{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 31.807226181030273,
      "learning_rate": 4.5e-06,
      "loss": 10.2322,
      "step": 10
    },
    {
      "epoch": 0.02,
      "grad_norm": 14.19406795501709,
      "learning_rate": 9.5e-06,
      "loss": 8.6686,
      "step": 20
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.947138786315918,
      "learning_rate": 1.45e-05,
      "loss": 6.6091,
      "step": 30
    },
    {
      "epoch": 0.04,
      "grad_norm": 7.508649826049805,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 5.1112,
      "step": 40
    },
    {
      "epoch": 0.05,
      "grad_norm": 5.768527030944824,
      "learning_rate": 2.45e-05,
      "loss": 4.4247,
      "step": 50
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.962294578552246,
      "learning_rate": 2.95e-05,
      "loss": 3.7838,
      "step": 60
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.343990802764893,
      "learning_rate": 3.45e-05,
      "loss": 3.3172,
      "step": 70
    },
    {
      "epoch": 0.08,
      "grad_norm": 5.001032829284668,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 3.1734,
      "step": 80
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.4707157611846924,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 2.9784,
      "step": 90
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.321723461151123,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 2.6301,
      "step": 100
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.037026405334473,
      "learning_rate": 4.98448275862069e-05,
      "loss": 2.8441,
      "step": 110
    },
    {
      "epoch": 0.12,
      "grad_norm": 6.042785167694092,
      "learning_rate": 4.967241379310345e-05,
      "loss": 2.5558,
      "step": 120
    },
    {
      "epoch": 0.13,
      "grad_norm": 7.331098556518555,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 2.5997,
      "step": 130
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.285608291625977,
      "learning_rate": 4.932758620689655e-05,
      "loss": 2.3912,
      "step": 140
    },
    {
      "epoch": 0.15,
      "grad_norm": 5.633673191070557,
      "learning_rate": 4.915517241379311e-05,
      "loss": 2.3576,
      "step": 150
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.462876796722412,
      "learning_rate": 4.898275862068966e-05,
      "loss": 2.3879,
      "step": 160
    },
    {
      "epoch": 0.17,
      "grad_norm": 5.2996015548706055,
      "learning_rate": 4.8810344827586205e-05,
      "loss": 2.6634,
      "step": 170
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.960510730743408,
      "learning_rate": 4.863793103448276e-05,
      "loss": 2.3228,
      "step": 180
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.18018913269043,
      "learning_rate": 4.846551724137931e-05,
      "loss": 2.4035,
      "step": 190
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.5356884002685547,
      "learning_rate": 4.8293103448275865e-05,
      "loss": 1.9488,
      "step": 200
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.8790781497955322,
      "learning_rate": 4.812068965517242e-05,
      "loss": 2.2974,
      "step": 210
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.502866506576538,
      "learning_rate": 4.794827586206897e-05,
      "loss": 2.1646,
      "step": 220
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.9694302082061768,
      "learning_rate": 4.777586206896552e-05,
      "loss": 2.1186,
      "step": 230
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.986870050430298,
      "learning_rate": 4.7603448275862073e-05,
      "loss": 2.1907,
      "step": 240
    },
    {
      "epoch": 0.25,
      "grad_norm": 7.217125415802002,
      "learning_rate": 4.743103448275862e-05,
      "loss": 2.0278,
      "step": 250
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.887138843536377,
      "learning_rate": 4.725862068965518e-05,
      "loss": 2.269,
      "step": 260
    },
    {
      "epoch": 0.27,
      "grad_norm": 14.621356964111328,
      "learning_rate": 4.7086206896551726e-05,
      "loss": 2.0509,
      "step": 270
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.5131711959838867,
      "learning_rate": 4.6913793103448275e-05,
      "loss": 2.2239,
      "step": 280
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.753391265869141,
      "learning_rate": 4.674137931034483e-05,
      "loss": 2.0629,
      "step": 290
    },
    {
      "epoch": 0.3,
      "grad_norm": 4.106902599334717,
      "learning_rate": 4.656896551724138e-05,
      "loss": 2.0625,
      "step": 300
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.522510528564453,
      "learning_rate": 4.6396551724137935e-05,
      "loss": 1.9146,
      "step": 310
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.454258441925049,
      "learning_rate": 4.622413793103448e-05,
      "loss": 1.7679,
      "step": 320
    },
    {
      "epoch": 0.33,
      "grad_norm": 12.172571182250977,
      "learning_rate": 4.605172413793103e-05,
      "loss": 1.762,
      "step": 330
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.5643630027771,
      "learning_rate": 4.587931034482759e-05,
      "loss": 1.9651,
      "step": 340
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.371229648590088,
      "learning_rate": 4.5706896551724136e-05,
      "loss": 2.1326,
      "step": 350
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.6412742137908936,
      "learning_rate": 4.553448275862069e-05,
      "loss": 1.955,
      "step": 360
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.9563775062561035,
      "learning_rate": 4.536206896551725e-05,
      "loss": 1.8654,
      "step": 370
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.498210906982422,
      "learning_rate": 4.5189655172413796e-05,
      "loss": 2.0221,
      "step": 380
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.9855848550796509,
      "learning_rate": 4.501724137931035e-05,
