{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 100000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005,
      "grad_norm": 0.645419180393219,
      "learning_rate": 3.326666666666667e-06,
      "loss": 0.7848,
      "step": 500
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.199676513671875,
      "learning_rate": 6.660000000000001e-06,
      "loss": 0.5106,
      "step": 1000
    },
    {
      "epoch": 0.015,
      "grad_norm": 2.394542694091797,
      "learning_rate": 9.993333333333333e-06,
      "loss": 0.4416,
      "step": 1500
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8768391609191895,
      "learning_rate": 1.3326666666666667e-05,
      "loss": 0.3916,
      "step": 2000
    },
    {
      "epoch": 0.025,
      "grad_norm": 2.688706874847412,
      "learning_rate": 1.666e-05,
      "loss": 0.3804,
      "step": 2500
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.755723476409912,
      "learning_rate": 1.9993333333333337e-05,
      "loss": 0.3683,
      "step": 3000
    },
    {
      "epoch": 0.035,
      "grad_norm": 2.6755011081695557,
      "learning_rate": 1.9998694075951004e-05,
      "loss": 0.3715,
      "step": 3500
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.1302595138549805,
      "learning_rate": 1.9994766172915367e-05,
      "loss": 0.3406,
      "step": 4000
    },
    {
      "epoch": 0.045,
      "grad_norm": 1.452305555343628,
      "learning_rate": 1.9988217315672814e-05,
      "loss": 0.343,
      "step": 4500
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.0988876819610596,
      "learning_rate": 1.997904922154766e-05,
      "loss": 0.326,
      "step": 5000
    },
    {
      "epoch": 0.055,
      "grad_norm": 3.2653279304504395,
      "learning_rate": 1.99672642947137e-05,
      "loss": 0.3305,
      "step": 5500
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.114853858947754,
      "learning_rate": 1.995286562556374e-05,
      "loss": 0.3307,
      "step": 6000
    },
    {
      "epoch": 0.065,
      "grad_norm": 4.524819374084473,
      "learning_rate": 1.993585698989922e-05,
      "loss": 0.3335,
      "step": 6500
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.944462537765503,
      "learning_rate": 1.9916242847940064e-05,
      "loss": 0.3287,
      "step": 7000
    },
    {
      "epoch": 0.075,
      "grad_norm": 2.251349687576294,
      "learning_rate": 1.9894028343155063e-05,
      "loss": 0.3319,
      "step": 7500
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.8081364631652832,
      "learning_rate": 1.9869219300913088e-05,
      "loss": 0.3294,
      "step": 8000
    },
    {
      "epoch": 0.085,
      "grad_norm": 2.8510942459106445,
      "learning_rate": 1.9841822226955496e-05,
      "loss": 0.3133,
      "step": 8500
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.03712797164917,
      "learning_rate": 1.9811844305690107e-05,
      "loss": 0.315,
      "step": 9000
    },
    {
      "epoch": 0.095,
      "grad_norm": 2.4897940158843994,
      "learning_rate": 1.977929339830722e-05,
      "loss": 0.3164,
      "step": 9500
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.979016900062561,
      "learning_rate": 1.9744178040718144e-05,
      "loss": 0.3087,
      "step": 10000
    },
    {
      "epoch": 0.105,
      "grad_norm": 3.3440868854522705,
      "learning_rate": 1.9706507441316814e-05,
      "loss": 0.3188,
      "step": 10500
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.8600707054138184,
      "learning_rate": 1.9666291478565047e-05,
      "loss": 0.3251,
      "step": 11000
    },
    {
      "epoch": 0.115,
      "grad_norm": 2.5654237270355225,
      "learning_rate": 1.962354069840208e-05,
      "loss": 0.3131,
      "step": 11500
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.2061824798583984,
      "learning_rate": 1.9578266311479082e-05,
      "loss": 0.3149,
      "step": 12000
    },
    {
      "epoch": 0.125,
      "grad_norm": 2.5924246311187744,
      "learning_rate": 1.9530480190219365e-05,
      "loss": 0.32,
      "step": 12500
