{
  "best_metric": 0.9823666485038896,
  "best_model_checkpoint": "FFPP-Raw_1FPS_faces-expand-40-aligned\\checkpoint-4044",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 4044,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 8.450878143310547,
      "learning_rate": 1.234567901234568e-06,
      "loss": 0.7645,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.411558628082275,
      "learning_rate": 2.469135802469136e-06,
      "loss": 0.6817,
      "step": 20
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.373965740203857,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.5998,
      "step": 30
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.776858329772949,
      "learning_rate": 4.938271604938272e-06,
      "loss": 0.5307,
      "step": 40
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.085646629333496,
      "learning_rate": 6.172839506172839e-06,
      "loss": 0.525,
      "step": 50
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.5710902214050293,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.4941,
      "step": 60
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.6829416751861572,
      "learning_rate": 8.641975308641975e-06,
      "loss": 0.5426,
      "step": 70
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.156679630279541,
      "learning_rate": 9.876543209876543e-06,
      "loss": 0.4969,
      "step": 80
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.465975284576416,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.5458,
      "step": 90
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.412788152694702,
      "learning_rate": 1.2345679012345678e-05,
      "loss": 0.5162,
      "step": 100
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.6743991374969482,
      "learning_rate": 1.3580246913580247e-05,
      "loss": 0.5121,
      "step": 110
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.343439102172852,
      "learning_rate": 1.4814814814814815e-05,
      "loss": 0.5369,
      "step": 120
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.486733913421631,
      "learning_rate": 1.604938271604938e-05,
      "loss": 0.5169,
      "step": 130
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.967243194580078,
      "learning_rate": 1.728395061728395e-05,
      "loss": 0.5097,
      "step": 140
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.3753137588500977,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.4909,
      "step": 150
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.698143005371094,
      "learning_rate": 1.9753086419753087e-05,
      "loss": 0.5047,
      "step": 160
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.35822057723999,
      "learning_rate": 2.0987654320987655e-05,
      "loss": 0.5101,
      "step": 170
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.044003009796143,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.4809,
      "step": 180
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.79054594039917,
      "learning_rate": 2.345679012345679e-05,
      "loss": 0.5084,
      "step": 190
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.470571041107178,
      "learning_rate": 2.4691358024691357e-05,
      "loss": 0.4427,
      "step": 200
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.30893611907959,
      "learning_rate": 2.5925925925925925e-05,
      "loss": 0.4598,
      "step": 210
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.693650245666504,
      "learning_rate": 2.7160493827160493e-05,
      "loss": 0.4888,
      "step": 220
    },
    {
      "epoch": 0.17,
      "grad_norm": 6.367429733276367,
      "learning_rate": 2.839506172839506e-05,
      "loss": 0.4479,
      "step": 230
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.16640567779541,
      "learning_rate": 2.962962962962963e-05,
      "loss": 0.431,
      "step": 240
    },
    {
      "epoch": 0.19,
      "grad_norm": 11.808082580566406,
      "learning_rate": 3.08641975308642e-05,
      "loss": 0.4055,
      "step": 250
    },
    {
      "epoch": 0.19,
      "grad_norm": 21.1939640045166,
      "learning_rate": 3.209876543209876e-05,
      "loss": 0.3949,
      "step": 260
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.935243606567383,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.3778,
      "step": 270
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.449581146240234,
      "learning_rate": 3.45679012345679e-05,
      "loss": 0.3449,
      "step": 280
    },
    {
      "epoch": 0.22,
      "grad_norm": 8.583562850952148,
      "learning_rate": 3.580246913580247e-05,
      "loss": 0.3957,
      "step": 290
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.4094133377075195,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.3931,
      "step": 300
    },
    {
      "epoch": 0.23,
      "grad_norm": 7.312207221984863,
      "learning_rate": 3.82716049382716e-05,
      "loss": 0.3241,
      "step": 310
    },
    {
      "epoch": 0.24,
      "grad_norm": 14.610054969787598,
      "learning_rate": 3.950617283950617e-05,
      "loss": 0.3262,
      "step": 320
    },
    {
      "epoch": 0.24,
      "grad_norm": 11.481966018676758,
      "learning_rate": 4.074074074074074e-05,
      "loss": 0.3301,
      "step": 330
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.101536750793457,
      "learning_rate": 4.197530864197531e-05,
      "loss": 0.3068,
      "step": 340
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.435379981994629,
      "learning_rate": 4.3209876543209875e-05,
      "loss": 0.413,
      "step": 350
    },
    {
      "epoch": 0.27,
      "grad_norm": 26.36928939819336,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.3646,
      "step": 360
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.532124996185303,
      "learning_rate": 4.567901234567901e-05,
      "loss": 0.4034,
      "step": 370
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.976860046386719,
      "learning_rate": 4.691358024691358e-05,
      "loss": 0.3523,
      "step": 380
    },
    {
      "epoch": 0.29,
      "grad_norm": 11.547489166259766,
      "learning_rate": 4.814814814814815e-05,
      "loss": 0.3585,
      "step": 390
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.796356201171875,
      "learning_rate": 4.938271604938271e-05,
      "loss": 0.3474,
      "step": 400
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.1311421394348145,
      "learning_rate": 4.993129980763946e-05,
      "loss": 0.3042,
      "step": 410
    },
    {
      "epoch": 0.31,
      "grad_norm": 18.114513397216797,
      "learning_rate": 4.979389942291839e-05,
      "loss": 0.3492,
      "step": 420
    },
    {
      "epoch": 0.32,
      "grad_norm": 14.068829536437988,
      "learning_rate": 4.965649903819731e-05,
      "loss": 0.2967,
      "step": 430
    },
    {
      "epoch": 0.33,
      "grad_norm": 22.513200759887695,
      "learning_rate": 4.951909865347623e-05,
      "loss": 0.3142,
      "step": 440
    },
    {
      "epoch": 0.33,
      "grad_norm": 17.25227165222168,
      "learning_rate": 4.9381698268755155e-05,
      "loss": 0.3429,
      "step": 450
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.3096699714660645,
      "learning_rate": 4.924429788403408e-05,
      "loss": 0.2944,
      "step": 460
    },
    {
      "epoch": 0.35,
      "grad_norm": 7.90877628326416,
      "learning_rate": 4.9106897499313e-05,
      "loss": 0.3029,
      "step": 470
    },
    {
      "epoch": 0.36,
      "grad_norm": 14.439558982849121,
      "learning_rate": 4.896949711459192e-05,
      "loss": 0.2862,
      "step": 480
    },
    {
      "epoch": 0.36,
      "grad_norm": 7.571442127227783,
      "learning_rate": 4.883209672987085e-05,
      "loss": 0.3197,
      "step": 490
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.224518775939941,
      "learning_rate": 4.8694696345149774e-05,
      "loss": 0.3173,
      "step": 500
    },
    {
      "epoch": 0.38,
      "grad_norm": 19.048572540283203,
      "learning_rate": 4.8557295960428687e-05,
      "loss": 0.3749,
      "step": 510
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.699029922485352,