      "loss": 1.9202,
      "step": 390
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.402889728546143,
      "learning_rate": 4.48448275862069e-05,
      "loss": 1.6582,
      "step": 400
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.9488422870635986,
      "learning_rate": 4.467241379310345e-05,
      "loss": 1.9576,
      "step": 410
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.99562406539917,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 1.6823,
      "step": 420
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.2172203063964844,
      "learning_rate": 4.432758620689655e-05,
      "loss": 1.7448,
      "step": 430
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.7004073858261108,
      "learning_rate": 4.415517241379311e-05,
      "loss": 1.687,
      "step": 440
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.6527938842773438,
      "learning_rate": 4.398275862068966e-05,
      "loss": 1.6152,
      "step": 450
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.223691463470459,
      "learning_rate": 4.3810344827586206e-05,
      "loss": 1.7911,
      "step": 460
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.228793144226074,
      "learning_rate": 4.363793103448276e-05,
      "loss": 1.9451,
      "step": 470
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.724231243133545,
      "learning_rate": 4.346551724137931e-05,
      "loss": 1.791,
      "step": 480
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.219877243041992,
      "learning_rate": 4.3293103448275865e-05,
      "loss": 1.692,
      "step": 490
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.977465867996216,
      "learning_rate": 4.3120689655172414e-05,
      "loss": 1.886,
      "step": 500
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.566962242126465,
      "learning_rate": 4.294827586206896e-05,
      "loss": 1.802,
      "step": 510
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.129835844039917,
      "learning_rate": 4.2775862068965525e-05,
      "loss": 1.659,
      "step": 520
    },
    {
      "epoch": 0.53,
      "grad_norm": 3.913256883621216,
      "learning_rate": 4.2603448275862074e-05,
      "loss": 1.5919,
      "step": 530
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.7998114824295044,
      "learning_rate": 4.243103448275862e-05,
      "loss": 1.7564,
      "step": 540
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.310154676437378,
      "learning_rate": 4.225862068965518e-05,
      "loss": 1.7347,
      "step": 550
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.835707426071167,
      "learning_rate": 4.208620689655173e-05,
      "loss": 1.802,
      "step": 560
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.9488534927368164,
      "learning_rate": 4.191379310344828e-05,
      "loss": 1.9981,
      "step": 570
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.861155986785889,
      "learning_rate": 4.174137931034483e-05,
      "loss": 1.9519,
      "step": 580
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.403691053390503,
      "learning_rate": 4.156896551724138e-05,
      "loss": 1.6711,
      "step": 590
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.9013047218322754,
      "learning_rate": 4.1396551724137935e-05,
      "loss": 1.6679,
      "step": 600
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.6646413803100586,
      "learning_rate": 4.1224137931034484e-05,
      "loss": 1.7117,
      "step": 610
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.6011579036712646,
      "learning_rate": 4.105172413793103e-05,
      "loss": 1.6351,
      "step": 620
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.9296406507492065,
      "learning_rate": 4.087931034482759e-05,
      "loss": 1.8179,
      "step": 630
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.2965588569641113,
      "learning_rate": 4.070689655172414e-05,
      "loss": 1.7164,
      "step": 640
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.7480602264404297,
      "learning_rate": 4.053448275862069e-05,
      "loss": 1.5552,
      "step": 650
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.9659781455993652,
      "learning_rate": 4.036206896551724e-05,
      "loss": 1.79,
      "step": 660
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5911238193511963,
      "learning_rate": 4.018965517241379e-05,
      "loss": 1.7139,
      "step": 670
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.8339680433273315,
      "learning_rate": 4.0017241379310345e-05,
      "loss": 1.6479,
      "step": 680
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.8804699182510376,
      "learning_rate": 3.98448275862069e-05,
      "loss": 1.6501,
      "step": 690
    },
    {
      "epoch": 0.7,
      "grad_norm": 5.505954742431641,
      "learning_rate": 3.967241379310345e-05,
      "loss": 1.549,
      "step": 700
    },
    {
      "epoch": 0.71,
      "grad_norm": 3.225405693054199,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 1.5231,
      "step": 710
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.4233756065368652,
      "learning_rate": 3.932758620689655e-05,
      "loss": 1.5577,
      "step": 720
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.9218379259109497,
      "learning_rate": 3.915517241379311e-05,
      "loss": 1.8394,
      "step": 730
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.7937753200531006,
      "learning_rate": 3.898275862068966e-05,
      "loss": 1.4735,
      "step": 740
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.696561336517334,
      "learning_rate": 3.8810344827586206e-05,
      "loss": 1.7371,
      "step": 750
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6280263662338257,
      "learning_rate": 3.863793103448276e-05,
      "loss": 1.6161,
      "step": 760
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7006651163101196,
      "learning_rate": 3.846551724137931e-05,