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.142591714859009,
      "learning_rate": 1.948019486570503e-05,
      "loss": 0.3053,
      "step": 13000
    },
    {
      "epoch": 0.135,
      "grad_norm": 3.261009693145752,
      "learning_rate": 1.942742352439093e-05,
      "loss": 0.3025,
      "step": 13500
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.1369099617004395,
      "learning_rate": 1.937218000464673e-05,
      "loss": 0.3112,
      "step": 14000
    },
    {
      "epoch": 0.145,
      "grad_norm": 3.518235445022583,
      "learning_rate": 1.931447879312804e-05,
      "loss": 0.3122,
      "step": 14500
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.573578119277954,
      "learning_rate": 1.9254335020977542e-05,
      "loss": 0.3092,
      "step": 15000
    },
    {
      "epoch": 0.155,
      "grad_norm": 2.894007921218872,
      "learning_rate": 1.9191764459857124e-05,
      "loss": 0.3033,
      "step": 15500
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.554785966873169,
      "learning_rate": 1.9126783517812005e-05,
      "loss": 0.2888,
      "step": 16000
    },
    {
      "epoch": 0.165,
      "grad_norm": 2.739440679550171,
      "learning_rate": 1.9059409234968034e-05,
      "loss": 0.2966,
      "step": 16500
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.247506022453308,
      "learning_rate": 1.8989659279063214e-05,
      "loss": 0.2891,
      "step": 17000
    },
    {
      "epoch": 0.175,
      "grad_norm": 3.7641541957855225,
      "learning_rate": 1.8917551940814615e-05,
      "loss": 0.2955,
      "step": 17500
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.1178364753723145,
      "learning_rate": 1.8843106129121993e-05,
      "loss": 0.2982,
      "step": 18000
    },
    {
      "epoch": 0.185,
      "grad_norm": 29.405061721801758,
      "learning_rate": 1.8766341366109214e-05,
      "loss": 0.2907,
      "step": 18500
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.0723931789398193,
      "learning_rate": 1.8687277782004954e-05,
      "loss": 0.2789,
      "step": 19000
    },
    {
      "epoch": 0.195,
      "grad_norm": 3.0163700580596924,
      "learning_rate": 1.8605936109863872e-05,
      "loss": 0.2994,
      "step": 19500
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.1900928020477295,
      "learning_rate": 1.8522337680129728e-05,
      "loss": 0.2907,
      "step": 20000
    },
    {
      "epoch": 0.205,
      "grad_norm": 2.2802605628967285,
      "learning_rate": 1.843650441504186e-05,
      "loss": 0.2931,
      "step": 20500
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.1744939088821411,
      "learning_rate": 1.8348458822886437e-05,
      "loss": 0.2895,
      "step": 21000
    },
    {
      "epoch": 0.215,
      "grad_norm": 2.543396472930908,
      "learning_rate": 1.8258223992094055e-05,
      "loss": 0.2997,
      "step": 21500
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0287576913833618,
      "learning_rate": 1.816582358518521e-05,
      "loss": 0.2889,
      "step": 22000
    },
    {
      "epoch": 0.225,
      "grad_norm": 2.6123692989349365,
      "learning_rate": 1.80712818325652e-05,
      "loss": 0.2905,
      "step": 22500
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.583251476287842,
      "learning_rate": 1.7974623526170113e-05,
      "loss": 0.2957,
      "step": 23000
    },
    {
      "epoch": 0.235,
      "grad_norm": 1.5247690677642822,
      "learning_rate": 1.78758740129656e-05,
      "loss": 0.2853,
      "step": 23500
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.5719194412231445,
      "learning_rate": 1.777505918830004e-05,
      "loss": 0.2855,
      "step": 24000
    },
    {
      "epoch": 0.245,
      "grad_norm": 2.6126811504364014,
      "learning_rate": 1.767220548911395e-05,
      "loss": 0.2705,
      "step": 24500
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.8367679119110107,
      "learning_rate": 1.756733988700737e-05,
      "loss": 0.2751,
      "step": 25000
    },
    {
      "epoch": 0.255,
      "grad_norm": 2.151132106781006,