      "learning_rate": 4.841989557570761e-05,
      "loss": 0.3774,
      "step": 520
    },
    {
      "epoch": 0.39,
      "grad_norm": 10.619318008422852,
      "learning_rate": 4.828249519098654e-05,
      "loss": 0.3161,
      "step": 530
    },
    {
      "epoch": 0.4,
      "grad_norm": 12.99872875213623,
      "learning_rate": 4.814509480626546e-05,
      "loss": 0.2833,
      "step": 540
    },
    {
      "epoch": 0.41,
      "grad_norm": 12.019877433776855,
      "learning_rate": 4.800769442154438e-05,
      "loss": 0.2371,
      "step": 550
    },
    {
      "epoch": 0.42,
      "grad_norm": 17.415578842163086,
      "learning_rate": 4.7870294036823306e-05,
      "loss": 0.2497,
      "step": 560
    },
    {
      "epoch": 0.42,
      "grad_norm": 10.935409545898438,
      "learning_rate": 4.7732893652102225e-05,
      "loss": 0.2674,
      "step": 570
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.685445785522461,
      "learning_rate": 4.759549326738115e-05,
      "loss": 0.3019,
      "step": 580
    },
    {
      "epoch": 0.44,
      "grad_norm": 6.198949337005615,
      "learning_rate": 4.745809288266007e-05,
      "loss": 0.2686,
      "step": 590
    },
    {
      "epoch": 0.45,
      "grad_norm": 14.602446556091309,
      "learning_rate": 4.7320692497939e-05,
      "loss": 0.2582,
      "step": 600
    },
    {
      "epoch": 0.45,
      "grad_norm": 23.308094024658203,
      "learning_rate": 4.718329211321792e-05,
      "loss": 0.3482,
      "step": 610
    },
    {
      "epoch": 0.46,
      "grad_norm": 19.93767547607422,
      "learning_rate": 4.7045891728496844e-05,
      "loss": 0.2962,
      "step": 620
    },
    {
      "epoch": 0.47,
      "grad_norm": 15.824058532714844,
      "learning_rate": 4.6908491343775764e-05,
      "loss": 0.2796,
      "step": 630
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.9304609298706055,
      "learning_rate": 4.6771090959054684e-05,
      "loss": 0.2712,
      "step": 640
    },
    {
      "epoch": 0.48,
      "grad_norm": 14.579902648925781,
      "learning_rate": 4.663369057433361e-05,
      "loss": 0.2783,
      "step": 650
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.262262344360352,
      "learning_rate": 4.649629018961254e-05,
      "loss": 0.2541,
      "step": 660
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.660795211791992,
      "learning_rate": 4.6358889804891456e-05,
      "loss": 0.2636,
      "step": 670
    },
    {
      "epoch": 0.5,
      "grad_norm": 24.701696395874023,
      "learning_rate": 4.6221489420170376e-05,
      "loss": 0.3344,
      "step": 680
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.420050621032715,
      "learning_rate": 4.60840890354493e-05,
      "loss": 0.3001,
      "step": 690
    },
    {
      "epoch": 0.52,
      "grad_norm": 18.686073303222656,
      "learning_rate": 4.594668865072822e-05,
      "loss": 0.2415,
      "step": 700
    },
    {
      "epoch": 0.53,
      "grad_norm": 13.247358322143555,
      "learning_rate": 4.580928826600714e-05,
      "loss": 0.3413,
      "step": 710
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.456878662109375,
      "learning_rate": 4.567188788128607e-05,
      "loss": 0.2764,
      "step": 720
    },
    {
      "epoch": 0.54,
      "grad_norm": 13.153331756591797,
      "learning_rate": 4.5534487496564995e-05,
      "loss": 0.2365,
      "step": 730
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.744629859924316,
      "learning_rate": 4.5397087111843915e-05,
      "loss": 0.2239,
      "step": 740
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.051612854003906,
      "learning_rate": 4.525968672712284e-05,
      "loss": 0.2357,
      "step": 750
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.455211639404297,
      "learning_rate": 4.512228634240176e-05,
      "loss": 0.2401,
      "step": 760
    },
    {
      "epoch": 0.57,
      "grad_norm": 14.734637260437012,
      "learning_rate": 4.498488595768068e-05,
      "loss": 0.2987,
      "step": 770
    },
    {
      "epoch": 0.58,
      "grad_norm": 13.91230583190918,
      "learning_rate": 4.484748557295961e-05,
      "loss": 0.1869,
      "step": 780
    },
    {
      "epoch": 0.59,
      "grad_norm": 11.488323211669922,
      "learning_rate": 4.4710085188238534e-05,
      "loss": 0.2134,
      "step": 790
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.693485736846924,
      "learning_rate": 4.4572684803517453e-05,
      "loss": 0.2026,
      "step": 800
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.261735916137695,
      "learning_rate": 4.443528441879637e-05,
      "loss": 0.2343,
      "step": 810
    },
    {
      "epoch": 0.61,
      "grad_norm": 6.214202404022217,
      "learning_rate": 4.42978840340753e-05,
      "loss": 0.2183,
      "step": 820
    },
    {
      "epoch": 0.62,
      "grad_norm": 7.414583683013916,
      "learning_rate": 4.4160483649354226e-05,
      "loss": 0.1992,
      "step": 830
    },
    {
      "epoch": 0.62,
      "grad_norm": 13.354372024536133,
      "learning_rate": 4.402308326463314e-05,
      "loss": 0.2622,
      "step": 840
    },
    {
      "epoch": 0.63,
      "grad_norm": 13.756591796875,
      "learning_rate": 4.3885682879912066e-05,
      "loss": 0.2207,
      "step": 850
    },
    {
      "epoch": 0.64,
      "grad_norm": 16.665782928466797,
      "learning_rate": 4.374828249519099e-05,
      "loss": 0.2232,
      "step": 860
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.784102439880371,
      "learning_rate": 4.361088211046991e-05,
      "loss": 0.2174,
      "step": 870
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.463679313659668,
      "learning_rate": 4.347348172574883e-05,
      "loss": 0.2443,
      "step": 880
    },
    {
      "epoch": 0.66,
      "grad_norm": 12.118477821350098,
      "learning_rate": 4.333608134102776e-05,
      "loss": 0.2259,
      "step": 890
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.19271183013916,
      "learning_rate": 4.319868095630668e-05,
      "loss": 0.2127,
      "step": 900
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.1488037109375,
      "learning_rate": 4.3061280571585604e-05,
      "loss": 0.2007,
      "step": 910
    },
    {
      "epoch": 0.68,
      "grad_norm": 6.0573320388793945,
      "learning_rate": 4.2923880186864524e-05,
      "loss": 0.2268,
      "step": 920
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.142855644226074,
      "learning_rate": 4.278647980214345e-05,
      "loss": 0.2263,
      "step": 930
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.926745414733887,
      "learning_rate": 4.264907941742237e-05,
      "loss": 0.2133,
      "step": 940
    },
    {
      "epoch": 0.7,
      "grad_norm": 12.452362060546875,
      "learning_rate": 4.25116790327013e-05,
      "loss": 0.2,
      "step": 950
    },
    {
      "epoch": 0.71,
      "grad_norm": 7.340023994445801,
      "learning_rate": 4.2374278647980216e-05,
      "loss": 0.1764,
      "step": 960
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.337459564208984,
      "learning_rate": 4.2236878263259136e-05,
      "loss": 0.2465,
      "step": 970
    },
    {
      "epoch": 0.73,
      "grad_norm": 15.628137588500977,
      "learning_rate": 4.209947787853806e-05,
      "loss": 0.2005,
      "step": 980
    },
    {
      "epoch": 0.73,
      "grad_norm": 5.059557914733887,
      "learning_rate": 4.196207749381699e-05,
      "loss": 0.2389,
      "step": 990
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.46613073348999,
      "learning_rate": 4.18246771090959e-05,
      "loss": 0.2134,
      "step": 1000
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.257927894592285,
      "learning_rate": 4.168727672437483e-05,
      "loss": 0.1958,
      "step": 1010
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.179327011108398,
      "learning_rate": 4.1549876339653755e-05,
      "loss": 0.1929,
      "step": 1020
    },
    {
      "epoch": 0.76,
      "grad_norm": 22.463008880615234,
      "learning_rate": 4.1412475954932675e-05,
      "loss": 0.2206,
      "step": 1030
    },
    {
      "epoch": 0.77,
