.0770750650094335e-06,
      "loss": 0.2718,
      "step": 100
    },
    {
      "epoch": 0.4334763948497854,
      "grad_norm": 2.300833225250244,
      "learning_rate": 7.008477123264849e-06,
      "loss": 0.2815,
      "step": 101
    },
    {
      "epoch": 0.43776824034334766,
      "grad_norm": 2.1259753704071045,
      "learning_rate": 6.939425380063924e-06,
      "loss": 0.2544,
      "step": 102
    },
    {
      "epoch": 0.44206008583690987,
      "grad_norm": 2.159365177154541,
      "learning_rate": 6.869935437168449e-06,
      "loss": 0.2632,
      "step": 103
    },
    {
      "epoch": 0.44635193133047213,
      "grad_norm": 2.122368812561035,
      "learning_rate": 6.800022995348381e-06,
      "loss": 0.2562,
      "step": 104
    },
    {
      "epoch": 0.45064377682403434,
      "grad_norm": 1.8842718601226807,
      "learning_rate": 6.729703850834381e-06,
      "loss": 0.2687,
      "step": 105
    },
    {
      "epoch": 0.45493562231759654,
      "grad_norm": 1.6606192588806152,
      "learning_rate": 6.65899389174876e-06,
      "loss": 0.2289,
      "step": 106
    },
    {
      "epoch": 0.4592274678111588,
      "grad_norm": 2.395519256591797,
      "learning_rate": 6.587909094515663e-06,
      "loss": 0.2487,
      "step": 107
    },
    {
      "epoch": 0.463519313304721,
      "grad_norm": 1.7725951671600342,
      "learning_rate": 6.5164655202513135e-06,
      "loss": 0.2341,
      "step": 108
    },
    {
      "epoch": 0.4678111587982833,
      "grad_norm": 2.764538288116455,
      "learning_rate": 6.444679311135112e-06,
      "loss": 0.2704,
      "step": 109
    },
    {
      "epoch": 0.4721030042918455,
      "grad_norm": 2.5796592235565186,
      "learning_rate": 6.372566686762427e-06,
      "loss": 0.2878,
      "step": 110
    },
    {
      "epoch": 0.47639484978540775,
      "grad_norm": 2.3967843055725098,
      "learning_rate": 6.300143940479881e-06,
      "loss": 0.2606,
      "step": 111
    },
    {
      "epoch": 0.48068669527896996,
      "grad_norm": 2.054126739501953,
      "learning_rate": 6.227427435703997e-06,
      "loss": 0.2278,
      "step": 112
    },
    {
      "epoch": 0.48497854077253216,
      "grad_norm": 2.0056793689727783,
      "learning_rate": 6.154433602223979e-06,
      "loss": 0.2131,
      "step": 113
    },
    {
      "epoch": 0.4892703862660944,
      "grad_norm": 2.2301571369171143,
      "learning_rate": 6.0811789324895365e-06,
      "loss": 0.2437,
      "step": 114
    },
    {
      "epoch": 0.49356223175965663,
      "grad_norm": 2.192481517791748,
      "learning_rate": 6.0076799778845105e-06,
      "loss": 0.2484,
      "step": 115
    },
    {
      "epoch": 0.4978540772532189,
      "grad_norm": 1.9841991662979126,
      "learning_rate": 5.933953344987215e-06,
      "loss": 0.2561,
      "step": 116
    },
    {
      "epoch": 0.5021459227467812,
      "grad_norm": 2.0496819019317627,
      "learning_rate": 5.860015691818292e-06,
      "loss": 0.2586,
      "step": 117
    },
    {
      "epoch": 0.5064377682403434,
      "grad_norm": 1.8923070430755615,
      "learning_rate": 5.78588372407695e-06,
      "loss": 0.2329,
      "step": 118
    },
    {
      "epoch": 0.5107296137339056,
      "grad_norm": 1.734017252922058,
      "learning_rate": 5.711574191366427e-06,
      "loss": 0.2484,
      "step": 119
    },
    {
      "epoch": 0.5150214592274678,