      "learning_rate": 1.7460489881166986e-05,
      "loss": 0.2855,
      "step": 25500
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.9020886421203613,
      "learning_rate": 1.7351683491154984e-05,
      "loss": 0.297,
      "step": 26000
    },
    {
      "epoch": 0.265,
      "grad_norm": 2.8366949558258057,
      "learning_rate": 1.7240949249561402e-05,
      "loss": 0.2825,
      "step": 26500
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.195739507675171,
      "learning_rate": 1.7128316194521923e-05,
      "loss": 0.2697,
      "step": 27000
    },
    {
      "epoch": 0.275,
      "grad_norm": 3.6873130798339844,
      "learning_rate": 1.7013813862103163e-05,
      "loss": 0.2751,
      "step": 27500
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.151796579360962,
      "learning_rate": 1.689747227855733e-05,
      "loss": 0.2797,
      "step": 28000
    },
    {
      "epoch": 0.285,
      "grad_norm": 2.8154728412628174,
      "learning_rate": 1.677932195244836e-05,
      "loss": 0.2804,
      "step": 28500
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.2176313400268555,
      "learning_rate": 1.6659393866651585e-05,
      "loss": 0.2699,
      "step": 29000
    },
    {
      "epoch": 0.295,
      "grad_norm": 3.4895665645599365,
      "learning_rate": 1.6537719470228993e-05,
      "loss": 0.2759,
      "step": 29500
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.498751401901245,
      "learning_rate": 1.6414330670182278e-05,
      "loss": 0.2769,
      "step": 30000
    },
    {
      "epoch": 0.305,
      "grad_norm": 2.554347038269043,
      "learning_rate": 1.6289259823085765e-05,
      "loss": 0.2839,
      "step": 30500
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.8352137804031372,
      "learning_rate": 1.6162539726601477e-05,
      "loss": 0.2804,
      "step": 31000
    },
    {
      "epoch": 0.315,
      "grad_norm": 3.4627280235290527,
      "learning_rate": 1.603420361087849e-05,
      "loss": 0.2793,
      "step": 31500
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.391278862953186,
      "learning_rate": 1.5904285129838924e-05,
      "loss": 0.2807,
      "step": 32000
    },
    {
      "epoch": 0.325,
      "grad_norm": 2.66107439994812,
      "learning_rate": 1.577281835235279e-05,
      "loss": 0.2721,
      "step": 32500
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.1572623252868652,
      "learning_rate": 1.563983775330399e-05,
      "loss": 0.27,
      "step": 33000
    },
    {
      "epoch": 0.335,
      "grad_norm": 1.8128657341003418,
      "learning_rate": 1.5505378204549907e-05,
      "loss": 0.2781,
      "step": 33500
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.731980800628662,
      "learning_rate": 1.536947496577684e-05,
      "loss": 0.2789,
      "step": 34000
    },
    {
      "epoch": 0.345,
      "grad_norm": 2.5709681510925293,
      "learning_rate": 1.5232163675253793e-05,
      "loss": 0.2833,
      "step": 34500
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.9191770553588867,
      "learning_rate": 1.5093480340486922e-05,
      "loss": 0.2701,
      "step": 35000
    },
    {
      "epoch": 0.355,
      "grad_norm": 2.5929877758026123,
      "learning_rate": 1.495346132877721e-05,
      "loss": 0.2694,
      "step": 35500
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.3980495929718018,
      "learning_rate": 1.4812143357683773e-05,
      "loss": 0.256,
      "step": 36000
    },
    {
      "epoch": 0.365,
      "grad_norm": 2.3195793628692627,
      "learning_rate": 1.4669563485395315e-05,
      "loss": 0.2733,
      "step": 36500
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.2807137966156006,
      "learning_rate": 1.4525759101012249e-05,
      "loss": 0.2779,
      "step": 37000
    },
    {
      "epoch": 0.375,
      "grad_norm": 2.4870142936706543,
      "learning_rate": 1.4380767914742074e-05,
      "loss": 0.2714,
      "step": 37500
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.8473994731903076,
      "learning_rate": 1.4234627948010497e-05,