      "loss": 1.4593,
      "step": 770
    },
    {
      "epoch": 0.78,
      "grad_norm": 3.765646457672119,
      "learning_rate": 3.8293103448275866e-05,
      "loss": 1.5506,
      "step": 780
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.9314448833465576,
      "learning_rate": 3.8120689655172415e-05,
      "loss": 1.5805,
      "step": 790
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.736654281616211,
      "learning_rate": 3.794827586206896e-05,
      "loss": 1.5675,
      "step": 800
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.3542542457580566,
      "learning_rate": 3.777586206896552e-05,
      "loss": 1.5955,
      "step": 810
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.1898491382598877,
      "learning_rate": 3.760344827586207e-05,
      "loss": 1.3708,
      "step": 820
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.2105820178985596,
      "learning_rate": 3.743103448275862e-05,
      "loss": 1.6554,
      "step": 830
    },
    {
      "epoch": 0.84,
      "grad_norm": 4.901267051696777,
      "learning_rate": 3.725862068965517e-05,
      "loss": 1.3071,
      "step": 840
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.991741418838501,
      "learning_rate": 3.708620689655173e-05,
      "loss": 1.3876,
      "step": 850
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.1567299365997314,
      "learning_rate": 3.691379310344828e-05,
      "loss": 1.5718,
      "step": 860
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.741442084312439,
      "learning_rate": 3.674137931034483e-05,
      "loss": 1.3982,
      "step": 870
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.704728364944458,
      "learning_rate": 3.656896551724138e-05,
      "loss": 1.3117,
      "step": 880
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5768979787826538,
      "learning_rate": 3.6396551724137936e-05,
      "loss": 1.6102,
      "step": 890
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.6916723251342773,
      "learning_rate": 3.6224137931034484e-05,
      "loss": 1.6576,
      "step": 900
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.1157259941101074,
      "learning_rate": 3.605172413793104e-05,
      "loss": 1.4144,
      "step": 910
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.7090058326721191,
      "learning_rate": 3.587931034482759e-05,
      "loss": 1.3267,
      "step": 920
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.5148727893829346,
      "learning_rate": 3.570689655172414e-05,
      "loss": 1.7362,
      "step": 930
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.4861910343170166,
      "learning_rate": 3.553448275862069e-05,
      "loss": 1.4764,
      "step": 940
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.909857988357544,
      "learning_rate": 3.536206896551724e-05,
      "loss": 1.8828,
      "step": 950
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.5957543849945068,
      "learning_rate": 3.51896551724138e-05,
      "loss": 1.4019,
      "step": 960
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.6423611640930176,
      "learning_rate": 3.5017241379310345e-05,
      "loss": 1.2464,
      "step": 970
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.7054439783096313,
      "learning_rate": 3.4844827586206894e-05,
      "loss": 1.3908,
      "step": 980
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.5509027242660522,
      "learning_rate": 3.467241379310345e-05,
      "loss": 1.6855,
      "step": 990
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.0527048110961914,
      "learning_rate": 3.45e-05,
      "loss": 1.41,
      "step": 1000
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.912597417831421,
      "learning_rate": 3.4327586206896554e-05,
      "loss": 1.696,
      "step": 1010
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.4118075370788574,
      "learning_rate": 3.415517241379311e-05,
      "loss": 1.4162,
      "step": 1020
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.8533530235290527,
      "learning_rate": 3.398275862068966e-05,
      "loss": 1.3809,
      "step": 1030
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.950042724609375,
      "learning_rate": 3.381034482758621e-05,
      "loss": 1.3845,
      "step": 1040
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.852569341659546,
      "learning_rate": 3.363793103448276e-05,
      "loss": 1.3189,
      "step": 1050
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.6454888582229614,
      "learning_rate": 3.346551724137931e-05,
      "loss": 1.4498,
      "step": 1060
    },
    {
      "epoch": 1.07,
      "grad_norm": 2.481353759765625,
      "learning_rate": 3.3293103448275866e-05,
      "loss": 1.4578,
      "step": 1070
    },
    {
      "epoch": 1.08,
      "grad_norm": 2.739814519882202,
      "learning_rate": 3.3120689655172415e-05,
      "loss": 1.4949,
      "step": 1080
    },
    {
      "epoch": 1.09,
      "grad_norm": 2.799699068069458,
      "learning_rate": 3.2948275862068964e-05,
      "loss": 1.1835,
      "step": 1090
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.6168289184570312,
      "learning_rate": 3.277586206896552e-05,
      "loss": 1.4057,
      "step": 1100
    },
    {
      "epoch": 1.11,
      "grad_norm": 3.1404354572296143,
      "learning_rate": 3.260344827586207e-05,
      "loss": 1.2438,
      "step": 1110
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.4057466983795166,
      "learning_rate": 3.2431034482758623e-05,
      "loss": 1.5332,
      "step": 1120
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.1605113744735718,
      "learning_rate": 3.225862068965517e-05,
      "loss": 1.3803,
      "step": 1130
    },
    {
      "epoch": 1.1400000000000001,
      "grad_norm": 2.467651605606079,
      "learning_rate": 3.208620689655172e-05,
      "loss": 1.3769,
      "step": 1140
    },
    {
      "epoch": 1.15,
      "grad_norm": 2.0736987590789795,