      "grad_norm": 1.7440507411956787,
      "learning_rate": 5.637103883409525e-06,
      "loss": 0.231,
      "step": 120
    },
    {
      "epoch": 0.51931330472103,
      "grad_norm": 2.1365606784820557,
      "learning_rate": 5.562489626255104e-06,
      "loss": 0.2572,
      "step": 121
    },
    {
      "epoch": 0.5236051502145923,
      "grad_norm": 2.0460870265960693,
      "learning_rate": 5.487748278476342e-06,
      "loss": 0.2535,
      "step": 122
    },
    {
      "epoch": 0.5278969957081545,
      "grad_norm": 1.7586556673049927,
      "learning_rate": 5.412896727361663e-06,
      "loss": 0.2112,
      "step": 123
    },
    {
      "epoch": 0.5321888412017167,
      "grad_norm": 2.070552110671997,
      "learning_rate": 5.337951885099167e-06,
      "loss": 0.2179,
      "step": 124
    },
    {
      "epoch": 0.5364806866952789,
      "grad_norm": 1.604686975479126,
      "learning_rate": 5.262930684955439e-06,
      "loss": 0.2321,
      "step": 125
    },
    {
      "epoch": 0.5407725321888412,
      "grad_norm": 2.326280117034912,
      "learning_rate": 5.187850077449604e-06,
      "loss": 0.2568,
      "step": 126
    },
    {
      "epoch": 0.5450643776824035,
      "grad_norm": 2.1093199253082275,
      "learning_rate": 5.112727026523461e-06,
      "loss": 0.22,
      "step": 127
    },
    {
      "epoch": 0.5493562231759657,
      "grad_norm": 1.8710286617279053,
      "learning_rate": 5.03757850570861e-06,
      "loss": 0.242,
      "step": 128
    },
    {
      "epoch": 0.5536480686695279,
      "grad_norm": 1.5934395790100098,
      "learning_rate": 4.9624214942913916e-06,
      "loss": 0.2369,
      "step": 129
    },
    {
      "epoch": 0.5579399141630901,
      "grad_norm": 1.6365737915039062,
      "learning_rate": 4.88727297347654e-06,
      "loss": 0.2463,
      "step": 130
    },
    {
      "epoch": 0.5622317596566524,
      "grad_norm": 1.8989615440368652,
      "learning_rate": 4.8121499225503974e-06,
      "loss": 0.2265,
      "step": 131
    },
    {
      "epoch": 0.5665236051502146,
      "grad_norm": 1.747003436088562,
      "learning_rate": 4.737069315044562e-06,
      "loss": 0.2292,
      "step": 132
    },
    {
      "epoch": 0.5708154506437768,
      "grad_norm": 2.0357205867767334,
      "learning_rate": 4.662048114900837e-06,
      "loss": 0.2605,
      "step": 133
    },
    {
      "epoch": 0.575107296137339,
      "grad_norm": 2.0515100955963135,
      "learning_rate": 4.587103272638339e-06,
      "loss": 0.2586,
      "step": 134
    },
    {
      "epoch": 0.5793991416309013,
      "grad_norm": 1.8800376653671265,
      "learning_rate": 4.512251721523659e-06,
      "loss": 0.2161,
      "step": 135
    },
    {
      "epoch": 0.5836909871244635,
      "grad_norm": 2.1146693229675293,
      "learning_rate": 4.437510373744897e-06,
      "loss": 0.2449,
      "step": 136
    },
    {
      "epoch": 0.5879828326180258,
      "grad_norm": 1.944589376449585,
      "learning_rate": 4.362896116590475e-06,
      "loss": 0.2438,
      "step": 137
    },
    {
      "epoch": 0.592274678111588,
      "grad_norm": 2.022193431854248,
      "learning_rate": 4.2884258086335755e-06,
      "loss": 0.2528,
      "step": 138
    },
    {
      "epoch": 0.5965665236051502,
      "grad_norm": 1.5787066221237183,
      "learning_rate": 4.214116275923051e-06,
      "loss": 0.2198,