      "loss": 0.2607,
      "step": 38000
    },
    {
      "epoch": 0.385,
      "grad_norm": 2.424431085586548,
      "learning_rate": 1.4087377523491016e-05,
      "loss": 0.2756,
      "step": 38500
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.063149929046631,
      "learning_rate": 1.3939055255055419e-05,
      "loss": 0.2603,
      "step": 39000
    },
    {
      "epoch": 0.395,
      "grad_norm": 3.3082692623138428,
      "learning_rate": 1.3789700037647988e-05,
      "loss": 0.268,
      "step": 39500
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.0695600509643555,
      "learning_rate": 1.363935103708598e-05,
      "loss": 0.259,
      "step": 40000
    },
    {
      "epoch": 0.405,
      "grad_norm": 4.004266738891602,
      "learning_rate": 1.3488047679789052e-05,
      "loss": 0.2739,
      "step": 40500
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.9311223030090332,
      "learning_rate": 1.3335829642440372e-05,
      "loss": 0.2634,
      "step": 41000
    },
    {
      "epoch": 0.415,
      "grad_norm": 2.7343368530273438,
      "learning_rate": 1.3182736841582109e-05,
      "loss": 0.2726,
      "step": 41500
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.1848347187042236,
      "learning_rate": 1.3028809423147997e-05,
      "loss": 0.2793,
      "step": 42000
    },
    {
      "epoch": 0.425,
      "grad_norm": 2.7985005378723145,
      "learning_rate": 1.2874087751935756e-05,
      "loss": 0.2702,
      "step": 42500
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.09635853767395,
      "learning_rate": 1.2718612401022148e-05,
      "loss": 0.2684,
      "step": 43000
    },
    {
      "epoch": 0.435,
      "grad_norm": 2.377711534500122,
      "learning_rate": 1.2562424141123364e-05,
      "loss": 0.2701,
      "step": 43500
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.9348018169403076,
      "learning_rate": 1.2405563929903614e-05,
      "loss": 0.2693,
      "step": 44000
    },
    {
      "epoch": 0.445,
      "grad_norm": 3.3448026180267334,
      "learning_rate": 1.2248072901234715e-05,
      "loss": 0.2637,
      "step": 44500
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.9979649782180786,
      "learning_rate": 1.2089992354409415e-05,
      "loss": 0.268,
      "step": 45000
    },
    {
      "epoch": 0.455,
      "grad_norm": 2.439824342727661,
      "learning_rate": 1.1931363743311377e-05,
      "loss": 0.2646,
      "step": 45500
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.9513328075408936,
      "learning_rate": 1.1772228665544634e-05,
      "loss": 0.2629,
      "step": 46000
    },
    {
      "epoch": 0.465,
      "grad_norm": 1.9054509401321411,
      "learning_rate": 1.1612628851525313e-05,
      "loss": 0.2691,
      "step": 46500
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.5216522216796875,
      "learning_rate": 1.1452606153538577e-05,
      "loss": 0.2638,
      "step": 47000
    },
    {
      "epoch": 0.475,
      "grad_norm": 1.3118677139282227,
      "learning_rate": 1.129220253476358e-05,
      "loss": 0.2656,
      "step": 47500
    },
    {
      "epoch": 0.48,
      "grad_norm": 3.147639274597168,
      "learning_rate": 1.1131460058269342e-05,
      "loss": 0.2543,
      "step": 48000
    },
    {
      "epoch": 0.485,
      "grad_norm": 2.3028457164764404,
      "learning_rate": 1.0970420875984453e-05,
      "loss": 0.2633,
      "step": 48500
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.496572256088257,
      "learning_rate": 1.0809127217643462e-05,
      "loss": 0.2629,
      "step": 49000
    },
    {
      "epoch": 0.495,
      "grad_norm": 2.1966664791107178,
      "learning_rate": 1.0647621379712843e-05,
      "loss": 0.2653,
      "step": 49500
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.8232370615005493,
      "learning_rate": 1.0485945714299487e-05,
      "loss": 0.2539,
      "step": 50000
    },
    {
      "epoch": 0.505,
      "grad_norm": 3.019258975982666,
      "learning_rate": 1.0324142618044612e-05,
      "loss": 0.2688,
      "step": 50500