      "grad_norm": 6.543388366699219,
      "learning_rate": 4.1275075570211595e-05,
      "loss": 0.1832,
      "step": 1040
    },
    {
      "epoch": 0.78,
      "grad_norm": 6.427690029144287,
      "learning_rate": 4.113767518549052e-05,
      "loss": 0.1845,
      "step": 1050
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.30201244354248,
      "learning_rate": 4.100027480076945e-05,
      "loss": 0.2071,
      "step": 1060
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.301946640014648,
      "learning_rate": 4.086287441604837e-05,
      "loss": 0.2033,
      "step": 1070
    },
    {
      "epoch": 0.8,
      "grad_norm": 12.093053817749023,
      "learning_rate": 4.072547403132729e-05,
      "loss": 0.1897,
      "step": 1080
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.62564754486084,
      "learning_rate": 4.0588073646606214e-05,
      "loss": 0.1791,
      "step": 1090
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.241898536682129,
      "learning_rate": 4.045067326188513e-05,
      "loss": 0.2097,
      "step": 1100
    },
    {
      "epoch": 0.82,
      "grad_norm": 6.859925746917725,
      "learning_rate": 4.031327287716406e-05,
      "loss": 0.192,
      "step": 1110
    },
    {
      "epoch": 0.83,
      "grad_norm": 13.545909881591797,
      "learning_rate": 4.017587249244298e-05,
      "loss": 0.2016,
      "step": 1120
    },
    {
      "epoch": 0.84,
      "grad_norm": 9.510741233825684,
      "learning_rate": 4.0038472107721906e-05,
      "loss": 0.2266,
      "step": 1130
    },
    {
      "epoch": 0.85,
      "grad_norm": 12.568158149719238,
      "learning_rate": 3.9901071723000826e-05,
      "loss": 0.2498,
      "step": 1140
    },
    {
      "epoch": 0.85,
      "grad_norm": 14.593520164489746,
      "learning_rate": 3.976367133827975e-05,
      "loss": 0.1981,
      "step": 1150
    },
    {
      "epoch": 0.86,
      "grad_norm": 5.609714508056641,
      "learning_rate": 3.962627095355867e-05,
      "loss": 0.1987,
      "step": 1160
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.593751907348633,
      "learning_rate": 3.948887056883759e-05,
      "loss": 0.1862,
      "step": 1170
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.881361961364746,
      "learning_rate": 3.935147018411652e-05,
      "loss": 0.2006,
      "step": 1180
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.768789291381836,
      "learning_rate": 3.9214069799395445e-05,
      "loss": 0.2235,
      "step": 1190
    },
    {
      "epoch": 0.89,
      "grad_norm": 7.065619468688965,
      "learning_rate": 3.907666941467436e-05,
      "loss": 0.1968,
      "step": 1200
    },
    {
      "epoch": 0.9,
      "grad_norm": 12.627870559692383,
      "learning_rate": 3.8939269029953284e-05,
      "loss": 0.1783,
      "step": 1210
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.900853157043457,
      "learning_rate": 3.880186864523221e-05,
      "loss": 0.2126,
      "step": 1220
    },
    {
      "epoch": 0.91,
      "grad_norm": 3.763246536254883,
      "learning_rate": 3.866446826051113e-05,
      "loss": 0.2136,
      "step": 1230
    },
    {
      "epoch": 0.92,
      "grad_norm": 7.94022274017334,
      "learning_rate": 3.852706787579005e-05,
      "loss": 0.1912,
      "step": 1240
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.354190826416016,
      "learning_rate": 3.8389667491068977e-05,
      "loss": 0.1925,
      "step": 1250
    },
    {
      "epoch": 0.93,
      "grad_norm": 12.671570777893066,
      "learning_rate": 3.82522671063479e-05,
      "loss": 0.1832,
      "step": 1260
    },
    {
      "epoch": 0.94,
      "grad_norm": 18.882457733154297,
      "learning_rate": 3.811486672162682e-05,
      "loss": 0.1868,
      "step": 1270
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.26636028289795,
      "learning_rate": 3.797746633690574e-05,
      "loss": 0.1815,
      "step": 1280
    },
    {
      "epoch": 0.96,
      "grad_norm": 5.035627841949463,
      "learning_rate": 3.784006595218467e-05,
      "loss": 0.1912,
      "step": 1290
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.453719139099121,
      "learning_rate": 3.770266556746359e-05,
      "loss": 0.1617,
      "step": 1300
    },
    {
      "epoch": 0.97,
      "grad_norm": 6.647568225860596,
      "learning_rate": 3.7565265182742515e-05,
      "loss": 0.1969,
      "step": 1310
    },
    {
      "epoch": 0.98,
      "grad_norm": 7.901888370513916,
      "learning_rate": 3.7427864798021435e-05,
      "loss": 0.1893,
      "step": 1320
    },
    {
      "epoch": 0.99,
      "grad_norm": 5.35443115234375,
      "learning_rate": 3.7290464413300355e-05,
      "loss": 0.1674,
      "step": 1330
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.808985710144043,
      "learning_rate": 3.715306402857928e-05,
      "loss": 0.1762,
      "step": 1340
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.947412963585564,
      "eval_loss": 0.1263597458600998,
      "eval_runtime": 141.4931,
      "eval_samples_per_second": 609.62,
      "eval_steps_per_second": 38.108,
      "step": 1348
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.0969462394714355,
      "learning_rate": 3.701566364385821e-05,
      "loss": 0.1727,
      "step": 1350
    },
    {
      "epoch": 1.01,
      "grad_norm": 11.357718467712402,
      "learning_rate": 3.687826325913713e-05,
      "loss": 0.1529,
      "step": 1360
    },
    {
      "epoch": 1.02,
      "grad_norm": 16.987056732177734,
      "learning_rate": 3.674086287441605e-05,
      "loss": 0.192,
      "step": 1370
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.532151222229004,
      "learning_rate": 3.6603462489694974e-05,
      "loss": 0.1439,
      "step": 1380
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.81009578704834,
      "learning_rate": 3.64660621049739e-05,
      "loss": 0.1541,
      "step": 1390
    },
    {
      "epoch": 1.04,
      "grad_norm": 6.317430019378662,
      "learning_rate": 3.632866172025282e-05,
      "loss": 0.1602,
      "step": 1400
    },
    {
      "epoch": 1.05,
      "grad_norm": 4.5247039794921875,
      "learning_rate": 3.619126133553174e-05,
      "loss": 0.1465,
      "step": 1410
    },
    {
      "epoch": 1.05,
      "grad_norm": 6.828434944152832,
      "learning_rate": 3.6053860950810666e-05,
      "loss": 0.2143,
      "step": 1420
    },
    {
      "epoch": 1.06,
      "grad_norm": 9.505870819091797,
      "learning_rate": 3.5916460566089586e-05,
      "loss": 0.2331,
      "step": 1430
    },
    {
      "epoch": 1.07,
      "grad_norm": 7.402731418609619,
      "learning_rate": 3.577906018136851e-05,
      "loss": 0.1997,
      "step": 1440
    },
    {
      "epoch": 1.08,
      "grad_norm": 13.928895950317383,
      "learning_rate": 3.564165979664743e-05,
      "loss": 0.1778,
      "step": 1450
    },
    {
      "epoch": 1.08,
      "grad_norm": 7.864881992340088,
      "learning_rate": 3.550425941192635e-05,
      "loss": 0.1937,
      "step": 1460
    },
    {
      "epoch": 1.09,
      "grad_norm": 8.725811004638672,
      "learning_rate": 3.536685902720528e-05,
      "loss": 0.166,
      "step": 1470
    },
    {
      "epoch": 1.1,
      "grad_norm": 19.298831939697266,
      "learning_rate": 3.5229458642484205e-05,
      "loss": 0.2126,
      "step": 1480
    },
    {
      "epoch": 1.11,
      "grad_norm": 6.058204650878906,
      "learning_rate": 3.5092058257763125e-05,
      "loss": 0.1524,
      "step": 1490
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.551276206970215,
      "learning_rate": 3.4954657873042044e-05,
      "loss": 0.1853,
      "step": 1500
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.204520225524902,
      "learning_rate": 3.481725748832097e-05,
      "loss": 0.1951,
      "step": 1510
    },
    {
      "epoch": 1.13,
      "grad_norm": 2.948188543319702,
      "learning_rate": 3.46798571035999e-05,
      "loss": 0.1559,
      "step": 1520
    },
    {
      "epoch": 1.14,
      "grad_norm": 12.01566219329834,
      "learning_rate": 3.454245671887881e-05,
      "loss": 0.1783,
      "step": 1530
    },
    {
      "epoch": 1.14,