      "learning_rate": 3.1913793103448276e-05,
      "loss": 1.4032,
      "step": 1150
    },
    {
      "epoch": 1.16,
      "grad_norm": 2.9988651275634766,
      "learning_rate": 3.1741379310344825e-05,
      "loss": 1.4226,
      "step": 1160
    },
    {
      "epoch": 1.17,
      "grad_norm": 2.2302157878875732,
      "learning_rate": 3.156896551724138e-05,
      "loss": 1.5159,
      "step": 1170
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.2775397300720215,
      "learning_rate": 3.1396551724137936e-05,
      "loss": 1.2539,
      "step": 1180
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.2107908725738525,
      "learning_rate": 3.1224137931034485e-05,
      "loss": 1.3425,
      "step": 1190
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.783532977104187,
      "learning_rate": 3.105172413793104e-05,
      "loss": 1.2874,
      "step": 1200
    },
    {
      "epoch": 1.21,
      "grad_norm": 2.1126065254211426,
      "learning_rate": 3.087931034482759e-05,
      "loss": 1.5065,
      "step": 1210
    },
    {
      "epoch": 1.22,
      "grad_norm": 2.010965347290039,
      "learning_rate": 3.070689655172414e-05,
      "loss": 1.3932,
      "step": 1220
    },
    {
      "epoch": 1.23,
      "grad_norm": 5.611945152282715,
      "learning_rate": 3.053448275862069e-05,
      "loss": 1.4646,
      "step": 1230
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.9755380153656006,
      "learning_rate": 3.0362068965517242e-05,
      "loss": 1.1764,
      "step": 1240
    },
    {
      "epoch": 1.25,
      "grad_norm": 2.106905221939087,
      "learning_rate": 3.0189655172413794e-05,
      "loss": 1.3547,
      "step": 1250
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.7636513710021973,
      "learning_rate": 3.0017241379310346e-05,
      "loss": 1.2024,
      "step": 1260
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.3909709453582764,
      "learning_rate": 2.9844827586206898e-05,
      "loss": 1.2256,
      "step": 1270
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.8174681663513184,
      "learning_rate": 2.967241379310345e-05,
      "loss": 1.4106,
      "step": 1280
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.857674241065979,
      "learning_rate": 2.95e-05,
      "loss": 1.1851,
      "step": 1290
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.9900033473968506,
      "learning_rate": 2.932758620689655e-05,
      "loss": 1.2333,
      "step": 1300
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.5615085363388062,
      "learning_rate": 2.9155172413793103e-05,
      "loss": 1.3969,
      "step": 1310
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.7219326496124268,
      "learning_rate": 2.8982758620689655e-05,
      "loss": 1.2629,
      "step": 1320
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.7303168773651123,
      "learning_rate": 2.881034482758621e-05,
      "loss": 1.4391,
      "step": 1330
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.7237037420272827,
      "learning_rate": 2.8637931034482763e-05,
      "loss": 1.3606,
      "step": 1340
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.575646162033081,
      "learning_rate": 2.8465517241379315e-05,
      "loss": 1.1907,
      "step": 1350
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 2.1866729259490967,
      "learning_rate": 2.8293103448275863e-05,
      "loss": 1.3406,
      "step": 1360
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.5900951623916626,
      "learning_rate": 2.8120689655172415e-05,
      "loss": 1.3234,
      "step": 1370
    },
    {
      "epoch": 1.38,
      "grad_norm": 2.729886293411255,
      "learning_rate": 2.7948275862068968e-05,
      "loss": 1.3068,
      "step": 1380
    },
    {
      "epoch": 1.3900000000000001,
      "grad_norm": 7.489139556884766,
      "learning_rate": 2.777586206896552e-05,
      "loss": 1.2874,
      "step": 1390
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.6766670942306519,
      "learning_rate": 2.7603448275862072e-05,
      "loss": 1.369,
      "step": 1400
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.6040589809417725,
      "learning_rate": 2.743103448275862e-05,
      "loss": 1.4296,
      "step": 1410
    },
    {
      "epoch": 1.42,
      "grad_norm": 2.0411570072174072,
      "learning_rate": 2.7258620689655173e-05,
      "loss": 1.2803,
      "step": 1420
    },
    {
      "epoch": 1.43,
      "grad_norm": 2.604081869125366,
      "learning_rate": 2.7086206896551725e-05,
      "loss": 1.2784,
      "step": 1430
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.4787765741348267,
      "learning_rate": 2.6913793103448277e-05,
      "loss": 1.3021,
      "step": 1440
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.665315866470337,
      "learning_rate": 2.674137931034483e-05,
      "loss": 1.1502,
      "step": 1450
    },
    {
      "epoch": 1.46,
      "grad_norm": 2.4420969486236572,
      "learning_rate": 2.6568965517241378e-05,
      "loss": 1.3531,
      "step": 1460
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.961679220199585,
      "learning_rate": 2.639655172413793e-05,
      "loss": 1.413,
      "step": 1470
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.6548084020614624,
      "learning_rate": 2.6224137931034482e-05,
      "loss": 1.2579,
      "step": 1480
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.22232985496521,
      "learning_rate": 2.6051724137931037e-05,
      "loss": 1.4794,
      "step": 1490
    },
    {
      "epoch": 1.5,
      "grad_norm": 2.133786916732788,
      "learning_rate": 2.587931034482759e-05,
      "loss": 1.4226,
      "step": 1500
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.7223315238952637,
      "learning_rate": 2.570689655172414e-05,
      "loss": 1.3799,
      "step": 1510
    },
    {
      "epoch": 1.52,
      "grad_norm": 2.187749147415161,
      "learning_rate": 2.5534482758620693e-05,
      "loss": 1.3415,
      "step": 1520