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.8619256019592285,
      "learning_rate": 1.0162254521005956e-05,
      "loss": 0.2633,
      "step": 51000
    },
    {
      "epoch": 0.515,
      "grad_norm": 2.05895733833313,
      "learning_rate": 1.0000323875531241e-05,
      "loss": 0.26,
      "step": 51500
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.8814473152160645,
      "learning_rate": 9.838393145125787e-06,
      "loss": 0.2601,
      "step": 52000
    },
    {
      "epoch": 0.525,
      "grad_norm": 3.4948761463165283,
      "learning_rate": 9.676504793317163e-06,
      "loss": 0.2584,
      "step": 52500
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.9963233470916748,
      "learning_rate": 9.514701272519918e-06,
      "loss": 0.2542,
      "step": 53000
    },
    {
      "epoch": 0.535,
      "grad_norm": 1.5181660652160645,
      "learning_rate": 9.353025012903112e-06,
      "loss": 0.2635,
      "step": 53500
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.8904043436050415,
      "learning_rate": 9.191518411263777e-06,
      "loss": 0.2419,
      "step": 54000
    },
    {
      "epoch": 0.545,
      "grad_norm": 2.7805871963500977,
      "learning_rate": 9.03022381990908e-06,
      "loss": 0.269,
      "step": 54500
    },
    {
      "epoch": 0.55,
      "grad_norm": 4.442915439605713,
      "learning_rate": 8.869183535550167e-06,
      "loss": 0.2721,
      "step": 55000
    },
    {
      "epoch": 0.555,
      "grad_norm": 3.1872241497039795,
      "learning_rate": 8.708439788210573e-06,
      "loss": 0.2685,
      "step": 55500
    },
    {
      "epoch": 0.56,
      "grad_norm": 3.0732016563415527,
      "learning_rate": 8.548034730152153e-06,
      "loss": 0.2622,
      "step": 56000
    },
    {
      "epoch": 0.565,
      "grad_norm": 2.145470380783081,
      "learning_rate": 8.388010424821354e-06,
      "loss": 0.2647,
      "step": 56500
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.276179313659668,
      "learning_rate": 8.22840883581883e-06,
      "loss": 0.252,
      "step": 57000
    },
    {
      "epoch": 0.575,
      "grad_norm": 2.082150459289551,
      "learning_rate": 8.069271815895196e-06,
      "loss": 0.2804,
      "step": 57500
    },
    {
      "epoch": 0.58,
      "grad_norm": 4.978368282318115,
      "learning_rate": 7.910641095975886e-06,
      "loss": 0.2577,
      "step": 58000
    },
    {
      "epoch": 0.585,
      "grad_norm": 0.9249425530433655,
      "learning_rate": 7.752558274217919e-06,
      "loss": 0.2536,
      "step": 58500
    },
    {
      "epoch": 0.59,
      "grad_norm": 3.2407662868499756,
      "learning_rate": 7.595064805101538e-06,
      "loss": 0.2502,
      "step": 59000
    },
    {
      "epoch": 0.595,
      "grad_norm": 1.8381527662277222,
      "learning_rate": 7.438201988559449e-06,
      "loss": 0.2648,
      "step": 59500
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.238192081451416,
      "learning_rate": 7.282010959146662e-06,
      "loss": 0.2644,
      "step": 60000
    },
    {
      "epoch": 0.605,
      "grad_norm": 4.041945457458496,
      "learning_rate": 7.1265326752536545e-06,
      "loss": 0.2633,
      "step": 60500
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.593888521194458,
      "learning_rate": 6.971807908365737e-06,
      "loss": 0.2582,
      "step": 61000
    },
    {
      "epoch": 0.615,
      "grad_norm": 2.6872119903564453,
      "learning_rate": 6.817877232371453e-06,
      "loss": 0.2648,
      "step": 61500
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.848517894744873,
      "learning_rate": 6.66478101292277e-06,
      "loss": 0.25,
      "step": 62000
    },
    {
      "epoch": 0.625,
      "grad_norm": 3.348888635635376,
      "learning_rate": 6.5125593968498855e-06,
      "loss": 0.245,
      "step": 62500
    },
    {
      "epoch": 0.63,
      "grad_norm": 3.613619089126587,
      "learning_rate": 6.361252301633422e-06,
      "loss": 0.2632,
      "step": 63000
    },
    {
      "epoch": 0.635,
      "grad_norm": 2.6312239170074463,