      "step": 139
    },
    {
      "epoch": 0.6008583690987125,
      "grad_norm": 2.127326726913452,
      "learning_rate": 4.1399843081817085e-06,
      "loss": 0.2428,
      "step": 140
    },
    {
      "epoch": 0.6051502145922747,
      "grad_norm": 1.6171480417251587,
      "learning_rate": 4.066046655012786e-06,
      "loss": 0.2361,
      "step": 141
    },
    {
      "epoch": 0.6094420600858369,
      "grad_norm": 1.6328740119934082,
      "learning_rate": 3.992320022115492e-06,
      "loss": 0.2333,
      "step": 142
    },
    {
      "epoch": 0.6137339055793991,
      "grad_norm": 1.5752973556518555,
      "learning_rate": 3.918821067510464e-06,
      "loss": 0.2159,
      "step": 143
    },
    {
      "epoch": 0.6180257510729614,
      "grad_norm": 2.487964630126953,
      "learning_rate": 3.845566397776022e-06,
      "loss": 0.2619,
      "step": 144
    },
    {
      "epoch": 0.6223175965665236,
      "grad_norm": 2.5639259815216064,
      "learning_rate": 3.7725725642960047e-06,
      "loss": 0.2361,
      "step": 145
    },
    {
      "epoch": 0.6266094420600858,
      "grad_norm": 1.714277982711792,
      "learning_rate": 3.6998560595201188e-06,
      "loss": 0.2218,
      "step": 146
    },
    {
      "epoch": 0.630901287553648,
      "grad_norm": 1.72953462600708,
      "learning_rate": 3.627433313237576e-06,
      "loss": 0.237,
      "step": 147
    },
    {
      "epoch": 0.6351931330472103,
      "grad_norm": 1.7776756286621094,
      "learning_rate": 3.555320688864889e-06,
      "loss": 0.2291,
      "step": 148
    },
    {
      "epoch": 0.6394849785407726,
      "grad_norm": 1.651568055152893,
      "learning_rate": 3.483534479748688e-06,
      "loss": 0.2326,
      "step": 149
    },
    {
      "epoch": 0.6437768240343348,
      "grad_norm": 1.5333776473999023,
      "learning_rate": 3.4120909054843375e-06,
      "loss": 0.1999,
      "step": 150
    },
    {
      "epoch": 0.648068669527897,
      "grad_norm": 1.4129211902618408,
      "learning_rate": 3.3410061082512422e-06,
      "loss": 0.2014,
      "step": 151
    },
    {
      "epoch": 0.6523605150214592,
      "grad_norm": 1.9737615585327148,
      "learning_rate": 3.2702961491656197e-06,
      "loss": 0.2184,
      "step": 152
    },
    {
      "epoch": 0.6566523605150214,
      "grad_norm": 1.5220732688903809,
      "learning_rate": 3.1999770046516198e-06,
      "loss": 0.2179,
      "step": 153
    },
    {
      "epoch": 0.6609442060085837,
      "grad_norm": 1.649389386177063,
      "learning_rate": 3.130064562831553e-06,
      "loss": 0.2097,
      "step": 154
    },
    {
      "epoch": 0.6652360515021459,
      "grad_norm": 1.6441725492477417,
      "learning_rate": 3.0605746199360755e-06,
      "loss": 0.2267,
      "step": 155
    },
    {
      "epoch": 0.6695278969957081,
      "grad_norm": 1.4296993017196655,
      "learning_rate": 2.991522876735154e-06,
      "loss": 0.2149,
      "step": 156
    },
    {
      "epoch": 0.6738197424892703,
      "grad_norm": 2.1046576499938965,
      "learning_rate": 2.9229249349905686e-06,
      "loss": 0.2447,
      "step": 157
    },
    {
      "epoch": 0.6781115879828327,
      "grad_norm": 1.5854638814926147,
      "learning_rate": 2.8547962939308187e-06,
      "loss": 0.2071,
      "step": 158
    },
    {
      "epoch": 0.6824034334763949,
      "grad_norm": 1.6980681419372559,