      "grad_norm": 15.104211807250977,
      "learning_rate": 3.440505633415774e-05,
      "loss": 0.1621,
      "step": 1540
    },
    {
      "epoch": 1.15,
      "grad_norm": 5.36481237411499,
      "learning_rate": 3.426765594943666e-05,
      "loss": 0.172,
      "step": 1550
    },
    {
      "epoch": 1.16,
      "grad_norm": 5.6568827629089355,
      "learning_rate": 3.413025556471558e-05,
      "loss": 0.1645,
      "step": 1560
    },
    {
      "epoch": 1.16,
      "grad_norm": 11.129305839538574,
      "learning_rate": 3.39928551799945e-05,
      "loss": 0.1862,
      "step": 1570
    },
    {
      "epoch": 1.17,
      "grad_norm": 6.849048137664795,
      "learning_rate": 3.385545479527343e-05,
      "loss": 0.1507,
      "step": 1580
    },
    {
      "epoch": 1.18,
      "grad_norm": 8.361126899719238,
      "learning_rate": 3.371805441055235e-05,
      "loss": 0.1668,
      "step": 1590
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.8484487533569336,
      "learning_rate": 3.3580654025831275e-05,
      "loss": 0.1965,
      "step": 1600
    },
    {
      "epoch": 1.19,
      "grad_norm": 11.742258071899414,
      "learning_rate": 3.3443253641110195e-05,
      "loss": 0.1384,
      "step": 1610
    },
    {
      "epoch": 1.2,
      "grad_norm": 7.576124668121338,
      "learning_rate": 3.330585325638912e-05,
      "loss": 0.1316,
      "step": 1620
    },
    {
      "epoch": 1.21,
      "grad_norm": 15.978646278381348,
      "learning_rate": 3.316845287166804e-05,
      "loss": 0.1324,
      "step": 1630
    },
    {
      "epoch": 1.22,
      "grad_norm": 2.425368070602417,
      "learning_rate": 3.303105248694697e-05,
      "loss": 0.1429,
      "step": 1640
    },
    {
      "epoch": 1.22,
      "grad_norm": 8.935554504394531,
      "learning_rate": 3.289365210222589e-05,
      "loss": 0.1087,
      "step": 1650
    },
    {
      "epoch": 1.23,
      "grad_norm": 6.281069755554199,
      "learning_rate": 3.275625171750481e-05,
      "loss": 0.1681,
      "step": 1660
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.907856941223145,
      "learning_rate": 3.2618851332783734e-05,
      "loss": 0.1431,
      "step": 1670
    },
    {
      "epoch": 1.25,
      "grad_norm": 5.795341968536377,
      "learning_rate": 3.248145094806266e-05,
      "loss": 0.1575,
      "step": 1680
    },
    {
      "epoch": 1.25,
      "grad_norm": 12.340773582458496,
      "learning_rate": 3.234405056334158e-05,
      "loss": 0.1537,
      "step": 1690
    },
    {
      "epoch": 1.26,
      "grad_norm": 9.601103782653809,
      "learning_rate": 3.22066501786205e-05,
      "loss": 0.1618,
      "step": 1700
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.238292694091797,
      "learning_rate": 3.2069249793899426e-05,
      "loss": 0.15,
      "step": 1710
    },
    {
      "epoch": 1.28,
      "grad_norm": 6.925644874572754,
      "learning_rate": 3.193184940917835e-05,
      "loss": 0.1913,
      "step": 1720
    },
    {
      "epoch": 1.28,
      "grad_norm": 4.840887069702148,
      "learning_rate": 3.1794449024457266e-05,
      "loss": 0.1547,
      "step": 1730
    },
    {
      "epoch": 1.29,
      "grad_norm": 11.813157081604004,
      "learning_rate": 3.165704863973619e-05,
      "loss": 0.1488,
      "step": 1740
    },
    {
      "epoch": 1.3,
      "grad_norm": 16.899797439575195,
      "learning_rate": 3.151964825501512e-05,
      "loss": 0.1665,
      "step": 1750
    },
    {
      "epoch": 1.31,
      "grad_norm": 10.905778884887695,
      "learning_rate": 3.138224787029404e-05,
      "loss": 0.1484,
      "step": 1760
    },
    {
      "epoch": 1.31,
      "grad_norm": 14.453968048095703,
      "learning_rate": 3.124484748557296e-05,
      "loss": 0.1327,
      "step": 1770
    },
    {
      "epoch": 1.32,
      "grad_norm": 9.780988693237305,
      "learning_rate": 3.1107447100851885e-05,
      "loss": 0.103,
      "step": 1780
    },
    {
      "epoch": 1.33,
      "grad_norm": 4.881911277770996,
      "learning_rate": 3.0970046716130804e-05,
      "loss": 0.1604,
      "step": 1790
    },
    {
      "epoch": 1.34,
      "grad_norm": 4.467998027801514,
      "learning_rate": 3.083264633140973e-05,
      "loss": 0.1167,
      "step": 1800
    },
    {
      "epoch": 1.34,
      "grad_norm": 11.12286376953125,
      "learning_rate": 3.069524594668865e-05,
      "loss": 0.139,
      "step": 1810
    },
    {
      "epoch": 1.35,
      "grad_norm": 8.770483016967773,
      "learning_rate": 3.055784556196758e-05,
      "loss": 0.13,
      "step": 1820
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.070440292358398,
      "learning_rate": 3.0420445177246497e-05,
      "loss": 0.1492,
      "step": 1830
    },
    {
      "epoch": 1.36,
      "grad_norm": 8.814356803894043,
      "learning_rate": 3.0283044792525423e-05,
      "loss": 0.1872,
      "step": 1840
    },
    {
      "epoch": 1.37,
      "grad_norm": 7.875304222106934,
      "learning_rate": 3.014564440780434e-05,
      "loss": 0.1561,
      "step": 1850
    },
    {
      "epoch": 1.38,
      "grad_norm": 6.7557692527771,
      "learning_rate": 3.0008244023083266e-05,
      "loss": 0.1345,
      "step": 1860
    },
    {
      "epoch": 1.39,
      "grad_norm": 11.669951438903809,
      "learning_rate": 2.987084363836219e-05,
      "loss": 0.1003,
      "step": 1870
    },
    {
      "epoch": 1.39,
      "grad_norm": 9.418144226074219,
      "learning_rate": 2.9733443253641112e-05,
      "loss": 0.1245,
      "step": 1880
    },
    {
      "epoch": 1.4,
      "grad_norm": 5.591629505157471,
      "learning_rate": 2.9596042868920032e-05,
      "loss": 0.1536,
      "step": 1890
    },
    {
      "epoch": 1.41,
      "grad_norm": 10.448158264160156,
      "learning_rate": 2.9458642484198955e-05,
      "loss": 0.1217,
      "step": 1900
    },
    {
      "epoch": 1.42,
      "grad_norm": 8.35586166381836,
      "learning_rate": 2.932124209947788e-05,
      "loss": 0.1517,
      "step": 1910
    },
    {
      "epoch": 1.42,
      "grad_norm": 10.111129760742188,
      "learning_rate": 2.9183841714756805e-05,
      "loss": 0.1271,
      "step": 1920
    },
    {
      "epoch": 1.43,
      "grad_norm": 3.2071399688720703,
      "learning_rate": 2.9046441330035725e-05,
      "loss": 0.118,
      "step": 1930
    },
    {
      "epoch": 1.44,
      "grad_norm": 22.524978637695312,
      "learning_rate": 2.8909040945314648e-05,
      "loss": 0.1511,
      "step": 1940
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.672737121582031,
      "learning_rate": 2.877164056059357e-05,
      "loss": 0.1224,
      "step": 1950
    },
    {
      "epoch": 1.45,
      "grad_norm": 10.548683166503906,
      "learning_rate": 2.8634240175872494e-05,
      "loss": 0.1307,
      "step": 1960
    },
    {
      "epoch": 1.46,
      "grad_norm": 8.707009315490723,
      "learning_rate": 2.8496839791151414e-05,
      "loss": 0.1184,
      "step": 1970
    },
    {
      "epoch": 1.47,
      "grad_norm": 4.6339874267578125,
      "learning_rate": 2.8359439406430337e-05,
      "loss": 0.143,
      "step": 1980
    },
    {
      "epoch": 1.48,
      "grad_norm": 11.373931884765625,
      "learning_rate": 2.8222039021709263e-05,
      "loss": 0.1468,
      "step": 1990
    },
    {
      "epoch": 1.48,
      "grad_norm": 5.647881031036377,
      "learning_rate": 2.8084638636988186e-05,
      "loss": 0.1188,
      "step": 2000
    },
    {
      "epoch": 1.49,
      "grad_norm": 9.399796485900879,
      "learning_rate": 2.7947238252267106e-05,
      "loss": 0.1484,
      "step": 2010
    },
    {
      "epoch": 1.5,
      "grad_norm": 11.064074516296387,
      "learning_rate": 2.780983786754603e-05,
      "loss": 0.1471,
      "step": 2020
    },
    {
      "epoch": 1.51,
      "grad_norm": 5.499329090118408,
      "learning_rate": 2.7672437482824952e-05,
      "loss": 0.136,
      "step": 2030
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.6922025680542,
      "learning_rate": 2.753503709810388e-05,
      "loss": 0.1292,
      "step": 2040
    },
    {
      "epoch": 1.52,
      "grad_norm": 6.548159599304199,