      "learning_rate": 6.210899404936761e-06,
      "loss": 0.253,
      "step": 63500
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.473121166229248,
      "learning_rate": 6.061540134201248e-06,
      "loss": 0.2563,
      "step": 64000
    },
    {
      "epoch": 0.645,
      "grad_norm": 3.9563801288604736,
      "learning_rate": 5.913213656307049e-06,
      "loss": 0.2603,
      "step": 64500
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.275865077972412,
      "learning_rate": 5.765958867302311e-06,
      "loss": 0.2513,
      "step": 65000
    },
    {
      "epoch": 0.655,
      "grad_norm": 2.183001756668091,
      "learning_rate": 5.619814382203341e-06,
      "loss": 0.244,
      "step": 65500
    },
    {
      "epoch": 0.66,
      "grad_norm": 3.7053382396698,
      "learning_rate": 5.474818524868482e-06,
      "loss": 0.2729,
      "step": 66000
    },
    {
      "epoch": 0.665,
      "grad_norm": 3.171715021133423,
      "learning_rate": 5.331009317948359e-06,
      "loss": 0.267,
      "step": 66500
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.751267671585083,
      "learning_rate": 5.188424472915091e-06,
      "loss": 0.2707,
      "step": 67000
    },
    {
      "epoch": 0.675,
      "grad_norm": 2.5205554962158203,
      "learning_rate": 5.047101380173111e-06,
      "loss": 0.2445,
      "step": 67500
    },
    {
      "epoch": 0.68,
      "grad_norm": 3.4921984672546387,
      "learning_rate": 4.907077099254194e-06,
      "loss": 0.2596,
      "step": 68000
    },
    {
      "epoch": 0.685,
      "grad_norm": 2.795581102371216,
      "learning_rate": 4.7683883490991985e-06,
      "loss": 0.2624,
      "step": 68500
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9924126863479614,
      "learning_rate": 4.6310714984292096e-06,
      "loss": 0.2619,
      "step": 69000
    },
    {
      "epoch": 0.695,
      "grad_norm": 0.9289371371269226,
      "learning_rate": 4.495162556208438e-06,
      "loss": 0.2499,
      "step": 69500
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.991086483001709,
      "learning_rate": 4.360697162201508e-06,
      "loss": 0.2542,
      "step": 70000
    },
    {
      "epoch": 0.705,
      "grad_norm": 3.306328296661377,
      "learning_rate": 4.2277105776275395e-06,
      "loss": 0.2491,
      "step": 70500
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.5418620109558105,
      "learning_rate": 4.096237675913521e-06,
      "loss": 0.2611,
      "step": 71000
    },
    {
      "epoch": 0.715,
      "grad_norm": 2.180978298187256,
      "learning_rate": 3.966312933549325e-06,
      "loss": 0.24,
      "step": 71500
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.262224197387695,
      "learning_rate": 3.837970421046881e-06,
      "loss": 0.2494,
      "step": 72000
    },
    {
      "epoch": 0.725,
      "grad_norm": 3.8794145584106445,
      "learning_rate": 3.7112437940057543e-06,
      "loss": 0.2497,
      "step": 72500
    },
    {
      "epoch": 0.73,
      "grad_norm": 3.2217772006988525,
      "learning_rate": 3.5861662842875534e-06,
      "loss": 0.2417,
      "step": 73000
    },
    {
      "epoch": 0.735,
      "grad_norm": 4.515202522277832,
      "learning_rate": 3.4627706913014503e-06,
      "loss": 0.2543,
      "step": 73500
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.5757079124450684,
      "learning_rate": 3.3410893734031046e-06,
      "loss": 0.2623,
      "step": 74000
    },
    {
      "epoch": 0.745,
      "grad_norm": 2.459864377975464,
      "learning_rate": 3.2211542394092443e-06,
      "loss": 0.2519,
      "step": 74500
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.33870267868042,
      "learning_rate": 3.102996740230131e-06,
      "loss": 0.2439,
      "step": 75000
    },
    {
      "epoch": 0.755,
      "grad_norm": 1.7703856229782104,
      "learning_rate": 2.986647860622105e-06,
      "loss": 0.2603,
      "step": 75500
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.727771282196045,
      "learning_rate": 2.8721381110623724e-06,
      "loss": 0.2486,