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.8895548582077026,
      "learning_rate": 2.5362068965517246e-05,
      "loss": 1.2736,
      "step": 1530
    },
    {
      "epoch": 1.54,
      "grad_norm": 2.2825796604156494,
      "learning_rate": 2.5189655172413794e-05,
      "loss": 1.4374,
      "step": 1540
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.5137097835540771,
      "learning_rate": 2.5017241379310346e-05,
      "loss": 1.1751,
      "step": 1550
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.8762792348861694,
      "learning_rate": 2.48448275862069e-05,
      "loss": 1.2994,
      "step": 1560
    },
    {
      "epoch": 1.5699999999999998,
      "grad_norm": 1.4601935148239136,
      "learning_rate": 2.467241379310345e-05,
      "loss": 1.6414,
      "step": 1570
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.7618972063064575,
      "learning_rate": 2.45e-05,
      "loss": 1.3107,
      "step": 1580
    },
    {
      "epoch": 1.5899999999999999,
      "grad_norm": 3.77205753326416,
      "learning_rate": 2.432758620689655e-05,
      "loss": 1.4596,
      "step": 1590
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.3197537660598755,
      "learning_rate": 2.4155172413793103e-05,
      "loss": 1.2752,
      "step": 1600
    },
    {
      "epoch": 1.6099999999999999,
      "grad_norm": 1.2479321956634521,
      "learning_rate": 2.398275862068966e-05,
      "loss": 1.2671,
      "step": 1610
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.6044553518295288,
      "learning_rate": 2.3810344827586208e-05,
      "loss": 1.4017,
      "step": 1620
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.746628999710083,
      "learning_rate": 2.363793103448276e-05,
      "loss": 1.1989,
      "step": 1630
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 1.4533942937850952,
      "learning_rate": 2.3465517241379312e-05,
      "loss": 1.3846,
      "step": 1640
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.4432073831558228,
      "learning_rate": 2.3293103448275864e-05,
      "loss": 1.385,
      "step": 1650
    },
    {
      "epoch": 1.6600000000000001,
      "grad_norm": 1.7605477571487427,
      "learning_rate": 2.3120689655172416e-05,
      "loss": 1.3516,
      "step": 1660
    },
    {
      "epoch": 1.67,
      "grad_norm": 2.024019718170166,
      "learning_rate": 2.2948275862068965e-05,
      "loss": 1.2414,
      "step": 1670
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 1.9421080350875854,
      "learning_rate": 2.2775862068965517e-05,
      "loss": 1.3778,
      "step": 1680
    },
    {
      "epoch": 1.69,
      "grad_norm": 2.7231059074401855,
      "learning_rate": 2.2603448275862072e-05,
      "loss": 1.3709,
      "step": 1690
    },
    {
      "epoch": 1.7,
      "grad_norm": 4.394734859466553,
      "learning_rate": 2.2431034482758624e-05,
      "loss": 1.0418,
      "step": 1700
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.7055786848068237,
      "learning_rate": 2.2258620689655173e-05,
      "loss": 1.3787,
      "step": 1710
    },
    {
      "epoch": 1.72,
      "grad_norm": 2.871880292892456,
      "learning_rate": 2.2086206896551725e-05,
      "loss": 1.2104,
      "step": 1720
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.9549953937530518,
      "learning_rate": 2.1913793103448277e-05,
      "loss": 1.3271,
      "step": 1730
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.724934458732605,
      "learning_rate": 2.174137931034483e-05,
      "loss": 1.1641,
      "step": 1740
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.8861603736877441,
      "learning_rate": 2.1568965517241378e-05,
      "loss": 1.3302,
      "step": 1750
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.9307113885879517,
      "learning_rate": 2.139655172413793e-05,
      "loss": 1.4253,
      "step": 1760
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.860857605934143,
      "learning_rate": 2.1224137931034486e-05,
      "loss": 1.4241,
      "step": 1770
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.8345351219177246,
      "learning_rate": 2.1051724137931038e-05,
      "loss": 1.2214,
      "step": 1780
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.1715213060379028,
      "learning_rate": 2.0879310344827586e-05,
      "loss": 1.3256,
      "step": 1790
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.8184287548065186,
      "learning_rate": 2.070689655172414e-05,
      "loss": 1.1173,
      "step": 1800
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.9528814554214478,
      "learning_rate": 2.053448275862069e-05,
      "loss": 1.1872,
      "step": 1810
    },
    {
      "epoch": 1.8199999999999998,
      "grad_norm": 2.1913580894470215,
      "learning_rate": 2.0362068965517243e-05,
      "loss": 1.2677,
      "step": 1820
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.9390571117401123,
      "learning_rate": 2.0189655172413795e-05,
      "loss": 1.2469,
      "step": 1830
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 1.8685649633407593,
      "learning_rate": 2.0017241379310343e-05,
      "loss": 1.2809,
      "step": 1840
    },
    {
      "epoch": 1.85,
      "grad_norm": 2.4311530590057373,
      "learning_rate": 1.98448275862069e-05,
      "loss": 1.2982,
      "step": 1850
    },
    {
      "epoch": 1.8599999999999999,
      "grad_norm": 1.5639426708221436,
      "learning_rate": 1.967241379310345e-05,
      "loss": 1.2996,
      "step": 1860
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.2009373903274536,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 1.2812,
      "step": 1870
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.017490267753601,
      "learning_rate": 1.9327586206896552e-05,
      "loss": 1.1055,
      "step": 1880
    },
    {
      "epoch": 1.8900000000000001,
      "grad_norm": 2.3106815814971924,
      "learning_rate": 1.9155172413793104e-05,
      "loss": 1.2474,