      "learning_rate": 2.787152346749173e-06,
      "loss": 0.2211,
      "step": 159
    },
    {
      "epoch": 0.6866952789699571,
      "grad_norm": 1.6970704793930054,
      "learning_rate": 2.720008377125682e-06,
      "loss": 0.2156,
      "step": 160
    },
    {
      "epoch": 0.6909871244635193,
      "grad_norm": 1.894504427909851,
      "learning_rate": 2.6533795557739407e-06,
      "loss": 0.2499,
      "step": 161
    },
    {
      "epoch": 0.6952789699570815,
      "grad_norm": 1.8777060508728027,
      "learning_rate": 2.5872809370133704e-06,
      "loss": 0.2158,
      "step": 162
    },
    {
      "epoch": 0.6995708154506438,
      "grad_norm": 3.1802759170532227,
      "learning_rate": 2.5217274553677975e-06,
      "loss": 0.2248,
      "step": 163
    },
    {
      "epoch": 0.703862660944206,
      "grad_norm": 1.714854121208191,
      "learning_rate": 2.4567339221911086e-06,
      "loss": 0.2338,
      "step": 164
    },
    {
      "epoch": 0.7081545064377682,
      "grad_norm": 1.6652601957321167,
      "learning_rate": 2.3923150223207176e-06,
      "loss": 0.2411,
      "step": 165
    },
    {
      "epoch": 0.7124463519313304,
      "grad_norm": 1.497910976409912,
      "learning_rate": 2.328485310759635e-06,
      "loss": 0.2089,
      "step": 166
    },
    {
      "epoch": 0.7167381974248928,
      "grad_norm": 1.593087911605835,
      "learning_rate": 2.265259209387867e-06,
      "loss": 0.2284,
      "step": 167
    },
    {
      "epoch": 0.721030042918455,
      "grad_norm": 1.6965844631195068,
      "learning_rate": 2.202651003703885e-06,
      "loss": 0.2139,
      "step": 168
    },
    {
      "epoch": 0.7253218884120172,
      "grad_norm": 1.6895867586135864,
      "learning_rate": 2.140674839596931e-06,
      "loss": 0.209,
      "step": 169
    },
    {
      "epoch": 0.7296137339055794,
      "grad_norm": 1.9336439371109009,
      "learning_rate": 2.0793447201508288e-06,
      "loss": 0.2235,
      "step": 170
    },
    {
      "epoch": 0.7339055793991416,
      "grad_norm": 1.726787805557251,
      "learning_rate": 2.01867450248011e-06,
      "loss": 0.2363,
      "step": 171
    },
    {
      "epoch": 0.7381974248927039,
      "grad_norm": 1.6296881437301636,
      "learning_rate": 1.9586778945990785e-06,
      "loss": 0.2111,
      "step": 172
    },
    {
      "epoch": 0.7424892703862661,
      "grad_norm": 1.867713212966919,
      "learning_rate": 1.8993684523245842e-06,
      "loss": 0.2041,
      "step": 173
    },
    {
      "epoch": 0.7467811158798283,
      "grad_norm": 1.8173270225524902,
      "learning_rate": 1.8407595762131814e-06,
      "loss": 0.254,
      "step": 174
    },
    {
      "epoch": 0.7510729613733905,
      "grad_norm": 1.9783226251602173,
      "learning_rate": 1.7828645085333645e-06,
      "loss": 0.2478,
      "step": 175
    },
    {
      "epoch": 0.7553648068669528,
      "grad_norm": 1.5777461528778076,
      "learning_rate": 1.7256963302735752e-06,
      "loss": 0.2382,
      "step": 176
    },
    {
      "epoch": 0.759656652360515,
      "grad_norm": 1.6963618993759155,
      "learning_rate": 1.6692679581866334e-06,
      "loss": 0.2268,
      "step": 177
    },
    {
      "epoch": 0.7639484978540773,
      "grad_norm": 1.4376635551452637,
      "learning_rate": 1.6135921418712959e-06,
      "loss": 0.1918,
      "step": 178