      "step": 76000
    },
    {
      "epoch": 0.765,
      "grad_norm": 2.4445149898529053,
      "learning_rate": 2.7594975197481546e-06,
      "loss": 0.2497,
      "step": 76500
    },
    {
      "epoch": 0.77,
      "grad_norm": 3.971421480178833,
      "learning_rate": 2.64875562472232e-06,
      "loss": 0.2476,
      "step": 77000
    },
    {
      "epoch": 0.775,
      "grad_norm": 4.322368144989014,
      "learning_rate": 2.5399414661275444e-06,
      "loss": 0.2585,
      "step": 77500
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.7902648448944092,
      "learning_rate": 2.433083578591029e-06,
      "loss": 0.2473,
      "step": 78000
    },
    {
      "epoch": 0.785,
      "grad_norm": 3.231074094772339,
      "learning_rate": 2.3282099837418004e-06,
      "loss": 0.257,
      "step": 78500
    },
    {
      "epoch": 0.79,
      "grad_norm": 3.8117921352386475,
      "learning_rate": 2.2253481828625144e-06,
      "loss": 0.2655,
      "step": 79000
    },
    {
      "epoch": 0.795,
      "grad_norm": 3.7752602100372314,
      "learning_rate": 2.1245251496777274e-06,
      "loss": 0.2548,
      "step": 79500
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.201462745666504,
      "learning_rate": 2.025767323280502e-06,
      "loss": 0.2546,
      "step": 80000
    },
    {
      "epoch": 0.805,
      "grad_norm": 5.425679683685303,
      "learning_rate": 1.929100601199211e-06,
      "loss": 0.2383,
      "step": 80500
    },
    {
      "epoch": 0.81,
      "grad_norm": 3.4955568313598633,
      "learning_rate": 1.834550332606365e-06,
      "loss": 0.2347,
      "step": 81000
    },
    {
      "epoch": 0.815,
      "grad_norm": 2.8629684448242188,
      "learning_rate": 1.742141311671227e-06,
      "loss": 0.2489,
      "step": 81500
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.249149799346924,
      "learning_rate": 1.6518977710579708e-06,
      "loss": 0.2682,
      "step": 82000
    },
    {
      "epoch": 0.825,
      "grad_norm": 2.124053716659546,
      "learning_rate": 1.5638433755710892e-06,
      "loss": 0.2506,
      "step": 82500
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.384835720062256,
      "learning_rate": 1.4780012159497037e-06,
      "loss": 0.2636,
      "step": 83000
    },
    {
      "epoch": 0.835,
      "grad_norm": 2.0018503665924072,
      "learning_rate": 1.3943938028124216e-06,
      "loss": 0.2565,
      "step": 83500
    },
    {
      "epoch": 0.84,
      "grad_norm": 6.473081588745117,
      "learning_rate": 1.3130430607543166e-06,
      "loss": 0.2561,
      "step": 84000
    },
    {
      "epoch": 0.845,
      "grad_norm": 5.408079147338867,
      "learning_rate": 1.2339703225975818e-06,
      "loss": 0.2511,
      "step": 84500
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.099701404571533,
      "learning_rate": 1.1571963237973693e-06,
      "loss": 0.2457,
      "step": 85000
    },
    {
      "epoch": 0.855,
      "grad_norm": 3.5256073474884033,
      "learning_rate": 1.0827411970042723e-06,
      "loss": 0.2531,
      "step": 85500
    },
    {
      "epoch": 0.86,
      "grad_norm": 3.8921525478363037,
      "learning_rate": 1.010624466784892e-06,
      "loss": 0.2483,
      "step": 86000
    },
    {
      "epoch": 0.865,
      "grad_norm": 4.602254867553711,
      "learning_rate": 9.40865044501853e-07,
      "loss": 0.2496,
      "step": 86500
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.7740793228149414,
      "learning_rate": 8.734812233546319e-07,
      "loss": 0.2526,
      "step": 87000
    },
    {
      "epoch": 0.875,
      "grad_norm": 3.1108462810516357,
      "learning_rate": 8.084906735824771e-07,
      "loss": 0.2532,
      "step": 87500
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.351101875305176,
      "learning_rate": 7.45910437830697e-07,
      "loss": 0.2497,
      "step": 88000
    },
    {
      "epoch": 0.885,
      "grad_norm": 3.348724842071533,
      "learning_rate": 6.857569266815233e-07,
      "loss": 0.2463,
      "step": 88500
    },
    {
      "epoch": 0.89,