      "step": 1890
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.609084129333496,
      "learning_rate": 1.8982758620689656e-05,
      "loss": 1.1917,
      "step": 1900
    },
    {
      "epoch": 1.9100000000000001,
      "grad_norm": 1.7436444759368896,
      "learning_rate": 1.8810344827586208e-05,
      "loss": 1.3704,
      "step": 1910
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.7709165811538696,
      "learning_rate": 1.863793103448276e-05,
      "loss": 1.3904,
      "step": 1920
    },
    {
      "epoch": 1.9300000000000002,
      "grad_norm": 1.775133490562439,
      "learning_rate": 1.8465517241379312e-05,
      "loss": 1.2368,
      "step": 1930
    },
    {
      "epoch": 1.94,
      "grad_norm": 2.224012613296509,
      "learning_rate": 1.8293103448275864e-05,
      "loss": 1.1834,
      "step": 1940
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.8021095991134644,
      "learning_rate": 1.8120689655172416e-05,
      "loss": 1.2385,
      "step": 1950
    },
    {
      "epoch": 1.96,
      "grad_norm": 2.179215431213379,
      "learning_rate": 1.7948275862068965e-05,
      "loss": 1.4039,
      "step": 1960
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.850145697593689,
      "learning_rate": 1.7775862068965517e-05,
      "loss": 1.1953,
      "step": 1970
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.6872076988220215,
      "learning_rate": 1.760344827586207e-05,
      "loss": 1.2247,
      "step": 1980
    },
    {
      "epoch": 1.99,
      "grad_norm": 2.2350351810455322,
      "learning_rate": 1.743103448275862e-05,
      "loss": 1.0938,
      "step": 1990
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.5278306007385254,
      "learning_rate": 1.7258620689655173e-05,
      "loss": 1.216,
      "step": 2000
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.9336029291152954,
      "learning_rate": 1.7086206896551722e-05,
      "loss": 1.1958,
      "step": 2010
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.2718322277069092,
      "learning_rate": 1.6913793103448278e-05,
      "loss": 1.047,
      "step": 2020
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.8943592309951782,
      "learning_rate": 1.674137931034483e-05,
      "loss": 1.3197,
      "step": 2030
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.7184983491897583,
      "learning_rate": 1.6568965517241382e-05,
      "loss": 1.1611,
      "step": 2040
    },
    {
      "epoch": 2.05,
      "grad_norm": 2.1850178241729736,
      "learning_rate": 1.639655172413793e-05,
      "loss": 1.2405,
      "step": 2050
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.6850008964538574,
      "learning_rate": 1.6224137931034483e-05,
      "loss": 1.119,
      "step": 2060
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.895009160041809,
      "learning_rate": 1.6051724137931035e-05,
      "loss": 1.3074,
      "step": 2070
    },
    {
      "epoch": 2.08,
      "grad_norm": 2.0490798950195312,
      "learning_rate": 1.5879310344827587e-05,
      "loss": 1.0361,
      "step": 2080
    },
    {
      "epoch": 2.09,
      "grad_norm": 2.465575695037842,
      "learning_rate": 1.570689655172414e-05,
      "loss": 1.1015,
      "step": 2090
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.3066885471343994,
      "learning_rate": 1.553448275862069e-05,
      "loss": 1.1267,
      "step": 2100
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.7399309873580933,
      "learning_rate": 1.5362068965517243e-05,
      "loss": 1.203,
      "step": 2110
    },
    {
      "epoch": 2.12,
      "grad_norm": 2.804007053375244,
      "learning_rate": 1.5189655172413793e-05,
      "loss": 1.0919,
      "step": 2120
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.44864022731781,
      "learning_rate": 1.5017241379310346e-05,
      "loss": 1.4365,
      "step": 2130
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.5640863180160522,
      "learning_rate": 1.4844827586206898e-05,
      "loss": 1.1694,
      "step": 2140
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.847272276878357,
      "learning_rate": 1.4672413793103448e-05,
      "loss": 1.1713,
      "step": 2150
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.4313805103302002,
      "learning_rate": 1.45e-05,
      "loss": 1.108,
      "step": 2160
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.906491994857788,
      "learning_rate": 1.432758620689655e-05,
      "loss": 1.2404,
      "step": 2170
    },
    {
      "epoch": 2.18,
      "grad_norm": 2.265270948410034,
      "learning_rate": 1.4155172413793106e-05,
      "loss": 1.1744,
      "step": 2180
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.927332878112793,
      "learning_rate": 1.3982758620689656e-05,
      "loss": 1.1907,
      "step": 2190
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.7866955995559692,
      "learning_rate": 1.3810344827586208e-05,
      "loss": 1.1983,
      "step": 2200
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.738226056098938,
      "learning_rate": 1.3637931034482759e-05,
      "loss": 1.2034,
      "step": 2210
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.8509528636932373,
      "learning_rate": 1.3465517241379311e-05,
      "loss": 1.2246,
      "step": 2220
    },
    {
      "epoch": 2.23,
      "grad_norm": 2.776275634765625,
      "learning_rate": 1.3293103448275861e-05,
      "loss": 1.0615,
      "step": 2230
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.994899868965149,
      "learning_rate": 1.3120689655172413e-05,
      "loss": 1.0704,
      "step": 2240
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.3992385864257812,
      "learning_rate": 1.2948275862068966e-05,
      "loss": 1.1459,
      "step": 2250
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.4067262411117554,
      "learning_rate": 1.277586206896552e-05,
      "loss": 1.104,
      "step": 2260
    },
    {
      "epoch": 2.27,