    },
    {
      "epoch": 0.7682403433476395,
      "grad_norm": 1.6802223920822144,
      "learning_rate": 1.5586814608915673e-06,
      "loss": 0.2403,
      "step": 179
    },
    {
      "epoch": 0.7725321888412017,
      "grad_norm": 1.6209694147109985,
      "learning_rate": 1.5045483219344387e-06,
      "loss": 0.2145,
      "step": 180
    },
    {
      "epoch": 0.776824034334764,
      "grad_norm": 1.8394389152526855,
      "learning_rate": 1.4512049560066837e-06,
      "loss": 0.2147,
      "step": 181
    },
    {
      "epoch": 0.7811158798283262,
      "grad_norm": 1.6125578880310059,
      "learning_rate": 1.3986634156713418e-06,
      "loss": 0.2106,
      "step": 182
    },
    {
      "epoch": 0.7854077253218884,
      "grad_norm": 1.5914397239685059,
      "learning_rate": 1.3469355723245303e-06,
      "loss": 0.2064,
      "step": 183
    },
    {
      "epoch": 0.7896995708154506,
      "grad_norm": 2.1572067737579346,
      "learning_rate": 1.2960331135131826e-06,
      "loss": 0.2461,
      "step": 184
    },
    {
      "epoch": 0.7939914163090128,
      "grad_norm": 1.6950761079788208,
      "learning_rate": 1.245967540294329e-06,
      "loss": 0.2076,
      "step": 185
    },
    {
      "epoch": 0.7982832618025751,
      "grad_norm": 1.7178866863250732,
      "learning_rate": 1.1967501646365147e-06,
      "loss": 0.2116,
      "step": 186
    },
    {
      "epoch": 0.8025751072961373,
      "grad_norm": 1.9565534591674805,
      "learning_rate": 1.1483921068639353e-06,
      "loss": 0.1959,
      "step": 187
    },
    {
      "epoch": 0.8068669527896996,
      "grad_norm": 2.3265628814697266,
      "learning_rate": 1.1009042931438784e-06,
      "loss": 0.222,
      "step": 188
    },
    {
      "epoch": 0.8111587982832618,
      "grad_norm": 1.572929859161377,
      "learning_rate": 1.0542974530180327e-06,
      "loss": 0.216,
      "step": 189
    },
    {
      "epoch": 0.8154506437768241,
      "grad_norm": 1.3902196884155273,
      "learning_rate": 1.00858211697822e-06,
      "loss": 0.1998,
      "step": 190
    },
    {
      "epoch": 0.8197424892703863,
      "grad_norm": 1.702994465827942,
      "learning_rate": 9.637686140871121e-07,
      "loss": 0.2223,
      "step": 191
    },
    {
      "epoch": 0.8240343347639485,
      "grad_norm": 1.6051976680755615,
      "learning_rate": 9.198670696444339e-07,
      "loss": 0.2146,
      "step": 192
    },
    {
      "epoch": 0.8283261802575107,
      "grad_norm": 1.641396403312683,
      "learning_rate": 8.768874028992431e-07,
      "loss": 0.2132,
      "step": 193
    },
    {
      "epoch": 0.8326180257510729,
      "grad_norm": 1.9438247680664062,
      "learning_rate": 8.348393248087289e-07,
      "loss": 0.2347,
      "step": 194
    },
    {
      "epoch": 0.8369098712446352,
      "grad_norm": 1.9501413106918335,
      "learning_rate": 7.937323358440935e-07,
      "loss": 0.2492,
      "step": 195
    },
    {
      "epoch": 0.8412017167381974,
      "grad_norm": 1.5952423810958862,
      "learning_rate": 7.535757238439939e-07,
      "loss": 0.2213,
      "step": 196
    },
    {
      "epoch": 0.8454935622317596,
      "grad_norm": 1.4453738927841187,
      "learning_rate": 7.143785619160026e-07,
      "loss": 0.1981,
      "step": 197
    },
    {
      "epoch": 0.8497854077253219,
      "grad_norm": 1.3863543272018433,