      "learning_rate": 2.7397636713382795e-05,
      "loss": 0.0878,
      "step": 2050
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.498106002807617,
      "learning_rate": 2.726023632866172e-05,
      "loss": 0.1235,
      "step": 2060
    },
    {
      "epoch": 1.54,
      "grad_norm": 14.733744621276855,
      "learning_rate": 2.7122835943940645e-05,
      "loss": 0.1243,
      "step": 2070
    },
    {
      "epoch": 1.54,
      "grad_norm": 4.867096424102783,
      "learning_rate": 2.6985435559219568e-05,
      "loss": 0.1553,
      "step": 2080
    },
    {
      "epoch": 1.55,
      "grad_norm": 18.455617904663086,
      "learning_rate": 2.684803517449849e-05,
      "loss": 0.161,
      "step": 2090
    },
    {
      "epoch": 1.56,
      "grad_norm": 7.310939311981201,
      "learning_rate": 2.671063478977741e-05,
      "loss": 0.1245,
      "step": 2100
    },
    {
      "epoch": 1.57,
      "grad_norm": 6.81366491317749,
      "learning_rate": 2.6573234405056334e-05,
      "loss": 0.1673,
      "step": 2110
    },
    {
      "epoch": 1.57,
      "grad_norm": 10.377093315124512,
      "learning_rate": 2.643583402033526e-05,
      "loss": 0.1255,
      "step": 2120
    },
    {
      "epoch": 1.58,
      "grad_norm": 8.426204681396484,
      "learning_rate": 2.6298433635614183e-05,
      "loss": 0.158,
      "step": 2130
    },
    {
      "epoch": 1.59,
      "grad_norm": 4.7572760581970215,
      "learning_rate": 2.6161033250893103e-05,
      "loss": 0.1573,
      "step": 2140
    },
    {
      "epoch": 1.59,
      "grad_norm": 12.727173805236816,
      "learning_rate": 2.6023632866172026e-05,
      "loss": 0.1426,
      "step": 2150
    },
    {
      "epoch": 1.6,
      "grad_norm": 8.446662902832031,
      "learning_rate": 2.588623248145095e-05,
      "loss": 0.1689,
      "step": 2160
    },
    {
      "epoch": 1.61,
      "grad_norm": 5.482067584991455,
      "learning_rate": 2.5748832096729876e-05,
      "loss": 0.1257,
      "step": 2170
    },
    {
      "epoch": 1.62,
      "grad_norm": 12.713666915893555,
      "learning_rate": 2.5611431712008792e-05,
      "loss": 0.1682,
      "step": 2180
    },
    {
      "epoch": 1.62,
      "grad_norm": 6.469407081604004,
      "learning_rate": 2.547403132728772e-05,
      "loss": 0.1454,
      "step": 2190
    },
    {
      "epoch": 1.63,
      "grad_norm": 14.35875129699707,
      "learning_rate": 2.5336630942566642e-05,
      "loss": 0.1278,
      "step": 2200
    },
    {
      "epoch": 1.64,
      "grad_norm": 9.981734275817871,
      "learning_rate": 2.5199230557845565e-05,
      "loss": 0.1292,
      "step": 2210
    },
    {
      "epoch": 1.65,
      "grad_norm": 5.275669574737549,
      "learning_rate": 2.5061830173124485e-05,
      "loss": 0.1384,
      "step": 2220
    },
    {
      "epoch": 1.65,
      "grad_norm": 8.421156883239746,
      "learning_rate": 2.4924429788403408e-05,
      "loss": 0.0778,
      "step": 2230
    },
    {
      "epoch": 1.66,
      "grad_norm": 6.505790710449219,
      "learning_rate": 2.478702940368233e-05,
      "loss": 0.1316,
      "step": 2240
    },
    {
      "epoch": 1.67,
      "grad_norm": 13.836162567138672,
      "learning_rate": 2.4649629018961254e-05,
      "loss": 0.1066,
      "step": 2250
    },
    {
      "epoch": 1.68,
      "grad_norm": 13.854897499084473,
      "learning_rate": 2.4512228634240177e-05,
      "loss": 0.1359,
      "step": 2260
    },
    {
      "epoch": 1.68,
      "grad_norm": 8.590911865234375,
      "learning_rate": 2.43748282495191e-05,
      "loss": 0.1295,
      "step": 2270
    },
    {
      "epoch": 1.69,
      "grad_norm": 3.8553526401519775,
      "learning_rate": 2.4237427864798023e-05,
      "loss": 0.1308,
      "step": 2280
    },
    {
      "epoch": 1.7,
      "grad_norm": 5.948857307434082,
      "learning_rate": 2.4100027480076946e-05,
      "loss": 0.1335,
      "step": 2290
    },
    {
      "epoch": 1.71,
      "grad_norm": 10.528166770935059,
      "learning_rate": 2.396262709535587e-05,
      "loss": 0.1331,
      "step": 2300
    },
    {
      "epoch": 1.71,
      "grad_norm": 3.8531999588012695,
      "learning_rate": 2.382522671063479e-05,
      "loss": 0.1208,
      "step": 2310
    },
    {
      "epoch": 1.72,
      "grad_norm": 15.348362922668457,
      "learning_rate": 2.3687826325913716e-05,
      "loss": 0.1525,
      "step": 2320
    },
    {
      "epoch": 1.73,
      "grad_norm": 9.51009464263916,
      "learning_rate": 2.3550425941192635e-05,
      "loss": 0.118,
      "step": 2330
    },
    {
      "epoch": 1.74,
      "grad_norm": 13.064322471618652,
      "learning_rate": 2.341302555647156e-05,
      "loss": 0.1434,
      "step": 2340
    },
    {
      "epoch": 1.74,
      "grad_norm": 5.5084614753723145,
      "learning_rate": 2.327562517175048e-05,
      "loss": 0.1387,
      "step": 2350
    },
    {
      "epoch": 1.75,
      "grad_norm": 10.431999206542969,
      "learning_rate": 2.3138224787029405e-05,
      "loss": 0.1147,
      "step": 2360
    },
    {
      "epoch": 1.76,
      "grad_norm": 14.113656044006348,
      "learning_rate": 2.3000824402308328e-05,
      "loss": 0.1442,
      "step": 2370
    },
    {
      "epoch": 1.77,
      "grad_norm": 8.030630111694336,
      "learning_rate": 2.286342401758725e-05,
      "loss": 0.1294,
      "step": 2380
    },
    {
      "epoch": 1.77,
      "grad_norm": 5.992265224456787,
      "learning_rate": 2.272602363286617e-05,
      "loss": 0.0894,
      "step": 2390
    },
    {
      "epoch": 1.78,
      "grad_norm": 10.719809532165527,
      "learning_rate": 2.2588623248145097e-05,
      "loss": 0.107,
      "step": 2400
    },
    {
      "epoch": 1.79,
      "grad_norm": 13.229023933410645,
      "learning_rate": 2.2451222863424017e-05,
      "loss": 0.1236,
      "step": 2410
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.23493766784668,
      "learning_rate": 2.2313822478702943e-05,
      "loss": 0.1576,
      "step": 2420
    },
    {
      "epoch": 1.8,
      "grad_norm": 12.941469192504883,
      "learning_rate": 2.2176422093981863e-05,
      "loss": 0.1393,
      "step": 2430
    },
    {
      "epoch": 1.81,
      "grad_norm": 7.724160194396973,
      "learning_rate": 2.2039021709260786e-05,
      "loss": 0.1096,
      "step": 2440
    },
    {
      "epoch": 1.82,
      "grad_norm": 5.428783416748047,
      "learning_rate": 2.190162132453971e-05,
      "loss": 0.1035,
      "step": 2450
    },
    {
      "epoch": 1.82,
      "grad_norm": 14.180803298950195,
      "learning_rate": 2.1764220939818633e-05,
      "loss": 0.1444,
      "step": 2460
    },
    {
      "epoch": 1.83,
      "grad_norm": 7.265615940093994,
      "learning_rate": 2.1626820555097556e-05,
      "loss": 0.1496,
      "step": 2470
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.01385498046875,
      "learning_rate": 2.148942017037648e-05,
      "loss": 0.1124,
      "step": 2480
    },
    {
      "epoch": 1.85,
      "grad_norm": 7.296875476837158,
      "learning_rate": 2.13520197856554e-05,
      "loss": 0.1387,
      "step": 2490
    },
    {
      "epoch": 1.85,
      "grad_norm": 12.550304412841797,
      "learning_rate": 2.1214619400934325e-05,
      "loss": 0.105,
      "step": 2500
    },
    {
      "epoch": 1.86,
      "grad_norm": 5.9940385818481445,
      "learning_rate": 2.1077219016213245e-05,
      "loss": 0.1368,
      "step": 2510
    },
    {
      "epoch": 1.87,
      "grad_norm": 5.355954170227051,
      "learning_rate": 2.093981863149217e-05,
      "loss": 0.086,
      "step": 2520
    },
    {
      "epoch": 1.88,
      "grad_norm": 8.216740608215332,
      "learning_rate": 2.080241824677109e-05,
      "loss": 0.0973,
      "step": 2530
    },
    {
      "epoch": 1.88,
      "grad_norm": 9.250962257385254,
      "learning_rate": 2.0665017862050014e-05,
      "loss": 0.103,
      "step": 2540
    },
    {
      "epoch": 1.89,
      "grad_norm": 12.288225173950195,
      "learning_rate": 2.0527617477328937e-05,
      "loss": 0.1405,
      "step": 2550
    },
    {
      "epoch": 1.9,
      "grad_norm": 8.24113941192627,
      "learning_rate": 2.039021709260786e-05,
      "loss": 0.1371,