      "grad_norm": 2.0203895568847656,
      "learning_rate": 1.260344827586207e-05,
      "loss": 1.2588,
      "step": 2270
    },
    {
      "epoch": 2.2800000000000002,
      "grad_norm": 2.0324246883392334,
      "learning_rate": 1.2431034482758622e-05,
      "loss": 1.2133,
      "step": 2280
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.5362598896026611,
      "learning_rate": 1.2258620689655174e-05,
      "loss": 1.2062,
      "step": 2290
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.9725372791290283,
      "learning_rate": 1.2086206896551724e-05,
      "loss": 1.2211,
      "step": 2300
    },
    {
      "epoch": 2.31,
      "grad_norm": 2.6382522583007812,
      "learning_rate": 1.1913793103448276e-05,
      "loss": 1.3856,
      "step": 2310
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.9707180857658386,
      "learning_rate": 1.1741379310344828e-05,
      "loss": 1.1416,
      "step": 2320
    },
    {
      "epoch": 2.33,
      "grad_norm": 2.1676788330078125,
      "learning_rate": 1.156896551724138e-05,
      "loss": 1.2297,
      "step": 2330
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.6406278610229492,
      "learning_rate": 1.1396551724137931e-05,
      "loss": 0.9923,
      "step": 2340
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.8479793071746826,
      "learning_rate": 1.1224137931034483e-05,
      "loss": 1.0925,
      "step": 2350
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.7905402183532715,
      "learning_rate": 1.1051724137931035e-05,
      "loss": 1.2361,
      "step": 2360
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.006901979446411,
      "learning_rate": 1.0879310344827587e-05,
      "loss": 1.144,
      "step": 2370
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.4485629796981812,
      "learning_rate": 1.0706896551724138e-05,
      "loss": 1.1374,
      "step": 2380
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.953250765800476,
      "learning_rate": 1.053448275862069e-05,
      "loss": 1.222,
      "step": 2390
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.4904489517211914,
      "learning_rate": 1.0362068965517242e-05,
      "loss": 1.1942,
      "step": 2400
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.9679901599884033,
      "learning_rate": 1.0189655172413794e-05,
      "loss": 1.2024,
      "step": 2410
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.6501275300979614,
      "learning_rate": 1.0017241379310344e-05,
      "loss": 1.2754,
      "step": 2420
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.617769479751587,
      "learning_rate": 9.844827586206896e-06,
      "loss": 1.2086,
      "step": 2430
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.9600815773010254,
      "learning_rate": 9.672413793103448e-06,
      "loss": 1.1549,
      "step": 2440
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.5949968099594116,
      "learning_rate": 9.5e-06,
      "loss": 1.2698,
      "step": 2450
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.2073543071746826,
      "learning_rate": 9.327586206896553e-06,
      "loss": 1.1885,
      "step": 2460
    },
    {
      "epoch": 2.4699999999999998,
      "grad_norm": 1.0354326963424683,
      "learning_rate": 9.155172413793103e-06,
      "loss": 1.2975,
      "step": 2470
    },
    {
      "epoch": 2.48,
      "grad_norm": 4.918059825897217,
      "learning_rate": 8.982758620689657e-06,
      "loss": 1.161,
      "step": 2480
    },
    {
      "epoch": 2.49,
      "grad_norm": 2.176772117614746,
      "learning_rate": 8.810344827586207e-06,
      "loss": 1.1096,
      "step": 2490
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.9638526439666748,
      "learning_rate": 8.63793103448276e-06,
      "loss": 1.1893,
      "step": 2500
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.3812869787216187,
      "learning_rate": 8.46551724137931e-06,
      "loss": 1.1958,
      "step": 2510
    },
    {
      "epoch": 2.52,
      "grad_norm": 2.1832292079925537,
      "learning_rate": 8.293103448275863e-06,
      "loss": 1.1582,
      "step": 2520
    },
    {
      "epoch": 2.5300000000000002,
      "grad_norm": 1.416548728942871,
      "learning_rate": 8.120689655172414e-06,
      "loss": 0.9992,
      "step": 2530
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.465884804725647,
      "learning_rate": 7.948275862068966e-06,
      "loss": 1.1773,
      "step": 2540
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.754629373550415,
      "learning_rate": 7.775862068965516e-06,
      "loss": 1.2893,
      "step": 2550
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.3299590349197388,
      "learning_rate": 7.60344827586207e-06,
      "loss": 1.1175,
      "step": 2560
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.6379165649414062,
      "learning_rate": 7.431034482758621e-06,
      "loss": 1.1522,
      "step": 2570
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.038396954536438,
      "learning_rate": 7.258620689655173e-06,
      "loss": 1.1105,
      "step": 2580
    },
    {
      "epoch": 2.59,
      "grad_norm": 3.7200427055358887,
      "learning_rate": 7.086206896551724e-06,
      "loss": 1.2392,
      "step": 2590
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.996336817741394,
      "learning_rate": 6.913793103448277e-06,
      "loss": 1.2586,
      "step": 2600
    },
    {
      "epoch": 2.61,
      "grad_norm": 2.2930266857147217,
      "learning_rate": 6.741379310344828e-06,
      "loss": 1.2299,
      "step": 2610
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.5807350873947144,
      "learning_rate": 6.568965517241379e-06,
      "loss": 1.1705,
      "step": 2620
    },
    {
      "epoch": 2.63,
      "grad_norm": 2.1657156944274902,
      "learning_rate": 6.3965517241379305e-06,
      "loss": 1.008,
      "step": 2630
    },
    {
      "epoch": 2.64,
      "grad_norm": 2.036822557449341,
      "learning_rate": 6.224137931034483e-06,