      "grad_norm": 3.5071520805358887,
      "learning_rate": 6.2804591435072e-07,
      "loss": 0.2609,
      "step": 89000
    },
    {
      "epoch": 0.895,
      "grad_norm": 1.3788470029830933,
      "learning_rate": 5.727925345510644e-07,
      "loss": 0.243,
      "step": 89500
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.7715322971343994,
      "learning_rate": 5.200112765237985e-07,
      "loss": 0.2501,
      "step": 90000
    },
    {
      "epoch": 0.905,
      "grad_norm": 1.8194293975830078,
      "learning_rate": 4.697159812390739e-07,
      "loss": 0.245,
      "step": 90500
    },
    {
      "epoch": 0.91,
      "grad_norm": 3.7787585258483887,
      "learning_rate": 4.21919837766398e-07,
      "loss": 0.2481,
      "step": 91000
    },
    {
      "epoch": 0.915,
      "grad_norm": 2.616973400115967,
      "learning_rate": 3.766353798160283e-07,
      "loss": 0.243,
      "step": 91500
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.718681812286377,
      "learning_rate": 3.338744824522266e-07,
      "loss": 0.2478,
      "step": 92000
    },
    {
      "epoch": 0.925,
      "grad_norm": 2.087679386138916,
      "learning_rate": 2.9364835897922606e-07,
      "loss": 0.2513,
      "step": 92500
    },
    {
      "epoch": 0.93,
      "grad_norm": 5.528986930847168,
      "learning_rate": 2.5596755800073593e-07,
      "loss": 0.2471,
      "step": 93000
    },
    {
      "epoch": 0.935,
      "grad_norm": 2.7250564098358154,
      "learning_rate": 2.208419606537593e-07,
      "loss": 0.2524,
      "step": 93500
    },
    {
      "epoch": 0.94,
      "grad_norm": 3.059098720550537,
      "learning_rate": 1.8828077801743006e-07,
      "loss": 0.256,
      "step": 94000
    },
    {
      "epoch": 0.945,
      "grad_norm": 2.8698060512542725,
      "learning_rate": 1.5829254869756395e-07,
      "loss": 0.2565,
      "step": 94500
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.0529208183288574,
      "learning_rate": 1.3088513658756962e-07,
      "loss": 0.2488,
      "step": 95000
    },
    {
      "epoch": 0.955,
      "grad_norm": 1.9833836555480957,
      "learning_rate": 1.060657288062672e-07,
      "loss": 0.25,
      "step": 95500
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.4886512756347656,
      "learning_rate": 8.384083381319464e-08,
      "loss": 0.2439,
      "step": 96000
    },
    {
      "epoch": 0.965,
      "grad_norm": 4.0352091789245605,
      "learning_rate": 6.421627970187971e-08,
      "loss": 0.2462,
      "step": 96500
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.697331190109253,
      "learning_rate": 4.719721267152033e-08,
      "loss": 0.2584,
      "step": 97000
    },
    {
      "epoch": 0.975,
      "grad_norm": 4.4305739402771,
      "learning_rate": 3.2788095677480644e-08,
      "loss": 0.2513,
      "step": 97500
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.2790756225585938,
      "learning_rate": 2.099270726096503e-08,
      "loss": 0.252,
      "step": 98000
    },
    {
      "epoch": 0.985,
      "grad_norm": 4.5093183517456055,
      "learning_rate": 1.1814140558158438e-08,
      "loss": 0.2452,
      "step": 98500
    },
    {
      "epoch": 0.99,
      "grad_norm": 3.447608709335327,
      "learning_rate": 5.254802489100819e-09,
      "loss": 0.2532,
      "step": 99000
    },
    {
      "epoch": 0.995,
      "grad_norm": 1.7088748216629028,
      "learning_rate": 1.3164131265197555e-09,
      "loss": 0.2532,
      "step": 99500
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.227078437805176,
      "learning_rate": 5.2447679532718895e-15,
      "loss": 0.246,
      "step": 100000
    },
    {
      "epoch": 1.0,
      "step": 100000,
      "total_flos": 1.1225588428538726e+18,
      "train_loss": 0.2772226203155518,
      "train_runtime": 23565.4339,
      "train_samples_per_second": 4.244,
      "train_steps_per_second": 4.244
    }
  ],
  "logging_steps": 500,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1225588428538726e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}