      "learning_rate": 6.761497063866207e-07,
      "loss": 0.2005,
      "step": 198
    },
    {
      "epoch": 0.8540772532188842,
      "grad_norm": 1.553649663925171,
      "learning_rate": 6.388977948002406e-07,
      "loss": 0.2026,
      "step": 199
    },
    {
      "epoch": 0.8583690987124464,
      "grad_norm": 1.4505783319473267,
      "learning_rate": 6.026312439675553e-07,
      "loss": 0.1954,
      "step": 200
    },
    {
      "epoch": 0.8626609442060086,
      "grad_norm": 2.247232675552368,
      "learning_rate": 5.673582480638395e-07,
      "loss": 0.2119,
      "step": 201
    },
    {
      "epoch": 0.8669527896995708,
      "grad_norm": 1.8728052377700806,
      "learning_rate": 5.330867767775333e-07,
      "loss": 0.2398,
      "step": 202
    },
    {
      "epoch": 0.871244635193133,
      "grad_norm": 1.493640422821045,
      "learning_rate": 4.998245735095459e-07,
      "loss": 0.2228,
      "step": 203
    },
    {
      "epoch": 0.8755364806866953,
      "grad_norm": 1.5458440780639648,
      "learning_rate": 4.6757915362368567e-07,
      "loss": 0.211,
      "step": 204
    },
    {
      "epoch": 0.8798283261802575,
      "grad_norm": 1.659397840499878,
      "learning_rate": 4.363578027486187e-07,
      "loss": 0.2086,
      "step": 205
    },
    {
      "epoch": 0.8841201716738197,
      "grad_norm": 1.587699055671692,
      "learning_rate": 4.0616757513173123e-07,
      "loss": 0.2016,
      "step": 206
    },
    {
      "epoch": 0.8884120171673819,
      "grad_norm": 1.5061396360397339,
      "learning_rate": 3.7701529204526856e-07,
      "loss": 0.2068,
      "step": 207
    },
    {
      "epoch": 0.8927038626609443,
      "grad_norm": 1.7324118614196777,
      "learning_rate": 3.4890754024512254e-07,
      "loss": 0.2208,
      "step": 208
    },
    {
      "epoch": 0.8969957081545065,
      "grad_norm": 1.465354084968567,
      "learning_rate": 3.2185067048259245e-07,
      "loss": 0.2151,
      "step": 209
    },
    {
      "epoch": 0.9012875536480687,
      "grad_norm": 1.370495080947876,
      "learning_rate": 2.9585079606947843e-07,
      "loss": 0.1964,
      "step": 210
    },
    {
      "epoch": 0.9055793991416309,
      "grad_norm": 1.866836428642273,
      "learning_rate": 2.7091379149682683e-07,
      "loss": 0.1929,
      "step": 211
    },
    {
      "epoch": 0.9098712446351931,
      "grad_norm": 1.388830304145813,
      "learning_rate": 2.470452911076227e-07,
      "loss": 0.2049,
      "step": 212
    },
    {
      "epoch": 0.9141630901287554,
      "grad_norm": 1.3401647806167603,
      "learning_rate": 2.242506878237538e-07,
      "loss": 0.1975,
      "step": 213
    },
    {
      "epoch": 0.9184549356223176,
      "grad_norm": 1.3348737955093384,
      "learning_rate": 2.0253513192751374e-07,
      "loss": 0.2076,
      "step": 214
    },
    {
      "epoch": 0.9227467811158798,
      "grad_norm": 1.2677830457687378,
      "learning_rate": 1.8190352989793325e-07,
      "loss": 0.1966,
      "step": 215
    },
    {
      "epoch": 0.927038626609442,
      "grad_norm": 1.590033769607544,
      "learning_rate": 1.6236054330219853e-07,
      "loss": 0.2217,
      "step": 216
    },
    {
      "epoch": 0.9313304721030042,
      "grad_norm": 1.4973030090332031,
      "learning_rate": 1.439105877423963e-07,
      "loss": 0.2107,
      "step": 217
    },
    {