      "loss": 1.36,
      "step": 2640
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.8225266933441162,
      "learning_rate": 6.051724137931035e-06,
      "loss": 1.0531,
      "step": 2650
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.9631621837615967,
      "learning_rate": 5.879310344827586e-06,
      "loss": 1.1185,
      "step": 2660
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.3313068151474,
      "learning_rate": 5.706896551724138e-06,
      "loss": 1.0997,
      "step": 2670
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.7591108083724976,
      "learning_rate": 5.53448275862069e-06,
      "loss": 1.2264,
      "step": 2680
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.669311761856079,
      "learning_rate": 5.362068965517242e-06,
      "loss": 1.0538,
      "step": 2690
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.9697110652923584,
      "learning_rate": 5.1896551724137934e-06,
      "loss": 1.2217,
      "step": 2700
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.5783755779266357,
      "learning_rate": 5.0172413793103455e-06,
      "loss": 1.1666,
      "step": 2710
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 2.451660394668579,
      "learning_rate": 4.844827586206897e-06,
      "loss": 1.2254,
      "step": 2720
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.6855318546295166,
      "learning_rate": 4.672413793103449e-06,
      "loss": 1.1644,
      "step": 2730
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.955734372138977,
      "learning_rate": 4.5e-06,
      "loss": 1.2558,
      "step": 2740
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.5705198049545288,
      "learning_rate": 4.327586206896552e-06,
      "loss": 1.3614,
      "step": 2750
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.878604769706726,
      "learning_rate": 4.1551724137931034e-06,
      "loss": 1.1227,
      "step": 2760
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.3818405866622925,
      "learning_rate": 3.9827586206896555e-06,
      "loss": 1.0261,
      "step": 2770
    },
    {
      "epoch": 2.7800000000000002,
      "grad_norm": 1.960067629814148,
      "learning_rate": 3.8103448275862068e-06,
      "loss": 1.2194,
      "step": 2780
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.6252312660217285,
      "learning_rate": 3.637931034482759e-06,
      "loss": 1.0916,
      "step": 2790
    },
    {
      "epoch": 2.8,
      "grad_norm": 2.1915979385375977,
      "learning_rate": 3.4655172413793105e-06,
      "loss": 0.9632,
      "step": 2800
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.7106447219848633,
      "learning_rate": 3.2931034482758626e-06,
      "loss": 1.1671,
      "step": 2810
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.840908169746399,
      "learning_rate": 3.120689655172414e-06,
      "loss": 1.1216,
      "step": 2820
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.71086585521698,
      "learning_rate": 2.9482758620689655e-06,
      "loss": 1.1521,
      "step": 2830
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.500496745109558,
      "learning_rate": 2.775862068965517e-06,
      "loss": 1.1021,
      "step": 2840
    },
    {
      "epoch": 2.85,
      "grad_norm": 2.563215494155884,
      "learning_rate": 2.603448275862069e-06,
      "loss": 1.0535,
      "step": 2850
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.2782319784164429,
      "learning_rate": 2.4310344827586205e-06,
      "loss": 1.1942,
      "step": 2860
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.2960996627807617,
      "learning_rate": 2.258620689655172e-06,
      "loss": 1.1178,
      "step": 2870
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.6457912921905518,
      "learning_rate": 2.0862068965517243e-06,
      "loss": 1.0038,
      "step": 2880
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.8096386194229126,
      "learning_rate": 1.913793103448276e-06,
      "loss": 1.142,
      "step": 2890
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.7525629997253418,
      "learning_rate": 1.7413793103448276e-06,
      "loss": 1.254,
      "step": 2900
    },
    {
      "epoch": 2.91,
      "grad_norm": 2.7859292030334473,
      "learning_rate": 1.5689655172413793e-06,
      "loss": 1.4235,
      "step": 2910
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.4601234197616577,
      "learning_rate": 1.3965517241379312e-06,
      "loss": 1.0771,
      "step": 2920
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.9630273580551147,
      "learning_rate": 1.2241379310344828e-06,
      "loss": 1.2312,
      "step": 2930
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.5725574493408203,
      "learning_rate": 1.0517241379310345e-06,
      "loss": 1.026,
      "step": 2940
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.9632741212844849,
      "learning_rate": 8.793103448275863e-07,
      "loss": 1.1516,
      "step": 2950
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.4151031970977783,
      "learning_rate": 7.068965517241379e-07,
      "loss": 1.2166,
      "step": 2960
    },
    {
      "epoch": 2.9699999999999998,
      "grad_norm": 1.4842921495437622,
      "learning_rate": 5.344827586206897e-07,
      "loss": 1.1279,
      "step": 2970
    },
    {
      "epoch": 2.98,
      "grad_norm": 2.3842453956604004,
      "learning_rate": 3.6206896551724143e-07,
      "loss": 1.0998,
      "step": 2980
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.8654992580413818,
      "learning_rate": 1.8965517241379312e-07,
      "loss": 1.2559,
      "step": 2990
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5230703353881836,
      "learning_rate": 1.7241379310344825e-08,
      "loss": 1.1137,
      "step": 3000
    }
  ],
  "logging_steps": 10,
  "max_steps": 3000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2786102083584000.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}