      "step": 2560
    },
    {
      "epoch": 1.91,
      "grad_norm": 24.87588882446289,
      "learning_rate": 2.0252816707886783e-05,
      "loss": 0.1137,
      "step": 2570
    },
    {
      "epoch": 1.91,
      "grad_norm": 13.435257911682129,
      "learning_rate": 2.0115416323165706e-05,
      "loss": 0.1034,
      "step": 2580
    },
    {
      "epoch": 1.92,
      "grad_norm": 18.450393676757812,
      "learning_rate": 1.9978015938444626e-05,
      "loss": 0.1475,
      "step": 2590
    },
    {
      "epoch": 1.93,
      "grad_norm": 8.536957740783691,
      "learning_rate": 1.9840615553723553e-05,
      "loss": 0.0997,
      "step": 2600
    },
    {
      "epoch": 1.94,
      "grad_norm": 9.833585739135742,
      "learning_rate": 1.9703215169002472e-05,
      "loss": 0.1198,
      "step": 2610
    },
    {
      "epoch": 1.94,
      "grad_norm": 6.295921802520752,
      "learning_rate": 1.9565814784281396e-05,
      "loss": 0.1138,
      "step": 2620
    },
    {
      "epoch": 1.95,
      "grad_norm": 9.540913581848145,
      "learning_rate": 1.942841439956032e-05,
      "loss": 0.0946,
      "step": 2630
    },
    {
      "epoch": 1.96,
      "grad_norm": 5.700273036956787,
      "learning_rate": 1.9291014014839242e-05,
      "loss": 0.1178,
      "step": 2640
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.3849515914917,
      "learning_rate": 1.9153613630118165e-05,
      "loss": 0.1156,
      "step": 2650
    },
    {
      "epoch": 1.97,
      "grad_norm": 15.1934232711792,
      "learning_rate": 1.9016213245397088e-05,
      "loss": 0.1263,
      "step": 2660
    },
    {
      "epoch": 1.98,
      "grad_norm": 8.725594520568848,
      "learning_rate": 1.887881286067601e-05,
      "loss": 0.1151,
      "step": 2670
    },
    {
      "epoch": 1.99,
      "grad_norm": 9.016326904296875,
      "learning_rate": 1.8741412475954934e-05,
      "loss": 0.1424,
      "step": 2680
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.650193214416504,
      "learning_rate": 1.8604012091233854e-05,
      "loss": 0.1407,
      "step": 2690
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.972523969069177,
      "eval_loss": 0.07045618444681168,
      "eval_runtime": 140.2638,
      "eval_samples_per_second": 614.963,
      "eval_steps_per_second": 38.442,
      "step": 2696
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.413318634033203,
      "learning_rate": 1.846661170651278e-05,
      "loss": 0.1072,
      "step": 2700
    },
    {
      "epoch": 2.01,
      "grad_norm": 6.745368003845215,
      "learning_rate": 1.83292113217917e-05,
      "loss": 0.1047,
      "step": 2710
    },
    {
      "epoch": 2.02,
      "grad_norm": 7.095730304718018,
      "learning_rate": 1.8191810937070623e-05,
      "loss": 0.1024,
      "step": 2720
    },
    {
      "epoch": 2.03,
      "grad_norm": 13.053765296936035,
      "learning_rate": 1.805441055234955e-05,
      "loss": 0.1113,
      "step": 2730
    },
    {
      "epoch": 2.03,
      "grad_norm": 10.390790939331055,
      "learning_rate": 1.791701016762847e-05,
      "loss": 0.1106,
      "step": 2740
    },
    {
      "epoch": 2.04,
      "grad_norm": 8.856596946716309,
      "learning_rate": 1.7779609782907393e-05,
      "loss": 0.0861,
      "step": 2750
    },
    {
      "epoch": 2.05,
      "grad_norm": 4.328437805175781,
      "learning_rate": 1.7642209398186316e-05,
      "loss": 0.0798,
      "step": 2760
    },
    {
      "epoch": 2.05,
      "grad_norm": 4.761692523956299,
      "learning_rate": 1.750480901346524e-05,
      "loss": 0.0879,
      "step": 2770
    },
    {
      "epoch": 2.06,
      "grad_norm": 12.073725700378418,
      "learning_rate": 1.7367408628744162e-05,
      "loss": 0.0962,
      "step": 2780
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.6801289319992065,
      "learning_rate": 1.7230008244023085e-05,
      "loss": 0.0877,
      "step": 2790
    },
    {
      "epoch": 2.08,
      "grad_norm": 12.418718338012695,
      "learning_rate": 1.7092607859302008e-05,
      "loss": 0.0984,
      "step": 2800
    },
    {
      "epoch": 2.08,
      "grad_norm": 3.915785312652588,
      "learning_rate": 1.695520747458093e-05,
      "loss": 0.108,
      "step": 2810
    },
    {
      "epoch": 2.09,
      "grad_norm": 6.9176435470581055,
      "learning_rate": 1.681780708985985e-05,
      "loss": 0.1026,
      "step": 2820
    },
    {
      "epoch": 2.1,
      "grad_norm": 3.9952585697174072,
      "learning_rate": 1.6680406705138778e-05,
      "loss": 0.0622,
      "step": 2830
    },
    {
      "epoch": 2.11,
      "grad_norm": 11.399799346923828,
      "learning_rate": 1.6543006320417697e-05,
      "loss": 0.1219,
      "step": 2840
    },
    {
      "epoch": 2.11,
      "grad_norm": 9.78951358795166,
      "learning_rate": 1.640560593569662e-05,
      "loss": 0.1127,
      "step": 2850
    },
    {
      "epoch": 2.12,
      "grad_norm": 5.477469444274902,
      "learning_rate": 1.6268205550975543e-05,
      "loss": 0.0797,
      "step": 2860
    },
    {
      "epoch": 2.13,
      "grad_norm": 8.457229614257812,
      "learning_rate": 1.6130805166254467e-05,
      "loss": 0.1041,
      "step": 2870
    },
    {
      "epoch": 2.14,
      "grad_norm": 11.861204147338867,
      "learning_rate": 1.599340478153339e-05,
      "loss": 0.1057,
      "step": 2880
    },
    {
      "epoch": 2.14,
      "grad_norm": 10.859254837036133,
      "learning_rate": 1.5856004396812313e-05,
      "loss": 0.0783,
      "step": 2890
    },
    {
      "epoch": 2.15,
      "grad_norm": 7.593705177307129,
      "learning_rate": 1.5718604012091236e-05,
      "loss": 0.101,
      "step": 2900
    },
    {
      "epoch": 2.16,
      "grad_norm": 11.803435325622559,
      "learning_rate": 1.558120362737016e-05,
      "loss": 0.0835,
      "step": 2910
    },
    {
      "epoch": 2.17,
      "grad_norm": 10.460546493530273,
      "learning_rate": 1.544380324264908e-05,
      "loss": 0.0951,
      "step": 2920
    },
    {
      "epoch": 2.17,
      "grad_norm": 14.229741096496582,
      "learning_rate": 1.5306402857928005e-05,
      "loss": 0.1109,
      "step": 2930
    },
    {
      "epoch": 2.18,
      "grad_norm": 4.7770233154296875,
      "learning_rate": 1.5169002473206925e-05,
      "loss": 0.1304,
      "step": 2940
    },
    {
      "epoch": 2.19,
      "grad_norm": 9.118112564086914,
      "learning_rate": 1.503160208848585e-05,
      "loss": 0.0583,
      "step": 2950
    },
    {
      "epoch": 2.2,
      "grad_norm": 5.745917797088623,
      "learning_rate": 1.4894201703764771e-05,
      "loss": 0.0851,
      "step": 2960
    },
    {
      "epoch": 2.2,
      "grad_norm": 8.237546920776367,
      "learning_rate": 1.4756801319043694e-05,
      "loss": 0.106,
      "step": 2970
    },
    {
      "epoch": 2.21,
      "grad_norm": 12.706246376037598,
      "learning_rate": 1.4619400934322616e-05,
      "loss": 0.1187,
      "step": 2980
    },
    {
      "epoch": 2.22,
      "grad_norm": 6.276710033416748,
      "learning_rate": 1.448200054960154e-05,
      "loss": 0.1031,
      "step": 2990
    },
    {
      "epoch": 2.23,
      "grad_norm": 4.847949981689453,
      "learning_rate": 1.4344600164880462e-05,
      "loss": 0.0694,
      "step": 3000
    },
    {
      "epoch": 2.23,
      "grad_norm": 5.567623615264893,
      "learning_rate": 1.4207199780159385e-05,
      "loss": 0.0809,
      "step": 3010
    },
    {
      "epoch": 2.24,
      "grad_norm": 5.528212070465088,
      "learning_rate": 1.4069799395438306e-05,
      "loss": 0.1021,
      "step": 3020
    },
    {
      "epoch": 2.25,
      "grad_norm": 10.893386840820312,
      "learning_rate": 1.3932399010717231e-05,
      "loss": 0.0918,
      "step": 3030
    },
    {
      "epoch": 2.26,
      "grad_norm": 13.184591293334961,
      "learning_rate": 1.3794998625996153e-05,
      "loss": 0.1005,
      "step": 3040
    },
    {
      "epoch": 2.26,
      "grad_norm": 7.119046688079834,
      "learning_rate": 1.3657598241275078e-05,
      "loss": 0.0996,
      "step": 3050
    },
    {
      "epoch": 2.27,
      "grad_norm": 12.941669464111328,