      "epoch": 0.9356223175965666,
      "grad_norm": 1.597288727760315,
      "learning_rate": 1.2655783185784253e-07,
      "loss": 0.2267,
      "step": 218
    },
    {
      "epoch": 0.9399141630901288,
      "grad_norm": 1.720969319343567,
      "learning_rate": 1.1030619638320805e-07,
      "loss": 0.2075,
      "step": 219
    },
    {
      "epoch": 0.944206008583691,
      "grad_norm": 1.4152257442474365,
      "learning_rate": 9.51593532626538e-08,
      "loss": 0.2003,
      "step": 220
    },
    {
      "epoch": 0.9484978540772532,
      "grad_norm": 1.540600299835205,
      "learning_rate": 8.11207248201834e-08,
      "loss": 0.2211,
      "step": 221
    },
    {
      "epoch": 0.9527896995708155,
      "grad_norm": 1.7658554315567017,
      "learning_rate": 6.819348298638839e-08,
      "loss": 0.2296,
      "step": 222
    },
    {
      "epoch": 0.9570815450643777,
      "grad_norm": 1.7942489385604858,
      "learning_rate": 5.638054858177644e-08,
      "loss": 0.2077,
      "step": 223
    },
    {
      "epoch": 0.9613733905579399,
      "grad_norm": 1.565927505493164,
      "learning_rate": 4.568459065683206e-08,
      "loss": 0.1982,
      "step": 224
    },
    {
      "epoch": 0.9656652360515021,
      "grad_norm": 1.7370845079421997,
      "learning_rate": 3.610802588895845e-08,
      "loss": 0.2303,
      "step": 225
    },
    {
      "epoch": 0.9699570815450643,
      "grad_norm": 1.3400768041610718,
      "learning_rate": 2.765301803645426e-08,
      "loss": 0.1816,
      "step": 226
    },
    {
      "epoch": 0.9742489270386266,
      "grad_norm": 1.6028944253921509,
      "learning_rate": 2.0321477449619098e-08,
      "loss": 0.2329,
      "step": 227
    },
    {
      "epoch": 0.9785407725321889,
      "grad_norm": 1.635040044784546,
      "learning_rate": 1.411506063912882e-08,
      "loss": 0.2365,
      "step": 228
    },
    {
      "epoch": 0.9828326180257511,
      "grad_norm": 1.5071089267730713,
      "learning_rate": 9.035169901754902e-09,
      "loss": 0.1913,
      "step": 229
    },
    {
      "epoch": 0.9871244635193133,
      "grad_norm": 1.8444362878799438,
      "learning_rate": 5.082953003528457e-09,
      "loss": 0.2008,
      "step": 230
    },
    {
      "epoch": 0.9914163090128756,
      "grad_norm": 1.3740102052688599,
      "learning_rate": 2.2593029204076578e-09,
      "loss": 0.1962,
      "step": 231
    },
    {
      "epoch": 0.9957081545064378,
      "grad_norm": 1.8837482929229736,
      "learning_rate": 5.648576365169245e-10,
      "loss": 0.2168,
      "step": 232
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.6341578960418701,
      "learning_rate": 0.0,
      "loss": 0.2153,
      "step": 233
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.36973223090171814,
      "eval_runtime": 3.4823,
      "eval_samples_per_second": 66.049,
      "eval_steps_per_second": 2.872,
      "step": 233
    },
    {
      "epoch": 1.0,
      "step": 233,
      "total_flos": 8.156803263102976e+16,
      "train_loss": 0.28537011958754627,
      "train_runtime": 757.3361,
      "train_samples_per_second": 14.709,
      "train_steps_per_second": 0.308
    }
  ],
  "logging_steps": 1,
  "max_steps": 233,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 800,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8.156803263102976e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}