      "learning_rate": 1.3520197856553999e-05,
      "loss": 0.084,
      "step": 3060
    },
    {
      "epoch": 2.28,
      "grad_norm": 8.232301712036133,
      "learning_rate": 1.3382797471832922e-05,
      "loss": 0.0887,
      "step": 3070
    },
    {
      "epoch": 2.28,
      "grad_norm": 9.092889785766602,
      "learning_rate": 1.3245397087111843e-05,
      "loss": 0.0794,
      "step": 3080
    },
    {
      "epoch": 2.29,
      "grad_norm": 7.75855827331543,
      "learning_rate": 1.3107996702390768e-05,
      "loss": 0.0795,
      "step": 3090
    },
    {
      "epoch": 2.3,
      "grad_norm": 14.936914443969727,
      "learning_rate": 1.297059631766969e-05,
      "loss": 0.0687,
      "step": 3100
    },
    {
      "epoch": 2.31,
      "grad_norm": 9.405156135559082,
      "learning_rate": 1.2833195932948613e-05,
      "loss": 0.113,
      "step": 3110
    },
    {
      "epoch": 2.31,
      "grad_norm": 10.729270935058594,
      "learning_rate": 1.2695795548227534e-05,
      "loss": 0.1129,
      "step": 3120
    },
    {
      "epoch": 2.32,
      "grad_norm": 11.434041976928711,
      "learning_rate": 1.2558395163506459e-05,
      "loss": 0.0989,
      "step": 3130
    },
    {
      "epoch": 2.33,
      "grad_norm": 8.561604499816895,
      "learning_rate": 1.2420994778785382e-05,
      "loss": 0.0948,
      "step": 3140
    },
    {
      "epoch": 2.34,
      "grad_norm": 6.7873406410217285,
      "learning_rate": 1.2283594394064305e-05,
      "loss": 0.0843,
      "step": 3150
    },
    {
      "epoch": 2.34,
      "grad_norm": 13.5293607711792,
      "learning_rate": 1.2146194009343227e-05,
      "loss": 0.1164,
      "step": 3160
    },
    {
      "epoch": 2.35,
      "grad_norm": 9.595284461975098,
      "learning_rate": 1.200879362462215e-05,
      "loss": 0.0833,
      "step": 3170
    },
    {
      "epoch": 2.36,
      "grad_norm": 7.143069267272949,
      "learning_rate": 1.1871393239901073e-05,
      "loss": 0.1036,
      "step": 3180
    },
    {
      "epoch": 2.37,
      "grad_norm": 12.40601921081543,
      "learning_rate": 1.1733992855179996e-05,
      "loss": 0.0641,
      "step": 3190
    },
    {
      "epoch": 2.37,
      "grad_norm": 5.299369812011719,
      "learning_rate": 1.1596592470458917e-05,
      "loss": 0.0835,
      "step": 3200
    },
    {
      "epoch": 2.38,
      "grad_norm": 3.2412006855010986,
      "learning_rate": 1.145919208573784e-05,
      "loss": 0.0985,
      "step": 3210
    },
    {
      "epoch": 2.39,
      "grad_norm": 9.663202285766602,
      "learning_rate": 1.1321791701016764e-05,
      "loss": 0.0725,
      "step": 3220
    },
    {
      "epoch": 2.4,
      "grad_norm": 9.806108474731445,
      "learning_rate": 1.1184391316295687e-05,
      "loss": 0.0976,
      "step": 3230
    },
    {
      "epoch": 2.4,
      "grad_norm": 12.536150932312012,
      "learning_rate": 1.104699093157461e-05,
      "loss": 0.1035,
      "step": 3240
    },
    {
      "epoch": 2.41,
      "grad_norm": 19.915842056274414,
      "learning_rate": 1.0909590546853531e-05,
      "loss": 0.0878,
      "step": 3250
    },
    {
      "epoch": 2.42,
      "grad_norm": 3.88319993019104,
      "learning_rate": 1.0772190162132454e-05,
      "loss": 0.0788,
      "step": 3260
    },
    {
      "epoch": 2.43,
      "grad_norm": 6.242338180541992,
      "learning_rate": 1.0634789777411378e-05,
      "loss": 0.096,
      "step": 3270
    },
    {
      "epoch": 2.43,
      "grad_norm": 10.742528915405273,
      "learning_rate": 1.04973893926903e-05,
      "loss": 0.0954,
      "step": 3280
    },
    {
      "epoch": 2.44,
      "grad_norm": 4.696013927459717,
      "learning_rate": 1.0359989007969224e-05,
      "loss": 0.0945,
      "step": 3290
    },
    {
      "epoch": 2.45,
      "grad_norm": 4.943475723266602,
      "learning_rate": 1.0222588623248145e-05,
      "loss": 0.1001,
      "step": 3300
    },
    {
      "epoch": 2.46,
      "grad_norm": 10.133197784423828,
      "learning_rate": 1.0085188238527068e-05,
      "loss": 0.0921,
      "step": 3310
    },
    {
      "epoch": 2.46,
      "grad_norm": 11.798482894897461,
      "learning_rate": 9.947787853805991e-06,
      "loss": 0.0675,
      "step": 3320
    },
    {
      "epoch": 2.47,
      "grad_norm": 5.647496700286865,
      "learning_rate": 9.810387469084915e-06,
      "loss": 0.0721,
      "step": 3330
    },
    {
      "epoch": 2.48,
      "grad_norm": 3.0277674198150635,
      "learning_rate": 9.672987084363836e-06,
      "loss": 0.0798,
      "step": 3340
    },
    {
      "epoch": 2.49,
      "grad_norm": 5.510377407073975,
      "learning_rate": 9.535586699642759e-06,
      "loss": 0.1071,
      "step": 3350
    },
    {
      "epoch": 2.49,
      "grad_norm": 7.622246265411377,
      "learning_rate": 9.398186314921682e-06,
      "loss": 0.0813,
      "step": 3360
    },
    {
      "epoch": 2.5,
      "grad_norm": 12.018797874450684,
      "learning_rate": 9.260785930200605e-06,
      "loss": 0.0849,
      "step": 3370
    },
    {
      "epoch": 2.51,
      "grad_norm": 5.821876525878906,
      "learning_rate": 9.123385545479528e-06,
      "loss": 0.0701,
      "step": 3380
    },
    {
      "epoch": 2.51,
      "grad_norm": 12.286214828491211,
      "learning_rate": 8.98598516075845e-06,
      "loss": 0.0825,
      "step": 3390
    },
    {
      "epoch": 2.52,
      "grad_norm": 5.903075218200684,
      "learning_rate": 8.848584776037373e-06,
      "loss": 0.0981,
      "step": 3400
    },
    {
      "epoch": 2.53,
      "grad_norm": 5.024692058563232,
      "learning_rate": 8.711184391316296e-06,
      "loss": 0.0905,
      "step": 3410
    },
    {
      "epoch": 2.54,
      "grad_norm": 14.582296371459961,
      "learning_rate": 8.573784006595219e-06,
      "loss": 0.1047,
      "step": 3420
    },
    {
      "epoch": 2.54,
      "grad_norm": 4.112821578979492,
      "learning_rate": 8.436383621874142e-06,
      "loss": 0.0831,
      "step": 3430
    },
    {
      "epoch": 2.55,
      "grad_norm": 8.43770980834961,
      "learning_rate": 8.298983237153064e-06,
      "loss": 0.0961,
      "step": 3440
    },
    {
      "epoch": 2.56,
      "grad_norm": 10.004687309265137,
      "learning_rate": 8.161582852431987e-06,
      "loss": 0.08,
      "step": 3450
    },
    {
      "epoch": 2.57,
      "grad_norm": 4.888688564300537,
      "learning_rate": 8.02418246771091e-06,
      "loss": 0.0741,
      "step": 3460
    },
    {
      "epoch": 2.57,
      "grad_norm": 6.9248881340026855,
      "learning_rate": 7.886782082989833e-06,
      "loss": 0.0918,
      "step": 3470
    },
    {
      "epoch": 2.58,
      "grad_norm": 5.887302398681641,
      "learning_rate": 7.749381698268756e-06,
      "loss": 0.0693,
      "step": 3480
    },
    {
      "epoch": 2.59,
      "grad_norm": 8.469816207885742,
      "learning_rate": 7.611981313547678e-06,
      "loss": 0.0911,
      "step": 3490
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.2210373878479004,
      "learning_rate": 7.474580928826601e-06,
      "loss": 0.0667,
      "step": 3500
    },
    {
      "epoch": 2.6,
      "grad_norm": 10.158321380615234,
      "learning_rate": 7.337180544105524e-06,
      "loss": 0.0658,
      "step": 3510
    },
    {
      "epoch": 2.61,
      "grad_norm": 8.220787048339844,
      "learning_rate": 7.199780159384446e-06,
      "loss": 0.088,
      "step": 3520
    },
    {
      "epoch": 2.62,
      "grad_norm": 10.971105575561523,
      "learning_rate": 7.062379774663369e-06,
      "loss": 0.1002,
      "step": 3530
    },
    {
      "epoch": 2.63,
      "grad_norm": 8.404732704162598,
      "learning_rate": 6.924979389942292e-06,
      "loss": 0.0869,
      "step": 3540
    },
    {
      "epoch": 2.63,
      "grad_norm": 8.824265480041504,
      "learning_rate": 6.7875790052212145e-06,
      "loss": 0.0705,
      "step": 3550
    },
    {
      "epoch": 2.64,
      "grad_norm": 3.96179461479187,
      "learning_rate": 6.650178620500138e-06,
      "loss": 0.0603,
      "step": 3560
    },
    {
      "epoch": 2.65,
      "grad_norm": 8.119577407836914,
      "learning_rate": 6.51277823577906e-06,
      "loss": 0.0719,
      "step": 3570
    },
    {
      "epoch": 2.66,
      "grad_norm": 7.250158786773682,
      "learning_rate": 6.375377851057983e-06,
      "loss": 0.0906,
      "step": 3580
    },
    {
      "epoch": 2.66,
      "grad_norm": 4.2896623611450195,
      "learning_rate": 6.237977466336906e-06,
      "loss": 0.0977,
      "step": 3590
    },
    {
      "epoch": 2.67,
      "grad_norm": 2.2607102394104004,
      "learning_rate": 6.100577081615829e-06,
      "loss": 0.0627,
      "step": 3600
    },
    {
      "epoch": 2.68,
      "grad_norm": 4.072081565856934,
      "learning_rate": 5.9631766968947515e-06,
      "loss": 0.0909,
      "step": 3610
    },
    {
      "epoch": 2.69,
      "grad_norm": 7.080986022949219,
      "learning_rate": 5.825776312173675e-06,
      "loss": 0.0982,
      "step": 3620
    },
    {
      "epoch": 2.69,
      "grad_norm": 9.746908187866211,
      "learning_rate": 5.688375927452598e-06,
      "loss": 0.0614,
      "step": 3630
    },
    {
      "epoch": 2.7,
      "grad_norm": 8.818674087524414,
      "learning_rate": 5.55097554273152e-06,
      "loss": 0.124,
      "step": 3640
    },
    {
      "epoch": 2.71,
      "grad_norm": 13.782171249389648,
      "learning_rate": 5.413575158010443e-06,
      "loss": 0.0926,
      "step": 3650
    },
    {
      "epoch": 2.72,
      "grad_norm": 8.21611499786377,
      "learning_rate": 5.276174773289365e-06,
      "loss": 0.0693,
      "step": 3660
    },
    {
      "epoch": 2.72,
      "grad_norm": 10.9508695602417,
      "learning_rate": 5.1387743885682885e-06,
      "loss": 0.0974,
      "step": 3670
    },
    {
      "epoch": 2.73,
      "grad_norm": 14.169780731201172,
      "learning_rate": 5.001374003847211e-06,
      "loss": 0.1044,
      "step": 3680
    },
    {
      "epoch": 2.74,
      "grad_norm": 3.944868564605713,
      "learning_rate": 4.863973619126134e-06,
      "loss": 0.0705,
      "step": 3690
    },
    {
      "epoch": 2.74,
      "grad_norm": 6.739749908447266,
      "learning_rate": 4.726573234405057e-06,
      "loss": 0.0679,
      "step": 3700
    },
    {
      "epoch": 2.75,
      "grad_norm": 13.893156051635742,
      "learning_rate": 4.589172849683979e-06,
      "loss": 0.0596,
      "step": 3710
    },
    {
      "epoch": 2.76,
      "grad_norm": 8.715542793273926,
      "learning_rate": 4.451772464962902e-06,
      "loss": 0.0716,
      "step": 3720
    },
    {
      "epoch": 2.77,
      "grad_norm": 11.366250038146973,
      "learning_rate": 4.314372080241825e-06,
      "loss": 0.0949,
      "step": 3730
    },
    {
      "epoch": 2.77,
      "grad_norm": 5.514875411987305,
      "learning_rate": 4.176971695520748e-06,
      "loss": 0.0752,
      "step": 3740
    },
    {
      "epoch": 2.78,
      "grad_norm": 14.893842697143555,
      "learning_rate": 4.03957131079967e-06,
      "loss": 0.108,
      "step": 3750
    },
    {
      "epoch": 2.79,
      "grad_norm": 12.94680118560791,
      "learning_rate": 3.902170926078593e-06,
      "loss": 0.1044,
      "step": 3760
    },
    {
      "epoch": 2.8,
      "grad_norm": 8.319701194763184,
      "learning_rate": 3.7647705413575158e-06,
      "loss": 0.0713,
      "step": 3770
    },
    {
      "epoch": 2.8,
      "grad_norm": 17.827917098999023,
      "learning_rate": 3.6273701566364385e-06,
      "loss": 0.0807,
      "step": 3780
    },
    {
      "epoch": 2.81,
      "grad_norm": 6.6090312004089355,
      "learning_rate": 3.489969771915361e-06,
      "loss": 0.0821,
      "step": 3790
    },
    {
      "epoch": 2.82,
      "grad_norm": 3.432737112045288,
      "learning_rate": 3.3525693871942843e-06,
      "loss": 0.0702,
      "step": 3800
    },
    {
      "epoch": 2.83,
      "grad_norm": 6.590944766998291,
      "learning_rate": 3.215169002473207e-06,
      "loss": 0.0575,
      "step": 3810
    },
    {
      "epoch": 2.83,
      "grad_norm": 7.214500427246094,
      "learning_rate": 3.07776861775213e-06,
      "loss": 0.0532,
      "step": 3820
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.6727337837219238,
      "learning_rate": 2.9403682330310528e-06,
      "loss": 0.0604,
      "step": 3830
    },
    {
      "epoch": 2.85,
      "grad_norm": 5.265418529510498,
      "learning_rate": 2.8029678483099755e-06,
      "loss": 0.0831,
      "step": 3840
    },
    {
      "epoch": 2.86,
      "grad_norm": 2.7963054180145264,
      "learning_rate": 2.665567463588898e-06,
      "loss": 0.0854,
      "step": 3850
    },
    {
      "epoch": 2.86,
      "grad_norm": 3.112823963165283,
      "learning_rate": 2.528167078867821e-06,
      "loss": 0.0575,
      "step": 3860
    },
    {
      "epoch": 2.87,
      "grad_norm": 7.16569185256958,
      "learning_rate": 2.390766694146744e-06,
      "loss": 0.081,
      "step": 3870
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.7017323970794678,
      "learning_rate": 2.2533663094256666e-06,
      "loss": 0.082,
      "step": 3880
    },
    {
      "epoch": 2.89,
      "grad_norm": 15.373734474182129,
      "learning_rate": 2.1159659247045893e-06,
      "loss": 0.0652,
      "step": 3890
    },
    {
      "epoch": 2.89,
      "grad_norm": 15.30858039855957,
      "learning_rate": 1.978565539983512e-06,
      "loss": 0.0755,
      "step": 3900
    },
    {
      "epoch": 2.9,
      "grad_norm": 10.219656944274902,
      "learning_rate": 1.8411651552624347e-06,
      "loss": 0.0774,
      "step": 3910
    },
    {
      "epoch": 2.91,
      "grad_norm": 6.792389869689941,
      "learning_rate": 1.7037647705413576e-06,
      "loss": 0.0773,
      "step": 3920
    },
    {
      "epoch": 2.92,
      "grad_norm": 13.666367530822754,
      "learning_rate": 1.5663643858202803e-06,
      "loss": 0.0614,
      "step": 3930
    },
    {
      "epoch": 2.92,
      "grad_norm": 7.227494716644287,
      "learning_rate": 1.4289640010992032e-06,
      "loss": 0.0831,
      "step": 3940
    },
    {
      "epoch": 2.93,
      "grad_norm": 7.196842670440674,
      "learning_rate": 1.2915636163781259e-06,
      "loss": 0.0511,
      "step": 3950
    },
    {
      "epoch": 2.94,
      "grad_norm": 6.138864994049072,
      "learning_rate": 1.1541632316570488e-06,
      "loss": 0.0722,
      "step": 3960
    },
    {
      "epoch": 2.95,
      "grad_norm": 7.251946449279785,
      "learning_rate": 1.0167628469359715e-06,
      "loss": 0.0812,
      "step": 3970
    },
    {
      "epoch": 2.95,
      "grad_norm": 9.087138175964355,
      "learning_rate": 8.793624622148942e-07,
      "loss": 0.075,
      "step": 3980
    },
    {
      "epoch": 2.96,
      "grad_norm": 7.0772199630737305,
      "learning_rate": 7.419620774938171e-07,
      "loss": 0.0713,
      "step": 3990
    },
    {
      "epoch": 2.97,
      "grad_norm": 13.134087562561035,
      "learning_rate": 6.045616927727397e-07,
      "loss": 0.0738,
      "step": 4000
    },
    {
      "epoch": 2.97,
      "grad_norm": 4.489412307739258,
      "learning_rate": 4.6716130805166254e-07,
      "loss": 0.0742,
      "step": 4010
    },
    {
      "epoch": 2.98,
      "grad_norm": 3.1018354892730713,
      "learning_rate": 3.2976092333058533e-07,
      "loss": 0.0673,
      "step": 4020
    },
    {
      "epoch": 2.99,
      "grad_norm": 10.936025619506836,
      "learning_rate": 1.9236053860950813e-07,
      "loss": 0.0674,
      "step": 4030
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.309031009674072,
      "learning_rate": 5.4960153888430885e-08,
      "loss": 0.0852,
      "step": 4040
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9823666485038896,
      "eval_loss": 0.046593740582466125,
      "eval_runtime": 138.9154,
      "eval_samples_per_second": 620.932,
      "eval_steps_per_second": 38.815,
      "step": 4044
    },
    {
      "epoch": 3.0,
      "step": 4044,
      "total_flos": 6.432009125858943e+18,
      "train_loss": 0.17701233503004327,
      "train_runtime": 1261.8873,
      "train_samples_per_second": 205.067,
      "train_steps_per_second": 3.205
    }
  ],
  "logging_steps": 10,
  "max_steps": 4044,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 6.432009125858943e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}